Language and cognition

Language and Cognition - Ph.D. Dissertation by Riccardo Fusaroli

Table of Contents 0. Introduction: Thinking Through Language.......................... 12 0.1. The Thesis of the Dissertation ............................................................................................... 12 0.2. The Reasons behind the Thesis .............................................................................................. 13 0.3. The Phenomena Analysed ...................................................................................................... 14 0.4. What does it take to be a Pragmatist? (Sketching the Overall Perspective) .................................... 15 0.4.1. A Theory of Cognition ......................................................................................... 17 0.4.2. A Theory of Semiosis ........................................................................................... 17 0.4.3. A Theory of Experience ....................................................................................... 19 0.5. The Structure of the Dissertation ............................................................................................ 24 0.5.1. Articulating Language and Cognition – An Overview on Chapter 1 ................. 24 0.5.2. Language and Conceptual Structures – An Overview on Chapter 2 .................. 26 0.5.3. Linguistically Interacting Minds – An Overview on Chapter 3 .......................... 27 0.5.4. Conclusions .......................................................................................................... 28

1. The Social Horizon of Embodied and Extended Language ..... 30 1.1. Introduction: Beyond the Informational Paradigm ................................................................... 30 1.2. The Cognitive Stance ............................................................................................................ 32 1.2.1. A thread through cognitive sciences .................................................................... 32 1.2.1.1. Bastion of the Old: Behaviourism .............................................................................. 32 1.2.1.2. The Cognitive Revolution: tackling the filling of the sandwich ................................ 33 1.2.1.3. Beyond Symbolic Rules ............................................................................................. 36 1.2.1.3.1. Neural Networks ................................................................................................................. 36 1.2.1.3.2. Dynamical Systems Theory ................................................................................................ 38

1.2.1.4. Beyond the Brain and the Nervous System ............................................................... 42 1.2.1.4.1. Embedding Cognition in the Body ..................................................................................... 43 1.2.1.4.2. Embedding Cognition in the Context ................................................................................ 45 1.2.1.4.3. Distributing cognition ....................................................................................................... 46

1.2.1.5. The subject of cognition ............................................................................................. 48

1.2.2. A Peircean view: Mediation as a (partial) Mark of the Cognitive ....................... 54 1.2.2.1. 1.2.2.2. 1.2.2.3. 1.2.2.4.

The Case of Iconism .................................................................................................. 55 Continuity as the Core of Mediation ........................................................................ 57 The Structure of Continuity ...................................................................................... 60 What to do with the concept of habit ........................................................................ 66

1.2.2.4.1. 1.2.2.4.2. 1.2.2.4.3. 1.2.2.4.4.

Habits as Flexible Tendencies............................................................................................ 67 Habits as flexible tendencies of distributed cognition ........................................................ 68 Habit as a community-supported flexible tendency of distributed cognition .................... 71 The tentative, temporal and manipulative nature of habits .............................................. 75

1.2.3. A Pragmatist Approach to Cognition .................................................................. 79 1.3. Cognitive stances on language ................................................................................................ 83 1.3.1. The embodied turn: Language in the brain/body ............................................. 83 1.3.2. The extended turn: language (and mind) into the world ..................................... 86 1.4. A habit-based stance on Language in Cognition and Cognition in Language ................................ 89 1.4.1. Embodiment as ecological enaction ..................................................................... 89 1.4.2. Experience – Materiality – Intersubjectivity ....................................................... 91 1.4.3. Language as a tool .............................................................................................. 93 1.4.4. Language as an environment and as normative instances ................................... 94

8

1.4.5. The flexibility of language: semantic forms and use potentials ............................ 97 1.5. Conclusions ......................................................................................................................... 99

2. The Life of Caimans in Society and what it can tell us about the roles of metaphor, language and thought ............................... 101 2.1. Introduction: Of the life of metaphors in society ...................................................................... 101 2.1.1. Conceptualising Metaphors and Methodological Consequences ...................... 104 2.2. Cognitive Linguistics ........................................................................................................... 106 2.2.1. The declared novelty of the cognitive approaches ............................................. 106 2.2.1.1. The importance of studying metaphors in Cognitive Linguistics............................. 108

2.2.1.1.1. Historical Reasons of CMT .............................................................................................. 109 2.2.1.1.2. Prototypicality of CMT .................................................................................................... 110

2.2.2. Basic Tenets of CMT ......................................................................................... 113 2.2.2.1. Metaphors We Live By: Experientialism and image-schemas ................................. 113 2.2.2.2. Philosophy in the Flesh and Beyond: Embodied Realism, Primary Metaphors and Neural Grounding ................................................................................................................. 120

2.2.2.2.1. Primary metaphors and conflation ................................................................................... 120 2.2.2.2.2. The neural grounding of CMT ........................................................................................ 122

2.2.2.3. Analytical Methodology ........................................................................................... 125 2.2.2.3.1. PEOPLE ARE ANIMALS as a conceptual formation..................................................... 126 2.2.2.3.2. PEOPLE ARE ANIMALS as a conceptual metaphor ..................................................... 128 2.2.2.3.3. Berlusconi is a caiman as an expression of PEOPLE ARE ANIMALS: part 1 ............... 133

2.2.2.3.3.1. Conventionality......................................................................................... 133 2.2.2.3.3.2. Functionality ............................................................................................. 134 2.2.2.3.3.3. Nature ....................................................................................................... 135 2.2.2.3.3.4. Generality ................................................................................................. 136

2.2.2.3.4. Berlusconi is a caiman and PEOPLE ARE ANIMALS: an overview .............................. 136

2.2.2.4. CMT: An assessment................................................................................................ 137

2.2.3. Conceptual Integration Theory ......................................................................... 138 2.2.3.1. Neuron Huggers versus Circle Worshippers: On the Nature of Mental Spaces...... 143 2.2.3.1.1. Cognitive grounding: mental spaces and memory ........................................................... 144 2.2.3.1.2. And there come the Neuron Huggers............................................................................... 146 2.2.3.1.3. The revenge of the Circle Worshippers: towards a pheno-cognitive approach ............... 146

2.2.3.2. Blending and the taxonomy of spaces ...................................................................... 148 2.2.3.3. Optimality Principles and the Human Scale ............................................................ 150 2.2.3.4. Metaphors as conceptual integrations ...................................................................... 151 2.2.3.5. “Berlusconi is a caiman” as a metaphorical conceptual integration ........................ 152 2.2.3.6. CIT: an assessment ................................................................................................... 153

2.2.4. Assessing the cognitive approaches to metaphors .............................................. 156 2.2.4.1. Mediating the directly meaningful ........................................................................... 157 2.2.4.1.1. Image-schemas and generality .......................................................................................... 160 2.2.4.1.2. Joint Attention and Culture in the formation of Image-Schemas .................................... 162 2.2.4.1.3. The Structure of Abstract Domains.................................................................................. 167 2.2.4.1.4. Cultural Metaphorical Projection ..................................................................................... 169 2.2.4.1.5. Reintroducing the role of language in conceptual ontogenesis ........................................ 173

2.2.4.2. Reintroducing language use as cognition ................................................................. 173

2.2.4.2.1. Exploring the metaphor from a linguistic perspective ...................................................... 175 2.2.4.2.2. Diagrammatic Reasoning ................................................................................................. 178 2.2.4.2.3. Diagrammatic Reasoning in Language ............................................................................ 183

2.3. The strange case of the Man who Became a Caiman or Of the Life of Caimans in Society .......... 187 2.3.1. The importance of being a caiman in the news ................................................. 190 2.3.2. Cognitive Linguistics and corpus approaches .................................................... 191 2.3.2.1. What is a corpus?...................................................................................................... 191

2.3.2.2. Different Kinds of Corpora ...................................................................................... 192 2.3.2.3. Methodology of selection.......................................................................................... 195 2.3.2.4. Investigating concepts via verbal patterns ................................................................ 196

2.3.3. The Overall Perspective ..................................................................................... 198 2.3.3.1. Searching for Berlusconi .......................................................................................... 198 2.3.3.2. Searching for the caiman ......................................................................................... 201

2.3.4. A Diachronic Approach ..................................................................................... 209 2.3.4.1. Caimans before the Caiman (1984-2001) ................................................................ 210

2.3.4.1.1. Literal uses ........................................................................................................................ 210 2.3.4.1.2. Metaphorical uses ............................................................................................................. 211

2.3.4.2. The Birth of the Caiman (2002) ............................................................................... 214 2.3.4.2.1. A short introduction to Berlusconi (by Cordero) .............................................................. 214 2.3.4.2.2. The textual dimension of the caimanic smile: oppositions and narrations ....................... 215 2.3.4.2.3. A distinctive regime of comparisons ................................................................................. 216 2.3.4.2.4. Agency and power: between actants and figures .............................................................. 217 2.3.4.2.5. The caiman in its co-text .................................................................................................. 220

2.3.4.3. The Growth of the Caiman (2003-2005) ................................................................. 221 2.3.4.4. Caimanic Contagion (2006) ..................................................................................... 224 2.3.4.5. The Caiman as a form of life (2007-now)................................................................. 224 2.3.4.6. Changing what is it like to be a caiman: unsuccessful attempts at explicit diagrammatic manipulation .................................................................................................. 226 2.3.4.7. The Caiman in The Great Chain of Being .............................................................. 229

2.3.4.7.1. Berlusconi as an animal - from the big grass snake to the Hyena Ridens ........................ 230 2.3.4.7.2. Caiman versus Shark and Other Animal Stories in Italian Newspapers .......................... 231

2.3.4.8. Virgins vs. Cayman: the hidden role of verbal patterns ........................................... 233

2.3.5. Assessing the life of caimans in society ............................................................... 234 2.3.5.1. Experiential Grounding of “Berlusconi is a Caiman” .............................................. 236 2.3.5.1.1. PEOPLE ARE ANIMALS -> RUTHLESS PEOPLE ARE PREDATORS ................. 236

2.3.5.2. The Social Arena for metaphors .............................................................................. 241

2.4. Conclusions ....................................................................................................................... 242

3. Language as Intersubjective Coordination: An Experimental Setting ............................................................................... 247 3.1. Introduction ...................................................................................................................... 247 3.1.1. From distributed cognition to experimental pragmatics .................................... 248 3.1.2. Distributing and integrating cognition: the role of intersubjectivity .................. 250 3.1.3. Between synchrony, repetition and complementarity: how does landing a plane differ from dancing the tango? ..................................................................................... 251 3.1.4. The Enactivist Perspective: organisms and social interaction............................ 254 3.1.4.1. Back to the organism as the ground of sense-making.............................................. 254 3.1.4.2. The minimal ground of social interaction ................................................................ 258 3.1.4.3. Social enactivism under a pragmatist eye ............................................................... 263

3.1.5. Language as a tool for interaction and experimental approaches to pragmatics and semiotics ................................................................................................................ 271 3.2. Language as a tool for joint cognition: An Experimental Setting................................................ 274 3.2.1. Describing the experiment ................................................................................. 274 3.2.2. Measuring the cognitive performance ............................................................... 280 3.3. Opening the Black Box of Language ..................................................................................... 286 3.3.1. The structure of the interaction ......................................................................... 290 3.3.2. Skills needed for successful interaction .............................................................. 292 3.3.3. Sharing experiences ........................................................................................... 294

3.3.3.1. Language as a scaling device: the case of confidence communication ..................... 294 3.3.3.2. Language as a scaling device: back to the data ........................................................ 296 3.3.3.3. Scaling through negotiation and alignment ............................................................. 303

3.3.4. Language and Performance ............................................................................... 309 3.3.4.1. 3.3.4.2. 3.3.4.3. 3.3.4.4.

Relating degree of morphological convergence and group benefit ......................... 309 The degree of morphological convergence is not an effect of the model ................ 314 Scales and interactions ............................................................................................ 315 Scales and learning .................................................................................................. 316

3.4. A Pragmatist Assessment ..................................................................................................... 317 3.5. Conclusions ....................................................................................................................... 321

4. Conclusions .................................................................... 323 5. Bibliographical References ............................................... 353


0. Introduction: Thinking Through Language This dissertation is the field report of three exciting years of research, the trajectories of which are still unfolding. Perhaps unsurprisingly, it is about meaning. More precisely it is about how language and linguistic meaning fit into a more general model of ecological and dynamic cognition. I investigated this by attempting to bring a militant pragmatist and interpretive perspective into the emerging fields of cognitive and experimental semiotics. These overlapping approaches to one of the most ancient philosophical quests – meaning – are respectively an attempt to construct models of language, signs and meaning in interplay with the research on the human cognitive system and an attempt to formulate and test semiotic assumptions experimentally.

0.1. The Thesis of the Dissertation The central thesis of this work which functions as an initial assumption to be both articulated and argued for in the unfolding of the dissertation is that language is not a neutral medium for thought and cognition: language and communication are indeed integral parts of cognition crossing the boundaries between the external and the internal, between the bodily and the socially motivated. Following a conception in which cognition is habit-grounded, verbal patterns have a sensorimotor shade, a psychological edge and, yet, are neither subjective nor simply neural. They weave our worlds through social and intersubjective domains where explicit negotiation, implicit alignment and self-corrective unfolding help coordinating linguistic and nonlinguistic affordances and therefore joint action and cognition. In other words, linguistic meaning is constituted by the coordinated conceivable consequences of encountering and actively dealing with a linguistic structure in a context, which is constituted by physical, social and intersubjective affordances. Therefore, linguistic 12


signs are a bundle of operational conditions for cognition that interact with the local situation and evolve in interplay with their own history, contextual affordances and orientations of the ongoing practices. Their use shapes at different time scales the perceived and conceivable affordances of the situation in a ductile way that is open to manipulation, negotiation and alignment1.

0.2. The Reasons behind the Thesis After anticipating the conclusions of this dissertation, it is time to spell out what the questions are that lead this inquiry and what the rationale of pursuing such a conception is. The reason is a generalised lack of reflection on how language creates an intersubjective and social arena for cognition (cf. Chapter 1). On the one hand most models of cognition do not involve intersubjective and social domains or when they do only as the results of individual skills, on the other even most models of social cognition – in particular Theory Theory and Simulation Theory (cf. de Bruin & de Haan 2010) – do not involve any explicit role for the intersubjective use of language2 (cf. Chapter 3). This dissertation thus aims at creating a theoretical and operative basis for a method to analyse linguistically mediated cognitive coordination at different time scales and to start investigating in which ways cognition supports, constrains and/or co-constitutes cognitive phenomena. This double orientation, philosophical and applicative, is due to the fact that, in order to contribute to the cognitive enterprise, an in-depth semiotic reflection on the conceptual and philosophical framework of the cognitive enterprise has to articulate tangible consequences for the experimental and analytical practices of such enterprise. Otherwise, it will risk being but a theoretical exercise and betray the spirit of pragmatism, always inciting us to make differences and develop concepts only in the 1

2

This conception in particular will be articulated in the dynamics of metaphoric evolution and stabilisation (Chapter 2) and in the dialogical mechanisms of alignment in task-solving oriented conversations (Chapter 3). Interesting exceptions are the theory of narrative practices in Gallagher and Hutto 2008 and the idea of language as a tool for interacting minds in Tylén et al 2010.

13


light of their conceivable consequences. A second line of motivations behind this research is the attempt at conjoining semiotic reflections and conceptual tools to cognitive research. The concepts of habit, continuity, diagrammatic reasoning and meaning potential that I will articulate and apply in the course of this dissertation have important conceivable consequences for the cognitive enterprise. At the same time their application in less usual domains – that will involve not simply texts but diachronic corpora, sometimes in conjunction with correlated measures of performance – can also contribute back to semiotics, giving it a role in the cognitive enterprise in the shape of a cognitive and experimental semiotics.

0.3. The Phenomena Analysed To argue for a pragmatist conception of language and to articulate it into application-oriented points I am going to travel through three interrelated lands. In Chapter 1 I will consider the way cognitive linguistics and the extended mind hypothesis have conceptualised the relation between language and cognition. In Chapter 2 I will explore what it is like to be a caiman, that is, how the metaphor “Berlusconi is caiman” is coined in the Italian newspaper La Repubblica, how it has evolved during its so far eight years of existence and what it implies for the conceptual representations at stake. The third and last chapter of this dissertation will push the pragmatist approach to a still widely unexplored frontier of cognitive studies of language: the free use of language in an intersubjective experimental situation. The experiment involves couples of participants having to reach a joint decision during a minimal perceptual discrimination task. The use of language significantly increases the performance of the participants both as pairs and as individuals even when compared to the performance of the best isolated participant. How does that happen? Can we automatically treat the linguistic data in order to find correlations between it and the joint performance? What does language afford and how are these 14


affordances made relevant and how are they negotiated? Already this brief exposition could cause the feeling that hopeless heterogeneity lures around the corner: the life of the concept of caimans in the social arena of the media and participants exchanging opinions in a darkened room textual analysis and corpus linguistics; philosophy of linguistics and cognitive sciences, experimental psychology and automatic treatment of language. How does all this hold together? The answer lies in the attempt at developing a pragmatist and semiotic approach to the articulation of language and cognition and at doing so by tackling both foundational and still unexplored boundaries of the disciplines involved. This engenders a disciplinary platypus – to use a metaphor coined by Umberto Eco (1997) – an animal that semioticians hold dear, apparently heterogeneous, apparently contradictory, nevertheless alive and whole and thought provoking. The image of language that emerges when taking into account these different angles and patches is that of a symbolic,

evolving

device

motivated

by

embodied

experiences

and

produced/interpreted through cognitive mechanisms largely overlapping with the sensorimotor systems, but at the same time it is of language as “public” and “external” and thus as intermingling our most basic and intuitively internal processes with both the continuously evolving complexity and with the stability of the social arena.

0.4. What does it take to be a Pragmatist? (Sketching the Overall Perspective) Let us start by defining the foundations of the Peirce-inspired pragmatist approach that weaves this research together keeping in mind that one of Peirce's crucial worries was “to construct a naturalistic but non-reductive account of the human mind” (Short 2007:ix). I am not aiming at making a philological reconstruction as many have attempted (cf. Hausman 1993, Fabbrichesi Leo 2002, Bergman 2009) but at sketching how a pragmatist perspective led my research practice through the 15


different issues tackled in the chapters of this dissertation. These principles will thus not be the premises for an unwarranted Peircean imperialism, they will be open for assessment and evolution in their applications and consequences, true to the spirit of pragmatism. To be a pragmatist it is necessary to equate one's ideas with the dispositions to action that they engender, with their conceivable consequences, with what they make us ready to do, consciously or not. It is a conception that aims at going beyond traditional oppositions like materialism vs. idealism, empiricism vs. rationalism, internalism vs. externalism. It is a methodological conception that forces us i) to define cognition – and meaning – in terms of their conceivable and evolving consequences; ii) to pay attention to the processes of mediation through which these consequences can be conceived and dealt with; and thus iii) to focus on the dimension of “living generality” (Peirce 1931-58 CP 6.151) that permeates any cognition, the constitution and evolution of any possible cognition. Let us go a bit more into details with how these positions are articulated in Peirce3 along the lines of his entrenched epistemology, phenomenology and semiotics, according to the principle that “reasoning should not form a chain that is no stronger than the weakest of its links, but a cable whose fibers may be ever so slender, provided they are sufficiently numerous and intimately connected” (EP 1:29). The term pragmatism appears officially for the first time in a lecture by William James in 1898. James himself, however, acknowledges Peirce's paternity of the term, while Peirce at a later point – in 1905 – would distance himself from "Schiller, James, Dewey, Royce, and the rest of us" (Peirce 1931-58 CP 8.205) choosing to define his own specific variation of pragmatism “pragmaticism”. Pragmaticism was in Peirce's eyes “a name ugly enough to be safe from kidnappers” (Peirce 1931-58 CP 5.414). While trying to find inspiration in Peirce's conceptions more than in those of others I

3

This is meant as an overview that should make the directions of this research more comprehensible from the beginning. More details will be given where relevant in the course of the next chapters.

16


will, however, use the term pragmatism to be safe from the accusation of philosophical kidnapping.

0.4.1.

A Theory of Cognition

The origins of pragmatism are often traced back to two papers from the late 70's – “The Fixation of Belief" (1877) and "How to Make Our Ideas Clear" (1878) (cf. Hausman 1993, ch. 1; Fabbrichesi Leo 2002) – our trip, however, will begin 10 years earlier, with the anti-intuitionist theory of knowledge that Peirce develops in 1868 in what the critics have called the “anti-Cartesian essays”. In "Questions Concerning Certain Faculties Claimed for Man" Peirce is extremely clear about what cognition is, that is, how we know something: •

We have no power of Introspection, but all knowledge of the internal world is derived by hypothetical reasoning from our knowledge of external facts

•

We have no power of Intuition, but every cognition is determined by previous cognitions

•

We have no power of thinking without signs

•

We have no conception of the absolutely incognizable (Peirce W2:213)

0.4.2.

A Theory of Semiosis

Cognition is thus an inferential and semiotic process. From sensations to complex and abstract thoughts, what we deal with is sign structures4. This calls for a definition of what a sign is. Signs are defined as the opposite of intuitions and intuition is “a term that by unanimous consent of philosophers stands for immediate cognition [, that is,] knowledge in which no other representation enters consciousness to mediate between it and its object” (Peirce 1981- W1: 515). Signs thus seem to constitute knowledge that is constitutively mediated by another representation. And this is exactly how Peirce articulates his most complete definition of signs:

4

Cf. Chapter 1 for an in-depth definition and exemplification of Peirce's conception of cognition.

17

Language and Cognition - Ph.D. Dissertation by Riccardo Fusaroli “A sign, or representamen, is something which stands to somebody for something in some respect or capacity. It addresses somebody, that is, creates in the mind of that person an equivalent sign, or perhaps a more developed sign. That sign which it creates I call the interpretant of the first sign. The sign stands for something, its object. It stands for that object, not in all respects, but in reference to a sort of idea, which I have sometimes called the ground of the representamen. 'Idea' is here to be understood in a sort of Platonic sense, very familiar in everyday talk; I mean in that sense in which we say that one man catches another man's idea, in which we say that when a man recalls what he was thinking of at some previous time, he recalls the same idea, and in which when a man continues to think anything, say for a tenth of a second, in so far as the thought continues to agree with itself during that time, that is to have a like content, it is the same idea, and is not at each instant of the interval a

new idea.” (Peirce 1931-58 CP 2.228). Figure 1 - The sign function in Peirce

We should notice a few things: i) The sign stands for its object not in general but only according to an interpretant and under a certain respect, or ground, or representamen. In other words it is irreducibly triadic. ii) The sign is a general, that is, a function and not an individual idea, a point strengthened by the fact that any “reference 'upon a person' is a sop to Cerberus, because I despair that my real conception can be understood” (Peirce 1977 LW: 80-1). iii) The distinction between Object, Representamen and Interpretant is entirely determined by this function: the Interpretant now stands in the same relation to the Object as did the 18


Representamen, that is, it is to the same degree a sign of the object. Thus, Peirce’s definition of the sign is recursive; the fact that the interpretant is in itself a sign implies that it has its own interpretant and that the previous Representamen is now an object5. The triadic sign function is a generic function that can be filled by any sort of phenomena including cognition at all levels, from basic sensations (§ 1.2.2.1) to higher level rational decision making (§ 3.4). Since cognition is constituted by sign functions, each instance of cognition stands for its object only under a certain respect, connecting it to a certain interpretant, that is, to a certain other cognition. This connection is not neutral or a side effect, but it is constitutive of the sign. Thus more than a triadic structure, the sign function seems to be an open network structure and it is wider network of cognitions that determines the single cognition or sign. How this works and what the role of generality in this is will be further clarified in the next paragraphs and throughout chapter 1.

0.4.3.

A Theory of Experience

After sketching Peirce’s theory of knowledge and his theory of signs, the description of a third fibre of his reasoning can be added for this purpose: this fibre being his phenomenological conception. Peirce's definition of signs, indeed, “involves explicit reference to the three phenomenological categories. The sign relates a quality (Firstness) to an actual object (Secondness) and, by doing so, involves a more or less general regularity (Thirdness)” (Stjernfelt 2007:23)6. Let us consider them briefly. 5

6

Short (2007) argues that this would lead to the domain of deconstructionism and complete autonomy of the semiotic domain from the world. However, as we will see, the notion of habit grounds this semiotic recursion to the domain of social practices enacted in the world. Cf. also “What is a sign, as such, as Peirce conceived it? [...] The answer, which I give to this, might be summarised in its most general form by saying that the idea of a sign is the idea of manifestation, that is, the idea of appearance. The world appears or manifests itself to us through signs: for Peirce, it is a mere tautology to say this. For that is what is meant by a sign, viz. that through which the world manifests itself. The various kinds of signs are, then, the various ways in which this can occur.” (Ransdell 1966:3-4). Since the world manifests itself through signs, signs are the object of phenomenology. This is a conjoining of semiotics and phaneroscopy coming from a direction opposite to the one sketched by Stjernfelt, but with analogous results.

19


Phenomenology – or as Peirce started calling it from 1904 “phaneroscopy” in order to distinguish it from Hegel's work – is: The Doctrine of Categories, whose business it is to unravel the tangled skein [of] all that in any sense appears and wind it into distinct forms; or in other words, to make the ultimate analysis of all experiences the first task to which philosophy has to apply itself. It is a most difficult, perhaps the most difficult, of its tasks, demanding very peculiar powers of thought, the ability to seize clouds, vast and intangible, to set them in orderly array, to put them through their exercises. The mere reading of this sort of philosophy, the mere understanding of it, is not easy” (Peirce 1931-58 CP 1.280).

André De Tienne (1996, 2004) and Tom Short (2007) have dedicated many pages to the task of constructing and investigating a phaneroscopy in its conjoining role between mathematics and the philosophical sciences of which it is first and ground. Such a task is quite daunting both because of the elusiveness of its object and because “phaneroscopy is still in the condition of a science-egg, hardly any details of it being as yet distinguishable, though enough to assure the student of it that [...] it surely will in the future become a strong and beneficent science” (Ketner & Peirce 1998: 328). For the task at hand, however, it is sufficient to point out that in a Kantian critical spirit Peirce defines three categories (Firstness, Secondness, Thirdness) that are conditions of possibility for any single cognition – and sign and phenomenon. These categories are logical forms to be disentangled but should not be confused with any real cognition, since they are always co-present, albeit in different proportions. Moreover, contrary to a strong trend in applications of Peircean semiotics and pragmatism, defining these categories and discovering them in the phenomena analysed, however, is not the end of the phaneroscopical analysis. “The phaneroscopist would also want to exhibit how these categories actually combine and cooperate to shape experience” (De Tienne 2004: 26). This said I can thus move to a definition of the three categories. Firstness is a quality in its pure possibility that has no reference to anything else and that is not even actualised yet. Firstness constitutes the quality of experience: in order for something 20


to appear at all, this something must present a certain constellation of qualitative properties. Secondness is “the sense of reaction between an ego and a non-ego” (EP: 268); it is the way the world opposes a resistance. Secondness consists in the insistency with which the individuated and actualised existent phenomenon appears. It is the domain of Existence and of the uncontrollability of experience. Thirdness is the generality of experience, its regularity and its being expectable. “The events of Secondness are never completely unique, such an event would be inexperientiable, but relates to other events due to certain features in them; Thirdness is thus what facilitates understanding as well as pragmatic action, due to its continuous generality” (Stjernfelt 2007:14). It is only through Thirdness – in its mediating of Secondness and Firstness – that we can define Reality. To better understand Thirdness I need to refer to another doctrine which is involved in Pragmaticism as an essential consequence of it, but which the writer defended […] before he had formulated, even in his own mind, the principle of pragmaticism, is the scholastic doctrine of realism. This is usually defined as the opinion that there are real objects that are general, among the number being the modes of determination of existent singulars (Peirce 1931-58 CP 5.453)

This is an important point: Reality is not pure Existence, but it is imbued with Thirdness, that is, with generalities, dispositions and tendencies. These dispositions are not theoretical abstraction; they are real, crossing and defining objects and events. The world is not neutral but imbued with and defined by affordances (Gibson 1979, cf. § 1.4). The mature Peirce is a strong opponent of nominalism, that is, of the doctrine according to which concepts and generalities are just linguistic labels. For Peirce – inspired in this by Scholastics – generalities are real and an important part of our experience and reality and have consequences and therefore their status has to be acknowledged (the whole dissertation is an attempt at showing the importance of doing so, cf. in particular §1.2.2.4, § 2.4.4.2, § 3.4 and § 4). Even more radically, in Peirce's pragmatism, experience, cognition and meaning are only possible through 21


the mediation of Thirdness, through the idea of generality that constitutes the ground on which qualities (Firstness) can be actualised (Secondness) on the ground of general dispositions (Thirdness). It is the conditional and purely potential generality of Thirdness that enable us to receive any singularity. With this emphasis on generality as a disposition that enables the constitution of the experiential phenomenon and that in some way anticipates it without determining it, we are finally close to the core of Peirce's pragmatism, a fourth fibre in our reasoning. Indeed, the pragmatic maxim that constitutes the foundation of this conception states: “consider what effects that might conceivably have practical bearings you conceive the objects of your conception to have. Then, your conception of those effects is the whole of your conception of the object.” (Peirce 1931-58 CP 5.2). What Peirce is stating here and will keep on stating through much of his work is that what we think – the cognition, the sign – is to be understood in terms of what we deliberately choose to do as well as what we are ready to do consciously or unconsciously. In other words, meaning resides in the pragmatic efficacy, which the object of our conception acquires during a certain process thanks to a disposition to respond, a disposition shared on a collective level and crucial in the production of sense. Or, recollecting Peirce's Biblical quotation: “by their fruits ye shall know them” (Peirce 1931-58 CP 5.13). One possible misunderstanding that should be dispelled immediately is in underestimating the “conceivable” property of these effects. William James (1907) would claim that the meaning of something is in what we concretely do with it; it is in its practical bearings. James claims that, had it not been for Peirce, he would have used the word “practicalism” instead (cf. Peirce & James 2000, Sini 2000). However, Peirce's pragmatism never gives predominance to action predominance, it states instead that meaning has to be connected to the whole potential and conditional series of the resolutions to act that the human organism in its situatedness is prepared to enact to display its understanding of a concept. In this sense Peirce insists so much 22


on the terms “conceivably”: “This employment […] of derivatives of concipere must then have had a purpose. […] I did not mean to say that acts, which are more strictly singular than anything, could constitute the purport or adequate proper interpretation of any symbol” (Peirce 1931-58 CP 5.402, note 3).

We have all we need on the plate and maybe even more than can easily be chewed. Cognition is a triadic structure only possible through mediation of previous cognitions that create a ground of generality on which the single cognition can be constituted. But what is cognition? Nothing else than the sum of the conceivable consequences of its objects, that is, the mediating generality is nothing else than a disposition towards further possibilities. Cognition is a pragmatic structure that stands for an object according to its conceivable consequences, thus constituting it as meaningful. Cognition is semiosis. Its meaning, its structure and its flow are structured by the pragmatic maxim, that is, by the conceivable consequences and their evolution. However, how do consequences become conceivable? Or, in other words, what is the structure of this mediation? And, given the scope of this dissertation, how does language enter this picture? I will develop these themes in the next chapters (for the structure of mediation and language cf. §1.4, for concrete examples of how the conceivable consequences evolve § 2.3 and § 3.3). It is enough for now to re-state – now with Peirce backing us – that to be a pragmatist means a) to define cognition in terms of its conceivable consequences; b) to pay attention to the processes of mediation through which these consequences can be conceived and dealt with; and thus c) to focus on the dimension of “living generality” (Peirce 1931-58 CP 6.151) that permeates any cognition. We will see this better through the articulation of the concept of habit in Chapter 1. These are the principles that led my investigations and that will make possible a unitary picture and a development of such a pragmatist view in dealing with the articulation of language and cognition. 23


0.5. The Structure of the Dissertation 0.5.1. Articulating Language Overview on Chapter 1

and

Cognition

–

An

Chapter 1 is an investigation of the main concepts at stake in this dissertation: language and cognition. First I will sketch the developments that the notion of cognition has undergone since its inception by cognitive sciences in the late 40's with the aim of understanding its recent situated turn: the embodied, embedded and distributed conceptions of cognition. This will thus be re-framed in the pragmatist framework already partially unfolded in this introduction. Afterwards two of the most debated approaches to language in cognitive sciences – cognitive linguistics and the extended mind hypothesis – will be debated. Both have interesting philosophical reflections behind them. Both present interesting applications. However, they seem to stand on opposite sides: cognitive linguistics attempts to explain language through pre-linguistic cognition; the extended mind hypothesis conceptualises language through the idea of external material symbols allowing humans to go where no prelinguistic cognizer had gone before. To solve this opposition it will be argued that these approaches need to thematise the symbolic and social horizon for language as crucial nodes for a distributed and pragmatist conception of cognition more explicitly. Habits are considered as the basic semiotic structure of cognition, a structure that relies widely – phylogenetically, ontogenetically and/or through its unfolding – on an environment that is physical but also social and intersubjective. Language, as an integral part of human cognition, dwells upon and deeply reconfigures non-linguistic cognitive processes. Thanks to its symbolic and public nature, language widens the social horizon of cognition, allowing the intersubjective and cultural construction and manipulation of affordances at different time scales. The sensorimotor grounding of habits and language is not forgotten here –what 24


happens is that its engagement in the public and symbolic arena of human action and cognition re-articulates it. The environment and the body itself are reshaped by these symbolic dynamics, and at same time cognition – that relies on environment and on language – is reshaped at its very basis. Human habits come to fit the social horizon, sometimes through the interiorisation of these new structures (the possibility of inner speech for instance, cf. Magnani 2009 for other examples), sometimes through the co-constitution of cognition in interactions with the environment (like in the adaptive coordination required in dancing, cf. Kimmel 2010, Kirsh 2010). Language emerges from this chapter as deeply grounded in an ecological conception of pre-linguistic cognition and at the same time as a way of engaging cognition in a public arena and deeply reshaping it. This chapter lays the foundations for the theoretical and linguistic analyses of the successive chapters. It articulates a pragmatist semiotic perspective on the concept of experience and the role of language in it. It sketches the fundamental aspects of the concept of habit, suggesting a grid of elements to be taken into account when analysing meaningful cognitive processes, in particular their distributedness, their flexibility and evolution. It also begins to integrate the notion of meaning potential with the notion of symbols. All these elements will be crucial in the second chapter. As related to the third chapter, the first one emphasises the social and intersubjective arena through which language engages cognition, in all its dimensions and time scales. It tackles as well the issue of the radical change that language triggers in cognition, even in its non-directly linguistic aspects. All these elements will be crucial in the analysis of the experimental setting in the third chapter. This chapter constructs a framework for the analysis of symbolic dynamics in action through time, in their being crossed and constituted by different semiotic and cognitive systems.

0.5.2.

Language

and

Conceptual 25

Structures

–

An


Overview on Chapter 2 To see this interplay of conceivable consequences (affordances) sedimented and coordinated at different time scales I tackled one of the cornerstones of cognitive linguistics and of its philosophical foundations: the analysis of metaphors. In particular I focused on the two main approaches to metaphors in cognitive linguistics: Conceptual Metaphor Theory (CMT) and Conceptual Integration Theory (CIT). The questions asked are: Which conceptions of cognition and experience are brought into play in these model? What exactly are conceptual metaphors and where are they located? When one considers the widespread metaphorical patterns we can see in action in language and cognition, how do these general patterns unfold locally? How do the local specificities of metaphors reverberate and interact with these patterns? What is the role of language and of the social arena in all that? The careful consideration and critical assessment of CMT's and CIT's philosophical and methodological aspects is then gonna be complemented by the analysis of “Berlusconi is a caiman”, a family of linguistic metaphors based on the conceptual metaphor PEOPLE ARE ANIMALS7. “Berlusconi is a caiman” was coined in 2002 by Franco Cordero and has gone through a particularly intense “social life” – as de Saussure would have put it. The possibility of collecting a diachronic corpus – and therefore of investigating its local metaphorical variations due to genre, style and time – seemed to me a unique possibility to highlight some of the dynamics between a more embodied motivation and a social arena, thus covering some of the conceptual cramps and blindspots of Conceptual Metaphor Theory. This will thus enable me to further articulate the pragmatist perspective on the role of language in cognition sketched in chapter 1. Which verbal patterns are involved in the linguistic deployment of an innovative metaphor? How do they evolve in time? Which insights 7

The wider semantic and conceptual pattern underlying a set of linguistic metaphors – a conceptual metaphor – is according to a widespread convention from now on indicated in capitalised letters.

26


do they grant the conceptualising mechanisms at work? The notions of meaning potential, evolution of symbols and implicit diagrammatic manipulation in particular are in the foreground.

0.5.3. Linguistically Interacting Minds – An Overview on Chapter 3 If the second chapter was an incursion into the social aspects of language use as enacted in a very specific media landscape, the third chapter focuses more on local intersubjectivities, the local dynamics in them and how they interrelate with nondirectly linguistic cognitive processes. The research exposed in this chapter is a wonderful example of Secondness and chance irrupting in the phenomenal world while being at the same time more implicitly prepared by self-corrective dispositions (Thirdness). In the course of my visiting period at the University of Aarhus I happened to start collaborating with the Center for Functionally Integrated Neurosciences

and

in

particular

with

the

project

“Interacting

Minds”

(http://www.interacting-minds.net). I was struggling unsuccessfully with a few experimental designs that would enable me to observe and manipulate in the lab some of the aspects of language and meaning I had been articulating in the messy notes what would have become Chapter 1 and Chapter 2 of this dissertation. Were they overly ambitious or too simple, they simply did not act as the bridge between semiotic and cognitive practices and models that I was hoping to develop. Until one day, during one of the weekly meetings of Interacting Minds I was asked to become a subject for the pilot of an experiment (initially designed by Bahador Bahrami) on how intersubjective linguistic interaction modifies the ability of making decisions based on perceptual discrimination. My participation to the experiment led to a lengthy discussion on the initial results and on the linguistic and intersubjective dynamics at play. This led to a more direct involvement in the project – with the collaboration of Kristian Tylén – through the 27


collection of videos of the interacting participants, the transcription of the interaction, the development of qualitative and quantitative methodologies of analysis that would highlight correlations between interactional patterns and indexes of performance, as well as the development of variations of the experiment (still ongoing) that could test these correlations. This chapter develops the problem of constructing and coordinating a distributed cognitive system and the related issue of the role of individual agencies in the system. An initial part tackles the notion of intersubjectivity, articulating a critical assessment of the enactivist approach and its way of conceptualising social interactions. In order to introduce the role of language in the coordination, distribution and integration of cognitive systems, I then re-assess a crucial experiment in arguing for the role of material symbols in (chimpanzee) cognition and introduce the field of experimental pragmatics. This creates the theoretical ground on which to present the experiment to be then analysed and the findings a pragmatist framework can lead to, focusing on how linguistic affordances are employed, negotiated and aligned during the solution of this task and how the construction of this shared linguistic common ground reshapes non-linguistic cognitive processes. This incursion to the still untamed frontiers of cognitive linguistics, where the production and understanding of linguistic material can be manipulated and where measures of cognitive performance can be compared to linguistic patterns, thus enables me to develop the pragmatist framework not only in the direction of a cognitive, but also of an experimental semiotics.

0.5.4.

Conclusions

In the conclusions of the dissertation I will thus draw an account of this enterprise. In which ways a pragmatist semiotic framework enabled me to develop and apply conceptual tools to understand the role of language in cognition? Which tools can be made available towards the goal of a cognitive and experimental semiotics? How can 28


such a semiotics contribute to cognitive linguistics and cognitive sciences in general? And

what

can

this

dialog

contribute

29

back

to

semiotics?


1. The Social Horizon of Embodied and Extended Language 1.1. Introduction: Beyond the Informational Paradigm In a traditional information-processing view – as prototypically portrayed by the information theory developed by Shannon (1948) – language is simply a code that allows the transmission of previously elaborated information from one cognizer to another, just another kind of input and output. This model was developed to describe and improve the transmission of information through telephonic cables where it had a perfect adequacy. It became, however, the prototype and legitimising model for a widespread

conceptual

metaphor

–

or

ideology

(Leezenberg

2011)

-,

COMMUNICATION AS A CONDUIT, that is still widely implied in commonsense (Reddy 1979, Lakoff & Johnson 1980, Grady 1998 and Semino 2006), as well as in philosophical and scientific conceptions (Ronchi 2003, Leezenberg 2001)8. “Think before you speak” - an often offered suggestion or order gets 60 million of hits on Google – with its normative dimension displays the tragedy of a cognizer having to stop and think before any communicative act, not being able to engage in fluent interaction and conversation. But what if we do think through language and not in isolation but through conversations – linguistic or not – between ourselves and others, between ourselves and cognitive artefacts like books, software and devices, between ourselves and the culture we are embedded in? Indeed, both the conception of cognition as a coding/decoding function cushioned between input and output and this radical divide between cognition and language are

8

The conceptual metaphor was around long before Shannon and Weaver, albeit in different forms. Cf. for instance John Locke's Essay Concerning Human Understanding: “For language being the great conduit, whereby men convey their discoveries, reasonings, and knowledge, from one to another, he that makes an ill use of it, though he does not corrupt the fountains of knowledge, which are in things themselves, yet he does, as much as in him lies, break or stop the pipes whereby it is distributed to the public use and advantage of mankind.” (Locke 1698: 3.11.5).

30


being questioned in a crescent number of philosophical and empirical studies9. On the one hand cognitive linguistics – following the lead of authors like Ronald Langacker, Leonard Talmy, George Lakoff, Mark Johnson, Mark Turner, Gilles Fauconnier, etc.10 – attempts to reintegrate – and often reduce – language production and comprehension in cognition, showing that such processes are shaped at different levels by the sensorimotor system. On the other hand, situated cognition and enactivist approaches attempt to investigate what language adds to – and how it reshapes – non-linguistic cognition. Particularly representative of such approaches, Andy Clark (2006) developed the concept of material symbols as a crucial element of the “extended mind” hypothesis; Shaun Gallagher and Dan Hutto (2007) sketched how shared linguistic narrative practices could explain higher level understanding of the minds and actions of others and Kristian Tylén and colleagues (2010) investigated language as a tool for interacting minds in establishing a shared level of cognition. The aim of this first chapter is thus to compare explicitly and articulate the cognitive linguistics approach to the extended mind using the Peirce-inspired pragmatist framework that I sketched in the introduction, and to do so elaborating on the yet-tobe-introduced concept of habit. The symbolic nature of language and its embodied motivation can be held together within an intersubjective and social horizon11 that 9

10

11

I do not claim that this is the first time such a critique is brought forth in the history of Western ideas. However, it is only with Lakoff and Johnson that there is a lasting change in research practices in different disciplinary fields, therefore being significative enough as a starting point. In certain contexts "cognitive linguistics" refers to Chomsky's syntactical theories. I refer here exclusively to the tradition born from the authors just referred to. For an up-to-date treatment of this cognitive linguistics: Geeraerts & Cuyckens Eds. 2007. With “intersubjectivity” and “intersubjective” I will from now on refer to interactions where one or more people are directly involved in a dialogic process of coordination (cf. Matusov 1996; Rosero et al. 2010). With “social” I will refer to a normative dimension which is due to an agency that transcends the individual (a social institution for instance): this normative dimension does not, anyway, directly determine the individual, it is instead a tendency that, while transcending the individual, can be accessed through critical reflection and employed locally as a resource or manipulated. Two people having to pass through a door will reciprocally coordinate in a dialogic way since both of them are active participants in the interaction. This is the local intersubjective aspect. The social dimension can be found, for instance, in the tradition of letting the woman pass first, a tendency that can be locally engaged following it, or consciously accessed and made object

31


articulates, constrains, negotiates and sometimes denies motivations and symbolicities at play. Thus a pragmatist view of cognition can become the horizon within which cognitive linguistics and the extended mind hypothesis can complement each other. To follow up on this aim, i) I will investigate the evolution of the notion of cognition which in crossing different trends in cognitive sciences can be articulated with Peirce's pragmatism and his concept of habit. ii) This framework will be put to the test through re-articulating the debate between two apparently opposite positions on the relation between language and cognition: that of cognitive linguistics and that of the extended mind hypothesis. Finally, iii) the concept of habit will be applied in order to re-construct the articulation of language as social cognition.

1.2. The Cognitive Stance 1.2.1.

A thread through cognitive sciences

What is cognition? Or better, how to distinguish something that is cognitive from something that is not, that is, how to define the “mark of the cognitive” (Adams & Aizawa 2008)? This is a task that is far from trivial – Clark's and Chalmers' (1998) attempt at extending cognition and mind engendered a rich and complex discussion on the definition of what is extended (cf. Menary 2006, Adams & Aizawa 2001, Adams 2010, Clark 2008, 2009, Rowlands 2009a, Wheeler ms., Di Paolo 2009, Rupert 2009, Gallagher 2010). I cannot hope to solve the problem here: a whole dissertation could be dedicated to it. It is, however, possible to briefly go through the history of cognitive sciences to see the developments that the concept of cognition has gone through, both regarding the nature of cognitive processes and their location. This will be compared to and articulated with Peirce's conception of cognition, in order to produce an operative definition that, far from being definitive, will nevertheless allow us to see which role language could have in it. of a (lame) joke: “I'd let you pass but we would become a cliché”.

32

Language and Cognition - Ph.D. Dissertation by Riccardo Fusaroli 1.2.1.1. Bastion of the Old: Behaviourism

Cognitive sciences were born in the late 40's – and received this name in the 70's – as a reaction to behaviourism, an approach that tried to analyse action, thought and perception in terms of behaviours, that is, in terms of publicly observable processes without recourse either to internal physiological events or to hypothetical constructs such as the mind. In other words, behaviourism was the scientific study of organisms – and human beings – in terms of relations between the input of stimuli and the output of physical behaviours. Inner mental states were considered an unnecessary element in the theory – independently from their actual existence – since no knowable difference could be argued between two states of mind unless there was a demonstrable difference in the behaviour associated to each state. “Perception became discrimination, memory became learning, language became verbal behaviour, intelligence became what intelligence tests test” (Miller 2003:141). 1.2.1.2. The Cognitive Revolution: tackling the filling of the sandwich

Behaviourism was the dominant paradigm in the scientific study of human psychology12 until two conferences took place: “Cerebral Mechanisms in behaviour” hosted in September 1948 at CalTech and “Symposium on Information Theory” hosted in September 1956 at MIT. These conferences were crucial in getting together some of the people who had been trying to articulate scientific alternatives to behaviourism like von Neumann, McCullough, Lashley, Newell, Simon, Chomsky, Miller, Bruner, Goodnow and Austin13. This led to the foundation of a journal – Cognitive Science – in 1977 by Roger Schank, Gene Charniak and Allan 12

13

An interesting alternative to behaviourism was Gestalt Theory. However, the ascent of the Third Reich disrupted its German initial momentum and the Gestaltist refugees in the States did not find fertile ground for their perspective, especially considering the fact that they were not allowed to have PhD students (Ash 1999). Some of the Gestaltist insights, however, would pass into ecological psychology through the connection between Koffka and Gibson (Niveleau 2006), whose concept of affordance we will encounter later on in this dissertation. It has to be noted that European psychology had not complied to the behaviourist, and that European scholars like Sir Frederic Bartlett, A. R. Luria and Jean Piaget constituted an inspiration for the cognitive revolution.

33


Collins and in 1979 to the foundation of the Cognitive Science Society. The basic idea pushed forward by this growing movement was that rather than being imposed from the outside – as a reaction to an external input – the organisation of behaviour emanates from within and this internal organisation, cognition, is what we should be studying, tackling it from as many disciplinary perspectives as possible:

Figure 2 – Connections among the Cognitive Sciences according to the Sloan Report

How to make sense, though, of this crossroad of disciplines? Howard Gardner in his still insightful reconstruction of the origins of the discipline describes the new

34


paradigm in the following way: “Cognitive Science14 is a contemporary empirically based effort to answer long-standing epistemological questions – particularly those concerned with the nature of knowledge, its components, its sources, its development and its deployment” (Gardner 1987:6). Following Gardner's well-researched investigations and interviews with the pioneers of the field, five shared methodological beliefs can be seen as defining the cognitive enterprise: •

Mental representations have to be analysed at a wholly separate level from the biological or neurological on the one hand and the sociological or cultural on the other.

•

The computer is central to any understanding of the human mind, both as a tool for data analysis and simulation and as a model for cognition.

•

There is a need to deliberately de-emphasise certain factors, which may be important for cognitive functioning but whose inclusion at this point would complicate the cognitivescientific enterprise unnecessarily. These factors include the influence of affective factors or emotions, the contribution of historical and cultural factors and the role of the background context in which particular actions or thoughts occur.

•

Much is to be gained from interdisciplinary studies.

•

A key ingredient in contemporary cognitive science is the agenda of issues and the set of concerns, which have long exercised epistemologists in the Western philosophical tradition.

The revolutionary object of cognitive science is thus cognition: a level of “mental representation”, of internal mediation between input and output, perception and action, which was largely ignored by behaviourism. In the effective words of Susan Hurley (1998) cognition as storage and manipulation of representation is the filling of a sandwich, a filling to which cognitivists are finally getting. Mental representation has to be analysed at a level wholly separate from a biological or neurological level on the one hand and a sociological or cultural level, on the other15. The idea – 14 15

The singular form is programmatic, in that it is an attempt at constructing a unitary identity for this interdisciplinary field. It has to be noted, however, how Gardner himself underlines both the philosophical situatedness of the research in cognitive science and the temporary nature of the exclusion of issues like context,

35


inspired and supported by the incredible success of computer science and artificial intelligence – is that cognition follows the principles laid by Frege – reasoning strictly follows the form of propositions and their content does not matter – and Turing – this form is symbolic and manipulated according to rules. Concepts – the bricks of mental representation – are often represented as lists of features (cf. the “classical view” in Smith and Medin 1981) organised and symbolically manipulated in terms of predicates that stand for features of the external world. In Jerry Fodor's words: “What I am selling is the Representational Theory of Mind. […] At the heart of the theory is the postulation of a language of thought: an infinite set of 'mental representations' which function both as the immediate objects of propositional attitudes and as the domains of mental processes” (Fodor 1987: 16-17). To think means to store and manipulate symbols in a particular manner regardless of the domain. The formal manner of manipulating symbols constitutes the syntax of the domain, syntax that is thus not only a unique formal property of linguistic computation but also a more general property of all cognitive domains from vision to motor to problem solving ones. Cognition is conceived as a language-like16 inner layer that mediates between action and perception thus enabling human beings to not simply react to the world but to represent it, to manipulate these representations and as a consequence to come up with more complex patterns of behaviour that better serve their aims and desires. This initial cognitivist approach is quickly attacked from at least two inner fronts: i) its formal mechanisms; ii) its internalism, with the related issue of the ownership of cognitive processes. 1.2.1.3. Beyond Symbolic Rules

16

history and culture from cognition. It has to be noted, however, how language-like and formal do not mean “explicit to the possessor” as much as “explicit to the theorist”. Human beings can act according to proposition-like reasoning without explicitly thinking in syllogisms. The focus is on the cognitive system and not on the consciousness of the subject.

36

Language and Cognition - Ph.D. Dissertation by Riccardo Fusaroli 1.2.1.3.1.

Neural Networks

An important attempt at going beyond such a computational approach was made through the re-discovery of neural networks17 as models of the mind/brain complex (cf. Clark 2001). Initially developed by Warren McCulloch and Walter Pitts (1943), neural networks were brought back to the mainstream scene by David Rumelhart, David McClelland and the PDP Research Group (1986). While classical computationalists explain cognition by reference to language-like structures and their formal manipulation, connectionists explain cognition in terms of neural networks, that is, networks of simple processing units, a strong simplification of the networks of neurons found in the brain. None of these simple processing units need to represent anything, at least not in the strong symbolic sense of classical cognitivism. Nevertheless, connectionist models are able to behave in ways that mimic the behaviour of human beings even in the linguistic – and most prototypically symbolic – domain (cf. Elman 1995, Zlatev 1997). These models criticise the conception of formal representations while not – at least not in the beginning – moving beyond the focus on internal structures of mediation between perception and action. Symbolic representations and the direct programming of algorithmic rules are replaced by connections between sub-symbolic nodes and learning processes. Symbols are not outright rejected – however, their role changes: from being the constituents of cognition they become a by-product of constraint satisfaction processes, that is, of statistical pattern recognition, completion and disambiguation, a present response to an input of endogenous or exogenous origin. Symbols as transient constructs internally reflect the local context in their very structure. To justify the claim that there are context-free atoms acting as the representational baseline of such 17

With this definition I aim at emphasising the common ground of connectionism and computational neurosciences: The term ‘connectionism’ is primarily used for neural network models of cognitive phenomena constrained solely by behavioural (as opposed to neurophysiological) data. By contrast, the term ‘computational neuroscience’ is primarily used for neural network models constrained by neurophysiological and possibly also behavioural data (cf. Piccinini & Scarantino 2010).

37

Language and Cognition - Ph.D. Dissertation by Riccardo Fusaroli systems, we would need to discover syntactic structures which persist unaltered and carry a fixed content. The only contenders for such persistent content bearers are the weights which moderate the various context-sensitive activity patterns. But a given weight, or set of weights, cannot be identified with any fixed content in these super-positional systems, for each weight contributes to multiple representational abilities. It makes no sense to ask what this or that weight means; it is only relative to some specific inputs that the weights give rise to activation patterns which can be seen as expressing this or that (context-involving) content. (Clark 1993: 33)

1.2.1.3.2.

Dynamical Systems Theory

A second attempt at overcoming classical computationalist approaches comes from the mathematics of physical systems. Already in the 50's cybernetics adopted the mathematical apparatus now known as Dynamical Systems Theory – a theory which is based on systems of differential equations – to model cognitive behaviour as adaptive real-time embedded activity (Ross Ashby 1962). This conception was taken out of the shelves again in the late 80's and reached its peak in the 90's (Port & van Gelder Eds. 1995, Kelso 1995, Keijzer 2001). Dynamical Systems Theory describes how global order (a pattern) emerges from the collective action of a large amount of interacting components. Such a pattern – i.e., the collective action of the components – in turn restricts the degrees of freedom of the components of the system (“global-tolocal influence” or “downward causality”). Consider, for instance, the following analogy of the design of human minds. The human mind is like the architecture of a termite nest rising several feet above the ground, often reminiscent of typical buildings for humans (Oudeyer 2006). Termites build their complex nests not because each one has some overall architectural concept for the nest’s design and the steps needed to build it. Instead, each termite engages in very simply behaviours such as ‘‘If I come along a lump of earth, I should pick it up, and place it where the pheromone signal is strongest’’. The eventual superstructure that is built is the result of dynamic interaction in the specific environment of thousands of individual termites and not a projection to the macroscopic level (i.e., the nest) of information encoded at the microscopic level (e.g., the minds of individual termites). This is an 38


instance of “reciprocal causality” (Varela & Thompson 2003) given the downward force that an emergent feature exerts on lower-level behaviours. The notion of reciprocal causality emphasises that the global or macroscopic order that emerges from the non-linearly interacting local components has its own intrinsic features, duration and domain of causal efficacy. This does not mean that the existence of the macroscopic order is somehow independent of the microscopic processes. The macroscopic order is the relation of the components; global-to-local influence is the influence on the system’s individual components as it is exerted by the way in which the components of the system relate to each other (cf. Thompson and Varela 2001: 420; Thompson 2007: 426–427). The basic innovations of such a perspective are two: 1. The idea that cognitive processes may arise as products of self-organisation without there necessarily being specialised internal mental representations that serve as the causal basis for complex behaviours (e.g., speaking metaphorically), since self-organisation can occur within the individual mind but also within minds through dynamic activation and inhibition of lowerlevel cognitions; 2. The idea that cognition is a question of trajectories more than states. In the previous symbolic and in most sub-symbolic approaches “the mind works in a staccato fashion of entertaining one discrete stable non-overlapping representational state for a period of time, and then instantaneously flipping to entertain a different discrete stable non-overlapping representational state for another period of time” (Spivey 2007: 8). On the contrary Dynamical System Theory claims that the brain works as a predicting statistical device, keeping open more than one possibility at once and pondering the relative probability of each through the unfolding of the situation. Each brain neuron corresponds to one dimension of that space, which thus has a billion or so

39

Language and Cognition - Ph.D. Dissertation by Riccardo Fusaroli dimensions. At any given moment, the total state of brain activity corresponds to a single point in the space. Changes in that activity over time then produce trajectories through the space. Regions of the space to which many trajectories go (and where they sort of stay) are called attractor basins. In many contexts a given attractor basin corresponds to a fully developed percept — to a word understood, a face recognised, a stable perceived version of the Necker cube. The attractors are thus very important, but [DST] is even more interested in the trajectories themselves. The basic units of [its] thinking are events, not states (Neisser in Spivey 2006:ix).

The behaviour of a system over time is often portrayed as a continuous tracing of a line in a three-dimensional space. Time is represented in the continuing trace and other variables are embodied in the different axes. The term phase space refers to the set of possible states of the system. As the system changes states over time, it traces a trajectory in its phase space landscape – a path of the successive states it occupies. When a system’s behaviour is observed over an extended period, it sometimes happens that certain regions of the phase space are occupied often, others occasionally, and others never. An area of phase space the system occupies or approaches more frequently than others is called an attractor. An attractor exerts a kind of pull on the system, bringing the system’s behaviour close to it. Each attractor can be seen as a basin or valley in the phase space landscape, its region of attraction. Trajectories that enter the basin or valley move toward that attractor. Most psychological models assume that attractors are created by repeated experiences of a particular state so that the state becomes ‘‘engraved’’ in the person’s relevant psychological system. A shift from one attractor to another is called a ‘‘phase transition’’ or ‘‘phase change’’. For example, looking at an ambiguous figure such as a Necker cube or the classic vase/face silhouette produces bistable behaviour as the image shifts back and forth when you look at it. This shifting in the figure’s appearance (e.g., between a vase and a silhouette) is consistent with a dynamical account of a nonlinear trajectory settling into one attractor basin and then into the other, repeatedly (Spivey 2006). In systems with more than one attractor, the system’s trajectory typically approaches each one periodically but is never fully captured by 40


any of them. Plotting the behaviour of such a system over time shows a tendency to approach the various attractors, but often unpredictably. Shifts from one attractor to another may even seem random. A central consequence of a dynamical account of human behaviour is the realisation that the majority of the trajectory’s time is spent in intermediate regions of state space that gravitate toward multiple semi-stable attractor basins. Many of DST most well-known proponents (van Gelder 1995, Beer 1995, Thelen & Smith 1994) have pushed anti-internalism, to one degree or another, as a consequence of the dynamical view: Since the nervous system, body, and environment are all continuously evolving and simultaneously influencing one another, the cognitive system cannot be simply the encapsulated brain; rather, it is a single unified system embracing all three. […] inner and outer process are coupled, so that both sets of processes are continually influencing each other. (Port & van Gelder 1995:13).

This is not, however, a logical consequence of the dynamical systems mathematical approach. It is possible to study the brain's evolving activity without modelling the environment, as every single example in Spivey's (2006) The Continuity of Mind demonstrates. Moreover, a too strict coupling between the nervous system and the environment does not seem auspicable. After all, I am here writing about abstract concepts – models of cognition – while the sun is shining and all sorts of not related enjoyable things happen outside of my window18. Cognition is characterised by the capacity to at least partially decouple itself from the environment. The opposition between internalism and externalism is thus orthogonal to the critique of the formal 18

After writing these lines I stumbled upon a similar and better formulated example by Rick Grush: “The problem is that the tight coupling thesis is simply and demonstrably and (I would have thought, obviously) false. […] Pre-theoretically […] many paradigmatically cognitive capacities seem to have nothing at all to do with being in a tightly coupled relationship with the environment. I can think about the St. Louis Arch while I'm sitting in a hot tub in southern California or while flying over the Atlantic Ocean. […] The lesson is that in at least some cases, representing and cognizing is a matter of breaking the coupling, of getting out of the causal loop. But one cannot just split a coupled dynamic system in half and expect each half, especially the agent half, to do anything at all” (Grush 1997: 10-11).

41


mechanisms of classical computationalism (Wilson 2004). To better understand this I shall use the concept of “vehicle”. The distinction between vehicle and content is seminally introduced by Dennett (1969, 1978) and then popularised by Cussins (1990), Dennett (1991), Millikan (1991) and Hurley (2001). “We all know about the vehicle content distinction. We shouldn't confuse properties represented in content with properties of vehicles of content. In particular, we shouldn't confuse the personal and the sub-personal level. Vehicles of content are causally explanatory sub-personal events or processes or states. We shouldn't suppose that the properties of vehicles must be projected into what they represent for subject/agents, or vice versa. This would be to confuse the personal and sub-personal level” (Hurley 1998:1)19. What the previous discussion is about is the nature of vehicles. For traditional computationalism the vehicles of cognition are symbols that have formal – or syntactic – properties in virtue of which they are processed. For connectionism they are patterns of activation distributed across nodes in a network. But what about the location of such vehicles? 1.2.1.4. Beyond the Brain and the Nervous System

It is thus time to tackle this second branch of critiques. The internalistic conception of cognition would quite soon reach a crisis. The attempt at studying human and animal behaviours and cognitive performances and at simulating them via algorithmic software and robots runs into serious issues. Simple everyday tasks like riding a bike require complex loops of anticipations and feedbacks that are hard to implement in a sandwich like system. Animal behaviours relying on relatively simple neurobiological structures reveal themselves to be impossible to reproduce in robots and computer simulations of comparable or even superior complexity, as Beer and Chiel (1993) wonderfully demonstrate it with their analysis and simulation of cockroach behaviours. While much research retracts to safer and more abstract domains like chess strategies, a growing number of researchers start acknowledging 19

On personal and sub-personal, cp. 2.2.4.3

42


that – extending on Hurley's metaphor – the filling intermingles with the bread, no clear-cut boundary between cognition, action and perception is possible anymore if we want to understand what is really going on20. Clark dubs this alternative vision ‘Escher spaghetti’: “a seething mass comprising not just multiple criss-crossing strands (ordinary spaghetti), but strands whose ends feed back into their own (and others) beginnings, making ‘input’ and ‘output’, and ‘early’ and ‘late’ into imprecise and misleading visions of complex recurrent and re-entrant dynamics” (Clark 2009). The vehicles – supports – of cognition start expanding and catching in their “spaghetti grasp” more and more components beyond the neural system with consequences that do not leave the contents of cognition unshaken.

1.2.1.4.1. Embedding Cognition in the Body A first boundary that was crossed, to go beyond internalism21 was to bring the body into the picture. Cognition happens through the body and not just at an algorithmic or internalistic level. Embodiment becomes a key term in several strands of cognitive research that tries to overcome the “sandwich” conception of cognition and it is a term so effective that by the 1990s several authors had declared embodied cognitive science to be a new paradigm in cognitive science (cf. Varela, Thompson and Rosch 1991; Clark 1997; Pfeifer and Scheier 1999). The way the body is brought into the picture, however, is controversial. Several scholars (Nunez 1999, Clark 1999, Wilson 2002, Gibbs 2006, Gallagher 2007, Rohrer 2007, Violi 2007, etc.) have attempted to discriminate between the different uses of “embodied”. Relying on their work and

20

21

Similar and more radical positions had been expressed already in the 60's and the 70's by Jerome Gibson and his school of ecological psychology influenced by Gestalt psychology. As Jerome Gibson (1979) provokingly put it, it is not what is inside the head, but what the head is inside of. However, it is only with the turn I am describing now that the situatedness of cognition starts leading mainstream research. The three steps I am going to describe do not represent a temporal development, but a post-hoc taxonomy of the internalism/externalism debate. Different strands of research investigated different elements, sometimes overlapping them: body, context, sociocultural and the role other agencies in cognitive processes at different times and in different orders.

43


developing it we can distinguish at least four basic uses. The first – and most trivial – way of talking about embodied cognition is to claim that any cognitive process has an instantiation through a material support (as in Churchland 1992). A particularly salient case is embodiment as the biological materiality of the cognitive processes: cognitive processes happen through the nervous system and the brain and an understanding of the specificity of these neurobiological structures is necessary in order to understand cognitive processes. A second and more interesting way of talking about embodied cognition is to consider more basic bodily functions like perception and motor control as primary shaping forces in the neural system both at a phylogenetic and at an ontogenetic level. Sensorimotor neural mechanisms and pathways thus act as a basis for the implementation of high level cognition as well. Lakoff and Johnson's embodied realism and conceptual metaphors (1999, cf. also § 2.2.2) and Barsalou's perceptual symbols (1999, 2009), for instance, largely rely on such a conception. Beside its role in the ontogenesis of cognitive abilities and besides its being a condition for its taking place the body can also play an active role in the unfolding of online cognition, thus leading to a third conception of embodiment. Thelen and Smith (1994) brilliantly demonstrated the crucial role that the size, elasticity and weight of muscles and bones play in learning how to walk and in walking. To use an example that is more accessible to our awareness, we often use our fingers to count and remember the number of occurrences of an event, offloading the burden of working memory. If already in the second conception of embodiment the vehicles of cognition – sensorimotor circuits supporting and structuring abstract cognition – crossed sides and shapes the contents of cognition – abstract cognition involves structures from basic sensorimotor experience – the fourth conception of embodiment makes of this crossing its explicit theme. The fourth and probably most complex conception consists indeed in taking into account the role of the phenomenal experience of the lived body that can be found 44


for instance in Gallagher's (2004, cf. also Zahavi & Gallagher 2007) distinction between body schema and body image. Body schemas include the almost automatic system of processes that constantly regulates posture and movement to serve intentional action and our pre-reflective and non-objectifying body-awareness. Body images are instead conscious representations of some sensorimotor aspect of the body or, in phenomenological terms, a system of experiences, attitudes and beliefs where the object of such intentional states is one’s own body. Experiencing and cognizing through a body often constitutes and shapes the contents of cognition and experience itself. 1.2.1.4.2.

Embedding Cognition in the Context

The body, however, does not seem to be enough. The above mentioned investigations by Thelen and Smith (1994) brought these researchers beyond considering just the role of bodily structures, showing how the immersion of babies in water with the consequent difference in weight due to the different pressure deeply influenced the developing of walking movements. Indeed, the body is not static, it is a body in action, widely interacting with and integrating the (immediate) environment. This context-embedded view of cognition focuses on the way we actively use the environment to solve the problems we face: “Our problem-solving performances take shape according to some cost function or functions that, in the typical course of events, accord no special status or privilege to specific types of operations (motoric, perceptual, introspective) or modes of encoding (in the head or in the world)” (Clark 2008: 21). There is no inherent reason for the biological control system to care about differences of location or type of resource instead of simply using whatever it can to get the job done, following one or another of the most convenient cost-benefit tradeoffs available. The infant body in Thelen's and Smith's analysis takes advantage of the different density of the water and relying on this it develops an earlier ability to walk something with it would not have done outside of water. If this use of the (physical) context can be seen to a higher degree as a condition for (motor) cognition, 45


it can be shown easily how the environment can have important consequences in higher order cognition. The different piles of papers on my desk and the order of the papers in them translate strategic priorities into perceivable priorities, thus offloading part of the functions of my working and long-term memory. Constantly using a notebook to overcome memory deficits due to Alzheimer (Clark & Chalmers 1998), or, less drastically, using an abacus to count the times I have written “cognition” in this introduction or using a counter to keep track of the score of a basketball match are instances of a use of the environment that is analogous to the use of our fingers mentioned in the previous paragraph, a use of the environment that extends and integrates our biological cognitive skills. In all these cases there is the simple offloading of information storage faculties with the consequent possibility of relying on the different features and possibilities of manipulation afforded by these external components. But what happens if we consider an even more active role of the context like in the creation of the statistically-driven graphs from chapter 3? I could not have calculated and drawn and calculated them myself. It took a certain amount of time inputing data and scripts and a certain and unavoidable amount of swearing debugging the mentioned scripts. It took the encoded expertise of a couple of software programs and the calculating power of a central processing unit – and a few other hardwaremediated processes – to finally make them appear on the screen enabling me to see patterns and formulate further hypotheses to be tested. How much of that was “my” cognition? Could I even have even conceived them without such active helpers, not to mention those human beings who helped me through discussions and tutorials? How much, not only the vehicles, but also the contents of such cognitive processes, are distributed in the environment? These are the sorts of questions that constitute the ground for the third step beyond internalism, as we are on our way towards a fully distributed conception of cognition. 1.2.1.4.3.

Distributing cognition

46


The third step beyond internalism consists in the inclusion of the role of language and socio-cultural norms as well as of other cognitive agencies. Including language and socio-cultural norms in cognitive processes does not mean to assume or predict a complete internal representation of them, but the possibility to engage with them within the horizon of a community and the ability of doing so. Ruth Millikan explains this with an analogy: “I no more carry my complete cognitive systems around with me as I walk from place to place than I carry the U.S. currency system about with me when I walk with a dime in my pocket” (Millikan 1993: 170 cf. also Hutchins 1995; Latour 2005; Wilson & Clark 2009; and Fusaroli, Granelli & Paolucci (eds.) 2010). We do not need to carry the U.S. currency system with us as long as we know how to use a dime. We do not need to know all the details of the Italian law to be a lawyer as long as we can consult them, possibly through the help of a computer system. A PhD student does not need to be an experienced researcher as long as s/he can access her/his advisor for rethinking the most troublesome parts of his/her research. Such a direction brings us towards the acknowledgement of the active role of what was initially defined as the context and support of cognition and the attempt at overcoming the idea of cognition as inherently centred on the individual. One of the first and best examples of distributed cognition is still Hutchins' (1995) indepth analysis of the coordination of navigation on US navy ships around San Diego. His work showed how the whole process in its planning as well as in its unfolding is not present in its entirety in any of the minds of the single individuals; on the contrary, it emerges through the coordination and partial overlapping of knowledge and actions enacted across different individuals with different roles, competences and hierarchical positions, plus artefacts and tools in the environment. Another example is that of a constructive dialogue that, employing social linguistic competences unfolded in the specific context, engages knowledge and ideas from different individuals, giving rise to new decisions and information (Cf. Chapter 3). A third 47


example can be that of projects being developed in active research centres: far from being the effect of individual cognitive processes somehow supported by the environment, they are constituted and developed through discussions, influences, different contributions, different available methodologies and technologies. Just as embedded cognition shows us the body in action in a context, distributed cognition shows how the body and its defining actions are crossed by cultural practices and social horizons (§ 1.2.2.4.3). Cognition, initially conceived of as localised – in the brain, in the nervous system, in the body, etc. – is now to be understood as distributed and situated in a bundle of practices and emerges as a mediated process between a plurality of instances that cross and redefine the biological barriers of the individual. It is important to notice the emphasis on the terms “practices” and “process”. The shift is one from a focus on ‘‘things’’ such as stored representations to a concern with ‘‘activities’’ such as the act of representing. Such activities are often bodily, and they are often world involving in their nature. A full acknowledgement of such a distributed and activity-based perspective would seem to lead to the claim that distributed processes give rise to local meaning effects of agency and consciousness, motivated and constrained by global processes and, once established, contributing to them. That is, distributed systems come before the fully realised components, that only inside the system get their full definition. Such drift away from individual agencies is not, however, taken lightly and brought to extreme consequences. While certain authors, like Hutchins and Latour, seem more extreme, others like Adams, Rowlands and Gallagher try to reclaim the primacy of the subject. 1.2.1.5. The subject of cognition

Such attempts are based on the question of who or what a cognitive agency? While embodied and – to a certain degree – embedded cognition maintained a certain centrality of the organism, a centrality that then gets “extended”, distributed 48


cognition threatens this centrality. It is important to understand the radicality of this threat because it evades many of the critiques that have been made to embodied and embedded approaches. Following De Jaegher, Di Paolo & Gallagher (2010)'s discussion of the role of social interactions in social cognition I can hypostatise two roles that the active engagement of body, environment and socio-cultural norms can play in cognition: Given a phenomenon X, •

F is a contextual factor for X if F has an effect on X,

•

P is a constitutive process of X if P is part of the mechanisms that produce X 22.

Therefore, the heterogeneous practices we have considered as part of cognition can play different roles in the cognitive process: •

They can consist in a simple extension of pre-existing cognitive skills belonging to an organisms such the use of a notebook to memorise an address: If you are coupled to your pocket notebook in the sense that you always have it at the ready, use it a lot, trust it implicitly, and so forth, then you can infer that the pocket notebook constitutes a part of your memory store in a way that is analogous to your biological memory (Clark 2008: 46);

•

They can become part of the process that gives rise to the cognitive process itself – while using a computer to edit a PhD dissertation simply adds a contextual factor, this very cognitive task would not arise without language, writing and an active engagement with an academic system.

While often associated with a distributed cognition approach, many of the initial formulations of the extended mind hypothesis lean towards a more embedded approach. Let us just take the Parity Principle:

22

De Jaegher et al also debate a third possible role: C is an enabling condition of X if C is necessary for X to occur. While an important element, I keep it aside from this work for simplicity's sake.

49

Language and Cognition - Ph.D. Dissertation by Riccardo Fusaroli “If, as we confront some task, a part of the world functions as a process which were it done in the head, we would have no hesitation in recognising as part of the cognitive process, then that part of the world is (so we claim) part of the cognitive process”. (Clark & Chalmers 2007: 29)

While this principle certainly challenges our internalistic Cartesian intuitions and produces clear functionalist credentials, it still seems to bring in a principle of analogy. The cognitive agent is already there and its features define what we perceive as cognitive tasks and cognitive abilities. A more interesting articulation is the one that is produced by Richard Menary: The manipulation of external vehicles is importantly different from the manipulation of internal vehicles and their integration is the unit of cognitive analysis. We are not just coupling artefacts to preexisting cognitive agents; the organism becomes a cognitive agent by being coupled to the external environment (Menary 2006: 342).

If we follow Menary's perspective, the parity principle needs to be reformulated in the following way: For any process, whether it occurs inside or outside the body, if it meets the specifications that cognitive processes must meet, we should consider it cognitive without respect to its location. This brings us to the last point to discuss before comparing these conceptions to a pragmatist stance. If classical cognitivism relies on a notion of internalism that implies a strongly agentive centre – the organism or the computer23 – the progressive extension of the vehicles of cognition seems to progressively erode this centre. Hutchins, for instance, embraces a pure functionalism: as long as there is the storage and manipulation of information, there is cognition. A warship is a cognitive entity, at least at a certain level of description. Similar positions are expressed in Bruno Latour's works to the point that in his Science in Action (1987) the French sociologist 23

As an interesting twist it should be noted that defining computers and robots as cognitive devices either denies the role of orientation, value and intentionality in cognition or introduces a strongly external element: the definition of the cognitive tasks and of the aims to be achieved in computers are always defined directly or indirectly by the programmers and engineers, therefore presupposing already some intelligent agency.

50


and semiotician calls for a moratorium on cognitive explanations relying on individuals, focussing instead on externalistic networks of inter-defined roles (his actor-network theory) the nodes of which define the individuals occupying them. However, the focus on the organism is still pursued by many authors, because of its intuitive attractiveness for defining the limits of cognition. Without claiming completeness we can define two most important lines of argument that the literature in the field relies upon in one variation or another: one relying on the notion of nonderived contents and the other relying on the notion of ownership of the cognitive process. The first line of arguments comprises the argument of non-derived or intrinsic contents as the mark of the cognitive. The general idea – which is quite popular in analytical philosophy (cf. Adams & Aizawa 2008; Fodor 2009; Jacob 2010) is that inner representations have a content that is intrinsically due to the specificity of the physiological and neural system: if we look at a tree, the internal representation of it – be it symbolic or in terms of neural patterns – is not derived from the content of other representational states of a cognizing subject or from the social conventions that constitute that agent’s cultural and linguistic environment. There is something specific to perception and to the organism that constitutes the core of meaning and content. Therefore, situated and distributed conceptions have to be rejected or at least shrunk to simple “extensions” of proper cognition. A critique of the non-derived conception of perception will follow from § 1.2.2. The second line of argument – relying on ownership – seems to go back to a centre of cognition in the organism is developed amongst the followers of situated and distributed approaches: the concept of ownership. The first to introduce it into the contemporary debate is Mark Rowlands: Ownership is to be understood in terms of the appropriate sort of integration into the life – and in particular, the psychological life – of a subject. […] Ownership will prove to be derivative upon

51

Language and Cognition - Ph.D. Dissertation by Riccardo Fusaroli ownership of personal level cognitive processes: for cognitive processes, integration is ultimately integration into the conscious life of a subject (Rowlands 2009a: 17).

Not incidentally, a recent paper by Andy Clark, the herald of the extended mind hypothesis, claims that while cognitive processes are distributed, “consciousness is (still) in the head” (Clark 2009) as well as is “the machinery supporting it”. To underline the importance of this integration at a personal level in the extended mind hypothesis, I can quote Clark's phenomenological argument for the role of tools in cognition implicitly borrowed from Heidegger: “if the coupling between an agent and an external tool becomes intimate enough, the agent may no longer feel as though she is using a tool, but instead experience the (former) tool as part of herself” (Clark 2008:74). That is, the integration of agent and tool happens at a personal level, where the agent is still central. Clark is not alone in these positions. Crisafi and Gallagher (2009) and Gallagher (2010) from different angles show in a persuasive way how the presence of a legal system not only supports existing cognitive processes but also creates new processes. Since social institutions – as well as technologies like Google and the Web altogether – seem to extend far beyond the grasp of the individual and on time scales that escape the single cognitive process, Gallagher proposes to adopt a principle of Lockean ownership: “ownership is constituted by the work invested – i.e., it is the fact that I am engaged in the right way with mental institutions, and this engagement makes them a constituent part of my cognitive process. Only so far as I am engaged with these institutions (or with notebooks or pieces of technology), do they contribute to the constitution of my cognitive processes; and if I am not engaged with them (just as some neuronal processes in my brain may remain unactivated in specific circumstances) then they are not cognitively activated” (Gallagher 2010).

Just like Clark, Gallagher does not question the strong presence of a subject in his definition. The strong emphasis on the “I” clearly brings back a strong centrality of the individual. This looks like a paradox for an approach that makes of functionalism 52


its banner and that in more than one occasion (Wheeler & Cappuccio 2010) found itself at odds with enactivism for its excessive focus on the organism. Since I will come back to these issues in Chapter 3, I now limit myself to expose an additional element that enactivism brings to our discussion of the notion of cognition. The enactivist perspective (Varela 1979, 1997; Varela et al. 1991; Thompson 2007; Di Paolo 2005, 2009; Rowlands 2009b) defines cognition as the activity of regulating the interactions oriented by the perception of a valence. Cognition can thus be defined as an organism’s ability to transform contingent environmental modifications into opportunities of controlling those internal biological compensatory processes that are necessary for the well-being of that organism24. In other words, a sensemaking system is sensitive to graded differences between states, so that the organism can monitor how it is doing with regard to maintaining its viability and so it may thus regulate its behaviour accordingly in order to improve its situation. Enacting cognition means to constitute a space “of valence, of attraction and repulsion, approach or escape” (Thompson 2004:386), in a process that is dubbed “sensemaking” (Thompson & Stapleton 2009). This approach adds to the trends in cognitive sciences that we have previously considered of a crucial concern to the why of cognition. Enactivism explicitly thematises the emergence of cognition as an adaptive strategy for the preservation of the organism (in its coupling with the environment) and it is this perspective that allows the introduction of the concept of valence and of a basic semiotic conception. Articulating such an enactive perspective with a more fully distributed approach, the undeniable role of the single agency and the distributed habits and constraints that cross it, is one of the aims of this dissertation and cannot be fully tackled yet. I can, 24

Barandian and colleagues push forward the enactive approach to life and cognition in order to transform the question of the location of agency into an empirical one. “We identify three conditions that a system must meet in order to be considered as a genuine agent: a) a system must define its own individuality, b) it must be the active source of activity in its environment (interactional asymmetry) and c) it must regulate this activity in relation to certain norms (normativity)” (Barandian, Di Paolo and Rohde 2009:1).

53


however, already point out that this will require a more articulated attention to the different time scales at which socio-cultural norms and other agents participate to cognitive processes in constituting and shaping them; it will require a more articulated attention to how different resources and agencies are coordinated, not necessarily harmonically. I have all too briefly sketched the progressive and much debated deconstruction of the defining features of the initial conception of cognition. If classical cognitivism saw cognition as characterised by very specific vehicles (symbols) situated in the head, we have ended up with the idea of using trajectories in phase space as vehicles and arguing that location is not that relevant and that individual agency has to be considered within a wider network. However, in the sixty years and more of cognitive researches through neural networks, dynamical system theory, embodiment, embedded and distributed approaches something seems to have stayed constant. What has stayed constant in these approaches is the presence of a certain amount of complex and heterogeneously articulated mediation in the human – or animal – approach to the world. Even in the most radical distributed approaches the constitution of agencies or institutions with goals and internal structures as well as the active debate on who owns the cognitive processes indicates that cognition is used to achieve certain goals whatever the agency and its degree of distribution. This mediating role of cognition – whether it is called representation or not – is explicitly articulated by, for instance, Francisco Varela, Evan Thompson and Eleanor Rosch (1992), Andy Clark (1997), Susan Hurley (1998) and Lorenzo Magnani (2009). Even the brain is redefined in these terms: "the brain should not be seen as primarily a locus of inner descriptions of external states of affairs; rather, it should be seen as a locus of inner structures that act as operators upon the world via their role in determining actions” (Clark 1997:47).

54


1.2.2. A Peircean view: Mediation as a (partial) Mark of the Cognitive A similar conception was at the centre of Peirce's pragmatism and semiotics, a conception that – as we have seen it already in the introduction – a connected meaning to possible action (cf. Rosenthal 1994). Discussing this conception – in particular the concept of habit – can give us some useful conceptual tools to work on these themes conceptually and analytically. Mediation plays a crucial role – for our purposes – in Peirce's strenuous opposition to the idea of dyadic relations, of unmediated relations of action and reaction, of stimulus and response. Therefore, he would have been highly critical of any naïve behaviourism25 as were indeed many of his disciples, from William James (1900: 3) to John Dewey (1925: 198, 212-213). As mentioned in the introduction, the basic structure of any semiosis, including cognition down to perception, is a triadic relation, a mediated relation. This applies to even the most basic sensorimotor phenomena like for example the burning feeling due to an incautiously handled Italian moka espresso maker and the automatic withdrawing of our hand (cf. Eco 2007), an example that was subject to quite some debate centred on the concept of iconism (cf. Eco 1975, 1997, 2007) and that can help us to develop the positions sketched quickly in the introduction and to better understand in what “mediation” consists. 1.2.2.1. The Case of Iconism

Umberto Eco – after a long flirtation with cultural relativism concerning iconism (Eco 1968, 1975) – tackles the need for a basis of interpretation, for a beginning of semiosis: "a pure perceptual state that is free from any interpretation and that in some way precedes it" (Sini 2004: 24), through which the world “kicks the subject” 25

There are of course more interesting approaches to behaviourism, that simply imply the attempt at explaining as much as possible without recurring to complex internal cognition, in a not dissimilar fashion from the distributed , enactive and ecological approaches I am feeding my arguments with.

55


(Eco 1997) thus triggering her/his interpretations. This position is less trivial than it seems, because it engages the phaneroscopical framework composed of three categories: Firstness (the pure possibility of a quality), Secondness (the resistance of the world to the subject) and Thirdness (the generality of experience, the active dispositions). In the Introduction (§ 0.4.3) these three categories were examined, but Eco gives a different reading of them. In Eco's reading of Peirce, the way in which the world penetrates the subject is a primary iconism (Firstness, Ground), that is, an icon of sensory stimuli below the semiotic threshold (Eco 1997: 88). There is no mediation: every time we touch a boiling pot, we experience a similar pure sensory quality invading us; a disposition to retract that does not need any interpretation. It just happens, every time in a similar way (Eco 1997: 93) and its iconicity lies in this adequacy of the answer. From a pure possibility, we enter into the domain of Secondness already: “We find Secondness in occurrence, because an occurrence is something whose existence consists in our knocking up against it. A hard fact is of the same sort; that is to say, it is something which is there, and which I cannot think away, but am forced to acknowledge as an object or second beside myself, the subject or number one, and which forms material for the exercise of my will” (Peirce 1931-58 CP. 1. 358)

The pure possibility is noticed and passes into awareness as opposed to other possible perceptual qualities: the pot is boiling, not cold or just lukewarm. The third step of this perceptual process consists in formulating a perceptual judgement and thus in creating a proper percept: "I am feeling warmth, I am feeling pain" (Thirdness in Peirce's terms). This perceptual judgement is an inference based on some sort of premise, in this case previous experience. We could be holding a freezing pot instead or we could be tricked by our imagination. Still, the primary iconism cannot be fouled: in the moment we retracted, we were feeling pain. Summing up, the world imposes its qualities and from this “kick” the subject can start perceiving and interpreting. Through the quality of Firstness, the world is experienced as a resistance 56


and a force and thus cognised as a percept and a perceptual judgement. Mediation is something that happens after this kick, after Firstness and Secondness have had their – admittedly instantaneous – course. However, if we compare this reading with the presentation of phaneroscopy from the Introduction (§ 0.4.3) we find a basic misconception: Eco defines the three categories as separated moments of the perceptual process, in sharp contrast with Peirce's take that defines them as logical forms always co-present in the phenomenon. I am not the first to notice this incoherence (cf. Bonfantini 2003; Paolucci 2005; Stjernfelt 2007 and even Eco 2007) and I refer to these previous works for a more philological reading. What is more interesting is that Eco's interpretation cannot withstand a closer look, while Peirce's seems to be able to do so. Eco begins his argument dealing with Firstness and defines the perceptual process as an isolated moment sterilised from previous expectations and ongoing activities. This is not how perception happens. Perception is from the start deeply structured and mediated. Recent research on attention and perception (Noë 2004; Rensink, O’Regan & Clark 1997; O’Regan, Deubel, Clark & Rensink 2000; O’Regan & Noë 2001; O'Regan, Rensink & Clark 1999; Simons & Chabris 1999) has shown a high degree of attentional and change blindness. For subjects given a task (counting how many times a basketball is passed around in a team of basketball players) irrelevant elements tend to go unperceived, even when they are as ludicrous as a person in a gorilla costume dancing through the basketball field. Changing elements in an image while the eye, or better, the retinal fovea is focused elsewhere results in the lack of perception of this change. Leaving aside the discussion of what is consciously or unconsciously perceived and what is not, we can certainly state that the event of perception, the “percussivity” of the world on the subject, emerges only on a ground established by former practices. Even at a most basic level, the activation of neurons in V1 and in the extra-striate cortex (the cerebral areas correlated to the treatment of visual information) due to microstimulations reveals important variations in the treatment 57


of the same stimulus (Tolias et al. 2005) according to the ongoing cerebral activations as related to the activities and meaning construction of the individual. Perception always happens as a process embedded in other processes. Perception is not passive but active, both in actual movements of the perceptual systems and in anticipation. 1.2.2.2. Continuity as the Core of Mediation

In order to avoid an excessive focus on Firstness or Secondness we need to make Thirdness, that is, mediation, that is the continuity of the single cognitive act with other cognitive acts, the starting point of the analysis: Thirdness intended as the constitutive possibility of constructing a third entity that enables the organism to pass from a stimulus to a mediated reaction, from one thought to another and even to search for and construct specific stimuli instead of being a passive receptor of stimulations. This conception of continuity is one of the key concepts of pragmatism and it is worth it to spend a few more words on it. Peirce develops his model of continuity in opposition to the algebraic-geometrical model that seems to emerge in certain Aristotelian passages, that through Kant is developed by Cantor and Dedekind and that shaped a large deal of philosophical and mathematical contemporary thought (cf. Salanskis et al. 1992, Mameli 1997, Paolucci 2004, Fabbrichesi Leo & Leoni 2005). In the Cantorian model of continuity the geometrical line is composed by an infinite number of points so that no matter how close two points are individuated, a third will be found in between. What happens is thus that an infinite number of points gives rise to a continuous geometrical line. Can cognition be conceived as continuous in this sense? There are indeed several hints that point to a temporal discontinuity of the supports of perceptual and cognitive processes.

58


Figure 3 - A simplified picture of fixations in visual perception (Melcher & Colby 2008)

By way of example, the apparent continuity of the visual field and experience in human subjects is grounded on discrete sequences of saccadic fixations. There is only a small area of the retina, namely the fovea that has a high visual resolution. In order to adequately consider a wider visual field, every 250-300 milliseconds there is a saccadic movement that focuses the fovea elsewhere, a movement during which no signal is transmitted. The phenomenological continuity of our experience would thus emerge from a patchwork of local snapshots, of discrete acts that precede it. However, a second model is possible, a model that only occasionally surfaced in the history of western thought: in fragments by the Stoics, in Leibniz, in Peirce (Fabbrichesi Leo & Leoni 2005, Havenel 2008) and onwards in Thom (1988), Petitot (1995), Kelso (1995), Spivey (2006). Here the fundamental conceptualisation is not space as in the previous model but movement and relation. In movement the single instant is defined by its being crossed by trajectories, that is, by its history and by where it could lead. It is only with a subsequent act of abstraction that we can define the instant and the point. Re-using a previous example we may thus ask whether we can really define the single fixation as an isolated moment. If we consider the behavioural level we have to admit that the fixation belongs to a trajectory, a scan-path, that is, a sequence of saccadic movements, which is determined at the same time by the ongoing visual practice and by the salience structures of the environment. At the level of neural correlates the population code – the set of neurons – that is correlated with the recognition of a visual object requires almost 500 ms to reach a stable activated state. However, as we have seen, due to saccadic movements there cannot be exposure more than every 59


250-300 ms. How is recognition possible then? Recognition is possible due to preactivations and anticipations. At both levels the perception of the instant correlated to the single fixation is already informed by what has been seen before, by the objectives and by the anticipations, by what we are ready to do and cognize (cf. Noë 2004). Cognition – at the basic perceptual levels already – is not a sequence of discrete representations, easily isolated the one from another but “partially overlapping fuzzy gray areas that are drawn over time” (Spivey 2006:3), a combination of pre-activations and ongoing activities informing each other26. We can thus re-read the automatic withdrawing of our hand as already triadic and mediated, since it connects an object (the proximal stimulations) to an interpretant (withdrawing the hand) through the selection or the construction of a common ground (the tendency to avoid damage). The burning Firstness is actualised (Secondness) only because of a more general ground (Thirdness) – the practice of making a (very small) cup of coffee and of avoiding burns and the expectation of not being burned – that we withdraw the hand from the moka. Were we aware of the temperature and had we nevertheless the need to move the burning moka into the sink quickly, a difference mediation would be in place and we would be able to respond to the proximal stimulations in a different way and perform the task, although probably not entirely unharmed. Continuity - and therefore mediation thus emerges “not anymore as in Aristotle as the infinite divisibility of a substance, but as an operation, a genetic principle, the continuation of a practice […]. It is the complexity of the act of thought […] an operative-praxeologic principle” (Fabbrichesi & Leoni 2005: 37). Mediation can be thus described as concrete dynamic tendencies toward actualisations of varying types under varying types of 26

To emphasise activity is crucial to distinguish this position from the ones that consider the visual world a grand illusion, that believe that we do not experience its richness, but we only think we do: Ballard 1991; Churchland, Ramachandran, and Sejnowski 1994; Dennett 1969, 1991; O’Reagan 2010. An indirect argument against the grand illusion position is that when the participant's movements – including saccades and micro-saccades – are artificially blocked and the scene is static the participant ceases to see (Ditchburn & Ginsborg 1952) thus strengthening an active conception of perception and – crucially – perceptual experience.

60


circumstances – as a principle that through its living generality precedes and leads the single instants and states of cognition. 1.2.2.3. The Structure of Continuity

Cognition – in Peirce's eyes as well as in the eyes of many contemporary cognitive scientists – is the way through which reactions to the world are mediated and constituted in order to escape being immediately determined by the environment. The organism perceives the regularities and affordances of the organism, manipulates them, and – sometimes – representing and storing them – although in ways that we still have to fully understand (cf. Glenberg 1997, Sutton 2009). Cognition is thus a way to perceive and to manipulate conceivable consequences. However, already in the introduction I left unanswered the issue of how consequences become conceivable. If cognition is constituted by ”partially overlapping gray areas", it is always already a pre-activation of possible successive moments and in this way it defines constraints and possibilities for successive cognitions. This leaves the question open: which pre-activations are constituted and how are they selected? In Peirce the semiosic and cognitive movement between object, sign and interpretant has a tendency to stabilise: “Thought in action has for its only possible motive the attainment of thought at rest [that is belief]; and whatever does not refer to belief is no part of the thought itself. And what, then, is belief? It is the demi-cadence which closes a musical phrase in the symphony of our intellectual life. We have seen that it has just three properties: First, it is something that we are aware of; second, it appeases the irritation of doubt; and, third, it involves the establishment in our nature of a rule of action, or, say for short, a habit” (Peirce 1931-58 CP 5.396-7).

Habits can thus be defined as the structure of continuity, as dispositions to act/interpret/perceive in a similar manner in similar situations (Peirce 1931-58, CP 1.148, 1.157, 6.612) that appease a doubt, a gap in the texture of the habitude. 61


Let me give a further example of the pervasiveness of habits, of Thirdness, even at the most basic levels of perception. In his “The Modularity of Mind”, Jerry Fodor (1983) present the view that “low-level” or “basic” cognitive processes such as vision are automatic and not influenced by other cognitive processes (for a debate of this “cognitive impenetrability” cf. McCauley & Henrich 2006). The case he makes is based on the Müller-Lyer illusion: two segments of the same length that, due to opposite arrows juxtaposed at their extremities, look as though of different lengths.

Figure 4. Müller-Lyer Illusion.

The experience of seeing different lengths, argues Fodor, is direct and unavoidable no matter the fact that we know that these segments are of different length. Unfortunately for Fodor the Müller-Lyer illusion had already – long before his book came out – been subjected to one of the few rigorously controlled cross-cultural experimental projects in the history of anthropology and psychology. Building on W. H. R. Rivers’ pioneering work, Segall, Campbell & Herskovits (1966) coordinated an interdisciplinary team that systematically gathered data on the susceptibility of both children and adults from a wide range of human societies to five “standard illusions”. The researchers manipulated the length of the two lines in the Müller-Lyer illusion and estimated the magnitude of the illusion by determining the approximate point at which the two lines were perceived by the participants as being of the same length. The results were quite astonishing: 62


Figure 5 – Cultural effects on the Müller-Lyer illusion (Elaboration of Segall et al 1956 by Henrich 2008.

The horizontal axis shows the society in which the test has been performed. The vertical axis shows the “point of subjective equality” (PSE), and thus the extent to which segment “a” must be longer than segment “b” before the two segments are judged equal in length. In other words, PSE measures the strength of the illusion. The results show substantial differences among populations, with American participants from Evanston constituting the extreme end of the distribution, followed by the South African-European sample from Johannesburg. On average, the 63


Americans required that the first line be about a fifth longer than the second line before the two segments were perceived as equal. At the other end, the San foragers of the Kalahari were unaffected by the illusion and perceived the correct length of the segments. These findings suggest that visual exposure during ontogeny to factors such as the “carpentered corners” of modern environments may favour certain optical calibrations and visual habits that create and perpetuate this illusion.

Figure 6 – Müller-Lyer illusion in context

This means that the visual system adapts ontogenetically to the presence of recurrent features relevant for community-supported practices in the local visual environment. At the core of a supposedly cognitively impenetrable basic perceptual process we see the constitutive mediation of a cultural trajectory – the widespread presence of carpentered corners. The perception of two segments of the same length as being of 64


unequal length is a direct experience, which is, however, only made possible through the operational mediation of having perceived and acted in a prevalently orthogonal environment, something which is due to our architectural practices. We therefore interpret the arrows as square angles distorted by different distances and adjust the length perception. A more anecdotical example also shows how the omnipresent “carpentered” perspective shapes the human visual experience. Colin Turnbull was an anthropologist who lived amongst the the Ba'Mbuti of the Ituri forests (Uganda), a tribe whose daily environment was comparable to the one of the Zulus investigated by Segal and colleagues – a tribe showing only a moderate effect of the illusion. The Ba'Mbuti are used to forest environments, not to urban environments, and their

hunting and fishing practices call for careful aiming within few meters of distance. Figure 7 – Ba'Mbuti

During his stay with the Ba’Mbutis Turnbull had to travel for a few days through a region characterised by planes and one of his informants – Kenge – decided to go 65


with him. And then he saw the buffalo, still grazing lazily several miles away, far down below. He turned to me and said “What insects are those?”. At first I hardly understood, then I realised that in the forest vision is so limited that there is no great need to make an automatic allowance for distance when judging size. Out here in the plains Kenge was looking for the first time over apparently unending miles of unfamiliar grasslands, with not a tree worth the name to give him any basis for comparison. When I told Kenge that those insects where buffalo, he roared with laughter and told me not to tell such stupid lies. (Turnbull 1961: 217).

We see thus that acts of perception or cognition do not happen in the void; they rely on socially shared practices, they are enacted through sensorimotor skills and constraints and/or imbued with conceptual knowledge, they are primed and framed by previous local cognitions. Habits are exactly what, on the ground of wider practices, structures the possibility of these different experiences. On this more general ground of possibilities the single cognition/perception attains increased informative power (iconicity) that enables the subject to proceed further with her perception, action and interpretation (Eco 2007; Stjernfelt 2007; Fusaroli & Morgagni 2009). In brief, “immediate experience is not the experience of pure immediacy; it is shot through with the dispositional structure orderings of objectivity, for the appearances apprehended in 'immediate experience' are generated indirectly through the functioning of habit” (Rosenthal 1994: 32). In the same way we can deny the non-derived nature of perceptual contents. If a socio-cultural architectonic practice can shape the automatic perception of length, it is hard to claim that conventions and symbols do not enter perception. 1.2.2.4. What to do with the concept of habit

Habits are a tough concept in Peirce's work. Given the cosmological scope of his semiotics and phaneroscopy, habits and the related concept of Thirdness end up in certain passages as “Laws of Nature” or structures of rational thought. In other words habits end up as stable mechanisms that can be described in the form of a 66


material conditional: if p, then q. p triggers q (cf. Hoffmann 2008). However, another conception is possible and indeed explicitly pointed out by Peirce himself as it will be shown during the next pages. In particular, I am going to argue that habits – as structures of cognition – are flexible, that they are distributed (with a strong reliance on a social and intersubjective horizon) and that they are tentative, manipulative and temporal27. 1.2.2.4.1.

Habits as Flexible Tendencies

We have already seen how cognitive acts are situated in an interpretive practice, that is, in a trajectory connecting objects and interpretants, thus being open to an interplay between existing tendencies, practices at play, local contexts and phenomena. In Peirce's words: this law of habit seems to be quite radically different [...] from mechanical law, inasmuch as it would at once cease to operate if it were rigidly obeyed: since in that case all habits would at once become so fixed as to give room for no further formation of habits. In this point of view, then, growth seems to indicate a positive violation of law (Peirce 1931-58, CP 6.613)28

Taking a more cognitive perspective, this can be translated into the fact that a habit 27

28

However, Peirce was not the first to discuss the concept of habit. Even avoiding the usual references to Aristotle, I can point to Camic's reconstruction of the debate in the 18th century: “To many, the notion of habit immediately conjures up behaviour that consists in a fixed, mechanical reaction to particular stimuli and is, as such, devoid of meaning from the actor's point of view. In sociology, this image is one that became fairly widespread early in this century, though it was already current in the 1780s (see Reid 1788, pp. 114-17) and alive during the interim as well. The point to note, though, is that the image has also met with substantial opposition. In place of the idea of a fixed, mechanical reaction to stimuli, it has been held that habit creates a stable inner core that affords immunity from external sensations and impetuous appetites (Ferguson 1792, p. 225; Hegel 1821, p. 260; 1830, p. 144); that it is not by such stimuli as these, but by the ego itself, that habit is called into play and allowed to proceed, with leeway for situational adaptation (Hartmann 1939, p. 88; James 1890, p. 116; Tonnies 1909, pp. 32-33); and that, however much habitual action may be removed from "hesitation and reflection," such action is still no more "mechanical" than action of the same type that emerges from wholly reflective processes (Stewart 1792-1827, pp. 54, 55-57)” (Camic 1986:1046). Cf. also "Were the tendency to take habits replaced by an absolute requirement that the cell should discharge itself always in the same way, or according to any rigidly fixed condition whatever, all possibility of habit developing into intelligence would be cut off at the outset; the virtue of Thirdness would be absent." (Peirce 1931-58 CP 5.390).

67


does not fully determine actual behaviour, but “under different circumstances, it leaves open different possibilities of means of reaching certain goals” (Krauser 1977: 209). In other words, a habit does not consist in applying a rule or in subsuming under one concept or reaction the manifold of perception or experience. A habit consists instead in the tendency to interpret new tokens through relations that have been established between already cognised tokens through a process of flexible recognition and reconstruction of an analogy (Peirce W2 :230). There is a “living idea” that pervades and structures cognitive phenomena (Peirce 1931-58 CP 6.1512). I may thus define habits as regularities and not rules in that I put to use a dynamical systems theory metaphor, habits are attractors, that is, tendencies whose profiles depend “on the overall state of the organism involved in some activity and past basins of attractions created within the system” (Gibbs 2006: 115). Thus habits are constitutively situated, adaptable to contexts and open to learning and modifications. This is the basis of what has been called Peirce's pluralist realism, an attempt to expand realism to the growth and evolution of knowledge and organisms: A habit or disposition is a living meaning that generates acts of response in relation to criteria for grasping the situation in terms of which such activity is appropriate. These rules arise through the cumulative effect of past experience, and the creative synthesis or fixation, within the ongoing course of experience, of dispositionally organised relationships among possible experiences (Rosenthal 2000:81, cf also Rosenthal 1994 and § 1.2.2.4.4).

Therefore, the generality and the conceivable consequences of single cognitions do not have a deductive nature - that is do not follow a syllogistic rule – but they have an abductive nature, being attempts to infer and prepare for a plurality of not necessarily well defined events. This is actually a strong point of habits: their being translatable in different contexts29.

29

For an analysis of the translation of habits ethical judgment, gaming practices and didactic

68


Habits as flexible tendencies of distributed cognition

In the previous paragraphs we have compared habits to attractors, attractors being “basins of attractions created within the system”, where “the system” is not the same thing as “the organism” mentioned a few words earlier. This is a crucial element of the theory. Peirce himself is quite critical of any psychologism that focuses on individuals, and he claims that any theoretical concession in that direction is “a sop to Cerberus”. In a vein similar to that of Husserl – and, it has to be stressed, to those of many cognitive scientists (cf. Marconi 2001) – he focuses is on the general mechanisms of cognition and his definition of cognition does not pay any special regard to the organism. Cognition is a special case of semiosis and semiosis is pure function (§ 0.4.2). Therefore, cognition is pure function. In Peirce's words: To begin with the psychologists have not yet made it clear what Mind is. […] Feeling is nothing but the inward aspect of things, while mind on the contrary is essentially an external phenomenon. The error is very much like that which was so long prevalent that an electrical current moved through the metallic wire; while it is now known that that is just the only place from which it is cut off, being wholly external to the wire. Again, the psychologists undertake to locate various mental powers in the brain; and above all consider it as quite certain that the faculty of language resides in a certain lobe […]. In my opinion it is much more true that the thoughts of a living writer are in any printed copy of his book than that they are in his brain. A psychologist cuts out a lobe of my brain and then, when I find I cannot express myself, he says, 'You see your faculty of language was localized in that lobe.' No doubt it was; and so, if he had filched my inkstand, I should not have been able to continue my discussion until I had got another. Yea, the very thoughts would not come to me. So my faculty of discussion is equally localized in my inkstand. […] It is plain enough that the inkstand and the brainlobe have the same general relation to the functions of the mind. (Peirce 1931-58, CP 7.364-6. For a discussion and an extensive investigation of the echoes of this conception in Peirce's works cf. Skagestad 1999).

Indeed, “operational mediation” is not necessarily explicit internal computation, be it performed on the spot or ontogenetically developed modifications as in the case of

practices cf. Ferri & Fusaroli 2009.

69


the Müller-Lyer illusion. Wearing glasses changes our perceptual and cognitive responses and constantly writing papers with a word processor creates habits, that is, ways of thinking and acting that are constitutively computer-supported, for instance the readiness to undo an action or copy and paste, something which becomes an integral part of the habits of any skilled computer user (plus her computer), sometimes making itself explicit when the author of the present dissertation deals with printed out papers and automatically feels ready to push Cmd-c30 and paste some quotations into the open word processor on the screen. Dancing tango happens only when there is somebody dancing with us (Kimmel 2010, Kirsh 2010, § 3.1.3). Participating to a successful joint action creates stronger propensity and motivation to cooperate even on tasks that would be better performed by a single individual (Richardson et. al. 2007). Moreover, organisms tend to modify the physical and cognitive environment and thus the system they form with it. Inserting a post-it in a book will act as a focus for successive attention, making certain pages and passages more pregnant for our perceptual and cognitive process. Using certain linguistic patterns will influence the overall structure of a conversation and its affordance for ongoing practices (cf. § 3.3). Habits emerge through ecological interactions within and manipulations of a physical environment that is profoundly shaped by social and cultural structures. Habits rely on this environment and stabilise regularities of interactions as a “demi-cadence which closes a musical phrase in the symphony of our intellectual life” (Peirce 1931-58, CP 5.397), in order to tentatively make action and cognition more effective, less costly or in other ways fitter. And this often implies relying widely on the physical and social environment and on the ongoing interactions with it. This implies that a proper description of habits will have to deal with the distributed nature of cognition and behaviour described above. How much does the habit rely on the environment – be it physical, social, cultural or intersubjective? How much 30

Therefore revealing himself to be a Mac user.

70


has the habit been interiorised and how much does it still rely on wider systemic dynamics? To make this clearer let us take Vygotsky's (1962) work on what he termed “Zone of Proximal Development”. Vygotsky considered those cases in which children are temporarily able to succeed at designated tasks only by courtesy of the guidance or help provided by another human being (usually a parent or teacher). Habits have a systemic nature. When a more experienced agent talks the child, confronted by a tricky challenge, through the problem, the child can often succeed at tasks that would otherwise prove impossible for her to solve. We may all consider the time when we had to learn how to tie our shoelaces. The first couple of times the child needs to do this by herself, the task seems impossible. However, what happens is that a conspicuous number of children start repeating the verbal instructions previously provided by the adult. These children are able to tie their shoelaces. In time this process will be completely interiorised and we all – hopefully – know how to tie our shoelaces without whispering to ourselves. This example illustrates two points at the same time: i) the distributedness of habits, in this case the distributed structure that guides behaviour, focuses attention, and guards against common errors; ii) the translatability of habits – in this case from intersubjective interaction to the individual use of whispered or inner speech. 1.2.2.4.3.

Habit as a community-supported flexible tendency of distributed cognition

The social and intersubjective distribution of habits is but a subset of the general distributedness of habits I dealt with in the previous paragraph. However, these intersubjective and social dimensions will be further articulated for two reasons: i) many treatments of Peirce's crucial concepts of habits and diagrammatic reasoning as applied to cognition neglect this aspect of human habits; ii) when language enters the picture, the already present intersubjective and social horizon of cognition becomes so pervasive that it is impossible to ignore it. In a few places in this chapter I showed how different scholars argued how cognitive processes do not seem to care at all for the traditional boundary of the skull, which is 71


why the path to their being intersubjective or even social is open. Unfortunately, Peirce does not31 elaborate extensively on that. He does – anyhow – insist on the importance of the community of inquirers. The community has a cosmological importance: The real is, then, that which, sooner or later, information and reasoning would finally result in, and which is therefore independent of the vagaries of you and me. Thus, the very origin of the conception of reality shows that the conception essentially involves the notion of COMMUNITY, without definite limits, and capable of a definite increase in knowledge. (Peirce 1931-58, CP 5.311)

Moreover, the community has a normative importance in the definition of the proper fixation of belief and therefore an ideal, ethically responsible and ultimately adequate cognitive reasoning: “logic is rooted in the social principle” (Peirce 1981-, W: 284). Peirce sometimes goes further than this, moving from explicitly stating the normative and ideal nature of the social principle to naturalising it. In “The fixation of belief”, indeed, Peirce states that the method of (individual) tenacity which consists in stubbornly holding on to dogmas and previous beliefs “will be unable to hold its ground in practice. The social impulse is against it [...], an impulse too strong in man to be suppressed, without danger of destroying the human species.” (Peirce 1981-, W: 248). What happens is that stubborn certainty cannot hold its ground against a social instinct, against the opening of an intersubjectively grounded doubt and negotiation and potentially against the growth of signs (Peirce 1932-58, CP 6.613). This social instinct has two sides: i) it is already there, as “the [unavoidable] price we pay for awareness of the limits of rationality, of our inability to have immediate access to truth and reality” (Calcaterra 2001:1), as the price for the constitutively semiotic and mediated nature of cognition; ii) but it is also an ideal horizon “something on which to graft the very use and meaning, the multiple possibilities for theoretical and practical developments that accompany our figure as "artificers of knowledge" (ibid.). 31

There is, however, an explicit treatment of the dialogical nature of logic in the development of the existential graphs, more specifically in the passage between the Graphist and the Grapheus. Cf. Pietarinen 2006.

72


This social dimension in the texture itself of the habit, a social dimension which is a necessary ground and support for its existence, is a continuum in which three categories of habits can be hypostatised and articulated: i) habits as a normative and explicit layer; ii) habits as implicit principles of social normativity; iii) habits as very local principles of negotiation, coordination and synchronisation. In the first case habits are a normative and explicit layer that is socially defined can either be inherited from social institutions (from “do not carry a gun unless authorised”, to “open the door for the lady”, and so on) or locally established and explicitly negotiated (from “Do not use my milk from the common fridge” to “Do not eat the fruits on that tree”)32. In the second case habits are a principle of social normativity analogous to Bourdieu's habitus33, something which is defined as shaping individual performances without necessarily being explicitly formulated or present in the awareness of the individual. We can find empirical evidence of this, for instance in the fact that even at a basic level of motor resonance (the unconscious motor coordination between individuals) there is an important shaping effect exerted by more global factors: established social roles and the memory of previous interactions, amongst other things (for a review cp. Smith & Conrey 2009). Other examples are explicit changes in behaviour and cognitive practices due to different social contexts and difference in social and functional roles (Hutchins 1995) or the concept of affordance as expressed

32

33

A recent experiment shows how the presence of the picture of eyes over the departmental fridge at the Department of Psychology in Newcastle doubled the payments for the use of milk from that same fridge – as opposed to a control situation where a picture of flowers was displayed (Bateson et. al 2006). It thus seems that if this very simple trick had been applied to a certain tree carrying certain forbidden fruits the whole – frankly embarrassing for all sides – situation of the original sin could easily have been avoided. Social groups create “systems of durable, transposable dispositions, structured structures predisposed to function as structuring structures, that is, as principles which generate and organise practices and representation that can be objectively adapted to their outcomes without presupposing a conscious aiming at ends or an express mastery of the operations necessary in order to attain them. Objectively 'regulated' and 'regular' without being in any way the product of obedience to rules, they can be collectively orchestrated without being the product of the organising action of a conductor” (Bourdieu 1980:53)

73


by Gibson (1979: 139) that – at least in some passages – seems to be motivated exactly by this social dimension: “the real postbox [...] affords letter-mailing to a letter-writing human in a community with a postal system”. It is thus on habits as socially pre-established components that local interactions rely while potentially remodulating them. Habits can also be very local principles of negotiation, coordination and synchronisation. The cognitive role of motor resonance, joint attention and joint action is by now widely accepted (cf. Sebanz, Bekkering, Knoblich 2006). These mechanisms support the prediction of the actions of others and the natural integration of one's own behaviour with that of others, they allow building the actions of others into the structure of the common task and, as is argued by Hurley (2008), they cross the divide between the personal and the sub-personal. Such distinction can be ascribed to McDowell (1994) according to whom personal level explanations involve the constitutive character of agency and will – elements that dominate our folk theory of mind. Sub-personal level explanations are instead involved with states and mechanisms that causally underpin personal level phenomena and that can, according to our arguments, lie beyond the boundary of the individual and of the individual agency. To tie it up with our argument: synchronisation and co-ordination are not (only) the result of a conscious effort of a subject but also act at a more basic cognitive level below the habitual level of consciousness. Research shows that successful conversations involve the coordination of postural sway of participants, even in the absence of visual contact (Shockley et al. 2009) and that successful collaboration on a task shifts perception of affordances of the environment from those appropriate for individual capabilities to those matched with the capabilities of the group (Richardson et al. 2007). Whatever "operational mediation" is needed to perform the tasks, it has to be intersubjective in an important way since the participants are constantly tuned to each other and to the system that they constitute

74


together and they rely on this constant attunement in order to perform34. Habits are community-supported flexible tendencies that deeply shape our reality both because they are the inescapable mediation that allows us to access and act in the world, but also because they reshape the world in its social and even physical aspects through the creation of material artefacts (cf. the role of carpentered corners in the perception of the Müller-Lyer illusion § 1.2.2.3) and propagate behaviour and ways of thinking. Habits, as structures of a praxeologic operative continuity, define cognition in terms of its “conceivable consequences” (Peirce 1931-58 CP 5.402, CP 6.481), in terms of opening and constraining possibilities to further act, perceive and interpret: possibilities that are due to the systems to which individual cognitive systems are or have been or could be coupled. Briefly reconsidering the previous paragraphs I can thus explicitly articulate the concept of community along three different and interrelated dimensions: 1. The individual awareness of a social horizon in which the individual operates; 2. The implicit regulative effects given by repeated interactions between subjects belonging to a somehow delimited social group; 3. The vertical regulative effects due to the sedimented background of practices and more generally to cultural artefacts and norms inherited from previous generations. 1.2.2.4.4.

The tentative, temporal and manipulative nature of habits

The social and intersubjective dimension of habits should not, however, make us categorise the pragmatist stance I am building as relativist or idealist. The dimension of shared generality that constitutes habits is always unfolded, deployed and evolved 34

Given the nature of the habit and the fact that this taxonomy is in fact a continuum, I'd like to stress i) the fact that any normativity previously mentioned is somehow flexible, negotiable and social and therefore – potentially – in evolution; ii) that the conscious/unconscious divide is relative and this it is often possible to cross it – at the very least through the aid of scientific investigation as do the authors mentioned. (cf. Adamson 2005).

75


in interplay with Secondness, with the resistances of the world. Moreover, exactly because it implies a dimension of sharedness, of community, Thirdness is subjected to repeated and crossed deployments, making the role of Secondness even more prominent. Let us take our last (pseudo) perceptual example to highlight these aspects – with a smile:

Figure 8. Strip appeared on 16th May 2007 at www.wulffmorgenthaler.com © - Used by kind permission of the authors.

We can see exceptionally depicted in this strip - drawn by Wulffmorgenthaler, a Danish duo of comedians – all the dimensions of the habit that I have been discussing. The very basic percept experienced by Werner is not a simple proximal stimulation falling on his retina and then processed from scratch in his visual cortex and so on. An important part of Werner's perception is in his being ready to do something. He is employing sophisticated sensorimotor skills in eye, head and body 76


movements to focus on the blotch on the window. He is ready to hit. A readiness prompted by his holding a fly-catcher. If he had been bare handed, his possible attempts at grasping the fly would have made him more attentive to other affordances and probably to the third spatial dimension. Moreover, Werner's readiness to do is shaped by the intersubjective ongoing interaction. Werner's wife displays a not so friendly attitude, both present and, we can infer, established through time as a general style of interaction. And Werner displays a corresponding inclination towards considering her to be wrong, maybe even displaying in his willingness to hit the fly a projected desire to hit her. And yet: after 20 minutes Werner considers that his wife might be right and that what he has been trying to smash is not a fly on the window pane but a horse in the distance. His readiness to act is likely to change now. If this happens a few more times, or, more probably, if his wife keeps on reminding him of this mistake, his readiness to act upon blotches on the window pane will also change, becoming more circumspect, possibly informing his perceptual exploration of the blotch itself35. Each habit of an individual is a law, but these laws are modified so easily by the operation of selfcontrol, that it is one of the most patent of facts that ideals and thought generally have a great influence on human conduct (Peirce 1931-58 CP 1.348).

Peirce might be here far too optimistic about the easiness in changing habits. However, this emphasis can be pardoned if we consider that this self-control is for Peirce the mark of human cognition and semiosis as opposed to natural and animal semiosis36. We have thus reached an aspect of habits grounded on Peirce's realism, sometimes called pragmatic pluralism (Rosenthal 1994), sometimes evolutionary realism (Hausman 1993). We have already seen as the constitution of habits does not happen 35 36

For an investigation of how habit change can be instigated through computer games the strategies of which make the player fail repeatedly, cf. Ferri & Fusaroli (2009). We will see at the end of this chapter how language can play a role here.

77


in a solipsistic subject: the world, culture and the community are part of it, part of the mediating structure, both through their actual participation and their history of participation. Therefore, the constitution of habits relies on the world37, this being a first component of Peirce's realism. Analogously we have seen the role of Secondness in the semiosic process. The Italian moka burns, the fly-catcher ineffectively bounces against the empty window pane. The world thus acts as a resistance that does not disconfirm the habit directly. Werner still beats on the windowpane for 20 minutes. However, this resistance opens a space for doubt, a space in which habits can grow and modify themselves. We are thus facing the core of the notion of habit: the fact that they are living generalities in an active interaction and intermingling with the world. As Rosenthal (1994: 48) aptly notes: “the specificity of meaning that lies in the disposition or habit as the rule of generation includes within itself a basic indeterminateness both in relation to the total meaning and in relation to the specificity of the concrete, existing object denoted”. This indeterminateness is what makes the habit open to evolution and at the same time what allows its grounding the single act of cognition. This leads us to the last important point: habits are a complex compenetration of different temporalities. Their being grounded in previous experiences and in social practices brings to the present i) the wider temporality through which habits can stabilise. To this we can add ii) the more trivial temporality of the cognitive act, the time it takes to happen. A iii) third complex temporality can be added: the fact that this living and tentative generality projects the present further, towards a further action and modification of the cognitive trajectories – but also potentially towards the modification of the practices on which the habit is grounded: “Meaning as habit is a rule for the production of the relational generality of schematic aspects that set the conditions of recognition for what will count as verifying instances” (Rosenthal 1994:33). 37

A world and a reality, it has to be noted, that are already deeply semiotic and nevertheless “out there”, a complex of Secondness, Thirdness and Firstness (§ 0.4).

78

Language and Cognition - Ph.D. Dissertation by Riccardo Fusaroli “A habit as creatively structuring always brings something more to the organization of past experiences, though it is these past experiences that in fact nourished the habit in which they are now contained” (Rosenthal 1994:33).

The habits constitute a possibility for further conceivable action and cognition and for a reassessment of the cognitive trails. Therefore, we may add, cognition and habits have to be seen as manipulative and as actually or conceivably restructuring the cognitive environment both in its situated configuration and at the larger scale of practice sedimentation and evolution. Through the habit of organising my papers in piles, through the child's habit to repeat the verbal instructions in order to tie its own shoes and through the constitution of cognitive processes in and through the environment we modify the environment and thus the conceivable consequences we can perceive38.

1.2.3.

A Pragmatist Approach to Cognition

I could easily keep going even more in depth in the conception of habit and its evolutions through Peirce's works, through his fellow pragmatists, psychology – through behaviourism and then into clinical perspectives –, but also through more sociological approaches – starting from Durkheim and Weber to Bourdieu's habitus39. However, it is time to articulate how this pragmatist perspective contributes to the finding of a path towards a definition of an operative – although not comprehensive – notion of cognition40.

38 39

40

Relatedly Peirce specifically developed the notion of diagrammatic reasoning (cf. Stjernfelt 2007) that I will expose later in the dissertation, when the case studies make it necessary (§ 2.2.4.2). The concept of habit outside of Bourdieu's theoretical framework is surprisingly underinvestigated. For exceptions, cf. Camic 1986, but also Di Paolo’s & Barandian’s research project on the topic. However, the difficulty of dealing with the notion of cognition is notorious and little has been done to define it. Cf. Adams (2010:1-2): “The next time you are with friends in the cognitive sciences or neurosciences, ask them the following question: what makes something a cognitive process? Then stand back. You will hear a plethora of ideas that have very little in common and from a wide variety of backgrounds. You will hear anything from the view that something is cognitive if it happens in the brain (even on the retina), to the view that any kind of information processing

79


In the previous paragraphs we have seen how classical cognitivism characterised cognition as the storage and manipulation of formal symbols, possibly inside the head, and I have sketched the subsequent evolution of a quite heterogeneous landscape – through neural networks, embodiment, embedded and distributed approaches. The nature of the vehicles of cognition passed from being purely symbolic to include predictive neural patterns and non-symbolic bundles of trajectories in a phase space. The location of the vehicles and therefore the relevant constraints to these trajectories moved from being purely internal to emerging from the integration and/or active engagement of brain, body, socio-cultural norms and other cognitive agencies. The pragmatist perspective I have sketched is quite compatible with these turns. Cognition is constituted by inferences and inferences can rely on external elements like the act of writing (cf. the example of the inkstand, Peirce 1931-58 CP 7.364-6), the use of scientific instruments (on the use of cucurbits and logical machines, Peirce 1981-, W 6:69-70), the drawing and manipulation of diagrams (Peirce 1976 NEM IV: 316–19) and the enaction of socio-cultural normativity (on the horizon of the community and the social instinct, Peirce 1981-, W 3:276–90) and of intersubjectivity (on the interaction between Graphist and Grapheus in existential graphs, Peirce 1931-58 CP 4.431). The concept of mind in Peirce is the concept of a “quasi-mind” according to a purely functionalist perspective: cognition is constituted by sign functions. However, as we have seen, Peirce sketches an evolutionary realism: the habits that make possible the triadic movement of the sign function – and therefore of cognition – are unfolded and clash and grow through time according to their

constitutes cognition. I’m not suggesting that anyone will have trouble pointing out which processes are cognitive. By and large everyone will agree that inference, reasoning, perception, memory, learning, language processing, concept formation, and many aspects of emotion (and so on) are all cognitive processes. No the question to ask them is not which processes are cognitive. The question to ask is of the processes which are cognitive, what (exactly) makes them cognitive? This is the question that will really irritate, and, I’ve discovered, really interest them. It will interest them because it is a central question to the entire discipline of the cognitive sciences, and it will irritate them because it is a question that virtually no one is asking”

80


adequacy to the ongoing activity, to socio-cultural

– think of the resistance of

religiously motivated habits to physical and biological constraints and to knowledge about them. Moreover, both with the discussion of the pragmatic maxim41 (§ 0.4) and of iconicity (§ 1.2.2.1) I have highlighted how the meaning of any cognition consists – in an important way – in its conceivable consequences. Therefore, a more processual perspective – as the one presented by a truly distributed approach – will define cognition as a process through which it is possible: •

to perceive the regularities and affordances (possibilities for further action, perception and, more generally, cognition) of the physical and socio-cultural environment, affordances that are due to the systems and the activities in which the individual cognitive system is, was and could conceivably be;

•

to manipulate these regularities and affordances , either by mental simulation or by actual engagement with the environment

•

to represent and memorise these regularities and affordances in terms of modified dispositions to act, react and cognise but also in terms of dispositions to use internal and external symbols.

Cognition is thus a multifaceted mediation in which a plurality of instances crosses the individual both in its initial formation and in its ongoing activities, imbuing the single cognitive moment with its conceivable consequences and already preparing the organism to further act and cognize in order not only to preserve itself but also to further structure and modify the conceivable consequences and therefore the cognitive trajectories. All this in a way that is potentially in continuous evolution as the components of this process – biological, socio-cultural, material, etc. – change due to the actual situation, 41

“Consider what effects that might conceivably have practical bearings you conceive the objects of your conception to have. Then, your conception of those effects is the whole of your conception of the object.” (Peirce 1931-58 CP 5.2). In other words, meaning resides in the pragmatic efficacy which the object of our conception acquires during a certain process thanks to a disposition to respond, shared on a collective level and crucial in the production of sense.

81


the ongoing practice and the history of the system. In this process – as we will see – the development of semiotic systems, language in particular, catalyses the flexibilities and the possibilities of the cognitive system (Deacon 1997, Donald 2001, Thibault 2004, Tylén et al. 2010) and in a more explicit way opens the social arena of cognition where inheritance of constraints, negotiation, clashes and alignment play a crucial role. I can thus sum up this long and in certain passages complex – but hopefully not obscure – excursion into the history of cognitive sciences and into the way that a Peirce-inspired pragmatism can complement them into a few points that should be considered in order to adequately understand how language fits in cognition: 1. The different temporalities involved in the cognitive processes under analysis: a. the history of the system in its shaping of the habits that sustain the cognitive processes, b. the intrinsic temporality of the habit c. in its making us ready to act, in its projecting us into the future through the mediation of the past d. in its unfolding, e. the ongoing assessment and evolution of the habit and with it of the system. 2. The different components coupled in the cognitive process: a. the neural system b. the body c. the material environment d. the other cognitive agencies e. the socio-cultural and sedimented knowledge 82


3. The modalities and mechanisms of integration from a. the interiorisation of a component (such as language in inner speech) to b. the actual reliance on such components (such as the use of material supports for memory) Having thus defined the framework, it is time to turn our gaze specifically to language in order to sketch two of the main approaches to it – cognitive linguistics and the notion of “material symbols” in the extended mind hypothesis – and to evaluate and articulate them through the pragmatist and distributed conceptual framework just sketched.

1.3. Cognitive stances on language 1.3.1.

The embodied turn: Language in the brain/body

Cognitive linguistics originally emerged in the 1970s (Fillmore 1967; Lakoff & Thompson 1975; Rosch 1975) and arose out of dissatisfaction with formal approaches to language which were dominant in the disciplines of linguistics and philosophy. Its origins being strongly philosophically inspired, cognitive linguistics has nevertheless always been strongly influenced by theories and findings from the other cognitive sciences as they emerged during the 1960s and 1970s, particularly cognitive psychology (cf. the topic of categorisation in Fillmore 1976 and Lakoff 1987) and Gestalt psychology (cf. the definition of the perceptual bases of semantics in Talmy 2000 and Langacker 1987). As part of the cognitive sciences, cognitive linguistics considers language a psychologically real phenomenon. Anyway, as opposed to previous cognitive approaches – Chomsky's models in primis – language is not conceived as the effect of an autonomous brain module: "language is an integrated part of human cognition which operates in interaction with and on the basis of the same principles as other cognitive faculties” (Evans & Green 2006: 50). 83


According to this approach cognitive semantics "is chiefly concerned with the study of the relationship between experience, embodied cognition and language" (ibid.). Linguistic usage and interpretation is a product of cognitive systems, which is why we need to take into account cognitive domains and faculties such as bodily and mental experiences, image-schemas, perception, attention, memory, frames, categorisation, abstract thought, emotion, reasoning, inferences, etc. These positions have provoked perplexities and hard criticism, since they have been interpreted as an attempt to reduce the specificities of language in its social dimension to a conceptual – or pre-conceptual – level, which is internal to the individual (cf. Rastier 2001; Sinha 1988; Zlatev 1997 and 2010). Certainly certain statements from authors like the following may point in that direction: •

“Cognitive Linguistics focuses on mental, conceptual entities as legitimate objects of description in their own right” (Harder 2007:1247) or even “qualitative mental phenomena” (Talmy 2000);

•

"There is a unique representational level, the conceptual structure, in which linguistic, sensory and motorial information are compatible. […] For this reason to study semantics is to study cognitive psychology" (Jackendoff 1983:19);

•

“An ordered conception necessarily incorporates the sequenced occurrence of cognitive events as one facet of its neurological implementation” (Langacker 1986: 455);

•

The brain is thus the seat of explanation for cognitive linguistic results” (Dodge & Lakoff 2005: 71);

Such claims are reactions to both autonomous perspectives on language as well as to purely logical representations of cognition: However, they present a striking lack. Where is the social dimension of language? Language implies a proper or at least shared system of regularities of use (cf. Itkonen 2009) and neglecting this dimension implies the risk of a complete reduction of meaning and language either to a subjective domain or to a neurophysiological domain. But as we have already started to see in the previous paragraphs this is not the only possible articulation of language 84


and meaning in a cognitive linguistic framework. According to cognitive linguistics language is embodied in at least three ways (following on § 1.2.1.3.4): i) it is physiologically uttered (material embodiment) and ii) accompanied with gestures (the active role of the body in unfolding linguistic cognition)42; iii) the meaning it conveys is grounded in the sensorimotor system. Indeed, one of the basic and well argued for claims of cognitive linguistics is that linguistic understanding and production, thought and interaction with the world all rely on widely the same circuits (cf. Barsalou 1999, 2009). Concepts – or more precisely conceptualising systems – are based on sensorimotor activity: we keep track of interactions with the world and re-activate these traces in context both to act and to think. Language is a sequence of stimuli that orchestrate the retrieval of experiential traces of people, places, objects, events, and actions. [... L]anguage comprehension is grounded in the same knowledge and processes that are used to support comprehension and conceptualisation in many other domains. [...] Through language we re-activate action potentials, mentally simulating these actions and thus grounding our capacity to plan and execute actions and to understand the actions of others” (Zwaan & Kaschak 2009: 368, but cf. also Gallese 2007; Gallese & Goldman 1998; Glenberg & Robertson 2000).

While the focus is still on the mental and potentially inner simulation of previous experiences, the “orchestration” metaphor – resonating with the idea that semantics is about “the [non-reductive] relationship between experience, embodied cognition and language" – seems to lead to a more articulated conception of linguistic mediation not just as a trigger, but as a creative element in the simulation. Moreover, the article goes on: “The experiential traces reflect the comprehenders’ past experience with particular objects, actions, and events, as well as their previous experience with language” (Zwaan & Kaschak 2009: 368). Language creeps into the explanation of language. This paves the way for the idea that once language is in 42

On the cognitive role of gesture in language: Clark 2007, Goldin-Meadow 2005, McNeill 2005.

85


place, the picture changes and it is impossible to reduce it to anything else. But what is the specificity of language?

1.3.2. The extended turn: language (and mind43) into the world In mainstream cognitive linguistics and in most embodied strands of cognitive sciences (cf. Gibbs 2006), language seems to be – methodologically at least – an internal matter, that is, it has to be explained via processes that happen in the head and that were in place before language was. This is exactly the view that is opposed by the extended mind hypothesis (Clark & Chalmers 1998, Clark 2008). The extended mind hypothesis is based on a functionalist perspective supported by a strong evolutionary claim. The functionalist perspective – a principled indifference to the details of the physical structures that realise mental processes, that is, what is crucial to a mental state or process is its functional role, not its physical realisation – is effectively expressed through the Hypothesis of Cognitive Impartiality (HCI) – that we have seen already at work when defining the context-embeddedness of cognition: “Our problem-solving performances take shape according to some cost function or functions that, in the typical course of events, accord no special status or privilege to specific types of operations (motoric, perceptual, introspective) or modes of encoding (in the head or in the world) [...] It states that the biological control system does not care about differences of location or type of resource but simply uses whatever it can relative to some cost-benefit trade-off, to get the job done. (Clark 2008: 2122)

43

The “extended mind” - it has to be noted – is a theory that deals almost exclusively with cognition – where cognition is intended as some sort of information processing based tasks led (an aspect often left implicit) by some kind of conscious or unconscious human or organism-based intentionality – and it hardly accounts for consciousness and the phenomenological aspects of experience. An ironic stance since the extended mind hypothesis is – at the same time – partly grounded on a phenomenological principle, the Parity principle: “If, as we confront some task, a part of the world functions as a process which, were it to go on in the head, we would have no hesitation in accepting as part of the cognitive process, then that part of the world is (for that time) part of the cognitive process” (Chalmers and Clark 1998, my italics). Cf. Steffensen 2009 and Paolucci 2010.

86


This stance is backed by an implicit evolutionary claim: given that both evolution and development lead to optimised organisms and, in this case, cognizers, the most efficient way of cognising will be implemented44. If the external environment can be co-opted in a way that results in less cognitive effort, then it will be co-opted45. Thus, while a traditional perspective on cognition can be defined as brain-bound – “the model of mind as brain (or perhaps brain plus nervous system)” (ibid: xxvii) – the extended perspective claims that when parts of the environment are coupled to a cognitive system in the right way, they become parts of the mind. Cognition is not necessarily extended but it may and often does depend directly or constitutively on non-neural structures, including - and extending beyond – the body. The actual local operations that realize certain forms of human cognizing include inextricable tangles of feedback, feed-forward, and feed-around loops: loops that promiscuously criss-cross the boundaries of brain, body and world. The local mechanisms of mind, if this is correct, are not at all in the head. Cognition leaks out into body and world (ibid.: xxviii)

This formulation implies and expresses that “the thesis of the extended mind is an ontic thesis, of partial and contingent composition of some mental processes” (Rowlands 2009a). Cognition, as a matter of fact, in certain cases, but not in all of 44

45

This is a “given” that pervades so much of the extended mind literature, but which should not be taken for granted, if we consider the intricate paths and interactions of evolution, exaptation and culture, cf. Dennett 1996 and Gould 1997 for a discussion on these topics. Optimisation in evolution is always a compromise between different constraints at very different levels and it is not teleological. The engineering metaphor of optimisation as the organising principle of cognition is often backed up by and blended with capitalist and industrial metaphors: “From the perspective of HEC, the ancient biological skinbag is the handy container of persisting recruitment processes and of a batch of core data, information, and body-involving skills: Thus equipped, the mobile human organism is revealed as a kind of walking BIOS, ever ready to bootstrap into existence the larger softassembled cognitive systems that are quite literally, the information processing engines of much advanced thought and reason” (Clark 2008: 38); “It is the brain's great plasticity and thirst for cheap, outsourced labor that drives the distributed engines of socio-technological adaptation and change. It is true, too, that by subtracting those meaty islands of wet organismic plasticity, the whole process grinds to a standstill” (ibid.: 62); “[we are] factory tweaked and primed” (ibid. 66); etc. Cf. Steffensen 2009.

87


them, can happen through a coupling of the subject and the environment. The conditions for this coupling are four: i) the elements involved have to be constantly available; ii) their involvement must not encounter difficulties; iii) the elements involved must be automatically endorsed, possibly due to iv) a conscious endorsement happened in the past46. In this picture, language – or as Clark says, material symbols – can easily become an element coupled to the cognitive system: i) human beings can generally talk and think linguistically after they have learned a language; ii) they do so without excessive difficulties once they have mastered it; iii) the use of language is perceived as part of one's mental toolbox47. More than that, in Clark's words language is the “ultimate artefact” (Clark 1997:218; Wheeler 2004): "a form of mind-transforming cognitive scaffolding: a persisting, though never stationary, symbolic edifice whose critical role in promoting thought and reason remains surprisingly ill understood" (Clark 2008:4; cf. also Clowes & Morse 2005). Non- or pre-linguistic cognition is non-arbitrary, modality-rich and context-sensitive: a deeply embodied and deeply situated activity. The use of language allows human beings to circumvent certain limits of this system, granting cognition “a public system of essentially context-free, arbitrary symbols that push, pull, tweak, cajole and eventually cooperate with the biological skills” (Clark 2006: 47). Language is thus “a self-constructed cognitive niche” (Clark 2006:370) consisting of structures that combine with appropriate culturally transmitted practices to enhance problem-solving. Thanks to its relative independence from the context, language can be i) a source of additional targets for attention and learning (perceptually simple tokens that reify 46

47

Clark, Chalmers and the other debaters do not elaborate too much on these points. We could reread them on the terms that a habit has to be created, that is, an automatic disposition of the system that involves environmental components. A Heideggerian approach would use the “readyto-hand” category. For an experimental paradigm to test – in limited circumstances – this readiness to hand: Dotov et al. 2010. It thus becomes interesting to analyse in this framework cases where this endorsement and this readiness to hand is still in the making or breaks down: the performances of language learners and of patients with brain damages or people under the effect of adequately applied TMS and so on.

88


complex ideas); ii) a resource for directing and maintaining attention on complex conjoined cues; iii) together with more basic representational resources, the basis for new forms of hybrid thoughts48. A very important point in this re-definition of the role of language is that at the same time: i) Language creates a mediational layer that can select and renegotiate aspects of the immediate content as well as define a completely different context, thus enriching reality and cognition. ii) Language is constitutively imbricated in practice, in doing something, in coordinating something else and not just a representation or a transfer of information49.

1.4. A habit-based stance on Language in Cognition and Cognition in Language Comparing the previous two paragraphs we might get the impression that we are faced by two opposite stances on language: •

On the one hand language is grounded on and expresses non-linguistic and embodied cognition that pre-existed it, and linguistic meaning seems to be a – potentially internal – embodied experiential resonance, an inner content only triggered by language.

•

On the other hand language is an abstract external device that is used exactly because it does well what modality-rich cognition could not do, and linguistic meaning seems to be external, defined by (social) convention and grounded in the materiality of the symbol.

48

49

Bermudez (2003:151) recasts this into six points: “1) memory augmentation, 2) environmental simplification, 3) coordination of activities through control of attention and resource allocation, 4) the activity of transcending path-dependent learning (the learning of linguistic organisms is not constrained by complicated cognitive paths that are circumvented thanks to language), 5) control loops that act for our future behaviour: for example writing plans difficult to keep in one’s head), 6) data manipulation and representation”. We are reminded here of the pragmatist focus on the “conceivable consequences” that define the meaning of any entity or relation.

89


1.4.1.

Embodiment as ecological enaction

A closer look, however, seems to reveal the possibility of a reciprocal articulation of the two perspectives. Cognitive linguistics opens the possibility of a space of manoeuvre. The fact that language “orchestrates” and “informs” the experiential resonance may be read as follows: language has an active role in the shaping of experience and grounds the way it is intersubjectively constructed50 (§ 2.4.4 and § 3.3.3). In the same way, the extended mind hypothesis does not negate the more embodied and modality-rich dimensions of language and meaning, indeed claiming that it is through language that human beings are able to construct “new forms of hybrid thoughts”, which are still grounded in human embodied experience, but along dimensions that are more socially and linguistically articulated and that potentially reconfigure and articulate the grounding itself. If cognitive linguistics mainly focuses on two dimensions of embodiment – the sensomotricity of utterance and perception and the sensorimotor foundation of concepts (second and third kind of embodiment according to 1.2.4.1 – there is, indeed, at least a third additional dimension that is often neglected and fits a more distributed approach. The body is indeed not just a physical body, with its size and weight, articulations and biological structures. The body is an active presence in the world: it shapes and at the same time it is defined through our activities. When acquiring new skills, for example, we may begin by paying close attention to certain rules of performance, and when doing so we typically focus on and monitor our own bodily performance to an unusually high degree. But a successful acquisition of a new ability will lead to the ability to perform it without explicit monitoring of bodily movement; the skill becomes fully embodied and embedded within the proper context. Practice makes perfect because it habitualizes the skill. That which is practised in the past becomes embedded in my

50

This is not to say that every single author in cognitive linguistics share this view or that it is not locally contradicted in certain theoretical statements. I am pointing out a theoretical possibility – in more than a few cases tackled via analysis and experiments – that has to be further articulated, but it is not as such incompatible with the cognitive linguistic enterprise, quite the contrary.

90


present bodily repertoire and allows me to cope differently with new arising situations. As Merleau-Ponty writes: “The blind man’s stick has ceased to be an object for him, and is no longer perceived for itself; its point has become an area of sensitivity, extending the scope and active radius of touch, and providing a parallel to sight” (Merleau-Ponty 1962:143), a perspective quite similar to the enactive approaches (§ 1.2.1.4 and § 3.1.4): We propose as a name enactive to emphasise the growing conviction that cognition is not the representation of a pre-given world by a pre-given mind but is rather the enactment of a world and a mind on the basis of a history of the variety of actions that a being in the world performs. (Varela et al. 1991:9)

This notion of enaction, though, will have to fully integrate a social dimension (§ 1.4 and § 3.1.4.3). I have already extensively argued for the crucial importance of the intersubjective systemic dimension in the foundations of our cognitive system – and more arguments will be brought forth in the next chapters. The body is defined and shapes human intersubjective coordination through posture, facial expression and consciously or unconsciously performed gestures. Language is thus bodily grounded because it dwells on and takes further these active and intersubjective dimensions of the body, forming an articulated set of tools that allows us to act and operate in an individual (self directed internal language) or in a social context, extending the human possibilities of acting in and manipulating the physical and social world51. Exactly this kind of embodiment defines a third way for meaning to emerge, a way that operates through crossing the divide between internal and intrinsic contents on one side and explicit social conventions on the other52. Sense is constituted by the conceivable consequences of a linguistic structure, consequences that are certainly grounded on our physically and socially embodied experience but which also evolve 51 52

For a further review of how the body is not simply a physical device but a social and semiotic entity: Ziemke, Zlatev & Frank eds. (2007) and Frank, Dirven, Ziemke & Bernárdez Eds (2007). Sambre (2010) analyses Merleau-Ponty's conception(s) of body and language along similar lines, reaching the conclusion that in this author linguistic use is the fullest expression of the body,

91


– and are crucially co-constituted – through further interactions with the world, including individual as well as intersubjective and social reflection on and manipulation of language.

1.4.2.

Experience – Materiality – Intersubjectivity

Let us try to see how this perspective develops from our framework. We have seen cognition as structured by habits, as an operational mediation and a triadic structure that enables us to open new spaces for possible actions that are not unmediated responses and that allows organisms to actively search for stimuli and to manipulate the environment. This structure is not a priori confined to the boundary of the individual, but it forces upon us a systemic perspective in crossing the inner/outer dichotomy and in redefining it as at most a post-hoc boundary. We have seen, for instance, that individual actions and perception of affordances might be grounded on intersubjective interaction like motor resonance and joint attention, and that a socionormative dimension plays a role in these processes as well. In this conceptual framework, the articulation between cognitive linguistics and extended mind hypothesis finds a natural context: language becomes an additional extremely ductile layer through which systemic interactions are constituted, a layer that feeds back onto the other layers and re-shapes cognitive systems. Habits can find new arenas of deployment, evolve and new habits can emerge. This may be represented roughly in the following figure:

92


Figure 9 - Dynamics of language and cognition53.

Three interrelated factors can thus be (re)spelled out to better understand the role of language in cognition: •

Experience: experience includes both crucial embodied and non-linguistic dimensions, pointed out by cognitive linguistics, and intersubjective, sociocultural and often already linguistic dimensions.

•

Materiality: as pointed out in the extended mind hypothesis the sheer materiality of language creates a whole new horizon of manipulability – affordances – that was not present – or only to a lesser degree – in nonlinguistic cognition54.

•

Intersubjective and social horizon: materiality opens up to a public horizon that widens and qualitatively changes the intersubjective grounding of cognition

53 54

The terms in {} refer to the correspondent elements in Dynamical System Theory (Kelso 1995) that – used as a metaphor – inspired the drawing of this diagram. As it is implied in the notion of affordance (cf. Niveleau 2006, Morgagni 2010) materiality is conceived as embedded in socio-cognitive dynamics, in Peircean terms as Secondness (the resistances of the world) emerging on the ground of Thirdness (the expectations and general ways cognitive systems approach it with).

93


that we have already seen at play. At the same time, it is through this dynamical social horizon that the materiality of language makes sense and cognition is reshaped. To introduce a social horizon does not mean to open up to a radical culturalism according to which the sign system of a society or a culture constrains, defines and determines the individual into the finest detail of his actions. On the contrary, as we have seen in the previous paragraphs, the social horizon is always put to play, locally engaged and potentially verified55.

1.4.3.

Language as a tool

If a hypothetical initial use of language might have been a mere prothesis and immediate expression of the active body56, it becomes language proper when it functions as the vehicle and support of shared habits structuring cognitive interactions. The presence of words and other linguistic structures as public signs creates an anchor for the co-construction and attunement of cognitive systems – even at a spatial and temporal distance – and a shared external support that these systems can rely upon intersubjectively and which they can both explicitly negotiate and implicitly align. A non-exhaustive list of functions that language fulfils “as a tool for interacting minds” is provided by Tylén and colleagues (2010) in order to: i) extend the possibility field for interaction in space and time, ii) facilitate the structuring, profiling and navigation of joint attentional scenes, iii) enable the alignment and sharing of situation models and action plans and in order to iv) mediate the cultural shaping of interacting minds. In this sense, language is a tool or – not to forget Wittgenstein's (1953) lesson – a set of different tools. 55

56

This does not mean that there cannot be quite radical social influences as in fundamentalist societies that are not verified at least not by sensible standards. Still these influences evolve in time can rarely avoid enter a more complex arena of interactions with other discursive and not practices. Cf. for instance Vico's narration of the origins of language or Deacon's (1997) and Donald's (2001) more scientifically fed genealogy of language. Cf. Also the idea of first order languaging in Love 2007.

94


1.4.4. Language as an environment and as normative instances At the same time we should not forget that language pre-exists and transcends both the single use and the single speakers. Language is culturally passed from one generation to the next and is thus learnt again and again just through exposure to a sample of it and then suitably generalised and developed through further interactions with the linguistic and non-linguistic environment. Language is an important part of the cognitive niche built by human beings a long time ago and it forms a permanent and evolving artificial environment for the development and the unfolding of our cognitive abilities. The sedimented constraints to cognitive interactions constitute the environment in which the speaker develops his cognitive skills and on which she relies to deploy them. The distributed deposit of articulated linguistic uses and idioms – what de Saussure hypostatised in the notion of langue – has to be both at least in part interiorised and continuously coped with by the speaker in order for her to express herself and for her linguistic habits to consolidate and evolve within their social horizon. This is exactly what Peirce meant when he said that signs and men reciprocally educate each other. Language precedes the speakers as a bundle of predisposed linguistic habits and linguistically shaped non-linguistic habits that inform while not completely determining the expressive action of the individual situated in a specific community. Thus I can point out – again – the double social nature of language: at the same time it dwells in the basic skills of intersubjective tuning57 and extends the range of intersubjectivity widely in space, time and pervasivity. Linguistically socialised subjects do not simply engage with their own thoughts, but interiorise and dwell on this public dimension – which means, for instance, that they are able to interact with absent controlling agencies. The linguistic individual is thus subjected and defined by a plurality of more or less anonymous “thirds”, by a pluralistic dimension of flexible 57

Cf. Tomasello 1999 for a discussion of how symbolic activity might need joint attentional practices in order to develop.

95


and negotiated normativity. On the one hand the individual monitors and adjusts these “thirds” as she sees fit58, on the other they constitute the network of norms and values of the social reality through which the individual is formed and informed59. In this framework it becomes easy to point out some of the potential shortcomings of the previously mentioned approaches when cognitive linguistics and an extended perspective on cognition are not reciprocally articulated. Experience is deeply shaped by an intersubjective and socio-cultural horizon that often relies on language. Therefore language cannot be a mere expression of non-linguistic schemas; it engages them in a more ductile arena, it aligns and negotiates them intersubjectively and socially and it contributes to the conditions of their formation. Two examples come to mind, already within the field of cognitive linguistics. Jordan Zlatev (2005) showed the influence of mimesis and Chris Sinha and Kristina Jensen de López (2000) the role of material artefacts and socio-cultural practices in the formation of one of the most basic structures of cognition: image schemas. Image schemas are specific and recurring action paths formed through time by – usually early – embodied experience that act as regularities to orient future experiences (§ 2.2.2.1) and they are supposed to constitute the ground for understanding language through pre-linguistic conceptual and pre-conceptual structures. According to Sinha and Jensen de Lopez (2000: 31) children employ social knowledge of the canonical use of objects in conjunction with their innate capacity for schematising spatial relations. The image-schematic nature of cultural objects may be a prototypical ecological affordance that influences language (Sinha & Jensen de Lopez 2000: 22). Thus, Zapotec children are not as quick as Danish or English children to notice linguistic

58 59

It is enough to point out here the manifold of excuses and justifications that permeates our mental life especially when it comes down to the procrastination of our duties. In an initial footnote I stated my interest for cognition, leaving mind and phenomenology out of the scope of this article. Nevertheless, they become relevant here, because it is exactly in this interplay of individual and social normativities that the problem of meaning is articulated. In a linguistic community we are ready to do something together according to a shared understanding of what we are doing, what we should be doing, what we are expected to do. It is this retroflected aspect of meaning that allows it to “make sense”.

96


differences between senses of “under” and “in” because they are not encouraged to play with upright cups and more generally because the Zapotecs use a smaller variety of containers while they tend to use them more multi-functionally. Some other very persuasive examples are: i) the way different languages influence the way in which spatial relations are represented and the time line is conceptualised (Boroditsky 2001, but cf. Chen 2007), ii) as well as the different performance in absolute and relative spatial orientation tasks by native speakers of Guugu Yimithirr and English speakers that privilege the one or the other system (Levinson 1997); iii) and the evidences of the role of language in constituting a fully developed sense of body (Borghi and Cimatti 2010a). Intersubjectivity, language and culture change the environment in which children develop, therefore having a shaping influence even their most basic physical interactions and cognitive processes.

1.4.5. The flexibility of language: semantic forms and use potentials At the same time, exactly because cognitive processes are grounded on habits community-supported flexible and potentially evolving tendencies – the linguistic use of material symbols does not follow once and for all established conventions, as certain passages of Clark's articles could lead to think. The meaning of linguistic structures (be it semantic configurations, morphemes, words, idioms, etc.) is flexible and situated. Let us follow the Dynamical Systems Theory metaphor (§ 1.2.1.3.2), that is, verbal patterns can be imagined as attractors for the experiential traces connected to it (cf. Zwaan & Kaschak 2009: 368). An attractor is a trajectory in phase space towards which all of the trajectories of a non-linear dynamic system are attracted. The meaning of the word [as an attractor] being uttered does not belong simply to the individual but to the community to which the individual belongs [. . . ] and emerges in the context in which it is being used (Logan 2006: 153).

The variability of the context explains that “the attractor is a strange attractor 97


because the meaning of a word never exactly repeats itself” for instance because of the variability of the constraints imposed by the medium, the context and the practices at hand. These kinds of dynamics in linguistic and non-linguistic meaning have been increasingly investigated: Eco's Encyclopædia (1975); Violi's contextual instructions (1997); Rastier's interpretive semantics (2001, 2006); Visetti and Cadiot's theory of semantic forms (2001, 2006); Zlatev's use potential (1997); the model of meaning potential (Allwood 2003) are examples hereof. In all these cases, the meaning of a situated linguistic structure is defined by an ongoing interaction between sedimented usages (a more normative but pluralistic dimension), co-textual and contextual constraints and expectations. The whole idea behind interpretive semantics – for instance – is to produce a model that complements verbal form stability with the possibility of stability in meaning configurations. Therefore, interpretive semantics studies how the forms come into being: for instance by dissimilation (same verbal pattern, different meaning: “there is music and music!”) or assimilation (different verbal patterns, similar semes: “women, fire and dangerous things”). Another important point is the study of how semantic forms propagate (cf. how in “Tristesse de lune” by Baudelaire a form composed by “rondeur”, “mollesse”, “clarté” keeps propagating through terms like “lune”, “coussins”, “contour de seins”, “dos satiné”, etc.). The idea is to capture how the situatedness of language, together with the relevant norms (genre-related for instance) define a ground on which more or less stable semantic patterns are created and negotiated. Thus we see the local interaction of different time scales at which meaning is sedimented. Meaning, in this view, must be conceived of as a constraint on the type of local interaction that is the result of dynamics between the community-supported affordances of the verbal patterns and the local affordances constructed in situation. Chapter 2 and 3 will be dedicated to the further articulation of these points through the analyses of the temporal evolution of “Berlusconi is a caiman” on the pages of an Italian newspaper and of the local alignment of confidence descriptors in an 98


experiment of joint perceptual decision. However, we can tackle a quick example. Under the Third Reich German shops displayed on their windows a sign saying “Hunden und Juden ist der Eintritt verboten”: “the entrance is forbidden to dogs and Jews”. In this terrible sentence the meaning of the word “Juden” is not given simply by the single lexeme. The unfolding of the sentence in the specific shared cultural ground constructs a local frame of negativity through the forbidding, and through the lack of hygiene implied in the notion of dogs in a shop. Through the conjunction “und” that homologates Jews and dogs these negative traits are projected, provoking a certain moral judgment but also implying a certain physical disgust60. In this process we can see the importance of sedimented conceptual schemes – the concept of dogs, the widespread conceptual metaphor PEOPLE ARE ANIMALS (§ 2.2.2.3) – strengthening the overlapping of Jews and dogs but sedimented dispositions and competences are put to play and modified in the context in interaction with the syntactical structures and the semantic co-text in order to determine the exact meaning of its components. The mechanisms at work in the sentence and the meaningful structures they configured, sufficiently repeated and stabilised contributed to modify the conceptual formations of that culture and the cognitive dispositions of the population. The flexibility and evolution of meaning potentials, grounded on the flexibility and tentativity of habits, is due to the use of language for situated and intersubjective coordination.

1.5. Conclusions Applying a habit-based framework to the articulation between language and cognition, I can thus sketch some conclusions concerning the role of language in a distributed and pragmatist framework. What makes language so difficult to deal with in this context is that it crosses the external/internal divide in a plurality of ways. While being motivated by embodied experiences and produced/interpreted through

60

Two aspects that are strongly related as experimentally demonstrated in Schnall et al. 2008.

99


cognitive mechanisms largely overlapping with the sensorimotor systems, its being “public” and “external” articulates our most basic and intuitively internal processes to both the continuously evolving complexity and the stability of the social arena. I should like to restate that this does not mean I am denying the role of the prelinguistic and sensorimotor but to articulate and complement it with how it is put to play and partially re-shaped in and through the public arena. On the one hand the public nature of language creates structures that can act as additional foci of attention and – presenting features of abstraction, stability and manipulability – complement non-linguistic cognitive processes, thus widening the experiential field. Remember the functions of language according to Clark, Bermudez and Tylén, which can be summed up as follows: to augment memory, to simplify the environment, to coordinate activities through control of attention and resource allocation, to establish control loops and to represent and manipulate data. All this happens in a potentially public, therefore intersubjective and social way. Alignment and coordination become increasingly possible both in the situation and outside of it – which means when the subjects are not co-present and/or when the situation is past or future or just imagined; socio-cultural are established and constitute a resource and an environment for further cognitive activity. On the other hand this materiality and abstraction is not static and ideal, it is defined through practices within a social horizon and thus it is subjected to evolution and negotiation at the different time scales at which language acts in local intersubjective communication, in the development of idio- and sociolects, in general diachronic linguistic change and in the development of civilisation and science. Language is a tool whose use in many ways transcends and defines the individual within a social and intersubjective horizon that partakes in an historical evolution on different interacting time scales. Language is thus not a simple conduit, it is a crucial part of cognition, catalytically expanding its social and historical dimension and opening up at a wider scale even the most intuitively private cognitive processes to a more 100


flexible, adaptable and sometimes economical and powerful systemic dimension. A manifold of conceivable consequences can be drawn from these positions, including the following: the need to focus on the role played by linguistic structures in their co-text and in the practices through which they are employed, aligned and negotiated and the need to explore the resonance, negotiation and diffusion of linguistic structures as well as the cognitive consequences of this hereof. This is exactly what I am going to investigate in the next two chapters.

101


2. The Life of Caimans in Society and what it can tell us about the roles of metaphor, language and thought “a metaphor is an affair between a predicate with a past and an object that yields while protesting” (Goodman 1976:69)

2.1. Introduction: Of the life of metaphors in society Once sketched a pragmatist framework for the analysis of language as a crucial part of (human) cognition, it is time to see this framework at work when applied to one of the cornerstones of cognitive linguistics: the study of metaphors through Conceptual Metaphor Theory (CMT) and Conceptual Integration Theory. Can our framework fit the already existing results? Can it enrich our understanding of metaphors in a cognitive perspective by adding new observations and/or by re-assessing previous positions and results? I will dedicate a first part of this chapter to sketch the basic tenets and the methodological applications of the approaches at stake in dealing with metaphors: cognitive linguistics with Conceptual Metaphor Theory (CMT) and Conceptual Integration Theory (CIT) as well as extended mind hypothesis with its notion of material symbols61. This analysis of CMT's and CIT's theoretical and analytical positions will be complemented in the second part of the chapter by the analysis of the linguistic “surface” of the life of a metaphor using the peculiar case of the man who (metaphorically) became a caiman. In 2002 Franco Cordero – an editorialist for La Repubblica, one of the main Italian newspapers – described Berlusconi as having the smiles of a caiman. That colourful but minor image in his article slowly got a life 61

It has to be noted that to our knowledge nobody has used the concept of material symbols to deal with metaphorical phenomena up to now.

102


of its own and generated a whole family of related metaphors that may be provided with the more generic label “Berlusconi is a caiman”. These metaphors were being repeated and used in different variations at least 600 times in the pages of La Repubblica alone and, in addition, countless times more in the Italian mediascape62. Following one of the most traditional semiotic goals – “to study the life of signs in society” (de Saussure 1972:33) –, we will see how cognitive theories of metaphors, material symbols and pragmatism fare in the study of “the life of caimans in society”. In particular, I will focus on finding out how a man can be called a caiman and how the interpretation of such a metaphor can be constructed and constrained, on the role of pre-existing conceptual structures enabling this interpretation and on the specific role of language in the unfolding of such a metaphor and on its meaning and its temporal evolution. Moreover, the quick spreading and evolution of the “Berlusconi is a caiman” metaphorical pattern through the pages of La Repubblica endows us with a window on a “road less travelled” in cognitive linguistics (cf. Rastier 1999, Allan 2009), extended mind and even semiotics63 – the diachronic evolution of a metaphor. I will thus bring to the foreground the social and intersubjective dynamic horizon that was so central in chapter one. If this second part already presents some important aspects of the linguistic, social and temporal dimensions of metaphors, the third and final part of the chapter will try more radically to propose a pragmatist re-assessment of the theories presented, in particular CMT and material symbols, in that it combines the results of the analysis with an expanded exposition of the pragmatist concepts of icon, symbol, habit and 62

63

Mediascape, in Appadurai's words, "refers both to the distribution of electronic media capabilities to produce and spread information (newspapers, magazines, tv stations, movie production companies that are now available to an increasing number of public and private interests throughout the world, and to the images of the world created by these media […] what is most important for these mediascapes is that they grant their audiences large complex repertoires of visual images and narratives […] in which the worlds of goods, news and politics are blended" (Appadurai 1996). In most structuralist developments of semiotics, a much needed – emphasis on the textual dimension of meaning has resulted in an almost complete neglect of the temporal dimensions of meaning that was much present in thinkers like C. S. Peirce and J. Lotman. Temporal dimensions that it is high time to re-integrate in semiotic methodologies and conceptual frameworks.

103


diagrammatic reasoning. How embodied is experience? How do established conceptual metaphors situate themselves in our habit-based framework? Which kind of cognitive resource are they? How are they deployed? Which role does the distributed and social nature of habits play in their formation and linguistic deployment? Which role do verbal patterns and social use assume in an ecological and distributed cognitive approach? Discussing cognitive approaches to metaphors, Laurence Kirmayer (1992) stated that “metaphors are tools for working with experience”, not mere expressions of it, and in 1999 Raymond Gibbs followed up with the claim that it was high time to: […] move metaphor out of our heads and put it into the embodied and public world […] (This move) does not make metaphor any less cognitive than if we had long lists of metaphors nicely encoded in our heads. All this move attempts to do is acknowledge the culturally embodied nature of what is cognitive and to suggest that there is much less of a difference between what is cognitive and what is cultural than perhaps many of us have been traditionally led to believe. (Gibbs 1999:162).

Acknowledging – as does cognitive linguistics – that metaphors are “a species of perceptually guided adaptive action in a particular cultural situation” (Gibbs 1999: 162)64, their social linguistic expression, alignment and negotiation becomes a crucial mechanism that social communities have in order to share, align, negotiate and guide their experience. Thus, on the level where cognitive and social dynamics are inextricably interrelated I think that the pragmatist framework can play its role at best.

2.1.1. 64

Conceptualising Metaphors and Methodological

To be faithful to the original we need to add the whole quotation: “Metaphor is as much a species of perceptually guided adaptive action in a particular cultural situation as it is a specific language device or some internally represented structure in the mind[s] of individuals” (Gibbs 1999: 162). The reference to internally represented structures has, however, to be re-interpreted in the light of the pragmatist perspective of Chapter 1. The “inner” representation has to be re-conceptualised as simply a sign in a semio-cognitive process for which the inner-outer divide is only relative and a post-hoc effect. Positions that Gibbs is getting closer and closer to (cf. Gibbs & Tendahl 2006; Gibbs & Cameron 2008).

104


Consequences The Encyclopaedia Britannica defines a metaphor as a “figure of speech that implies comparison between two unlike entities, as distinguished from simile, an explicit comparison signalled by the words ‘like’ or ‘as’ ”65. The study of this kind of phenomenon extends back in time, at the very least – in the Western canon – until to Aristotle (some reconstructions in de Bustos 2002, Cacciari 1991; Ghiazza 2005; Leezenberg 2001). It is often said (Lakoff and Johnson 1980, 1999, Kövecses 2010) that metaphors have always been conceived as denotational fallacies and rhetorical ornaments, with the extreme result that twentieth-century Anglo-American thinking about metaphor has been emasculated, narrowed and inhibited by logical positivist views of language and is therefore either hostile or patronizing toward figurative expression (Johnson 1981:16). This position is largely acknowledged since even in the words of one of Anglo-American philosophers of language, “nobody would claim that the study of metaphor has been one of analytical philosophy's brighter achievements” (Blackburn 1984:180). However, the importance of analogy for the actual growth of knowledge was – for instance – acknowledged and investigated by Aristotle (Eco 200566), Vico (Danesi 1993), Kant, Cassirer, Blumemberg, Weinrich (Jäkel 1999), Locke, Hobbes (Mouton 2010), Breal (1897), Nietzsche (1979; cf. Brigati 2009, Fabbrichesi Leo 2009), Peirce – in whose pragmatism metaphors are “vital vehicles in the process of understanding, because it is only through the discovery of 65

66

Definitions vary on a common core, cf. for instance The Oxford English Dictionary (3rd edition) “a figure of speech in which a name or descriptive word or phrase is transferred to an object or action different from, but analogous to, that to which it is literally applicable” or the American Heritage Dictionary “ a figure of speech in which a word or phrase that ordinarily designates one thing is used to designate another, thus making an implicit comparison”. Eco notices how the cognitive power of metaphors can be inferred when Aristotle writes in Poetics (1459a8): “metaphor is the best of all tropes since to understand metaphors means to know how to see similarities or related concepts” (Eco 2004: 5). Compare also: “It may be that some of the terms thus related [i.e. in an analogy-based metaphor] have no special name of their own, but for all that they will be metaphorically described in just the same way. Thus, to cast forth seed-corn is called ‘sowing’; but to cast forth its flame, as said of the sun, has no special name. [Assuming the standard schema of analogy A:B = C:D,] this nameless act (B), however, stands in just the same relation to its object, sunlight (A), as sowing (D) to the seed-corn (C). Hence the expression in the poet, ‘sowing around a god-created flame’. (De Poetica: 25–30)

105


and the inquiry into parallelisms in other fields of investigation and or other media that new knowledge can be gained” (Johansen 1995: 73, cf. also Sørensen et al. 2007) – and more recently by the interactionist view on metaphors (Richards 1936 and Black 1962, 1993). Once we question – as all these authors do – that metaphor is simply a substitution of one term for another or a comparison, a superficial effect of language, we have on the foreground the interplay between pre-established knowledge and creativity, of how new knowledge can be produced through language. This interplay is the reason for which metaphor plays a central role in most reflections on meaning – with the important exception of analytic philosophy. The importance of metaphors for the development of contemporary semiotics can only be hinted at here. Limiting ourselves to explicit treatments of the topic67, we can find metaphors in Peirce's system of signs (Peirce 1931-58 CP 2.222; 2.277), at the core of Jakobson's articulation of paradigm vs. syntagm (Jakobson 1956), in Lotman's (1993) conception of explosion, in Greimas' (1966) structural semantics and in its interpretive development by Rastier (1987), as in the case study that allows Eco (1985) to explode the Porphyrian Tree into the Encyclopaedia, in the rhetorical system and in the name of Group Mu (Dubois 1970) and in Fontanille's and Zilberberg's (1998) tensive semiotics, and it would not be difficult to find further examples68. However, it is only with the cognitive turn in the study of metaphors (Lakoff and Johnson 1980; Ortony 1993; Fauconnier and Turner 2002, Grady 2007) – that the central role metaphors play in thought and cognition has entered mainstream research. And it is on these attempts and on the complementary perspective of the extended mind that I will focus now.

67

68

One could spend many pages in considering the metaphors structuring semiotic theories: the clear spatial structure of Greimas' generative path, energy, matter and the proliferation of containers in Fontanille's tensive semiotics and in his treatment of the body, Eco's Encyclopaedia and so on. As Basso (1999) does in a semiotically oriented extensive review of metaphor studies in the XXth century.

106


In order to present and evaluate the approaches at stake, there are some elements that have to be made explicit: i) the more general role of metaphor, ii) the level of description and explanation of metaphors, iii) the kind of conceptual and cognitive structures involved and iv) the methodological consequences for the analysis of metaphors.

2.2. Cognitive Linguistics 2.2.1.

The declared novelty of the cognitive approaches

An interesting – and for the topics at stake useful – phenomenon is that cognitive linguists, and the founders of CMT in particular, present their work as a drastic revolution in the study of metaphors, language and cognition (but cf. § 2.1.1). For instance – in the preface of Metaphors We Live By – Lakoff and Johnson state: Within a week [from starting our joint project on metaphors] we discovered that certain assumptions of contemporary philosophy and linguistics that have been taken for granted within the Western tradition since the Greeks precluded us from even raising the kind of issues we wanted to address (Lakoff and Johnson 1980:x).

This stance has been widely criticised (cf. Leezenberg 2001, Haser 2005, Rakova 2003) and has severely hindered cognitive linguistics – and CMT in particular – from engaging in a productive dialogue with philosophers (cf. Adamson 2005; Haser 2005). If it is sometimes true “that the success of ideas relies more heavily on clarity of purpose than on completeness and that giving predecessors a short shrift may well be essential for a new paradigm to take hold” (Lukes 2004, referring to a Kuhnian vision of science), the fact that Lakoff’s and Johnson’s strong claims are still largely unchallenged in the cognitive linguistic literature almost 30 years after the development of the theory is not a comforting sign. However, this allows us to sketch a caricatural and revealing dichotomy between so-called traditional theories of metaphors and CMT. In a recent systematisation of CMT by Zoltan Kövecses 107


(2010: viii) the opposition is spelled this way: Traditional Theory of Metaphors

Conceptual Metaphors Theory69

Metaphors are properties of words

Metaphors are properties of concepts

Metaphors are mainly used for artistic and Far from being ornaments, metaphors are used to rhetorical purposes Metaphors

are

the better understanding of certain concepts based

on

a

pre-existing Metaphors are often not based on similarity

resemblance between the two entities that are compared and identified Metaphors are a conscious and deliberate use of Metaphors are used effortlessly in everyday life by words

ordinary people

Metaphors are a figure of speech we can do Metaphors are an inevitable process of human without

thought and reasoning

According to these “traditional” theories of metaphors, metaphors are a question of linguistic enunciates used in a non-literal way (tropes). Metaphors would thus be tropes characterised by a basic linguistic form to which they can be reduced: "X is Y", where X (the tenor) is represented through Y (the vehicle) thanks to a pre-existing resemblance. X and Y can be single terms or more complex semantic configurations. The non-literality of metaphors is due to the presence of a semantic tension between X and Y: X cannot literally be or anyway is not Y; it is only temporarily considered such for rhetorical or aesthetic reasons and it relies on pre-existing similarities. Borrowing an example from the second part of this chapter, "Berlusconi is a caiman", the metaphorical nature of this statement appears as given, since the form

69

It has to be noted that Kövecses is not one of the most daring exponents of CMT. While being de facto its main supporter at the moment, his description of CMT does not seem to push its potentialities into the open. Conceptualising metaphors as properties of concepts and focusing on their role in “understanding” concepts, seems for instance to minimise the active role that metaphors have in constructing new concepts.

108


X is Y is respected and a person cannot literally be a caiman, it can only be temporarily considered such, namely due to the common voracity. “In metaphor [according to the classical theory] there is an implicit denotative mistake, the sentence cannot mean what it directly states since it is anomalous from the logical point of view”70 (Cacciari 1991:8) . In cognitive linguistics – following CMT – this view is radically subverted. “Berlusconi is a caiman” becomes the expression of a deeper way of conceptualising the world: the recurrence of the more general PEOPLE ARE ANIMALS conceptual pattern. In other words, a complex phenomenon like human morality and behaviour is understood through our conceptualisation of apparently simpler forms of life, animals71. Basic sensorimotor schemas are used to construct and understand such conceptual pattern. This is what talking about conceptual metaphors means: to investigate linguistic metaphors only as an access to wider conceptual configurations and to focus on deeper mechanisms for understanding and constructing knowledge. This brings into question the nature of concepts and their relation to meaning. Such considerations already emerged in chapter one and a partial answer was sketched (concepts are a specific subset of the operational mediation that constitutes cognition). Instead of producing here a more detailed account of what is meant by “concept” we will see at a later point in this work how the analyses allow us to speak of this. 2.2.1.1. The importance of studying metaphors in Cognitive Linguistics

The importance of metaphors for cognitive linguistics could be superficially attributed to purely historical reasons: the personal academic path of George Lakoff, one of the founders of cognitive linguistics, and the huge role of CMT in divulging 70 71

“Nella metafora [secondo la teoria classica] è implicito un errore denotativo, la frase non può significare ciò che afferma direttamente in quanto anomala dal punto di vista logico” As we will see, “Berlusconi is a caiman” can actually be seen as an expression of a sub-sub-type of that conceptual metaphor: PEOPLE ARE ANIMALS > TYPES OF PEOPLE ARE TYPES OF ANIMALS > RUTHLESS PEOPLE ARE PREDATORS.

109


and catalysing resources for cognitive linguistics. However, two additional reasons can be unfolded. The first one is that the prototypicality of many CMT inspired analyses lies in reducing linguistic occurrences to mere hints of underlying conceptual structures and cognitive mechanisms. The second reason is that the debate between CMT and CIT embodies the two main approaches of cognitive linguistics to semantics: investigating knowledge representation (the focus of CMT on conceptual structures)

and

investigating

meaning

construction

(CIT's

focus

on

conceptualisation): investigating structure vs. investigating processes. 2.2.1.1.1.

Historical Reasons of CMT

While many previous works laid the ground for cognitive linguistics (Fillmore 1975; Lakoff & Thompson 1975; Rosch 1975), it is CMT that has launched the development of the discipline into the mainstream of research. CMT has its immediate roots in the pages of Metaphor and Thought (Ortony 1979 [II ed. 2003]), especially in Michael Reddy's article on COMMUNICATION AS A CONDUIT. But its grand baptism was through the short ground breaking book Metaphors We Live By by George Lakoff and Mark Johnson (1980). The wide diffusion of the book created a public image for a second generation of cognitive scientists or at least of linguists. Lakoff – like Charles Fillmore and Ronald Langacker, two of the other founders of cognitive linguistics – started his academic career working on syntax following in the footsteps of Noam Chomsky, linguist and one of the founders of cognitive science in its inner representationalist initial flavour72. Lakoff's initial project was an attempt to develop a generative syntax-like semantics, an enterprise that forced him to re-direct his focus away from language-like mental and formal representations and towards a more embodied conception of cognition and meaning (cf. Chapter 1; Gibbs 2006). This conversion was received as paradigmatic. At the same time, the success of the book meant that many scholars and students in different disciplines, as well as politicians, professionals and educators, became interested in 72

On the traces of this initial collaboration still present in Lakoff's approach, cf. Pawelec 2007

110


the conceptual mechanisms behind linguistic expressions. So much so that in a recent paper Mark Johnson is able to say that: “Many studies have now shown the metaphorical constitution of basic concepts in the sciences (Magnani & Nersessian 2002), law (Winter 2001), mathematics (Lakoff & Núnez 2000), ethics (Fesmire 2003; Johnson 1994), medicine (Wright 2007), politics (Lakoff 1996), psychology (Fernandez-Duque & Johnson 2002; Gibbs 1994), music (Johnson & Larson 2003), and many other fields” (Johnson 2008: 44). To these fields it is easy to add cultural anthropology (Holland 1982, Kimmel 2005), literary studies (Fauconnier and Turner 2002; Giora 2003; Turner, 1987, 1991), religion (Soskice 1987), economics (Eubanks 1999, 2000; Kecskes 2004), education (Goulden and Griffin 1995; Staton and Peeples 2000), family interaction (Buzzanell and Burrell 1997; Hayden 2003), illness and disease (Bradac 2001), journalism (Kitis and Milapides 1997), organisational communication (Koch and Deetz 1981; Deetz 1984; Hill & Levenhagen 1995), social science (Danaher 1998, Turner 2001, Mouton 2010) and war (Kuusisto 2002; Lule 2004; Medhurst et al. 1998). Thus, it does not seem to be an exaggeration to claim that cognitive linguistics started its ascension to mainstream academic discipline thanks to CMT and that analysing metaphors means to tackle one of the cornerstones of cognitive linguistics. 2.2.1.1.2.

Prototypicality of CMT

There is a second and more relevant reason for CMT to occupy a big part of the second chapter of this work and that is its prototypicality, its embodying a certain style of doing cognitive linguistics, that is by far not the only one, but that still acts as implicit ideology for many analyses in the field73. Going through Metaphors we live by, Philosophy in the Flesh and many of the applications of the principles of CMT, a dominant analytical structure can be found. There are two main components in the analysis: a conceptual metaphor, presented in its main internal mappings and a more 73

“Conceptual metaphor is probably the best known aspect of Cognitive Linguistics: if you’ve heard only vaguely about Cognitive Linguistics, conceptual metaphor is likely to be the notion that you’ve come across” (Geeraerts 2006: 11)

111


or less organised list of linguistic expressions of it. Let us take COMMUNICATION AS A CONDUIT as presented in Metaphors We Live By, Ch. 374. We are presented with some internal mappings articulating the structure of the conceptual metaphor: IDEAS (or MEANINGS) ARE OBJECTS. LINGUISTIC EXPRESSIONS ARE CONTAINERS. COMMUNICATION IS SENDING75.

Then we are given a list of – anecdotical76 – linguistic expressions:

74

75

76

•

It's hard to get that idea across to him.

•

I gave you that idea.

•

Your reasons came through to us.

Variations of the analysis of these conceptual metaphors are to be found in Reddy 1979, Brugman 1995, Grady 1998 and de Leon 2004. In all these cases the structure of the argumentation is basically the same. For a more linguistically minded assessment and critique of the metaphor Semino 2004. While giving more importance to the linguistic aspects of the metaphor this last article does, however, not question and develop the nature of conceptual metaphors as resources for linguistically extended cognition. Cf. also note 76. An alternative – but for our purpose equivalent – reading is given by Grady 1998: CONSTITUENTS ARE CONTENTS, ACHIEVING A PURPOSE IS ACQUIRING A DESIRED OBJECT; INFORMATION IS CONTENTS, TRANSMISSION OF ENERGY IS TRANSFER, LEARNING IS ACQUIRING. Semino provides instead a more sophisticated scenario: “Within this space: – interactants can move in or out, towards or away from other participants, speech acts, conversational goals (e.g. ‘join the condolences’); – interactants can be positioned in different ways in relation to each other (e.g. ‘back’); – interactants can come into physical contact with each other in different ways, i.e. with or without pressure (e.g. ‘press’, ‘support’) or engaging in different types of physical conflict (e.g. ‘rap’, ‘hit out’, ‘bombard with questions’); – texts/utterances, their contents, or their illocutionary force can become visually accessible to the addressee via different types of movement (e.g. ‘came out’, ‘raise doubts’), via pointing (e.g. ‘point to’), or via visual representation (e.g. ‘outline’); – speech acts and texts/utterances are physical objects that can be constructed (‘make a plea’) and transferred from addressers to addressees (e.g. ‘give an order’, ‘deliver a speech’).” (Semino 2004:58) Lakoff and Johnson, to be honest, take these examples from Reddy (1979). However, Reddy's procedure when collecting them was far from scientific. We can discern from some not so explicit hints that he is using linguistic data collected from writing classes in his own institution, in particular from the assessment and feedback on written assignments. The specificity of the practice at stake, as well as the need for a more systematic and explicit procedure of data collection are not further investigated by either Reddy or Lakoff and Johnson.

112

Language and Cognition - Ph.D. Dissertation by Riccardo Fusaroli •

It's difficult to put my ideas into words.

•

When you have a good idea, try to capture it immediately in words.

•

Try to pack more thought into fewer words.

•

You can't simply stuff ideas into a sentence any old way.

•

The meaning is right there in the words.

•

Don't force your meanings into the wrong words.

•

His words carry little meaning.

•

The introduction has a great deal of thought content.

•

Your words seem hollow.

•

The sentence is without meaning.

•

The idea is buried in terribly dense paragraphs.

The linguistic “surface” of the metaphors has a merely exemplificatory function and in the work of the analyst it is merely a hint to be used to reconstruct a deeper and more essential structure. This primacy of concepts is re-stated by the fact that sometimes this analysis is followed by an attempt at showing how the conceptual metaphor is not purely a linguistic phenomenon but that it is actually enacted in many different contexts as a ready disposition for behaviour and thought. While in this case neither Reddy nor Lakoff and Johnson nor the other authors mentioned do so – we can find an argument for the presence of COMMUNICATION AS A CONDUIT in gestures in Mittelberg's study: “Phrases, nouns, verbs, morphemes, etc. are predominantly represented as imaginary physical objects held by two hands more or less apart with palms facing each other. Another possibility is to hold the thumb and index finger of the dominant or both hands up with the remaining fingers relaxed, reminding of how people take measure. Phrases, or ‘chunks’ of structure, are represented as larger entities than words and morphemes. Another variant is to extent one or both open hands with palm upwards, held out towards the audience to provide a surface on which an object, e.g. a grammatical

113

Language and Cognition - Ph.D. Dissertation by Riccardo Fusaroli category, is presented. A container with an explicit inside is evoked when speakers refer to infixes that go into the middle of a morpheme or to constituents into which the speaker seems to reach to pull out a sub-constituent. Choices between a solid object and a container or between the parts or actions that are highlighted are thus motivated by considerations of function and economy.” (Mittelberg 2006)

Indeed, CMT focuses on the discovery of large patterns of invariant projections behind both dead and innovative metaphors, thus moving the attention away from linguistic complexity in favour of a deeper understanding of the cognitive principles behind it. CMT is an attempt to go beyond language and discover underlying – and largely unconscious77 – cognitive structures that motivate and explain it78. And this happens even when CMT is applied to fields like literature (Lakoff and Turner 1989, Stockwell 2002) and politics (Lakoff 1996, 2004, 2006), where the linguistic “surface” is carefully crafted and manipulated79. Why then study metaphors and conceptual metaphors? “If Cognitive Linguistics is the study of ways in which features of language reflect other aspects of human cognition, then metaphors provide one of the clearest illustrations of this relationship” (Grady 2008:219). A richer and nonreductive articulation of this relationship being the aim of this dissertation, metaphor 77

78

79

“We can have no direct conscious awareness of what goes on in our minds. Neither [philosophical reflection nor phenomenology] can adequately explore the cognitive unconscious - the realm of thought that is completely and irrevocably inaccessible to direct conscious introspection. Conscious thought is the tip of an enormous iceberg. It is the rule of thumb among cognitive scientists that unconscious thought is 95 percent of all thought - and that may be a serious underestimate. Moreover, the 95 percent below the surface of conscious awareness shapes and structures all conscious thought. (Lakoff and Johnson 1999:12; cf. also ibidem: 81) “All of our knowledge and beliefs are framed in terms of a conceptual system that resides mostly in the cognitive unconscious. Our unconscious conceptual system functions like a “hidden hand” that shapes how we conceptualize all of our experience.” (ibidem: 13) ”What is startling is that, even for these most basic of concepts, the hidden hand of the unconscious mind uses metaphor to define our unconscious metaphysics, the metaphysics used not just by ordinary people, but also by philosophers to make sense of these concepts” (ibidem: 14) “Automatic Conceptualization: Because our conceptual systems are instantiated neurally in our brains in relatively fixed ways, and because most thought is automatic and unconscious, we do not, for the most part, have control over how we conceptualize situations and reason about them.” (ibidem: 556). Interestingly enough we can find a precursor of this view in Miller and Johnson-Laird (1976) where an explicit argument for “cognitive economy” is built, according to which the mind borrows the semantic structure of simple concepts to organise aspects of complex concepts that might be too computationally expensive to be represented in a stand-alone fashion. On the other hand, as we will see, these openings to politics and literature bring in the potential to develop CMT in the direction of a model that more ecological and more language-aware.

114


seems thus the proper place to start.

2.2.2.

Basic Tenets of CMT

2.2.2.1. Metaphors We Live By: Experientialism and image-schemas

Conceptual Metaphor Theory (CMT) acknowledges roots in Reddy's (1979) article on COMMUNICATION AS A CONDUIT, which ideas are developed and generalised to a wide variety of cases in Metaphors We Live By by Lakoff and Johnson (1980). With the publication of this book, CMT’s popularity began growing and it was further articulated and developed (Paprotté and Dirven 1985; Johnson 1987, 1993; Lakoff 1987; Lakoff and Turner 1989; Sweetser 1990; Turner 1991). There is quite a radical turning point with the publication of Philosophy in the Flesh (Lakoff and Johnson 1999) in which the developments of CMT were reassessed and redirected towards the “Neural Theory of Language” (Feldman 2006, Lakoff 2008). The flag of a more traditional approach to CMT is, however, still kept flying in the work of authors like Kövecses (2008, 2010). These developments and the gaining of such popularity did not take place without provoking debates, discussions and criticism (cf. Boroditsky 2000, Brandt and Brandt 2005, Bundgaard 1995, Evans 2004; Evans & Zinken 2008, Fauconnier and Turner 2002, Gibbs 1994, Haser 2005, Kövecses 2008, Leezenberg 2001, Murphy 1996, Musolff 2004, Rakova 2004, Stern 2000, Stjernfelt 2003, Zinken, Hellsten, & Nerlich 2011) in the most different disciplines: from linguistics to philosophy, from semiotics to neurosciences and from discourse analysis to psycholinguistics. These reactions rarely get the consideration they would deserve in contemporary presentations and discussions of CMT (cf. for instance Kövecses 2010, Grady 2008). However, to keep it simple I will here sketch the basic tenets of CMT in the form that was developed by Lakoff and Johnson, with occasional references to Kövecses. The criticism aimed at CMT will be active in the background, appearing more explicitly, when relevant, in the pragmatist reassessment of the theory and throughout the corpus analysis of “Berlusconi is a 115


caiman”. As previously mentioned, CMT is a prototypical case of potentially reductionist cognitive linguistic approach in which most linguistic metaphors are considered expressions of a very pervasive cognitive mechanism for understanding the abstract through the concrete, a normalising device operating a unidirectional projection from phenomenologically close experiential domains to less familiar ones. Our basic sensorimotor knowledge becomes central for the understanding of almost everything else. Such strongly embodied domain through a process of bootstrapping and projection enables human beings to create bodily grounded stable structures of abstract knowledge and to generate creative conceptual and linguistic constructions. Previously in the brief presentation of the metaphor “Berlusconi is a caiman” I showed the recurrence of the general conceptual pattern PEOPLE ARE ANIMALS80: we understand a complex phenomenon like human morality and behaviour through a conceptualisation of apparently simpler forms of life namely animals. To investigate conceptual metaphors does not mean to investigate language as an autonomous realm of scientific enquiry (as for instance in de Saussure 197281), and least of all to investigate linguistic metaphors in themselves. On the contrary, it means to investigate the more general cognitive mechanisms linguistic metaphors display and that make them possible. This is not a purely descriptive project: Lakoff and Johnson's quest is ultimately about meaning: how do linguistic expressions make sense? How do abstract conceptual structures make sense? To assume and demonstrate the existence of conceptual metaphors is a crucial step towards answering these issues, it is a complex and articulated move in the search for symbol 80

81

As already mentioned in the introduction, in order to distinguish conceptual metaphors from linguistic metaphors I will follow the notational conventions of Lakoff and collaborators and use UPPER CASE characters to identify conceptual metaphors. But cf. Thibault 1997 and Paolucci 2010 who present alternative readings of the Saussurean autonomy of language and linguistics, developing conceptions that are far more attuned to the pragmatist framework I am developing.

116


grounding (Harnad 1990). Therefore CMT, far from being a theory of linguistic metaphors is an all-round model consisting of some – never fully endorsed – philosophical assumptions (“experientialism”, then renamed “embodied realism”)82, a cognitive theory of concepts and a methodological apparatus implicitly defining the way metaphors can be analysed and what to search for them. Indeed, Lakoff and Johnson constructed their implicitly phenomenological83 "experientialist philosophy" in opposition to "objectivism", a fictional84 philosophical position according to which: •

Symbols get all their meaning through direct correspondence to things in the external world: meaning is “a relation between symbolic representations and objective (mind-independent) reality” (Johnson 1987: 18). As a consequence hereof

•

Nature and structure of bodies are irrelevant to meaning. As a consequence hereof

82

83

84

On at least one occasion, Lakoff and Johnson (2002:249) seem to take a step back, stating that they do not aim at developing a philosophical framework, on the contrary they simply push forward an "embodied realism" which "is not a philosophical doctrine tacked onto our theory of conceptual metaphor. It is the best account of the grounding of meaning that makes sense of the broadest range of converging empirical evidence that is available from the cognitive sciences’’ (Lakoff and Johnson 2002:249). However, this defence does not seem viable, not only because of the amount of philosophical statements present in CMT’s foundational texts, but especially because of the need of the theory and of the analytical methodologies for a sound philosophical foundation, the theory being –as I will show - as much a theory of metaphors as a theory of meaning and cognition. Experientialism is phenomenological in the intents if not in the results. The attempt at grounding meaning and concepts in experience, especially perceptual experience, and to define the conceptgenerating experiential mechanism are shared with many of the phenomenologists, from Husserl to Merleau-Ponty. Lakoff and Johnson are certainly not shy to point derisively at thinkers and to ascribe them inexcusably faulty objectivist position: from Husserl and Frege, Kant and Wittgenstein, not neglecting Aristotle and Plato, the foundations of Western thought are all naively objectivist. However, it is hard to find even one philosopher who would accept the simplistic objectivism Lakoff and Johnson describe (cf. Haser 2005 for a detailed analysis of this). One could argue that the real target of Lakoff and Johnson's criticism is the naivety with which some of these thinkers have sometimes been received and/or folk theories of truth and reality (cf. Lukes 2008). In this case, their criticism would come to gain more credibility and its importance would be enhanced, but more detailed references and analyses would be needed.

117


•

Thought is abstract and disembodied and its meaning is a function of truthconditions, of adequacy to an external reality

Beginning with Metaphors we live by, Lakoff and Johnson engage in the attempt to discover invariants behind linguistic idioms and dead metaphors. “Berlusconi is a caiman” is just one example out of many (“The police officer barked them out of the room”, “Achilles is a lion” and so on) metaphors derived from the general (or conceptual) metaphorical pattern PEOPLE ARE ANIMALS. Then comparing different patterns like this (THEORIES ARE BUILDING, etc.), Lakoff and Johnson individuated their common embodied grounding and found out that when studied in depth, they revealed shared abstract experiential patterns that were thus deemed to constitute the primitives of experience and meaning. Turning – for clarity's sake – Lakoff’s and Johnson’s path on its head we can follow the generation of linguistic expressions from these primitives and map the crucial points of the development of CMT. Making use of an implicitly phenomenological approach named “experientialism” Lakoff and Johnson employ the notion of experience to explain meaning. Experience emerges from the interaction of the subject with the world, an interaction that gives rise to directly meaningful structures: basic concepts and kinaesthetic image schemas85. Basic concepts are concepts for middle-sized objects, the objects most easily grouped together because of their perceptual salience (Lakoff 1987:133). Kinaesthetic image schemas are specific and recurring action-paths formed through time in our everyday interaction with and observation of the world around us (Evans & Green 2006:176). Lakoff argues that ‘‘these patterns emerge primarily as meaningful structures for us chiefly at the level of our bodily movements through space, our manipulation of objects, and our perceptual interactions’’ (1987: 29). For example, the containment 85

The concept is developed in Johnson (1987) and is based on Talmy’s (1985, 2000) work on the role of force-dynamics patterns in shaping syntactic constructions.

118


schema structures our regularly recurring experiences of putting objects into and taking them out of a bounded area. We experience the tactile version of this pattern when handling physical containers, we experience it visually as we track the movement of some object into or out of some bounded area or container. Furthermore, it is particularly important to note that an image schema can also be experienced cross-modally: for example, we can use the visual modality to guide our tactile and kinaesthetic experience when we reach into a container and grasp an object. In other words, it is experience in all its sensorial richness, meaningful to us by virtue of our embodiment that forms the basis of many of our most fundamental concepts. The universal character of kinaesthetic structuring follows from the ‘‘gross patterns’’ of human experience such as ‘‘our vertical orientation, the nature of our bodies as containers and as wholes with parts’’, etc. (Lakoff 1987: 303). Image schemas are thus relatively abstract conceptual representations that act as regularities in order to orient future experiences86. When a new experiential domain (emotions, abstract concepts, etc.) has to be dealt with, image schemas and basic concepts are projected onto it, constructing a basic intelligibility. We thus have some directly meaningful structures derived from bodily interaction with the world that are systematically related and projected onto abstract conceptual structures that only make sense through this process. Here we find the true significance of metaphors: these projections are metaphorical mappings, they ground a good deal of our cognition and conceptual structures. Abstract domains are claimed to be derived from concrete domains and it is claimed that this experiential grounding - the raison d'être of conceptual metaphors - comes in two flavours: experiential co-occurrence, in which two terms are seen one through the other because they often concur simultaneously in experience, and experiential similarity, where a certain similarity is used as a ground for a more extensive mapping. 86

We will deal more with image-schemas when discussing the conception of experience in CMT.

119


Recurring mappings give rise to more stable metaphorical conceptual structures (conceptual metaphors) Conceptual metaphors thus constitute a good deal of our conceptual structures and therefore are a crucial part of the way we “operationally mediate” and constitute the world. Where is language in all this? In order to produce and interpret language we rely widely on these structures. Even better: conceptual metaphors being a natural ground for cognition, they ensure the possibility of reciprocal understanding by motivating our linguistic production. How can we understand what somebody else’s linguistic utterance mean? “Conceptual metaphors are part of contemporary speakers' knowledge in long-term memory that is immediately recruited (i.e., accessed or activated) during online metaphorical language production and comprehension, as well as different reasoning tasks” (Gibbs 2006). Language is thus simply used to activate concepts that are shared by the interlocutors, because of their analogous bodily seminal interactions with the world. Linguistic idioms and dead metaphors are the expression of these underlying stabilised structures and to show that, Lakoff and Johnson entertain their reader with endless lists of conceptual metaphors followed by examples of their linguistic expression. As we have already noted, language is but one of the ways conceptual metaphors shape our cognition. Just to take another example, McNeill (1992, but cf. also Cienki 1998, Casasanto & Jasmin 2010) persuasively shows how gestures are also shaped by conceptual metaphors. He analyses for instance a gesture used by a mathematician during a conversation with a colleague about the technical concept of limits: while committing a speech error by mentioning ‘‘inverse limits’’ rather than the direct limits he has in mind, the speaker nonetheless makes the hand gesture associated with direct limits (an abrupt motion and stopping of the hand, at the ‘‘end point’’), showing that his gesture is in fact motivated by his conceptualisation of the topic— that is, an understanding in which a quantitative limit is treated as a physical obstacle or stopping point—and not by the word he is uttering at the time. It would be 120


difficult to argue that in an example like this one gestures are merely motivated by analogy with language; instead, the same underlying patterns of conceptualisation appear to motivate both gestural and verbal metaphorical uses. Other fields that have been explored in order to assess the pervasiveness of CMT beyond language are pictorial representations (Forceville 1994, 1998; Phillips 2003; El Refaie 2003; Rothenberg 2008), comics (Eerden 2009; Forceville 2005; Shinohara & Matsunaka 2009), videos (Fahlenbrach 2005, 2007; Forceville 2006), sign languages for the deaf (Wilcox 1993, Taub 2001) and cultural knowledge encoded as body habitus or action structure (Bailey et al. 1998, Kimmel 2005, Casasanto 2009). Most of these analyses conform perfectly to the diagrams we have drawn. 2.2.2.2. Philosophy in the Flesh and Beyond: Embodied Realism, Primary Metaphors and Neural Grounding

When 19 years after their ground-breaking work, Lakoff and Johnson defined the state of the art on conceptual metaphors in a magnus opus, Philosophy in Flesh (1999), they relaunched CMT in two bold directions. The first one is a more careful consideration of the developmental and neurological grounding of the theory. The second one is to use conceptual metaphors to centuries of Western history of ideas explaining it (away) through the embodied metaphorical bases of the main philosophical systems and terminologies87. Amidst this re-assessment of CMT we can observe that “experientialism” has been replaced by a partially renewed philosophical foundational theory “embodied realism”. This latest development of CMT relies on two main additions: a more sophisticated developmental stance based on primary metaphors and conflation and the neural grounding of conceptual metaphors and mappings. 2.2.2.2.1.

Primary metaphors and conflation

Christopher Johnson's PhD dissertation (1999) attempts a study of how children learn 87

An attempt that has greatly limited the role that CMT plays in the philosophical debate due to its undue simplifications.

121


the literal and metaphorical uses of the verb “to see”88. The natural context in which children learn these uses is one where the two conceptual domains of seeing and knowing are regularly correlated and where the difference between them is not experienced experienced. If parents and others regularly use “see” in contexts where it can mean either ‘perceive visually’ or ‘learn; find out’—as in “Let’s see what’s in the box” versus “Let’s see what this bell sounds like” - children may hypothesise a sense of the term which conflates literal and metaphorical meanings from the adult point of view, and later need to perform a process of deconflation before they understand that there are two distinct senses of the word and that they are linked by a conventional pattern of metaphor. Observing the evolution of the way children use the verb “to see” Johnson postulates two stages in linguistic and conceptual development: an initial conflation followed by differentiation. During the stage of conflation, “connections between coactive domains are established and the domains are not experienced as separate” (Lakoff and Johnson 1999: 49). Johnson observes that children learn to use source domain words – “to see” – with target domain meanings and grammar – as related to “to understand”. Lakoff and Johnson push the interpretation further and assume that the first stage of cognitive development involves the conflation of various domains in order to establish “primary metaphors”. These primary metaphors emerge “automatically and unconsciously” as we experience various concrete and “subjective” events together, e.g. our feeling warm while being held, the discomfort when lifting heavy objects, getting information through vision, etc. “Early conflations in everyday experience should lead to the automatic formation of hundreds of primary metaphors that pair subjective experience and judgment with sensorimotor

88

These studies were not at all new, though: as shown in Landau and Gleitman’s study (1985) on the acquisition of visual verbs by sighted and blind children, the verbs “look” and “see” are acquired as if “the meaning component ‘visual’ is never encoded in the syntax of these verbs, which is observable by both the blind and the sighted children, but rather is learned from extralinguistic observation alone, which varies for the blind and sighted children” (Fisher, Gleitman, & Gleitman 1992: 377).

122


experience” (Ibid: 49). This finding – interpreted by Lakoff and Johnson as the definitive naturalisation of association and ultimately meaning via experiential cooccurrence – is further developed by Joseph Grady. Grady investigates the complex, mature conceptual metaphors we find at work in scientific theories, literature, politics, etc. Complex and mature conceptual metaphors develop when conflation is replaced by a stage of differentiation ‘‘during which domains that were previously coactive are differentiated into metaphorical sources and targets’’ (ibid: 49): as adults we can distinguish the different domains. Even when we talk of these late conceptual stages, Grady argues, conceptual metaphors can be shown being reducible to articulated sets of more basic primary metaphors. The consequence is that all metaphoric projections can be grounded on involuntary and automatic primary metaphors, that is, experiential co-occurrence. Thus, according to Lakoff and Johnson endorsing Johnson’s and Grady’s research, our everyday cognition continues to depend on primary metaphors and on various extensions of them, mostly in the form of unconscious, neural activation. 2.2.2.2.2.

The neural grounding of CMT

This conflationary model of experience of the conceptual domain emphasises the apparent naturality of basic experiences and heavily relies on their neural grounding for which co-occurrence generates association. Indeed, already in 1988 Lakoff had started an ongoing cooperation with Jerome Feldman89 aimed at the construction of neural networks able to simulate the processing of metaphors. His research based on the initial joint work of Lakoff and Feldman, Srini Narayanan is credited with developing a viable and plausible approach to construct such networks. According to Narayanan's model, when sensorimotor and more abstract90 experiences are 89

90

“In 1988, Jerome Feldman came to the University of California, Berkeley, as director of the International Computer Science Institute, and he and I formed the NTL (Neural Theory of Language) group. Feldman is one of the founders of the theory of neural computation, and we have been working together since then” (Lakoff 2008:17). Narayan tries to conflate these two categories with “objective” and “subjective”. I will not follow this undue conflation.

123


conflated, permanent neural connections in the networks serving the two domains become established. Primary metaphors thus appear as a result of simultaneous activation in the two networks. “Via the Hebbian principle that neurons that fire together wire together, neural mapping circuits linking the two domains will be learned. Those circuits constitute the metaphor” (Lakoff 2008:26, cf. also Feldman 2006). The directionality of metaphors is due to the fact that the sensorimotor neural system has more inferential connections and therefore a greater inferential capacity than neural systems characterising abstract experience in itself91. People have the same bodies and basically the same relevant environments. Therefore, we will have very much the same experiences in childhood in which two domains are simultaneously active, and so we will learn neural metaphorical mappings linking those domains naturally, just by functioning in the world. Everyday physical interactions with the world give human beings the experience and the suitable brain activations to give rise to a huge system of primary metaphorical mappings. Human beings all over the world share these basic interactions and therefore, each on her own, developed a largely overlapping basic conceptual system without any awareness. By best fit, different cultural frames will combine with those primary metaphors and give rise to different metaphor systems. The technical details of the implementation of conceptual metaphors through neural networks do not interest us here (for more details cf. Feldman 2006, Lakoff 2008), what is interesting is: i) the shift in the epistemological framework; ii) which parts of the theory Lakoff still considers valid after 20 years of research into the Neural Theory of Language. As regards the first point, from Philosophy in the Flesh onward there is a shift in the primary explanatory power, from introspective linguistic analysis to computational

91

Narayanan, Lakoff and Feldman all seem to take this “fact” for granted, but not many reasons are given for it. As we have seen in chapter 1 a great deal of research as well as the pragmatist framework we are developing point to a more comprehensive conception of embodiment and experience, including social and abstract dimensions.

124


modelling of hypothesised neural activity92. While before the focus was on making largely implicit experiential structures of conceptualisation explicit, now what matters is the neural binding that follows from co-occurrence. What matters seems to be the neural co-activation, activation which flows along neural connections between distinct parts and which stimulates synapses to chemically change and grow stronger. The “mapping” in metaphor is neural circuitry strengthened and made permanent. This direct focus on the neural aspects of experiential correlations and metaphorical projections seems to lead outside of the focus of this dissertation, a dissertation which aims at investigating meaning and experience from a pragmatist perspective and that is only indirectly supported by cognitive mechanisms. Anyway, a more careful reading of the Neural Theory of Language literature shows that Lakoff, Narayanan and Feldman still widely recur to experience in order to train the neural networks into giving the right answer and in order to judge the results of such process. Indeed, virtually any neural network is but a diagrammatic way of inferring the mechanisms underlying linguistic and conceptual metaphors, it is not an explanation in itself. Experience leads the assumptions to be tested. Overly simplified neural structures are hypothesised and constructed. Inputs selected or constructed by the researchers as relevant are fed to the machine or the algorithms during the training sessions and the evaluation of the configurations is human-driven and relies on the researchers' understanding of their own experiences, in particular when the process is focused on linguistic meanings. As regards the second point, I will here paraphrase Lakoff's latest words on the topic (Lakoff 2008). Metaphors are a huge system of fixed conventional mappings that 92

“As of 1997, the old cognitive metaphor theory became the neural theory of metaphor with Srini Narayanan's work, as described in Philosophy in the Flesh (1999), and updated in my paper in Ray Gibbs' Cambridge Metaphor Handbook (2008). In the past year, neural linguistics has become a reality. Within it, blends are special cases of ordinary neural binding. When analyses of blends are done in NL, neural binding suffices. Within the current theory of neural best-fit mechanisms, the real properties of blends as described by Fauconnier and Turner fall out as consequences. There is no neural mechanism of "blending." The blending literature is superb on examples of real phenomena, but the old blending theory — like the old conceptual metaphor theory — does go deep enough, as the neural foundations reveal”. (Lakoff, personal communication).

125


exists physically in our brains93. The mappings happen typically across conceptual domains and operate partially by projecting source domain frame and image-schema structure onto the target domain and by acting as premises for reasoning about it. These mappings are often motivated by correlations in embodied experience, a motivation implied by the concept of image schema. Simpler correlations – primary metaphors – are composed in commonplace frames and complex metaphors. These pervasive conceptual structures are part of the cognitive unconscious and are learned and used automatically without our being aware of it, structuring a plurality of linguistic expressions that only involve parts of them and thus creating the possibility of novel metaphorical language. Being part of the cognitive unconscious, conceptual metaphors are commonly taken “as defining reality, and [we] live according to them” (Lakoff 2008). Two points have to be emphasised. In the “experientialist” phase of CMT the issue of consciousness was only sporadically touched upon, conceptual metaphors were represented as mainly working automatically, but they also crossed the boundary between conscious and unconscious cognition: “We draw inferences, set goals, make commitments, and execute plans, all on the basis of how we in part structure our experience, consciously and unconsciously, by means of metaphor” (Lakoff & Johnson 1980: 159). With the focus on the – still uncontroversially unconscious – neural networks, as we said, conceptual metaphors become “part of the cognitive unconscious and are learned and used automatically without awareness”94. The second point is that the focus on neural networks and their mechanisms brings Lakoff to postulate one and only one mechanism as a grounding principle for conceptual 93

94

Interestingly enough, Lakoff rarely quotes more interesting attempts at locating conceptual metaphors correlates in the brain like Rohrer (2002, 2005) and Pulvermüller (2005), authors who are more concerned with how sensorimotor areas of the brain are co-opted in the understanding of literal and metaphorical. This insistence on conceptual metaphors as unconscious leads to a contradiction in Lakoff's work on politics. On the one hand, Lakoff spends many words in trying to persuade the Democrats to develop more coherent and pervasive conceptual metaphors for their values and regrettably less on creating a critical awareness in the electorate. On the other hand, if conceptual metaphors are constitutively unconscious how is this work supposed to be able to produce any effect?

126


metaphors, at least at the root of the metaphoric cognitive mechanism: association due to co-occurrence. GOOD IS WARM because the infant experiences sensations of well-being in concomitance to being hugged. We will come back to this. 2.2.2.3. Analytical Methodology

According to this grounding framework, to understand a metaphor thus means to individuate the mappings between the two conceptual domains that it involves and to individuate their deeper experiential grounding. When individuating these mappings four principles have to be kept in mind: 1) the static nature of mappings, 2) the unidirectionality of mappings, 3) the invariance principle and 4) the partiality of the mappings. As regards the first point: “mappings should not be thought of as processes, or as algorithms that mechanically take source domains inputs and produce target domain outputs. Each mapping should be seen instead as a fixed pattern of ontological correspondences across domains” (Lakoff 1993:210). Mappings are not constructed locally, they are pre-existing and relatively invariant conceptual structures that precede and constrain linguistic production and understanding. The second important point to observe is that these mappings are unidirectional since our understanding usually proceeds from the more concrete (source) to the more abstract (target). The third point is that these mappings have to follow the “Invariance principle” (developed by Lakoff and Turner 1989: ch. 4): “Metaphorical mappings preserve the cognitive topology (that is, the image schema structure) of the source domain, in a way consistent with the inherent structure of the target domain” (Lakoff 1993: 215). But preserving the cognitive topology does not mean preserving the whole structure of the source conceptual domain. Mappings are usually incomplete. To say that Berlusconi is a caiman does not mean to say that he is green and lives in a river, for instance. The projected traits are the result of a selection. So which selective mappings motivate the production and understanding of 127


“Berlusconi is a caiman”? According to CMT to do so we need to take the larger perspective and understand what happens when people are compared to animals in a wide variety of cases. 2.2.2.3.1.

PEOPLE ARE ANIMALS as a conceptual formation

The choice of Berlusconi is a caiman as a case study has been partially motivated by the extensive studies that have been conducted and published around the PEOPLE ARE ANIMALS conceptual metaphor. Human cultures are rich in different forms of associations between people and animals, already from totemism and animism. But people are not conceptualised as animals also only in the so-called primitive ways of thinking: in everyday language plenty of idioms define people through animal figures just as Aesop did and Disney still does. Even the most sophisticated forms of thought do not seem to escape this metaphorical conceptualisation. Sophisticated philosophers from Heidegger to Derrida (2008), from Deleuze and Guattari (1980) to Agamben (2002), have struggled with the definition of what an animal is and to figure out how this definition relates to and defines human beings. Some branches of cognitive sciences are based on and at the same time debate the adequacy of experimenting on animals in order to be able to explain human cognition and meaning-making (cf. Crist 1999). Therefore, it was a very obvious step for CMT to point to the conceptual and pervasive structure of these associations: “much of human behaviour seems to be metaphorically understood in terms of animal behaviour” (Kövecses 2010: 124). Focusing on the linguistic expressions of PEOPLE ARE ANIMALS, there appears to be ‘‘so many animal phrases and expressions that a book listing them all would become a small dictionary’’ (Degler 1989:xiii). Indeed, such dictionaries are not uncommon. Brinkmann (1878) and Riegle (1907) investigate animal expressions in German, English, Italian, Spanish and French, carefully describing the frequency of such expressions. More recent works include Lyman (1983), Ammer (1989) and Palmatier (1995) for English; Wang (1991), Tang (2002) for Mandarin Chinese; 128


Röhrich (1991) for German; Davies and Bentahila (1989) for Moroccan Arabic; Fraser (1981) for insulting animal expressions in languages other than English. According to Low (1988) and Newmark (1988), animal metaphors are to a great extent used to describe inferior or undesirable human habits and attributes. O’Donnell (1990) focuses on the description of common and creative figurative meanings assigned to animal names and animal metaphors in languages. Holmes (2008) gives examples of the chicken metaphor in her sociolinguistic analysis of sexism in language. Sutton (1995) and Fontecha and Jiménez Catalán (2003) show how the main metaphorical meanings of female terms connote worse qualities than those of male terms. Other insightful investigations on animal expressions and their negative connotations can be found in Whaley and Antonelly (1983), Funk (1985), Norrick (1986), Claiborne (1988) and Nesi (1995). 2.2.2.3.2.

PEOPLE ARE ANIMALS as a conceptual metaphor

More Than Cool Reason by Lakoff and Turner (1989: Ch 4) presents the first cognitive linguistic treatment of the conceptual metaphor underlying animal expressions. Some examples are provided: “1) Pigs are dirty, messy, and rude; 2) Lions are courageous and noble; 3) Foxes are clever; 4) Dogs are loyal, dependable, and dependent from their masters; 5) Wolves are cruel and murderous; 6) Gorillas are aggressive and violent” (Lakoff and Turner 1989:194). Then the authors proceed to analyse the underlying conceptual formation as a specific case of the more general GREAT CHAIN OF BEING conceptual structure: non-human-made physical things, complex objects, plants, animals and humans are part of a continuum and better understood one through the other, in particular the more complex (humans) are commonly understood through the less complex (animals). “At any level in the basic Great Chain, the highest properties of beings at that level characterise that being. For example, the highest level properties of animals are their instincts [...] Similarly, the mental, the moral, and the aesthetic are generic level parameters of human beings” (Lakoff and Turner 1989: 167). 129


Trying to bring this treatment of PEOPLE ARE ANIMALS up to date with the more general CMT framework and its attempt at explaining meaning through basic experiences we can say that specific classes of animals – dogs, cats, and so on – and their behaviours are basic level concepts: concepts for middle-sized objects grouped together because of their perceptual salience (Lakoff 1988: 133). In a parallel discussion Wierzbicka (1985, 1996) argues that in people’s conceptualisation of animals the following aspects of animal life appear to be highly significant: “habitat,” “size,” “appearance,” “behaviour,” and “relation to people”. These significant aspects of human conceptualisation of animals form “a body of culture-dependent, automatically retrievable (frame) knowledge about animals, in which “relations to people” appear to be the most fundamental” (Martsa 2003:4) and they are projected onto human beings in order to categorise and understand them (better). This projection follows the communicative principle of The Maxim of Quantity – inspired by Grice (1989): “It is only the essential, culturally and psychologically salient properties, such as behaviour, internal states, desires, emotions, limited cognitive abilities of animals that are mapped onto humans, and, consequently, it is these properties that are lexicalized in the form of various linguistic constructions (Martsa 2003:5). After having introduced basic concepts, I have now to take into account Grady (1997:219)’s attempt at disimplicating the primary metaphors that motivate conceptual metaphors. Grady shows how the GREAT CHAIN OF BEING is structured by more basic primary metaphors like SPECIFIC IS GENERIC (Berlusconi shares features with the class of caimans and not with a specific one), MORE IS UP (more complex beings are higher beings) and MORE IS BETTER (more complex beings are better beings). These qualitative observations have been tested both through cross-cultural (Kövecses 2005, Hsie 2006; Talebinejad & Dastjerdi 2005; Ahrens & Say 1999) and quantitative corpus linguistics approaches (Deignan 2003, Hsie 2006; Ahrens & Say 1999; Dobrovol'skij & Pirainen 2005, Allan 2009). 130


Let us briefly consider two representative examples of this procedure. Kövecses, in his (qualitative) attempts at capturing the interplay between a very general invariant level and a cultural level motivating more superficial variations, suggests that the general level HUMAN IS ANIMAL contains at least the following four articulations: •

OBJECTIONABLE HUMAN BEHAVIOUR IS ANIMAL BEHAVIOUR

•

OBJECTIONABLE PEOPLE ARE ANIMALS

•

DIFFICULT-TO-HANDLE THINGS ARE DOGS

•

SEXUALLY ATTRACTIVE WOMEN ARE KITTENS (Kövecses 2010:125)

Kövecses concludes that the conceptual traits of /objectionability/ and /undesirability/ are the main meaning focus of such metaphors. Ahrens and Say, using a more rigorous quantitative approach, compare English and Mandarin Chinese. The statistical tendencies of PEOPLE ARE ANIMALS are: In English •

appearance -> appearance

•

behaviour -> behaviour

•

sounds -> appearance/behaviour

In Chinese: •

appearance (body part) -> appearance (body part)

•

behaviour -> behaviour

Both languages share another feature of this mapping: "The mapping of negativity is quite pronounced (except perhaps cat and squirrel in English, dragon and tiger in Mandarin Chinese), which is the natural result of most English and Chinese speakers feeling that humans are a priori better than animals". This is a point underlined by many other cognitive and non-cognitive studies. Low (1988:133), for instance, points out that expressions relating humans to animals emphasise undesirable traits,

131


reflecting human views of animals as lower forms of life95. Drawing on these findings I can further elaborate on three points: i) the relation between conceptual metaphors and linguistic expressions; ii) the directionality of the mappings and iii) the processes of abstraction at play. In all cases what is attempted is the construction of a type that subsumes the manifold of linguistic manifestations and their cultural variations under a (set of) rule(s) (cf. for a critique § 1.2.2.4.1). Such type and rules are constructed on the basis of the analysis of a limited set of samples – analysis performed by introspection and intersubjective control of it. Then it is operationalised and tested on a large indiscriminate corpus of linguistic occurrences in order to specify the internal articulations of the type. In short, while CMT tends to presents itself as a theory of reflective, interactive engagement between bodies and environments, in its linguistic analysis – exactly because it considers language as an expression of more basic cognitive processes – it risks turning into a determinative, unidirectional theory of conceptual determination. The metaphors that emerged from a reflective interaction with the world now seem to act upon language and experience as automatic, determinative structures. The second point to underline is the directionality of the mappings. As we have seen mappings in CMT are unidirectional, however, the traits projected on human beings (rude, cruel, murderous, etc.) are human already, or at least not exactly animal instincts. When this issue was noted by Lakoff and Turner as early as 1989, they argued that daily experiences with animals and with cultural representations of animals make them plausible candidates for highlighting and classifying more complex human behaviours through three steps: i) a personification of the animal, ii) a propensity to see entities in terms of a single defining feature of the animal which is selected on the basis of the feature that appears most typical or distinguishing; iii) a projection of the thus individuated anthropomorphic animal trait onto the human 95

It has to be noted, however, that although many animal metaphors are applied as abuse, quite a few are of another nature. Some animal metaphors are used positively (Richard is a lion) and some neutrally. These variations do not get explained considered nor explained.

132


being. This means that animals were personified first, and then these “human-based animal characteristics” were used to understand human behaviour”. (Martsa 2003:125). The positive side of this explanation is that culture (the cultural representation of animals) enters the experiential basis of metaphors – and thus concepts – in full. The negative side of it, however, is the explaining away of the already human dimension of the animal traits projected. This issue is not new, however. Just to mention relatively recent examples, I. A. Richards (1936) coped with this when analysing “The poor are the negroes of Europe” and Max Black (1962) when describing “Man is a wolf”. The perspective they developed is called “interaction view of metaphor” and has as its cornerstone the idea that there is a reciprocal interaction between the conceptual systems at stake: “negroes” and “wolf” are not left unchanged by the metaphor, but the properties they project onto “the poor” and “man” are informed by these same terms96. The unidirectionality of conceptual metaphors, as well as the conception of source domains as more primitive than target domains, are put in discussion by these observations,. If our concepts of animals are anthropomorphic to begin with they are certainly not simpler than our concepts of human beings. The selection of the relevant properties and in many cases the properties themselves are not simply bodily motivated regularities as certain simplistic views on CMT would want but regularities embodied and constructed through interaction with the intersubjective, social and encultured environment. The ground for the formation of schemas and conceptual structures reveals its complexity. The third point to be noted is the importance of the dimension of abstraction and generality that enters the scene. All sorts of animals are represented in the corpora analysed: domestic animals as well as exotic ones, most frequently encountered animals as well as mythological ones. Some animals are certainly more salient than

96

“In the given context the focal word [...] obtains a new meaning which is not quite its meaning in literal uses, nor quite the meaning which any literal substitute would have” (Black 1962:28). Such a view has been an unmentioned source for CMT and it involves a richer model of meaning construction in metaphors than CMT.

133


others and this is reflected in the frequency of the related linguistic expressions in the corpus. However, the general schema of PEOPLE ARE ANIMALS is claimed to hold good for all animals, including some not present in everyday experience. A dimension of abstraction and generality seems to be present already in the source domain, since a general trait of “animality” is selected and constructed97. Again the experiential bases of metaphorical mappings reveal important and already semiotic dimensions that have to be accounted for. 2.2.2.3.3.

Berlusconi is a caiman as an expression of PEOPLE ARE ANIMALS: part 1

No in depth analysis of Italian linguistic expressions of PEOPLE ARE ANIMALS has been done. It is, however, possible to extend the general schema of the conceptual metaphor from the intercultural studies previously mentioned to the Italian language as well. Elements salient to human beings (appearance, behaviour, relation to human beings) will be selected and projected and a strong possibility of a negative connotation will be considered. Caimans are not a common experience for the average Italian speaker. Not much can be said about them. Relatedly the word “caiman” was not commonly used in everyday Italian language. To strengthen this intuition a corpus search on the use of the word “caiman” before the coining of the metaphor (that is from 1st January 1984 to 14th May 2002) has been performed. The occurrences are few – 79 – and mostly related to misspellings of the Cayman Islands. Interestingly enough, even in the few literal occurrences no articulated conceptual detail is expressed in the text98. While not a proper proof of the scarce presence of caimans in Italian experience and of the conceptual salience of the animal, these data are still an indirect strengthening of the 97

98

Heider and Simmel (1944) show the salience of the perceptual trait of animacy for defining a notion of agency that can be attributed even to inanimate objects and this could account for the generality of the folk definition of animals. However, the trait of animality is attributed also to animals that have not been experienced and perceived. For instance dragons and phoenixes. This strengthens my emphasis on linguistic, abstract and social dimensions of experience. A detailed analysis of the presence of the lexeme “caiman” will be presented later on, in the section of the chapter dedicated to the corpus analysis.

134


intuitions of the author99. To proceed with our analysis we will rely on Kövecses (2010) analytic criteria to describe metaphors: conventionality; functionality; nature and generality100. 2.2.2.3.3.1.

Conventionality

While “conventional” in many linguistic and semiotic traditions means “arbitrary”101, in cognitive linguistics it assumes the meaning of “level of entrenchment” of the metaphor in everyday life and in everyday language use in a specific linguistic community. This is a conception that is very compatible with the conception of habit as developed in chapter 1. Conventional metaphors can be deeply entrenched in linguistic usage since they are experientially motivated and since their frequent recurrence contributes to their experiential salience. PEOPLE ARE ANIMALS is clearly a very common conceptual formation (cf. § 2.1.5.1). However, the use of the specific linguistic metaphor that relies on the concept of “caiman” is definitely unconventional (§ 2.1.5.2). 2.2.2.3.3.2.

Functionality

We have seen that the function of conceptual metaphors is to structure our understanding of and interaction with the world. This structuring can assume three different forms: 99

100 101

Informal enquiries on the conceptual knowledge of caimans in laypeople have been conducted by the author in Spain, Denmark and England: very few of the participants at the conferences and seminars where previous versions of this work have presented had any knowledge of the term and even in those few cases “caiman” was defined as a “crocodile” or “alligator”. While the use of the terms “nature” and “generality” does comply with the rest of my framework, I am maintaining Kövecses’ terminology for clarity’s sake. The notion of arbitrariness is central in structural linguistic and semiotics, for instance. De Saussure defines the nature of the sign as constitutively arbitrary, that is, the relation between signified and signifier is unmotivated. This grounds the notion of value. However, Benveniste convincingly argues that in Saussure arbitrariness is only limited to the relation between signs and external reality. Indeed, each sign is defined by the system (langue), therefore its intra-linguistics relations are necessary and motivated. This brings the notion of convention close to the notion of habit that we have developed in the previous chapter (although devoid of its socially embodied nature).

135


1. Structural metaphors: The mappings not only explain why the particular expressions mean what they do, but they also provide a basic overall structure and hence understanding. For instance in LOVE IS A JOURNEY the sedimented and articulated knowledge of journeys is systematically used to make sense of life, for instance inferring that life has a direction as in “He has not a direction in life”. 2. Ontological metaphors: Instead of providing an internal articulation to the concepts, these metaphors give abstract aspect a concrete form, enabling the speakers to conceptualise and manipulate them more easily. This is the case with for instance IDEAS ARE THINGS. 3. Orientational metaphors: The mappings establish coherence in a set of target concepts. This is the kind of metaphor behind MORE IS UP, HEALTHY IS UP, CONTROL IS UP, HAPPY IS UP, VIRTUE IS UP and so on, granting to all these families of expressions a common spatial orientation. PEOPLE ARE ANIMALS seems to be a case of structural metaphor. However, as we have already noted, the structure projected from animals to humans does directly derive from the animal domain. 2.2.2.3.3.3.

Nature

By nature we mean here the experiential grounding of metaphors. In the developments of CMT we find three main kinds of metaphorical grounding. The first one is conceptual knowledge: metaphors built upon our basic knowledge of concepts, PEOPLE ARE ANIMALS for instance, seem to be based on our knowledge of animals. The second possible grounding is that of more abstract image schemas. A very clear example of this can be found in the use of prepositions: “pass out; space out; zone out; tune out; veg out; conk out; rub out; snuff out; out of order; be out of something” (Kövecses 2010:43). All of these uses give access to the representation of a lack. While in knowledge-based metaphors the mapping between domains is rich, 136


in this case only a very schematic topological structure is mapped102. The third kind of grounding is less conceptual and more perceptual: it is called one-shot image (Lakoff and Turner 1989) or image-metaphor (Lakoff and Johnson 1999). The classical example given is “my wife [...] whose waist is an hourglass” by André Breton cited by Lakoff and Johnson 1999. The status of image-metaphors as conceptual metaphors is debated (cf. Grady 1999 that denies them any conceptual status relegating them to mere “resemblances”), however, it is not difficult to see how in “Berlusconi is a caiman” image metaphors can easily be found: The seemingly sleeping caiman on the banks, the sudden snap of the caiman's jaws, and so on. We will come back to this. 2.2.2.3.3.4.

Generality

The last element to be considered is the level of generality of the metaphor. By generality it is usually meant the level of abstraction from details. Therefore, generic conceptual metaphors would be THE GREAT CHAIN OF BEING or IDEAS ARE THINGS while specific conceptual metaphors would be PEOPLE ARE ANIMALS or LIFE IS A JOURNEY. Even more specific are image-metaphors with their figurative overlappings of terms and domains. While these three points are useful reference points, the generality of the metaphorical conceptual constructions should be considered as a continuum. Considering “Berlusconi is a caiman” we could go into more specific details. PEOPLE ARE ANIMALS does not really convey more specific aggregations of conceptual structures. We can argue for the existence of TYPES OF PEOPLE ARE TYPES OF ANIMALS and even more specifically of RUTHLESS PEOPLE ARE PREDATORS and each of these sub-formations is supported by countless metaphorical linguistic expressions, from “crocodile tears” to “that lawyer is a shark” and more complex blends like vampires, werewolves and talking snakes. Even 102

As we will see in the second part of this chapter, the metaphor Berlusconi is a caiman involves also an image schema.

137


specific types of animals end up being associated to very specific properties, showing how conceptual structures can be fine-grained. 2.2.2.3.4.

Berlusconi is a caiman and PEOPLE ARE ANIMALS: an overview

We have thus seen that Berlusconi is a caiman is the linguistic expression of a deeply entrenched (conventional) conceptual metaphor: PEOPLE ARE ANIMALS. Relying on the very generic metaphor THE GREAT CHAIN OF BEINGS, PEOPLE ARE ANIMALS is the result of systematic comparisons between animals and human beings. These mappings select experientially relevant traits (behaviour, appearance, relation to humans) of the animals and attribute them to human beings, generally over-imposing a negative, objectionable or at least caricatural connotation. Such structure with different saliences and focalisations seems to hold good in a wide variety of languages, investigated through corpus methodologies. While relying on sub-formations like TYPES OF PEOPLE ARE TYPES OF ANIMALS and even RUTHLESS PEOPLE ARE PREDATORS, Berlusconi is a caiman is an unusual expression of such a conceptual metaphor, especially because of the low recurrence of the term “caiman”. We might predict that it projects a negative light on Berlusconi and that there might be some imagistic element involved. However, CMT does not seem to be able to predict the specific properties projected – and this, to be fair, is not its focus. 2.2.2.4. CMT: An assessment

In § 2.0 we established some points to be considered in order to evaluate the approaches examined. 1. The more general role of metaphor for the approach at stake. Linguistic metaphors are just a way to access wider conceptual formations and basic mechanisms of human cognition. In this sense CMT has paved the way for other cognitive linguistic approaches also attempting this sort of methodological reduction; 138


2. The level of description and explanation of metaphors. Metaphors are initially explored at an introspective level using as data isolated examples; such initial exploration is then extended via corpus methodologies on the one hand and to neural network simulations on the other. Anyway, although sometimes the explanatory level seems to be set at a neural level and the description to be distributed throughout the corpora, conceptual metaphors are essentially an experiential and meaningful phenomenon that allows the analyst to explain both abstract and local meaning through directly meaningful experiential primitives. 3. The kind of conceptual and cognitive structures involved. Conceptual metaphors are based on experiential associations giving rise to primary metaphors and image schemas. These allow us to operate metaphorical projections involving basic concepts and generating abstract and articulated domains. 4. The methodological consequences for the analysis of metaphors. The analysis is thus aimed at showing this grounding of linguistic (or otherwise semiotic) metaphors in a wider and deeper conceptual system where experiential and directly meaningful primitives motivate a large deal of everyday and abstract conceptual formations. This excursus, the diagrammatic representations and the still barely sketched analysis of “Berlusconi is a caiman” as a linguistic metaphor has led us to underline some initial problems in the orthodox presentation of CMT, but also some directions to be pursued to integrate and enrich it. Embodied realism needs a sounder philosophical articulation, for instance when it comes to its definitions of experiential association, conceptual domain and abstraction. The concept of mapping needs to be supported by a more articulated model of which selections are made. The idea of conceptual metaphors, if initially the expression of a fully engaged interaction between the subject and the environment ends up being applied in a deductive and determining way that does not seem to account for local specificities and sedimented interpretive 139


practices. Language is rarely explicitly thematised as anything else than an expression of underlying dynamics, in clear contrast with the framework sketched in Chapter 1.

2.2.3.

Conceptual Integration Theory

Already within the cognitive linguistic authors who focus on meaning construction, some of these problems were noticed. The emergence of meaning and inference in blended spaces was overlooked as a theoretical issue in earlier work on basic metaphor, probably because the focus on abstracts mappings at the superordinate level obscured some of the principles of on-line construction of meaning in actual, specific cases. (Turner & Fauconnier 2002: 134)

CIT is exactly an attempt to focus on this online construction of meaning, an attempt that focuses on a more micro-grained analysis of how meaning emerges in non-linear ways integrating stable and unstable conceptual configurations. In his recent overview of the state of the art of CMT, Kövecses declares – almost dismissively – that CIT overcomes some CMT issues and, in particular, through CIT “we can provide more refined analyses of literary texts, and we can better handle certain problems that arise in connection with the metaphor analysis as presented so far” (Kövecses 2008:232). Ruining the suspense, I can anticipate that CIT contributes to the discussion of metaphors – far beyond the mere providing a more refined analysis of literary texts – along the following dimensions:

•

In CMT the focus is on structure of the source domain that gets unidirectionally mapped. In CIT the focus is on the interaction of mappings from multiple spaces and on the structures emerging from it.

•

In CMT conceptual mappings are static and pre-existing. CIT shows how metaphors are the effect of local “cobbling and sculpting”. On the one hand, cultures build networks over long periods of time that get transmitted over generations. Techniques for building particular networks are also transmitted. 140


On the other hand, “people are capable of innovating in any particular context. The result is integration networks consisting of conventional parts, conventionally structured parts, and novel mappings and compressions.” (Fauconnier & Turner 2008:53) •

In CMT we have seen conflation at work on a developmental time scale to be then replaced by differentiation. In CIT we see compression at work in ongoing cognitive processes: “For example, a cause–effect relation connecting different mental spaces in the network may be compressed into a representation relation or an identity relation within the integration network” (Fauconnier & Turner 2008: 54). The difference lies in the fact that compression is an online phenomenon that can be reverted, while conflation is a developmental phenomenon.

•

In CMT the answer to the question of what the metaphor means lies in making explicit any underlying conceptual formations and directly meaningful pre-conceptual structures. CIT answers differently, in that it aims at exposing the conceptual process of interpretation. Metaphors are more like a bundle of conditions to be enacted locally in a context and they do not have a static, predictable meaning irrespective of their actual use103.

This radical change in the analytical perspective has been seen sometimes as a complement to CMT (Grady et al 1999, Fauconnier and Turner 2008104), sometimes 103

104

Pushing the argument further than Fauconnier and Turner do I will later argue that this opens to the possibility that meaning construction occurs in a space of shared attention and that participants are collaborators in the act of “making sense”. Cf. also a recent debate on CogLing, following the call for papers for a Companion of Cognitive Semiotics on the state of the art of the study of conceptual metaphors: “As stated many times in the last few years, we view the work on conceptual blending as a scientific generalization of conceptual metaphor theory, not as an alternative to it. The generalization applies to the many metaphors not covered by a simple source-target model.” (Fauconnier); “As of 1997, the old cognitive metaphor theory became the neural theory of metaphor with Srini Narayanan's work, as described in PHILOSOPHY IN THE FLESH, 1999, and updated in my paper in Ray Gibbs' Cambridge Metaphor Handbook, 2008 In the past year, neural linguistics has become a reality. Within it, blends are special cases of ordinary neural binding.” (Lakoff on CogLing).

141


as a subtle paradigm shift that is not compatible with CMT (Brandt & Brandt 2005, Brandt 2010). More recently the founders of these two approaches have defined “the research programs developed for metaphor and blending as mutually reinforcing and often deeply intertwined, rather than at odds with each other” (Lakoff & Fauconnier 2010). The argument goes on pointing to the empirical focus of both approaches as what makes the two models two sides of the same coin and assuming the divergences only as an effect of focusing on different phenomena and in any case as not generating any deep incompatibility. The common horizon of empirical analysis – while in many respects quite lacking in both theories as we will see – is an excellent principle that could never be stressed enough. This said, a few points should be kept in mind: the choice of the phenomena to be analysed and their very constitution as phenomena depends crucially on the questions asked and the methodologies employed, two factors that depend on the theoretical framework. Exactly because of this I must stress that – while not forgetting the common empirical horizon – the (constrained) multiplication of theories and approaches is a necessary thing to avoid being trapped only in the interests of one theory. This multiplication helps us to compare and test the reciprocal limits of each theory, something which, to a certain degree, makes it worth it to emphasise differences. As we will see in chapter 3, convergence and complementarity must go hand in hand. We can easily see that CIT pushes into the foreground most of the crucial issues that we have seen at work in CMT: i) the role of linguistic metaphors, a central case for a more general cognitive mechanism; ii) the sometimes conflicting interplay of phenomenological, cognitive and neural perspectives; iii) the attempt at finding more general cognitive structures involved in metaphor production and interpretation, only this time mechanisms and not conceptual structures; iv) the not fully articulated interplay of linguistic and conceptual levels. But let us proceed with order. CIT was born from the initial work of Gilles 142


Fauconnier (1984, 1994, 1997) on mental spaces as a theoretical device to solve problems of reference and presupposition. Initially the model created a dialogue with more analytical approaches to semantics and adopted a certain mathematical rhetoric: A cette fin nous introduisons la notion d’espaces mentaux, distincts des structures linguistiques, mais construits dans chaque discours en accord avec les indications fournies par les expressions linguistiques. Dans le modèle, les espaces mentaux seront représentés par des ensembles structurés et modifiables — des ensembles [sets] avec des éléments a, b, c, ..., des relations satisfaites par ces éléments (R1ab, R2a, R3cbf, ...), et tels qu’on puisse leur ajouter de nouveaux éléments, ou établir de nouvelles relations entre leurs éléments. (Techniquement, un ensemble modifiable est une suite ordonnée d’ensembles ordinaires — il sera commode de parler d’espace mental construit au fil du discours, plutôt que de mentionner la suite correspondante d’ensembles.) / Des expressions comme “Ra1a2 ... an est valide dans l’espace mental M ” signifieront que a1,,a2, ..., an sont des éléments de M et que (a1, a2, ..., an) satisfait la relation R. (Fauconnier 1984: 32, quoted in Brandt 2004)

However, the model quickly moved towards a more cognitive perspective and through a collaboration with Mark Turner – a former collaborator of George Lakoff – mental spaces gave rise to Conceptual Integration Theory in which much of cognition is due to blending, the creative integration of two or more mental spaces. Fauconnier and Turner have been very prolific in their attempt to show that blending can be applied to various phenomena ranging from perception and conceptualisation to the origin of language105 and, not least, to the conceptual and linguistic phenomenon of metaphor (Fauconnier 1997; Turner 1996, 2001, 2003, 2006; Turner & Fauconnier 1995a, 1999, 2000, 2003; Fauconnier & Turner 1996, 1998, 2000, 2002). The wide scope of the model has provoked harsh criticism and it escapes our line of argument here. Therefore, we will focus only on conceptual

105

The human brain acquired the skill to blend some 50.000 years ago (Fauconnier and Turner 2002:389) and now “living in the human world is 'living in the blend' or, rather, living in many coordinated blends”(ibid.:390). Blends in CIT end up explaining the basic mechanisms of human cognitive processes from perceptual integrations to identity and categorisation. This over-extension of the concept falls out of our scope of analysis and will thus not be considered.

143


integration as a model for metaphors, both because of its potential descriptive and explanatory significance for our analysis and because it is a reaction to CMT internal to cognitive linguistics. A mental space is thus : A partial and temporary representational structure which speakers construct when thinking or talking about a perceived, imagined, past, present, or future situation. Mental spaces (or, ‘spaces’, for short) are not equivalent to domains, but, rather, they depend on them: spaces represent particular scenarios which are structured by given domains. For instance, a CIT account of example 1 [“The committee has kept me in the dark about this matter.”] would involve a space in which the agent is standing in the dark. While this representation appeals to our knowledge of visual experience, the recruited structure is only a small subset of knowledge of that domain. In short, a mental space is a short-term construct informed by the more general and more stable knowledge structures associated with a particular domain. (Grady, Oakley & Coulson 1999: 102)

Domains are replaced by partial and temporary representational structures (mental spaces) that speakers construct when thinking or talking. Mappings are local and situational and give rise to local, emergent and transient conceptual configurations. Pre-established patterns of embodied meaning or basic experiential associations are replaced as primitive by general mental operations which actively construct patterns and spaces of meaning. In order to understand the implications of this model for our discussion we will go through i) a definition of mental spaces; ii) a taxonomy of the functions that these mental spaces assume in order to generate a blend that iii) follows optimality principles and a human scale. This to finally iv) re-define metaphors and conceptual metaphors in terms of CIT and v) unfold the analytical methodology. 2.2.3.1. Neuron Huggers versus Circle Worshippers: On the Nature of Mental Spaces

Since mental spaces and the operations that construct, manipulate and integrate them constitute the basis of the difference between CMT and CIT, we need to spend 144


some time defining what mental spaces are and at what level of description they are situated. At first sight, not much attention is paid to the definition of mental spaces: surprisingly little attention is dedicated to discussing what mental spaces in fact are. Fauconnier and Turner (2002) only dedicate 1 out of 400 pages to this issue. Many different things are put into mental spaces: conceptual structure, perceptual structure, linguistic form, single objects, structured scenarios, unstructured scenarios, very rich and complex scenarios, very simple scenarios, sound, physical form, color, emotion, etc. However, what do all these things have in common? The answer may of course be that they are all mental, but this then potentially entails that practically all mental processes are also mental spaces. This is a gross generalization, and what insight does it give if it places everything in the same category? (Hougaard 2005: 57).

This situation might be due to the formal heritage of mental spaces. They were born as logical descriptive structures, therefore they can contain any sort of element. However, a more careful reading of CIT literature, in particular of the CIT manifesto “The Way We Think” (Fauconnier & Turner 2002) –, discloses several attempts at a cognitive and neural grounding of mental spaces. These attempts have created a ferocious intellectual debate labelled in CogLing – the most subscribed to mailing list on Cognitive Linguistics – the “Circle Worshippers vs. Neuron Huggers war”. 2.2.3.1.1.

Cognitive grounding: mental spaces and memory

In a cognitive understanding, mental spaces are structures of the working memory: Mental spaces operate in working memory but are built partly by activating structures available from long-term memory. Mental spaces are interconnected in working memory, can be modified dynamically as thought and discourse unfold, and can be used generally to model dynamic mappings in thought and language (Fauconnier & Turner 2002:102).

But what is working memory? In the words of Halligan and colleagues, working memory is “a temporary retention of information for the performance of an act that is contingent on that information” (Halligan et al 2003: 760). In other words, 145


working memory is a functional definition of the mediation of cross-temporal contingencies. Such a cognitive function has been modelled in several ways. Some models define the possibility for specific contents only for working memory: Baddeley for instance describes only two representational formats – visuo-spatial and auditory which are continuously re-enacted by the corresponding “slave modules” and when needed integrated in the episodic buffer (Baddeley 2000). However, other scholars contest these models and advance a more generally accepted and purely functionalist definition: working memory can include any sort of information as long as it is relevant. Defining mental spaces as structures of the working memory does thus not seem to constrain their potential contents in any way as long as they are “mental”106. The only constraint deriving from this cognitivist reformulation lies in the limited capacity of working memory: As a matter of on-line processing the megablend must maintain active connection across the whole network to give us a sense of overall global insight. Our hypothesis is that the megablend and its connections run into limits on working memory (Fauconnier & Turner 2002: 165).

To partially override this limit, mental spaces “can also become entrenched in long term memory. For example, frames are entrenched mental spaces that we can activate all at once” (Fauconnier & Turner 2002: 103). A statement that seems quite compatible with Cowan's (2007) model of working memory as a temporary and manipulable indexical structure for information stored in the long term memory system. This would integrate with the dialectics of entrenched and emerging structures in conceptual integration and possibly help motivating some of the dynamics and selections. However, Fauconnier and Turner do not seem to be 106

Unless, following the first chapter of this work, we consider a distributed perspective on working memory: ‘‘Instead, it [working memory] must be viewed as essentially hybrid, made up of two distinct components. In particular, the processes involved in working memory must be viewed as made up of both biological processes and processes of external manipulation of relevant information-bearing structures in the environment’’ (Rowlands 1999b: 147) and ‘‘Remembering, on this view, involves exploiting internal, bodily, and environmental resources in order to produce some sort of action, often social in nature’’ (Wilson 2004: 191). For an initial approach to these topics and CIT, cf. Hutchins 2004.

146


excessively concerned with following up on the actual integration of their model with state-of-the-art cognitive models or with exploring and making testable consequences of such integration explicit107. The second important cognitive grounding operation that Fauconnier and Turner attempt is in their emphasis on mental spaces and blending as unfolding in the cognitive unconscious, while their results flash into the consciousness without any apparent effort. Blending “seems magical precisely because the elaborate imaginative work is all unconscious” (Fauconnier & Turner 2002:44). Again we see the idea of the all-pervasive cognitive unconscious that, officially introduced in psychology by Freud, has then been translated into cognitive sciences and adopted as a banner by most scholars in the field (cf. Wilson 2002) and not least by Lakoff and Johnson. 2.2.3.1.2.

And there come the Neuron Huggers

In the last few decades, finally, cognitive methodologies and results have increasingly been tested with neuroscientific methodologies and led by neurobiological models. This is process is known as the Cognitive Neuroscience revolution which is crucially supported by brain imaging, a true technological wonder child (cf. Frith 2007). Fauconnier and Turner – just like Lakoff and Johnson – do not want to be left out and pursue a neurobiological grounding of mental spaces and blending: “[i]n the neural interpretation of these cognitive processes, mental spaces are sets of activated neuronal assemblies, and the lines between elements correspond to coactivationbindings of a certain kind.” (Fauconnier & Turner 2002: 40) And in blending “[w]hat counts as a “natural” match will depend absolutely on what is currently activated in the brain” (Fauconnier & Turner 2002: 21, cf. also Grady 2000). However, this line of research is not pursued further than the making of these generic statements and the emphasis is on how such grounding will grant more credibility to qualitative descriptions without replacing them: “Needless to say, I applaud the quest for deeper explanations that would take us all the way to the

147

Language and Cognition - Ph.D. Dissertation by Riccardo Fusaroli neural level. Such an advance would reinforce, not invalidate, theoretical accounts of conceptual mappings and grammatical constructions achieved within cognitive linguistics. We don't throw away the laws of thermodynamics because it turns out they stem from properties of kinetic particles.” (Fauconnier on CogLing 27/05/2010).

2.2.3.1.3.

The revenge of the Circle Worshippers: towards a pheno-cognitive approach

However, CIT seems to be moving beyond the purely internalistic dimension to where the cognitive and neural grounding could – but do not necessarily – lead. One way in which it does this is represented by the so-called Århus reformulation of CIT: “Blending is understood as (a form of) semiotic integration, and a mental space is conceived as an imagined gestaltually integrated representation of some scenario”. (Brandt 2003: 3, but cf. also Brandt & Brandt 2005; Brandt 2010). The move towards phenomenology opens up to all sort of semiotic grounding of mental spaces, following the lines of Chapter 1. In particular the most recent developments (Brandt 2010) richly articulate the problem of attention and relevance in a context that is social, intersubjective and semiotically defined. A second attempt at escaping the risk of internalism has been partially enacted by Seana Coulson and Todd Oakley (2008). Coulson and Oakley tackle the analysis of conversations in terms of blending and immediately they are forced to redefine mental spaces as ways “to package information about an interlocutor's center of interest within an interactive context (Coulson and Oakley 2008). This definition is intended to capture the extent to which language users formulate and understand concepts by focusing on simulating physical, social, and introspective scenes and situations in a perspective that is not too different from the one presented in the previous chapter (cf. Barsalou 2009, Zwaan & Karschak 2009). Therefore, mental spaces are the representation/simulation of distinct physical, social and/or introspective scenes and situations in which our attention is focused on a few salient elements. These representations/simulations are

148

Language and Cognition - Ph.D. Dissertation by Riccardo Fusaroli “determined by semantic frames for structuring the micro-features of a scene, such as role assignments, action and event sequences […] and can represent events unfolding in time without having to posit a different space as each change occurs” (Coulson and Oakley 2008).

What is more interesting is that readers and discourse participants while not necessarily conscious of mental spaces per se have potential conscious access to the content of each mental space in the network. “Following Mandler (2004:59-89) we suggest that mental spaces comprise implicit and explicit declarative knowledge, with different facets of declarative knowledge made explicit and left implicit as discourse proceeds” (Coulson and Oakley 2008). Mental spaces are thus temporary and unstable representations with a certain degree of internal coherence that are grounded in the ongoing discourse and realised in the cognitive system of the individuals at the boundary of consciousness, a cognitive system that in some ways attempts to account for an interactive and intersubjective context. None of these two attempts really puts us in the social arena of language, where truly distributed cognition happens. Both the attempts seem to posit that social cognition consists in the internal imagination and conceptualisation of the other and only secondarily in the activity of integrated interaction between subjects, linguistic structures and environment that follows from a distributed cognition approach. In Gallagher’s, Di Paolo’s and de Jaegher’s (2010) terms we could say that internal construction of the other constitutes the possibility for social interaction and not vice versa as a truly distributed approach would try to argue (cf. § 3.4.2). However, these attempts point to how CIT is posing the question of the role of interaction and context and that it is struggling to develop a more aware perspective on these issues. But we will come back to this after having dealt with the mechanisms through which mental spaces are blended and give rise to new meanings in context. 2.2.3.2. Blending and the taxonomy of spaces

Fauconnier's initial model of mental spaces includes a certain amount of “discourse management” primitives (Fauconnier 1997: 49, 73), that is, functions enacted by 149


mental spaces107. However, these functions tend to recede into the background as soon as the model of conceptual integration is developed. Building an integration network involves setting up mental spaces, matching across spaces, projecting selectively to a blend, locating shared structures, projecting backwards to the inputs or the blend, and running various operations in the blend itself (Fauconnier & Turner 2002: 44).

The basic model of conceptual integration is defined by a minimum of two input spaces, a generic space, and the blend. In order for the conceptual integration to be accomplished the two input spaces must share a partial common structure. This common structure defines the generic space108. Thanks to this common structure partial trans-spatial mappings can be made, mappings that generate the blending space. Mappings as in the case of CMT are partial but, differently from CMT, do not go from one space to another, but involve two input spaces generating an emerging structure, that was not contained in its entirety in any of the input spaces, and that often violates some of their structures. This emerging structure is produced through three interconnected dynamic operations: 1. Composition: the elements projected from the input spaces interact making new relations available that were not contained in the two input spaces. 107

108

Base space (being the space from which the discourse starts is the ‘anchor’ for the configuration and remains easily accessible); Viewpoint space (the space from which other spaces are currently being built or accessed); Focus space (the space where content is currently being added or that gets structured internally); Event space (a space representing the time of the event or of the state focused in the utterance, often the same as the focus space); Foundation space (a space created by a counterfactual, it is built as factual only inside the hypothetical configuration); Expansion space (a space created by the previsions made possible by the foundation space); Speech space (a space created by the speech verbs); Belief space; and I could go on. The nature of the generic space is quite interesting. Referring to Husserl's phenomenology, Bundgaard (1999) defines it as the schematic transcendental substrate that rules the correlation and the compatibility between the semiotic interpretant and the natural interpretant, between the intentional subjective instance and the morphological objective structure of reality. In any case the generic space can easily be traced back to Lakoff & Turner (1989) conceptual metaphor GENERIC IS SPECIFIC that is often at play in metaphorical formations to explain phenomena of generalisations and hypostatisation of feature. As a further hint of this filiation we can notice how the metaphor was further developed in Turner (1991) and Turner (1996).

150


2. Completion: the schematic structure in the common space is enriched by background knowledge: frames, conceptual domains and other cognitive and cultural models. 3. Elaboration or running the blend: on this growing structures cognitive work can be made by manipulating the different events as an integrated unit. “Elaboration” refers to the mental simulation of the situation depicted by the blend. New features may be included in the blend through simulating and running the blend. As Gibbs and Tendahl remark: “This offers the possibility to construct very creative blends, because elaboration is not governed by the linguistic form anymore” (Tendahl & Gibbs 2008: 1845). 2.2.3.3. Optimality Principles and the Human Scale

It should be clear by now that conceptual integration is far from a deterministic mechanism; it is not a tool for predicting meaning but for producing and interpreting it. However, in the course of the years some optimality principles (Fauconnier & Turner 2002: 327-333) have been developed that seem to guide blending. 1. The Topology Principle: Other things being equal, the blend and the inputs have to be set up so that useful topology in the inputs and their outer-space relations is reflected by inner-space relations in the blend. 2. The Pattern Completion Principle: Other things being equal, elements in the blend have to be completed by using existing integrated patterns as additional inputs. A completing frame has to be used that has relations that can act as compressed versions of the important outer-space vital relations between the inputs. 3. The Integration Principle: An integrated blend has to include a coherent scenario. 4. The Maximization and Intensification of Vital Relations Principle: Other things being equal, vital relations in the network – Change, Identity, Time, Space, Cause151


Effect, Part-Whole, Representation, Role, Analogy, Disanalogy, Property, Similarly, Category, Intentionality, and Uniqueness (Fauconnier & Turner 2002: 101) – have to be maximised and intensified, in particular in the blended space and in its projections and feedbacks. 5. The Web Principle: Other things being equal, the manipulation of the blend as a unit must maintain the web of appropriate connections to the input spaces easily and without additional surveillance or computation, this way the emerging structures and blending dynamics can retroact on the input spaces and on their mappings. This means that even when the blend is achieved the other spaces do not disappear. 6. The Unpacking Principle: Other things being equal, the blend should prompt for the reconstruction of the entire network all by itself. 7. The Relevance Principle: Other things being equal, an element in the blend should have relevance, including relevance for establishing links to other spaces and for running the blend. Conversely, an outer-space relation between the inputs that is important for the purpose of the network should have a corresponding compression in the blend. The more these optimality principles are developed and employed the clearer it becomes that the different mechanisms involved in blending are not built into the void. Blending aims at a Gestalt-like coherence, a coherence that is grounded upon basic experiential dimensions, the vital relations. Unfortunately there is no further reflection in the blending literature, yet, on why these dimensions are basic (or vital) for the human cognitive system. However, in more than one passage Fauconnier and Turner stress how when employing these principles and emphasising the vital relations, what we get is blends at a “human scale” (Fauconnier and Turner 2002: 312). The embodied dimensions that seemed to have disappeared from CIT due to its abandoning of the unidirectional mapping of CMT as ground of the blending itself have here an excellent opportunity to reacquire a central role in the theory. 152

Language and Cognition - Ph.D. Dissertation by Riccardo Fusaroli 2.2.3.4. Metaphors as conceptual integrations

Once sketched the potentially all-pervasive framework that conceptual integration gives to conceptualisation, it is time to zoom in on define the role of metaphors in it. We have already mentioned some attempts at normalising CMT inside CIT. Fauconnier and Turner (2000) define conceptual metaphors as a case of deeply entrenched single-scope blending: The single-scope network has two input spaces with different organising frames. One input projects topical elements to the blend, while relational structure, or predicates, are projected from the other. The organising frame of the blend is an extension of the organising frame of one of the inputs but not of the other. However, a closer look shows how in The Way We Think metaphors also appear in the definition of simplex integrations (XYZ-metaphors like “Necessity is the mother of invention”) and how they are also a prototypical case of double-scope blending (“this surgeon is a butcher”). Without entering the details of this internal articulation of kinds of blending that does not really help our enterprise, we can mention a second attempt: Grady et al (1999) defined metaphors (conceptual and linguistic) as conceptual integrations in which (1) a single element in the blend corresponds to an element in each of the input spaces; (2) certain very salient aspects of the input domains are prohibited from entering the blend; (3) some salient structure in the blended space is prevented from floating back to the inputs. Interestingly the authors notice that there are different degrees of metaphoricity, that is, the three previous principles constitute a prototype of metaphoricity whereto single metaphors can adhere more or less strictly. In this structure entrenched conceptual metaphors appear as pre-existing structures leading but not determining the integration. Linguistic metaphors are conceptual integrations as well, relying at different degrees on entrenched structures be they stable conceptual formations: conceptual metaphors, frames, etc. How does this then apply to “Berlusconi is a caiman”? Does this bring us further in the understanding of this metaphor? 153

Language and Cognition - Ph.D. Dissertation by Riccardo Fusaroli 2.2.3.5. “Berlusconi is a caiman” as a metaphorical conceptual integration

“Berlusconi is a caiman” clearly sets up two input spaces: one with the relevant knowledge about Berlusconi, the other with the relevant knowledge about the caiman. And already we face huge problems. What is the relevant structure of the knowledge regarding Berlusconi and the caiman? We can certainly define some very basic dictionarial facts. And through this we can define a common schematic structure for the generic space: both are animated beings, both belong to the GREAT CHAIN OF BEINGS. However, as we have already shown, there is no sedimented and structured knowledge regarding the concept of “caiman”. There is, however, the knowledge of PEOPLE ARE ANIMALS and the probable projection of salient traits (not the length of the intestines for instance) and of negative characterisation. As Fauconnier himself said: A sentence in itself has no fixed number of readings. It has a potential for generating connections in mental-space configurations. The number of readings will be a product of this potential and the spaces available (and accessible) in a particular context.” (Fauconnier 1983:54)

We need to take the context into account in order to select the content of the mental spaces and the interplay between entrenched and novel structures. We may, however, already notice how the issues we had highlighted in analysing PEOPLE ARE ANIMALS can be overcome. The problem of directionality that we had noticed in considering the anthropomorphism of the animal traits projected does not require a double mapping to be solved anymore. On the contrary it is the expectable result of the metaphor happening in the blend where partial structures from the relevant portions of the two domains actualised in the mental spaces give rise to new meaning. 2.2.3.6. CIT: an assessment

154


In paragraph 2 we established some points according to which the models examined should be evaluated: 1. The more general role of metaphor for the approach at stake. Metaphors are central to CIT because they define the arena in which CIT confronts itself with CMT and because they represent a very salient case of conceptual integration that tests the weaknesses and the strengths of the theory. 2. The level of description and explanation of metaphors. CIT is essentially a theory of the conditions for meaningful experiences. Its analyses mainly operate at a phenomenological level even when they call for a cognitive and neurobiological grounding of their basic mechanisms. However, CIT does not seem to tackle the issue of how meaning emerges in the first place, which CMT does instead. 3. The kind of conceptual and cognitive structures involved. Conceptual integrations are based on a detailed – and functionally defined – theory of how mental spaces are blended, their contents are selectively composed in a way to give rise to emerging structures that in return modify the whole space network. CIT is a theory of ongoing and dynamic cognition, of ever-evolving conceptual structures. There seems to be room for an experiential grounding of the results of conceptual integration through the role of pre-established conceptual structures (frames, etc.) and through the human scale implied by the optimisation principles, although a more explicit thematisation of that is due. 4. The methodological consequences for the analysis of metaphors. The analysis is thus aimed at showing how entrenched conceptual structures are put to play in order to generate novel and local meaning configurations and paying special attention to the role of the context in selecting and foregrounding relevant structures and information. 155


Both CMT and CIT are cognitive approaches by theoretical import. That is to say, they adopt a kind of informed common-sense perspective on the human mind – informed in that such approach takes into account some important developments in 20th century psychology (for example, gestalt principles) and common-sense in that it defines its concepts in terms that do not require any knowledge of psychology or neurobiology (cf. Stefanowitsch 2010b). This strategy has the advantage that it does not require a continuous and frantic update as more specific models of cognitive functions keep on changing and replacing each other. On the other hand this strategy often impedes cognitive linguistics from actually contributing to these developments109. CMT is principally a theory of meaning that focuses on the way linguistic (and nonlinguistic) superficial phenomena are motivated by an underlying conceptual level situated in the long term memory and ultimately grounded on directly meaningful bodily experience and experiential association. CMT aims at uncovering these conceptual structures (in an initial phase) and then at showing their neural plausibility (in a later phase). Thus, CMT gives us a lead on what pre-existing knowledge – embodied directly meaningful experiences – and what kind of motivation underlies most local metaphorical uses. CIT has a different focus: it does not make assumptions the origin of meaning, but it investigates how information is locally assembled and on the basis of that it constructively criticises the unidirectional and static dimensions of conceptual metaphors, showing how they are only a limit case. In doing so, CIT grants us a sophisticated system of mental spaces and mental operations that should account for how novel structures emerge and for how conceptual structures are potentially reshaped. CIT primitives are not directly meaningful experiential schemas, but cognitive operations – evolved approximately 50.000 years ago and defining,

109

The situation is slowly changing: cf. Gonzalez-Marquez, Mittelberg, Coulson and Spivey (eds) 2007, Casasanto 2009, Turner 2010, Stefanowitsch 2010a for excellent methodological directions.

156


amongst other things, human cognition (Fauconnier & Turner 2009). CIT displays thus a more dynamic nature that is open – but not yet fully developed in that direction (cf. Pascual 2008, Liddel 2003, Hutchins 2005, Williams 2005, Brandt 2010) – to a wider variety of semiotically mediated motivations through the inclusion of a context. This model shows a welcome focus on the creativity of the “surface” of linguistic behaviour that was lacking in CMT and it introduces a few critical issues. The first one is the concept of mental space itself. While introducing a welcome criterion of contextual relevance in the construction the local semantic scenario, it does not further elaborate either on the mental or on the spatial nature of mental spaces. Where is “mental” situated? Does it consist of purely internal representations? Does the context only intervene through priming the relevance of its content or does it participate more actively in its definition? Which kind of entities does it contain? Are they traditional symbolic representations in a classical cognitivist sense (§ 1.2.1.2) or are they more akin to operational mediators (§ 1.2.2)? An additional concern is the lack of any motivation behind the predilection for meaningful Gestalts at a human scale. How does that relate, for instance, to experiential realism as sketched by Lakoff and Johnson?

2.2.4.

Assessing the cognitive approaches to metaphors

Up to know CMT and CIT have been considered from an internal perspective: what they are trying to describe and explain and which inconsistencies emerge in this process. However, in Chapter 1 I have sketched – and argued for – the development and use of a pragmatist framework in order to articulate and understand the role of language in cognition. Following up on that it is time to actively engage in showing the risks of excessive reductionism and, relatedly, internalism at work in CMT and CIT, risks from which some of the inconsistencies highlighted derive. Cognition – we have seen – is not (only) what goes on in the head. Cognition happens (also) through 157


the reliance on body, environment, other agencies and socio-cultural norms at different time scales: philogenesis, ontogenesis, online cognition. Moreover, cognition is not as much a question of subsuming under a type, but a more flexible question of adaptive and distributed habits Which repercussions do these positions have on our understanding of metaphors? I will thus argue that i) the directly meaningful experiences on which CMT grounds meaning are not direct and immediate since they rely on heterogeneous mediating elements (cognitive abilities to abstract, joint attention, nurturing practices and cultural artefacts); ii) metaphorical projections can result in various degrees of blending and difference between the terms and the domains; iii) the public nature of language allows for diagrammatic manipulation, alignment and negotiation, processes that trigger conceptual evolution and change. 2.2.4.1. Mediating the directly meaningful

We have seen how CMT is in fact a theory of meaning. Meaning is defined as something meaningful to somebody and meaningfulness is simply stated as emerging directly from motor action and perception (cf. Lakoff & Johnson 1980:227). Abstract and social conceptual domains are given shape via concrete – that is, sensorimotor – ones and the metaphorical projection between them is described as originating in the individual’s “sensorimotor experiences so regularly they become neurally linked” (Lakoff and Johnson 1999: 555). To be fair, when Lakoff and Johnson describe this embodied grounding, they seem to point to a more nuanced conception: "there are at least three levels to what we are calling the embodiment of concepts: the neural level, phenomenological conscious experience and the cognitive unconscious" (Lakoff and Johnson 2002: 102). Let us take these levels into account one by one110. At a phenomenological level, embodiment is 110

I am loosely following here Zlatev's (2007) attentive work of reconstruction of embodiment in CMT.

158

Language and Cognition - Ph.D. Dissertation by Riccardo Fusaroli the way we schematise our own bodies and things we interact with daily (Lakoff and Johnson 2002: 36)

everything we can be aware of, especially our own mental states, our bodies, our environment and our physical and social interactions. This is the level at which we speak of the "feel" of experience (Lakoff and Johnson 2002: 103).

Phenomenological embodiment ends up swallowing up the whole domain of conscious experience. However, in Lakoff and Johnson's theory conscious experience does not amount to much, since the cognitive unconscious constitutes the realm of thought that is completely and irrevocably inaccessible to direct conscious introspection […] the 95 percent below the surface of conscious awareness [that] shapes and structures all conscious thought (Lakoff and Johnson 2002: 12-13)

Such liminarity of conscious experience is the reason for which Lakoff and Johnson claim that we must go beyond phenomenology to employ standard explanatory methods of linguistics, psychology, and neuroscience that allow us to probe structures within our unconscious thought processes (Johnson 2005: 21).

This unconscious level is thus at least partially constituted by the neural level – a level that "significantly determines [...] what concepts can be and what language can be" (Lakoff and Johnson 2002: 104), and “an embodied concept is a neural structure that is part of, or makes use of the sensorimotor system of our brains. Much of conceptual inference is, therefore, sensorimotor inference" (Lakoff and Johnson 2002: 20). However, this neural level is a higher-level generalisation that is heavily dependent on "an important metaphor to conceptualise neural structure in electronic terms" (Lakoff and Johnson 2002: 103). It does not involve physiological details like "ion channels and glial cells" (Lakoff and Johnson 2002: 103); on the contrary, it is based

159


on a computer simulation (§ 2.2.2.2.2)111. Nevertheless, when Lakoff develops his Neural Theory of Language the neural level overtakes it all: “The point is that, for me, CMT ceased to exist in 1997 when the neural theory of metaphor was formulated and integrated into work in turning cognitive linguistics in general into a general neural theory of language. NL – neural linguistics – is a detailed theory of how neural control, structure, and binding works in the brain, with special attention to all the details of language” (Lakoff, private communication).

In Chapter 1 (§ 1.2.1.3.4) I presented four main ways of considering embodiment which are as follows: i) material support; ii) sensorimotor ground of cognition; iii) bodily structures as actively involved in cognizing; iv) the phenomenological feel of being a body and its consequences for cognition. I have also added that when a fully distributed conception of cognition is in place, also the role of the body gets reshaped by the social and intersubjective environment. The embodiment described by CMT covers only a subset of these possibilities. In the initial formulations – and, to be fair, still in most of Johnson's solo works (cf. Johnson 2008) – the phenomenological dimension had a strong importance as well as had the pluralistic mediation of culture and history. However, in its latest developments CMT seems to have narrowed its main focus on embodiment as a shaping force in the neural system and the bodily structures in interaction with the world have a role only in the initial formation of sensorimotor neural structures. The body and the way it constrains and leads our experiences in context give rise to the schemas that determine our conceptual system. This situated embodiment – explicit in the moment of emergence of the directly meaningful primitives – disappears in the successive phase in which these interiorised conceptual schemas determine higher order cognitive processes. Anyway, the situatedness of the body does not seem to play 111

It has to be noted how in almost every single (rhetorical) use of the neural system to embody language and meaning there is no mention at all of any crucial components of it except for the neural circuitry. No mention at all of the role of hormones and blood vessels is to my knowledge ever made, although no serious neuroscientist would underplay the role of these factors.

160


a strong role in re-shaping conceptual systems during everyday cognitive processes if not through the "direct meaningfulness" of certain experiences that enable us to understand new conceptual formations. Not by chance, much work in CMT has been dedicated to i) the ontogenesis of conceptual structures, showing how the infant can interiorise the primary experiential structures that will then determine her way of thinking, talking and behaving; ii) how these structures are expressed in the widest possible range of language and conceptual domains; and iii) reduce them as much as possible to a possible neural level grounded on neural networks and on the notion of binding. However, our articulation of this grounding differed on at least one crucial point that opens the door to a radical re-assessment of the theory: experience is not direct, it has instead an intrinsic and pluralist dimension of mediation and it relies on a distributed notion of cognition. From this point follows the need to integrate a consideration of the elements and processes implicated in this mediation. As a starting point I will sketch here how this notion of mediation can be applied to the notion of imageschema. In § 2.3 we will see how it can be applied 2.2.4.1.1.

Image-schemas and generality

Marina Rakova, a Russian analytical philosopher, launched a campaign against CMT in 2002 – first on the pages of Cognitive Linguistics (Rakova 2002), then in a whole book (Rakova 2004). She attempts to show how embodied realism is not a credible philosophical path given its inconsistencies with established philosophical categories, in particular in its confused references to embodiment and to its role in shaping experience. While I do not embrace the spirit of her critiques and her reductive views on what cognition is, her reflection on the notion of image-schemas can help us highlight how mediation should be better thematised. According to Rakova a theory of meaning and concepts should fit in either a (caricatural) empiricist or in a (caricatural) rationalist perspective: Are image-schemas learned or innate? Embodied realism is hard to fit in in either. Certainly Women, Fire and 161


Dangerous Things leaves some space for innate concepts (Lakoff 1987: 165), but at the same time Lakoff and Johnson (2002) state clearly that image-schemas are learned universals. Thus Rakova starts inquiring: which mechanisms enable human infants to learn image schemas? Aren't those mechanisms innate? Which sequence does this learning follow? Which neural mechanisms is it grounded upon? Let us then take the example of the CONTAINER image-schema and follow up on this critique. According to Johnson's (1987) CONTAINER is naturally derived from the instances of putting something into oneself, of analysing the visual field into containers or of understanding oneself as being placed into a container. Such a process can hardly emerge without amongst other things: i) the notion of permanence of objects and matter; ii) a sense of agency; iii) the mastering of external reference points. These are all abilities that appear quite late in the cognitive development of the infant, thus undermining – Rakova argues – the plausibility of an empiricist genesis of image schemas and therefore of CMT, since embodied realism is supposed to be an extreme empiricism where “nothing is innate and everything is learned” (Rakova 2004:86). Rakova is both right and wrong. Right because the abstraction of image schemas from bodily interactions with the world does require the mediation of certain preexisting cognitive skills, as the developmental psychologist Jean Mandler (2004) has shown in a number of experimentally-backed articles112. Starting at an early age infants attend to objects and spatial displays in their environment. By attending closely to such spatial experiences, children are able to abstract across similar kinds of experiences, finding meaningful patterns in the process. For instance, the container image schema is more than simply a spatio-geometric representation. It is a ‘theory’ about a particular kind of configuration in which one entity is supported by another entity that contains it. Image schemas are simply the outcome of the way an infant’s immature input systems process the spatial structure that exists in the world (1992: 112

A lack of the theory that was debated in Bundgaard 1995 and Stjernfelt 2003.

162


592) through an individual’s ability to detect an inside, an outside, and a boundary between them (1994: 65). According to Mandler, this twist frees the infant from the need to manage very complex skills and requires only an innate mechanism of analysis, not innately given content (1992: 592). At the same time, Rakova is wrong in claiming that such a cognitive mediation is a deadly blow to embodied realism. One of the key elements of embodiment is precisely the attempt to go beyond the opposition between learned and innate. Already with connectionism (Elman et al. 1996) and even more so with situated approaches and dynamical system theory modelling the boundaries have been blurred. If already genetic inheritance to be expressed requires a mediation at multiple level – from cytoplasmatic environment to intrauterine experiences -, thus certainly still leading, but not determining development, the cognitive level shows an even greater plasticity, prompting Andy Clark to define human beings as "naturalborn cyborgs" (Clark 2004). "It is important to distinguish the mechanisms of innateness from the content of innateness" (Elman et al. 1996:359), and mechanisms can act at different levels, representations, architectures and timing. Experientialism can very well rely on a representational empiricism grounded – at least partially – on innate architectures and timing. As Mandler shows, there is evidence for innate attentional-perceptual mechanisms enabling infants to follow trajectories and paths and to discriminate between containment, occlusion, and support. This does not invalidate the fact that these mechanisms are used to learn schemas. Moreover, the innate architecture might be at least partially due to evolution, thus reintroducing a dimension of phylogenetic “blind learning”, that is, adaptation. As usual, the pragmatist insight that leads us to focus on the mediation occurring in cognitive processes does not solve the problem, but leads to further empirical questions: Which systems and which timing is put to play in the infant? Which predisposition enables infants to hypostatise interactions with the world in schemas? How is the validity and

163


the application of these schemas tested and re-shaped113? What is the role of the conceptual systems already at play in the culturally shaped environment and nurturing practices and, more explicitly, in language in structuring this learning process? 2.2.4.1.2.

Joint Attention and Culture in the formation of Image-Schemas

Introducing the mediating role of cognitive abilities in the emergence of imageschemas means to open up to a much wider set of possibilities than the simple presence of an innate or learned mechanism inside the head. It means to open up to the mediation of a potentially distributed cognitive system. Jean Mandler is aware of this possibility but chooses another focus otherwise: There is also little discussion in this book of parental influences on infant learning. This is due mainly to my conviction that the earliest conceptual learning and memory development are more a function of what infants observe and analyze than of what parents try to teach. It is possible that I underestimate parental influence in this regard, but in any case I hope readers will come to appreciate just how much infants can achieve on their own (Mandler 2004:ix).

However, much persuasive research has been done on the role that of intersubjectivity and nurturing practices play in the early cognitive activity of the infant (cf. Tomasello 1999, Zlatev 2005, Kimmel 2005, Zlatev et al. 2008; Trevarthen 2010, Violi 2010, De Jægher, Di Paolo, Gallagher 2010 for reviews of these issues). Trevarthen's research showed how pervasive the sensitivity to bodily movements, gazes and facial expressions (primary intersubjectivity) is as well as to interaction in pragmatic contexts (secondary intersubjectivity). Tomasello showed how constitutive of human interactions and of the ontogenesis of the human cognitive system joint attention is. If these intersubjective practices are so ubiquitous 113

Stjernfelt (20036), for instance, shows how the image-schema behind the conceptual metaphors for mental instability in Danish is a complicated thermodynamical schema whose physical details are still being worked out by scientific research. This does not invalidate the fact that infants can learn these image-schemas through interaction with the world, but it pushes us to further investigate the mechanisms of it and their interplay – or lack thereof – with our awareness.

164


and important, it is possible to hypothesise that they enter the process of selection and abstraction of image schemas from interactions that CMT implicitly postulated as individual. I have already mentioned the Sinha & De Lopez (2000) study on the role of nurturing practices and cultural artefacts in the formation and employment of the image schema of CONTAINER. According to Sinha and Jensen de Lopez (2000: 31) children employ social knowledge of the canonical use of objects in conjunction with their innate capacity for schematising spatial relations. The image-schematic nature of cultural objects may be a prototypical ecological affordance that influences language (Sinha and Jensen de López 2000: 22). Thus, Zapotec children are not as quick as Danish or English children to notice linguistic differences between the senses of “under” and “in” because they are not encouraged to play with upright cups and more generally because Zapotecs use a smaller variety of containers while tending to use them more multi-functionally. Widening our gaze, we can find more studies that highlight the shaping role of these intersubjective culturally shaped practices in the formation of image-schemas. Precursor of such investigations can be seen in Bourdieu (1977: 90-92) and Mauss (1973). Bourdieu’s ethnography of the Algerian Kabyles, while not introducing the notion of image schema itself yet, describes a complex system of gendered homologies in which postures, practices, and social space together define OUTWARD and UP schemas as male and INWARD and DOWN schemas as female. In general, Bourdieu's theory of embodied cultural knowledge – emblematised by the notion of habitus and concretely manifested in bodily posture and movement patterns – sees ritual and everyday activities as continuous structural exercises for particular schemas. On the same lines, Mauss produces a fascinating taxonomy of how culture shapes body postures and movement styles mostly within European populations. An additional case study can be found in Geurts' (2003) study of the Anlo-Ewe of 165


South Ghana. Geurts' study covers ethnographic, linguistic and developmental aspects of embodied engagement with specific focus on kinaesthetic and proprioceptive schemas. It turns out that BALANCE is a central and very nuanced schema. Cultivating proper balance is ubiquitous in Anlo-Ewe life. It is common to carry heavy loads on the head, therefore balance is crucial and nurtured in order for the individual to be able to move freely and naturally. However, achieving balance has not as its only goal the explicit learning of a specific transportation technique; someone's posture and gait also suggest something a person's moral fortitude and psychological disposition. Metaphors expressing this are linguistically more varied and elaborated than their Euro-American counterparts, an example of which is “having a backbone”. This is emphasised by rituals elaborating on the theme of head balancing of objects. Given the centrality of BALANCE, it is not surprising that special social practices ensure that children develop a proper sense of BALANCE: Infants often get their joints flexed to develop an awareness of graceful movement; and toddlers are continually encouraged to practice their balance. Through Geurts' detailed analysis and Kimmel’s (2002) re-reading of it, it is possible to observe how Anlo-Ewe BALANCE is not a general image schema, but strongly situated and modulated in contexts ranging from mundane ones such as the fetching of water to highly elaborate ones such as ritual. A detailed ethnographical investigation turns out to produce a vision quite different from the abstracted image schema of BALANCE that children all over the world acquire in a roughly similar way proposed by Lakoff and Johnson. The balance ethos of the Anlo-Ewe brings to the fore how an image schema like BALANCE may be culturally refined. Finally, acquisition of image-schemas can be seen as mediated through language itself (Bowerman 1996; Zlatev 1997). For example, probably due to linguistic marking, Yucatec Maya pay more attention to what something is made of in categorising, while English speakers pay attention to shape (Lucy 1996: 49-52). At a different level of analysis, Dewell (2005) argues that image-schemas cannot be 166


considered purely pre-verbal, because only through language do we develop a tendency to favour maximally precise and differentiated linear shapes that can be explicitly profiled and publicly accessed from a flexible perspective. This all too brief overview on how the notion of image-schemas as emblematic of basic conceptual or pre-conceptual structures has to consider a relevant socio-cultural and intersubjective horizon. The productivity of such a re-assessment of the notion does not wish to diminish the importance of physical and biological constraints in their emergence. It would like instead to show how they are deployed and interact with a plurality of mediating instances. Without explicitly taking this heterogeneous and yet constitutive complex mediation into account, an undue reductionism seems unavoidable. Investigating the neural bases of conceptual metaphors and binding (§ 2.2.2.2.2 and § 2.2.3.1.2) is a valuable enterprise, but only if it takes into account the rich experiential environment in which neural structures take shape, with which they interact and on which they rely. After all, in the text that introduced the notion of image-schemas social, historical and intersubjective dimensions were present: “image-schemata are related to sensorimotor patterns, but these patterns emerge in the blending of culture, language, history and bodily mechanisms” (Johnson 1987)114. Unfortunately this blending was not systematically explored (but cf. Hampe 2005, Kimmel 2005, 2010). However, the reflection on conceptual integration – less focused on the sensorimotor grounding of meaning and experience – shows how even the more basic domains seem to be integrated with other domains and potentially reshaped. Social, emotional, intersubjective and cultural dimensions intrude on the sensorimotor domain, potentially enriching the experiential basis of meaning, shaping the value and the aspects according to which image-schemas are abstracted. Moreover, if the traditional focus in cognitive linguistics on image schemas as 114

Or again “what we call 'direct physical experience’ is never merely a matter of having a body of a certain sort; rather, every experience takes place within a vast background of cultural presuppositions” (Lakoff and Johnson 1980:57).

167


abstractions emerging from bodily experience sometimes fools people into thinking that they function without much engagement from the rest of the body, we should take into account the possibility of a more distributed nature of image schemas, not only at an ontogenetic scale but also in online cognition. Image-schemas can be conceived as simulations of partially abstracted bodily actions that happen on the fly during cognitive activity and not as encoded structures in the head that are passively activated as part of unconscious linguistic processes of understanding. Linguists and psychologists be cautious in making concrete claims about how and where image schemas might be mentally represented. It is even possible that image schemas are not specific properties of the mind but reflect experiential gestalts that never get encoded as explicit mental representations. A different possibility is that image schemas might be characterized as emergent properties of our ordinary conceptual systems and therefore are not explicitly represented in any specific part of the mind. (Gibbs and Colston 1995: 370).

In the pragmatist framework this is equivalent to ascribing a more active role of environmental elements, socio-cultural norms and other agencies in the online constitution and deployment of recurring image schemas. A particularly promising domain for such investigations seems to be improvisational dances like tango where learning processes seem to consist – at least in part – in acquiring the ability to create on-the-fly shared image-schemas with the partner (cf. Kimmel 2010 for some examples and Kirsh 2010 for additional analyses of the mechanisms at play and of the role of bodily representation, as well as § 3.1.3). 2.2.4.1.3.

The Structure of Abstract Domains

If this extremely promising line of investigations is valid for basic conceptual structures, it is not that far out to hypothesise that this must be even more valid for less basic notions like abstract domains and metaphorical projection. Abstract domains like “love” or “life” or “ideas” or “theories” are delimited and supported by language and cultural practices in important ways, since language and cultural practices dynamically interact with the sensorimotor or otherwise embodied 168


structures metaphorically projected into them. It is not possible to construct basic metaphors like IDEAS ARE OBJECTS on the basis sensorimotor actions if we do not already have a notion of “idea” – most likely made possible by language. Already in a paper from 1996 Murphy highlighted how two different versions of metaphorical concepts could be implied in CMT: a strong one and a weak one. In the strong version representations of target concepts are mediated in that speakers are able to refer to and grasp such abstract concepts only via reference to the relevant source domain. Abstract concepts are represented entirely as a set of mappings from the representational structure of more concrete concepts (Murphy 1996).

This version – that anticipates conflation of a couple of years – makes it hard to tease source and target apart as well as to distinguish between different conceptual metaphors with the same target. If the connection between target and source is a determining one, then how comes LOVE can be determined both by LOVE IS A JOURNEY

and

by

LOVE

IS

MAGIC

(both

listed

by

Lakoff:

http://cogsci.berkeley.edu/lakoff/targets/Love.html)? The weak version of metaphorical concepts involves an independent representation of source and target. Therefore, the motivation behind the mapping is called into question: is it due to experiential grounding (with all the problems that have already been reported) or to linguistic usage? If the domains pre-exist the mapping, then the mapping itself can be due to a manifold of reasons. Indeed, many cognitively oriented corpus linguists (cf. Kintsch & Mangalath 2010, Louwerse 2010, Stefanowitsch 2010a) by mapping the frequency and structure of lexical cooccurrences attempt to reconstruct the different kinds of information there encoded, adding social and linguistic experience to embodied grounding. This is done by investigating how much of our implicit understanding of words is due to the words usually associated with them, even in the absence of an embodied grounding.

169


Such a role of language in the construction of embodied domains would also explain the stability of the cultural variations of conceptual metaphors through time. Children do not have to make up their own conceptual metaphors from scratch, nor do they have to figure out their embodied grounding in details since they are exposed to language and other expressions of existing conceptual metaphors. Not only can both children and adults alike via language participate to a wider experiential grounding than they would be able to on their own with their bodies, but language and symbols in general open up the possibility to develop conceptual systems far beyond bodily experience. Think of quantum mechanics and relativity theory and of the challenges they pose to our understanding of the world because of the extreme distance of the notions involved and of their structure from bodily experience. Not only we need to acknowledge a certain independence of abstract conceptual structures from embodiment, but also we have to acknowledge that they are crucial to experience, since abstract structures structure it and serve as norms of reference for it. Adamson gives us some examples of this: “With an erroneous sense of justice, we learn, one can make serious errors in judgment and cause great harm. Thus we need norms that are true (an idea which itself proves elusive to reflection) and which work well. And, when we realize that some norms already influence our thinking, we are compelled at times to examine our norms carefully. But, as any reader of Plato’s dialogues can attest, approaching the ideas of justice, piety, friendship, or love directly proves extremely difficult! Secondly, abstract ideas can approach extreme generality, thus transcending all concrete meanings. Such generality promises a “universal” standard by which we can potentially understand and assess many specific situations. However, we struggle to say what these ideas are – what they mean in themselves – and our thinking easily loses its moorings at this level of abstraction. It is only through later reflection or analysis that we realize that these familiar words are difficult to specify and define. Once we attend to them directly, they seem vague, extremely remote from concrete experience, and elusive. (Recall Augustine on time: 'What then is time? If no one asks me, I know what it is. If I wish to explain it to him who asks, I do not know')” (Adamson 2007: 9).

Abstract domains deserve a less derivative treatment than it is often given. Our 170


accounts of concrete forms and their entailments cannot exhaust our account of cognition. Certainly, it is through these sensorimotor schemas or patterns that we can grasp much of the abstract domains; however, there are also other dynamics are at play, specifying and re-shaping the modalities of such more embodied grounding. Our understanding of the world crucially depends the way language makes available to us social experiences beyond our immediate environment as well as abstract concepts. I will show examples of this in the analysis of the evolution of “Berlusconi is a caiman” (§ 2.3) and in the analysis of online engagement and linguistic alignment (§ 3.2). 2.2.4.1.4.

Cultural Metaphorical Projection

We have seen how, according to CMT, abstract domains are determined – or at least shaped – by an initial process of conflation: “Early conflations in everyday experience should lead to the automatic formation of hundreds of primary metaphors that pair subjective experience and judgment with sensorimotor experience” (Lakoff & Johnson 1999: 49). Such conflations constitute more complex conceptual metaphors as well, since they are “built out of primary metaphors” (Lakoff & Johnson 1999: 60). It is only at a second stage that concrete and abstract domains are separated conceptually (Lakoff & Johnson 1999:46). However, our everyday cognition continues to depend on primary metaphors and on various extensions of them, mostly in the form of unconscious, neural activation. Instead of focusing on the pervasive repetition of pre-established patterns of embodied meaning CIT focuses on general mental operations, which actively construct patterns and “spaces” of meaning. Such highly flexible mechanisms with their capacities for compressing, expanding, selecting, integrating, unpacking, scaling and completing semantic domains definitely escape the potentially reductionist perspective brought in by conflation. The mental spaces at stake are constituted and put to play through dynamics in which fusion, creation and difference co-exist opening up for the possibility of interesting cultural, genre-related and textual 171


variations in the nature of the metaphorical projection itself. Let us take into consideration again the conceptual metaphor PEOPLE ARE ANIMALS to highlight this. As we have previously argued, such a conceptual formation is pervasive cross-culturally and across genre and cannot be reduced to a simple conflation. Indeed, the animal traits projected onto human beings seem to be anthropomorphic already. “Achilles is a lion” individuates a trait of “courage” that is not intrinsic to the animal domain. Grady (1997) redefines the problem in terms of resemblance, that is, in terms of culturally perceived similarity. Achilles and the lion are not similar in terms of their physical or behavioural aspects but in terms of our perception of them as sharing a certain disposition to act in an impulsive way without fear. However, this perception must be socially and linguistically mediated, since lions are not commonly experienced, at least in Western societies, and even when they are – in circuses and zoos – they are rarely engaged in courageous acts. This commistion between human and animal can be even more problematic. Let us take The Metamorphosis by Kafka. Gregor Samsa is transformed into a cockroach. Is this a literal or a metaphorical transformation or maybe both? It depends on how we decide to read the story. How about the case of the anthropomorphic and zoomorphic animals in the imaginary worlds by Walt Disney? Considering Mickey Mouse, Goofy and Pluto we can find a graded continuity in accordance to which Mickey is rather human, Goofy less so, though he is certainly granted speech and reason he displays dependency and the nature of a (dog) companion. Pluto is definitely a dog, but with more personality and human traits than real world dogs. Such confusion as regards the terms (vehicle and tenor, source and target) and spaces is not confined to fairy tales and comic books. Rane Willerslev reports analogous experiences from his fieldwork amongst the Siberian Yukaghirs: Watching Old Spiridon rocking his body back and forth, I was puzzled whether the figure I saw before me was man or elk. The elk-hide coat worn with its hair outward, the headgear with its

172

Language and Cognition - Ph.D. Dissertation by Riccardo Fusaroli characteristic protruding ears, and the skis covered with an elk’s smooth leg skins, so as to sound like the animal when moving in snow, made him an elk; yet the lower part of his face below the hat, with its human eyes, nose, and mouth, along with the loaded rifle in his hands, made him a man. Thus, it was not that Spiridon had stopped being human. Rather, he had a liminal quality: he was not an elk, and yet he was also not not-an-elk. He was occupying a strange place in between human and nonhuman identities. A female elk appeared from among the willow bushes with her offspring. At first the animals stood still, the mother lifting and lowering her huge head in bewilderment, unable to solve the puzzle in front of her. But as Spiridon moved closer, she was captured by his mimetic performance, suspended her disbelief, and started walking straight toward him with the calf trotting behind her. At that point he lifted his gun and shot them both dead. Later he explained the incident: “I saw two persons dancing toward me. The mother was a beautiful young woman and while singing, she said: ‘Honored friend. Come and I’ll take you by the arm and lead you to our home.’ At that point I killed them both. Had I gone with her, I myself would have died. She would have killed me.” (Willerslev 2007)

This peculiar form of confusion between man and animal, due to their closeness in nature, is typical of certain strands of animism. In another basic spiritual stance, totemism, the relation is more stereotypically metaphorical: animals are used as categories for social distinction and social behaviour, often showing the blending between human and animal traits I have already highlighted: “[This is the] Achuar's conception of the toucans' domestic life. They say that these birds form lasting couples and that there is a strong bond between the male and the female. When one member of the couple is killed, the other cries out heartrendingly for several days and then hastens to find a new mate. Such behaviour, far from being stigmatised is regarded as entirely legitimate, for widowhood must not drag on, since man can fulfill himself only within marriage. […] The highly socialised behaviour imputed to the toucan is also attributed to the woolly monkey, which is supposed to observe scrupulously the rules prescribed for contracting a marriage. […] This behaviour contrasts with that of other monkeys like the howler who has what seems to be a frantic sex life, coupling unhesitatingly with mother or sister.” (Descola 1994:96).

And we could easily go on. Drawing inspiration from the work of Giovanni Bottiroli (1993) on literary metaphors, I introduce a graded continuum of distances between the two terms and domains at stake in metaphors: i) metaphors that connect two 173


terms without causing any lasting redefinition of their identity; ii) metaphors that connect two terms defining a similarity between them, redefining – but not confusing – their identity; iii) metaphors that create a total overlapping between two terms, with a complete destruction of identity boundaries115. If conceptual metaphors shape our understanding of and interaction with the world, they play with the distance between domains. Calling Berlusconi a caiman may not confuse the identity of the two terms, but it certainly changes Berlusconi's identity and reshapes the way he is represented and the surrounding discourse. As a brief example we can mention here how recently a legal clash between Berlusconi and Rupert Murdoch has been characterised as a struggle between a caiman and a shark116 and how more generally the latest moves of Berlusconi are interpreted as caiman-like from the beginning, as caused by caiman nature. To discuss the distance articulated by a metaphor is not just a taxonomic exercise since as we will see it can influence the life of the metaphor deeply, and it can be played with. 2.2.4.1.5.

Reintroducing the role of language in conceptual ontogenesis

It thus possible to reformulate the way in which from directly meaningful bodily interactions with the world basic conceptual structures emerge and end up constituting our conceptual knowledge and determining our linguistic usage by metaphorical projections. This re-introduction of language and cultural practices in conceptual formation is not 115

116

In another work (Fusaroli 2009) I investigated these mechanisms more thoroughly. Literary metaphors often follow the second regime, but cases in which metaphors more radically escape the realm of rhetoric and start modifying (the narrative) reality (third regime) are not uncommon. The "Orpheus and Eurydice" metaphor in The Ground Beneath her Feet by Salman Rushdie, or the "angel and demon" metaphor in The Satanic Verses by the same author are good examples of this. These metaphors – born in the words of the narrator – take possession of the behaviour and even of the appearance of the characters. In other cases there can be metaphors so strictly interrelated with the narration that their metaphorical nature becomes dubious: "Three metamorphoses of the spirit have I designated to you: how the spirit became a camel, the camel a lion, and the lion at last a child" (Thus Spoke Zarathustra, Friederich Nietzsche). “L'approccio del caimano è inizialmente stato amichevole, dopo i primi morsi dello squalo [...]”; “La resa dei conti tra il caimano e lo squalo sta per arrivare” Loris Mazzetti sul Fatto del 19 aprile 2010.

174


an attempt to sneakily introduce a radical cultural relativism into the very foundations of cognitive linguistics. It is instead an attempt to fully recognise the complex interplay of physical, biological, intersubjective and socio-cultural constraints that shapes and constitutes human cognition and meaning. The interplay of the body with the physical world presents certain affordances that are manipulated, constrained and modified by the engagement with other agencies and with socio-cultural norms (§ 1.2.2 & 1.4). 2.2.4.2. Reintroducing language use as cognition

However, this re-assessment of the cognitive linguistic approach to metaphor is not fully accomplished. If we have seen the pervasive role of language and cultural practices in shaping the grounding of the human conceptual system, we still have not seen language at work as an integral part of cognitive processes. Reducing the whole of cognition to an ideally pure and individual-centered sensorimotor domain, however, is not the only risk to which CMT could lead us. In chapter 1 (§ 1.2.3) I criticised the distinction between internal and external when debating the nature of cognition. Cognition happens – in many cases – as a distributed and integrated process. Mind and experience participates in it, therefore crossing the border between internal and external. CMT and CIT, however, like many works in cognitive linguistics, lean towards a strong internalist position. In particular, CMT’s and CIT’s attempts at neurally grounding their models seem to lead towards locating conceptual metaphors and integrations at the level of the single individual. Language and linguistic metaphors in particular seem to have the role of “simply” transmitting enough information to trigger internal “computation” so that subjects separately can reach the same or an analogous “mental state”117 separately. 117

There are a few exceptions, however. Gibbs (1997) asks us to bring metaphors out into the culture. Kimmel (2010) engages image-schemas at a cultural and intersubjective level. Stefanowitsch (1997, 2010) and Louwerse (2010) argue for the importance of linguistic cooccurrences in the online engagement and stabilisation of conceptual formation, including conceptual metaphors. Oakley and Coulson (2008) show the relevance of a dialogic unfolding of

175


In addition to the many citations in chapter 1 and in the previous paragraphs I can add a couple here, just to strengthen the point: In modern linguistics, word meanings are held to be conceptual entities, which is to say, they are held to constitute mental units, paired with phonetically-realisable forms, and stored in long-term memory. The repository of such form-meaning pairings forms a structured inventory which is commonly referred to as the mental lexicon (Evans 2010:1).

It is unanimously agreed in Cognitive Linguistics that meaning does not reside in linguistic units but is constructed in the minds of the language users. For the listener this means that he takes linguistic units as prompts and constructs from them a meaningful conceptual representation. In fact, this principle of meaning construction is not confined to language. Every transformation of a sensory stimulus into a mental representation is an instance of meaning construction (Radden et al 2007: 1).

When we start considering language in its intersubjective and social dimensions seriously (§ 1.2.3) these assumptions on the individual nature of mental and of cognition are not satisfactory anymore. In particular, we have to start considering the role of diagrammatic manipulation, alignment and negotiation that are afforded by the symbolic nature of language. Remember the functions of language according to Clark, Bermudez and Tylén: to augment memory, to simplify the environment, to coordinate activities through control of attention and resource allocation, to establish control loops and to represent and manipulate data. All this in a potentially public, therefore intersubjective and social way. Let us take it step by step. First I will introduce the notion of diagrammatic reasoning via the analysis of “Man is a reed, the weakest of beings, but a thinking reed” by Blaise Pascal, then I will export it to a textual and intersubjective arena via a more complex analysis of “Berlusconi is a caiman” in the “La Repubblica” corpus. 2.2.4.2.1.

Exploring the metaphor from a linguistic perspective

We have seen at work the conceptual formation of THE GREAT CHAIN OF conceptual integrations.

176


BEING that supposedly underlies PEOPLE ARE ANIMALS. It is certainly a pervasive formation and it would be easy to content oneself with interpreting the following passage simply as an instance of it: [§] L'homme n'est qu'un roseau le plus faible de la nature; mais c'est un roseau pensant. Il ne faut pas que l'univers entier s'arme pour l'écraser. Une vapeur, une goutte d'eau suffit pour le tuer. Mais quand l'univers l'écraserait, l'homme serait encore plus noble que ce qui le tue; parce qu'il sait qu'il meurt; et l'avantage que l'univers a sur lui, l'univers n'en sait rien. Ainsi toute notre dignité consiste dans la pensée. C'est de là qu'il faut nous relever, non de l'espace et de la durée. Travaillons donc à bien penser. Voilà le principe de la morale. [§] Il est dangereux de trop faire voir à l'homme combien il est égal aux bêtes, sans lui montrer sa grandeur. Il est encore dangereux de lui faire voir sa grandeur sans sa bassesse. Il est encore plus dangereux de lui laisser ignorer l'un et l'autre. [175] Mais il est très avantageux de lui représenter l'un et l'autre. (Pascal 1670: XXIII:174-175)118

However, such a reduction would betray the effort and the intention of Blaise Pascal. Pascal is employing language to fixate his thought and to develop it. There is certainly the projection from a lower being, a reed, to a higher one, man. However, the metaphor is more elaborated than that and it has a temporal development. First of all there are genre-related and co-textually built semantic potentialities, “isotopies” Greimas (1968) would call them. These lines unfold philosophical reflections by a Jansenist of the 17th century and they appear in a chapter called “greatness of man”. Thus, the cloud of potential traits to which "man" gives access is constrained by sedimented philosophical discussions that have been evoked in the course of the text and of the book and that are also easily available to the philosophical reader: the notion of "man" as a thinking being, as a moral being, and so on. When the 118

Curiously, the official English translation tells quite a different story: “Man is but a reed, the most feeble thing in nature; but he is a thinking reed. The entire universe need not arm itself to crush him. A vapour, a drop of water suffices to kill him. But, if the universe were to crush him, man would still be more noble than that which killed him, because he knows that he dies and the advantage which the universe has over him; the universe knows nothing of this. All our dignity consists, then, in thought. By it we must elevate ourselves, and not by space and time which we cannot fill. Let us endeavour, then, to think well; this is the principle of morality. A thinking reed. -- It is not from space that I must seek my dignity, but from the government of my thought. I shall have no more if I possess worlds. By space the universe encompasses and swallows me up like an atom; by thought I comprehend the world. (Pascal, Thoughts, VI:347-348)

177


“roseau”, the “reed” enters the scene many common features could be highlighted – the being singular, the being a being –and many mappings could happen – the river in which the reed lives could be the flow of ever-changing phenomena and experiences that constitutes the environment of the human subject or the man could be green. This is why Pascal does not stop here, but goes on in elaborating and specifying the metaphor: “the weakest of beings”. Human beings are mortal. Reeds are fragile. The semantic traits at work are reconfigured. Fragility is mapped onto the human domain, making mortality a foreground trait. We can spot an image-schema at work here that enables the mapping: the experience of breaking, of a force dynamics schema in which the Antagonist is far stronger than the Agonist. Human beings are fragile beings, exposed to the forces of nature, to diseases, violence and death. The image-schema is basic and motivates our understanding, but we should not forget that it is important to analyse how it is made pertinent by the ongoing process linguistic elaboration. Again, the metaphor is further elaborated: “But a thinking reed”. In the traditional exegesis of this metaphor this “thinking” is in contrast with the reed. The fragility of the reed is ennobled by this mental capacity. The situation of man, so frail and yet capable of thought that enables him to rise beyond nature, is touching. And the salience of this reversal is enhanced by the potential reversal of the force-dynamics schema involved: by thinking the Agonist becomes free, under certain respects, from nature, from the Antagonist. Pascal shows us how: i) via language it is possible and relatively easy to elaborate on and extend existing conceptual metaphors. Conceptual metaphors can be seen as habits of “perception” of the world that have to be enacted locally. Pascal sets the conditions for man being a reed. He sketches the relevant trait – feebleness –, makes it salient by exemplifying its consequences – “a vapour, a drop of water suffices to kill him”. Then, on this ground, he draws a distinction through adding the following “but a thinking reed”, showing us how ii) metaphors – at least in a distinctive regime (§2.2.4.14) – also contain the germ of differences between the terms at play. 178


This mechanism is not only available for the explicit treatment of single metaphors. "Juliet is the sun", one of the most quoted metaphors, widely relies on this. Single analysts have pointed a wide variety of interpretations as the one that comes out naturally from the metaphor: Juliet has been described as peerless, nurturing, warm, radiant, the centre of Romeo's world, the object of Romeo's admiration, beautiful, life-giving and so on. An attentive analysis of the play shows how Shakespeare sets up a careful background for the interpretation of this metaphor. Romeo is portrayed as escaping the sun, creating "artificial nights", then persuaded by Benvolio to go to a party. He carries a torch, a figure of light, soon replaced by chandeliers, in an increase of luminosity that reaches its peak when, spotting Juliet, Romeo describes her as teaching the torches to burn bright, as a jewel on the cheeks of the night, as a snowy dove in a flock of dark ravens. We have thus a pre-selection of features, an escalation of light that overcomes resistance, that will make Romeo’s artificial nights vanish and that has no peers. Furthermore, these traits form the common ground through which to interpret the complex figure of Juliet being the sun and the maid of the moon and having at the same time stars in her eye-sockets. In Juliet all figures of light blend in an attempt to explain brightness beyond description. If we keep on reading the play, new manipulation of these metaphors are made salient. The death of Romeo makes it possible to interpret the sun as something that is too risky to get close to and the stars in Juliet’s eye-sockets reverberate of this danger, becoming a reminder the inauspicious stars that, according to Shakespeare, doomed the two young lovers. Thus Shakespeare explores – admittedly in a poetic and not always coherent way – a configuration of metaphors with a common theme and displays their consequences narratively. 2.2.4.2.2.

Diagrammatic Reasoning

The linguistic deployment of metaphors looks more and more like what Peirce would have called a case of diagrammatic reasoning and since this notion is going to accompany the next analyses, it is time to unfold it. The notion of diagrammatic 179


reasoning dwells on the notion of iconism and becomes the “centerpiece of a Peircean epistemology” like the one we have tried to sketch in these pages. We have already dealt with the issue of iconism (§ 1.2.2.1), so it will suffice here to repeat that a great distinguishing property of the icon is that by the direct observation of it other truths concerning its object can be discovered than those which suffice to determine its construction. (...) Given a conventional or other general sign of an object, to deduce any other truth than that which it explicitly signifies, it is necessary, in all cases, to replace that sign by an icon. This capacity of revealing unexpected truth is precisely (...) the iconic character (...).(Peirce 1931-58 CP 2.279).

In this operational conception of iconism, diagrams are that particular kind of sign that expresses a skeleton-like sketch of relations and that “is aided to be so by conventions.”(Peirce 1931-58 CP 4.418). But a diagram is not simply a sketch – it needs to be manipulable: “As soon as an icon is contemplated as a whole consisting of interrelated parts whose relations are subject to experimental change, we are operating on a diagram” (Stjernfelt 2000). Indeed, By diagrammatic reasoning, I mean reasoning which constructs a diagram according to a precept expressed in general terms, performs experiments upon this diagram, notes their results, assures itself that similar experiments performed upon any diagram constructed according to the same precept would have the same results, and expresses this in general terms. (Peirce 1976 NEM 4:47-48)

When Peirce began to develop his notion of diagrammatic reasoning he was arguing that mathematics is an experimental science: theorems are often demonstrated through the use of points, lines and sketches in the form of an ideal – or diagrammatic – experimentation. Therefore, he stresses the general validity of the results to be achieved. However, we can settle for a less demanding definition in which diagrammatic reasoning constructs sketches of relations – prototypically via their material representation – that can be manipulated and thus afford new insights and hypotheses to be tested further. All this happens through the mediation of “conventions”, that is, skills and conventions are needed to structure the 180


construction, the manipulation and the interpretation of the diagram. Not everybody can demonstrate the Pythagorean theorem graphically. In the same way, Pascal had to know French and how to write and a good deal of philosophy to sketch his metaphorical diagram of the human nature. Peirce readily admits that such a use of the word “diagram” as presented above employs it in “a wider sense than is usual” (Peirce 1976 NEM IV, 315), but as Stjernfelt notes precisely this is the great advantage of this diagram concept: a whole series of semiotic processes – the tropisms studied by biosemiotics, the contemplation of pictures, metaphorical, analogical, and poetical reasoning, linguistic and narratological syntax, basic sensorymotor schemata, as well as mathematics proper – become understandable as different realizations of one and the same basic rational semiotic behaviour, namely, diagram experimentation. (Stjernfelt 2000)

Before applying the notion of diagrammatical thinking to metaphors and language more specifically let us delve a bit more into the details of how Peirce exemplifies his notion of diagrammatic thinking and extends it beyond mathematics through such a delicious narrative that I will be forgiven if I report it all: “But why do that [build a System of diagrammatization], when the thought itself is present to us?” Such, substantially, has been the interrogative objection raised by more than one or two superior intelligences, among whom I single out an eminent and glorious General. Recluse that I am, I was not ready with the counter-question, which should have run, "General, you make use of maps during a campaign, I believe. But why should you do so, when the country they represent is right there?" Thereupon, had he replied that he found details in the maps that were so far from being "right there," that they were within the enemy's lines, I ought to have pressed the question, "Am I right, then, in understanding that, if you were thoroughly and perfectly familiar with the country (...), no map of it would then be of the smallest use to you in laying out your detailed plans?" To that he could only have rejoined, "No, I do not say that, since I might probably desire the maps to stick pins into, so as to mark each anticipated day's change in the situations of the two armies." To that again, my sur-rejoinder should have been, "Well, General, that precisely corresponds to the advantages of a diagram of the course of a discussion. Indeed, just there, where you have so clearly pointed it out, lies the advantage

181

Language and Cognition - Ph.D. Dissertation by Riccardo Fusaroli of diagrams in general”. “Namely, if I may try to state the matter after you, one can make exact experiments upon uniform diagrams; and when one does so, one must keep a bright lookout for unintended and unexpected changes thereby brought about in the relations of different significant parts of the diagram to one another. Such operations upon diagrams, whether external or imaginary, take the place of the experiments upon real things that one performs in chemical and physical research. Chemists have ere now, I need not say, described experimentation as the putting of questions to Nature. Just so, experiments upon diagrams are questions put to the Nature of the relations concerned." (Peirce 1931-58 CP 4.530).

We have now all the elements we need in order to frame diagrammatic reasoning in the conception of distributed cognition that I am developing. Michael May (1999) has represented diagrammatic reasoning in the following way:

Figure 9 - Diagrammatic Reasoning

While this diagram usefully emphasises the iterated constructions and manipulations and the heuristic purpose of such a diagrammatic endeavour, it under-represents a few elements: the skilful engagement of the agents involved, that is, the habits at play situating such operation in specific practices and styles. To strengthen the case let us just think about the skills and instruments needed to draw a map as well as about the difficulty in assessing reciprocal visibility of two armies on a military map – unless we are very familiar with altimetric lines. Given such a reliance on skills, diagrammatic 182


reasoning is practice-based: different maps and charts will be drawn for different purposes as well as manipulated and interpreted in different ways. Diagrammatic reasoning can thus be defined as a paradigmatic case of distributed cognition. The cognitive process is constituted via the integration of various individual skills, socio-cultural norms (conventions in drawing a map) and external artefacts through a process developed in time. The need of having rich internal models of the situation depicted in the map is weakened because various aspects are discharged from the agent and delegated to the external representation, something which offers both a proximal perceptual environment and a manipulative one that contains all the resources needed to effectively perform the creative task of finding new properties of a certain portion of territory. This happens in a dynamic coupling between such an external representation and cognitive skills, conjoining partial external manipulations and partial reliance on working and long term memory as mediated and constituted by the situation and by the skills at work. A whole new set of cognitive processes is possible where high order cognition and sensorimotor skills are inextricably and continuously intermingled. To tie this up with the previously introduced pragmatist categories I can say that as any other sign a diagram constitutes a particular kind of mediating function that emphasises its gnoseologic power, its telling us something more. In such a mediating function we see Thirdness at work both in the habits that support, lead and constitute diagrammatic reasoning and in the more general value of the results of this process, Secondness in the resistance that the diagram has towards certain manipulations and interpretations as well as in its referring to what it is representing and Firstness in its heuristic, that is, iconic power. Furthermore, diagrammatic reasoning perfectly resonates with the pluralist realist stance of pragmatism since it allows us to create hypotheses and to try them out, modifying our representations and concepts without crystallising them, passing from one set of relations to another and opening new possibilities of meaning and inquiry. 183


Diagrammatic Reasoning in Language

This said, how does diagrammatic reasoning apply to language and metaphorical structures in particular? Let us add to Pascal and Shakespeare the anonymous author of the Song of Solomon: “Thy teeth are like a flock of sheep119 that are even shorn, which came up from the washing; whereof every one bear twins, and none is barren among them” (Song of Solomon, 4:2).

One could easily argue that a diagrammatic analysis – be it conscious or not – precedes any metaphor, since metaphors consist in the recognition of a diagrammatic schema in one phenomenon which may be used in understanding another (cf. Stjernfelt 2000). However, in the Song of Solomon – as in Pascal’s unfolding of “man is a reed” – we can see the diagrammatic process unfold in time and rely on successive linguistic elaborations of the metaphor. The comparison between teeth and sheep is not so obvious. The poet elaborates on that through the following “that are even shorn”. The traits of regularity and lack of distinction to which “even shorn” gives access resonate with the word “flock”. Encouraged perhaps by this neat echo, the poet goes on: “which came up from the washing”. The effect is spoiled for us and it is hard not to conjure an image of “shaggy, dripping creatures (bleating and smelly as well!)” (Eco 1983:100). However, Eco acutely observes that if we situate the poem in its original cultural context: the biblical poet drops all those properties of sheep negatively identified above, so as to preserve only the characteristic of their aequalitas numerosa, their splendid unity in variety – as well as their whiteness. It is understood that the poet is able to do so because within his culture these most probably were the properties associated with sheep, at least within the poetic tradition. And it is also clear that the qualities chosen to define the beauty of a healthy and sturdy country girl, destined to tend the flocks among the rocky Palestinian hills, single out her upright solidity, her unbroken state of perfection (Eco 1983:101).

119

We are technically speaking in front of a simile and not a metaphor. However, this distinction does not have conceivable consequences in the following analysis and will be therefore dropped.

184


So here we can see the interpretive habits at play. The diagrammatic unfolding of a metaphor – be it expressed linguistically or only in our reasoning and understanding – constitutively dwells on and plays with such interpretive habits, with how we are used to perceive the world. Furthermore, the expansion of the metaphor leads and supports the habits indicated by Eco, adding to the initial diagram “Thy teeth are like a flock of sheep” a more detailed depiction of which aspects of the sheep are involved. However, while Eco's analysis stops here, the metaphor is not over: “whereof every one bear twins, and none is barren among them”. What does the fertility of sheep have to do with teeth? Certainly we can find the repetition of the semantic trait “regular” in “every one” and a remodulation of it in the symmetry of geminal pregnancies. But still the pregnancy of the sheep is not explained. To do so we need to take a wider look at the poem and see how a “skeleton-like sketch of relations” is disseminated through it and to re-assess the metaphor in the more holistic semantic regime set in the poem in order to better understand the way differences and projections are articulated. The whole poem is disseminated with comparisons between the lovers or parts of them and animals, plants and inanimate objects. The girl in the poem is compared to “a company of horses” (1:9) and her breasts are like “two young does that are twins” (7:3) if we stay in the animal domain, and her stature is compared to that of a palm tree (7:8), her breasts to a “cluster of the vines” (ibid.) if we stay in the plant domain. Her cheeks “are comely with rows of jewels” (1:10), her neck is “with chains of gold” (1:10). In all these examples we see different semantic traits becoming relevant: livelihood, greenness, youth, richness. While they all maintain their specific local resonances, they can also be seen at a more abstract level as figures for fertility, a fertility emphasised by the repetitions of the figure of the twins. Referring to the model of figurative regimes (§ 2.2.4.1.4) we can say that there is a blending of confusive and distinctive regimes. The local metaphors seem to take on a 185


distinctive regime. They are at least partially innovative, they open up to a plurality of blending possibilities. However, such metaphors and comparisons leak onto each other, echoing and resonating through shared terms. The breasts are both does and vines, twins are the breasts and the lambs to be given birth to by the teeth, almost excessively rich are the jewels constituting the cheeks and the pregnancies, regular and even are the sheep and the pillars. And so on. And all these parts leak onto the girl – and vice versa – creating the semantic configuration of a healthy and fertile. companion. More could be written on how this metaphorical dissemination interacts with the narrative structure of the poem, with the depiction of the fictional minds of the characters and so on. However, this was enough to show how via an integration of cultural stereotypes, genre-related norms and textual features that support and modulate them semantic grounds are constituted and stabilised. Through these semantic grounds the local semantic configurations are constituted via a diagrammatic and tentative process attempting inferences and verifying them. This mechanism – adequately situated in the relevant practices - is valid both for the activity of the writer, for the interpretation of the analyst and, more importantly, the interpretation of the reader. For the sake of simplicity I have started with a text. However, as argued extensively in Chapter 1, a most prominent affordance of language is its being public and open to intersubjective and social manipulations. Let us take an example [collected and introduced to me by Anders Hougaard] from a very specific discursive practice: the passage of speech between the news anchorman and the weather forecast anchorman on Danish TV (DR1) on February 6th 2009 at 18.30.

A: og det er jo georgia o'keefes billeder at de er sanselige (.) de er sensuelle (.) og de er erotiske .hhh og

186

Language and Cognition - Ph.D. Dissertation by Riccardo Fusaroli det (0.6) de er malet sån så de damper ud af dem 1.3 og nu vil metrologen søren jakobsen vise sine nyeste billder , .hh de handler vist mest om (.) vejret 0.4 B: det gør de ja og det fortsætter 0.8 owse (.) fugtigt skulle jeg måske sige .hh der er øh=en stor frontzone her jo som (.)

glider mod nordøstlige ....

… A: and now Georgia o'Keefe's pictures that are sensuous (.) that are sensual (.) and that are erotic .hhh and that (0.6) that are painted so that it steams out of them 1.3 and now the meteorologist søren jakobsen will show his newest pictures , .hh they are probably mostly about (.) the weather 0.4 B: they are so, yes and it continues 0.8 also (.) damp/wet should I maybe say .hh there is øh=a big frontal zone here yes that (.) slides towards the northeast ....

What we see here is the joint – although probably not completely willing – construction of a metaphor. The news anchorman is constructing a semantic ground, as sketch of relations that act as the interpretive premise for what is coming after. Sensuality and sexuality inhabit the interpretive space ready to be further modulated. Therefore, when the meteorologist pronounces the very innocent sentence “it continues also damp” the humidity is projected onto a complete different plan than weather forecasting, therefore prompting the “wet” translation of it. And the same happens, although with attenuated strength, with “big frontal zone” and “slides”. Hougaard (2011) produces a skilled analysis of the pauses and the role they have in triggering the metaphorical interpretation. For our purposes here, however, it is sufficient to show how via language the metaphor – weather for sexuality – is intersubjectively distributed120. An intersubjective interpretive habit is established

120

Other studies on metaphors in conversations, although not so focused on distribution, are Cameron (2003, 2008), Cameron and Deignan (2006), and Gibbs and Cameron (2008). These studies focus on the life of metaphors in conversation: once a linguistic metaphor is introduced into

187


and it is through it that metaphors are defined and meaning is re-constructed and shaped. Many more examples could be produced – showing how weather forecasting is indeed an underestimated and underanalysed discursive practice – it is time, however, to pass to a wider public arena for metaphors, the corpus of a newspaper, to investigate the life of a metaphor in society: “Berlusconi is a caiman”.

2.3. The strange case of the Man who Became a Caiman or Of the Life of Caimans in Society What is it like to be caiman? Or better, what is it like to become a caiman? I have already mentioned many times how this is what happened to Berlusconi, the most (in)famous Italian prime minister and entrepreneur in the last 9 years. It all began in late 2001 when Franco Cordero, a renowned law and philosophy professor, began collaborating with La Repubblica – one of the main Italian newspapers, and one of the few with a moderate leftist perspective. Cordero sets upon analysing in detail Berlusconi's legal "reforms" – also called ad personam laws – and Berlusconi's personal behaviour in politics and in the tribunal. In doing so, he often produces syncopated lists of Mr. B. deeds: he fights non-existent communism, he promises the moon, he exploits popular antipathy against the very same old governments to which he owes his own privileges and he raises the flag of shameless anti-politics. He collects electors in need of something new – headless masses, ultra-moderates, ultras, orphans, survivors of shipwrecks, rampant ones, trans-crossers, phobic congregations, quails ready to jump, and he wins121 (26/02/2002).

121

the discourse it can be redeployed, developed or dropped (Cameron 2008). In redeployment, the same metaphorical word or phrase is used by either the same or the other speaker with a different topic. In the case of metaphor development, the same word or phrase is used again with the same topic. Metaphor development can be achieved by means of a variety of processes such as repetition, relexicalization, explication, and contrast (Cameron 2008). "combatte l' inesistente comunismo; promette la luna; cavalca l' antipatia verso i vecchi governi, ai quali deve i privilegi; inalbera insegne d' antipolitica spregiudicata; raccoglie elettori bisognosi d' un quid novi, masse acefale, più che moderati, ultras, orfani, naufraghi, rampanti, trasvolatori, congreghe fobiche, quaglie pronte al salto; e vince".

188


On May 15th, 2002 in one of the more and more frequent articles by Cordero this passage appears: the only rank that can be attributed to him, anthropologically, is that of prince of the crooks, one of those crooks that charmed the peasants at the local fairs: fluvial speech, smiles as a caiman, overabundant gestures, cheerful effects; nothing made him predestined to a political dominance. If he were forced to participate to fair dialectical discussions (on TV discussions are rigged), he would drown in half a glass of water”122.

This is, however, only one image that is easily lost in the wider Pantagruelic description. And indeed it does not seem to have any influence. We have to wait more than a year before we see the metaphor again which we do on May 20th, 2003: Berlusconi “displays the hunger of the caiman, not to mention the taste, intellect, moral feelings, irony as the Leviathan that sneezes fire in the dialogue between Yahweh and Job”123.

This new occurrence is soon followed by other ones like “the jaws of a caiman” and again “the smiles of a caiman”. That colourful but minor image in Cordero’s article slowly got a life of its own and in just a couple of years we can find a whole family of related metaphors which can be given the more generic label “Berlusconi is a caiman” – metaphors being repeated and varied at least 600 times only on the pages of La Repubblica and countless times more in the Italian mediascape. The first article in which the metaphor appears would on its own constitute a rich field see sedimented conceptual metaphors, pre-existing linguistic usages and cotextual dynamics at play in the coinage of a new metaphor. However, a more extended corpus approach allows us to ask – and maybe even start to answer – new 122

123

"L' unico rango che gli competa, antropologicamente, è quello da principe dei «bagalun dl' lüster», gl' imbonitori che incantavano i contadini nelle fiere: loquela fluviale, sorrisi da caimano, gesto sovrabbondante, effetti ilari; niente lo predestinava al dominio politico. Costretto a partite dialettiche leali (sugli schermi gliele addomesticano), annegherebbe in mezzo bicchiere d' acqua." “espone appetiti da caimano, nonché gusti, intelletto, sentimenti morali, ironia, conformi al Leviathan che starnuta fuoco nel dialogo Yahweh-Giobbe"appetiti da caimano"

189


questions and to investigate a different time scale than the ontogenetic one seen at work in the ontogenetic development grounding conceptual metaphors and the online unfolding of linguistic metaphors in texts and conversations of the previous paragraphs. What is the role of verbal patterns, pre-existing or locally established? How does language use engage and reshape concepts? What is the life of caimans in society, to put it more lightly? A question that is particularly interesting since caimans as words and as animals did not use to have a current currency in Italian everyday life, a question therefore enabling us to see the variations of the concept behind “caiman”. A second set of questions that we can address but not completely solve is about the reasons behind the incredible success of the metaphor. Because this was not the first time Berlusconi was compared to an animal. Michele Serra, a satirical author, called him “billionaire ridens”. Francesco Cossiga – a former president of the Republic – for some time tried to portray Berlusconi as a small tail-wagging dog with a huge smile. Berlusconi is also, with a certain frequency (200 occurrences on La Repubblica) called “biscione”, that is, the “big grass snake”. However, neither of these metaphors achieved the prevalence of “The Caiman”. “Billionaire ridens” never became more than an appellative used exclusively by Serra and did not produce consequences in the general structure and content of the articles it appeared. Cossiga's drawing cannot even be found on the Internet. “Big grass snake” is still around but only in newspaper articles dealing with television and conflict of interest issues always with a strong connection to Fininvest. Additionally, just like “billionaire ridens”, “big grass snake” is rarely further elaborated in the articles in which it appears. “Caiman”, instead, lives a healthy life in all sorts of media and discourse domains, and from its initial application to Berlusconi it has spread to refer to all sort of individuals and behavioural styles.

190


2.3.1.

The importance of being a caiman in the news

The possibility of asking the questions listed in the previous paragraph was my main reason to choose this metaphor. Tackling a popular metaphor used in the media allowed me to see how a metaphor is used in a social and public arena over a period of time. This protected me from the risk of using the exempla ficta that seem to be far too frequent in linguistics (cf. Gibson and Fedorenko 2010) and semiotics (Fabbri 1998). A fact that is very clearly exemplified by Chomsky’s approach: “If you sit and think for a few minutes, you‘re just flooded with evidence” (Chomsky 1984:44). The use of corpora, instead, floods us with real life examples in a fully textual dimension that are not just representative of individual sensitivity and biases and exposure to this or that genre of texts. Moreover, the fast life cycle of the media gives us the possibility to observe in a relatively short time-span the life and career of the metaphor. What does it take for a freshly coined metaphor to be effectively presented? How does it evolve – if so – through repetition? How does it change when moved to new contexts? How does it interact with previous uses of the words and expressions of the concepts involved? Last but not least, investigating metaphors in the pages of a newspaper with a freely accessible online archive of all the issues since 1984 saves us from the need to digitalise our corpus before being able to partially automatise the search for occurrences and for their frequency, variations, co-occurrences and colocations. However, employing a corpus approach requires to be very aware of the criteria used to create the corpus and to search in it.

2.3.2.

Cognitive Linguistics and corpus approaches

I am not the first one to propose to coordinate corpus approaches with cognitive linguistics. Indeed, when Kövecses (2008) attempted his very partial – and defensive – summary of the criticism to which CMT has been subjected, corpus approaches were one of the main targets. Authors like Allan (2009), Boers and Demecheleer 191


(1997), Cameron (2003), Cameron and Deignan (2003), Charteris-Black (2000, 2003, 2004), Deignan (1999, 2005, 2008), Fernando (1996), Gries (2003), Gries & Stefanowitsch (2006), Grondelaers, Geeraerts & Speelman (2007), Koller (2003) Lindstromberg (1997), Semino (2002), Skorczynska (2001), Skorczynska and Deignan (2006) Speelman, Grondelaers & Geeraerts (2003), Stefanowitsch (2003) and Stefanowitsch & Gries (2003) have, using conceptual, epistemological and analytical argumentations, called for a more empirical approach to the linguistic analysis of (conceptual) metaphors. Such an approach has to take into account real occurring expressions embedded in their context of occurrence, as well as the frequency of their occurrence as compared to similar and related expressions. Another important source of inspiration for this analysis is the work of Francois Rastier and his attempts at conjugating computer aided corpus research with a strong hermeneutically and semiotically oriented perspective (Rastier et al 1994, Rastier 2002, 2005). 2.3.2.1. What is a corpus?

A corpus is usually defined as “any collection of spoken or written texts” (Deignan 2008:282) or even as a “wide set of words” (Sinclair 1991:ix). Corpora are often built around the work of a single author, a set of issues of one or more newspapers, collections of transcribed spoken data or more broadly based collections of a range of text types. However, a closer consideration of most corpora reveals a more accurate set of principles in the construction of it that reflects research interests – the specific questions of the investigation and so on. Damon Mayaffre (2002) writes about “reflexive corpora”, that is, corpora that are constructed following explicitly formulated interpretive principles. Inspiring himself to Mayaffre Rastier defines a typology for the data investigated: i) the archive, constituting the complete set of the available documents; ii) the corpus of reference, constituting the set of texts against which the findings of the investigation will be compared in order to test their general relevance and the specificity of the text analysed; iii) the working corpus, defined by the needs and aims of the investigation; iv) the subcorpus, which is the object of more 192


detailed work, and that is constituted by the relevant passages of the text(s) analysed (Rastier 2004). This taxonomy makes the (ideal) hermeneutic foundations that a corpus should have explicit. A corpus should avoid being a set of words, of examples or of fragments. A corpus should be a structured set of integral texts that are documented, possibly enriched by tags and articulated i) in a reflexive and theoretical way that takes into account discourse and genre and ii) in a practical way according to a possible range of investigations. 2.3.2.2. Different Kinds of Corpora

This brings us to the issue of how to assemble our corpus. Some corpora have already been assembled – “ready-made” corpora – but most of the time researchers have to rely on corpora that they have assembled themselves. Ready-made corpora are large and they usually include a wide range of text types and genres, in an attempt to enable the analyst to get results as representative of the language as a whole as possible. However, it would obviously be unsafe to claim that any corpus can ever be truly representative of the language experience of all speakers. Moreover, any study that attempts to take into account the role of context – and history, genre, etc. – or to go more into details with a specific linguistic development – as in our case – has to create its own corpus to which these very specific questions can be asked. Some examples of this kind in metaphor studies include Cameron’s corpus of educational discourse (2003), Koller’s corpus of newspaper reports and journal articles on the topic of mergers and acquisitions (2003) and Santa Ana’s corpus of issues of the Los Angeles Times (1999). There are a few “ready-made” corpora for Italian124, one of the most renowned

124

Cf. Lessico di frequenza dell'italiano parlato (LIP, 1993), Lessico di frequenza della lingua italiana contemporanea (LIF,1972), Vocabolario elettronico elettronico della lingua italiana. Il vocabolario del 2000 (VELI, 1989), Corpus di italiano parlato (Cresti 2000) and LIZ ( Letteratura Italiana Zanichelli in cd-rom (1993,1995, 1997), the Italian Reference Corpus (1991) and the Italian

193


being the CORIS/CODIS developed at the University of Bologna125. Even more specifically the SSLIMIT in Forlì has collected a corpus with all La Repubblica’s articles from 1984 to 2000126. However, these corpora were not fit to be working corpora for our investigation for two main reasons. CORIS/CORDIS does not classify its texts and results according to any specific social uses, genres and diachronic developments. Moreover, an explorative search resulted in no occurrence of the word “caiman” and of the class of linguistic metaphors I am investigating. While better structured to allow for the search of genre and discourse related parameter, La Repubblica/SSLIMIT does not include the historical period in which the metaphor emerges and unfolds. For these reasons these corpora will only be used as loose reference corpora in order to create a background for the further examination and comparisons of my findings. In the same way, other corpora have been sporadically consulted: Corpus-eye (http://corp.hum.sdu.dk/cqp.it.html) which consists of the Italian transcriptions of the European Parliament sessions and on a dump

of

Wikipedia

made

in

2005,

the

Bank

of

English

(http://titania.cobuild.collins.co.uk/boe_info.html), the Internet archives of La Repubblica, il Corriere della Sera, l'Unità, Il Fatto, il Giornale, Le Monde, plus various versions of Wikipedia in Italian as cached at more or less regular time spans by

125

126

Corpus Documentation PAROLE (1998) both developed at the ILC by the the Pisa CNR. “A [synchronic] corpus of written Italian - CORIS/CODIS is available on-line for research purposes. The project, designed and co-ordinated by R. Rossini Favretti, was started in 1998, with the purpose of creating a representative and sizeable general reference corpus of written Italian which would be easily accessible and user-friendly. CORIS contains 120 million words and has been updated every three years by means of a built-in monitor corpus. It consists of a collection of authentic and commonly occurring texts in electronic format chosen by virtue of their representativeness of modern Italian”. (http://corpora.dslo.unibo.it/coris_eng.html) “The "la Repubblica" corpus is a very large corpus of Italian newspaper text (approximately 380M tokens). The corpus is tokenized, pos-tagged (with the Treetagger trained with ad-hoc resources), lemmatized (with Morph.it) and categorised in terms of genre and topic (with SVMLight trained with ad-hoc resources). The genre labels used in the corpus are news-report and comment; topic labels are church, culture, economics, education, news, politics, science, society, sport, and weather. Articles in the corpus are structured into the following (mostly optional) parts: title, subtitle, summary, and text. Meta-data information about article author and year is also available. The corpus has been indexed using the IMS Corpus WorkBench. (http://dev.sslmit.unibo.it/corpora/corpus.php?path=&name=Repubblica)

194


internetarchive.org, as well as the Web in general. However, for the present purpose I delved specifically into the unstructured Internet archives of La Repubblica (www.repubblica.it)127 that hosts the birth of the metaphor and most of its occurrences. The archives cover the period from October 1st, 1984 to today. For practical reasons the corpus has been considered up until August 31st, 2009. In May 2010 the results of the analysis were double-checked on the Internet version of the archives, covering articles up until May 13th, 2010. From this (reference) corpus I selected the relevant texts containing occurrences of the metaphor to form the working corpus. I have carefully considered the entirety of each article when analysing the metaphor appearing in it, even when this textual dimension is not mentioned explicitly in the following pages. This procedure, in which working corpus and sub-corpus coincide, was remarkably time consuming but necessary for me to be able to map the semantic configurations constructed through the metaphors. Since, by corpus linguistics standard, a corpus like mine is quite limited not reaching the recommended 10 million words, I believe that the textual dimension of the analysis – a careful analysis of the entire co-text focusing its holistic constraints on local occurrences – was an essential criterion of scientificity. Moreover, such an analysis allowed me to avoid undue reductions of meaning into purely formal and/or statistical patterns and to individuate less obvious evolutions of the meaning of the metaphor. 2.3.2.3. Methodology of selection

Researchers in corpus linguistics have individuated five main methods to select a proper corpus from an unstructured archive when investigating (conceptual) metaphors: i) manual searching (Semino & Masci 1996), ii) searching for source domain vocabulary (Deignan 1999, Hanks 2004, 2006, Hilpert 2006, Partington

127

The other newspapers were also initially investigated but did not produce enough material for any conclusion that went beyond being just anecdotic. However, where those findings can help strengthening or cautioning us about one of my analytical points, they will be mentioned.

195


2003, 2006, Köller 2006, Markert and Nissim 2002, 2006), iii) searching for target domain vocabulary (Koivisto-Alanko 2000, Tissari 2003, Stefanowitsch, 2004, 2006, Koivisto-Alanko and Tissari 2006), (iv) searching for sentences containing lexical items from both the source domain and the target domain (Martin 2006)and v) searching for metaphors based on ‘markers of metaphor’ like “metaphorically speaking” (Goatly 1997). The fifth method did not yield any interesting results when applied to my La Repubblica reference corpus, so it was quickly discarded. The search for target domain vocabulary (in particular for Berlusconi and related terms) generated far too many irrelevant results and therefore I opted for a search that combined lexical expressions related to the source domain, lexical expressions from the relevant regions of both the source and the target domains and a manual search through the selected texts in order to carefully verify and in some cases adjust the relevant keywords and queries defining the proper corpus. Finally, I organised the texts into a temporal sequence and manually identified i) more prototypically literal occurrences of the morpheme “ca*man*” (and their two main referents, the animal and the islands), ii) metaphorical occurrences of the morpheme “ca*man*” and iii) specific occurrences of the “Berlusconi is a caiman” metaphor128. 2.3.2.4. Investigating concepts via verbal patterns

The use of corpus methodology in a semiotic investigation is not unproblematic. The risk is to lend too much attention to the expression – the verbal surface of language – and not enough to the content and the nuances of the ongoing meaning construction processes. This risk is especially strong when employing an onomasiological approach as opposed to a semasiological one as we did – in other words when the

128

For the present purpose no intercoding protocol (Neuendorf 2002) employing multiple coders to ensure the reliability of introspective evaluations of linguistic and semantic aspects has been employed. However, all the results have been discussed with both native and non-native speakers of Italian.

196


corpus is not semantically tagged, but the searches are driven by lexical and morphological criteria. However, a main argument for the relevance of such approach may be adduced here. The first is that introspection and intersubjective control and discussion of the interpretation of corpus results is very much in focus in much of corpus linguistics: “Because meanings do not present themselves directly in the corpus data, will introspection not always be used in any cognitive analysis of language?” (Geeraerts & Cuyckens 2007: 18). What a carefully crafted corpus linguistic approach can give us is a more refined perspective on the data, on the sedimentation and evolution of linguistic and semantic associations as well as on the interactions of semantic constructions and the linguistic expressions of them. This wealth of data can complement, lead, constrain and stimulate the sensibility of the analyst and grant additional ways to test and falsify hypotheses. A second issue that has been raised a couple of times when presenting my findings is the relation between semantic configurations and concepts. The underlying argument in the questions was relying on two oppositions: • Human beings have concepts in the head, while language use is out in the world, cannot be reduced to individual cognitive processes. • Concepts

are

imagistic

and/or

propositional

structures,

semantic

configurations are instead the complex effects of differential structures (Rastier 1991, Paolucci 2010). What does thus entitle me to investigate conceptual metaphors and the concept of caiman via corpus linguistics? These questions certainly point out effectively that there is more to concepts than the way they are expressed in words. We have seen this already in the attempts of cognitive linguistics to analyse conceptual metaphors in behaviours, movies, images, gestures and so on129. However, there is a closer 129

The first conclusion these studies support is that relationships between non-linguistic domains of knowledge cannot necessarily be inferred from metaphors in language. Linguistic metaphors reveal only a subset of the conceptual metaphors that appear to structure our mental representations of

197


interplay between conceptualising operations and linguistic uses and interactions than these oppositions would lead us to believe. In the pragmatist framework developed in chapter one conceptualisation goes on in many important ways in the public arena and linguistic uses and interactions are one of the most important ways of creating conceptual structures, putting them to play, tuning and negotiating them, thus semiotically grounding a position on concepts that sees concepts as locally assembled constructions participating in the ongoing practice and in the surrounding context and co-text and not as entrenched representational structures. A conception that, indeed, is not that far from much contemporary research in cognitive sciences: Conceptual structures emerge as temporary constructions in working memory constructed on the spot from generic and episodic information in long-term memory, as well as from contextual interactions, rather than as stable structures stored in long-term memory (Gibbs 2003: 32)130.

2.3.3.

The Overall Perspective

How to look at this wealth of data? I first approached the corpus achronically, that is, considering the totality of the data available, searching for the occurrences of the target – Berlusconi – and the source – caiman –, for their most frequent cooccurrences and for which relations they entertain with them. While it is not a direct analysis of the meaning configurations constructed through these morphemes, this search can give us some hints as to how much and how the morphemes at stake are

130

similarity and time. The second conclusion is that even when linguistic metaphors fail to predict the exact relationships revealed by behavioural tests, they nevertheless point to important links between the source and target domains. Space and time, speed and time, proximity and similarity are not unrelated: rather, they appear to be related in more complex ways than linguistic analyses alone can discover. As such, linguistic metaphors should be treated as a source of hypotheses about the structure of abstract concepts. Evaluating these hypotheses – determining when a linguistic metaphor reflects an underlying conceptual metaphor – requires both linguistic and extralinguistic methods and calls for cooperation across disciplines of the cognitive sciences (Casasanto 2009:143). Cf. for example Raczaszek-Leonardi & Kelso 2008 and their psycholinguistic experiments highlighting the context and co-text situatedness of prototypes as activated by sentences like “the bird is on the table” vs. “the bird is on the branch.

198


used, which semantic traits are commonly activated in them (in Rastier's terms, which afferent semes are actualised in the semantic molecule constructed interpreting the morpheme). As we will see, such an approach will not yield enough information for my enterprise and I will have to take a diachronic stance. 2.3.3.1. Searching for Berlusconi

The popularity of Berlusconi as a topic of conversation, his newsability and his talent for occupying the centre of the scene is well portrayed by the amount of occurrences of his name: 107.366 occurrences in the La Repubblica corpus alone. Quite an achievement if we consider how other illustrious politicians fare in this department. Amongst the evergreen we can mention Francesco Cossiga (11.835) and Giulio Andreotti (18.636). Amongst newer politicians the numbers are still significantly lower: Umberto Bossi (23.410), Pierferdinando Casini (17.118), Francesco Rutelli (21.482), Massimo D'Alema (27.304), Walter Veltroni (34.456). Even Romano Prodi, for many years the only credible alternative to Berlusconi and a former president of the European Commission, only scores 47.661 occurrences. The occurrences of “Berlusconi” show a steady increase through time and huge peaks coinciding with electoral campaigns and periods in which he is prime minister: 14000 12000

Occurrences

10000 8000 6000

Berlusconi

4000 2000 0 1980 1985 1990 1995 2000 2005 2010 2015

Year

199

Language and Cognition - Ph.D. Dissertation by Riccardo Fusaroli Figure 10 –The Diachronic evolution of occurrences of “Berlusconi”

When Berlusconi is mentioned his name appears most often in the genitive case: around 25.000 occurrences. Berlusconi's government, Berlusconi's companies, Berlusconi's television channels, Berlusconi's empire, Berlusconi's allies, Berlusconi's minions and so on. He is quite often the subject of an active verb (blabber, win, declare) and more rarely in the accusative case, although the reports of his legal vicissitudes (accused, under investigation, under process) ensure a significant presence of this case. An interesting comparison can be made by listing the co-occurrences of both Berlusconi and Prodi131 and thereby showing how this semantic structure is a prerogative of Berlusconi and how Prodi does not share his generalised semantic trait of control, quite the contrary: Berlusconi

Prodi

1

Blabber (straparla)

Led by (presieduto)

2

Sir (Cavalier)

Fails (boccia)

3

Prime Minister (Premier)

Government (governo)

4

Leadership

Prime Minister (Premier)

5

Attacks (attacca)

Promises

6

Televisions (Televisioni)

Debate (discussione)

7

Allies (Alleati)

Meeting (colloquio)

131

The lists are not produced according to frequency but according to association strength. The problem with a frequency-based definition would have been that it would not have taken into account the base frequency of the linguistic items in question and thus would not have allowed us to assess whether an item occurs more or less frequently with the node word than would be expected by chance alone (see e.g. Church and Hanks 1990, see also Stefanowitsch and Gries 2003).

200

Language and Cognition - Ph.D. Dissertation by Riccardo Fusaroli 8

Ally (Alleato)

Informs (informa)

9

Under investigation (indagato)

Falls (cade)

10

Government (governo)

Compromise (compromesso)

11

Under process (imputato)

Invites (invita)

12

Fininvest

Replies (replica)

13

Party (partito)

Catholic (cattolico)

14

Men (uomini)

Meets (incontra)

A second control on the semantic configuration of Berlusconi comes from checking the archive of Le Monde for Berlusconi’s co-occurrences132. This yields similar results, with an even stronger predominance of Berlusconi’s position of control as expressed by the genitive case. “Berlusconi”’s co-occurrences

“Berlusconi”’s case

Fininvest

genitive

Government (gouvernement)

genitive

Group (groupe)

genitive

Party (parti)

genitive

To have to (devoir)

nominative

Italy (Italia)

genitive

132

Due to the different structure of the archive of Le Monde, I was only able to check for raw frequency and not for association strength. However, given the purely qualitative point – the constancy in a predominance of nominative and genitive cases, this does not constitute a crucial problem.

201

Language and Cognition - Ph.D. Dissertation by Riccardo Fusaroli To be able to (pouvoir)

nominative

Victory (victoire)

genitive

Movement (mouvement)

genitive

To control (contrôler)

nominative

Empire (empire)

genitive

To do (faire)

nominative

To manage (diriger)

nominative

Ally (allié)

genitive

Friend (ami)

genitive

Without trying to infer too much from these data and co-occurrences, we can easily argue that “Berlusconi” is used almost exclusively in a referential way, as a proper name. This does not devoid it of semantic content. There is the recurrence of certain semantic traits and cases: possession, controlling agency, legal troubles. Conjoined with the high number of occurrences, this allows us to argue that there is a sedimented conceptual knowledge of the public figure Berlusconi both as an entrepreneur and as a politician, and that knowledge depicts him as an agent in control and in possession of many things. 2.3.3.2. Searching for the caiman

If we try to apply this process to the word "caiman" [caimano] we do not get many results. Standard retrieval strategies for co-occurrences do not yield any significant result. Caiman is a strange word in Italian. In general, the term “caiman” which according to the dictionary refers to alligators, a species of South-American crocodiles, is infrequent. If we check a copy of the Italian version of Wikipedia from 202


2003 we can find two occurrences due to the description of symbols on shields – “ a rampant jaguar and a rampant caiman” [un giaguaro e un caimano rampanti] and “Caiman:

Symbol

of

the

Department

of

Choco”

[Caimano:

Stemma~del~Dipartimento~del~Choco] – as well as two passing references to a specific kind of caimans (Baba, or caiman from Llanos): “the giant anaconda and the Baba” [l'anaconda gigante e il caimano dagli occhiali] and “analogous to the one on the Baba” [simile a quello riscontrabile nel caimano dagli occhiali]. Until the coinage of the metaphor “Berlusconi is a caiman”, a search on La Repubblica generates 36 occurrences that will be analysed more in detail in the diachronic analysis (while “crocodile” [coccodrillo], for instance, produces around a thousand occurrences). The CORIS/CODIS corpus does not list a single occurrence. Analogous results appear when similar searches are carried out using corpora of other European languages (English, French and Danish). These searches show very limited amounts of occurrences. Taking the Bank of English, one of the largest corpora for both written and oral English, as an example, it lists only three occurrences: •

“Mr Lapointe hopes to begin in Bolivia, where poaching threatens the caiman crocodile”.

•

“Yet with the right kind of controls, and the development of proper management for the species, the caiman could be conserved and still create income for local people and foreign exchange for the government.”

•

“So, for instance within Random House (under Gaily Rebuff) we find Jonah Cape, once run by Tom Marcher, now by David Godwit and Gearing Camel; Chatty and Winds, then as now run by Caiman Calix; Hutches; Barrier & Jerkins; Bolder Head; etc, etc.”

My working corpus, however, yields very different results: even subtracting those occurrences that refer to the Cayman Islands (including the many Italian misspellings that account 494 occurrences) we find 808 occurrences of the morpheme “caiman*”. 203


What does it mean, then, to be a caiman in the Italian news? If we have a look at our subcorpus, the use of the word gives access to a plurality of referents: SouthAmerican crocodiles; exotic and slightly disgusting talismans, dictators from Central Africa, Italian football players, the Caribbean Islands, Italian water polo players, Italian politicians and Italians in general. These different referents are represented through complex and different semantic configurations. Caimans come out as exotic and out of place; voracious and dangerous, they have to be killed since they cannot be controlled. When mummified they testify to a culture and a place that is other, alien, barely conceivable. When in plural the word often appears on a long list of exotic animals without further specification. The Cayman Islands, a British Overseas territory located in the Caribbean Sea, is a place out of control and often opposed to an internationally regulated capital market and what happens there can, on the one hand, be deemed illegal when seen from the outside but on the other hand this is something that due to the constitutive smoothness of the legal and financial space of the Islands is in fact technically legal, or at least impossible to judge. After these more literal occurrences we have a few cases where the figure of the caiman refers to human beings, a list of which follows below: •

Félix Huphouet-Boigny, the first president of the Ivory Coast (“the students that have long been demonstrating against Boigny, the grand caiman of Yamoussoukro [gli studenti che protestano da tempo contro Boigny, il grande caimano di Yamoussoukro] 11/09/1990).

•

Pierluigi Casiraghi, a famous striker from the Italian football scene (“Fiorentina has been defeated by a goal of that great opportunist that is Caimano Casiraghi” [La Fiorentina è stata messa sotto da un gol di quel grande opportunista che è Caimano Casiraghi] 03/09/1991).

•

Eraldo Pizzo, an almost deified waterpolo player (“Many believe that Pizzo “the caiman” will start playing again” [sono in molti a credere che Pizzo, "il 204


caimano", scenda ancora in vasca a giocare]), and the waterpolo players led by him. •

Berlusconi and his party.

•

The mythical communist opposition to Berlusconi (“the red Caimans” March 25th, 2006); and from the late 2008 more generic groups of people (“the Italy of the caimans”; “little caimans at the airport”).

•

Valentino Rossi (“Le Grand Prix de France was now the theatre of Vale’s revenge, [since Vale had become] once again an insatiable caiman” [Il Gran Premio di Francia era ormai il teatro della vendetta di Vale tornato caimano insaziabile] May 22nd, 2006)

•

Cesare Previti (“ignoring the danger he had stopped a caiman which had escaped from its pond at Euroflora […] It was Cesare Previti visiting” [con sprezzo del pericolo aveva bloccato un caimano evaso dal suo stagno all' Euroflora. [...] Era Cesare Previti in visita” April 23rd, 2001])

•

Aaron Peirsol (“Peirsol ,the caiman” [il caimano Peirsol] August 18th, 2008)

•

Orrico ("When one hears about Orrico, one expects a caiman ready to eat its prey. Then one discovers he is a pet good for the courtyard” [Quando sente parlare di Orrico, uno si aspetta un caimano, pronto a mangiare la preda. Dopo invece si accorge che é un buon animale da cortile] July 21st, 1991)

•

Casillas (“If Casillas wins it will not be because they call him the Caiman” [Se vince Casillas non sarà perchè lo chiamano il Caimano] June 25th, 2008)

•

Achille Lauro (“Comparing them? Difficult. Two different ages. Well, certainly Lauro was the billionaire of his age. He was a populist like Berlusconi. He was in his own way a little caiman. But at least he was from Napoli” [Accostarli? Un azzardo. Due epoche diverse. Beh, certo, Lauro era il miliardario dei suoi tempi. Era populista come Berlusconi. Era a suo modo 205


un piccolo 'caimano'. Ma almeno era napoletano]) 24/05/2006 •

Rupert Murdoch (“the Australian caiman of the media” [il caimano australiano dei media] 19/04/2006)

In the disparity of the individuals listed, there seems to be a constellation of possible motivations. As in all the re-enactments of the PEOPLE ARE ANIMALS conceptual metaphor, the metaphor constructs an anthropomorphic perspective on the animal, selects one or more physical (behavioural or appearance relater) traits and projects them on the human being creating in the process a moral connotation. Every now and then the image of the snapping jaws of the caiman are turned into a vision of voracity, of danger, of breaking any opposition or boundary. Predictably, the Cayman Islands do not seem to share such semantic structure, displaying an absence of boundaries instead. In order not to opt for a simple reductionism, trying to identify a common core to all the metaphorical occurrences, I went for a more in-depth and chronological approach. The occurrences are distributed as follows.

206


400 350 300

Occurrences

250 Caimano Caimani Caimana Caiman*

200 150 100 50 0 1980

1985

1990

1995

2000

2005

2010

Years

Figure 11 – The diachronic evolution of occurrences of “caiman*”

If we check the development of the cluster of morphemes denoting the Cayman Islands we get a surprisingly similar pattern133:

Cayman Islands in Time Figu re 12 Dia chro

100 80

Occurrences

133

120

Caimano (isole del) 60 Isole And a pleasant surprise when seeing how, in time, Italian journalists learned to Caimane spell the name of Cayman* the islands properly. Compound 40

207

20 0 1980

1985

1990

1995

Year

2000

2005

2010

Language and Cognition - Ph.D. Dissertation by Riccardo Fusaroli nic Evolution of Cayman Islands

However, the surprise due to such an analogy in the pattern should give place to a more careful

400

analysis

350

of

the

data.

300 250

Figure 200 13 Diachr 150 onic Evoluti 100 on of Caiman 50 AND Cayma 0 n 1980 Islands

Cayman Islands Caiman

1985

1990

1995

2000

2005

2010

The two main peaks in the graphs happen in different years: the caiman peaks in 2006 when Moretti's movie “Il Caimano” is released, filling the pages of the newspapers; the Cayman Islands peak in 2004, when the giant food company Parmalat declares bankruptcy and huge sums of money are discovered in a Cayman Islands based offshore entity134. The two peaks seem to be unrelated, but the general increase of the use of these words after those peaks could actually be related, as we will see. 134

Parmalat was the largest Italian food company and the fourth largest in Europe, controlling 50% of the Italian market for milk and milk-derivative products. In 2000 it was discovered that the 4 billion Euro in liquidity Parmalat claimed to have at its disposal did not exist and that 8 million Euros in bonds of investors' money had evaporated as well. Parmalat is the largest bankruptcy in European history, the losses representing 1.5% of Italian GNP—proportionally larger than the ratio of the combined Enron and WorldCom bankruptcies to the GNP of the United States. Behind Parmalat's facade as a productive agro-industrial company with 34,000 employees, hid a giant financial speculative scheme to attract investors' money and syphon it off through a network of 260 international offshore speculative entities, whereto the money disappeared. At the receiver end was a Cayman Islands-based offshore entity called Bonlat supposed to have received at least 8 billion Euro. This is the reason for the high frequency of references to the Cayman Islands.

208


A last thing that we can examine in this gross-grained approach is how metaphorical and literal uses relate to each other. In order to do so we have manually coded all the occurrences, excluding ambiguous cases and the times where “caiman” refers to Moretti's movie from 2006. Metaphors and literal uses coexist in the beginning with a slight dominance of literal uses. However, metaphorical usages tend to take over, even when we eliminate all the occurrences directly referring to Berlusconi. 80 70

Occurrences

60 50 40

Caiman(s) Met Caiman(s) Lit

30 20 10 0 1980

1985

1990

1995

2000

2005

2010

2015

Year

Figure 14 - A diachronic perspective on metaphorical and literal uses of “caiman*”

2.3.4.

A Diachronic Approach

Just an overview of how the words involved are distributed in the text does not really help our analysis. It lacks a crucial dimension, namely temporality. And already in the end of the previous paragraph the diachronic evolution of the use of the morpheme “caiman*” both in its metaphorical and literal uses has been employed. It is thus time to develop a more fine-grained analysis. The evolution of the meaning of the morpheme “caiman*” and of its metaphorical uses can be divided into five indicative stages: i) caimans before The Caiman (from 1984 to 2001), ii) conception and birth of the Caiman (2002), iii) the growth of the Caiman (2003 to 2005), iv) caimanic contagion (2006) and v) little caimans in society (2007-2009). These stages have been individuated due to differences in both frequency of use and meaning configuration. Due to the limited amount of occurrences, no serious statistical 209


treatment of the text is possible. Moreover, such an approach alone would not be adequate for the theoretical framework we are dealing with. It is necessary to proceed along the old-fashioned path: to read the text and to carefully reconstruct their semantic textures. However, the corpus approach and the automated search in the text allows us to map variations and evolutions in time that an achronic approach does not and both to discover and to verify that there is indeed a systematic pattern. 2.3.4.1. Caimans before the Caiman (1984-2001)

2.3.4.1.1.

Literal uses

From January 1st, 1984 to May 14th, 2002 only 38 literal occurrences of the morpheme “caiman*” are found in my working corpus. Interestingly enough, even in most of the literal occurrences the articles describe hardly any detail that could structure the conceptual configuration of “caiman*”. Only on two occasions are there any details: • “a three metres long caiman that grabbed him with its jaws” [un caimano di tre metri e mezzo di lunghezza [sic] che l' ha afferrato con le sue fauci] in a brief chronicle on page 23 on June 21st, 1993 after a teenager was devoured by a caiman in Florida. However, the article concludes that in general caimans are not dangerous. • “the angered mother caiman made the river overflow with big tail movements” [la madre caimano in collera fece straripare il fiume a colpi di coda], again as part of a very short text on February 17th, 1997 recollecting an alternative myth to the Biblical deluge. In all the other cases the literal uses of the morpheme “caiman*” actualise only the motive of the exotic/out of place/different in an explicit way. A couple of examples: •

“I think that an author does nothing else than to talk about himself even if he talks about – I don't know – the life of a caiman” [Mi pare che un autore non fa che raccontare se stesso anche se racconta, che so, la vita di un caimano] 210


(February 2nd, 1989) •

“from a voodoo rite to a cheek to cheek dinner with a caiman” [dal rito vudu alla cenetta cheek to cheek con caimano] (August 9th, 1989)

The plural form “caimans” appears exclusively in lists of exotic animals, sometimes depicting illegally imported animals, sometimes just creating exotic, imaginary and excessive images: •

“ghosts and caimans, turtles with two heads and sexual violence on a bed of honey” [fantasmi e caimani, tartarughe con due teste e violenze sessuali su un letto di miele] (May 15th, 1987)

•

“big felines, caimans, monkeys and turtles” [grandi felini, caimani, scimmie e testuggini] (December 20th, 1990)

•

“rhinos, crocodiles and caimans” [rinoceronti, coccodrilli e caimani] (February 25th, 1992)

In all the cases applying a wider textual analysis makes it possible to reconstruct how the morpheme “caiman” is crossed and defined by the isotopy /exoticness/, an exoticness that is sometimes declined as /extraneous/, something that should not be here, sometimes as extreme otherness that you can encounter through tourism or on a fictional trip. Often, but rarely explicated, the otherness of the caiman is at the same time potentially risky and violent135. 2.3.4.1.2.

Metaphorical uses

The same period presents 24 metaphorical uses of the morpheme “caiman*” of which only one is in the plural form. Three macro-uses can be considered representative of the general trend: i) Pierluigi

135

Curiously enough, such a representation does not appear in specialised literature on etology and zoology where the caiman is represented as threatened and we learn that compared to the African crocodile the caiman is not that aggressive at all. To the point that when a few African crocodiles are released into the South-American niche of the caimans, the “ferocity and voraciousness of the crocodiles quickly pushes the caimans of the areas to the brink of extinction” (Leslie & Spotila 2004). However, we are dealing here with information that in all probability does not enter the Italian mediascape and any aspects of social life except for isolated and specialised niches.

211


“Caiman” Casiraghi (and other football players), ii) Eraldo Pizzo, called “the Caiman” (and his team of water polo players), iii) Félix Huphouet-Boigny, “the big caiman of Yamoussoukro”, the first president of the Ivory Coast and – almost as a precognition - iv) Cesare Previti, one of Berlusconi's most prominent lawyers. If we analyse the way the metaphor is constructed in the case of football players (Casiraghi on four occasions, but also Orrico), we see a recurring meaning configuration (a “semic molecule” or semantic form (§ 1.4.5), in Rastier's terms): •

/Unstoppable/

•

/individualistic-opportunistic/

•

/precise/effective/

•

/aggressive/

Indeed, the caiman “does not forgive”, the caiman “cynical and opportunistic”, it “scores” and it “does not pass the ball” (September 3rd, 1991), the caiman “has a sure strike”, it “grabs the prey and scores” (September 17th, 1991), the caiman is “furious” with “the grace of Jack the Ripper” (August 25th, 1991) and so on. However, the meaning constructed through the metaphor has a positive dimension even when indicating a certain fury and selfishness136. Concerning Eraldo Pizzo and his team of water polo players two very different components of the metaphor are textualised. The first is the very basic isotopic connection of “caiman*”, “water” and “swimming” that is also made explicit by Wikipedia: “named 'the caiman' for his ability in dominating the water”. However, also Eraldo Pizzo's past is often thematised. Eraldo Pizzo won his first olympic gold medal as a twenty-year-old at the Games in Rome 1960, participated in all the Olympic Games until 1972 and always performed impeccably according to the articles in the corpus. Now he is portrayed as the trainer of the Pro Recco team. There he brings his style that belongs to another age. He is the Caiman, “he is another species” (April 12th, 1998). This otherness of Pizzo is often mentioned; 136

For a refined analysis of how opportunism and individualism can be conceived of as positive traits in football discourse, cf. Demuru 2010, ch. 2.

212


nobody is like him, he demands a degree of discipline of his players “that is not of these days”. Thus otherness of the caiman and its exotic nature can be shown to play a role. Like in the football players’ case “caiman*” displays a positive dimension. Through the Caiman, Pro Recco is great again. Other sportsmen described as caimans display the same general patterns: Lennox Lewis for instance – a British world champion of boxing – is unstoppable and resilient: “when he resists your jabs and grabs you like a caiman” (November 15th, 1999) there is nothing his adversary – in this case Evander Holyfield – can do. If from the sport domain we pass to that of politics, immediately the evaluative dimension is reverted. Two are the persons described through this metaphor. Félix Huphouet-Boigny was a village chief from Ivory Coast that led his country to independence through the idea of a “Francafrique”, a French Africa economically and politically connected to its former colonialist. The description of HuphouetBoigny through the figure of the caiman is isolated and thus I could find no other occurrence of it in any of my other corpora. The only other result generated after several attempts was an article describing the creation by Houphouet-Boigny of a reserve with 600 crocodiles and caimans in the centre of his hometown Yamoussoukro. When employed to describe Hophouet-Boigny, the caiman is a figure of excess (the grand caiman) that has built the “biggest church of the world” “an architectonic kitsch” of European elements parodying St. Peter in Rome – and that has spent on such humongous church a sum of money comparable to the compounded public debt of Ivory Coast. The grand caiman could not content himself with that so he persuaded the Pope to consecrate the church overcoming both the protests of the students calling for a wiser use of that money and the cautions of the Pope. The grand caiman displays his power, ignores or overcomes both local – the students – and external – the Catholic church – logics to establish his own rationality of excess. In the third case, the use of “caiman*” to describe Cesare Previti, the dangerous dimension of the caiman and its not being supposed to be here are brought to the 213


foreground (“ignoring the danger he caught the escaped caiman. It was, however, Previti”). Considering these three metaphorical macro-uses of the morpheme “caiman*” I have highlighted important variations due to the context, the discourse genre and the specific individuals involved. Many elements, however, are constant. The metaphor is always applied to a specific individual, only by explicit transference to a group, Pizzo’s water polo player. Moreover, the evaluative dimension of the metaphor is constant inside specific discourse genre: it is negative in politics and positive in sport. Such an analysis shows us that i) the comparison of Berlusconi to a caiman is not without precedent and ii) a common core of the metaphor is hard to find, it is possible, however to uncover a motive of loosely related and flexible semantic traits137. 2.3.4.2. The Birth of the Caiman (2002)

2.3.4.2.1.

A short introduction to Berlusconi (by Cordero)

Just as the morpheme “caiman*” has a history of both literal and metaphoric use, so does the morpheme “Berlusconi”. In addition to the overall picture previously sketched (§ 2.4.2.1), we have to consider that from the beginning of his collaboration with La Repubblica in 2001, the author of the metaphor, Franco Cordero, has made Berlusconi the main target of his political and legal commentaries. Cordero has thus created a specific and repeated representation of Berlusconi that we need to take into account. Cordero never uses Berlusconi's name, instead he writes Mister B. or Sir B. The most recurring semantic traits are: •

137

the effortless defeating a (potential) resistance: “he fights an inexistent

We would like to re-state that this analysis of the metaphor in the way it is explicitly supported and reverberated by the context is not an attempt at exhausting the meaning of the metaphor. On the contrary, it is an attempt at showing how the context reveals and constrains certain potentialities that given the metaphor in isolation would have not been described.

214


communism”, “he wins”, ...; •

the crossing of boundaries: he exploits the general dislike of the same old governments that granted him privileges, he collects electors from every field under the banner of the new: headless masses, ultra-moderate people, ultras, orphans, shipwreck victims, yuppies, fearful congregations, trans-crossers; ...;

•

extreme extroversion and excess: “he promises the moon”; “he exploits the general dislike”; “he carries the flag of anti-politics”;

• 2.3.4.2.2.

he leads an excessive farce that is very dangerous. The textual dimension of the caimanic smile: oppositions and narrations

With this construction of Berlusconi’s figure in mind, we can finally approach the birth of “Berlusconi is a caiman” on May 15th, 2002: “the only rank that can be attributed to him, anthropologically, is that of prince of the crooks – one of those crooks that charmed the peasants at the local fairs: fluvial speech, the smiles of a caiman, overabundant gestures, cheerful effects; nothing made him predestined to political dominance. If he were forced to participate in fair dialectical discussions (on TV discussions are rigged), he would drown in half a glass of water”138.

Cordero’s article is extremely complex as is usually his style (cf. Appendix A for the whole article). References to political vicissitudes in the Italy of the XIV century and in fascist Italy plus to linguistic strategies from “1984” by Orwell all co-exist and overlap by contiguity creating a dense semantic texture139. I cannot dwell too long on 138

139

"L' unico rango che gli competa, antropologicamente, è quello da principe dei «bagalun dl' lüster», gl' imbonitori che incantavano i contadini nelle fiere: loquela fluviale, sorrisi da caimano, gesto sovrabbondante, effetti ilari; niente lo predestinava al dominio politico. Costretto a partite dialettiche leali (sugli schermi gliele addomesticano), annegherebbe in mezzo bicchiere d' acqua." Having Cordero’s style in mind I can refer to a conversation between Antonio Gnoli and Cordero himself on the pages of La Repubblica (20/02/2002): “[G:] I would like to discuss your writing style. You compare it to the segment of a line. [C:] Does that not persuade you? [G:] I would describe it as closer to a baroque fold. [C:] I hope not the Baroque style in the XVIIth century! [G:] Let us then say that your writing style is serpentine, not at all linear, extremely rich in deviations. [C:] Look, the important thing is that below those stylistic efflorescences, there is a conceptual texture. This does not admit deviations. Either it succeeds in doing this or it crumbles down. [G:] Ok, but your articles, your essays,are rich in suggestions. [C:] Probably that is due to

215


an articulated textual analysis, given that my scope here is to map the evolutions of the metaphor. However, I will mention elements that are directly relevant for uncovering the dynamics of such evolutions. The article displays three elements in particular that while not being directly present in the previous quotation seem crucial in supporting the specific meaning-construction processes involved in the metaphor: 1. the gallery of characters to which Berlusconi is implicitly or explicitly compared 2. the narration of which Berlusconi and his caimanic smiles are a part and through which he gets defined. 3. The semantic ground on which the semantic form of the caiman smile emerges What do they imply for our understanding of “Berlusconi is a caiman” and of “the smiles of a caiman” in particular? Let us consider them briefly. 2.3.4.2.3.

A distinctive regime of comparisons

The characters to which Berlusconi is compared are of two kinds: important political characters and intellectually stimulating authors. Among the first kind we find Mussolini and Hitler. Berlusconi is compared to them because of his ambitions and of his political role. Those common traits cannot, however, hide that Berlusconi does not share their greatness, their real charisma and their direct military power. He is a crook [bagalun] – extremely successful – but still a crook. Among the second kind of characters to which Berlusconi is compared we find Bartolo, Tigrini, Orwell, Moore, Guicciardini and Montesquieu. These authors have all developed and communicated crucial insights onto human politics, communication and nature. The comparison whatever experience I have had in the tribunal.” [[G:] Tornerei al suo modo di scrivere che lei equipara al segmento della linea. [C:] Non la convince? [G:] A me pare invece più vicino alla piega barocca. [C:] Spero non al Barocco seicentesco! [G:] Diciamo allora che la sua è una scrittura sinuosa, tutt'altro che lineare, ricchissima di deviazioni. [C:] Guardi, l' importante è che sotto queste efflorescenze stilistiche ci sia la trama concettuale. Questa non ammette varianti. O riesce oppure frana. [G:] D' accordo, ma i suoi articoli, i suoi saggi, sono ricchi di suggestioni. [C:] Probabilmente in ciò influisce quel tanto di esperienza forense che ho svolto.”

216


with them is even less favourable for Berlusconi. The Italian prime minister is described as relying on and spreading a cultural void or, at the most as a character belonging to The Bold and the Beautiful, one of the soap operas that made Berlusconi's TV channels successful. Amidst these comparisons the metaphor of the caiman constructs through its reductive and caricatural potential the perfect opposition to the previous all-round figure, that even when negative – Hitler, Mussolini – had real charisma and greatness. I have mentioned earlier (§ 2.2.4.1.4) how metaphors can be articulated along a continuum between three regimes: separative, distinctive and confusive, according to how they articulate the identity of source and target, vehicle and tenor, that is the elements involved in the metaphor. Cordero's entire article weaves a refined distinctive

texture.

Different

historical

periods

(XIVth

century,

Fascism,

Contemporaneity), different historical figures (Guicciardini, Hitler, Berlusconi, etc.) and different domains (intellectual production, political dominion) reverberate on each other and provide insights on each other without that leading to their identities being confused and their differences disappearing. This is, in Bottiroli’s (1993) terms, the regime of distinction (cf. § 2.4.7.1 for an example of separative regime in metaphors for Berlusconi). The figure of the caiman, even when it is limited to its smiles, potentially participates in this very same regime and it is ready to be expanded upon. 2.3.4.2.4.

Agency and power: between actants and figures

The article displays what narrative semiotics calls a “narrative schema”140, that is, the 140

In structural narratology the depiction of a meaningful action is the basic structure of the narration. Representative of this position is A. J. Greimas (1970) who, using Propp (1968) as a starting point, puts at the core of his model of narratives the Narrative Schema: “an internal structure which assigns a general form to the action and which distributes a limited number of general roles to be played by the protagonists” (Bundgaard 2007). A narrative schema is, in its most current formulation (Bertrand 2000), constituted by three phases: manipulation, action and sanction. It involves interdefined actantial positions like destinant and destinee, subject, antisubject and object. These positions endorse value stances and thus create an arena where a conflict can be played out and a moral inferred, that compares the outputs of the actions and the enunciational sanction. Even if the cultural origin of the schema is sometimes made explicit

217


narrative macro-articulation of a meaningful action. Berlusconi is portrayed as a subject that is characterised by his actions, but whose power and whose competence to act does at the same time not seem to be there, to be real. “B. acts as the owner; he employs parliament and government as if they were Mediaset shops; allies, satellites, clients worship him and follow whatever he says, and he says some quite crazy things” [B. agisce da padrone; adopera Parlamento e governo quasi fossero botteghe Mediaset; alleati, satelliti, clienti, gli servono messa; asseverano qualunque cosa dica, e gliene sfuggono d' enormi.]. However, “B. does not have [charisma or militias], nor does Italy have autocratical institutions (special tribunals, a prime minister with special powers, etc)”. “If he were forced to participate in fair dialectical discussions (on TV discussions are rigged), he would drown in half a glass of water. Not even money explains why he is so strong: he has much, he is the richest man in Italy but he cannot buy everything; when you pay your electors one by one, you go out of money.” [Costretto a partite dialettiche leali (sugli schermi gliele addomesticano), annegherebbe in mezzo bicchiere d' acqua. Nemmeno i soldi spiegano perché sia così forte: ne ha da scoppiare, l' uomo più ricco d' Italia, ma non può comprarsela; quando paghi gli elettori uno a uno, rimani al verde.]. So in what does his power consist? This is explained very clearly in the same sentence in which “caiman*” appears: “the only rank that can be attributed to him, anthropologically, is that of prince of the crooks, one of those crooks that charmed the peasants at the local fairs: fluvial speech, the smiles of a caiman, overabundant gestures, cheerful effects”. He has the power of words and tricks, an extremely real power. How is this enough to give Berlusconi his prominent political position, since, according to Cordero, this illusion-inducing power did not make “him predestined to a political dominance”?

(“cultural grid of narrative organisation sedimented in the collective memory by tradition as a primitive” Bertrand 2000), Greimas often seems to consider it as an almost transcendental category of human understanding, an ahistorical formal mechanism and Bundgaard (2007) attempts to ground it in basic – and implicitly universal – perceptual phenomenology, thus strongly naturalising it.

218


Before acquiring his political dominance, Berlusconi had to pass through a previous phase that Propp (1968) would call the acquisition of a magical gift: “through dark alliances/bargains he acquired the device through which he reigns over souls: through images and sounds he evokes virtual worlds; monopolist of the TV channels he penetrates people’s heads, he inoculates affections and phobias, he makes fake ideas real and suggests decisions”. [Attraverso oscure intese s'era acquisito l'ordigno grazie al quale regna sulle anime: a colpi d'immagini e suoni, evoca mondi virtuali; monopolista dei canali, entra nelle teste, inocula affetti e fobie, insuffla finte idee, instilla decisioni]. This power is a power over distinctions: 1. Berlusconi annihilates existing distinctions: “since thought is infective, a good profilaxis begins by controlling the lexicon (Orwell, Appendix to 'Nineteen eightyfour'). Controllers reduce the vocabulary to elementary words, so vague that, according to the directives they can mean one thing or the opposite”. Additionally, Berlusconi claims to be inspired by Thomas Moore but Utopia, More's masterpiece, is an apology of “anticapitalism: shared properties, six hours of honest daily work serving the public good, a quiet life, intellectual entertainment, interior growth” – the opposite of Berlusconi's style and values. Opposites are overlapped and confused, pre-existing distinctions do not really matter, Berlusconi relies on and spread “a cultural void”. 2. Berlusconi at the same time creates distinctions centred on himself: Mussolini exploited the Italians’ fear of Lenin, Berlusconi invented a communist conspiracy that would pervade Italian politics, juries and intellectual discussions, in order to cover the position of saviour. His first instinct is to devour – countless former Catholic and communist activists constitute the backbone of his party. When this process of assimilation is not possible, then this resistance is reshaped as the enemy of all that is good and fair. Berlusconi thus assumes values opposed to culture, dialogue and constructive debate – that characterised instead Prodi's representation in the media (§ 2.4.2.1). He has the 219


power to annihilate and recreate distinctions at will thanks to previous and never fully specified “dark alliances/bargains” that enabled him eventually to enter in possession of at least three TV channels141. What is thus the role of the metaphor “the smiles of a caiman” in all this? The smile of a caiman is an expression tricky to interpret. It has a lot of teeth for sure. It seems to stand for openness while it is in fact a full display of dangerousness. The danger is there and hides in full display. Berlusconi continuously exercises his manipulative power through television and words and tricks. And this very same exertion and the display he puts on seem pleasant and open and persuasive. The caiman smiles as it occupies the position of the main character, prefiguring what could reveal his nature: the annihilation of differences in order to establish its own order. 2.3.4.2.5.

The caiman in its co-text

The metaphor is thus not a singularity that appeared out of nowhere; its semantic form gets shaped and led by a ground of previously established isotopies and at the same time it specifies and relaunches them. “The smiles of a caiman” as a semantic form appears on the ground of a bundle of recurring semantic traits (isotopies): •

/power/. Expressions like “regime”, “tyrant”, “master” and “overabundant” pre-figure /power/, especially in its crushing of distinction and overcoming of resistances.

•

/danger/

•

/water/. Expressions like “fluvial speech”, followed by “to drown” and “water”.

•

/out-of place/. Berlusconi is represented as the prince of crooks, where “crooks” is expressed with the only dialectal word in the article, “bagalun”. The prince of crooks does not belong to the social texture where he cheats

141

If we consider that in other articles Cordero explicitly identifies the dark alliances as Berlusconi’s dealings with previous political powers, amongst which Bettino Craxi, we see an interesting transfer of power from the political sphere to that of the media, that in turn, grant increased political power.

220


peasants. He comes from outside. What we see is thus an articulated interplay of these different isotopies with the figure of the caiman. The /water/ isotopy anchors the caiman to a figurative level making it a salient image more pervasive than its occurrence as a single expression would otherwise allow. The smiles, with their display of teeth, make the image even more salient. The power and the danger displayed in the smiles do not get acknowledged also due to Berlusconi’s diversity. His diversity and his power threaten to disrupt the local social and cultural texture, since he is able and willing to ignore and disrupt the local order and pre-established distinctions and norms, since he is able and willing to crush any opposition and norm to generate what Cordero calls “an amorphous Cesarism”. It can thus be said that although the metaphor can appear as marginal, being a single occurrence in a complex text, it is in fact perfectly fit to the complex semantic texture of the article. Following the general distinctive regime of the article, the metaphor “the smiles of a caiman” is entrenched with many crucial isotopies and it makes them concrete and suggestive through the imagery correlated to it. 2.3.4.3. The Growth of the Caiman (2003-2005)

We have to wait more than a year for the metaphor to appear again which it does in the following forms on May 20th, 2003 “the appetites of a caiman” [appetiti da caimano] and then again "smiles of a caiman” [sorrisi] (July13th 2003) and “jaws of a caiman” [fauci da caimano] (July 29th 2003). In 2004 more idiomatic expressions begin to appear and spread: “the habit of the caiman” [abito da caimano], “moves of the caiman” [mosse da caimano], “caiman tail, teeth and skin” [coda, denti e pelle da caimano] and so on. The following years see the multiplication of the appearances of the caiman under many guises, guises that are collected in the following table.

Body

Body in action

Extended Body / Caiman Style

221

Language and Cognition - Ph.D. Dissertation by Riccardo Fusaroli Jaws of a caiman Smiles of a caiman Habit of the caiman Choices of the caiman (2003)

(2002)

(2004)

(2008)

Tail of the caiman Appetites of a caiman Caiman effect (2004)

Syndrome of the caiman

(2004)

(2008)

(2003)

Skin of the caiman Moves of the caiman Army of the caiman Playing caiman (2008) (2004)

(2006)

(2006)

Teeth of the caiman Whisk of the caiman’s Mermaid [seduction] of Caiman spirit (2008) (2004)

tail (2006)

the caiman (2006)

Bite of the caiman Clothes of the caiman Cayman Style (2009) (2006)

(2010) Tones of the caiman Being in full caiman (2010) (2008) Caiman regime (2008)

Moral of a caiman (2009)

Aims of the caiman Taste of the caiman (2008) (2008) Caiman's tears (2008)

Culture of the caiman (2003)

For simplicity’s sake I limited the table to the expressions directly involving the morpheme “caiman*”. Many other expressions are actually used, from “baring his fangs” to “jumping on its prey” and they also follow the dynamics of evolution that I am going to illustrate now. The first thing to notice is that the morpheme “caiman*” when used to express the metaphor “Berlusconi is a caiman” is always presented through schematic linguistic patterns (smile of a *, culture of the *, etc.) that recur more frequently than the frequency of their components would lead us to expect. This idiomatic expressions display a constant structure: the caiman is presented in its singular form, they never present “caimans”; moreover, the expressions always foreground a specific aspect of the caiman. Initially the focus is on body parts and movements, later on the 222


foregrounding is of more abstract features like its style or its moral. The evolution of the foregrounded aspects proceeds in a way that is similar to my treatment of cognition in chapter 1: it initially conceives the caiman as a static body, then as an active body and finally it displays a more distributed conceptualisation in which the environment and cultural dimensions play explicitly a role in defining what a caiman is. The difference between “the jaws of a caiman” and “caiman style” makes this evolution quite evident. All of these metaphorical expressions in their difference contribute to profile a certain style that Berlusconi has, they contribute to define a general meaning of the “Berlusconi is a caiman” metaphor. Initially they do so by summoning vivid perceptual imagery, that is, by relying on salient physical traits and movements. However, the more the metaphor becomes entrenched in the use of Italian language, the less these physical aspects are directly mentioned and the more abstract feature are selected. We can also notice how the semantic traits relevant in the metaphor are always prepared by the more extended co-text, in ways that are analogous to the ones we have seen in the first article (§ 2.4.3.2.2). On May 20th, 2003 Cordero, who is again debating about the legal vicissitudes of Berlusconi describes Berlusconi’s “appetites of a caiman” preceded by the description of Berlusconi's successful attempts at increasing his income in spite of legal, economical or ethical obstacles as well as of his intentions not to stop his accumulation of positions and sources of incomes. On July 13th, 2003, Cordero again comments on Berlusconi's political behaviour at the European Parliament and describes his smiles as those of a caiman as part of a description of Berlusconi’s comical behaviour which according to Cordero, consists of obsessively repeating the same words, smiling as a caiman, laughing, jumping around and making obscene gestures. The excesses as well as the dimension of performance and fiction of the smile are thematised142. The dimension of danger is pushed to the foreground, to the point that as a final sentence Cordero feels the need 142

The smile of the caiman becomes thus the opposite of the crocodile’s tears that display something that is not there.

223


to add “we should take him seriously”. On July 16th, 2004 Cordero describes again a parliamentarian session. The Conflict of Interest is being debated. Berlusconi is at the same time prime minister and “the owner of commercial television channels, hegemonic master of the public television channels, arch-publisher, banker, merchant, insurance agent; his hands are wherever money flows”. But Berlusconi has “the jaws of a caiman” – he continues to hold on his powers and positions. Many of the semantic traits highlighted so far continue to appear in the background: the excesses, the danger, and the fictions. I could keep going here, but already all the elements are on the table. The general picture of the metaphor that emerges from the set of all the relevant articles of this period (2003-2005) can be defined along two lines. The first is that all the occurrences of the morpheme “caiman*” qualify types of behaviour that Berlusconi enacts either in the parliament or in the legal domain, most often in the legal court. The second is that the occurrences of the metaphor display a plurality of relevant isotopies (/excess/, /danger/, /fiction/, etc.). Each article – if we exclude Cordero’s Baroque ones – only foregrounds one of them. In other words, the metaphor in this period tends to emphasise the selection and construction of only one specific aspect of the figure of the caiman, without reflecting on the general figure and without articulating the different traits the one to the other. However, this is bound to change. 2.3.4.4. Caimanic Contagion (2006)

In 2004 La Repubblica starts reporting rumours about a movie by Nanni Moretti – an influential Italian leftist movie director – that will be called “The Caiman” [Il Caimano]. It is the beginning of a deep change for our metaphor. Most of the articles referring to the movie in this period have been read and consciously excluded from my working corpus, since their use of the metaphor is purely referential: it refers to the movie without further expanding on the metaphor. Such referential uses occur 33 times between 2004 and the beginning of 2006. However, in 2006 when the movie 224


reaches the theatres these use of “caiman*” reach the peak of 230 occurrences in a single year. This peak is not without consequences. The most evident one is the co-occurring proliferation of the metaphor. More and more authors start using the term “caiman” in a metaphorical acception. In 2006 it is not that hard to receive the label of “caiman” on the pages of La Repubblica. As in the period before 2002, sportsmen (Valentino Rossi, Alessandro Faiolhe Amantino detto Mancini, Casillas, etc.) and politicians (Giulio Andreotti, Achille Lauro, etc.) are mostly involved, with the exception of the media mogul Rupert Murdoch, labelled “the Australian caiman”. Without going into details – since many of these occurrences were listed and briefly analysed in the achronic account of “caiman” (§ 2.4.2.2) – the reader should be reminded of the double evaluative dimension in the metaphorical use: to be labelled a caiman is positive in discourse on sport, negative in discourses on politics. 2.3.4.5. The Caiman as a form of life (2007-now)

After the huge amount of recurrences of the metaphor in 2006 the numbers subside – although they maintain a higher frequency than in the previous periods. However, the use of the metaphor seems to have changed and evolved. We have seen that until 2005 “Berlusconi is a caiman” was a metaphor framed by idiomatic expressions and the explicit highlighting of both relevant physical features of the caiman and relevant semantic traits and in most cases situated in the debate of legal or parliamentarian debates. After 2006 and “The Caiman” by Moretti the metaphor is used in a much freer way. “Caiman” is used metaphorically without the need of framing it with syntagmatic structures like “the smiles of” or “the moves of”. For the first time the results present metaphorical uses of “caiman*” also in the plural form. More importantly, the semantic form of the metaphor changes as well. More physical features of the caiman are increasingly replaced by abstract qualities (the habit and the style of the caiman). Being a caiman begins to be a question of forms of life, instead of bringing one 225


isotopy to the foreground, that is, of selecting only one quality or feature as the actualised meaning of the metaphor. The different semantic forms to which the metaphor gave rise seem to sediment and overlap in the idea of an agency that locally breaks pre-existing (and global) boundaries to establish its own control. In this figure the notion of coming from another place (bringing a different logic) and the notion of excess (breaking local boundaries) are combined. Let us consider briefly three of the metaphorical uses of the plural form of the metaphor: “The Italy of caimans”, “little caimans at the airport”, “old vote-bringer caimans” and “caimans' tears”. These occurrences share the idea that caimans smoothen a space that was (or should be) striped by norms and customs, in order to establish their own local and personal advantage. From Italians keeping on voting for Berlusconi because they hope they will not have to pay taxes to Sicilian wannabe-godfathers and from politicians bending the rules to Italians skipping queues at the airport and questioning the need to extract their laptops at the security checks with the ensuing delays for everybody else. There are caimans everywhere and this form of life becomes more and more widespread in the pages of La Repubblica – and one would suspect in real life as well. What we have is thus the evolution of the meaning potential of the metaphors from a very unstable, but selectively structured according to the context, potential to a more abstract but structured schema that can be deployed effectively in a wide variety of contexts143. It might not be by chance that in the last months of 2009 the Financial Police discovered what the reporters called “the gang of the caiman” in the area around Napoli (cf. 16/10/2009; 17/10/2009; 29/11/2009). After some important arrests, there 143

Interestingly enough, after having presented these reflections and written these pages, an interview with Cordero got published in Il Fatto on April 25th, 2010. Cordero states that: “the metaphor of the caiman catches this biological monolithic power of the individual: he can only be this way. And he goes on guided not by reason but by reflexes” [La metafora del caimano coglie questa potenza biologica, monolitica dell’individuo: può essere solo così. E va avanti non a colpi di raziocinio, ma guidato da riflessi.] As in our reading, there is no reason, no recognition of external boundaries, of other logics, but only a self-absorbed destruction of any obstacle and opposition.

226


had been a momentary relapse of the Camorra, the local mafia, in Orta di Atella and the normal social texture of legality had been apparently shyly re-established. It was at this point that a new gang tried to create their own space of power crossing past Camorra orders and newly re-established legality. Local entrepreneurs were invited for a friendly visit. There a curious display was waiting for them: rabbits were fed to a caiman. Not a lion, not a big dog nor even a crocodile: but a caiman. The figure that we have analysed in the pages of La Repubblica penetrated either the actions and symbologies of the Camorra men or at least the words of the journalists telling us this story. 2.3.4.6. Changing what is it like to be a caiman: unsuccessful attempts at explicit diagrammatic manipulation

While the story we have seen is quite linear in its explosion of occurrences and sedimentation of a symbolic core across contexts and discourses, there are a few attempts at constraining and redirecting this course. We can qualify them as explicit diagrammatic manipulations, since they do acknowledge the metaphor, but they try to change the way it makes sense (the properties projected, the blending process and the afferent semes – to use the terms from some of the theories we have mentioned) These attempts are made by a couple of authors still criticising Berlusconi but also and especially by Berlusconi himself and his followers. We can individuate at least five different strategies: i) denials of the metaphor and its cognitive value, ii) denials of its applicability, iii) attempts at changing the semic molecule it constructs and iv) attempts at exceeding the metaphor either quantitatively or qualitatively. Let us look at this in detail. Whenever the metaphors “Berlusconi is a caiman” is opposed we see the call for more sobriety, for what Bottiroli calls a separative regime (§ 2.2.4.1.4), for the acknowledgement that metaphors are just rhetorical ornaments that distract us from the truth. A good example can be found in Giampaolo Rugarli's statement from March 31st, 2006: “caimans can be seen only at the movies or in the zoos” [i caimani 227


si vedono solo al cinema e nei giardini zoologici]. Along similar lines, we can observe denials of the applicability or of the effectiveness of the metaphor. For instance in setting up his new 2008 electoral campaign Berlusconi acknowledges the metaphor and implicitly defines it as a mere effect of communicative strategies – and not of his nature – and declares that from now on it will not be possible to apply it to him ever again: “I won't be the 'Caiman' ever again” [Non sarò mai più il 'Caimano'] (06 February 2008); “I do not want to be a caiman anymore” [Non voglio più essere un caimano] (11 February 2008). Slightly different is the strategy of an exponent of Berlusconi's party on March 28th, 2006: according to him what the left is doing is “to see caimans in the fountains of Milan” [vedere caimani nei Navigli]. The interpretive and cognitive power of the metaphor is not denied; what is denied is its referential appropriateness in the context. A more complex strategy is the attempt at changing the meaning making mechanisms activated by the metaphor and therefore the semantic form it contributes to construct. A very naïve attempt is made by Giuliano Ferrara, a former member of the Communist Party but since its start a loyal and outspoken member of Berlusconi's ever-changing party144: Berlusconi is “a generous and self-ironic caiman” [caimano generoso e autoironico] (15/03/2006). After a few derisory mentions this use is meritoriously consigned into oblivion. Mario Giordano – journalist and editor in chief of many of Berlusconi's family’s newspapers and TV news programs – employs a more sophisticated strategy: the metaphor is accepted but its internal dynamics are inverted. If the caiman has started out as being an agent insensitive to external barriers and crushing them, in Giordano's view the caiman gets an opposite property: Berlusconi “needs to be a thick skinned caiman to cope with all the crap the communists throw at him” [ha bisogno d'essere caimano pelle spessa per resistere alla merda che gli gettano addosso i comunisti] (12/05/2006). It is interesting in this case to see the inversion of agency: where in the normal 144

For an analysis in the terms of the cognitive pragmatist semantics here sketched of the evolution of Berlusconi's party, cf. Ferri & Fusaroli 2010.

228


metaphor Berlusconi is an agent breaking resistances and barriers, here he is portrayed as resisting attempts at shattering the truth and a correct representation of Berlusconi himself. However, the original semantic form of the metaphor is already too established for this attempt to make a lasting difference. Only a couple of mentions of this use are reported in La Repubblica, and a cursory research into more Berlusconi-friendly newspapers – admittedly constrained more limited access to their archives – Libero and Il Giornale did not yield any result. Slightly more successful but still not making a lasting difference are the attempts at exceeding the metaphor. One case of this being attempted consists in i) to acknowledge the potentially subversive power of the caiman and then ii) to apply the metaphor to political actors opposite to Berlusconi while iii) restating that Berlusconi is still most prototypical embodiment of the anthropomorphic caiman. This way Berlusconi can fight the subversive power of the very much dreaded – and yet largely fictional – Communist Left. This is what happens when Berlusconi declares: “I am the good Caiman that eats the red evil caimans” (02/10/2006); “we have to defeat the red caimans” (18/02/2006); “there are not only evil ogres like the communists in this world. There are also the good caimans: I am the caiman and I will eat them all! The good caiman that eats the evil Communists” (30/03/2006). A last interesting kind of manipulation that can be found with a certain frequency in the corpus is an attempt at mixing the metaphor with other metaphors. In his distinctive, but borderline confusive style (§ 2.4.3.2.2), Cordero creates sequences in which Berlusconi is: the caiman intended as the primordial Leviathan, alien to everyday life and human beings and without any consideration for external norms and distinctions that are not established by itself (tens for occurrences introduced for the first time in 20/05/2003), the caiman and the wolf with three throats, both animals characterised by being voracious and ferocious (13/07/2003; 27/05/2006; 12/12/2008; 24/02/2009) and the caiman and the shark again for the voracity (01/07/2008; 24/03/2009). There is also an attempt – made by Filippo Ceccarelli, a refined political scientist with an interest in political communication – at re229


qualifying Berlusconi's caimanhood during the new political campaign of 2008: “Berlusconi is a new animal with sharp teeth but also an ever-changing skin, he is a caimeleon!”. The occurrences of this new hybrid are only three and all by Ceccarelli. However, there is a promising – but yet not to be relied upon – increased frequency of the co-occurrences of the morphemes “caiman*” and “camaleon*” (e.g. 14/04/2010). The remarkable fact is, nevertheless, that none of these explicit manipulations leaves a significant trace in the social life of the metaphor. “Caiman” is still a negative term and a figure par excellence of the destructive power of a single agency over preexisting barriers and rules. 2.3.4.7. The Caiman in The Great Chain of Being

This hybridising of animals brings us to our last point in the description of the social life of caimans before discussing it in the light of our previous discussion of cognitive approaches to metaphors. The figure of the caiman, as CMT would predict is but one in a wider world of political anthropomorphisms, some inspired by the caiman, some pre-existing it. Berlusconi himself has previously been described through other animal figures. Let us briefly explore this animalistic world of politics. 2.3.4.7.1.

Berlusconi as an animal - from the big grass snake to the Hyena Ridens

Searching in the corpus for ways that Berlusconi has been understood and constructed through animal metaphors we can find only two such metaphors whose occurrences go beyond a handful: “big grass snake” [biscione] and “billionaire ridens” [miliardario ridens] that stages a clear reference to hyena ridens. While it is not possible to use my corpus to explain the difference in fates of these metaphors, it is still possible through textual and corpus analysis describe their different origins and evolutions and therefore to formulate some hypotheses. Michele Serra coined the expression “Billionaire ridens” on May 19th, 1994 in the pages of L'Unità. Serra refers to Berlusconi's love of jokes and pranks. However, we 230


can see some important differences. “The caiman” is born not as a simple appellation, it emerges in time through a constant interplay between different textual dimensions. Initially just an attribute, it slowly takes over the whole figure of Berlusconi. Billionaire ridens is an appellation that does not seem to resonate if not weakly through the texts of Michele Serra and then in the few occurrences on La Repubblica's pages. The two metaphors share the (initial) focus on the smile though through a different profiling of it. However, as we have seen in the textual and corpus analysis, in the case of the caiman the smile can endorse different isotopies from danger and power to fictionalised appearance. The expression “Billionaire ridens” does not share this fate – the affordances of the hyena figure do not seem to be exploited and the metaphor sticks to a role of rigid designator and it does not unfold its potential cognitive power. In the case of the metaphorical use of “big grass snake” it is a different story. The “big grass snake” is the symbol of the Visconti, an ancient dynasty of rulers in Milan. Their symbol is still reproduced on the shield of the city: a huge snake eating a baby. Berlusconi has replaced the baby with a flower and made this symbol into the logo of his financial masterpiece Fininvest. Fininvest is a company that owns eminent Italian companies (Mediaset, Mondadori, etc.) and that is in turn owned by 38 holders whose shares are in turn owned by Berlusconi and his family, in a complicated structure the documentation of which was lost “in a fire” (“Fininvest” on Wikipedia). Metonymically, the “big grass snake” is initially used to refer to Fininvest and, in a second moment, to Berlusconi. There are a few cases where this use assumes more explicit metaphorical tones and there is a semantic reverberation in the text. Let us see an example: “One may wonder about his true aims. On the other hand, he is a big grass snake” (12/03/2004). The article makes then explicit that Berlusconi is dominated by his hunger. Even the socio-political events surrounding the use of the two metaphors are similar (to the point that on June 27th, 1990 we find the occurrence “the government of the big grass snake”). However, the expression “big grass snake” did not spread as wildly as the “caiman” did outside of articles dealing 231


with Fininvest and its interests. The metonymical components stay strong, limiting the productive overlapping of the identities that we have seen in the case of the caiman metaphor. The use of “big grass snake” tends to fall under the separative regime, it does not extensively reverberate with Berlusconi's identity if not in very constrained domains. It seems reasonable to point to this reverberation, to the presence of a distinctive style, to an implicit recurrence in the co-text of the semantic traits relevant in the metaphor and to the schematic richness of the blend as crucial elements in the success of the metaphor “Berlusconi is a caiman” as compared to the others. This does not, however, exclude that other factors may have played a role, for instance Moretti's movie and the different salience of the grounding of the metaphor (grass snakes are probably less impressive than caimans. 2.3.4.7.2.

Caiman versus Shark and Other Animal Stories in Italian Newspapers

There are other animals on the political scene in Italy. Exploring the corpus it is possible to discover that Franco Marini is called “the wolf”, Lamberto Dini “the toad” or sometimes “the frog”. Francesco Cossiga is called “il gatto mammone”. PEOPLE ARE ANIMALS is, indeed, a pervasive conceptual mechanism on its own. However, it is possible to observe that “Berlusconi is a caiman” acts as a catalyst of new local metaphoric formations, both by contagion and by oppositions. For what regards contagion, in 2006 we see a proliferation of metaphorical uses of “caiman”: from Valentino Rossi to Giulio Andreotti. Regarding oppositions the two most remarkable ones are the one with Veltroni and Murdoch. Walter Veltroni, considered at that time as a possible alternative to Berlusconi, is a dog: “il veltro” described by Dante as a redemptive figure. And some of Veltroni’s supporters attempt to support this claim by reconstructing the genealogy of the Veltroni family and their shield, supposedly depicting a dog. Rupert Murdoch, instead competes with Berlusconi on the economic field. Murdoch is one of the most outstanding media moguls in the world and competition ensues when he tries to enter the Italian 232


pay-tv market that Berlusconi implicitly considers his own. Initially Murdoch is called “the Australian caiman”, however, as soon as the competition becomes more engaged Murdoch becomes “the shark”. Sharks and caimans are animals often associated in descriptions of voracious financial and (il)legal behaviour, compare for instance the following passage – also from La Repubblica - :“a gang of vile criminals made up of financiers, bankers, sharks, caimans, leeches and businessmen that inhabit the world of high finance” 26/02/2002. The metaphorical expression “shark” has some precedents in describing people close to Berlusconi. A recent example can be showed: “the 'mega-president' Augusto Minzolini, how he is confidentially called by the prime minister [Berlusconi]. The Caiman and the Shark, a perfect couple” (13/03/2010). However, in the battle with Murdoch the two animals are opposed in a titanic fight: “the Australian shark answers bite for bite” 20/03/2009). The two metaphorical constructions – the caiman and the shark – are based on a common ground: both Berlusconi and Murdoch were “born outsiders” but each of them by “being egocentric and expansionist” was able to “grow in power and richness without limits”. Initially they love each other, they cooperate with each other, although they envy each other: Berlusconi envies the global empire that Murdoch has built, Murdoch envies the political success of Berlusconi, that stricter Australian and American laws impeded him to replicate. The clash starts when Italian Television went digital. Murdoch is already in the market with a package of pay-tv channels: Sky. Berlusconi moves Mediaset – his package of both free and perview Television channels – in the market considering it its own by right. And then there is the fight, jaws against jaws and bite for bite. A textual analysis of the articles involved displays a mixture of embodied metaphorical terms – jaws, bits, etc. – and the more abstract traits I have already individuated in the caiman metaphor. The two moguls are described as voracious and as unable to suffer any barrier and resistance. The thing that afflicts Berlusconi the most in this fight is, according to Curzio Maltese (23/02/2009), his having to create boundaries by law to preserve the (his) status quo. But it is not an affliction strong in enough to stop the Italian prime 233


minister. Any means is good if it is of any use to maintain Berlusconi's own smooth space untainted by any rule that is not a direct expression of his agency. 2.3.4.8. Virgins vs. Cayman: the hidden role of verbal patterns

If the linguistic dimension of metaphors affords for implicit and explicit manipulation and evolution, there is also one last interesting phenomenon to be reported: the stability of the verbal patterns involved in the metaphor analysed tend to involve all other uses of the same verbal patterns. Given that Berlusconi has been associated to “caiman” this association seems to attract in its conceptual areas all possible occurrences of the morpheme “caiman” no matter their original meaning.

14 12

Occurrences

10 8 Berlusconi (virgin) Berlusconi (cayman)

6 4 2 0 1980 1985

1990 1995 2000

2005 2010 2015

Year

Figure 15 Diachro nic Evoluti on of the cooccurre nces of “Berlus coni” with “Virgin Islands ” and

with “Cayman Islands”

Berlusconi has often been associated with financial scandals and offshore fiscal paradises. During the electoral campaign in 1994 rumours started spreading: the 38 holdings owning Berlusconi's company, Fininvest, were not managed in a transparent 234


and legal way. The Virgin Islands were often mentioned as an example of offshore fiscal paradises employed to avoid taxes and uncomfortable laws. However, the more Berlusconi is associated with the figure of the caiman, the more the Cayman Islands are mentioned while the Virgin Islands lose their connection to Berlusconi.

2.3.5.

Assessing the life of caimans in society

It is now time to draw some conclusions from this long exploration of the corpus. Given that its dimensions did not allow a serious statistical treatment of the methodology used it has not been possible to formulate hypotheses and to test them, but it has been possible to employ automatic procedures in order to reconstruct the diffusion of the metaphors and its predecessors and therefore to identify the relevant texts to be analysed and compared. Not many studies are available in cognitively oriented literature on the life and carrier of a metaphor in society. Most studies are more interested in a philological (Allan 2009) or psycholinguistic perspective on the interplay between the literal and the metaphorical. Let us take a representative and authoritative example: Several stages can be recognised in the life history of a durable metaphor. When it is first coined, the only way to interpret it is to employ one’s innate metaphorical interpretive strategy, which is subject to a wide range of contextual and communicative constraints. Once a metaphor takes hold in a speech community and gets repeated sufficiently often, its character changes. First, its meaning becomes circumscribed relative to the freshly coined metaphor, becoming more determinate; second, it begins to be laid down as an item in the mental lexicon, so that in time, it can be retrieved in the same way as a literal expression; third, it begins a process of semantic drift, which can weaken or obscure its metaphorical origins.145 (Croft & Cruse 2004:205)

145

The quotation continues and focuses more on the phenomenological aspects of the use of metaphor: “At the beginning of its life, even if it is being laid down as an item in the lexicon, speakers are very conscious of its status as a metaphor, and they can recreate easily the metaphorical path of its derivation. As time passes, however, the sense of the expression’s metaphorical nature fades and eventually disappears (although it can be brought to life by means of Lakoffian elaborations etc.). Once that happens, the expression is no different from a literal expression, and only etymologists and historians of the language can recreate the path of

235


Another example – from psycholinguistics – is from the “career of metaphor” hypothesis by Bowdle and Gentner: individual metaphors evolve over the course of their lives from comparison – horizontal alignment between literal meanings – in the early stages to categorisation – vertical alignment between the literal target term and the base’s metaphorical abstraction – as they become conventionalised. Conventionalisation often results in local metaphoric categories, but it can also take the form of largescale conventional systems of metaphors (Gentner & Bowdle 2008:123).

What I have done instead – complementarily – is to look at the actual social life of a linguistic metaphor, at its changes in frequency, at the changes in the way the metaphor is expressed and supported through the co-text, finding out a different story whose articulation with the other perspectives has still to be investigated, since it would need a larger scale investigation on a high number of metaphors. This has led me to highlight how when the metaphor “Berlusconi is a caiman” was born the reader was hardly left to his “innate metaphorical interpretive strategy”. Previous linguistic usages were available. The text supported and led the reader's interpretation in multiple ways. This to the point that during the first few years of the life and growth of the metaphor, “Berlusconi is a caiman” is always strongly constrained via the co-text and expressed through idiom-like sentences that situate the actualised meaning of it along multiple dimensions. Only through the extreme explosion of its occurrences due to the movie “The Caiman” movie the word “caiman” acquires its autonomy. Without losing its textual and local dispersions, it displays a more stable schematic dimension: most occurrences of the metaphor explicitly actualise the figure of an agency that imposes its will over external preexisting barriers. The meaning of the metaphor is anchored to this unified abstract motive strongly enough to allow for a wider range of expressions without losing its understandability. In this process new idiomatic expressions are developed and both plural and singular form are used. “Caiman” becomes a contagious style and form of derivation. At some point along this path of change, the expression acquires a capability to act as a literal basis for further metaphorical extensions, which is not possible for a fresh metaphor”.

236


life keeping his prototypical centre in Berlusconi. This process happens through many voices and through many different genres. Such a temporal evolution allows us to move further with our pragmatist constructive critique of cognitive linguistics along at least three lines by i) investigating the embodied grounding of the metaphor and its evolution in time, ii) suggesting the role of the symbolic nature of language in crossing the line between the literal and the metaphoric and iii) articulating some additional possible roles of the social arena and of language in cognition, metaphorical and not. 2.3.5.1. Experiential Grounding of “Berlusconi is a Caiman”

2.3.5.1.1.

PEOPLE ARE ANIMALS -> RUTHLESS PEOPLE ARE PREDATORS

“Berlusconi is a caiman” is a family of expressions clearly related to the conceptual metaphor PEOPLE ARE ANIMALS and to RUTHLESS PEOPLE ARE PREDATORS, in particular. Behaviour and appearance of the caiman, according to predictions made by CMT, are privileged in the mapping, at least initially. So there is no reason to doubt the importance and effect of this conceptual formation. However, as we have seen not only is such a formation in many ways culturally shaped, but its linguistic expressions participate in a plurality of instances from idiomatic uses to co-textual leads and figurative regimes. In order to really figure out what was going on and to articulate the variations of the metaphor in a common motive, I had to examine the context, co-text and use, through a more detailed textual analysis: this brought me to deal with the construction of semantic forms, narrative structures and figurative regimes. All this highlights meaning construction processes that dwell upon but are only partially described by both conceptual integrations by sedimented conceptual metaphors. The selection of relevant features for the mapping is contextually and co-textually motivated and; framed by idioms, thus displaying a wide variation from one article to another. Narration also enters the scene, (§ 2.4.3.2.3) structuring our understanding 237


of the metaphor and organising the meaning of local metaphors even when apparently limited on a body part (the teeth of the caiman) in coherent Gestalten that comply with Optimality Principles. However, the diachronic investigation highlighted how the initial usage of the metaphor (up to 2006) privileged the projection of body parts and actions, employing only when the metaphor is widespread, more abstract properties: habits, style and so on. Indeed, initial linguistic expressions of the metaphor make explicit, although in a flexible way, what is deemed crucial in PEOPLE ARE ANIMALS (§ 2.2.2.3.2): appearance and behaviour. Later this evolves towards leaving space to more complex and abstract dimensions. What happens is that the metaphorical projection of skin, teeth, jaws, bites and smiles into Berlusconi is not without consequences. The necessity of translating from one universe of discourse into another pushes to the schematic properties of these physical aspects the foreground. Leonard Talmy spent many pages (Talmy 2000) highlighting the pervasiveness of force dynamics, basic images schemas that can be found in descriptions of physical motion, modal verbs expressing social dynamics and forces, psychological states and so on. Concerning the metaphorical expressions of “caiman” we repeatedly see repeatedly the crushing of resistance, what Talmy would call an agonist in motion overcoming an antagonist that is acting – relatively to the agonist – as a barrier. Once this element is repeated and varied in many contexts it becomes itself the starting point of the metaphor (caiman style). The caiman that did not originally have such an articulated conceptual structure is reshaped by these uses and this behavioural, social and moral style enters its definition explicitly and becomes its most salient trait. When we start considering one last element of the conceptual metaphor PEOPLE ARE ANIMALS, the “objectionable” trait that human beings tend to assume when compared to animals, these contextual and usage-bound dynamics are emphasised. Almost invariably we can see that a negative evaluative dimension is present in a political discourse, but when the metaphor is applied to the sport discourse, from football, to races, to water polo then a positive dimension is constructed. 238


A possible explanation for these observations focuses on the intermingling of practices, discourse and embodied motivations. The force dynamics schema we have previously highlighted unfolds in a discourse when the metaphor is used, a discourse that has its evaluative field already partially in place. There is a strong discursive tendency in contemporary chronicles on sport (cf. Demuru 2010, ch. 2) to emphasise the overcoming of limits, to crush the resistance of the adversary, to impose oneself and to focus on individuality. When the caiman figure is deployed here it fits this ideological structure perfectly, therefore acquiring a positive value. However, political discourse in newspapers and in La Repubblica in particular deals with a dimension of collectivity and social responsibility (for an analysis of this cf. Gangemi 2011, chap 1). This is implies certainly an idea of change, but through involvement and dialogue. This does not fit with the caiman style and therefore the negative evaluation it comes strongly to the foreground. Not even Berlusconi's welloiled machine of spin-doctors and propaganda is able to revert the metaphor effectively and assume the negative nickname as a positive characterisation (a strategy effectively described in Lotman 1985). On the one hand, I can certainly confirm the importance of the physically embodied motivations of the metaphor. However, these motivations do not seem to constitute an invariant layer that strictly determines more abstract, discursive and cultural domains. On the contrary, the selection of which embodied aspects are put to play, which figure in the foreground and which in the background, their translatability into the abstract domains – all these aspects are deeply discursive and defined in an interplay with the co-text, discursive rules and the structures of the more abstract domains and mental spaces involved. Conceptual metaphors are thus more like habits than like rules, types, or invariant and fixed conceptual structures. If this is already enough to show the importance of the social arena of language that was pushed forward in chapter 1, I can add one last point on how this arena contributes to the constitution of experience and concepts: the conceptual evolution of /caiman/ even in its non-metaphorical uses. The target (Berlusconi) reveals a 239


strong structure widely reiterated and articulated in our corpus and in particular in Cordero's production. On the contrary, the source (/caiman/) that should be the structuring element for the target has a much looser structure. Italian-speakers do not generally have any direct experiences with caimans. Even when the morpheme “caiman*” is used in a more literal way to refer to the animals it actually stands for /exoticness/. And at the end of my analysis, its overall use is co-textually supported by a dimension of agency and control more than by appearance-related traits. The schematic core of the metaphor and consequently of the term “caiman” is a core open to variations, crossing the literal and the metaphorical divide and calling for contextual and co-textual constraints. It behaves like what Visetti and Cadiot (2001) call a motive, that is, as a unifying principle participating in a history of usage, in discourses and transactions, that motivates the variety of uses of a linguistic unit and acting as a lead for further cognition. A motive i) is initially conceived as an unstable bundle of semantic dimensions, involving as such a dynamics of coalescences and of potential co-articulations, ii) constitutes a transposable organisation that is typical of a certain “phase” of the process of meaning making, iii) is finally defined by an open series of formulation that constrains the co-articulation and the conjoined elaboration of other motives and iv) co-implicates a repertoire of constructions, of collocations and of phraseologies (Visetti & Cadiot 2006:123)146

At the actual state of its career the metaphor involves a bundle of semantic dimensions as well as a general narration, a distribution of actantial roles and an action-oriented evaluation where the caiman can do, spreads, dominates together

146

“i) se concoit en première approche comme un faisceau instable de dimensions sémantiques, comprenant comme tel un jeu de coalescences, aussi bien que de co-articulations potentielles; ii) constitue une organisation transposable, caractéristique d'une certaine 'phase' du sens; iii) se “définit” éventuellement par une série ouverte de formulations, qui entraînent la co-articulation et l'élaboration conjointes d'autre motifs, et iv) co-implique un répertoire de constructions, de collocations et de phraséologies”. In other passages the author describe the motive as “une certaine forme de rassemblement ou d'unification qualitative, un 'parcours de l'instant', qui, si l'on pouvait transpore sur le plan de l'imagination, ne pourrait apparaître que comme une série chaotique et tourbillon ante de profils à peine entrevus et fondus les uns dans les autres” Cadiot e Visetti 2001: 114

240


with a call for a social and ethical evaluation in the discursive practice in which the metaphor occurs. The study of “Berlusconi is a caiman” is coherent with other existing attempts at conjugating the cognitive study of metaphors with a corpus methodology. The potential ambiguity of metaphors is often resolved by contextual constraints both in anticipating the metaphoric nature and the specific traits that will be relevant (Martin 2007). This happens at a wider semantic level but also through relatively fixed lexical and grammatical patterns (Deignan 2005; Stefanowitsch & Gries 2007). The target domain seems to strongly constrain the use of the source domain words (Gries & Stefanowitsch 2007). Metaphors display a great variability and change over time (Deignan 2005). However, my analysis with its attentive consideration of textual dimensions and the way they deploy and modify the mappings involved through time sometimes with lasting consequences, has constructed a case for spelling out a more complex conception of experience and of linguistic and conceptual dynamics. The analysis of “Berlusconi is a caiman” has directly opened up to the importance of the social dynamics of the linguistic “surface” and to the fact that these symbolic dynamics move in a space of Thirdness and generalities where more basic motivations do not lose they strength, but where the application of such motivations is mediated and unfolded through more cultural practices. 2.3.5.2. The Social Arena for metaphors

The analysis of “Berlusconi is a caiman” has shown how the social arena of language can constrain and expand the expression of conceptual metaphors and how it can display dynamics and evolutions that can be modelled as loosely repeated and distributed over time diagrammatic reasoning. Conceptual metaphors become habits in this sense, that is, adaptive and distributed patterns of meaningful interpretations and manipulations. Such habits are supported by exposure to language as structured

241


by use147. They are adaptive because they change in time through their being measured with political and mediatic events, with linguistic changes and with the history of use. They are distributed because rely not only on the individual, but on the mediascape and its history. In this sense, it could be argued as do Gibbs & Cameron (2008) that conceptual metaphors are best conceptualised as emerging self-organisation of (habit-based I would add) cognitive processes than as information-dense internal representations. Certainly my data can be interpreted in terms of the evolutions of a landscape of attractors (§ 1.4.5) where the different terms and semantic traits at play have to be seen as defined via bundles of semantic trajectories and as constraining the interpretive trajectories as well as the possibilities of further discursive constructions (cf. the unsuccessful attempts at manipulating the meaning of “Berlusconi is a caiman”) and where the conceptual metaphor RUTHLESS PEOPLE ARE PREDATORS is deeply situated in the discursive practice, etc. However, more work is due especially in investigating the neuro-physiological bases of metaphor processing over time in order to be able to argue further for the emerging nature of conceptual metaphors as well as to better understand the degree to which they are interiorised in long term memory. For now it is sufficient to have shown how linguistic metaphors participates in a plurality of not always intentional factors, of a plurality of semantic trajectories that go beyond the individual, relying on genre-conventions, on shared linguistic usages, on other people's utterances and so on.

147

A line of reasoning that I cannot follow here since it would require the construction of a different corpus but that would still fit with the current framework is one championed by Haser (2005). Haser shows how conceptual metaphors are not monolithic conceptual formations but looser formations of partially heterogeneous structures. PEOPLE ARE ANIMALS would thus only be a label covering heterogeneous and more local structures like RUTHLESS PEOPLE ARE PREDATORS, OBJECTIONABLE PEOPLE ARE ANIMALS, PEOPLE WE HAVE AN INTIMATE RELATION WITH ARE ANIMALS and so on that would construct blendings and distribute values in different ways.

242


2.4. Conclusions As it is always the case, more could be said about metaphors. However, it is time to draw some conclusions and to articulate the pragmatist framework we are developing further in order to turn to the third and last part of this investigation: the use of language as social coordination in experimental situations of joint perceptual assessment. We have seen that when debating metaphors in a cognitive linguistics framework there are different matters at stake: the origins of meaning and in general the mechanisms of cognition as well as the usage potential of idioms, words and morphemes and their role in structuring the conceivable linguistic and extralinguistic consequences of the ongoing cognitive processes and ways of conceptualising the world. All approaches seem to agree about the pervasiveness of metaphors and on their effectiveness and their structuring and reshaping of the human way of conceptualising the world. However, they do so from very different perspectives. CMT focuses on the origins of meaning and assumes an ontogenetic perspective

showing

metaphorical

projections

as

a

pervasive

conceptual

determination from embodied early interactions with the world to more abstract conceptual domains. CMT portrays the formation of conceptual structures as an ideal process where purely embodied experience precedes and determines social and linguistic experiences. Such an abstraction has been shown as excessively reductionist since even the emergence of image-schemas has to be situated in what Peirce would have considered a community-supported horizon. Such a notion of community – it has to be restated – should not be considered as a determining agency in itself, opening the door to cultural relativism. The horizon of a community can be individuated at least along three interrelated dimensions: 1. the individual awareness of a social horizon in which the individual operates 2. the implicit regulative effects given by repeated interactions between subjects belonging to a relevant social group 243


3. the vertical regulative effect due to the sedimented background of practices and more generally to cultural artefacts and norms inherited from previous generations (§ 1.2.2.4.3) These dimensions define a landscape of attractors, a network of partially predefined values that act as a pervasive ground on which the cognitive processes of the individual assume their orientation and their meanings. We have seen this in action in perception in § 1.2.2.1. In this chapter I applied this framework to the analysis of the emergence of image-schemas, showing how the size of the community is strongly related to the practices we are considering. In the case of primary and secondary intersubjectivity the primary dimension will be the implicit regulative effects given by repeated interactions between the infant and its parents. In the definition of the distance between the terms of a metaphor we will have a wider role of culturally shared norms and culturally defined artefacts, thus greatly increasing the dimension of the relevant “community”. CIT does not make any assumption about the different components of human experience and cognition, but it maintains a strong focus on the individual. Its focus is on blending, a general mechanism constituting “the way we think”, although CIT depicts blending as ruled by a never properly explained Gestaltic logic that results in blends at a human scale. It is the lack of focus on the constitution of the mechanisms through which a good blending is achieved and the lack of ecological settings in the metaphors considered (Brandt & Brandt 2005, Oakley & Coulson 2006) that prevented such a model from assuming a more distributed perspective and a sounder ground on a pragmatist theory of human experience, thus better embodying the joint consideration of form and content it heralds (Fauconnier & Turner 2002) and then forgets. Displaying at all levels the flexible and distributed nature of metaphorical processes, I have shown that it is possible to construct a pragmatist semiotic account in which elements from CMT and from CIT are constructively held together and developed. The basic human situation of being in and with the world is already constitutively 244


embedded within a social and cultural horizon. Abstract thinking – but more generally all semantic and conceptual domains – puts to play this concrete experiential ground, selecting out certain elements for certain purposes, negotiating and aligning them in the social arena. Therefore, new and/or more abstract experience is specified by a recourse to the basic human situation of being in the world. However, such a basic condition is embodied and social and at the same time and it develops via this constitution of new domains of experience. In this account the notion of habit as developed in chapter 1 plays a crucial role and seems to be able to more satisfactorily articulate the levels at which metaphors but more generally cognition are played: the linguistic, conceptual and neural levels. We have seen how Lakoff has attempted to turn CMT and the study of language into a Neural Theory of Language. What counts goes on in the neurons and their mechanisms, Lakoff repeatedly states. Such a conception is in deep contradiction with the pragmatist framework I have been constructing. Neural systems – in all their complexity that extends far beyond neural networks – are but a part of the story, since they are taken into and given a role through a distributed network the integration of which gives rise to cognition. Only through such networks do neural system accomplish a cognitive function intended as a function of mediation in terms of the structuring and manipulation of conceivable consequences. To re-tell CMT’s foundational story we could say that the initial patterns of movement, feeling and perception – to which regular patterns of neural activity become correlated and through which they are reinforced – develop as value-oriented patterns participating in distributed cognitive systems: primary and secondary intersubjectivity for instance in all their subtleties of coordination and joint attention in a culturally structured world. Habits – and thus image-schemas, and thus primary metaphors and thus conceptual metaphors – are not just neural patterns, but participate in wider horizons through which the neural patterns become part of meaningful activities. Image-schemas and conceptual metaphors are but habits of interpretation, more pregnant than others due to their ground in more usual experiences. However, they 245


are not monolithic bricks of our experience. They are more like a landscape of attractors to be enacted in situation. Language and more explicitly socio-cultural experiences widen such distribution. If already in the analysis of the constitution of image schemas and other basic conceptual structures language seems to have an effect, when we deal with linguistic behaviour and textual artefacts the complexity of the cognitive scenario is even greater. The analyses of linguistic metaphors and the more extensive corpus investigation of “Berlusconi is a caiman” have highlighted the fine-grained texture of habits at play at different time scales: conceptual metaphors, lexical history, genre-conventions, intertextual references, textual styles, etc. If certainly culture and language had to be born at a certain point and the enquiry into their origins is a worthy one, their appearance irrevocably changed the cognitive landscape. Meaning and conceptual structures now have to be conceived as constituted in the ongoing interplay of physical, biological, and socio-cultural constraints. The structures of the physical world in interaction with the biological specificities of human bodies constitute affordances, that is, pregnant possibilities of action, perception and interpretation. Linguistic patterns are motivated by such embodied and material affordances bringing them to a public and more manipulable scene. Individual experiences clash, get aligned and negotiated via language and the socio-cultural norms language allows. Such intersubjective and social linguistic dynamics retroacts upon our attentional structures and embodied and perceptual stance on the world. While the shaping influence of this interplay peaks during crucial periods in infancy and childhood it would be a mistake to underplay the role of conceptual evolution because the experiential of individuals and communities changes as well as the linguistic expression and the negotiation and alignment of it. In this ongoing interplay metaphors seem to play a crucial role at different levels: from that of basic conceptual formation to those of explicitly socially negotiated and aligned conceptualisation. Applying a pragmatist framework to such processes – from the pluralist conception of experience and mediation to the notions of habit and diagrammatic thinking – enables us to fully appreciate their cognitive role in a 246


distributed cognitive framework and their reshaping of conceivable consequences as well as the social horizon that vibrates in them.

247


3. Language as Intersubjective Coordination: An Experimental Setting 3.1. Introduction Imagine a dark room. Two participants in an experiment sit in front of two identical screens and the angle is such that they cannot see each other's screen. On the screens a sequence of two images is shown – identical on each screen. Six faint elements (Gabor patches) disposed in a circle on each of the images. One of the elements on one of the two pictures is slightly different from all the others, that is, it is an almost imperceptible contrast oddball. Each participant separately indicates in which image he or she believes to have seen the oddball. When their answers are the same, the task is repeated. When their answers differ, the participants are encouraged by a message on the screen to negotiate a joint decision freely and without a time limit (more details in § 3.2). This apparently simple procedure – designed by Bahador Bahrami (Bahrami et. al. 2010), a post.doc at Interacting Minds, CFIN, Århus – seems quite far from our everyday life and use of language. At the same time, since it involves free discussion, it also distances itself from the extremely controlled experimental settings to which much of cognitive sciences got us used to. These peculiarities notwithstanding – or maybe even because of them, as we will see – this paradigm discloses an incredible amount of surprising data and even more questions at all levels. In particular, it enables me to tackle from a new angle the central question of this dissertation: how does language – here in the shape of dyadic online conversation – affect and structure cognition – in this case visual perception and decision making?

248


3.1.1. From distributed pragmatics

cognition

to

experimental

Although chapter 1 was written taking into account the results of the other chapters, it was close to armchair philosophy, considering existing empirical findings and philosophical frameworks in order to sketch a vision of language that would integrate non-linguistic cognition without being reduced to it. Chapter 2 presented an application of such a vision, through which it is articulated and integrated through it the cognitive linguistic perspective on metaphors. The conceptual consideration of the issues at stake was complemented by textual semiotics and tempered by methodologies and perspectives from corpus linguistics. In this way I have analysed textual manifestations of metaphors, relying on a diachronic intertextual dimension to tackle the social horizon of language use. However, the choice of the corpus and the metaphor analysed made me map the developments of the metaphor “Berlusconi is a caiman” as a systemic evolution, showing the role of the social arena in shaping metaphor use. While individuals inventions and deviations from the norm were from time to time analysed, the focus was on the evolution of general and publicly shared tendencies in using the metaphor. To complete the sketch of the role of language in distributed cognition it is now time to move on to a more directly dialogic phenomenon where the (relative) autonomy of the individuals is more evident and the process of coordination of the different components of the distributed cognitive system comes to the foreground. Chapter 3 ventures into “the wild”, after a few theoretical considerations, testing and developing the framework in a setting that blends ecological ethnographical observations and experimentally measured conditions. The aim is to investigate how linguistically mediated habits are engaged and developed intersubjectively in real time and what influence they have on cognitive performance and therefore processes. This chapter can be seen as being located at the crossroads between three recent and thriving fields: Dynamical System Theory approaches to language (Raczaszek249


Leonardi & Kelso 2008), experimental pragmatics (Clark 1996) and experimental semiotics (Galantucci 2009). In other words it investigates linguistic dynamics in terms of attractors and trajectories under experimental conditions, paying particular attention to the interplay between and reciprocal modification of the affordances of the ongoing practice, the linguistic structures put to play and their consequences for the cognitive system and its performance. This intersection reveals phenomena and dynamics that are crucial for our framework. How are pre-existing linguistically mediated habits made pertinent and deployed? How do new habits develop? How does language effectively or ineffectively mediate the creation of a systemic intersubjective level for cognition? What is the role of symbolic patterns in that? How are they motivated, how do they evolve, how do they mediate cognition? Can we quantify this process in order to create a more productive dialogue between semiotics and cognitive sciences? Such questions are central in order to properly articulate and make productive the framework I am developing. I will not be able to answer all of them, but this study is meant to pave the way for such an enterprise and to show that some hypotheses can already be formulated and tested. A few pieces are still missing in our pragmatist framework before we can fully understand what is going on in the experiment that will constitute the centre of this chapter, there are still a few pieces missing in our pragmatist framework. The first is the notion of intersubjectivity and the process of coordination needed for different components to become parts of an integrated cognitive system. The consideration of these elements will lead us through the enactivist approach to social cognition. The second piece to be added is the experimental manipulation of the conditions of communication, that is, experimental pragmatics. Through experimental pragmatics it becomes possible to better understand the linguistic dimensions of social cognition and articulate it to the ongoing debate in cognitive sciences. However, there is still a gap in the literature in this field: a way of manipulating and quantifying the effect of 250


linguistic interactions, in all their complexity, in cognitive performance. After this rather theoretical introduction it will thus be time to go through Bahrami's paradigm. Initially linguistic interactions were considered a black box148: the only thing that mattered was the average effect of communication calculated by comparing joint performances with individual performances. I will thus proceed to unpack the black box, analysing how linguistic interactions unfold and how a more qualitative approach can lead to a quantitative analysis. Such a quantitative approach will make it possible to unveil a direct correlation between linguistic patterns and joint performance and to contribute to articulate a more fine-grained framework for analysing the role of language in cognition.

3.1.2. Distributing and integrating cognition: the role of intersubjectivity Up to now my gaze has rested on the systemic nature of distributed cognition, on how cognitive systems are indeed systems constituted by heterogeneous parts. However, this has left the problem of coordination and partial adaptation in the background. How do these parts come to participate to a coordinated system? Moreover, when different agencies are participating in a cognitive activity they do not necessarily merge into a system. They become coupled, but as if dancing, not locked in that they maintain a certain amount of autonomy like for instance the possibility to disagree. This problem has generally been neglected. On the one side, extended mind approaches leave the intersubjective and social dimension of cognitive systems in the background. On the other, more socially oriented distributed approaches like Latour's actor-network theory are more concerned with highlighting

148

It has to be noted that this presentation divides two distinct moments in the research for simplicity's sake. Already during the first round of data collection, I started a discussion with Bahador Bahrami. That discussion was followed by a few meetings with him and Kristian Tylén during which the role of language and how to analyse and manipulate it had been put on the table.

251


the way nodes of the network are holistically constituted by the network itself149. Even when intersubjectivity is thematised a narrow focus on agreement and shared understanding predominates, therefore limiting studies of intersubjectivity to analyses of consensus-oriented activities and processes of unification concerning the participant’s subjectivities. This strict focus on agreement results in implicit forms of reductionism: intersubjectivity viewed as a state of symmetry, intersubjectivity reduced to individual subjectivity (for a review, cf. Smolka et al. 1995), joint activity regarded as a simple sum of individual activities and joint activity progressing from heterogeneity to increasing symmetry between the individuals' perspectives150. In Chapter 1 intersubjectivity was defined as a dialogic process of coordination shaped by social norms, or, to use Matusov's definition: “Intersubjectivity is more usefully defined as a process of coordination of individual participation in joint sociocultural activity rather than as a relationship of correspondence of individual’s actions to each other” (Matusov 1996: 26). In § 1.2.1.5 I debated the issue of who is the subject of cognition. The further discussion of intersubjectivity will allow me to better integrate the role of individual agencies, their coordination and distributed cognitive processes.

3.1.3. Between synchrony, repetition and complementarity: how does landing a plane differ from dancing the tango? A pragmatist approach to cognition has to develop tools to describe the processes through which cognition is integrated and distributed via processes that involve

149

150

It has to be said, however, that if we were to take a deeper look at Latour's writings an alternative vision would seem possible. Actor-networks seem absolute and already there only because there is a tendency to hide the work of mediation through which these networks are constituted. Latour is very attentive to the progress of construction of science, an analytic perspective in which debates and controversies are crucial to expose the underlying practices of social mediation. Actor networks are the effect of "agonistic encounters" (Latour 1990) between different conceptions of the world and between opposite theories embodied in a plurality of social practices, technologies and material representations. This perspective is also reiterated in the predominance of the Naming Game and of the Category Game in experimental semiotics and pragmatics.

252


synchronisation

and

complementarity,

agreement

and

disagreement.

The

components of a cognitive system are engaged and integrated via an activity in which their heterogeneity, the different affordances and perspectives are coordinated in a process that could effectively be metaphorised as a dance. Certain activities require the difference be minimised, others involve difference as a productive component of the system. We can draw an interesting comparison between the cognitive distribution involved in landing a plane (inspired by Rosero, Lecusay and Cole 2010) and in tango dancing (inspired by Kimmel 2010 and Kirsh 2010). Landing a plane requires high levels of redundant constraints on the possible ways of interpreting and responding to the set of changing conditions encountered along the descent trajectory (Hutchins 1995). This is both ensured by a shared professional perspective (or vision, as Goodwin 1994 would call it) between pilot and co-pilot and by the constraining role of the cockpit environment. The coordination needed to land a plane is carefully planned in order to minimise ambiguity. Pilot and co-pilot have undergone years of training to ensure a high level of agreement about what is going on, both in their perceptions and in their actions. The cockpit environment itself is structured so as to keep this convergence as tight as possible, strictly constraining the correspondence between actions of the pilots and the various instruments on which they rely. This tight convergence is not only made possible by the material disposition of the cockpit, but it is also supported and constrained by do the social practices taking place in the cockpit: the procedures for landing an airplane are so minutely scripted and the correspondence between the pilots' actions and the instrument-centred representations so tightly coupled that the activity of landing the plane becomes highly automated. The attention of the pilots is mainly directed at enacting redundancy, that is, checking and cross-checking rather than at performing the impossible task of jointly computing and communicating the various metrics required on a second-by-second basis along with the necessary adjustments to the plane's trajectory and speed. 253


Tango argentino, on the other hand, emphasises complementarity as opposed to synchrony and redundancy. Tango argentino – to a larger degree than ballroom tango – is an improvisational pair dance with a clear distribution of roles: a leader and a follower. Tango argentino involves intersubjective coordination that crosses and defines the single moving body and more limited kinds of bodily interaction like cooperating to carry a load because the actions of both partners continuously interdefine each other. Analogously, tango argentino goes – in its way of constructing intersubjective coupling – beyond other examples of co-regulation like walking past a stranger on a narrow sidewalk, primary mother-child interaction (§ 3.1.4.2) or greeting an acquaintance since it comprises a clear distinction between the dancers' roles and a more sophisticated system that needs to be trained over years to work fast, in proper form, and allow for creativity. Overall, we could say that in order to dance the tango argentino at least two different levels of competence have to be developed: i) a repertory of specific step sequences, their elements, and the nodes at which these are connected as well as attentional, postural and kinetic techniques for “tango-like” movements and ii) intersubjectivity skills both in sensing the partner and in displaying coordination clues. This distinction in two levels emphasises the need for specific coordinating abilities that cannot be reduced to controlling one's own body. We thus see a continuum in the distribution and integration of cognition along at least two dimensions: •

The degree of convergence and the degree of complementarity of action and interpretation

•

The degree of dissolution of individual agencies in the system and the degree of coupling of components that do maintain a difference of perspective and orientation of actions.

In order to prevent a possible misunderstanding I need to specify that coordination 254


as well as distribution and integration of cognition happens even when the single individual or component does not share goals. In a heavily trafficked city pedestrians, traffic officers and car drivers all have different goals and yet some coordination takes place and allows a certain degree of successful interaction. Analogously, in a university department groups of students, administrative personnel, teachers of different ranks and researchers have to achieve a certain degree of coordination, even in the clear divergence of at least part of their goals, in order for such goals to be fulfilled and for the department to function. To better model such differences it is necessary to focus on the processes of distribution and integration found in cognition and on the ways the different components get engaged in these processes and unfold them. It is thus time to zoom in on the enactive approach which at least in the version developed by Di Paolo and his colleagues extensively reflected on the relation between organisms, coordination dynamics and sociality.

3.1.4. The Enactivist Perspective: organisms and social interaction The basic assumption of enactivism is that there is continuity between life and mind, that is, that the basic concepts needed to understand the organisation of life are central to a proper scientific understanding of the mind. Another and more radical way of putting it has been advanced by Evan Thompson (2007:129): the basic concepts needed to understand human experience turn out to be applicable to life itself in a way not too distant from Peirce's cosmological conceptions (cf. Favareau 2007). The enactivist positions that spring from the work of Varela (1979, 1997, Varela et al. 1991, di Paolo 2005, Thompson 2007) have been developed as an interesting alternative to cognitivism, an alternative that attempts to do justice to human experience and to be biologically grounded at the same time. 3.1.4.1. Back to the organism as the ground of sense-making

255


The starting point of enactivism is the model of metabolic autonomy that constitutes life, that is, autopoiesis (Maturana and Varela 1980). An autopoietic system is any self-organising biological system in which the constituent processes (i) recursively depend on each other for their generation and their realisation as a network, that is, by their activity continuously generate and realise the network that produces them, (ii) constitute the system as a unity in whatever domain they exist and (iii) determine a domain of possible interactions with the environment (cf. Varela 1979: 55; Weber and Varela 2002:115). Organisms are thus defined as autopoietic systems, systems that produce and distinguish themselves from the environment in a way that is not caused by external forces. Given that organisms produce and maintain themselves, they are in a constant precarious situation where external forces and metabolic consumption threaten to dissolve them and require both resistance to variations and continuous overcoming of them. In other words organisms need to be robust in order to be able to sustain a certain range of perturbations as well as a certain range of internal structural changes before they lose their autopoiesis. However, robustness is often not enough. An organism on a cliff preserves its own autopoiesis, even if the situation is dangerous. Even if being on a cliff makes self-preservation less likely in the long-run, autopoiesis is preserved even if the organism falls, at least until the organism crashes into the ground. Therefore the notion of autopoiesis needs to be integrated by some mechanism able to determine and constrain the conceivable consequences of the ongoing behaviour. Let us consider another example, one used by Varela on several occasions (1991, 1997): bacteria put in a sucrose solution. Bacteria are able to detect or discriminate the differential sucrose gradients and orient themselves such that their propulsion will maximise the exposure to sucrose. There is nothing intrinsic to the sucrose molecule that makes it "nutritional" or “positive”. Rather, it becomes so only in relation to the nature of the bacterium's specific metabolism151. As an autopoietic 151

A view that could be criticised, since there are differences between sucrose and helium for instance.

256


system bacteria select some particular physical correlations – the presence of sugar – as relevant as opposed to other chemical compounds. Sugar is an essential nutrient that keeps the bounded network of processes of production going, and it is defined by the autopoietic system as relevant for the possible interactions with the environment. However, bacteria do not simply swim around randomly in the sucrose solution but discriminate between different gradients of it in an oriented way. They swim up the gradient and they detect changes and not only states. This requires the ability of the system to monitor and regulate itself, its internal processes and its interactions, in order to appreciate graded differences between otherwise equally viable states. Bacteria display a special way of being tolerant to challenges, that is, they actively monitor perturbations and compensate for their tendencies. Bacteria thus possess what enactivists call “adaptivity”, the ability to generate a normativity for themselves, that is the ability to discriminate within their current set of viability conditions by attributing different values to them, value that depends on the current state of the system. Bacteria thus do not simply go after sugar, but they are able to detect in which direction the sugar is most likely to be more concentrated and to swim in that direction. Adaptivity is a basic appreciation of conceivable consequences of behaviours and behavioural patterns given the actual state of the system. Adaptivity is a system's capacity, in some circumstances, to regulate its states and its relation to the environment with the result that, if the states are sufficiently close to the limits of its viability: 1.

Tendencies are distinguished and acted upon depending on whether the states will approach or recede from these proximal limits and as a consequence,

2.

Tendencies that approach these limits are moved closer to or transformed into tendencies that do not approach them and so future states are prevented from reaching these limits with an outward velocity. (Di Paolo 2005:438)

Sucrose is easy to decompose in order to get energy out of that. Helium is not. At the same time even substances seemingly unsuitable to life have been found as nutritive to certain bacteria, cf. the recent discovery of bacteria living on arsenic. Anyway, to overcome these difficulties, a more pragmatist inspired perspective could be employed, arguing that “nutritive” and “positive” are relational valence where both the bacterium and the molecule play a role.

257


If autopoietic activity defines a purpose and a perspective, that is, an asymmetry between organism and environment, it is adaptivity that allows the system to anticipate and act upon that which it needs for its self-constitution and that which might threaten it. Some means of adaptivity are: self-monitoring, control of internal regulation, and control of external exchanges. In this way the bacterium can swim up the gradient and we can move away from the cliff. It has to be noted that such selfgeneration becomes a source of norms: different events will contribute differently to its continuation, some enhancing it, others putting it in danger. Adaptivity introduces a distinction between good, neutral and bad ways of realising autopoiesis. Through adaptivity the system is able to make sense of more indirect forms of coupling with the world and sustain the requirements of several organic norms. Through the notion of adaptive autopoiesis, we reach the core of enactivism: the notion of sense-making. Autopoietic adaptive organisms can be said to bring forth significance or relevance through the viable structural coupling of an autonomous system with its environment. In other words: the world of an organism is primarily a context of significance, significance that is defined in relation to that organism's particular manner of realising and preserving its precarious identity. Autopoietic adaptive systems indeed construct the environment as “a place of valence, of attraction and repulsion, approach or escape” (Thompson 2004: 386). The sucrose solution and its gradients are meaningful to the bacterium in the sense that they are enacted, that is, selected and acted upon according to norms that take into account the internal states of the organism itself. The environment makes sense via selfmonitored interactions aimed at improving the conditions of the organism. Agency and eventually meaningful experience is grounded on the ability of the organism to transform contingent environmental modifications into opportunities for controlling the internal biological compensatory processes that are necessary for the well-being of that same organism. Such ability, indeed, operates a shift from symmetric structural coupling – system and environment influence each other without loss of 258


viability – to behaviour, by definition asymmetric, since the organism originates the regulation of structural coupling. If this is valid for bacteria, the evolution of organisms with more complex form of perception, action and emotion, gives adaptivity – and therefore sense-making – the chance to acquires a wider range of possibilities and a temporal stance. Animals can appreciate right now the danger that is impinging on them from a distance. This is the origin of a special relation with the world, that of perception and action, which is charged with internal significance, and hence with the development of an emotional dimension (what might have been an inner life of need and satisfaction now becomes rich in possibilities such as fear, desire, apprehension, distension, tiredness, curiosity, etc.). (Di Paolo 2009:17)

The above is true to the point that at higher level of biological complexity we can observe migrating species that travel long distances to find nourishment, enduring long periods of hunger and navigating through difficult landscapes in the meantime. If traditional cognitivism saw the mind basically as an information processing system that manipulates mental representations of the outside world by means of specific rules (§ 1.2.1.3), enactivism sees the mind as rooted in the nonrepresentational capacity of an autonomous system to self-regulate its own situation in an environment in accordance with its capacity for sense-making. Cognition is sensemaking in interaction: the regulation of coupling with respect to norms established by the self-constituted identity that gives rise to such regulation in order to conserve itself. 3.1.4.2. The minimal ground of social interaction

These fascinating positions, however, still have to face a main problem: the “cognitive gap” (De Jaegher & Froese 2009). Enaction seems to work fine for unicellular organisms as well as for the minimal cognitive processes of insect-like behaviour (Beer 1995, 2003, Brooks 1999). However, it seems inadequate for human cognition to which representation and a plurality of value systems are crucial. The 259


enactivist answer, crucial for our arguments, is to introduce an intersubjective and social horizon into enaction and the consequent development of the notion of participatory sense-making (De Jaegher & Di Paolo 2007). The enactivist procedure for tackling these issues is analogous to the one we have already seen in grounding sense-making: how minimal can we go in defining sociality? The solution proposed relies mainly on two key experiments and on the subsequent robotic simulation of their results. The first is a simple experiment of minimal interaction by Auvray and colleagues (2009) who investigated the role of inter-individual interaction in organising appropriate task-solving behaviour under minimal experimental conditions. Two participants were asked to locate each other in a one-dimensional virtual environment. The participants were able to control their avatar's movements in the virtual space and they receive an all-or-nothing tactile feedback to the finger whenever their avatar is overlapping any other object within that space. They can encounter three types of objects: i) a static object, ii) the avatar of the other participant and iii) a ‘shadow’ copy of the other’s avatar that exactly mirrors the other’s movement at a displaced location. The task is non-trivial because all of these objects are of the same size and thus generate the same all-or-nothing tactile response. The only way to differentiate between the objects is through the distinct kinds of interaction dynamics that can result from their encounter. Nevertheless, participants did manage to locate each other successfully most of the time, despite the minimal conditions under which they had to complete the task. This is the case because ongoing perceptual crossing – i.e. when both avatars mutually interact with each other, affords the most stable interactive situation under these circumstances. “The surprising result of the study is thus that even though the participants ‘failed’ to achieve the task individually, i.e. there was no significant difference between their probabilities of recognising the other’s avatar and the other’s shadow object, they managed to solve the task together, finding most of the time the other’s avatar and 260


ignoring their shadow” (Auvray et al. 2009: 39). What this experiment shows is that even at these minimal conditions it is possible to combine human actions, to operate a cognitive integration between several acting organisms and in this way to create emerging behaviour and cognitive processes that involve more than one individual. The second experiment initially performed by Murray and Trevarthen (1985) and then more rigorously reproduced by Nadel and colleagues (1999) employed 2-monthold infants animated by their mothers to engage in coordination via a live double video link—mother and infant are each in a different room and interact with each other via a TV screen. In this condition the infant is lively and interacting via what is called bi-directional proto-conversations (Trevarthen 1979, 1993; Reddy 2008) consisting of smiles and vocalisations. However, when the live video of the mother is replaced with a video playback of her actions recorded previously, the infant becomes distressed or removed. These results indicate that 2-month-olds are sensitive to social contingency, that is, to the mutual responsiveness during an ongoing interaction, and that this sensitivity plays a fundamental role in the unfolding of coordination. As I have argued in the previous chapters the discovery of this primary intersubjectivity changes our perspective on the ontogenesis of the individual cognitive system showing how embedded it is in interactions that are not reducible to individual components. A few critics – Gergely and Watson (1999) and Russell (1996) – have suggested that these experiments only pushed the problem back to a more basic level and that an innate cognitive module aimed at detecting intentionality and at processing agency could explain these results. This criticism has been at least partially refuted by Froese’s and Di Paolo’s (2008) simulation of Trevarthen's results with simulated agents with a minimum amount of internal information processing and no preprogrammed way to deal with intentionality and agency. These agents were designed to cross their sensors as far away from their starting positions as possible, a task that requires mutual localisation, convergence on a target direction and movement in that 261


direction, while not losing track of each other. The agents were able to sustain patterns of coordination with reciprocal and rhythmic crossings of the sensors, a robotic proto-conversation. Moreover, coordination broke down every time an agent was forced to interact with a playback of movements from a previous, successful trial. Agents interacting with such a non-responsive “partner” did not have the capacity to generate and sustain the kind of oscillatory behaviour necessary for coordination: it is the interaction that does the trick, exactly like in Trevarthen's experiment. From these results Di Paolo and his group infer that at least in principle social interaction does not derive from individual social skills but on the contrary social skills and social cognition emerge from more basic intersubjective interaction. Such minimal interaction is defined as: the regulated coupling between at least two autonomous agents, where the regulation is aimed at aspects of the coupling itself so that it constitutes an emergent autonomous organisation in the domain of relational dynamics, without destroying in the process the autonomy of the agents involved (though the latter’s scope can be augmented or reduced) (De Jaegher & Di Paolo 2007: 493).

It has to be observed how the notion of coordination is necessary but not sufficient for social interaction to take place. Coordination is a phenomenon observed both in physical and biological systems (Buck & Buck 1976; Kelso 1995) in which the behaviours of two or more systems that are in sustained coupling or have been coupled in the past or have been coupled to another common system become non accidentally correlated. In other words the behaviours of the systems start to constrain and influence each other in real time to the point that under a certain respect they can be described as one system. However, in a social interaction the autonomy of the interaction does not imply that the coupled systems lose their autonomy completely. Social interactions are grounded in individual agencies and open up new possibilities for sense-making without their participants losing their identity as individuals.

262


On the one hand the interaction process is organisationally closed. That is, its organisation is characterised by processes such that 1. processes of inter-individual coordination are related as a network so that they recursively depend on each other in the generation and realisation of the processes themselves 2. they constitute the interaction as a unity recognisable in the space of relation dynamics On the other hand, such coordination can be extremely precarious, engaging and disengaging, whereas zones of absolute coordination are followed by zones of only relative coordination. As a consequence of the dynamic nature of this process we directly perceive the other as another subject in its own right and without having to engage in inference or simulation in order to perceive her as such. Of course sometimes it helps to reason about the actions of others or imaginatively to put ourselves in their shoes so as to understand their intentions better152. But this typically happens only under not so usual circumstances when normal social interaction breaks down, and even then this does not mean that we question the other's status as a subject as such. Such a coupled system view of social cognition focuses on reciprocal interaction and recurrent feedback loops. It is the active process of engaging with others itself that gives rise to our understanding them. Social interactions are not just the coordination 152

This is what happens according to the two mainstreams of social cognition: Theory Theory and Simulation Theory. Theory theory (TT) claims that we explain and predict another person’s behaviour by relying on an innate or acquired theory of how people generally behave and of the mental states such as beliefs or desires that cause their behaviour (Premack and Woodruff 1978; Baron-Cohen et al. 1986; Antonietti et al. 2006). On the basis of our theory of mind, we make inferences about others’ mental states. Simulation theory (ST), on the other hand, claims that we have no need for a theory like this because we have an inner model that we can use for simulating another person’s mental states, and this model is our own mind. Thus, we model the beliefs and intentions of others whom we deal with as if we were in their situation, or as if we were them (Dokic and Proust 2002; Gordon 1996; Goldman 2009). One could say that according to these theories the person who perceives another does not actually interact with him or her but deals with internal models or simulations of her actions.

263


of two embodied agents but also a mutual incorporation where joint sense-making becomes possible. In pragmatist terms via social interactions the space of meaning that is created is not simply what we are ready to do in the ongoing practice but it includes a plurality of possible perspectives made possible by the systems we are coupled to and the ones we were coupled to. Moreover, this space of meaning does not include only the plurality of virtualities due what I am doing now but also the plurality of virtualities due to the other possible actions I could be doing. In other words, it is through social interactions that we can conceive of the Dynamical Object. Or, put in a Husserlian way: Only when a subject experiences that the same object can be experienced by several subjects, and that it is given for them in various perspectives, that the subject is in a position to realise that there is a distinction between the object itself and its appearance, its being-for-me (Zahavi 1996:38-39)

Crucially for the issues tackled in this chapter objects and representations seem thus conceivable only thanks to a mix of convergence and complementarity. The different subjectivities at play do not simply share one same profile of the object but they share a plurality of profiles unfolding in the interaction and in a plurality of possible practices. That is, thanks to the experience of other subjectivities we get to conceive of objects at the same time as unitary and as constituted by a wide plurality of possibilities. The possibility of this “open intersubjectivity”, that is, of the experiencing others as others remains highly contentious even with regard to our closest primate relatives (Tomasello et al. 2003) and lacks scientific support for most other species (Tomasello 1999). Therefore, it could be inferred that it is this unique human ability to coordinate that enables the jump from bacteria and insect-like cognition to higher level cognition through the construction of shared objectivities or at least shared representations. 3.1.4.3. Social enactivism under a pragmatist eye

264


The enactivist account of cognition and intersubjectivity is not without problems and not without gaps. Social behaviour is observed in different animals, from social insects to wolves, so it could constitute a way of bridging the cognitive gap in a continuous way. What Di Paolo and colleagues are doing, though, it is to extrapolate a very specific kind of human interaction and remove as much as possible the social normativity in it. To the point that it is possible to find the same kind of abstract form of coordination also in physical and biological systems. However, then they define this as the ground for the perception of others, without first investigating the possibility of finding it in other organisms (continuity) and the differences separating humans from animals. The cognitive gap is still wide and to be further investigated. However, the discussion of the enactivist approach gives me the possibility both to restate some of the crucial aspects of a pragmatist approach and to bring to the foreground the issue of coordination and autonomy. The main difference between the enactivist approach and a pragmatist approach is that enactivists try to focus as much as possible on the origins of the organism and on the origins of social interaction, defining the minimal conditions for its emergence, while pragmatism in the way it has been developed in this dissertation focuses more on the mediation that makes cognition possible when cognition is already in place. If enactivists stress “the questions about individuation, autonomy, agency, normativity, and the nature of cognition” (Di Paolo 2009:10), a pragmatist approach can instead bring to the foreground the complexity of the functionalist network that constitutes all of these phenomena when they are already happening. Let me repeat the enactivist definition of social interaction: The regulated coupling between at least two autonomous agents, where the regulation is aimed at aspects of the coupling itself so that it constitutes an emergent autonomous organisation in the domain of relational dynamics, without destroying in the process the autonomy of the agents involved (though the latter’s scope can be augmented or reduced) (De Jaegher & Di Paolo 2007: 493).

265


I have already shown how Di Paolo and colleagues argue for this regulated coupling as being the minimal level of social interaction, a level that is autonomous from the individuals that compose it. However, the first and second chapter of this dissertation have already argued exhaustively that a far more complex social horizon is constitutive of human cognition. For instance, the enactivist definition does not even mention the learning processes that structure the human possibilities of and styles in interacting intersubjectively. Once coordination emerges it leaves traces that constrain the possibilities for further coordination of the same agents with each other and with other agents. In order to coordinate effectively both when dancing the tango argentino and when landing a plane strong skills and conventions are necessary. Analogously the enactivist definition of social interaction does not even mention the role of social norms, of cultural artefacts, of language. Why is there only one time scale at play – the ongoing interaction – and not others like cultural practices, the history of previous interactions and so on? If “cognitive processes by necessity involve multiple feedback loops and organizing activity across that boundary” (Di Paolo 2009:10) then we need to better thematise the role of the coupled systems outside that boundary and their lasting effects on the agents. This is exactly where the pragmatist framework came into the picture and the direction in which it articulated the notion of cognition. The environment, the interacting agencies, technologies and cultural practices modify which consequences are conceivable when the organism structures its behaviour in this or that direction. However, this mediation is far from free from tensions and clashes. We have seen in Chapter 1 how the habits that structure the cognitive system are not static dispositions but “living generality”, that is, they evolve and clash and re-adapt. It is via the notion of habit that we can see how the autopoiesis and the adaptivity of the organism can get re-organised far beyond the boundaries of the body and integrate environment, socio-cultural norms and other agencies. Thus in an intersubjective interaction it is possible to rely on the regularities and resistances of the other's 266


behaviour and on the regularities and resistances of the coordination in order to detach representations from the immediate affordances and rely on their more symbolic affordances. In 1997 Thompson and colleagues (Thompson et al. 1997) threw the cognitive science community into some excitement with the report that under determined circumstances chimpanzees could do symbolic-mediated reasoning without prior language training. In other words, Thompson and colleagues claimed that it was possible to “extend” the chimpanzee mind by training chimpanzees to treat physical tokens as symbols for abstract relationships and then to internally manipulate such symbols. Chimpanzees could thus represent and reason about abstract relationships. This experiment became highly quoted and ended up being a major support both to Terrence Deacon's (1997) evolutionary account of linguistic skills and – more to our point – to Andy Clark's notion of material symbols. It is thus interesting to consider it a bit more in detail to see what is really going on in the experiment and how our pragmatist perspective can make sense of it, bringing to the foreground the role of the social horizon and of agencies and coordination153. The study relies on the testing and observation of five animals, all traditionally reared captive chimpanzees (Pan troglodytes), in other words chimpanzees used to the interaction with humans and to being subjected to experiments but still living in somewhat natural conditions and not in a laboratory, a cage or a human house. Two tasks are involved in the experimental procedure: i) a conditional discrimination task and ii) a mix of a physical match-to-sample and a conceptual match-to-sample task. Let us look at them more closely. It is an established observation that chimpanzees can both match and discriminate objects on the basis of physical appearance. However, they cannot match conceptual relations among objects even when given extensive training. Chimpanzees, in other words, do not seem to be able to judge relations among relations among objects. And 153

I will in doing so partly rely on Hutchins' (2008) culturalist re-reading of these same findings.

267


this is where the use of plastic tokens seems to be playing a symbolic role. In the first task, the “conditional discrimination task”, the chimpanzee has to discriminate the abstract properties of same or different in the relation between pairs of objects, associating such abstract properties to plastic tokens. A window extends along one side of the experiment room separating a chimpanzee indoor area from an experimenter area. Stimuli can be presented by the experimenter on a shelf on the side of the window, and the chimpanzee can respond by touching images that appear on a touch-sensitive video monitor that is located in the chimpanzee's part of the test room at right angles to the window. Rewards can be dispensed via a plastic tube that projects through the window. The experimenter presents a pair of objects on the shelf. The chimpanzee then chooses one of the two tokens presented on the video monitor screen to go with the pair of objects that has been presented. If the objects in the pair are physically identical the animal is rewarded for choosing a particular token, in this case a heart-shaped token. When presented with a pair of non-identical objects the animal is rewarded for choosing a diagonally shaped token. All the chimpanzees tested perform well in this task. Thompson and colleagues interpret this fact as the acquisition of a code where the heart-shaped and diagonally shaped tokens stand for abstract relations of “sameness” and “difference” between pairs of objects. The second task is a mix of conceptual and physical match-to-sample trials. An experimenter places a sample object pair on a shelf on the experimenter’s side of the window. Two possible choice then appear on the video screen. In the physical match-to-sample trials one of the alternative pairs is a physical match to the sample pair. Choosing the physical match pair is considered to be the correct response. In the conceptual match-to-sample trial one possible choice is a pair of identical objects but not identical to the sample pair and the other is a pair of non-identical objects neither identical to the elements of the sample pair. The chimpanzee indicates its choice by touching an alternative pair on the video screen. In the conceptual matchto-sample trials choosing the identical pair alternative in response to an identical 268


sample pair and choosing the non-identical alternative pair in response to the nonidentical sample pair are considered correct responses. Since the alternative pairs never share objects with the sample pairs, correct performance requires the animal to match the relation between the objects in the sample object pair with the relation between the objects in the alternative object pair. Thus, it is conceptual relations rather than objects that must be matched. All responses, whether correct or not, were rewarded in the conceptual matching trials. All the chimps who had been trained in the conditional discrimination task perform surprisingly well at the conceptual discrimination task. Thompson and colleagues argue that having mastered a code, that is, having learned how to represent the concepts of sameness and difference, the chimps now have an intuitive way to encode in a perceptually-grounded symbol - heart or diagonal - an otherwise abstract propositional representation (Thompson et al. 1997: 41). What happens according to the authors is that the chimpanzee “imagines” choosing the token that would be chosen with this sample pair. But when it turns to the computer touch screen to choose a token, the tokens that were learned in the conditional discrimination task are not present. The chimpanzee encounters instead two alternative pairs of objects on the screen. The chimpanzee then “imagines” choosing the token that would be chosen with each of the alternative pairs. Finally, the chimpanzee is able to choose the alternative pair that is associated with the symbol that matches the symbol associated with the sample pair. The reason why the tag-trained chimps can perform this surprising feat is – so the authors suggest – because by mentally recalling the tokens that the chimps can reduce the higher-order problem to a lower-order one: all they have to do is spot that the relation of difference describes the pairing of the two recalled tokens (heart and diagonal). The learning made possible through the initial loop into the world of stable, perceptible plastic tokens has allowed the chimpanzee brain to build circuits that, perhaps by simply imaging the tokens themselves at appropriate moments, reduce the higher-order problem to a lower-order one of a 269


kind their brains are already capable of solving. However, an enactively informed pragmatist perspective endows us with the possibility of an alternative and partially complementary interpretation of the data. Chimpanzees do not develop a symbolic ability in general, but a symbolic ability related to the experimental situation. And it is through this very experimental situation that such ability is developed. What is crucial in the experiment is indeed the conditional discrimination training. Without that chimpanzees are unable to discriminate between abstract relations, at least for the purpose of the experiment. However, the conditional discrimination training should not be considered as a black box: it has structures we can highlight and it develops interaction sustained habits that can be then applied to the conceptual discrimination task. The conditional discrimination training is an interaction between experimenter and chimpanzee with the structure of a strict procedural game (Ferri 2010): it is a repeated interaction with a reliable script-like structure. A particular abstract property of the experience of the interaction with the pair of objects must be noted and such noticing is rewarded with a prize. The chimpanzee - widely enculturated to experimental practices - learns to rely on the structures of the interaction, of the human-chimpanzee coordination, in order to solve the task. The ability to abstract a specific aspect of the surrounding environment is thus crucially mediated both by the object (the token) and by the social interaction that reliably foregrounds the relevant properties and sustains the use of the token through time. Thompson and colleagues focus too much on the token and too little on the social dynamics involved in the task and on the way they support the symbolising habit. The perception of abstract relations is not too dissimilar to the perception of a constellation (Hutchins 1995): deeply encultured and relying on a process of interactive socialisation and symbolic practices. Indeed, two more observations strengthen this interpretation. The first observation is that social dynamics are repeated and structure the conceptual discrimination trials 270


that, although not rewarded, are supported by the alternated physical discrimination trials where rewards are available. Which effect would the removal of the physical discrimination task and of its support to the relevant symbolic habits have? To support this point, the symbolic activity and the discrimination of abstract relations does not seem to have any lasting influence outside of the specific experimental situations. Chimpanzees do not start employing symbols in their daily routines, to the least to indicate abstract relations. What Thompson and colleagues highlight is a pre-symbolic use, showing how the possibility of relying on the coordination dynamics and the perspective of the other to create a process of objectification (Froese & Di Paolo 2009) opens up a space for symbolic uses. However, the chimpanzees fail in fully developing symbols because they are not taken – probably because not fully equipped to do so (Tomasello 1999)154 – into linguistic dynamics, the next object of our argumentation. Similar arguments could be directed at the other examples used in developing the material symbols conception (e.g. Boysen, Bernston, Hannan and Cacioppo 1996). Before leaving the experiment, however, a last observation has to be made. The intentionality of the chimps is reshaped by the coordination process in the experimental practice. Via the repeated game-like interaction the chimp becomes motivated towards coordinating with the experimental practice and towards mediating its cognitive processes via material tokens and social interactions. The individuality of the chimpanzee and its intentionalities participate to and rely so pervasively on the distributed cognitive system155 that the analysts of the experiment ended up attributing to the chimpanzee an intentional symbolic activity more

154

155

There are more than a few interesting hypothetical reconstructions on how the development and the extended use of proto-symbols might have over time evolutionarily pushed the development of the human brain and culture (cf. Deacon 1997, Donald 1991), therefore enabling human beings to enter more sophisticated forms of linguistic dynamics. Without being reducible to it. The chimp is not absorbed and enslaved by the interaction that is fully social. The chimp can rely on certain structures of the interaction to perform tasks it would not be able to perform (and would not perform in other contexts) but it can still refuse to cooperate and it will still bring to the interactions its limitations (it cannot talk!) and idiosyncrasies.

271


properly attributable to the human part of the system. This does not mean to reduce symbols to the social dynamics. It means to show that symbolic activity in chimpanzees, that is the selection of relevant aspects in the stimuli, their hypostatisation and embodiment in material symbols is strongly supported by the "priming" of the training and of the interloped physical discrimination task and its reward structure. So the chimps become able to grasp and rely upon a certain degree of symbolism thanks to the distributed cognitive process. We have seen how enactivism inspires in the pragmatist framework a renewed interest in the interplay between individual agencies and processual coordination dynamics and brings into the foreground the fact that cognitive systems get coupled and uncoupled, break the coordination, converge and coordinate, develop different degrees of coordinated stability and move on, carrying the traces of such processes. Through such a reformed pragmatist framework I have shown how the inchoative use of symbols displayed in at least some cases of symbolically-extended ape minds deeply relies on the structures of the social interaction in which the material symbols, the chimps, the experimental setting and the human experimenters are embedded. While more could be said, in order not to stray too far from the points of focus in this dissertation what we need to concentrate on now is an aspect that is completely underdeveloped in the enactivist reflection. The processes of coordination are necessary in order to shape not only the single agency but also its ability to go beyond the affordances of the moment in order to introduce a distance – the possible perspective of the other – and enter the “shared world of significance” (Varela et al. 1991:207) that grounds the possibility of symbols. However, the enactivist perspective does not problematise the role of language and the question of how – once language is in place – these processes of coordination take place.

272


3.1.5. Language as a tool for interaction and experimental approaches to pragmatics and semiotics Many of the pieces necessary in order to define the role of language in constructing and coordinating distributed cognitive systems are already in place. In Chapter 1 we have seen that language use can i) augment memory, ii) simplify the environment, iii) coordinate activities through control of attention and resource allocation, iv) establish control loops and v) represent and manipulate data both in an open and in a covert context (§ 1.4.3). In Chapter 2 I introduced and applied the notion of diagrammatic reasoning via language as the way these functions are achieved both in single texts, in intersubjective linguistic exchanges and on a wider diachronic scale in a newspaper corpus. This investigation showed how language is both an inherited environment carrying pre-existing constraints to interaction and a ductile tool changing over time and adapting to ongoing interactions. However, the focus was on qualitative developments at a time scale of a few years. Can we map these dynamics in a more quantitative way and at a more detailed time scale in order to enable a more effective dialogue between the semiotic enterprise that is expressed in this dissertation and the wider disciplinary field of cognitive sciences? In particular, can we map more precisely the way language fits local needs, the way it gets changed by these needs and, more crucially, the way these adaptation processes interact with cognitive performances? A privileged observation point on these issues seemed to me to be experimental pragmatics: a minority of students of human communication has in the last forty years developed experimental investigations of spoken language dialogue (Krauss and Weinheimer 1964; Garrod and Anderson 1987; Clark 1996; Horton and Keysar 1996; Brennan and Hanna 2009; Garrod and Pickering 2009; Shintel and Keysar 2009). This growing line of research has more recently been complemented by experimental semiotics, that is, the study of interactions that occur in the absence of pre-established communicative conventions (Galantucci 2009; Galantucci and 273


Garrod 2010), an absence achieved either by impeding the use of linguistic communication (Galantucci 2005) or by employing software agents (Loreto and Steels 2007). The use of software agents in particular allows the researchers to implement the succession of generations, thus simulating the interaction of multiple time scales from sedimented and inherited semiotic and linguistic competences, through the history of previous interactions the individual has undergone, to the unfolding of the single local interactions. Compared to other approaches to language and other semiotic systems, experimental pragmatics and experimental semiotics: i) do not content themselves with descriptions and hypothetical explanations but attempt to experimentally manipulate and control different factors in human communication in order to uncover the causal relations behind it and ii) assume that the natural dimension to study human communication and language (or sign) use is social interaction and not just individual cognitive performances as in most psycholinguistics (Pickering and Garrod 2004). Experimental and simulative paradigms construct strongly goal-oriented situations where pairs or groups of subjects have to communicate in order to coordinate the joint solution of a problem: the subjects or agents need to construct a shared operative representational layer that enables joint action. While the situation is quite constrained by the need to solve the task at hand, the linguistic and semiotic exchanges produced tend to display at the same time a limited set of common strategies and wide variability concerning the development and the unfolding of these strategies. It is exactly in these common structures and variabilities that we can see the interplay of physical, biological, socio-cultural and iconic constraints with the local history of the dialogical interactions and the ductility of language adapted via alignment and negotiation to the affordances of the situated interactive practice (Tylén et al 2011). It is this sort of research that has inspired the more experimental investigation in the next pages. A caution should be added. Experimental pragmatics does not study language as 274


such and does not investigate the nature of language. Written language, for instance, has not been investigate up to now except in chat-like system. What experimental pragmatics does – and I will follow in the next paragraphs – is to select a limited aspect of language in a very specific situation and try to test some hypotheses which general validity will have to be further tested. Experimental pragmatics gives me, however, a unique way to test the pragmatist framework and to explore the possibility of supporting it not only conceptually and via qualitative analysis, but also quantitatively.

3.2. Language as a tool for joint cognition: An Experimental Setting What I am going to do is thus to go through the setup of the experiment, to present its results and then to show how the pragmatist framework developed so far can be used to develop some hypotheses about the linguistic interaction going on, to test them and therefore contribute to explain what is exactly going on in the results.

3.2.1.

Describing the experiment

The experiment we are going to deal with was born from a long tradition of testing perceptual decision-making in isolated individuals (cf. Ernst & Bülthoff 2004). The question asked is: how does information from different sensory modalities integrate? Does this integration add to the individual's perceptual performance? How does it do so? The literature shows quite persuasively that probability distributions arising from different sensory modalities integrate in a Bayes optimal way given full access not just to the magnitudes of sensory signals but also to their probability distributions or at least to their means and variances in the different sensory modalities (Ernst & Banks 2002, Ernst & Bülthoff 2004). In other words, isolated individual observers can accurately integrate different sources of sensory information according to the sources' relative reliability. Such a process can be mathematically modelled and simulated in accordance to the experimental data. 275


Figure 16 – The multi-sensory setting

An evident question follows from the discussion of cognitive distribution and integration in the previous chapters. Does this robust paradigm hold if we replace internal integration with a distributed system? How is information compared and processed if instead of two different sensory modalities we have two different individuals? Can we pass from multi-sensory integration to multi-agent integration? To answer these questions the setting described at the beginning of this chapter was designed. A dark room with two desks with a synchronised computer screen each. Two participants sit at the desks at an angle that does not make it possible for the one to see what the other is doing or seeing and vice versa. Two almost identical images 276


appear on a sequence and they have to be discriminated first individually and then as a pair. Eighty-eight participants were recruited from undergraduate, graduate and faculty members of Aarhus University, Denmark156. The members of each dyad157 knew each other before the experiment and no participant was recruited for more than one experiment.

Figure 17 – The experimental setting

The stimulus set displayed in each interval consisted of six vertically oriented Gabor patches158 organised around an imaginary circle with a radius of eight degrees at equal distances from each other.

156 157 158

44 pairs; mean age ± standard deviation: 28.30±6.27 From now on “group” and “dyad” will be considered interchangeable, since I am not going to deal with groups containing more than two participants. Gabor patches are patterns obtained through the application of a Gabor filter (Daugman 1988). Gabor patches are frequently used in texture representation and discrimination, since the filter presents frequency and orientation similar to the human visual system. In this particular case, The features of the Gabor patches were: 0.45 degrees of standard deviation of the Gaussian envelope; 1.5 cycles/degree of spatial frequency; and 10% of contrast).

277


Figure 18 – The disposition of Gabor patches

In order to create a minimal difference between the two sets of Gabor patches one of the six patches in one of the two images had its contrast elevated, therefore producing what is called a contrast oddball. The contrast oddball added one of four possible values 1.5%, 3.5%, 7.0% or 15% to the 10% contrast of the non-oddball items. The oddball location and interval were randomised across the experimental session.

Figure 19 – Displa y of the stimul us set

The partici pant respo nding with the keyboard after agreement with her/his partner initiates the trial. A black central fixation cross (width: 0.75 degrees visual angle) appears on the screen for a variable period, drawn uniformly from the range 500-1000 ms. The first set of six Gabor patches is shown for 85 ms. Then a new fixation cross is displayed for 1000 ms followed by 85 ms of exposure to the second set of Gabor patches. The fixation-cross turns into a question mark after the 2nd interval to prompt the participants to respond. The question mark stays on the screen until both participants have 278


responded. Each participant initially responds without consulting the other159. Once the participants have answered, both their individual decisions are displayed on the monitors, one in blue (keyboard) the other in yellow (mouse). The order of display is randomised to avoid spatial biasing. In case the two answers are the same, feedback is given as to their correctness. In case the two participants' answers are different, a joint decision is requested, with the request made in blue if the keyboard participant is to announce the decision and in yellow if the mouse participant is to input it. In odd trials this task is given to the keyboard participant, in even trials to the

159

mouse

participant.

The participant who uses the keyboard responded by pressing “N” and “M” for the first and second interval, respectively; the participant who used the mouse responded with a left and right click for first and second interval, respectively.

279


Figure 20 – The experimental script

The process of joint decision is as unstructured as possible: the participants are free to discuss their choice with each other as long as they want. Once they input their joint decision they receive feedback immediately as to the correctness of both their initial individual decisions (in blue and yellow) and their joint decision (in white). Participants are then free to discuss their decisions and to reflect on the whole process. It has thus to be stressed that stimulus exposure aside the pace of the experiment’s progress is set by the participants. This task is initially performed sixteen times to allow the participants to become familiar with the process, then the real experiment starts, being constituted by two main sessions of eight blocks of sixteen trials each. In order to balance for possible influences of position, input device and so on, participants switch places (and thereby response device) at the end of session one.

280


3.2.2.

Measuring the cognitive performance

The performance was measured in terms of correct answers given160 both during the individual and the joint decisions. The expectations were not high. Against the common-sense assumption that working in small groups is better than working alone, a substantial body of experimental research has shown that contrasting group and individuals performance does not reveal any significant improvement attributable to group work (cf. the reviews in Hastie 1983 and Hill 1982). While groups often outperform their average members, they perform worse than the pooled results of equivalent individuals working alone. Such results have been confirmed in tasks related to induction (Laughlin & Futoran 1985, Laughlin & Shippy 1983), problem solving (Kelly & Thibaut 1969), brainstorming, puzzle solving and writing. Overall, this body of experimental data seemed to suggest that there is no specific advantage in joining individuals into a group and that if any specific cognitive product of group work had to be pointed out that was the generalised criticism the members aimed at each other (Laughlin & Futoran 1985). Different explanations have been pointed out: a reduced effort in the presence of others (the presence of others (Latane et al. 1979), interpersonal competition (Hastie et al 2005) and groupthink (Turner & Pratkanis 1998). To answer the questions if participants in the experiment could effectively share information and of how their group performance would relate to their individual performance, Bahrami and colleagues developed four mathematical models on which to test the results: •

The Coin Flip model (CF): if the participants communicated only the decision taken, in case of disagreement the joint decision would not outperform the flipping of a coin to decide. Such a model was employed for instance by Hill (1982). No group benefit is predicted, that is, the joint decisions’ accuracy is expected not to be better than the average accuracy of the two individuals.

160

Reaction times were also measured but did not reveal any significant correlations.

281


•

The Behaviour and Feedback model (BF) also assumes that the participants only share information about which decision they have individually taken. However, it is also assumed that the groups learn from trial-to-trial feedback which individual is more accurate and as a consequence ends up using that individual’s decisions also for the joint decisions. Evidence for this behaviour in group decision-making is presented in Hastie and Kameda (2005) and Laughlin and Ellis (1986). The accuracy of the joint decisions is thus expected to

•

be

equal

to

the

accuracy

of

the

best

individual.

The Weighted Confidence Sharing model (WCS) assumes that participants not only share information about which decision they have individually taken but also their confidence161 in the decision itself. This is a new model not previously tested. It does not presuppose the presence of any feedback – as opposed to BF. The accuracy of the joint decisions is expected to be a function of the difference in accuracy between the two individuals and a group benefit is expected to arise when the ratio between the two is more than

•

0.5.

The Direct Signal Sharing model (DSS) is taken directly from the multisensory integration research (Ernst & Banks 2002; Körding & Wolpert 2004)

161

Confidence is defined as "a person's strength of belief about the accuracy or quality of a prediction, judgment or choice” (Peterson & Pitz 1988), or the individual’s estimate of the probability of being correct.

282


and assumes that the mean and standard deviation of the sensory response to the stimulus are communicated. Evidences for such model in group decision making have been presented in Sorking and Hays (2001). The accuracy of the joint decisions is expected to exceed the accuracy of the isolated individuals given

their

perfect

sharing

of

information.

In order to avoid burdening the reader with the mathematical details of the enterprise I will just spoil the suspense and reveal that the WCS model ended up being the best predictor, at least for the results presented in the Science paper (Bahrami et al. 2010). However, as we will see, the analysis of the linguistic interactions was to reveal some interesting twists and to make the model more complex. Let us take a closer look at the results, using group 21 as an example.

283


Figure 21 – Performance of Group 21

It is easy to see in blue and yellow in the lower part of the graph the performance of the two participants and observe that participant number 2 is tendentially better than participant number 1. What is surprising, however, is that the accuracy of joint decisions is significantly better than individual accuracy, an observation that is more easily assessed by normalising the performance of the best individual in the upper part of the graph and mapping the joint performance onto it. Even if not all the pairs reach such group benefit, the group benefit holds even when all the pairs are taken into account. Notice that in the upper part of the figure green represents the group benefit and in the lower part green represents the best individual performance.

284


Figure 22 – Average Group benefit for all the groups

A first result of the experiment is thus that on the average when two subjects share through language their perceptual information, their accuracy in discriminating between the percepts increase significantly. Moreover, the groups seem to behave according to the WCS model, that is, their joint accuracy in decision-making is a function of the difference in the individuals’ accuracy, the less difference the better joint accuracy.

Figure 23 Group benefit plotted against relative sensitivity (smin/smax).

285

Language and Cognition - Ph.D. Dissertation by Riccardo Fusaroli Each blue square represents one group162.

In order to control for these results a variation of the experiment was set up in which the participants were not allowed to communicate although they could see each other's choice. Whenever the participants disagreed in their decision, one of the two (chosen randomly by the computer) made a decision individually by arbitrating between their own choice and that of the other participant. Feedback about the correct choice was then given to both participants. The results were unequivocal. The accuracy of joint decisions did not go beyond the accuracy of the best individual. Thus linguistic interactions are somehow necessary to achieve a collaboration benefit, much in contrast with the results of the mentioned previous studies that show how groups do not perform better than pooled individuals. Finally, given that the model made the extremely counterintuitive prediction that giving feedback to the participants as regarding the correct answer was not making a difference, a further variation was developed: the participants were never told the correct answer. The results showed a significant collaboration benefit that was statistically indistinguishable from the results of the original version of the experiment. Group benefit seems thus to be achievable via communication alone, also in the absence of objective feedback. These results are a ground-breaking demonstration of the important role of language and communication in the construction of distributed cognitive systems, a demonstration that goes beyond the qualitative appreciation, to attempt instead at quantifying and factoring the actual import of linguistic interactions and the conditions under which they can make a difference. However, there is something deeply unsatisfactory in these models – at least for a 162

The attentive reader will notice that the predictions made by WCS and DSS for the range of groups observed are very similar. A second experiment was set up, feeding noise to the stimuli of the less accurate participant in order to increase the difference in individual accuracy between the participants. This variation allowed the researchers to discriminate between WCS and DSS leading to uncontroversially declare WCS the fittest model to the results.

286


semiotician. All the models assume an optimal degree of communication and the perfect coordination of the participants in the distributed cognitive system. Could it be that by applying a semiotic gaze and employing the tools developed in the previous chapter more insight could be gained both to further interpret the data produced by the existing experiments and to contribute to new experimental designs in order to better understand the factors playing a role in the phenomena observed. For instance around 20% of the groups did not show any increase in joint accuracy as compared to individual accuracy. While such a percentage does not destroy the statistical significance of the results, it raises the question of the reasons behind the lack of improvement. Is it just the effect of chance – some groups are bound to fail – or is there some more data to be mined in order to explain these failures in a more satisfactory way? This kind of investigation could help disclosing which factors in the linguistic interaction are crucial, in which ways language is employed and so on. Moreover, given the previous critical comments aimed at the experiments on chimpanzees in 3.1.4.3 it should not be forgotten to keep a critical eye on the social interactions at work in this experimental setting.

3.3. Opening the Black Box of Language In the results presented so far the effects of linguistic interaction on performance are clear. Linguistic interactions are crucial in allowing the construction of a distributed cognitive system able to perceptually discriminate visual stimuli with a greater accuracy than any of its components. Linguistic interactions seem so powerful that they can take over the role of objective feedback in strengthening effective cognitive habits in the individuals. We see what Peirce called the social instinct (1.2.2.4.3) at work, forcing individuals to express and compare their own perceptual and decisionmaking habits and to get them reshaped by the public arena. The findings do not, however, so far really consider what is going on in the interaction. Groups using language are considered as one condition and their results averaged. Linguistic 287


interaction is a black box. It is thus time to open the black box of such linguistic interactions and go beyond the intuition-based insights gained by experimenters while attending the experiments. The first published report carries traces of such insights, something which can be seen in the following quotations: the participants “could split the task spatially (e.g. left and right half of screen) or temporally (first and second intervals). Such strategies were frequently employed, but never sustained for more than a few trials” or “we never observed any dyads explicitly communicating contrast and reliability separately” (Bahrami et al. 2010: 1085). These insights, however, are not enough for a scientific study of the role of linguistic interactions. Therefore, what we163 did was to produce and analyse videos of the experimental sessions, beginning with the twenty-first group. The videos reveal against any naïve expectation rich multi-modal and multifaceted interactions. Given the richness of the data to analyse it was crucial to establish a relevance criterion. We started from the pragmatist framework in order to draw some predictions from it. We expected to see a language-mediated coordination process through which habits that structure the interaction and that possibly also have an impact on the individual cognitive systems would emerge, be tested and stabilised. Ideally such patterns, when identified, could be quantified and correlated with performance, therefore giving us clear indexes of linguistically mediated coordination and of its impact on the performance of the joint cognitive system. The null hypothesis – in a Popperian fashion, the conditions at which our hypothesis would be falsified – was thus that there would be no significant difference in linguistic and interactional patterns between the successful groups and the unsuccessful groups, meaning that such patterns do not have any effect on the performance. After an initial consideration of some of the videos and of their proxemic, 163

This part of the research has after an initial qualitative analysis performed by myself turned into a joint project with Kristian Tylén. Without the continuous discussions and his technical skills the quantitative and most interesting part of this project would not have been realised.

288


interactional and linguistic aspects we realised that if we wanted the analysis to be integrated in the experimental paradigm, we needed to treat the largest amount of data possible and to minimise the impact of the sensibility of the analyst. Therefore, and in continuity with the previous chapters, we decided to focus for the moment on two aspects: lexical choices and dynamics and the length and general structure of the interaction164. The hypothesis was that in order to properly coordinate and achieve a group benefit the participants had to: 1. Negotiate and stabilise a repertoire of linguistic codes that enable a tight coordination of the decision making task; 2. Negotiate and stabilise emergent, locally normative patterns of turn taking and role distribution. The null hypothesis would be then that successful and unsuccessful groups do not display significant differences in the stabilisation of a standard way of negotiating the perceptual information and the decision making process. Accordingly, we opted for a selective transcription of the videos165: the transcriptions report only the joint decision making linguistic interactions and they focus on the lexical aspects while intonation, prosody, pauses, pronunciation and overlaps are generally not reported. In particular, lexemes are transcribed using normative spelling (the written convention) and not taking into account the actual pronunciation which in spoken Danish can present a wide range of variations. As an example I report here the first three interactions transcribed (from group 21), with an English translation on the side for the comfort of the reader166:

164 165 166

It has to be noted that some additional preliminary analyses are in progress on prosodic aspects (pitch profiles in particular) and on the reciprocal bodily orientation and gaze of the participants. The transcription process was made possible by the cooperation of Peer Christensen, Nicolaj Hansen, Benjamin Riis and Kristina Broberg, which I heartily thank. Obviously the analysis was conducted on the original text.

289


Original

Translation

session 1 session 1 Interaction 01(Success) Interaction 01(Success) B (0:01:31.9) jeg er sgu ikke helt sikker B (0:01:31.9) I am not completely sure H* H*, damn A (0:01:33.3) jeg synes det var toern A (0:01:33.3) I also think it was the altså.. second B (0:01:35.2) [mhh ] B (0:01:35.2) [mhh ] A (0:01:35.2) [skal vi] prøve? A (0:01:35.2) [shall we] try? B (0:01:35.5) prøve toern B (0:01:35.5) try the second A (0:01:36.6) okay (.) A (0:01:36.6) okay (.) A (0:01:38.9) den var rigtig (?) A (0:01:38.9) was it right (?) A (0:01:40.6) næste! A (0:01:40.6) next! Interaction 02(Success) Interaction 02(Success) A (0:02:20.0) jeg var ikke helt sikker der A (0:02:20.0) I was not completely sure N* there N* B (0:02:21.2) næ B (0:02:21.2) nope B (0:02:22.2) jeg er 60% (mumler) B (0:02:22.2) I am 60% (mumbles) A (0:02:24.1) jeg tror du har har ret A (0:02:24.1) I believe you are right B (0:02:25.8) prøve? B (0:02:25.8) try? B (0:02:27.9) det var rigtigt B (0:02:27.9) that was right A (0:02:28.4) .hh ja næste. A (0:02:28.4) .hh yes next Interaction 03(Success) B (0:02:38.5) jamh... jeg synes den var svært at se A (0:02:39.5) ja det samme her B (0:02:41.4) mhh A (0:02:41.8) det var sådan lidt. B (0:02:42.6) jeg er fifty-fifty på (den her) A (0:02:45.9) skal vi prøve en toer B (0:02:46.8) eeeh ja eeh du prøver bare A (0:02:48.8) okay B (0:02:51.2) det var sgu rigtigt A (0:02:52.7) næste B (0:02:53.8) mmh

Interaction 03(Success) B (0:02:38.5) yeah... I found that difficult to see A (0:02:39.5) yes - the same here B (0:02:41.4) mhh A (0:02:41.8) it is kind of B (0:02:42.6) I am fifty-fifty on (this one) A (0:02:45.9) should we go for the second B (0:02:46.8) eeeh yeah eeh you try it out A (0:02:48.8) okay B (0:02:51.2) that was fucking right A (0:02:52.7) next B (0:02:53.8) mmh

All the interactions were numbered and – when possible – tagged with the success or the failure of the interaction. The beginning of each speech turn was timed. In 290


several cases we found it necessary to attribute more than one speech turn in a sequence to the same participant if a long pause enough separated the utterances. It should be noted also that the names of the participants were abbreviated for them to stay anonymous. At the time of the transcription in order to avoid biases none of the transcribers knew the general performance of the pair transcribed or the details of the analysis. These transcriptions were then subjected to a combination of qualitative and quantitative approaches. Two researchers (me and Kristian Tylén) independently identified key components in each interaction, compared their results with those of the other and with the videos and on the basis of this process of identification defined linguistic indexes in order to automate the reconstruction and the analysis of the distributive patterns of the key components across interactions, sessions and groups167. The aim was to obtain not only significant differences between successful groups and unsuccessful groups but also to be able to establish a proportional correlation between performances and linguistic indexes.

3.3.1.

The structure of the interaction

The interactions share a common structure. Interaction 3 (Group 21)

Interaction 13 (Group22)

B jamh... I think it was difficult to see A Yeah the same here B mhh A It was like a bit B I am fifty-fifty on (this here)

A Again I think I see a bit in Opening / Statement of uncertainty both places B ok B I think oeh.. I had such a.. B a good gaze at number 1 and then number 2 B it was not that good B and in the first one I don't think in any case I saw it

167

Phase

While writing these lines nine groups have been fully transcribed and analysed, approximately fourteen hours of videos and 818 relevant interactions. The chapter is based on these transcriptions, and the ongoing analyses of the rest of the corpus are confirming these results and increasing their statistical significance.

291

Language and Cognition - Ph.D. Dissertation by Riccardo Fusaroli A should we try a two B eeeh ja eeh just try A okay

A ok, so we take the second Negotiation B what... what was it... naa no, that was for sure A what are you saying? B yes, let us try the second one A yes B ok

B that was fucking right A Next B mmh

A

shit

Closure / Assessment

All the groups tend to lengthily negotiate a common state of uncertainty, even when the individuals did not hesitate in their individual decisions. While it is possible that this is simply an instance of the democratic game (Schelling 1960) in which individuals do not want to look like they are trying to establish their own leadership and therefore display their common state of uncertainty I will, nevertheless, come back to another possible and maybe complementary explanation for this. However, this structure becomes far shorter in the course of the experimental session. Interaction 103 (group 21)

Interaction 114 (group 22)

Phase

B: jah that is what do you say B: 15 A: I think it was more in the first A: shall we not take mine? B: ja I am very ( ) it's like that I think we B: yes take the first oeh I am not sure that two is right here

Negotiation

B: that was also right

Closure

B: nice

Such qualitative restructuring of the linguistic interaction is also visible in the evolution of the length of the interactions. Here we can see how the average number of speech turns in an interaction diminishes by 2.7 turns between the two parts of the experimental session.

292


Figure 24 – Reduction in speech turns in Group 21

And such result holds

for

all

the

groups.

3.3.2. S ki lls ne e ded for successful interaction What we see is the establishment and evolution of a social interaction with, at least partially, the features predicted by the enactivist model and in addition hereto the mediation of cultural artefacts and social norms that pragmatism has got us used to considering. The task at hand requires a certain degree of social coordination that seems to be picked up by all the groups. A few generic social practices are at work: the experimental practice in which participants – all volunteers – want to fulfil the role of good subjects performing well and the social interaction between the two participants – who already know each other – implying that they should leave the room in friendly terms. These social practices intertwine with a more specific practice dictated by the experimental paradigm and embodied by the physical set-up of the experiment and by the instructions on the computer screens. To solve the task the participants have to linguistically share their perceptual experience and reach a 293


shared decision168. We have thus an individual level of the experimental practice which consists in: 1. The capacity of comprehending the task 2. The capacity of perceptually discriminating the stimuli 3. The capacity of making a choice based on such discrimination. We also have an intersubjective level of the experimental practice which consists in: 4. The capacity of coordinating with the other participant The individuals analysed all clearly comprehended the task and were able to perceptually discriminate and operate their individual choices although at different levels and they were also clearly able to coordinate their social interaction. All the groups structured their interaction through a loose functional process apt to satisfy the democratic game, to reach a decision and to give the right emphasis to feedback. Through the session the coordination becomes more tightly coupled. The initial phase is skipped in favour of more direct negotiation – this will be further analysed in the next paragraph. However interesting this behaviour does not tell us anything about the capacity of only a subset of the groups to achieve a group benefit or about their different group benefit. We still miss a last point, a specific way of coordinating: 5. The capacity of effectively sharing the information related to perceptual experience and to confidence in decision.

3.3.3.

Sharing experiences

We have seen how this experiment has its roots in multi-modal sensory integration, in which the different modalities are led, supported and processed by a neural system 168

These social practices can be manipulated and the introduction of such manipulations in the experimental paradigm will be pursued.

294


which might be heterogeneous but with a long history of internal coordination. It is the neural system that ensures the common ground for balancing the respective reliabilities and integrating the different sensory modalities169. In the multi-agent system at work here that common ground is not available. It has to be constructed and agreed upon via language since there are no strongly established conventions for describing and negotiating the extremely short visual experience of Gabor patches. To understand how this happens we need to make a little detour. Language is widely employed in establishing conceptual models in order to scale and assess individual experiences and performances, which makes it possible to publicly share and integrate their informational content. This happens at – if we look at the literature – at least two different levels: the object of introspection and more interestingly for us the degree of confidence of such evaluations. 3.3.3.1. Language as a scaling device: the case of confidence communication

Indeed, the construction of a scale of commensurability is not at work only in sharing the contents of introspection but also in assessing the reliability of the information provided by introspection (Peterson, & Pitz 1988). Reliability is one of the key factors in modelling cognitive processes (Kruschke 2010) and indeed the different degrees of confidence communicated alongside information greatly affect decisions made both in experimental situations and in juries (Zarnoth & Sniezek 1997; Sniezek & Van Swol 2001; Pulford 2002), affecting adults and children in similar ways (Moore, Harris, & Patriquin 1993). There have been a few attempts trying to define a constrained language to assess confidence both about events and states of mind. Given that subjects feel more comfortable communicating probabilities and confidence using verbal expressions than numerical ones (Hamm 1991), studies have been produced assessing the numerical values attributable to linguistic labels like 169

To be more precise it is that neural system together with the stability of the external world that makes all that possible. In an unreliable world such as the one devised by Descartes' evil daemon multi-modal integration would be unreliable as well.

295


"absolutely impossible", "rather likely", "very probable" and "almost certain" (for a review and the list of the 178 expressions investigated cf. Druzdzel 1989, Clark 1990). The main conclusion is that although there is great deal of between subject variability in the numerical values assigned, the ordinal relations between these expressions are more or less consistent. However, Druzdzel (1989) shows that the numerical and ordinal value of the expressions is also highly context and task sensitive with an asymmetry and a large overlap between different phrases. There seem to be several co-existing and partially overlapping uncertainty concepts. "Which answer one will receive to a question about uncertainty or probability will be dependent upon the family of concepts being activated. Thus questions of ‘uncertainty’ may trigger a different line of thought than a question of ‘chances’; ‘probabilities’ may be assessed independently of questions about ‘usualness’ or ‘frequency of occurrence’; in hindsight, the appropriateness of ‘hopes’ and ‘doubts’ will be judged by other criteria than their corresponding probability values, etc." (Teigen 1988: 37-38)

It thus seems that the unidimensional assumptions in the studies reviewed by Clark and Druzdzel tend to obscure the rich and context-sensitive semantics of expressions of confidence (Fox 1986). 3.3.3.2. Language as a scaling device: back to the data

We now have at our disposal all the conceptual tools to understand how a common ground can be created. And indeed if we take an excerpt of the transcript we can notice how much of what is going on is an attempt at figuring out ways to describe one's own confidence in perception. Interaction 43 (S) A (0:02:42.1) B (0:02:43.7) A (0:02:46.2) B (0:02:47.2) A (0:02:47.3)

we take yours of course I saw nothing I didn’t see anything either – I saw … I took a bet [way to go! [way to go! 296


B

(0:02:48.5)

Interaction 44(F) B (0:02:58.3) A (0:02:59.4) A (0:03:00.3) right in both of them B (0:03:04.6) A (0:03:13.6) B (0:03:16.3) A (0:03:18.0) Interaction 45 (S) A (0:03:19.8) A (0:03:21.9) B (0:03:23.5) B (0:03:24.9) B (0:03:27.4) A (0:03:28.0)

mine was also just a bet there ((laughs)) I don’t know I don’t know either I saw something both in the left corner and in the centre on the okay, I think it was over in the left side, but oehm I’ll pass no! we ruin the scores – now we must… yeah, now we must pull our shit together fuck me no, I didn’t see it – we take yours yes, actually I think I was sure there wait a minute, I’m just – yes it’s me yes good!

While there is an attempt to describe the percept (“I saw something both in the left corner and in the centre on the right”) far more space is dedicated to the expression of confidence. Four descriptor-types – highlighted in different colours – are used and repeated across the interactions: to see (at se), to take a bet (at gætte), to know (at vide), to think (at synes). These descriptor-types present variation in degrees of confidence (I didn't see anything vs. I saw something). Similar patterns hold for the whole corpus analysed. As opposed in the studies on confidence previously mentioned these confidence scalings are spontaneously generated, debated and compared giving us a more ecological insight into the dynamics involved in the stabilisation and negotiation of a confidence scale. In the data analysed we have individuated sixty-two types of descriptors. The most reliable way to individuate them, a way developed through the discussion of the independent coding performed by the two researchers involved in the project, is based on a morphological level (for the importance of morphemes in automated linguistic analysis and more in general in the stabilisation of semantic 297


configurations, cf. Rastier et al. 1994, Cadiot & Visetti 2001). An in-depth analysis of the single transcript is performed individuating all the expressions of confidence. Only the expressions individuated by both researchers are then kept. These expressions are then grouped in families of variations according to a core morpheme, like “*sur*” (*sikk*) that is present in a wide range of variations from “fucking unsure” to “I am sure”. In case an expression of confidence contains more than two relevant morphemes as in “I think I am sure” we count it only once, considering one morpheme (think) a modifier of the other (sur*). A list of these sixty-two morphological cores declined in their most common form – be it verbal, substantival, or adjectival – is displayed in the following table.

At Ane

At Opleve

Fornemmelse

At basere (to base)

At passe

Den/Det

At bedømme (to judge)

At satse

was)

At bestemme Sig (to decide)

At se (to see)

Et Eller Andet (one

At få (to get)

At sidde forkert (to sit in a or the other)

At falme

wrong way)

Forskel (difference)

At fange (to catch)

At sige (to say)

Intuition (intuition)

At finde (to find)

At skelne (to discriminate)

Kraftig

At forsvare (to defend)

At skyde (to shoot)

Lost (lost)

At føle (to feel)

At stå fast (to sit steady)

Lys (light)

At gætte (to guess)

At synes (to think)

Markant

At give sig (to give oneself)

At tænke (to think)

Måske (maybe)

At hælde

At tage en chance (to take a Mening (intention)

At haenge

chance)

Mørk (dark)

At have (to have)

At turde (to dare)

Numbers

At indrømme

At tro (to believe)

Punkt (point)

298

Var

(it


At kigge (to look)

At vække interesse (to awake Sikker (sure)

At kunne lide (to like)

an interest)

Skarp (sharp)

At lægge mærke (to notice)

At være (to be)

Stærk (strong)

At miste/misse (to miss)

At vide (to know)

Svær (difficult)

At opdage (to discover)

Betydelig (meaningful)

Svagt (weak)

At overtale (to persuade)

Blank (blank)

Tvivl (doubt) Tydelig (clear)

While this list is growing along with the ongoing analysis and most groups present at least one use per most of the families there is only a limited set of widely recurring families of descriptors as the following table showing the two most frequent dominant groups by type displays170. Group

Group

Group

Group

Group

Group

Group

Group

Group

21

22

29

32

35

36

42

43

44

Sure

To see At se (59)

Sure

To see At se (111)

To think Sure At synes Sikker (157) (51)

To see At se (109)

To see At se (86)

To see At se (108)

Sure Sikker (47)

Clear Tydelig (86)

To believe To know To At tro (31) At vide believe (16) At tro (58)

Sikker (123)

Sikker (84)

To think

Number Numbers s (54) synes (65)

At

To see At se (48)

(49)

It is clear to see that at least taking the two most frequent families of descriptors the wild variability is reduced. The most recurring family is “to see” (at se) followed by “sure” (sikker). While intuitively the use of numbers to asses and express confidence would seem obvious the absence of numbers as most frequent family supports Hamm's (1991) findings that participants prefer verbal expressions to numerical

170

I report the number of occurrences only for a qualitative overview and they should not be used to compare groups since in this table they are not balanced by an index of the general verbosity of the group.

299


expressions of confidence. The dominance of a small numbers of families points to the importance of interplay between the use of pre-existing linguistic habits - the use of “to see” is quite natural in a visual discrimination task, for instance – and local dynamics priming, selecting and supporting one describing strategy instead of another. Certain morphemes display the structure of attractors. These morphemes tend to recur because of their affordances but given the local history of the linguistic interaction that tendency can be negated and replaced with another tendency in favour of other descriptors. For instance, group 29 has a very low frequency of occurrences of “to see” and instead it detailedly articulates a confidence scale based on numbers, a family of descriptors that is avoided by most other groups. We can observe the same phenomenon though to a lesser degree when we zoom in on any particular family. Individual families, with the exception of those that only occur once, tend to be organised along different degrees of certainty. The most frequent family articulates the whole scale, while the other families end up by covering only a certain area of the certainty scale as in the following example of groups 21 and 43.

300


Figure 25 – Two examples of confidence scales

The order of the single descriptors is approximative since the degrees of confidence covered are not discretely defined and therefore there is much overlapping and repetition. Two are the most interesting findings from this zoom in, displaying an interplay between common occurrences and the possibility for the single group to stabilise idiosyncratic linguistic habits. There are a few commonly recurring modifiers used to internally articulate the certainty scale across both families and groups – numbers excluded: 1. More traditional quantifiers (ikke [not], lidt [a little], helt [totally]) 2. What I call emotional quantifiers (sgu [a weaker version of “fucking”], fandme and satme [two versions of “damn”]). However, the same family of descriptors presents huge variations between groups. This is most remarkable in the confidence scale constituted by numbers. Only “55”, “60” and “not 100” (ikke 100) are shared by at least two groups. A less constrained scale such as for instance “sure” has, while being a very common family, less than half of the descriptors shared by at least two groups, only a handful shared by at least three groups and none shared by all groups. This means that the stability of the family does not revolve around a few crucial internal articulations, but that taken any two groups they will share items that are not shared by any other two groups. As an illustration of this point the following table compares four groups, highlighting the descriptors shared by at least two groups in red.

301


Figure 26 – Overlapping between group regarding the “*sur*” [*sik*] family

The idiosyncratic descriptors, that is to say, the descriptors used by no more than one group can be both quite common expressions like “almost sure” (næsten sikker), quite unusual expressions like “pink sure” (lyserød sikker) or “Brian Mikkelsen sure”171 (Brian Mikkelsen sikker) and even what could be deemed as grammatical and logical mistakes. In group 43 subject A says “I just took a chance” (Jeg gættede bare) to which B replies “I took a chance less [sic]” (Jeg gættede mindre). We can also observe 171

This expression refers to the Danish minister of Economic and Business Affairs. Translating the expression - and therefore partially changing its meaning – in an Italian context would mean to say “Tremonti-sure” which means been dead certain and still unavoidably doomed to be proved wrong.

302


two successive interactions from the second session of group 22: B: 15 A: shall we take mine, then? [skal vi ikke tage min saa?] B: yes [ja]

B defines his confidence as being “15”, to which A replies by discarding B’s choice due to her/his insufficient confidence in it. Everything seems normal. Except that in the next interaction something odd happens:

B: 5 A: okay, so we take 2 [B's choice] [okay, saa vi tager to]

Given an even lower confidence in B’s choice A decides to follow B’s choice anyway. However, 5 and 15 in group 22 have simply become an index of low confidence and – in the last third of the interactions of that group – any expression of low confidence will simply prompt the other subject to make a decision. Expressions of medium confidence will instead prompt the interlocutor to express her/his own confidence, while no recurring pattern can be found for the rarer cases of high confidence. Another case found in Group 32 is the specialisation of two families of descriptors over time: •

“To Believe” [At tro]: initially a generic quantifier of confidence, it becomes a quantifier used exclusively for the speaker’s own decisions in previous interactions.

•

“To Think” [At synes]: initially a generic and scalable quantifier of (un)certainty, it starts being used only as a modifier of “to see” [at se]

What we see at work is the creative reuse of existing norms and competences when 303


negotiating a set of inter-defined linguistic tools adapted to the ongoing practice. The external norms only play a limited role while internal norms of use are negotiated and developed. 3.3.3.3. Scaling through negotiation and alignment

The idiosyncrasy of many descriptors, however, prompted us to investigate more closely how they came to be. It seemed improbable that an explicit negotiation of which descriptors to use for articulating a confidence scale would lead to the peculiar and fuzzy scales described in the previous paragraph. Indeed, explicit negotiation does not seem to take place in any significant amount. Even less explicit questions like “are you sure?” or “what do you mean by …?” do not appear in most groups and even when they do, they do so in few occurrences without this having any consequence at all (as opposed to groups where such questions do not occur). What seems to happen is the spreading of implicit morphological and anaphoric alignment from one interlocutor to the other between speech turns and between interactions.

304


Figure 27 – Lineage patterns of descriptors

In the fragment here displayed it is easy to see a plurality of descriptors at play: to take a bet (at gætte), to know (at vide), to think (at synes) and to see (at se). It is also easy to see how one specific descriptor tends to be repeated more often and that this behaviour becomes more and more pronounced, that is, all descriptors tend to be used less and less except for “to see”. We labeled this phenomenon “pruning”.

305


Figure 28 – Lineage patterns of descriptors after pruning

The mechanism giving rise to such lineage patterns and their stabilisation can be described as lexical and anaphoric priming, that is, the fact that the use of an expression will make its further use more likely. In the case of lexical priming a lexical token introduced by interlocutor A is repeated by B in an adjacent speech turn: Group 21 (from interaction 66) A: I’m kind of fifty-fifty on this one B: I’m fifty-fifty too

In the case of anaphoric priming an expression introduced by A is referred to by B in an adjacent speech turn: Group 21 (Interaction 49)

306

Language and Cognition - Ph.D. Dissertation by Riccardo Fusaroli A I am not sure here, N* B Me fucking neither

Most theories of dialogue focus on displaying the conscious and heterogeneous strategies interlocutors use to accumulate a common ground, that is, “all the information that both interlocutors believe to be shared by themselves and their conversational partner” (Garrod & Pickering 2009: 443). This is clearly not what is happening in our case for two different reasons. First, what enables the interlocutors to perform the joint decision is lexical alignment and not propositional information. Second, such alignment seems to be implicit and hardly ever explicitly negotiated. However, our findings are compatible with at least some of the results in experimental pragmatics. In at least four different experimental paradigms ranging from elicitation of sentences via a confederate to the need of joint spatial coordination it has been shown that, while in theory interlocutors could achieve alignment of their models through explicit negotiation, in practice they normally do not (Brennan & Clark 1996; Clark & Wilkes-Gibbs 1986; Garrod & Anderson 1987; Schober 1993). Even more counter-intuitively actual negotiation seems to impair coordination and alignment (Galantucci & Steels 2008). It is quite unusual for people outside of the academic world to suggest a definition of a verbal expression and obtain an explicit assent from their interlocutor. Instead, “global” alignment of models seems to result from “local” alignment at the level of the linguistic representations being used. Pickering and Garrod (2004) have attempted to develop a model that can describe all the findings about this dialogic priming effectively, a model which they called “interactive alignment account”. The basic idea is that dialog is a form of joint action172 in which interlocutors have as their goal the aligning of their understanding 172

Following Sebanz, Bekkering and Knoblich (2006:70), in Chapter 1 I defined joint action as what happens when two or more individuals coordinate their actions in space and time to bring about a change in the environment. In the light of this chapter, however, I have to point out that Sebanz et

307


of the situation whether they agree about every detail or not (cf. Zwaan & Radvansky 1998). Interlocutors achieve such alignment by aligning at different linguistic levels concerned with words, grammar and sound, and such alignment leads to alignment of situation models. The alignment of situation models is achieved by a nonconscious and effortless priming mechanism. For example when describing pictures of objects to each other interlocutors use the same expressions to refer to them (e.g., Brennan & Clark 1996; see also Garrod & Anderson 1987). Similarly, they tend to repeat each other’s choices of grammatical constructions to describe pictures of events (Branigan, Pickering, & Cleland 2000). If the confederate173 were to describe a card as “The chef giving the jug to the swimmer”, the participant would tend to describe the next card as “The cowboy handing the banana to the burglar” but if the confederate were to say “The chef giving the swimmer the jug”, the participant would then tend to say “The cowboy handing the burglar the banana”. Such priming effects can be extremely strong and occur with other grammatical constructions (Cleland & Pickering 2003) and even between languages in bilinguals (Hartsuiker, Pickering, & Veltkamp 2004). The idea is that when listening to and interpreting the other’s utterances, the activation of the representation of a word or grammatical construction enhances its activation during production since the neural substrate for doing so is the same. Thus, alignment in dialog occurs automatically and without extensive negotiation between interlocutors or modelling of the other’s mental states. Additionally, it has been observed that alignment at one level enhances alignment at other levels. For example, grammatical alignment is enhanced when there is a repetition of words, in the case of which participants are even more likely to say “The cowboy handing the banana to the burglar” after hearing “The chef handing the jug to the swimmer” than after “The chef giving the jug to the

173

al. argue that such joint actions require the merging of both the individuals’ action plans to lead to shared representations: this is just yet another symptom of the exclusive emphasis on convergence. Joint action can – and most often does – involve complementary coordination as well. By “confederate” I refer to an accomplish of the experimenter posing as participant, in order to manipulate the conditions of the experiment.

308


swimmer” (Branigan et al., 2000). Indeed, grammatical alignment is enhanced for words that are semantically related, and thus the tendency to say “the sheep that’s red” (rather than the red sheep) is greater following “the goat that’s red” than following “the knife that’s red” (Cleland & Pickering 2003). In this case semantic alignment enhances grammatical alignment. Similarly the alignment of words leads to an alignment of situation models: people who describe things the same way tend to think about them in the same way too (Markman & Makin 1998). It can be claimed that these processes of alignment are not primarily intentional even though the interlocutors have as their goal to align their understanding. Potential misunderstandings can thus be solved in most cases by alignment at more behavioural levels - lexical or other - leaving negotiation as well as any other more sophisticated and potentially costly strategies that depend on modelling the interlocutor’s mental state for that minority of cases in which the former and more primitive mechanism fails (cf. Garrod & Pickering 2009). Our findings seem to fit with this model if we add a few specifications like the following: the already mentioned addition of anaphoric alignment, the extension of the alignment, strategies employed to reach it and the effect of having reached such alignment. In the specific task in which the participants are involved the main alignment we observe is a morphological one. There are two reasons the participants should not converge too strictly. The first is that they have to express and share their different confidences, therefore while sharing a common system to define confidence they should not also simultaneously align on the same descriptors. The second reason is that the participants do not have a pre-standardised conceptual or linguistic scale to graduate their confidence. They need flexibility and to be open to new situations. Therefore, while aligning at the morphological level they keep innovating the lexical level, generating – at least for the approximately 90 minutes of the interactions – an open ended repertoire of descriptors. In this case linguistic coordination implies a 309


combination of morphological convergence and lexical and syntagmatic creativity and complementarity. As regards the strategies employed to achieve this alignment we have already observed that initial linguistic interactions contain lengthy and repeated opening speech turns where uncertainty is declared again and again. These lengthy turns could – apart from being a consequence of the democratic game (§ 3.1) – also be an occasion to try out descriptors and partially align on them. Adding a last specification we can observe that alignment changes the ecology of the linguistic interactions. The number of families declines fast as well as does the number of repetitions, although at a slower rate. Convergence at a morphological level implies a decreased need to negotiate and/or restate alignment.

3.3.4.

Language and Performance

Given the results of the linguistic analysis our hypothesis would claim that the organisation of the interactions and the convergence on one family of descriptors should be positively correlated with the performance of the groups. We have already seen how independently from each other all the groups are able to effectively structure the interaction according to a common template that evolves and gets more efficient over time. Therefore, we could not find any correlation at that level. 3.3.4.1. Relating degree of morphological convergence and group benefit

However, the attempts at relating the degree of convergence on a dominant family lead to more interesting findings. If we take the number of occurrences of each family of descriptors in both the best and the worst performing group and if for each group we make a graph where the families are in order of number of occurrences we end up with the following two pictures:

310


Figure 29 – Distribution of families of confidence descriptors in Group 43

311

Language and Cognition - Ph.D. Dissertation by Riccardo Fusaroli Figure 30 – Distribution of families of confidence descriptors in Group 36

The distribution of families is clearly different between the two groups. In order to make them comparable we fitted a power function to those distributions, in other words, we calculated the power curve that would cross the y values (the number of occurrences expressed in percentages) with the highest approximation174.

Figure 31 – Degree of convergence on one family of descriptors: groups 43 and 36

It should be clear from the graph how there is quite a strong difference in the steepness of the two slopes. If we then draw such power functions for each group, we get the following result.

174

Such procedure – a standard transformation in statistical treatment of data – was needed to balance out the different degrees of verbosity (the different amounts of talk) between groups as well as to take into account the general distribution of all the families, no matter their number, through a single numerical index.

312


Figure 32 – Degree of convergence on one family of descriptors: all groups

We now have a single numerical index per group quantifying their degree of convergence on a single family of descriptors, that is, their developing of a tendency to use one family of descriptors much more than others. This index can be now tested for its correlation with the group benefit.

313


Figure 33 – Correlation between degree of convergence and group benefit175

The degree of correlation is 0.9763, that is very high, and the chances this correlation is an effect of chance (p value) is 0.0359, making the correlation statistically significant. We finally have a better grasp on what is happening throughout the linguistic interactions. The groups converge on one system to generate linguistic descriptors of confidence through a basic mechanism of alignment. Such a process relies on preexisting linguistic affordances, but it puts them to play according to the local history of the interaction. However, which family of descriptors acquires dominance in the

175

To the untrained eye, these results could look as a shapeless cloud, hardly displaying any correlation. This is, however, the reason why statistical techniques have been developed: to highlight patterns otherwise invisible to the eye and mind of the analyst.

314


conversational exchanges is not that important as long as there is convergence on one language to describe confidence the groups can achieve a group benefit, that is, they are able to share information and use it to outperform the accuracy even of the best of the two individuals in the group. There are, however, still a few elements to be considered: i) Could the degree of convergence just be an effect of the similarity of individual performance? ii) Is a shared scale the only factor that determines a group benefit or is there something else in the linguistic interaction that proves crucial? iii) What is the role of the shared scale not only in the performance of the distributed cognitive system but also – once developed – in the individual cognitive system? 3.3.4.2. The degree of morphological convergence is not an effect of the model

The first question can be answered by showing that there is no significant correlation between the difference in individual performance – that according to the WCS model would predict group benefit – and the degree of convergence.

315

Language and Cognition - Ph.D. Dissertation by Riccardo Fusaroli Figure 34 – Lack of correlation between degree of convergence and difference in individual performances

The correlation between the two sets of values presents a degree of correlation of 0,376 (quite weak) and a p value of 0,254, meaning that the correlation is not statistically significant. It becomes thus possible to hypothesise a more complex model to describe joint performance that would add the degree of convergence as a second independent variable: Group Benefit = K (difference in performance) + K (convergence) + C +

176

The group benefit becomes a function of both difference in individual performance and degree of morphological convergence. We could thus say not only that group benefits are achievable if the participants have a similar level of skills and expertise, as Bahrami and colleagues seem to conclude, but also that by aligning on a shared language participants can overcome the disparity of skills and successfully coordinate to achieve better results than they would alone. 3.3.4.3. Scales and interactions

The second question, a question related to the exact role of the shared scale, can be verified by constructing a further variation of the experiment177. The participants are now not allowed to interact linguistically. In case of disagreeing decisions, a graded bar appears on their screen on which they are asked to express their confidence by selecting one of the grades. Once both the participants have made their choice and they are shown each other’s confidence rate and one of the participants has to indicate the “joint” decision. In the original version of the experiment participants 176 177

This is of course just the very abstract form that such model could assume. This variation, the construction of which is carried on by Bahador Bahrami, Karsten Olsen and Dan Bang has not yet being published.

316


would developed a shared confidence scale in the course of their conversational interactions. In the present variation the graded confidence scale is pre-existing, culturally inherited. However, this situation does not lead to any group benefit, neither in the standard condition nor when the divergence in individual performances is increased by adding noise to the stimuli presented to the worse performer. This shows that the communication mediated via the graded bar is not rich enough. We can hypothesise that the major differences lie in the structure of the scale and in the role it plays in the interaction. In linguistic interactions we see the participants start from declarations of uncertainty and then negotiate their relative confidence, changing it over time. Moreover, the scale developed verbally is never a closed set of pre-established elements, but participants are able to carve locally fit nuances. The combination of these elements might help both motivation and the ability to reformulate one’s confidence in terms of the scale. Further tests are needed to better pinpoint the differences in the two interactions and kinds of scale that make a difference for the joint performance178. 3.3.4.4. Scales and learning

Finally comes the third question: once this scale has emerged via interaction which role does it play in individual performance? A further analysis has been done of the learning curve in individuals solving the task without communication as opposed the learning curve in individuals allowed linguistic interactions in case of disagreement. The results179 show clearly that the individuals allowed linguistic interactions present a steeper learning curve which means that they learn faster. The participants in the two groups start from the same level of individual performance. However, for interacting participants learning is so fast that the first point in the graph, collecting the results of the first 64 trials, shows their performance as already much higher that 178

179

Some of these tests are being planned. They involve the possibility for confidence to be negotiated via the bar as well as the manipulation of the competence of the participants through an initial training in a linguistic scaling system either explicitly or via exposure to videos of successful groups. Developed by Karsten Olsen (2010).

317


the performance of non-interacting participants.

Figure 35 – Comp arison betwe en indivi dual learni ng in intera ctive vs. nonintera ctive situati ons

Interestingly enough this increased learning effect is not replicated in the condition in which communication happens only via the bar. Again these results lead to further experimental variations to be tested. For instance, what would happen if the possibility of interacting were to be taken away after a scale had emerged? Would the individual performances still keep their accuracy? For how long would that last? However, it is now time to sum up the results and to more explicitly recast them in the pragmatist terms I have developed until now, in order to show how this experimental paradigm and the findings it has led to can articulate certain aspects of a pragmatist theory of language, and where in turn a pragmatist perspective can further contextualise and situate the experimental paradigm.

3.4. A Pragmatist Assessment This experimental paradigm allows us to observe and manipulate how language enables the progressive coupling of two subjects in a distributed cognitive system 318


aimed at the solution of a very specific task. Two cautions have to be put forward. The first one concerns the angle of analysis chosen: in order to develop quantifiable indexes that could be tested for correlation with joint performance other aspects have been left in the background. The emotional engagement for instance, the role of swearing, laughter and sometimes even flirting. Given the limited amount of groups analysed, personality profiles and the degree of acquaintance between the subjects has not been taken into account. But such is the nature of the scientific enterprise and of semiosis: only a limited amount of aspects can be focused upon at any given time. And I do believe that adding a quantitative relevance to the pragmatist framework developed so far can effectively contribute to its making a difference and to shape its conceivable consequences in a productive direction. The second caution concerns the situatedness of the linguistic and cognitive practice analysed. This – or indeed any – experimental paradigm and the ensuing findings do not grant us a grand theory of communication. It does not give us a model of language in general. However, they grant us an in-depth insight into a very specific aspect of language, namely its ability to be carved into a common ground between subjects in order to effectively share information. By constructing such a constrained niche it was possible to operationalise and apply the pragmatist framework. We have seen how individuals contribute to the solving of the task with their own understanding of the task180, their visual acuity and their ability of making a decision based on it. When disagreement occurs individual cognition and behaviour is not sufficient anymore and coordination and sharing of confidence has to be established. The participants choose to do so via linguistic interactions. The participants have to coordinate the structure of the interaction, a scale and the way to use it. All the groups in the experiment structured their interaction through a loose functional

180

Some preliminary data from the Chinese sessions of the experiment (introduced in the next paragraphs) show that a few subjects misunderstand the purpose of the study and instead of discriminating the image presenting the contrast oddball they attempt to discriminate differences of contrast in the fixation cross.

319


process apt to satisfy the democratic game, to reach a decision and to give the right emphasis to feedback. In this initial phase successful groups managed via alignment to converge on one morpheme-based family of descriptors for their confidence. These findings can be related to other experiments in which groups have shown to develop abstract and more stable representation than individuals (Schwartz 1995). Such convergence, however, goes hand in hand with at least two forms of complementarity. The first one is the lexical and syntagmatic variability of the descriptors motivated by the fact that the participants keep on negotiating the scale and have to express potentially divergent confidence and to renegotiate their confidence over time. The second form of complementarity lies in the way expressions of confidence are used as a tool for coordination. We have seen as, at least in certain groups, certain expressions of confidence assign the roles in the decision-making process, therefore expressions of low confidence will trigger a decision act in the interlocutor. This ongoing negotiation seems to be essential given the lack of group benefit shown by the bar variation of the experiment. This could be further related to the need of an intersubjective interaction in pain assessment in order to generate reliable results. However, clearly, not everything is up to the intersubjective dynamics. The successful groups all develop a graded one dimensional scale. The dynamics is needed to develop it, but the scale seems to have a stability that goes beyond the dynamics of interaction. Indeed, far from regulating only the group coordination, the development and enactment of a scale seems to feedback to individual cognition and behaviour, speeding up the learning process. This is a perfect example of the habit-led cognitive activity I have been sketching through this dissertation. The individuals have to put their habits in the public arena via linguistic expressions that are not reducible to purely abstract symbols but that are habit-supported. Such public arena reshapes the habits and the individual is changed in the process, at least temporarily. Linguistic patterns act as habitgrounded symbols (Nöth 2010), developing as constraints for the interaction and then feed-backing to more individual cognitive habits (Raczaszek-Leonardi & Kelso 2008, 320


Chapter 1). This public arena and linguistic patterns are however not prey of a radical relativism. For the participants to solve the task a scale has to be develop, a scale that in its heterogeneous linguistic manifestations maintains a strong constraining role on the possibilities of the interaction. The habits developing it and relying upon it have to comply to the constraints of the situation. We have also seen how the ongoing interaction is not the out-of-social-context kind of interaction that enactivists are positing. Social norms play a strong role, from preexisting linguistic affordances, to the democratic game – which refers to the constraints to the interactional moves due to the need of getting out of the experimental situation on friendly terms. The strong cultural component of the interaction highlighted by the analysis has led to the development of a handful of further variations. The repetition of the experimental paradigm in Beijing and Teheran is revealing very different patterns of interactions and very different patterns of joint performance181. Besides widening the range of cultural contexts in which the experimental paradigm is tested, we are also planning to manipulate the linguistic affordances for the construction of the scale either by preliminarily training the participants in the use of a specific linguistic family of descriptors or via exposure of the participants to other groups employing good practices. Another manipulation is to create a community of participants, each having to collaborate with each other in pairs over time, in order to see the development of a shared language beyond the single dyad. All this shows the complexity of human cognition even in the most basic perceptual processes. The contrast of the oddball is taken in a complex set of expectations and structured interactions. Its Secondness is emerges on the ground of a Thirdness that 181

The analysis is still in too early a phase for me to be able to report its results, since adapting our procedure of qualitative and quantitative analysis of the data to Mandarin Chinese and Farsi is even more time consuming than its Danish equivalent. However, Chinese participants seem particularly bad at coordinating and their joint performance is not significantly superior to the best individual’s performance. Moreover, for Chinese participants communication is not enough to overcome the lack of feedback on the correct answer. The Iranian data show that female participants are able to overcome even large differences in individual accuracy. And so on.

321


is distributed and evolving. Mapping these distributions and evolutions is the path on which the pragmatist perspective developed throughout this dissertation has led me. I have proved that the role and effectiveness of language in coordinating the two participants and their performance can be mapped and measured and can lead to further variations of the experimental paradigm. In this, I feel that my aim of developing an experimental side to a pragmatist cognitive semiotic approach has proved successful.

3.5. Conclusions Throughout Chapter 3 the pragmatist framework has been developed in order to better accommodate for the role of individualities, coordination processes and complementarity in the constitution, mechanisms and even dissolution of cognitive systems. This suggests that the conceptual framework at play should focus on the processes of emergence, distribution and integration of cognitive systems. To argue for such a view I have gone through that intellectual development that could be called social enactivism, articulating it to the socially grounded notion of habit. Since enactivism does not yet take language into account I attempted a partial integration of its themes first with the cases and insights from the material symbols literature and with experimental pragmatics. The pragmatist framework unveiled the proto-symbolic and habit based nature of chimpanzees’ use of tokens in lieu of abstract relations. Then I moved more fully into the linguistic domain. Via an experiment in multi-agent sensory integration I was able to investigate in detail the processes of linguistic coordination and alignment on a confidence scale. A group benefit was achieved through convergence on a morpheme-based confidence scale. Such convergence was however intertwined with lexical and syntagmatic creativity as well as with complementarity in the distribution of roles in the joint decision-making process. The construction of a somewhat fluid scale for confidence acted as a constraint for the individual cognitive systems both in the interacting and in the 322


individual condition. This scaling activity is not limited to the experimental paradigm I have thoroughly analysed but can be seen in other situations where confidence has to be communicated, in pain assessment and in aesthetic assessment. Scaling is, however, not the only function of language that can be analysed in this way. Categorisation and the construction of systems of references seem to highlight mechanisms analogous to the ones here analysed (Tylén et al 2011). This chapter has thus disclosed the possibility of applying the pragmatist framework developed in the previous chapters experimentally and it has shown the niche that it can fill both with respect to experimental psychology and to the more theoretical debate in distributed cognition and enactivism. The notion of habit can help focusing and analysing the role and evolution of regularities in behaviour and cognition, especially when the public arena of meaning is involved. It can point both to qualitative and quantitative patterns that can then be tested for correlation with more traditional measures in experimental psychology (like reaction time and success rate). When this approach is applied to linguistic interactions it can show the way linguistic patterns can emerge from, widen and constrain cognitive systems and processes.

323


4. Conclusions It is thus time to draw an account of this enterprise. In the introduction I stated that the aims of this dissertation were at the same time to investigate the role of language and communication in cognition and to develop the pragmatist semiotic framework towards a cognitive and experimental semiotics. The pragmatist framework endows us with an awareness of the complex interplay of Thirdness, Secondness and Firstness that constitutes human experience and cognition (and much more). It endows us then with a rich model of experience, that blends embodied, intersubjective and socio-cultural aspects. The crucial notion enabling this framework has been the notion of habit, that is, of a certain disposition to act, perceive and interpret in a certain way in certain circumstances. Habits are a way to enable behaviour and cognition to work smoothly and to harness uncertainty (“the irritation of doubt” Peirce would say). Habits are a flexible and relational structure where the resistance of the world and social interactions play a role. Thinking is – at least in certain occasions – like dancing tango; it is not done alone, without moving, without interacting, without making place for and relying upon the external world, for its structures, its resistances and its dynamics. This has two main related implications: the adaptivity (§ 3.1.4.1) of habits and their distributedness (§ 1.2.1.4). Organisms and humans in particular are plastic and adaptive: they do not simply rely on neural mechanisms inside the head. Human beings learn to deal with the world by interacting with it. In many occasions this interaction is interiorized. We learn our mother tongue by hearing it and talking it and we rely on those interactions by repeating parts of them and being corrected, but then we speak the language. When later on in life we learn a second (and a third) language such interiorisation is still happening, although at a minor degree, due to a decreased plasticity of the neural system. In other cases we keep on relying on those 324


interactions, like in diagrammatic reasoning to demonstrate a theorem or finding our way with a map (§ 2.4.2.2), or even dancing tango and landing a plane (§ 3.1.3). We have seen thus that distributed processes enter the formation of the habits that constitute the backbone of cognition, at different time scales along a continuum, from a phylogenetic, passing through an ontogenetic, to an online time scale. Cognition is thus a relational process aimed at •

perceiving the regularities and affordances (possibilities for further action, perception and, more generally, cognition) of the physical and socio-cultural environment, affordances that are due to the systems and the activities in which the individual cognitive system is, was, and conceivably could be.

•

manipulating these regularities and affordances, either by mental simulation or by actual engagement with the environment

•

representing and memorising these regularities and affordances in terms of modified dispositions to act, react and cognise but also in terms of dispositions to use symbols (be them material or internalised).

This perspective has grown through the discussion of the embodied, embedded, extended, enactive and distributed approaches to cognition (§ 1.2.1). What it contributes to them is a specific focus on the role of socio-cultural and intersubjective factors in shaping cognitive processes and even agencies. When we take language into account (§1.4) it relies on these adaptive and distributed habits, engaging them in a wider public arena. To explore this process I discussed: i) the distributed nature of the embodied grounding of language (§ 1.2.2, § 2.2.4, § 3.3.3), ii) the development of language in intersubjective interactions aimed at joint decision-making (§ 3.3), iii) the role socio-cultural norms (§ 1.4.4, § 2.4.3.2, § 3.3.3.3) and of the arena of the media (§ 2.3). The conceptual and empirical findings sketch a picture of language as shaping at different time scales the perceived and conceivable affordances of situations in a ductile way that is open to manipulation, negotiation 325


and alignment. In these investigations two dimensions became crucial: temporality and coordination. Distributed cognitive systems need to coordinate heterogeneous components through both convergence and complementarity. And this coordination happens over time through – sometimes reciprocal – adaptivity. A standard semantic form for “Berlusconi is a caiman” is converged over time in Chapter 2, dwelling on existing linguistic and semantic structures (idioms, conceptual metaphors, shared semantic traits, image-schemas) and putting them to play locally. The pairs of participants in Chapter 3 converge over time on standard scripts and morpheme-based scales of confidence. Their morphological convergence goes hand in hand with lexical syntagmatic creativity and complementary distributions of roles. They also display a complex interplay of pre-existing skills and local developments of the system. The cases investigated do not cover the whole spectrum of language structures and linguistic uses, however, they allowed me to show how a pragmatist semiotic framework can be usefully applied both to the cornerstone of cognitive linguistics, metaphors, and to its still untamed frontier, experimental pragmatics. Exploring these two cases has forced me to hybridise corpus linguistics (§ 2.4.1), textual (2.4.3.2.2) and philosophical analysis and experimental psychology (§ 3.3.3.4). This has raised many issues and many questions that while deserving further research, have found initial answers. The analysis of the evolution of semantic forms needs automated search procedures to produce relevant corpora beyond the single analyst's ability. On the other hand, however, it needs textual considerations to drive those procedures and to individuate dynamics that go beyond the plan of expression, the linguistic form. In the same way tracking the role of linguistic interactions and linguistic structures in the coordination of joint action and cognition requires a deep understanding of the experimental procedures and of the statistical treatments of the data, to which a semiotic sensibility can add the ability of opening the black box of the linguistic data and contribute to define the relevant factors for the measured 326


performance. By investigating the role of language and communication in cognition, both through the analysis of the conceptual structures at play in the mediascape, and through the analysis of joint decisions coordinated via language, it has been possible to show the applicability of pragmatist semiotic conceptual tools. This application integrates existing methodologies and conceptual frameworks with a specific attention to intersubjective and social dynamics through time, as well as the process of coordination via convergence and complementarity. This shows the possibility for a pragmatist cognitive and experimental semiotics in an interplay with more traditional approaches. Though the field of problems is the same – meaning and its transformations – the arena in which the game is played is different. This new arena requires a dialogue with our evolving knowledge of human cognitive systems and the ability of narrowing and testing hypotheses in the lab, even recurring to quantitative indexes, while not forgetting the dimension of meaning.

327


Appendix A - The first appearance of “Berlusconi is a caiman” Il Signore della propaganda e un paese senza passioni La Repubblica — 15 maggio 2002 pagina 15 sezione: PRIMA PAGINA

NEL Trecento svaniscono le libertà comunali, libertà facinorose, costellate da soprusi, in una faida permanente. Bartolo, cervello onnivoro (1313 o 12141357), dedica un trattatello alla tirannia, distinguendo due specie: «ex defectu tituli», l' usurpatore; «ex parte exercitii», chi abusa del potere acquisito legittimamente. Nonché due forme: una «manifesta»; l' altra «velata et tacita», sotto maschera costituzionale (sono grato al colto lettore napoletano che mi segnala i testi, molto attuali). En passant, due aneddoti su Bartolo: «subtilis sed non memoriosus», chiede le citazioni all' amico Francesco Tigrini; e mangia poco, pesando le quantità, perché non vuol offuscarsi la mente (Savigny, Storia del diritto romano nel Medio Evo, trad. Bollati, Gianini e Fiore, Torino 1857, II, 638). Veniamo all' establishment berlusconiano: gl' italiani l' hanno votato e, sinora almeno, rispetta le libertà fondamentali; ciò lo distingue dal governo manifestamente tirannico, ma l' anomalìa resta. Chiamiamolo regime personale: negli ultimi due secoli esiste un solo esempio; e il Mussolini 193640 naviga più cautamente, persino quando s' equipara al re nel maresciallato dell' Impero. Cosa vi sia d' abnorme, è presto detto: B. agisce da padrone; adopera Parlamento e governo quasi fossero botteghe Mediaset; alleati, satelliti, clienti, gli servono messa; asseverano qualunque cosa dica, e gliene sfuggono d' enormi. Ringraziamo il Cielo che non siano automi come Achille Starace, Joachim von Ribbentrop o Wilhelm Keitel. Nel loro piccolo ricordano piuttosto l' exvescovo Talleyrand o l' exoratoriano Joseph Fouché alle prese con Napoleone. Fosse meno assorbito dal culto a se stesso, coglierebbe punte fredde negli sguardi: è una corvée d' 328


ossequio; gliela prestano, coatti, aspettando le occasioni. Cominciamo ab ovo: come mai abbia tanto potere; e il fenomeno berlusconiano appare nuovo. Mussolini dispone d' una milizia al soldo statale. Hitler ne schiera due: le SA, vedove d' Ernst Roehm trucidato su ordine suo; e le SS, sotto Himmler, «Enrico il fedele». B. non possiede arnesi simili, né l' Italia 2002 ha istituzioni autocratiche (tribunali speciali o del Popolo, poteri eminenti nel capo del governo, cancelliereFührer, Gran Consiglio, organi legislativi castrati, ecc.). Mussolini gode d' un carisma naturale soverchiante, finché sciagure esterne non glielo spengano; Federzoni, Grandi, Ciano, Bottai, ruotano sommessi intorno all' astro; anche Farinacci ammutolisce; solo Balbo arrischia una cauta fronda in falsetto. L' uomo d' Arcore non irradia fluido magnetico. L' unico rango che gli competa, antropologicamente, è quello da principe dei «bagalun dl' lüster», gl' imbonitori che incantavano i contadini nelle fiere: loquela fluviale, sorrisi da caimano, gesto sovrabbondante, effetti ilari; niente lo predestinava al dominio politico. Costretto a partite dialettiche leali (sugli schermi gliele addomesticano), annegherebbe in mezzo bicchiere d' acqua. Nemmeno i soldi spiegano perché sia così forte: ne ha da scoppiare, l' uomo più ricco d' Italia, ma non può comprarsela; quando paghi gli elettori uno a uno, rimani al verde. Se tutto stesse lì, sarebbe un plutocrate qualunque, solo più grosso. Il suo potere è d' un genere etereo. Attraverso oscure intese s' era acquisito l' ordigno grazie al quale regna sulle anime: a colpi d' immagini e suoni, evoca mondi virtuali; monopolista dei canali, entra nelle teste, inocula affetti e fobie, insuffla finte idee, instilla decisioni; s' era allevata una massa contemplante, pronta al voto. Signorìa tanto meno resistibile, quanto più declinano le arti discorsive: siamo comunità tribali rispetto all' Atene dove disquisivano i sofisti, 24 secoli fa; e spiace dirlo, il tempo lavora dalla sua parte. Lo scenario suggerisce due rilievi. Primo: avendo sotto mano un arsenale, naturale che l' adoperi; è ridicolo aspettarsi del pudico selfrestraint. Secondo: se non vuol perderlo, deve difendere quel potere, e così l' accresce; ipnosi contro raziocinio, una partita diseguale. Libro e dialogo gli ripugnano. Nell' attesa, ne fabbrica d' innocui. Il pubblico ha 11 anni, ripete agli spacciatori. La politica berlusconiana sta in una frase: 329


quanto meno la gente pensa, tanto meglio; è un vizio da estirpare. Che sia a buon punto, consta dalle mercuriali: se l' uditorio fosse sveglio, nessun berlusconiano oserebbe dirsi liberale; invece battono bandiera. Ha una nemica, l' agorà o piazza nel senso ateniese, luogo della disputa, dove nessun illusionista sfida impunemente cose e fatti. Ora, essendo infettivo il pensiero, la profilassi comincia dal vocabolario (Orwell, Appendice a 'Nineteen eightyfour' ). I controllori lo riducono a parole elementari, così vaghe che, secondo le direttive, significhino una cosa o l' opposta: «amico», «nemico», «pace», «guerra», «ordine», «crimine», «liberale», «comunista», ecc.; mancando i vocaboli, le differenze svaniscono, fino all' atrofia dell' organo. I flussi verbali seguitano, anzi non s' era mai parlato tanto: lo stile laconico insospettisce, sintomo d' introversione; idem la sintassi, malfamata perché allarga proditoriamente gli spazi mentali, ma va deperendo a vista d' occhio e morrà da sola. Esistono sul mercato discorsi precostituiti, pronti all' uso, tanto comodi, recitabili come filastrocche. Gli anticorpi difendono dal contagio. L' integrato s' infuria quando sente frasi sature, ordinate, trasparenti: gli suonano ermetiche, stravaganti, oscure; e così vomita i veleni. Oltre alle sequele prêtàdire, dove non c' è niente da pensare, il regime promuove logorree farfalline, nonché gli usi viscerali della parola, quali invettive, bisbiglio estatico, ringhi, scherno. Veniamo alle opinioni soidisantes liberali. Ne sono sempre circolate ma l' attuale costellazione riesce nuova. Il precedente, molto diverso, risale agli anni 191924, quando persone serie applaudono il fascismoterapia anticomunista, illuse che poi restituisca i poteri alla decrepita classe dirigente (mummie come Salandra). Nella risposta al discorso della Corona (Palazzo Madama, 18 giugno 1921), Luigi Albertini, artefice e comproprietario del «Corriere», dà atto all' «opinione pubblica» d' avere «reagito», salvando l' Italia dai «gorghi del comunismo», ed è chiaro chi siano i salvatori, non opinanti ma squadristi. Mercoledì 9 agosto 1922 non ha ancora capito dove vada l' onda: rivendica il merito d' avere combattuto l' ipotesi d' un socialismo governativo; vitupera l' abortito «miserabile sciopero»; irride i riformisti Turati, Treves, Modigliani, ora questuanti un posto ministeriale mentre voltano «le spalle all' odiata 330


Destra e ai fascisti» suoi alleati (9 agosto 1922). Anche Croce riteneva terapeuticamente utili i fascisti: lo dicono tre interviste, 19224 (Pagine sparse, II, 37189); romperà i ponti nella veemente protesta 1 maggio 1925 contro il manifesto degl' intellettuali neri. Era paura quel consenso strumentale: paura della peste bolscevica; sindrome da cittadella assediata, sebbene l' occupazione delle fabbriche fosse innocua messinscena. Giolitti, infatti, al suo quinto ministero, non muove dito, lasciando che sbolla, salvo poi giocare la carta nera contro popolari e rossi, affinché rinsaviscano. I liberalconservatori professano bigottismo classista. L' ottuagenario cuneese commette un errore culturale: gli avventurieri interventisti l' avevano estromesso; è riapparso alla ribalta in piena tempesta; ragiona sul fenomeno fascista secondo vecchi modelli d' alchimia parlamentare. Mussolini contro Lenin, ma l' Italia 2002 non ha bisogno d' essere salvata; e siccome gli àuguri schivano l' autentico quesito, formuliamo chiaro: cosa bolle sotto la signoria berlusconiana. Col loro permesso, comincerei l' analisi dal vuoto culturale. Nell' autopanegirico B. è anche bibliofilo; impavido come al solito, indica due autori prediletti: niente meno che Tommaso Moro (14791535) e Guicciardini (14831540). Dio sa così abbia da spartire con l' utopista inglese amico d' Erasmo, cancelliere della Corona, inflessibile avversario d' Enrico VIII sino alla morte sul patibolo, o con l' algido storicopolitologo fiorentino. Deve averglieli nominati qualche consulente umanista, svagato o forse burlone, perché Utopia celebra l' anticapitalismo: beni in comune, sei ore d' onesto lavoro giornaliero al sevizio pubblico, vita quieta, divertimento intellettuale, crescita interiore. I referti d' autobiografia vanno presi con le molle, specie quando li firma un tycoon delle televisioni, nel tempo libero impresario calcistico: non incrudeliamo, quindi; certo, sarebbero archetipi più verosimili «Beautiful» o qualche storia su Tortuga, base della pirateria trionfante, nel cui Gotha, se i ricordi puerili non mi tradiscono, figura anche un Olonese. Pragmatismo, cantano i panegiristi. Converrà intendersi sulla parola: se «pragmatico» significa «attento ai fatti», «alieno dalle fantasmagorie» «abile nel calcolare serie causali e probabilità», Cavour lo era par excellence, avendo una dottrina molto chiara; l' aveva Depretis, computista del 331


lavoro politico minuto; Giolitti, altrettanto poco intellettuale, ne applicava una così aperta al futuro da spaventare gli uomini d' ordine. L' ha anche B., terribilmente povera e cruda: filosofia del successo, ossia Ego ipertrofico, sordità morale, dinamismi estroversi; il tutto aggravato dall' analfabetismo politico. Avviso ai caudatari, caso mai fossero tentati dall' idea d' indignarsi: non sono contumelie; ogni parola descrive fatti notori, dalle leggi pro domo sua all' assurda pretesa d' essere l' Unico, fuori d' ogni spazio legale, perché gl' italiani l' hanno votato. I suoi exploits non stupiscono più: vuole i tribunali genuflessi (immerso nei due autori, non aveva letto Montesquieu); e impone il silenzio sul groviglio d' interessi incompatibili. Eloquenti anche i gesti minuti: ad esempio, diserta il 25 aprile; nell' occasione eleva agli altari Edgardo Sogno, partigiano monarchico, poi golpista, suo confratello in P2. Tale possibile futuro dominus d' una repubblica presidenziale. Torniamo all' Italia post 1919. Spaventati dai rossi, i conservatori puntavano sul castigamatti nero, così ottimisti da credere che, domata la bestia, sarebbe rientrato nei ranghi: presupponevano un loro diritto naturale al governo; cabale da cervelli ammuffiti ma avevano una storpia razionalità. Ottant' anni dopo, avviene l' inverso. I moderni «liberal» saltano bravamente nel buio, sottomessi allo scorridore, senza l' alibi dei barbari alle porte: nessuno ha occupato le fabbriche, né bussano i soviet; barbaro semmai è l' unto dal popolo, con quel passato, mentre gli avversari avevano condotto al concerto europeo un' Italia decotta dalle consorterie dalle quali lui discende. L' alternativa attuale pende tra caute formule liberalsocialiste e un cesarismo amorfo, dove niente appare chiaro (meno che mai l' asserito liberismo, tollerabile solo fin dove gli venga comodo). I saltatori scelgono l' azzardo. Niente d' abominevole: a qualcuno piace l' avventura; né costituisce delitto inseguire l' en plein alla roulette berlusconiana, ma rispettino le parole. Non sta bene mistificare l' insegna, piantandola su un composto d' autocrazia, nichilismo affaristico, volgarità televisiva. FRANCO CORDERO

332


Appendix B – An example of transcript – Group 21 session 1 Interaction 01(S) B (0:01:31.9) A (0:01:33.3) B (0:01:35.2) A (0:01:35.2) B (0:01:35.5) A (0:01:36.6) A (0:01:38.9) A (0:01:40.6)

Jeg er sgu ikke helt sikker H* Jeg synes det var toern altsaa [Mhh ] [Skal vi] proeve? proeve toern Okay (.) den var rigtig (?) Naeste!

Interaction 02(S) A (0:02:20.0) B (0:02:21.2) B (0:02:22.2) A (0:02:24.1) B (0:02:25.8) B (0:02:27.9) A (0:02:28.4)

jeg var ikke helt sikker der N* nae jeg er 60 procent (mumler) Jeg tror du har har ret proeve? det var rigtigt .hh ja naeste.

Interaction 03(S) B (0:02:38.5) A (0:02:39.5) B (0:02:41.4) A (0:02:41.8) B (0:02:42.6) A (0:02:45.9) B (0:02:46.8) A (0:02:48.8) B (0:02:51.2) A (0:02:52.7) B (0:02:53.8)

jamh... jeg synes den var svaert at se ja det samme her mhh det var saadan lidt. jeg er fifty-fifty paa (den her) skal vi proeve en toern eeeh ja eeh du proever bare okay det var sgu rigtigt Naeste mmh

Interaction 04(F) A (0:03:23.1) B (0:03:24.9) A (0:03:29.9) B (0:03:30.9) A (0:03:33.9) B (0:03:34.0) A (0:03:34.8) B (0:03:35.7) A (0:03:38.7) A (0:03:45.2)

*hhh jeg var heller ikke sikker (mumler) jeg er ogs-oeh.. jeg haenger nok umid. 5 - 55 procent til ettern men saa tager vi ettern mmh det var rigtig det var rigtig (mumler) nej det er forke[rt det var] faktisk forkert [(mumler)] ja. naeste

Interaction 05(S) B (0:04:15.3)

.hhhh

333

Language and Cognition - Ph.D. Dissertation by Riccardo Fusaroli A (0:04:16.2) endelige afgoerelse til dig B (0:04:20.0) B (0:04:24.1) A (0:04:25.7) B (0:04:26.7) B (0:04:29.8) A (0:04:31.5) B (0:04:31.8) A (0:04:32.4) A (0:04:34.0) B (0:04:34.4)

jeg er ikke sikker (...?) jeg vil lade det vaere jeg vil overlade den end-den jaeeh jeg er ogsaa meget usikker paa min proev at tage toern som du sagde (?) fordi jeg eeh det er dig der skal tryk proever at tage toern (.) korrekt korrekt .hhhja det var toern den ja det sa jeg godt naeste mhhm

Interaction 06(S) A (0:04:42.3) B (0:04:44.8) B (0:04:49.7) A (0:04:50.4)

.hh lige ved at tro det var ettern jeg kom til at trykke forkert der proev at tryk (mumler) (.) ja det er rigtigt Naeste

Interaction 07(S) B (0:05:21.2) A (0:05:23.3) B (0:05:25.4) A (0:05:27.0) B (0:05:28.2) A (0:05:32.4) B (0:05:35.7) A (0:05:37.2) B (0:05:38.2) A (0:05:38.9) B (0:05:39.9)

jeg er meget i tvivl jeg synes de var naesten ens ja... hmm... hmm men oeh skal vi tage en ettern at koere paa proeve en ettern Det er op til dig H* jeg er sgu fifty-fifty paa den her saa tager vi toern wrong (word) saa var det forkert ja ettern var rigtigt ja naeste mh

Interaction 08(S) A (0:05:46.4) B (0:05:48.2) B (0:05:49.3) A (0:05:50.3) B (0:05:51.0) A (0:05:52.4) B (0:05:52.7) B (0:05:55.6) A (0:05:57.3) B (0:05:57.9)

det lader jeg vaere op til dig N*[ og traef den her endelige afgoerelse] [ja jeg tror og ] at der er et eller andet der siger mig at det er ettern ja jeg synes den virkede lidt jamen saa sys-saa det er dig der[ har du afgoer den her] [skal vi proeve den *KLIK*] ogsaa rigtig mhh naeste ja

Interaction 09(F) B (0:06:24.2) A (0:06:25.0) B (0:06:27.8) A (0:06:30.0) B (0:06:31.1) A (0:06:33.0) B (0:06:34.1)

.hh jae jeg synes det var ettern det her men ja proev at tage ettern det kunne godt vaere at jeg trykkede forkert jeg er ikke 100 procent sikker maaske oeh okay vi tager en ettern ja

334

Language and Cognition - Ph.D. Dissertation by Riccardo Fusaroli A B A B

(0:06:36.7) (0:06:38.0) (0:06:39.8) (0:06:40.3)

det var forkert ja det var toern ja naeste mh

Interaction 10(S) A (0:06:43.9) A (0:06:46.7) B (0:06:47.6) A (0:06:49.1) B (0:06:50.4) A (0:06:50.6) A (0:06:52.7) B (0:06:53.2) A (0:06:53.4) B (0:06:53.9)

*.hh(proev) den var svaer der er jeg ikke sikker der N* jeg er naesten sikker paa at det er ettern saa synes jeg det er [ dig de d afgoer den] [jeg synes den virkede] meget tydligere du afgoer den det var rigtigt ja naeste .hhh

Interaction 11(S) A (0:07:07.4) B (0:07:09.6) A (0:07:11.5) B (0:07:13.6) A (0:07:14.3)

der er jeg ikke helt sikker der N* jeg synes ettern den virkede mest tydlig [Klik] saa goer vi det saa'en, saa tager vi den det var ogsaa rigtigt ja naeste

Interaction 12(S) B (0:07:31.1) A (0:07:33.2) B (0:07:39.0) A (0:07:39.7) A (0:07:42.5) B (0:07:42.9) A (0:07:44.6) B (0:07:45.7)

der er jeg sgu usikker H* hvad (fanden vil du goere?) jeg synes det var toern altsaa men maaske 51 procent rigtig proev at tage toern okay korrekt= det var rigtig naeste mmh

Interaction 13(S) A (0:07:54.6) B (0:07:57.7) A (0:08:00.6) B (0:08:04.3) A (0:08:04.9) B (0:08:06.3) A (0:08:06.9) B (0:08:07.6) A (0:08:09.2) B (0:08:12.0) A (0:08:13.4) B (0:08:13.8)

*hvisker* den er svaer (daekker sin mund) njaeh jeg er sgu usikker paa jeg synes de var meget ens ja... eeh (.) hvad sa du jeg trykkede ettern du tror ettern [aabner munden men "afbrydes"] jeg er usikker skal vi proeve bare tage ettern he.hhja jeg er ogsaa usikker altsaa korrekt korrekt det er rigtigt okay naeste .hhh

Interaction 14(S) A (0:08:31.0) B (0:08:33.7) A (0:08:34.1)

jeg er naesten sikker paa at det er ettern det her toern det her er du det? ja

335

Language and Cognition - Ph.D. Dissertation by Riccardo Fusaroli B A B B B A B A A B A B

(0:08:35.6) (0:08:37.8) (0:08:38.6) (0:08:41.9) (0:08:43.3) (0:08:45.2) (0:08:45.8) (0:08:48.1) (0:08:51.2) (0:08:51.4) (0:08:52.1) (0:08:53.2)

Interaction 15(S) B (0:08:59.9) A (0:09:00.7) ettern saa ettern B (0:09:03.4) A (0:09:07.3) B (0:09:08.5) Interaction 16(S) B (0:10:05.4) A (0:10:06.4) B (0:10:08.5) A (0:10:08.9) B (0:10:09.4) B (0:10:11.7) A (0:10:12.2)

Okay (.) det er dig der... afgoer det emh (.) ja skal vi proeve at tage toern? Jeg er lidt usikker hmhhmh Jeg haelder mest til ettern men vi kan godt tage toern (klik) det var rigtigt det er sgu rigtigt ja ja naeste mmh jeg tror det er ettern ja jeg var heller ikke helt sikker paa den her altsaa[....] jeg trykker ogsaa [mmh]B (0:09:05.2) - her .hhh det var rigtigt, naeste den var rigtigt ja det er jeg fandme usikker paa der er jeg sikker paa at jeg synes jeg sa den i toern toern? ja okay den tager vi den var rigtig naeste

Interaction 17(S) A (0:10:18.1) der er jeg sikker paa toern igen N* B (0:10:19.4) er du det? A (0:10:19.8) ja det er jeg B (0:10:20.3) det er jeg sgu ikke A (0:10:22.1) er du ikke? B (0:10:23.5) neehjajeejajeeeoehhh JO oeh jo jo altsaa men jeg er usikker paa om jeg... jeg - kan ikke rigtigt bedoemme den A (0:10:27.8) skal vi proeve... B (0:10:29.5) ...bare tryk toern ikke' A (0:10:30.2) okay B (0:10:31.3) det er garanteret rigtigt B (0:10:32.7) det er det ogsaa A (0:10:33.6) naeste Interaction 18(S) A (0:11:07.3) B (0:11:09.2) A (0:11:12.9) A (0:11:14.6)

jeg er ikke sikker der jeg tror jeg trykkede forkert N* ja... jeg tror det er toern. det er jeg... 60 procent sikker paa (klik) ja ja du har ret naeste

Interaction 19(S)

336

Language and Cognition - Ph.D. Dissertation by Riccardo Fusaroli B (0:11:34.6) ja der er jeg godt nok lost A (0:11:35.9) det jaaa det er jeg ogsaa N* A (0:11:38.7) dues du sidder med afgoerelsen derover B (0:11:41.2) det er fifty-fifty A (0:11:41.9) jeg foelger-ja-men jegeh jeg er oeh jeg foelger med hvad du (klik) synes der er rigtigt du trykker nu. Det var det ogsaa. Naeste B (0:11:49.4) nnh Interaction 20(S) A (0:12:47.1) naeste. nej. (0:12:49.0)hov der var det forkert for en af os B (0:12:50.3) hmm (0:12:51.9)jeg vidste ikke-jeg oeh- jeg var usikker paa den det var derfor jeg tog - den (........) A (0:12:54.9) ja B (0:12:56.5) der er jeg virkelig fifty-fifty. saa det er din afgoerelse A (0:12:59.8) jeg proever at trykke toern B (0:13:00.9) ja A (0:13:00.9) jeg er ikke 100 procent men... naesten B (0:13:04.0) det skal man-det-jeg tror ogsaa det er rigtigt-jeg eroeh-jeg er usikker paa min A (0:13:07.3) det var rigtigt B (0:13:07.7) det var rigtigt A (0:13:08.8) naeste Interaction 21(S) A (0:13:14.7) B (0:13:16.7) A (0:13:18.9) B (0:13:22.1) A (0:13:22.6)

der var jeg ikke sikker der jeg er 55 procent sikker paa at det er-oeh-er oeh afgoerelsen er hos dig (klik) det var ogsaa rigtigt ja. Naeste.

Interaction 22(S) A (0:13:39.2) B (0:13:43.2) A (0:13:47.6) B (0:13:48.3) A (0:13:50.6) B (0:13:50.8) A (0:13:51.4)

der er jeg ikke sikker der N*, det er jeg ikke'. Du-du kan godt have ret der okay. Saa tror jeg jeg tror det er ettern… skal vi proeve og tage ettern? ja der e(t eller andet) der siger mig det var rigtigt det var rigtigt ja naeste

Interaction 23(S) B (0:13:56.6) A (0:13:57.7) B (0:14:00.8) A (0:14:03.4) B (0:14:03.5) A (0:14:04.3)

der er jeg lost ja. jeg [synes] jeg sa en lille forskel i toern saa tager vi den (.) ja det var ogsaa rigtigt naeste

Interaction 24(S) A (0:14:11.4) B (0:14:12.3) A (0:14:15.8) B (0:14:17.6)

jeg er ikke sikker der N* det er jeg sgu heller ikke hehhihihi det bliver fifty-fifty ja og det afgoer du okay (klik) saa tager vi toern

337

Language and Cognition - Ph.D. Dissertation by Riccardo Fusaroli A

(0:14:20.6)

Interaction 25(F) B (0:14:28.9) A (0:14:29.7) B (0:14:31.3) A (0:14:33.4) B (0:14:36.2) B (0:14:40.1) A (0:14:42.5) B (0:14:43.1) A (0:14:43.9)

*ja* det var rigtigt. naeste. u:ha ja der maa jeg indroemme der trykkede jeg ogsaa bare en jeg synes jeg sa det der en lille forskel i to'ern saa proever vi tager den (.) nej den var wrong. det var sgu ettern ja.naeste ja det var oeh. ja

Interaction 26(F) A (0:14:49.3) .hhheh A (0:14:50.4) jeg synes vi har mange split decisions nu her B (0:14:52.6) ja lige nu her... desvaerre. A (0:14:55.2) ja oeh B (0:14:55.4) hvad synes du H*? A (0:14:58.2) hh jamen jeg er ikke sikker N* det er jeg (.) det er jeg ikke'. Saa jeg trykkede -bare et eller andet B (0:15:04.2) det gjorde jeg sgu ogsaa (......).heh jeg kunne ikke se forskellen A (0:15:08.1) neeej B (0:15:09.4) ehm A (0:15:10.2) eeh altsaa det er. det er dig der skal trykke saa afgoerelsen er hos dig saa oeh -jeg ville lade det vaere op til dig og oeh afgoere det B (0:15:17.0) Saa (saetter jeg)oeh proever jeg at tage ettern (klik) A (0:15:18.3) ja B (0:15:21.2) wrong: A (0:15:21.7) mmh B (0:15:22.7) det var sgu forkert. det var toern A (0:15:25.0) ja. naeste Interaction 27(S) A (0:15:41.1) B (0:15:45.5) A (0:15:46.0) B (0:15:46.5) A (0:15:48.0) B (0:15:49.3) A (0:15:49.4) B (0:15:50.1) A (0:15:52.5)

jeg er ikke sikker der jeg synes der var i begge to der at-eh der varjeg synes den var -i begge to - mest (fremme) i toern okay (klik) proev at tage den ja det var ogsaa rigtigt det var rigtigt ja... .hh naeste

Interaction 28(F) A (0:16:10.9) B (0:16:11.7) B (0:16:13.2) A (0:16:14.7) B (0:16:17.8) A (0:16:22.2) B (0:16:22.6)

jeg er ikke sikker her N* det er jeg sgu heller ikke (helt) hmm hvad finder vi frem til hmm altsaa jeg er fifty-fifty. oeh. jeg kunne ikke se forskellen saa oeh nej jeg trykkede ettern

338

Language and Cognition - Ph.D. Dissertation by Riccardo Fusaroli A B A B A B B A B

(0:16:23.3) (0:16:24.4) (0:16:25.5) (0:16:26.3) (0:16:26.6) (0:16:28.1) (0:16:31.7) (0:16:32.9) (0:16:34.7)

ja bsh det samme her jeg ved det ikke skal vi proeve en ettern? ja som du trykkede det kan vi godt (.) jeg er ikke sikker det var sgu da dig der havde ret mmh okay. Naeste ja

Interaction 29(F) A (0:16:41.7) B (0:16:43.5) A (0:16:45.6) B (0:16:46.8) A (0:16:50.1) B (0:16:50.8) A (0:16:53.1) B (0:16:57.3)

saa er det dig der afgoer det men jeg tror altsaa at jeg trykkede forkert der (clausen) okay men saa skal du proeve en toer saa.. en toer (klik) aah det var sgu.... det var sgu forkert heh ja det var det. okay. vi (lavede) begge to fejl. naeste. ja- ja


der er jeg sikker paa at det er toern N* er du det ja det synes jeg jeg er sgu usikker saa den tager vi bare (klik) okay det var ogsaa rigtigt H* ja. Naeste

Interaction 31(F) A (0:18:05.7) B (0:18:07.5) sikker paa det A (0:18:11.6) B (0:18:12.4) B (0:18:14.6) A (0:18:15.2) B (0:18:16.4) Interaction 32(S) A (0:18:26.8) B (0:18:28.2) A (0:18:28.3) B (0:18:30.4) A (0:18:32.1) B (0:18:32.7) B (0:18:35.5) A (0:18:36.6) B (0:18:36.7) A (0:18:37.9)

jeg er ikke helt sikker der det er jeg sgu heller ikke men (.) jeg haelder mest til ettern men jeg er -ikke vi proever ettern (ehm) det var forkert ja. naeste jo jeg synes altsaa ettern den var... okay for jeg er usikker okay saa vi proever bare ettern hvis det er at du... det er dig der trykker okay (klik) (.) det var ogsaa rigtigt ja naeste (snoefter) naeste

Interaction 33(S)

339

Language and Cognition - Ph.D. Dissertation by Riccardo Fusaroli B (0:19:07.5) jeg er sgu jeg er totalt usikker der (mumler!) inden du trykkede A (0:19:10.0) ja jeg er ogsaa meget der oeh (.) jeg synes der var en lille forskel i ettern men - jeg er ikke 100- [100 procent sikker N*] B (0:19:16.1) -vi proever ettern [fordi jeg er ikke] B (0:19:18.5) okay. vi proever ettern A (0:19:19.4) yes (.) B (0:19:21.3) det var ogsaa rigtigt A (0:19:21.8) mmh A (0:19:23.0) naeste B (0:19:23.5) (snoefter) Interaction 34(F) A (0:19:45.9) B (0:19:51.7) B (0:19:58.7) A (0:20:00.5) B (0:20:05.6) A (0:20:06.0) B (0:20:06.7) A (0:20:07.5) B (0:20:09.0)

naeste. hov (.) (0:19:50.8)der er jeg ikke sikker det er jeg heller ikke (.) ehm (.) jeg tror faktisk din er rigtig (.) saa trykker du (klik) (.) naeh det var forkert *det var forkert ja* det var en ettern. Naeste ja

Interaction 35(F) A (0:20:14.2) B (0:20:14.5) A (0:20:16.5) B (0:20:17.6) A (0:20:20.8) B (0:20:21.1) A (0:20:21.7) B (0:20:22.4) A (0:20:24.0)

jeg er ikke sikker der er jeg heller der er jeg total usikker du ja jeg tror- vi proever en ettern der ja (.) det [ var ogsaa forkert] [det var (...) din der var rigtig] ja= =ja naeste

Interaction 36(F) A (0:20:39.4) B (0:20:40.8) A (0:20:43.8) B (0:20:44.1) A (0:20:44.5) B (0:20:48.7) A (0:20:50.6)

jeg er ikke sikker der N* jeg synes bare jeg sa lidt et eller andet der var tydligere i toern ja oppe i det ene hjoerne ja det proever vi (.) det var sgu forkert hrmh naeste

Interaction 37(S) A (0:21:30.3) B (0:21:32.6) A (0:21:34.4) B (0:21:35.0) A (0:21:37.1) B (0:21:38.4) A (0:21:40.3) B (0:21:41.5) A (0:21:42.0)

jeg synes altsaa to'ern den var lidt jeg kunne slet ikke... jeg synes toern (her den var) ja oeh jeg er usikker paa den det er fifty-fifty du afgoer du trykker saa tager vi toern hvis du er lidt mere sikker paa det ja det var ogsaa rigtigt ja naeste

340

Language and Cognition - Ph.D. Dissertation by Riccardo Fusaroli B

(0:21:42.8)

Interaction 38(F) A (0:22:02.1) B (0:22:03.4) sikker slet ikke A (0:22:10.9) B (0:22:11.8) A (0:22:11.9) B (0:22:15.4) A (0:22:16.4)

(snoefter) jeg er ikke sikker der N* det er jeg sgu heller ikke (.) ehm jeg haelder mest til ettern men jeg er ikke jep-je proever en ettern hmm fordi jeg er heller ikke det var sgu fork[ert] du havde ret jo [ja ] jae naeste

Interaction 39(S) A (0:22:38.8) jeg synes jeg eh sa en forskel i toern B (0:22:42.9) okay A (0:22:44.0) [(jeg ved ikke)] jeg ved ikke hvor sikker du var B (0:22:46.7) [ oeoehhhm ] men jeg er (.) (0:22:50.5) jeg er ikke sikker paa noget overhovedet oeh saa - vi tager vi tager bare toern A (0:22:57.9) (klik) ja B (0:22:58.2) det var sgu rigtigt H* A (0:22:58.7) ja naeste Interaction 40(S) A (0:23:19.2) jeg er ikke sikker der N* B (0:23:21.3) nnej det er jeg sgu heller ikke (.) (0:23:24.1) jeg kunne vitterligt ikke se det saa jeg -trykkede bare ettern A (0:23:26.9) skal vi proeve og give det en ettern B (0:23:28.1) ja A (0:23:28.6) okay (.) A (0:23:30.9) det var rigtigt B (0:23:31.6) det var sgu rigtigt A (0:23:32.5) naeste Interaction 41(S) A (0:23:37.6) B (0:23:39.5) A (0:23:45.8) B (0:23:47.7) A (0:23:49.9) B (0:23:51.7) A (0:23:52.7)

der er jeg er ikke sikker der det er jeg sgu heller ikke oehm *msk skal vi oeh skal vi proeve ja du trykker hvad du [synes] [jeg proever] at tage en ettern men jeg er satme ikke sikker nej (.) det var rigtigt ja naeste

Interaction 42(S) A (0:24:17.7) B (0:24:19.5) A (0:24:20.2) B (0:24:21.0) A (0:24:22.2) B (0:24:23.6) A (0:24:24.8) B (0:24:29.8) A (0:24:31.7)

der er jeg ikke sikker jeg tror jeg trykkede forkert okay du jeg trykkede toern [ men] jeg ved det sgu ikke [ ja ] men jeg tror [ogsaa du ] [ jeg synes bare] der var lidt tydligere i to'ern ja vi giver det en to'er (.) det var sgu rigtigt H* mmh naeste

341


Interaction 43(F) B (0:24:45.5) A (0:24:47.0) B (0:24:48.0) A (0:24:49.7) B (0:24:50.7) A (0:24:53.3) ikke B (0:24:57.1) A (0:24:59.7) B (0:25:00.2) A (0:25:01.0)

jeg er overhovedet ikke ehm nej jeg er ikke sikker paa noget som helst (i den her runde) jeg synes bare der var lidt mere[ paa oeh paa toern] [(gaber dybt) ja] - men det kan godt vaere at det maaske ikke er men jeg skal proeve to'ern jo ! jo bare proev wrong det var sgu forkert ja naeste


der er jeg ikke sikker N* der [(..........................)] [jeg synes jeg saa] en lille smule mere i ettern ja jeg er ikke sikker. skal vi proeve (...) ja (klik)det var det var rigtigt nok naeste

Interaction 45(S) B (0:25:23.9) A (0:25:25.4) B (0:25:27.2) B (0:25:31.2) A (0:25:33.4)

jeg er overhovedet ikke sikker H* jeg synes det var lidt mere i toern det oeh har du set det skal vi proeve at trykke det (.) det er korrekt ja ja naeste

Interaction 46(F) A (0:25:38.7) jeg er ikke sikker der det kunne lige saa godt vaere toern som B (0:25:42.2) jaeh A (0:25:43.4) som ettern som jeg [ trykkede ] B (0:25:43.8) [jeg er ogsaa meget] usikker paa dem A (0:25:44.9) du afgoer det (.) B (0:25:49.3) jeg synes bare der var en anelse mere i ettern det kan godt vaere du har -rjeg proever at trykke ettern (klik) (.) B (0:25:54.5) det var oeh A (0:25:55.2) det var din (....) B (0:25:55.8) det var to'ern ja A (0:25:58.1) naeste Interaction 47(S) A (0:26:03.3) B (0:26:04.3) B (0:26:06.6) A (0:26:08.4) B (0:26:12.1) A (0:26:15.3) B (0:26:18.9) A (0:26:21.4)

der er jeg ikke sikker det er jeg sgu heller ikke jeg synes de var-der var ikke ret meget (variable) i den mmhh (.) det er din afgoerelse om (det er)-jeg er fifty-fifty paa den her mhh jeg proever en to'er her jeg synes denmmi min intuition siger mig hja (.) det er korrekt naeste

342


Interaction 48(S) A (0:27:07.5) B (0:27:08.5) A (0:27:11.0) med -afgoerelsen saah B (0:27:16.0) A (0:27:19.8) B (0:27:21.2) B (0:27:23.6) A (0:27:25.2) B (0:27:25.5)

jeg er ikke sikker der N* jeg er sgu usikker paa det *msk* maaske det var en ettern men oeh jeg er ikke sikker. det er dig der sidder jeg er helt blank paa det oeh det er saadan ligesom fifty-fifty eller hvad tag hvad der foelger-falder dig ind (gaber og trykker) ettern den var rigtig ja naeste (snoefter)

session 2 Interaction 01(S) A (0:00:11.7) B (0:00:13.6) A (0:00:14.4) B (0:00:18.2) A (0:00:19.6) B (0:00:21.8) A (0:00:22.3)

jeg er ikke sikker N* Det er jeg sgu heller ikke du afgoer det (.) det var korrekt saa naar du trykker er du klar ja

Interaction 02(S) A (0:00:52.4) B (0:00:54.3) A (0:00:55.7) B (0:00:57.7) A (0:01:02.0)

jeg er helt sikker paa det var ettern okay saa er din nok *decision ja det var rigtigt ja

Interaction 03(S) B (0:01:04.4) A (0:01:07.7) B (0:01:11.7)

det er forkert jeg trykkede forkert det ettern det er dig der trykker ja .hhhh

Interaction 04(S) A (0:01:20.6) B (0:01:22.9) A (0:01:24.7) A (0:01:28.6) B (0:01:29.8) A (0:01:30.7) A (0:01:33.7) B (0:01:34.6)

jeg er ikke sikker (.) det er jeg sgu heller ikke hvad goer vi (.) skal vi proeve toern ja okay (klik) det var rigtigt ja

Interaction 05(S) B (0:01:54.8) A (0:01:55.9) A (0:02:01.1) A (0:02:02.2) B (0:02:03.4)

jeg er ikke sikker naeh det er jeg heller ikke (.) ja.h skal vi tage ettern. (.) jeg ved det ikke N* (kan vi det) bare

343

Language and Cognition - Ph.D. Dissertation by Riccardo Fusaroli A B A B A B A B

(0:02:06.2) (0:02:07.0) (0:02:08.1) (0:02:09.6) (0:02:10.3) (0:02:10.3) (0:02:13.4) (0:02:13.9)

d.hh okay vi [...] [jeg er flad oehh jeg ved det ikk'oehh] min indskydelse er ettern saa det okay skal vi proeve det ja det var rigtigt hhhja er du klar

Interaction 06(F) A (0:02:21.4) B (0:02:24.8) A (0:02:26.1) B (0:02:29.2) A (0:02:30.5) B (0:02:31.4) A (0:02:32.0)

der er jeg ikke sikker jeg tror jeg har forkert der okay skal vi proeve og tage ettern ja (.) det var forkert du havde ret H* mmhh hmm er du klar ja

Interaction 07(S) B (0:03:13.8) A (0:03:16.3) B (0:03:18.2) A (0:03:23.5)

jeg er naesten sikker paa at det er ettern den der vi proever ettern (klik) det var ogsaa rigtigt ja


jeg er sgu usikker jeg synes det var toern den her okay (.) proev at tage toern (.) rigtigt (ja)


jeg er sgu usikker ja ((banging noise)) hvad siger du jeg synes du var lidt mere jeg synes der var lidt i toern skal vi proeve og tage toern (.) det var forkert hmrja

Interaction 10(F) A (0:04:59.0) B (0:05:01.1) A (0:05:02.6) A (0:05:05.6) B (0:05:06.2) A (0:05:07.6)

der er jeg ikke sikker der N* Naesten sikker paa at det er ettern vi tager en ettern (.) det var forkert det var sgu forkert ja

Interaction 11(S) A (0:05:12.5) A (0:05:16.6) B (0:05:18.2) B (0:05:19.7)

jeg er ikke sikker maaske jeg trykkede forkert (.) der (.) du sidder [ med] afgoerelsen N* [hmmm] ( ) ta' toern

344

Language and Cognition - Ph.D. Dissertation by Riccardo Fusaroli A B A

(0:05:21.4) (0:05:23.1) (0:05:24.0)

Interaction 12(S) B (0:05:30.2) A (0:05:31.6) B (0:05:33.9) Interaction 13(S) A (0:05:52.4) B (0:05:56.1) hjoerne dernede A (0:05:59.9) A (0:06:02.5) B (0:06:03.3)

det goer jeg bare det var rigtigt ja der er jeg usikker der synes jeg der var noget i toern der bare tag den for jeg er meget usikker paa det det var ogsaa rigtigt der er jeg ikke sikker .hhhhhuh jeg synes der var et eller andet mere tydligt i toern i det ene vi proever toern [det var rigtigt] [( )]

Interaction 14(S) A (0:06:15.4) B (0:06:17.3) A (0:06:19.1) A (0:06:20.0) B (0:06:21.6)

der er jeg ikke sikker der jeg synes der var lidt ved ettern den var lidt [tydligere [vi proever den proev den ( den varr rigtig )


der er jeg usikker (.) helt vildt jeg synes der var lidt i toern der proev at tage toern du trykker ( ) toern (.) det var rigtigt hmm

Interaction 16(S) A (0:06:45.3) B (0:06:46.6) A (0:06:49.0) B (0:06:51.1) A (0:06:53.9) B (0:06:54.5)

jeg er ikke sikker det er jeg heller ikke proev ettern som du trykker proev ( ) jeg er meget usikker paa den det var rigtigt ja

Interaction 17(S) B (0:07:03.5) A (0:07:07.3) B (0:07:09.9) A (0:07:10.8) B (0:07:14.6) A (0:07:15.3) B (0:07:16.9) A (0:07:20.6)

ettern (.) ja hvis du synes det saa tager vi ettern jeg er ikke sikker H* jeg synes der var lidt oeh forskel i toern mmmh du sidder med afgoerelsen og trykker ( ) Tag toern jeg er usikker det var rigtigt ( ) ja

Interaction 18(S) B (0:07:34.2)

uha

345


(0:07:37.7) (0:07:39.2) (0:07:40.0) (0:07:42.0)

Interaction 19(F) B (0:07:56.6) A (0:07:58.3) B (0:07:59.7) A (0:08:02.9) B (0:08:03.4) Interaction 20(S) B (0:08:26.7) A (0:08:28.5) B (0:08:29.3) X (0:08:32.6) A (0:08:36.1) B (0:08:37.1) H* A (0:08:45.3) B (0:08:48.6) A (0:08:49.5) A (0:08:52.8) A (0:08:57.8) B (0:09:01.5) A (0:09:08.2) B (0:09:10.1) A (0:09:13.6) B (0:09:15.6) A (0:09:18.5) B (0:09:19.6)

jeg synes det var fifty-fifty den her det er jeg ogsaa fifty-fifty du sidder med afgoerelsen saa mmh det bliver fifty-fifty (.) ettern. det var rigtigt det var ogsaa fifty-fifty (skal vi saa sige ) toern der ja (klik) det var [forkert det var ( ) okay [hhmm okay ja naa ((B's telephone interrupts)) jeg tror... lige et oejeblik (.) jeg saetter den lige paa lydloes. Hvad tror du det er for en jeg tror det er ettern (.) ettern du havde toern lad os proeve toern (.) jeg tror du trykkede toern der ikke' (.) hmm hvad skal vi hvad skal vi afgoere her hvad skal vi sige oehm oerrhm (.)ja det ved jeg sgu ikke mnh hvad synes du mhhmmh maaske der var lidt i toern proev at tage toern (klik) det var rigtigt det var rigtigt

Interaction 21(F) B (0:09:42.6) A (0:09:44.7) B (0:09:46.1) A (0:09:48.8) A (0:09:53.0) B (0:09:53.5) A (0:09:54.1)

den synes jeg var rigtig svaer ja det synes jeg ogsaa oehm.... Jeg haelder mest til toern men jeg er ikke sikker vi... ja ehm det siger vi proev den det var saa [ikke rigtigt [den var ikke rigtig ja

Interaction 22(F) A (0:10:10.7) B (0:10:12.4) A (0:10:13.6) B (0:10:16.7) A (0:10:19.1) B (0:10:20.1) A (0:10:23.5) B (0:10:25.2)

hmm hhhjeg er ikke sikker der N* det er jeg sgu heller ikke det kan ligesaa godt vaere (.) ettern som toern synes jeg jeg synes de var meget ens ( ) ja mbh du sidder med afgoerelsen N* okay (.) saa tager vi toern okay det var sgu ettern ja

346

Language and Cognition - Ph.D. Dissertation by Riccardo Fusaroli Interaction 23(S) B (0:10:33.7) A (0:10:34.7) B (0:10:37.3)

der er jeg ogsaa usikker der synes jeg der var lidt i toern der proev at tage toern ja

Interaction 24(F) B (0:10:57.6) A (0:10:58.8) B (0:11:01.5) A (0:11:05.7) B (0:11:08.0) B (0:11:12.8)

jah jeg er heller ikke sikker her. det er jeg ikke det er jeg sgu heller ikke helt. saa maaske oehm hvad er din intuitions ( ) min intuition den siger ettern. men jeg er langt fra sikker (.) det var ogsaa forkert

Interaction 25(F) B (0:11:20.8) A (0:11:22.3) B (0:11:23.7) A (0:11:26.0) B (0:11:27.2) A (0:11:27.7)

jeg er totalt usikker ( ) jeg synes toern den var lidt proev proev at tage den (.) det var forkert ja okay

Interaction 26(F) B (0:11:37.9) A (0:11:41.4) B (0:11:43.0) A (0:11:43.8) B (0:11:45.3) A (0:11:47.1) B (0:11:49.0) A (0:11:52.3) B (0:11:53.0) A (0:11:53.6)

jeg synes det va:r jeg synes ettern var lidt tydligere .hh jamen saa lad os [tage ettern [men jeg er ikke helt sikker saa tag den okay du trykker det er mig okay. det var sgu forkert du mhh det var da vildt ja

Interaction 27(S) B (0:12:22.4) A (0:12:24.4) B (0:12:25.2) A (0:12:26.7) B (0:12:29.2)

der er jeg usikker jeg synes ettern proev at tage ettern det er dig der trykker det var rigtigt


den var svaer at se ja jeg synes der var lidt i toern proev at tage toern jeg er usikker. det var saadan fifty-fifty gaet ((sighs)) det var sgu rigtigt H* ja

Interaction 29(S) B (0:13:55.9) A (0:13:56.4) B (0:13:57.5)

.hhh jeg synes helt bestemt [( )[jamen det var det ogsaa

347


(0:13:58.4) (0:13:58.8) (0:14:01.3) (0:14:01.7)

-i ettern ja jeg skulle lige til at sige dernede at jeg kom tror jeg trykkede den forkert 'kay proev at tage ettern (klik) det var ogsaa rigtigt

Interaction 30(F) A (0:14:12.1) B (0:14:14.3) A (0:14:15.5) A (0:14:17.6) B (0:14:19.3) A (0:14:20.5)

der er jeg ikke sikker maa jeg sige jeg er naesten sikker paa det er toern vi proever en toer det var ettern alligevel det var sgu ettern jah

Interaction 31(F) A (0:14:31.1) B (0:14:33.9) A (0:14:34.5) A (0:14:36.6) B (0:14:38.1) A (0:14:38.7) B (0:14:39.5)

det kunne ligesaa godt vaere dig der har ret der tror jeg tror det er ettern ja (.) det var sgu toern det var fandme toern det var det sgu O?

Interaction 32(S) B (0:14:56.1) A (0:14:57.9) B (0:15:01.3) A (0:15:02.7) B (0:15:05.4) A (0:15:06.4) B (0:15:07.1)

det er jeg meget usikker paa det er jeg ogsaa. utroligt. og det er dig der: skal trykke saa du: det er mig der har afgoerelsen det har du min intuition siger mig det er ettern hmm det er det ogsaa


hmmm jeg tror det jeg tror jeg trykkede forkert. jeg tror det er en toer tror du det. proev. (klik) det var rigtigt ja

Interaction 34(S) A (0:16:05.4) B (0:16:06.9) A (0:16:08.7) en toer B (0:16:13.2) ikke 100 procent sikker A (0:16:16.7) B (0:16:19.9) B (0:16:21.0) A (0:16:22.5) Interaction 35(S) B (0:16:27.9) A (0:16:31.9)

((opens mouth)) er du sikker paa den her nej jeg er ikke sikker paa den overhovedet ikke (.) det kan ligesaa godt vaere hmm. Jeg synes der var lidt i toern jeg haelder mest til toern men jeg er sgu nej jeg tror jeg giver dig ret lad os proeve en toer ( ) det var ovre til venstre ( ) det var ogsaa rigtigt ja det var det uha den var svaer [jeg er i hvertfald total usikker heheh hahahah os' her...

348

Language and Cognition - Ph.D. Dissertation by Riccardo Fusaroli B A B A B A B A B

(0:16:32.4) (0:16:33.2) (0:16:34.9) (0:16:35.5) (0:16:38.5) (0:16:40.0) (0:16:41.4) (0:16:41.8) (0:16:42.9)

det kunne jeg i hvert fald ikke se nej. men du sidder med afgoerelsen er det mig ja det er saa (.) du faar... det bliver intuition? det her ja *det er i orden* ettern den var rigtig ja

Interaction 36(F) A (0:16:48.5) B (0:16:50.4) A (0:16:51.7) B (0:16:52.5) A (0:16:56.0) B (0:16:57.0) A (0:16:57.5) B (0:16:59.0) A (0:16:59.8) B (0:17:00.7) A (0:17:01.4)

aahha den var ogsaa svaer den her ja jeg synes bare den var lidt mere tydlig den foerste men jeg er sgu ikke sikker vi proever ettern mmh ( ) det var [forkert [det passede - det passede sgu ja *de:t*

Interaction 37(S) A (0:17:16.3) B (0:17:17.9) A (0:17:19.6) B (0:17:21.3) B (0:17:24.3) A (0:17:25.4)

jeg[( ) [jeg sgu jeg sgu ikke se oeh naesten sikker paa det var ettern proev at tage den (.) det var rigtigt ja


jeg er usikker der ja men jeg synes der var i toern der proev og-saa tager vi toern (.) [det var forkert [det var forkert [ja [ja

Interaction 39(F) A (0:17:54.7) B (0:17:55.5) A (0:17:56.1) B (0:17:56.9) A (0:17:58.5) B (0:17:59.9) A (0:18:03.0) B (0:18:05.1) A (0:18:07.4) B (0:18:08.8)

ej [den var svaer ej er ikke sikker der det er jeg heller ikke hvad goer vi det er dit-din oeh jeg er fifty-fifty jeg tror det du trykker er rigtigt jeg tror jeg haelder mest til ettern. det var forkert det var sgu det jeg trykkede hja

349

Language and Cognition - Ph.D. Dissertation by Riccardo Fusaroli Interaction 40(S) A (0:18:52.0) B (0:18:53.1) A (0:18:54.6) B (0:18:57.3) A (0:18:58.3) B (0:18:58.7)

[jeg ( ) [jeg er naesten sikker paa det er ettern men saa proever vi ettern for jeg var ikke sikker det var ogsaa rigtigt ja det var oppe i: hoejre hjoerne

Interaction 41(S) B (0:19:23.1) A (0:19:26.1) B (0:19:27.6) A (0:19:29.7) B (0:19:29.9) A (0:19:31.5) B (0:19:32.0)

ja det var saadan fifty-fifty for mig i den der a synes toern den var lidt saa proev at tage den oeh eller det er mig der har den ja det er dig jeg proever at tage toern saa for jeg er sgu usikker paa den ja det var sgu ogsaa rigtigt

Interaction 42(F) B (0:19:40.6) A (0:19:41.2) B (0:19:42.3) A (0:19:43.3) B (0:19:44.0) A (0:19:44.9) B (0:19:46.6) A (0:19:48.5) B (0:19:50.5)

[hmm] [a ]a ikke sikker her er du ikke sikker nej det er jeg heller ikke ((coughs)) jeg tror du har ret ( ) jeg tror det er toern okay. det var sgu ettern N* ja

Interaction 43(S) B (0:21:02.1) A (0:21:03.9) B (0:21:05.6) B (0:21:08.0)

jeg er ikke 100 procent sikker jeg tror det var toern N* saa tager du toern (.) det er rigtigt

Interaction 44(F) B (0:21:17.2) A (0:21:17.9) B (0:21:21.4) A (0:21:22.8) B (0:21:24.2) A (0:21:27.4)

jah jah du har du har afgoerelsen her N* for jeg er ikke sikker proev at tage toern det var sgu ikke [( ) [det er kraftedme oeh ukorrekt ( ) ( )

NB ... meta interaction E (0:21:27.7) I kan godt lige tage en 10 sekunders pause hvis det er B: det er okay A: jeg ved ikke... vil du have en pause N* B: ej det er sgu fint nok vi koerer A: vi koerer Interaction 45(S) A (0:21:42.1) B (0:21:44.2)

der er jeg naesten sikker paa en toer der okay den tager vi bare

350

Language and Cognition - Ph.D. Dissertation by Riccardo Fusaroli B A B A B A B A B B

(0:21:46.1) (0:21:48.1) (0:21:48.3) (0:21:49.2) (0:21:49.9) (0:21:50.3) (0:21:51.6) (0:21:52.9) (0:21:53.3) (0:21:56.9)

det er mit DAMP der spiller mig et puds okay jeg er alt for usikker okay yes saa traekker jeg laesset ja [ N* ] indtil du er oppe igen [heheheh] yes selvfoelgelig goer vi det

Interaction 46(S) A (0:22:09.6) B (0:22:12.1) A (0:22:12.7) B (0:22:13.4) A (0:22:14.5) B (0:22:16.9)

jeg tror du har ret jeg tror jeg trykkede forkert der N* ((sighs))ja er du sikker lidt toer toer toer *ja (.) det var rigtigt ja

Interaction 47(S) B (0:22:27.6) A (0:22:28.8) B (0:22:30.3) B (0:22:31.7)

der er jeg usikker H* jeg er naesten sikker paa ettern saa tager vi bare ettern (.) det var ogsaa rigtigt

Interaction 48(S) B (0:22:46.1) B (0:22:49.5)

Jah det er din afgoerelse (.) det var rigtigt

Interaction 49(F) A (0:23:01.2) B (0:23:04.7) B (0:23:07.4) A (0:23:09.6) B (0:23:12.5) A (0:23:13.1) B (0:23:14.9) A (0:23:16.4)

a er ikke sikker der det kunne ligesaa godt vaere dig okay. det er din afgoerelse. jeg er ogsaa fifty-fifty. Jeg haelder bare mest til toern men jeg er sgu ikke sikker ja vi proever toern det var fandme [forkert [det var fandme ikke ( ) okay det var helt vildt ja

Interaction 50(S) A (0:23:19.6) B (0:23:21.5) A (0:23:22.1) B (0:23:22.4) B (0:23:24.7) A (0:23:26.3) B (0:23:27.5) B (0:23:30.8) A (0:23:31.4)

der er jeg naesten sikker paa toern der N* er du det ja det er jeg det er utroligt jeg ikke kunne fokusere paa nogen. Hvad fanden sker der for mig du afgoer det okay vi tager toern (.) det er rigtigt ja

Interaction 51(S) A (0:24:07.1)

a er ikke sikker der

351

Language and Cognition - Ph.D. Dissertation by Riccardo Fusaroli B A B A

(0:24:08.5) (0:24:10.1) (0:24:11.2) (0:24:13.7)

det er jeg heller ikke du afgoer det okay (.) det var rigtigt ja

Interaction 52(S) B (0:24:24.2) A (0:24:25.6) B (0:24:27.6) A (0:24:28.4) B (0:24:29.5)

aahhh jeg trykkede forkert du har ret med ettern ikke ogsaa jo (.) jeps

Interaction 53(F) B (0:24:46.4) A (0:24:47.7) B (0:24:49.2) A (0:24:51.3) A (0:24:53.4) B (0:24:54.1) A (0:24:54.6) B (0:24:55.3)

der er jeg usikker ja det er jeg ogsaa jeg tror du har ret faktisk du trykker (.) naa [( ) [du havde sgu ret ja

Interaction 54(S) B (0:25:34.4) B (0:25:35.9) A (0:25:39.3) B (0:25:40.9) B (0:25:43.6) A (0:25:46.3)

ja:h (.) jeg haelder til toern men jeg er sgu ikke sikker ( ) langt fra okay jeg er sikker paa ettern okay saa trykker du ettern (.) det er godt (.) det er rigtigt mmh

Interaction 55(S) B (0:26:28.9) ja:h (.) det er (.) hvad siger du A (0:26:32.7) a synes der var lidt mere i ettern B (0:26:35.6) ja jeg er meget ( ) det er saadan (.) jeg tror vi tager ettern oe:h jeg er ikke sikker paa det er toern er rigtigt her. B (0:26:45.1) det var ogsaa rigtigt Interaction 56(F) B (0:26:51.4) A (0:26:52.3) B (0:26:53.5) A (0:26:54.4) B (0:26:58.9) A (0:27:00.4) A (0:27:03.8) B (0:27:05.0) A (0:27:05.4)

u:ha ja den var [godt nok [huhuhaha den var fandme svaer hhuhhhha oeh hvad goer vi jeg sidder med trykkern og hvad skal vi det er din afgoerelse jeg kunne slet ikke se det aa:h det er et skud jeg proever et skud i ettern (.) det var forkert ja det var toern

352

okay


5. Bibliographical References Adams, F. (2010). Why we still need a mark of the cognitive. Cognitive Systems Research, 11, 324-331. Adams, F., & Aizawa, K. (2001). The bounds of cognition. Philosophical Psychology, 14, 43-64. Adams, F., & Aizawa, K. (2008). The bounds of cognition. Malden, MA: Blackwell Publishing. Adamson, T. (2005). Metaphorical Cognition: Two Models. In S. Coulson and B. LewandowskaTomaszczyk (Eds.), The. Literal/Non-Literal Distinction. Berlin: Peter Lang. Adamson, T. (2007). Cognition and Conflation: Addressing a paradox in cognitive linguistics. Cognitive Semiotics, 2007(15), 87-101. Agamben, G. (2002). L’aperto. L'uomo e l'animale. Torino: Bollati Boringhieri. Ahrens, K., & Say, A. L. T. (1999). Mapping Image-Schemas and Translating Metaphors. In Proceedings of 13th Pacific Asia Conference on Language, Information and Computation (pp. 95-102). Academia Sinica. Allan, K. (2008). Metaphor and metonymy: a diachronic approach. Oxford: Wiley-Blackwell. Allwood, J. (2003). Meaning potentials and context: Some consequences for the analysis of variation in meaning. In H. Cuyckens, R. Dirven, J. R. Taylor. Cognitive approaches to lexical semantics (pp. 29-66). Ammer, C. (1989). It's raining cats and dogs--and other beastly expressions. New York: Paragon House. Anciano, D. (1985). Can Language Be Used to Assess Pain? A Review. Current Psychological Research & Reviews, 4, 167-182. Antonietti, A., Sempio, O. L., & Marchetti, A. (2006). Theory of mind and language in developmental contexts. New York: Springer. Appadurai, A. (1996). Modernity at large: Cultural dimensions of globalization. Minneapolis: University Of Minnesota Press. Aristotle. (1967). Poetics. Ann Arbor: University of Michigan Press.

353


Ash, M. (1997). Gestalt Psychology in German Culture, 1890-1967: Holism and the Quest for Objectivity. Cambridge & New York: Cambridge University Press. Ashby, W. R. (1952). Design for a brain. New York: Wiley. Auvray, M., Lenay, C., & Stewart, J. (2009). Perceptual interactions in a minimalist virtual environment. New ideas in psychology, 27(1), 32-47. Baddeley, A. D. (2000). Short-term and working memory. In A. D. Baddeley (Ed.), The Oxford Handbook of Memory (pp. 77-92). New York: Oxford University Press. Bailey, D., Chang, N., Feldman, J., & Narayanan, S. (1998). Extending embodied lexical development. In Proceedings of the 20th Conference of the Cognitive Science Society (p. 84). Madison (WI): Lawrence Erlbaum Associates. Ballard, D. H. (1991). Animate vision. Artificial intelligence, 48(1), 57-86. Baron-Cohen, S., Leslie, A. M., & Frith, U. (1986). Mechanical, behavioural and intentional understanding of picture stories in autistic children. British Journal of Developmental Psychology, 4(1), 113-125. Barsalou, L. W. (1999). Perceptual symbol systems. Behavioral and brain sciences, 22(04), 577-660. Barsalou, L. W. (2009). Simulation, situated conceptualization, and prediction. Philosophical Transactions of the Royal Society London B Biological Sciences, 364, 1281-9. Basso, P. (ms.) Metafora e scuole linguistiche dalla prima metà del secolo a oggi. Available on

http://www.sssub.unibo.it/documenti/seminario_metafora/basso.doc Bateson, M., Nettle, D., & Roberts, G. (2006). Cues of being watched enhance cooperation in a realworld setting. Biology Letters, 2(3), 412. Beer, R. D. (1995). A dynamical systems perspective on agent-environment interaction. Artificial Intelligence, 72(1-2), 173-215.

354


Beer, R. D. (2003). The dynamics of active categorical perception in an evolved model agent. Adaptive Behavior, 11(4), 209. Beer, R. D., & Chiel, H. J. (1993). Simulations of cockroach locomotion and escape. In Proceedings of the workshop on "Locomotion Control in Legged Invertebrates" on Biological neural networks in invertebrate neuroethology and robotics (pp. 267-285). San Diego, CA: Academic Press Professional, Inc. Bergman, M. (2009). Peirce's philosophy of communication: the rhetorical underpinnings of the theory of signs. London and New York: Continuum. Bermúdez, J. L. (2003). Thinking without words. Oxford: Oxford University Press. Bertrand, D. (2000). Précis de sémiotique littéraire. Paris: Nathan. Black, M. (1962). Models and metaphors. Ithaca: Cornell University Press. Black, M. (1993). More about Metaphor. In A. Ortony (Ed.) Metaphor and Thought (2nd ed.). Cambridge & New York: Cambridge University Press. Blackburn, S. (1984). Spreading the word: groundings in the philosophy of language. Oxford: Oxford University Press. Blanchard, E. B., Andrasik, F., Arena, J. G., & Teders, S. J. (1982). Variation in meaning of pain descriptors for different headache types as revealed by psychophysical scaling. Headache, 22, 137-9. Boers, F., & Demecheleer, M. (1997). A few metaphorical models in (Western) economic discourse. Discourse and perspective in cognitive linguistics, 151, 115-129. Bonfantini, M. A. (2003). Introduzione. In Opere (C. S. Peirce). Milano: Bompiani. Borghi, A. M., & Cimatti, F. (2010). Embodied cognition and beyond: Acting and sensing the body. Neuropsychologia, 48(3), 763-773. Boroditsky, L. (2001). Does Language Shape Thought? Mandarin and English Speakers' Conceptions of Time. Cognitive psychology, 43(1), 1-22.

355


Bottiroli, G. (1993). Retorica: l'intelligenza figurale nell'arte e nella filosofia. Torino: Bollati Boringhieri. Bourdieu, P. (1977). Outline of a Theory of Practice. Cambridge & New York: Cambridge University Press. Bourdieu, P. (1980). The logic of practice. Stanford: Stanford University Press. Bowerman, M. (1996). Learning how to structure space for language: A crosslinguistic perspective. Language and space, 385-436. Boysen, S. T., Bernston, G. G., Hannan, M. B., & Cacioppo, J. T. (1996). Quantity-based interference and symbolic representations in chimpanzees (Pan troglodytes). Journal of experimental psychology. Animal behavior processes, 22(1), 76. Bradac, J. J. (2001). Theory comparison: Uncertainty reduction, problematic integration, uncertainty management, and other curious constructs. Journal of Communication, 51(3), 456-476. Brandt, L. (2010). Language and enunciation – A cognitive inquiry with special focus on conceptual integration in semiotic meaning construction. Ph.D Dissertation, Aarhus Universitet. Brandt, L., & Brandt, P. A. (2005). Making sense of a blend: A cognitive-semiotic approach to metaphor. Annual Review of Cognitive Linguistics, 3(1), 216-249. Brandt, P. A. (2004). Spaces, domains, and meanings: essays in cognitive semiotics. Brüssel: Peter Lang. Branigan, H. P., Pickering, M. J., & Cleland, A. A. (2000). Syntactic co-ordination in dialogue. Cognition, 75(2), 13-25. Bréal, M. (1897). Essai de sémantique:(science des significations). Paris: Hachette. Brennan, S. E., & Clark, H. H. (1996). Conceptual pacts and lexical choice in conversation. Learning, Memory, 22(6), 1482-1493. Brennan, S. E., & Hanna, J. E. (2009). Partner Specific Adaptation in Dialog. Topics in Cognitive Science, 1(2), 274-291.

356


Brigati, R. (2009). La verità sospetta. Per una lettura di Verità e menzogna in senso extramorale. In G. Martufi & U. Spadoni (Eds.) Identità mobile (pp. 29-57). Pesaro: Edizioni Nobili. Brinkmann, F. (1878). Die Metapheren: Die Thierbilder der Sprache. A. Marcus. Brooks, R. A. (1999). Cambrian intelligence: the early history of the new AI. Cambridge, MA: MIT Press. Brugman, C. (1995). Give and Take in conversation: a reinvestigation of the conduit metaphor. In Proceedings of the 4th International Cognitive Linguistics Conference. Albuquerque. Buck, J., & Buck, E. (1976). Synchronous fireflies. Scientific American, 234(5), 74. Bundgaard, P. (2007). The cognitive import of the narrative schema. Semiotica, 165(1-4), 247-261. Bundgaard, P. (n.d.). Cognition and Event Structures. Almen Semiotik, 15, 78-106. Bundgaard, P. (n.d.). Morfologiske typer og intellektuelle tokens. Almen Semiotik, 9-10, 63-83. de Bustos, E. (1992). La metáfora y la Filosofía Contemporánea del Lenguaje. Anthropos, 37-42. Buzzanell, P. M., & Burrell, N. A. (1997). Family and Workplace Conflict Examining Metaphorical Conflict Schemas and Expressions Across Context and Sex. Human Communication Research, 24(1), 109-146. Cacciari, C. (2001). La metafora: da evento del linguaggio a struttura del pensiero. Milano: Raffaello Cortina. Cadiot, P., & Visetti, Y. M. (2001). Pour une théorie des formes sémantiques: motifs, profils, thèmes. Paris: PUF. Calcaterra,

R.

(2001).

Human

Nature

and

Rational

Norms.

Retrieved

from

http://www.digitalpeirce.fee.unicamp.br/calcater/p-humcal.htm Cameron, L. (2003). Metaphor in educational discourse. London & New York: Continuum. Cameron, L. (n.d.). Metaphor and talk. In R. W. Gibbs, Jr (Ed.) The Cambridge Handbook of Metaphor and Thought. New York: Cambridge University Press. Cameron, L., & Deignan, A. (2003). Combining large and small corpora to investigate tuning devices around metaphor in spoken discourse. Metaphor and Symbol, 18(3), 149-160. Camic, C. (1986). The matter of habit. American Journal of Sociology, 91(5), 1039-1087.

357


Cappuccio, M., & Wheeler, M. (2010). The sign of the hand. Versus, 113-115. Casasanto, D. (2009a). Embodiment of abstract concepts: Good and bad in right-and left-handers. Journal of Experimental Psychology: General, 138(3), 351-367. Casasanto, D. (2009b). When is a linguistic metaphor a conceptual metaphor. New directions in cognitive linguistics, 127-145. Casasanto, D., & Jasmin, K. (2010). Good and Bad in the Hands of Politicians: Spontaneous Gestures during Positive and Negative Speech. PloS one, 5(7), e11805. Charteris-Black, J. (2000). Metaphor and vocabulary teaching in ESP economics. English for Specific Purposes, 19(2), 149-165. Charteris-Black, J. (2003). Speaking with forked tongue: A comparative study of metaphor and metonymy in English and Malay phraseology. Metaphor and symbol, 18(4), 289-310. Charteris-Black, J. (2004). Corpus approaches to critical metaphor analysis. Basingstoke: Palgrave-Macmillan. Chen, J. Y. (2007). Do Chinese and English speakers think about time differently? Failure of replicating Boroditsky (2001). Cognition, 104(2), 427-436. Chomsky, N. (1984). Modular Approaches to the Study of the Mind. San Diego: State University Press. Church, K. W., & Hanks, P. (1990). Word association norms, mutual information, and lexicography. Computational linguistics, 16(1), 22-29. Churchland, P. M. (1992). Reconceiving Cognition. Minnesota Studies in the Philosophy of Science, 15(15), 475-480. Churchland, P. S., Ramachandran, V. S., & Sejnowski, T. J. (1994). A Critique of Pure Vision. Largescale neuronal theories of the brain, 23. Cienki, A. (1998). Metaphoric gestures and some of their relations to verbal metaphoric expressions. In J.-P. Koenig, Discourse and cognition: Bridging the gap (pp. 189-204). Stanford: CSLI Publications.

358


Claiborne, R. (1988). Loose Cannons and Red Herrings: A Book of Lost Metaphors. New York: Norton. Clark, A. (1991). Microcognition: Philosophy, cognitive science, and parallel distributed processing. Cambridge, MA: MIT Press. Clark, A. (1993). Associative engines: Connectionism, concepts, and representational change. Cambridge, MA: MIT Press. Clark, A. (1997). Being there: Putting brain, body, and world together again. Cambridge, MA: MIT Press. Clark, A. (1999). An embodied cognitive science? Trends in cognitive sciences, 3(9), 345-351. Clark, A. (2001). Mindware: An introduction to the philosophy of cognitive science. Oxford: Oxford University Press. Clark, A. (2004). Natural-born cyborgs: Minds, technologies, and the future of human intelligence. Oxford: Oxford University Press. Clark, A. (2006). Material symbols. Philosophical psychology, 19(3), 291-307. Clark, A. (2007). Curing cognitive hiccups: A defense of the extended mind. The Journal of Philosophy, CIV(4), 163-192. Clark, A. (2008). Supersizing the mind: Embodiment, action, and cognitive extension. New York: Oxford University Press. Clark, A. (2009). Spreading the Joy? Why the Machinery of Consciousness is (Probably) Still in the Head. Mind, 118(472), 963-993. Clark, A., & Chalmers, D. (1998). The extended mind. Analysis, 58(1), 7-19. Clark, D. A. (1990). Verbal uncertainty expressions: a critical review of two decades of research. Current Psychology: Research & Reviews, 9, 203-235. Clark, H. H. (1996). Using language. Cambridge & New York: Cambridge University Press. Clark, H. H., & Wilkes-Gibbs, D. (1986). Referring as a collaborative process. Cognition, 22(1), 1-39.

359


Cleland, A. A., & Pickering, M. J. (2003). The use of lexical and syntactic information in language production: Evidence from the priming of noun-phrase structure. Journal of Memory and Language, 49(2), 214-230. Clowes, R., & Morse, A. F. (2005). Scaffolding cognition with words. Proceedings of the Fifth International Workshop on Epigenetic Robotics: Modeling Cognitive Development in Robotic Systems. Coulson, S., & {Van Petten}, C. (2002). Conceptual integration and metaphor: An event-related potential study. Memory & Cognition, 30(6), 958. Craig, K. D., & Prkachin, K. M. (1978). Social modeling influences on sensory decision theory and psychophysiological indexes of pain. Journal of Personality and Social Psychology, 36, 805-15. Croft, W., & Cruse, D. A. (2004). Cognitive linguistics. Cambridge & New York: Cambridge University Press. Cussins, A. (1990). The connectionist construction of concepts. Oxford: Oxford University Press. Danaher, D. (1998). Peirce's semiotic and conceptual metaphor theory. Semiotica, 119, 171-208. Danesi, M. (1993). Vico, metaphor, and the origin of language. Bloomington: Indiana University Press. Daugman, J. G. (1988). Complete Discrete 2-D Gabor Transforms by Neural Networks for Image Analysis and Compression. IEEE Transactions on Acoustics Speech and Signal Processing, 36(7), 1169-1179. Davies, E. E., & Bentahila, A. (1989). Familiar and less familiar metaphors: an analysis of interpretations in two languages. Language & Communication, 9(1), 49-68. De Bruin, L., & De Haan, S. (2010). Enactivism & Social Cognition: In Search of the Whole Story. The Companion of Cognitive Semiotics, 1. De Jaegher, H., & Di Paolo, E. (2007). Participatory sense-making. Phenomenology and the Cognitive Sciences, 6(4), 485-507.

360


De Jaegher, H., Di Paolo, E., & Gallagher, S. (2010). Can social interaction constitute social cognition? Trends in cognitive sciences, 441-447. De Jaegher, H., & Froese, T. (2009). On the role of social interaction in individual agency. Adaptive Behavior, 17(5), 444. De Saussure, F. (1972). Cours de linguistique générale. Paris: Payot. De Tienne, A. (1996). L'analytique de la représentation chez Peirce: la genèse de la théorie des catégories. Brüssel: Facultés universitaires Saint-Louis. De Tienne, A. (2004). Is phaneroscopy as a pre-semiotic science possible? Semiotiche, 2. Deacon, T. W. (1997). The symbolic species. New York: Norton. Deetz, S. (1984). Metaphor analysis. In W. Gudykunst & Y. Kim (Eds.), Methods for Intercultural Communication Research (pp. 215-228). Beverly Hills: Sage. Degler, T. (1989). Straight from the horse's mouth--and other animal expressions. Basingstoke: PalgraveMacmillan. Deignan, A. (1999). Corpus-based research into metaphor. In G. Low, L. Cameron, Researching and applying metaphor (pp. 203-220). Cambridge & New York: Cambridge University Press. Deignan, A. (2005). Metaphor and corpus linguistics. Amsterdam & Philadelphia: John Benjamins. Deignan, A. (2008). Corpus Linguistics and Metaphor. In R. W. Gibbs, Jr (Ed.) The Cambridge Handbook of Metaphor and Thought (pp. 280-294). Cambridge & New York: Cambridge University Press. Deleuze, G., Guattari, F., & Deleuze, G. (1980). Mille plateaux. Paris: Éditions de minuit. Demuru, P. (2010). Brasile: il calcio, la storia, l'identità. Un esercizio di semiotica della cultura. Dennett, D. C. (1969). Content and consciousness. Atlantic Highlands, NJ: Humanities Press. Dennett, D. C. (1978). Brainstorms. Montgomery. Cambridge, MA: MIT Press. Dennett, D. C. (1991). Consciousness Explained. Boston: Little, Brown and Co.

361


Dennett, D. C. (1996). Darwin's dangerous idea: Evolution and the meanings of life. New York: Simon & Schuster. Derrida, J. (2008). The animal that therefore I am. New York: Fordham University Press. Descola, P. (1994). In the society of nature: a native ecology in Amazonia. Cambridge & New York: Cambridge University Press. Deubel, H., Clark, J. J., & Rensink, R. A. (2000). Picture changes during blinks: Looking without seeing and seeing without looking. Visual Cognition, 7(1), 191-211. Dewell, R. (2005). Dynamic patterns of CONTAINMENT. In B. Hampe (Ed.) From Perception to Meaning: Image Schemas in Cognitive Linguistics (pp. 369-393). Berlin & New York: Mouton de Gruyter. Di Paolo, E. (2005). Autopoiesis, adaptivity, teleology, agency. Phenomenology and the Cognitive Sciences, 4(4), 429-452. Di Paolo, E. (2009). Extended life. Topoi, 28(1), 9-21. Dines Johansen, J. (1995). Ikonicitet i litteratur. K&K, 78. Ditchburn, R. W., & Ginsborg, B. L. (1952). Vision with a stabilized retinal image. Nature, 170, 36-37. Dobrovol'ski, D. O., & Piirainen, E. (2005). Figurative language: cross-cultural and cross-linguistic perspectives. Bingley: Emerald. Dodge, E., & Lakoff, G. (2005). Image schemas: From linguistic analysis to neural grounding. In B. Hampe (Ed.) From Perception to Meaning: Image Schemas in Cognitive Linguistics (pp. 57-91). Berlin & New York: Mouton de Gruyter. Dokic, J., & Proust, J. (2002). Simulation and knowledge of action. Amsterdam & Philadelphia: John Benjamins. Donald, M. (1991). Origins of the modern mind: Three stages in the evolution of culture and cognition. Cambridge, MA: Harvard University Press.

362


Donald, M. (2001). A mind so rare: The evolution of human consciousness. New York: Norton. Dotov, D. G., Nie, L., & Chemero, A. (2010). A Demonstration of the Transition from Ready-toHand to Unready-to-Hand. PLoS ONE, 5(3). Druzdzel, M. J. (1989). Verbal uncertainty expressions: literature review. Technical Reports of Carnegie Mellon University, MU-EPP-1990-03-02. Eco, U. (1968). La struttura assente: introduzione alla ricerca semiologica. Milano: Bompiani. Eco, U. (1975a). Trattato di semiotica generale. Milano: Bompiani. Eco, U. (1975b). Trattato di semiotica generale. Milano: Bompiani. Eco, U. (1985). Sugli specchi e altri saggi. Milano: Bompiani. Eco, U. (1997). Kant e l'ornitorinco. Milano: Bompiani. Eco, U. (n.d.). Il limite e la soglia. In C. Paolucci, Studi di Semiotica Interpretativa. Milano: Bompiani. Eco, U., & Paci, C. (1983). The scandal of metaphor: Metaphorology and semiotics. Poetics Today, 4(2), 217-257. Eerden, B. (2009). Anger in Asterix. In C. J. Forceville & E. Urios-Aparisi (Eds.), Multimodal Metaphor (pp. 243-264). Berlin & New York: Mouton de Gruyter. Elman, J. L. (1995). Language as a dynamical system. In R. F. Port & T. Van Gelder (Eds.), Mind as Motion (pp. 195-223). Cambridge, MA: MIT Press. Ernst, M. O., & Banks, M. S. (2002). Humans integrate visual and haptic information in a statistically optimal fashion. Nature, 415(6870), 429-433. Ernst, M. O., & Bulthoff, H. H. (2004). Merging the senses into a robust percept. Trends in cognitive sciences, 8, 162-9. Evans, V. (2004). The structure of time. Amsterdam & Philadelphia: John Benjamins. Evans, V. (2010). Figurative language understanding in LCCM Theory. Cognitive Linguistics, 21(4), 601662.

363


Evans, V., & Green, M. C. (2006). Cognitive linguistics: An introduction. Edinburg: Edinburgh university press. Evans, V., & Zinken, J. (2011). Figurative language in a modern theory of meaning construction: a lexical concepts and cognitive models approach. Cambridge: Cambridge Scholars Press. Fabbrichesi Leo, R. (2004). Cosa significa dirsi pragmatisti. Milano: Clueb. Fabbrichesi Leo, R. (2009). Nietzsche and James: A Pragmatist hermeneutics. Europe and American Pragmatism: influences and interactions, 1, 25-35. Fabbrichesi Leo, R., & Leoni, F. (2005). Continuità e variazione: Leibniz, Goethe, Peirce, Wittgenstein, con un'incursione kantiana. Milano: Mimesis. Fahlenbrach, K. (2005). The Emotional Design of Music Videos. Approaches to Audiovisual Metaphors. Journal of Moving Images Studies, 3(1), 22-28. Fahlenbrach, K. (2007). Embodied Spaces: Film Spaces as (Leading) Audiovisual Metaphors. In J. D. Anderso & B. Fisher-Anderson (Eds.), Narration and spectatorship in moving images (p. 105). Cambridge, MA: Cambridge Scholars Press. Fauconnier, G. (1984). Espaces mentaux. Paris: Editions de Minuit. Fauconnier, G. (1994). Mental spaces: Aspects of meaning construction in natural language. Cambridge & New York: Cambridge University Press. Fauconnier, G. (1997). Mappings in thought and language. Cambridge & New York: Cambridge University Press. Fauconnier, G., & Turner, M. (1996). Blending as a central process of grammar. In A. Goldberg (Ed.), Conceptual structure, discourse and language (pp. 113-130). Stanford: CSLI Publications. Fauconnier, G., & Turner, M. (1998). Conceptual integration networks. Cognitive Science: A Multidisciplinary Journal, 22(2), 133-187.

364


Fauconnier, G., & Turner, M. (2001). Compression and global insight. Cognitive Linguistics, 11(3-4), 283-304. Fauconnier, G., & Turner, M. (2002). The way we think: Conceptual blending and the mind's hidden complexities. New York: Basic Books. Fauconnier, G., & Turner, M. (2008a). Rethinking metaphor. In R. W. Gibbs, Jr (Ed.) The Cambridge Handbook of Metaphor and Thought. Cambridge & New York: Cambridge University Press. Fauconnier, G., & Turner, M. (2008b). The origin of language as a product of the evolution of modern cognition. In B. Laks (Ed.), Origin and evolution of languages. London: Equinoz. Favareau, D. (2007). The evolutionary history of biosemiotics. In Introduction to biosemiotics (pp. 1-68). New York: Springer. Feldman, J. (2006). From molecule to metaphor: A neural theory of language. Cambridge, MA: MIT Press. Fernandez-Duque, D., & Johnson, M. (2002). Cause and effect theories of attention: The role of conceptual metaphors. Review of general psychology, 6(2), 153-165. Fernando, C. (1996). Idioms and idiomaticity. New York: Oxford University Press. Ferri, G. (2010). Giocare Con(tro) l'automa. Ph.D Dissertation, Università di Bologna. Ferri, G., & Fusaroli, R. (2009). Which Narrations for Persuasive Technologies? Habits and Procedures in Ayiti: The Cost of Life. Proceedings of AAAI 2009. Fesmire, S. (2003). John Dewey and moral imagination: Pragmatism in ethics. Bloomington: Indiana University Press. Fillmore, C. J. (1967). The Case for Case. In Bach and Harms (Ed.), Universals in Linguistic Theory. (pp. 188). New York: Holt, Rinehart & Winston. Fillmore, C. J. (1976). Frame Semantics and the nature of language. Annals of the New York Academy of Sciences, 280, 20-32.

365


Fisher, C., Gleitman, H., & Gleitman, L. R. (1991). On the semantic content of subcategorization frames. Cognitive Psychology, 23(3), 331-392. Fodor, J. (1983). The modularity of mind. Cambridge, MA: MIT press. Fodor, J. (1987). Psychosemantics : the problem of meaning in the philosophy of mind. Cambridge, MA: MIT Press. Fodor, J. (2009). Where is my mind. London Review of Books, 31(3), 13-15. Fontanille, J., & Zilberberg, C. (1998). Tension et signification. liege: Mardaga. Fontecha, A. F., & {Jiménez Catalán}, R. M. (2003). Semantic derogation in animal metaphor: a contrastive-cognitive analysis of two male/female examples in English and Spanish. Journal of pragmatics, 35(5), 771-797. Forceville, C. (1994). Pictorial metaphor in advertisements. Metaphor and Symbol, 9(1), 1-29. Forceville, C. (1998). Pictorial metaphor in advertising. London: Routledge. Forceville, C. (2005). Visual representations of the idealized cognitive model of anger in the Asterix album La Zizanie. Journal of pragmatics, 37(1), 69-88. Forceville, C. (2006). The SOURCE-PATH-GOAL schema in the autobiographical journey documentary. New Review of Film and Television Studies, 4(3), 241-261. Frank, R. M., Dirven, R., Ziemke, T., & Bernárdez, E. (2008). Body, Language and Mind, vol. 2, Sociocultural Situatedness. Berlin and New York: Mouton de Gruyter. Fraser, B. (1981). Insulting problems in a second language. TESOL Quarterly, 15, 435-441. Frith, C. D. (2007). Making up the mind: How the brain creates our mental world. Oxford: Wiley-Blackwell. Froese, T., & Di Paolo, E. (2008). Stability of coordination requires mutuality of interaction in a model of embodied agents. From Animals to Animats 10, 52-61. Froese, T., & {Di Paolo}, E. (2009). Sociality and the life–mind continuity thesis. Phenomenology and the Cognitive Sciences, 8(4), 439-463.

366


Funk, C. E. (1985). A hog on ice and other curious expressions. New York: Harper & Row. Fusaroli, R. (2009). Metaphor, Unnatural. In Dictionary of Unnatural Narratology. Aarhus Universitet. Fusaroli, R., & Ferri, G. (2009). L'arco e i trecento. Strategie semantiche di comunicazione politica nel Partito Democratico. Versus, 108. Fusaroli, R., Granelli, T., & Paolucci, C. (2010). The External Mind. Versus 113-115. Milano: Bompiani. Fusaroli, R., & Morgagni, S. (2009). Enacting Icons. ArCo 2009. Fusaroli, R., & Tylén, K. (2011). Language as social appraisal. Language Sciences. Galantucci, B. (2005). An experimental study of the emergence of human communication systems. Cognitive Science: A Multidisciplinary Journal, 29(5), 737-767. Galantucci, B. (2009). Experimental semiotics: A new approach for studying communication as a form of joint action. Topics in Cognitive Science, 1(2), 393-410. Galantucci, B., & Garrod, S. (2010). A new approach for studying the emergence and the evolution of human communication. Interaction studies, 11(1), 1-13. Galantucci, B., & Steels, L. (2008). The emergence of embodied communication in artificial agents and humans. In I. Wachsmuth (Ed.) Embodied communication in humans and machines (pp. 229256). Oxford: Oxford University Press. Gale, W. A., & Fox, J. (1986). Knowledge, Decision Making and Uncertainty. New York: AddisonWesley. Gallagher, S. (2004). How the body shapes the mind. New York: Oxford University Press. Gallagher, S. (2007). Phenomenological and experimental contributions to understanding embodied experience. In T. Ziemke, J. Zlatev, R. M. Frank (Eds.), Body, Language and Mind: Embodiment (Vol. 1). Berlin & New York: Mouton de Gruyter. Gallagher, S. (2010). The overextended mind. Versus, 113-115.

367


Gallagher, S., & Crisafi, A. (2009). Mental institutions. Topoi, 28(1), 45-51. Gallagher, S., & Hutto, D. D. (2008). Understanding others through primary interaction and narrative practice. The shared Mind: Perspectives on intersubjectivity, 17-38. Gallagher, S., & Zahavi, D. (2008). The phenomenological mind. London: Routledge. Gallese, V. (2007). Before and below ‘theory of mind’: embodied simulation and the neural correlates of social cognition. Philosophical Transactions B, 362(1480), 659. Gallese, V., & Goldman, A. (1998). Mirror neurons and the simulation theory of mind-reading. Trends in cognitive sciences, 2(12), 493-501. Gangemi, A. (2011). Mutazioni politiche. Abiti, invenzioni e stili nella leadership democratica. Ph.D Dissertation, Università di Bologna. Garrod, S., & Anderson, A. (1987). Saying what you mean in dialogue: A study in conceptual and semantic co-ordination. Cognition, 27(2), 181-218. Garrod, S., & Pickering, M. J. (2009). Joint action, interactive alignment, and dialog. Topics in Cognitive Science, 1(2), 292-304. Geeraerts, D. (2006). Cognitive linguistics: Basic readings. Berlin & New York: Mouton de Gruyter. Geeraerts, D., & Cuyckens, H. (2007). The Oxford handbook of cognitive linguistics. New York: Oxford University Press. Gentner, D., & Bowdle, B. (2008). Metaphor as structure-mapping. In R. W. Gibbs, Jr (Ed.) The Cambridge Handbook of Metaphor and Thought (pp. 109-129). Cambridge & New York: Cambridge University Press. Gergely, G., & Watson, J. S. (1999). Early socio-emotional development: Contingency perception and the social-biofeedback model. In P. Rochat (Ed.), Early social cognition: Understanding others in the first months of life (pp. 101-136). London: Lawrence Erlbaum Associates. Geurts, K. L. (2003). Culture and the senses. Berkeley: University of California Press.

368


Ghiazza, S. (2005). La metafora tra scienza e letteratura. Firenze: Le Monnier Università. Gibbs, R. W. J. (1994). The poetics of mind: Figurative thought, language, and understanding. Cambridge & New York: Cambridge University Press. Gibbs, R. W. J. (1999). Taking Metaphor Out of our Heads and Putting it into the Cultural World. In R. W. Gibbs, Jr, & G. J. Steen (Eds.), Metaphor in cognitive linguistics (p. 145). Amsterdam & Philadelphia: John Benjamins. Gibbs, R. W. J. (2000). Making good psychology out of blending theory. Cognitive Linguistics, 11(3/4), 347-358. Gibbs, R. W. J. (2003). Prototypes in dynamic meaning construal. In J. Gavins & G. Steen (Eds.), Cognitive Poetics in practice (pp. 27-40). London: Routledge. Gibbs, R. W. J. (2006). Embodiment and cognitive science. Cambridge & New York: Cambridge University Press. Gibbs, R. W. J., & Cameron, L. (2008). The social-cognitive dynamics of metaphor performance. Cognitive Systems Research, 9(1-2), 64-75. Gibbs, R. W. J., & Colston, H. L. (1995). The cognitive psychological reality of image schemas and their transformations. Cognitive Linguistics, 6(4), 347-378. Gibbs, R. W. J., & Tendahl, M. (2006). Cognitive effort and effects in metaphor comprehension: Relevance theory and psycholinguistics. Mind & Language, 21(3), 379-403. Gibson, E., & Fedorenko, E. (2010). Weak quantitative standards in linguistics research. Trends in cognitive sciences. Gibson, J. J. (1979). The ecological approach to visual perception. Boston: Houghton Mifflin. Giora, R. (2003). On our mind: salience, context, and figurative language. Oxford: Oxford University Press. Glenberg, A. M. (1997). What memory is for: Creating meaning in the service of action. Behavioral and brain sciences, 20(01), 41-50.

369


Glenberg, A. M., & Robertson, D. A. (2000). Symbol grounding and meaning: A comparison of highdimensional and embodied theories of meaning. Journal of Memory and Language, 43(3), 379401. Goatly, A. (1997). The language of metaphors. London: Routledge. Goldin-Meadow, S. (2005). Hearing gesture: How our hands help us think. Cambridge, MA: Harvard University Press. Goldman, A. (2009). Mirroring, mindreading, and simulation. In J. A. Pineda (Ed.), Mirror neuron systems: The role of mirroring processes in social cognition (pp. 1-20). New York: Human Press. Gonzalez-Marquez, M. (2007). Methods in cognitive linguistics. Amsterdam & Philadelphia: John Benjamins. Goodwin, C. (1994). Professional vision. American anthropologist, 96(3), 606-633. Gordon, R. M. (1986). Folk psychology as simulation. Mind & Language, 1(2), 158-171. Gould, S. J. (1997). The exaptive excellence of spandrels as a term and prototype. Proceedings of the National Academy of Sciences of the United States of America, 94(20), 10750-10775. Goulden, N. R., & Griffin, C. J. G. (1995). The meaning of grades based on faculty and student metaphors. Communication Education, 44(2), 110-125. Grady, J. (1997). Foundations of meaning: Primary metaphors and primary scenes. Grady, J. (1998). The Conduit Metaphor revisited: a reassessment of metaphors for communication. In J.-P. Koenig (Ed.), Discourse and cognition: Bridging the gap (pp. 205-218). Stanford: CSLI Publications. Grady, J. (1999). A typology of motivation for conceptual metaphor: Correlation vs. resemblance. In R. W. Gibbs, Jr, & G. J. Steen (Eds.), Metaphor in cognitive linguistics. Amsterdam & Philadelphia: John Benjamins.

370


Grady, J. (2007). Metaphor. In D. Geeraerts & H. Cuyckens (Eds.), The Oxford Handbook of Cognitive Linguistics. New York: Oxford University Press. Grady, J., Oakley, T., & Coulson, S. (1999). Conceptual blending and metaphor. In R. W. Gibbs, Jr, & G. J. Steen (Eds.), Metaphor in cognitive linguistics (pp. 101-124). Amsterdam & Philadelphia: John Benjamins. Greimas, A. J. (1966). Sémantique structurale: recherche de méthode. Paris: Larousse. Greimas, A. J. (1970). Du sens. Paris: Seuil. Grice, P. (1989). Studies in the Way of Words. Cambridge, MA: Harvard University Press. Gries, S. T. (2003). Towards a corpus-based identification of prototypical instances of constructions. Annual review of cognitive linguistics, 1(1), 1-27. Gries, S. T., & Stefanowitsch, A. (2007). Corpora in cognitive linguistics: corpus-based approaches to syntax and lexis. Berlin & New York: Mouton de Gruyter. Grofman, B., Owen, G., & Hastie, R. (1983). Experimental evidence on group accuracy. In B. Grofman & G. Owen (Eds.), Information Pooling and Group Decision Making (pp. 129-157). Greenwich, CT: JAI Press. Grondelaers, S., Geeraerts, D., & Speelman, D. (2007). A case for a cognitive corpus linguistics. In M. Gonzales-Marques, I. Mittelberg, S. Coulson & M. Spivey (Eds.), Methods in Cognitive Linguistics (pp. 149-169). Amsterdam & Philadelphia: John Benjamins. Groupe µ. (1970). Rhétorique générale. Paris: Larousse. Grush, R. (1997). Yet another design for a brain. Review of Port and van Gelder (Eds.), Mind as motion. Philosophical Psychology, 10, 233-242. Halligan, P. W., Fink, G. R., Marshall, J. C., & Vallar, G. (2003). Spatial cognition: evidence from visual neglect. Trends in cognitive sciences, 7(3), 125-133.

371


Hamm, R. M. (1991). Selection of verbal probabilities: a solution for some problems of verbal probability expression. Organizational Behavior and Human Decision Process, 48, 193-223. Hanks, P. (2004). The syntagmatics of metaphor and idiom. International Journal of Lexicography, 17(3), 245. Hanks, P. (2006). Metaphoricity is gradable. Trends in linguistics studies and monographs, 171, 17. Harder, P. (2003). Mental spaces: Exactly when do we need them? Cognitive Linguistics, 14(1), 91-96. Harder, P. (2007). Cognitive linguistics and philosophy. In D. Geeraerts & H. Cuyckens (Eds.), The Oxford Handbook of Cognitive Linguistics (pp. 1241-1265). New York: Oxford University Press. Harnad, S. (1990). The symbol grounding problem. Physica D: Nonlinear Phenomena, 42(1-3), 335-346. Hartsuiker, R. J., Pickering, M. J., & Veltkamp, E. (2004). Is syntax separate or shared between languages? Psychological Science, 15(6), 409. Haser, V. (2005). Metaphor, metonymy, and experientialist philosophy: challenging cognitive semantics. Berlin: Walter de Gruyter. Hastie, R., & Kameda, T. (2005). The robust beauty of majority rules in group decisions. Psychological Review, 112(2), 494-508. Hausman, C. R. (1993). Charles S. Peirce's evolutionary philosophy. Cambridge & New York: Cambridge University Press. Havenel, J. (2008). Peirce's Clarifications of Continuity. Transactions of the Charles S. Peirce Society: A Quarterly Journal in American Philosophy, 44(1), 86-133. Hayden, S. (2003). Family metaphors and the nation: Promoting a politics of care through the Million Mom March. Quarterly Journal of Speech, 89(3), 196-215. Heider, F., & Simmel, M. (1944). An experimental study of apparent behavior. The American Journal of Psychology, 57(2), 243-259. Hill, G. W. (1982a). Are n+ 1 heads better than 1? Psychological Bulletin, 91, 517-539.

372


Hill, G. W. (1982b). Group versus individual performance. Psychological Bulletin, 91, 517-539. Hill, R. C., & Levenhagen, M. (1995). Metaphors and mental models: Sensemaking and sensegiving in innovative and entrepreneurial activities. Journal of Management, 21(6), 1057. Hilpert, M. (2006). Distinctive collexeme analysis and diachrony. Corpus Linguistics and Linguistic Theory, 2(2), 243-57. Hoffmann, M. H. G. (2008). Cognitive conditions of diagrammatic reasoning. Semiotica, 172. Holland, D. (1982). All is metaphor: Conventional metaphors in human thought and language. Reviews in Anthropology, 9(3), 287-297. Holmes, J. (2008). An introduction to sociolinguistics. Reading, MA: Addison-Wesley. Horton, W. S., & Keysar, B. (1996). When do speakers take into account common ground? Cognition, 59(1), 91-117. Hougaard, A. (2005). Conceptual disintegration and blending in interactional sequences: A discussion of new phenomena, processes vs. products, and methodology. Journal of pragmatics, 37(10), 1653-1685. Hougaard, A. (2011). Metaphors and Conversation-Analysis. The Companion of Cognitive Semiotics, II. Hsieh, S. C. (2006). A corpus-based study on animal expressions in Mandarin Chinese and German. Journal of Pragmatics, 38(12), 2206-2222. Hurley, S. (1998). Consciousness in action. Cambridge, MA: MIT Press. Hurley, S. (2001). Perception and action: Alternative views. Synthese, 129(1), 3-40. Hurley, S. (2008). The shared circuits model (SCM): How control, mirroring, and simulation can enable imitation, deliberation, and mindreading. Behavioral and Brain Sciences, 31(01), 1-22. Hutchins, E. (1995). Cognition in the wild. Cambridge, MA: MIT Press. Hutchins, E. (2005). Material anchors for conceptual blends. Journal of pragmatics, 37(10), 1555-1577.

373


Hutchins, E. (2008). The role of cultural practices in the emergence of modern human intelligence. Philosophical Transactions B, 363(1499), 2011. Itkonen, E. (2008). The central role of normativity for language and linguistics. In J. Zlatev, T. Racine, C. Sinha & E. Itkonen (Eds.) The Shared Mind (pp. 279-306). Amsterdam & Philadelphia: John Benjamins. Jackendoff, R. S. (1983). Semantics and Cognition. Cambridge, MA: MIT Press. Jacob, P. (2010). Intentionality. The Stanford Encyclopedia of Philosophy. Jäkel, O. (1999). Kant, Blumenberg, Weinrich: Some forgotten contributions to the cognitive theory of metaphor. In R. W. Gibbs, Jr, & G. J. Steen (Eds.), Metaphor in cognitive linguistics (p. 9). Amsterdam & Philadelphia: John Benjamins. Jakobson, R. (1956). Two aspects of language and two types of linguistic disturbances. In Jakobson and Halle, Fundamentals of Language (pp. 69-96). Le Hague: Mouton. Johnson, C. (1999). Metaphor vs. conflation in the acquisition of polysemy: The case of SEE. Berkeley University. Johnson, M. (1987). The body in the mind: The bodily basis of meaning, imagination, and reason. Chicago & London: University of Chicago Press. Johnson, M. (1994). Moral imagination: Implications of cognitive science for ethics. Chicago & London: University of Chicago Press. Johnson, M. (2008a). Philosophy’s debt to metaphor. In R. W. Gibbs, Jr (Ed.) The Cambridge Handbook of Metaphor and Thought (pp. 39-53). Cambridge & New York: Cambridge University Press. Johnson, M. (2008b). The meaning of the body: Aesthetics of human understanding. Chicago & London: University of Chicago Press. Johnson, M., & Lakoff, G. (2002). Why cognitive linguistics requires embodied realism. Cognitive Linguistics, 13(3), 245-263.

374


Johnson, M., & Larson, S. (2003). " Something in the Way She Moves"-Metaphors of Musical Motion. Metaphor and symbol, 18(2), 63-84. Kecskes, I. (2004). Editorial: Lexical merging, conceptual blending, and cultural crossing. Intercultural pragmatics, 1(1), 1-26. Keijzer, F. A. (2001). Representation and behavior. Cambridge, MA: MIT Press. Kelso, J. A. S. (1995). Dynamic patterns: The self-organization of brain and behavior. Cambridge, MA: MIT Press. Ketner, K. L. (1998). His glassy essence: an autobiography of Charles Sanders Peirce. Nashville: Vanderbilt University Press. Kimmel, M. (2002). Metaphor, Imagery, and Culture. Spatialized Ontologies, Mental Tools, and Multimedia in the Making. Kimmel, M. (2005). Culture regained: Situated and compound image schemas. In B. Hampe (Ed.) From Perception to Meaning: Image Schemas in Cognitive Linguistics (pp. 285-312). Berlin & New York: Mouton de Gruyter. Kimmel, M. (2010). Intersubjectity at close quarters. The Companion of Cognitive Semiotics, I. Kintsch, W., & Mangalath, P. (2010). The construction of meaning. Topics in Cognitive Science. Kirmayer, L. J. (1992). The body's insistence on meaning: metaphor as presentation and representation in illness experience. Medical Anthropology Quarterly, 6(4), 323-346. Kirsh, D. (2010). How marking in dance constitutes thinking with the body. Versus, 113-115. Kitis, E., & Milapides, M. (1997). Read it and believe it: How metaphor constructs ideology in news discourse. A case study. Journal of Pragmatics, 28(5), 557-590. Koch, S., & Deetz, S. (1981). Metaphor analysis of social reality in organizations. Journal of Applied Communication Research, 9(1), 1-15.

375


Koivisto-Alanko, P. (2000). Abstract words in abstract worlds: directionality and prototypical structure in the semantic change in English nouns of cognition. Helsinki: Société néophilologique. Koivisto-Alanko, P., & Tissari, H. (2006). Sense and sensibility: Rational thought versus emotion in metaphorical language. In A. Stefanowitsch & S.T. Gries (eds.) Corpus-based approaches to metaphor and metonymy. Berlin & New York: Walter de Gruyter. Koller, V. (2002). " A Shotgun Wedding": Co-occurrence of War and Marriage Metaphors in Mergers and Acquisitions Discourse. Metaphor and Symbol, 17(3), 179-203. Koller, V. (2006). Of critical importance: Using corpora to study metaphor in business media discourse. Berlin: Walter de Gruyter. Körding, K. P., & Wolpert, D. M. (2004). Bayesian integration in sensorimotor learning. Nature, 427(6971), 244-247. Kövecses, Z. (2005). Metaphor in culture: Universality and variation. Cambridge & New York: Cambridge University Press. Kövecses, Z. (2008). Conceptual metaphor theory Some criticisms and alternative proposals. Annual Review of Cognitive Linguistics, 6(1), 168-184. Kövecses, Z. (2010). Metaphor: A practical introduction. New York: Oxford University Press. Krauss, R. M., & Weinheimer, S. (1964). Changes in reference phrases as a function of frequency of usage in social interaction: A preliminary study. Psychonomic Science, 1(113-114). Krausser, P. (1977). The Three Fundamental Structural Categories of Charles S. Peirce. Transactions of the Charles S. Peirce Society, 189-215. Kruschke, J. K. (2010). What to believe: Bayesian methods for data analysis. Trends in cognitive sciences. Kutas, M., & Federmeier, K. D. (2009). The N400. Annual Review of Psychology, 62(1). Kutas, M., & Hillyard, S. A. (1980). Reading senseless sentences: Brain potentials reflect semantic incongruity. Science, 207(4427), 203-205.

376


Kuusisto, R. (2002). Heroic tale, game, and business deal? Western metaphors in action in Kosovo. Quarterly Journal of Speech, 88(1), 50-68. Lakoff, G. (1987). Women, fire, and dangerous things: What categories reveal about the mind. Chicago & London: University of Chicago Press. Lakoff, G. (1993). The contemporary theory of metaphor. In A. Ortony (Ed.) Metaphor and Thought (2nd ed., pp. 202-251). Cambridge & New York: Cambridge University Press. Lakoff, G. (1996). Moral politics. Chicago & London: University of Chicago Press. Lakoff, G. (2004). Don't think of an elephant!: know your values and frame the debate: the essential guide for progressives. White River Junction, Vt: Chelsea Green. Lakoff, G. (2006). Whose freedom. The Battle over America’s Most Important Idea. New York: Farrar Straus & Giroux. Lakoff, G. (2008). The neural theory of metaphor. In R. W. Gibbs, Jr (Ed.) The Cambridge Handbook of Metaphor and Thought. Cambridge & New York: Cambridge University Press. Lakoff, G., & Fauconnier, G. (2011). On Metaphor and Blending. The Companion of Cognitive Semiotics, II. Lakoff, G., & Johnson, M. (1980). Metaphors we live by. Chicago & London: University of Chicago Press. Lakoff, G., & Johnson, M. (1999). Philosophy in the flesh: The embodied mind and its challenge to western thought. New York: Basic Books. Lakoff, G., & Nunez, R. (2000). Where mathematics comes from: How the embodied mind brings mathematics into being. New York: Basic Books. Lakoff, G., & Thompson, H. (1975). Introducing cognitive grammar. BLS, 1, 295-313. Lakoff, G., & Tumer, M. (1989). No cool reason: The power of poetic metaphor. Chicago & London: University of Chicago Press.

377


Landau, B., & Gleitman, L. R. (1985). Language and experience: Evidence from the blind child. Cambridge, MA: Harvard University Press. Langacker, R. W. (1986). Abstract motion. Proceedings of the annual meeting of the Berkeley Linguistics Society, 12, 455-471. Langacker, R. W. (1987). Foundations of cognitive grammar: Theoretical Prerequisites. Stanford: Stanford University Press. Lascaratou, C. (2007). The language of pain : expression or description? Amsterdam & Philadelphia: John Benjamins. Latane, B., Williams, K., & Harkins, S. (1979). Many hands make light the work. Journal of Personality and Social Psychology, 37, 822-832. Latour, B. (1987). Science in action: how to follow scientists and engineers through society. Milton Keynes: Open University Press. Latour, B. (1990). Visualisation and Cognition: Drawing things together. In M. Lynch & S. Woolgar (Eds.), Representation in Scientific Practice. Cambridge, MA: MIT Press. Latour, B. (2005). Reassembling the social: an introduction to actor-network-theory. Oxford: Oxford University Press. Laughlin, P. R., & Ellis, A. L. (1986). Demonstrability and social combination processes on mathematical intellective tasks. Journal of Experimental Social Psychology, 22(3), 177-189. Laughlin, P. R., & Futoran, G. C. (1985). Collective Induction:: Social Combination and Sequential Transition. Journal of Personality and Social Psychology, 48(3), 608-613. Laughlin, P. R., & Shippy, T. A. (1983). Collective Induction. Journal of Personality and Social Psychology, 45, 139-148. Leezenberg, M. (2001). Contexts of metaphor. Amsterdam & New York: Elsevier. Leezenberg, M. (2011). Cognitive Linguistics and Social Science. The Companion of Cognitive Semiotics, 2.

378


de León, C. M. (2004). Metonymic Motivation of the CONDUIT Metaphor. Metaphorik.de, 6, 79-90. Levinson, S. C. (1997). Language and cognition: The cognitive consequences of spatial description in Guugu Yimithirr. Journal of Linguistic Anthropology, 7(1), 98-131. Liddell, S. K. (2003). Grammar, gesture, and meaning in American Sign Language. Cambridge & New York: Cambridge University Press. Lindstromberg, S. (1997). English prepositions explained. Amsterdam & Philadelphia: John Benjamins. Lindzey, G., Aronson, E., Kelly, H. H., & Thibaut, J. W. (1969). Group Problem Solving (pp. 1-101). Reading, MA: Addison-Wesley. Locke, J. (1698). An essay concerning human understanding. Troutman & Hayes. Logan, R. K. (2006). The extended mind model of the origin of language and culture. In N. Gontier, J. van Bendegem & D. Aerts (Eds.), Evolutionary epistemology, language and culture (pp. 149-167). Dordrecht: Springer. Loreto, V., & Steels, L. (2007). Social dynamics: Emergence of language. Nature Physics, 3(11), 758-760. Lotman, J. M. (1993). La cultura e l'esplosione. Milano: Feltrinelli. Louwerse, M. M. (2010). Symbol interdependency in symbolic and embodied cognition. Topics in Cognitive Science. Love, N. (2007). Are languages digital codes? Language Sciences, 29(5), 690-709. Low, G. D. (1988). On teaching metaphor. Applied Linguistics, 9(2), 125. Lucy, J. A. (1996). Grammatical categories and cognition: A case study of the linguistic relativity hypothesis. Cambridge & New York: Cambridge University Press. Lukes, D. (2008). Review of Verena Haser, Metaphor, Metonymy and Experientialist Philosophy: Challenging Cognitive Semantics. Cognitive Linguistics, 19, 313-324. Lule, J. (2004). War and its metaphors: news language and the prelude to war in Iraq, 2003. Journalism Studies, 5(2), 179-190.

379


Lyman, D. (1983). The Animal Things We Say. New York: Jonathan David Publishers. Magnani, L. (2009). Abductive cognition : the epistemological and eco-cognitive dimensions of hypothetical reasoning. New York: Springer. Magnani, L., & Nersessian, N. J. (2002). Model-based reasoning. Dordrecht: Kluwer Academic Publishers. Mameli, M. (1997). Synechism: aspetti del pensiero di C.S. Peirce. Ph.D Dissertation, Università di Bologna. Mandler, J. M. (1992). How to build a baby: II. Conceptual primitives. Psychological Review, 99, 587. Mandler, J. M. (1996). Preverbal Representation and Language. In P. Bloom, M. Peterson, L. Nadel, & M. Garrett (Eds.), Language and Space (pp. 365-384). Cambridge, MA: MIT Press. Mandler, J. M. (2004). The foundations of mind: Origins of conceptual thought. New York: Oxford University Press. Marconi, D. (2001). Filosofia e scienza cognitiva. Bari: Laterza. Markert, K., & Nissim, M. (2002). Towards a corpus annotated for metonymies: the case of location names. In Proceedings of the 3rd international Conference on Language Resources and Evaluation (pp. 1385-1392). Markert, K., & Nissim, M. (2006). Metonymic proper names: A corpus-based account. Trends in linguistics studies and monographs, 171, 152. Markman, A. B., & Makin, V. S. (1998). Referential communication and category acquisition. Journal of Experimental Psychology, 127, 331-354. Martin, J. H. (2006). A corpus-based analysis of context effects on metaphor comprehension. Trends in linguistics studies and monographs, 171, 214. Martsa, S. (2003). Conceptual mappings in the ethnobiological categorization of animals. Cambridge, MA: MIT Press.

380


Maturana, H. R., & Varela, F. J. (1980). Autopoiesis and cognition: The realization of the living. New York: Springer. Matusov, E. (1996). Intersubjectivity without agreement. Mind, Culture, and Activity, 3(1), 25-45. Mauss, M. (1973). Techniques of the Body. Economy and Society, 2(1), 70-88. May, M. (1999). Diagrammatic reasoning and levels of schematization. In T. D. Johansson, M. Skov & B. B. Pedersen (Eds.) Iconicity - A fundamental problem in Semiotics (pp. 175-193). Århus: NSU Press. Mayaffre, D. (2002). Les corpus réflexifs: entre architextualité et hypertextualité. Corpus et recherches linguistiques, (1). McCauley, R. N., & Henrich, J. (2006). Susceptibility to the Müller-Lyer illusion, theory-neutral observation, and the diachronic penetrability of the visual input system. Philosophical Psychology, 19(1), 79-101. McCulloch, W. S., & Pitts, W. (1943). A logical calculus of the ideas immanent in nervous activity. Bulletin of mathematical biology, 5(4), 115-133. McDowell, J. (1994). The content of perceptual experience. The Philosophical Quarterly, 44(175), 190205. McNeill, D. (2005). Gesture and thought. Chicago & London: University of Chicago Press. Medhurst, M. J. (1998). The rhetorical renaissance: A battlefield report. Southern Communication Journal, 63(4), 309-314. Melcher, D., & Colby, C. L. (2008). Trans-saccadic perception. Trends in cognitive sciences, 12(12), 466473. Menary, R. (2006). Attacking the bounds of cognition. Philosophical Psychology, 19, 329-344. Merleau-Ponty, M. (1962). Phenomenology of Perception. London: Routledge.

381


Miller, G. A. (2003). The cognitive revolution: a historical perspective. Trends in Cognitive Sciences, 7, 141-144. Miller, G. A., & Johnson-Laird, P. N. (1976). Language and perception. Cambridge & New York: Cambridge University Press. Millikan, R. G. (1991). Perceptual content and Fregean myth. Mind, 100(400), 439-459. Millikan, R. G. (1993). White Queen Psychology and Other Essays. Cambridge, MA: MIT Press. Mittelberg, I. (2006). Making Grammar Visible: The Use of Metaphoric Gestures to Represent Grammatical Categories and Structures. Moore, C., Harris, L., & Patriquin, M. (1993). Lexical and prosodic cues in the comprehension of relative certainty. Journal of Child Language, 20, 153-67. Morgagni, S. (2010). Repenser la notion d’affordance dans ses dynamiques sémiotiques. Intellectica. Mouton, N. (2010). On the evolution of social scientific metaphors. Copenhagen. Murphy, G. L. (1996). On metaphoric representation. Cognition, 60(2), 173-204. Murray, L., & Trevarthen, C. (1985). Emotional regulation of interactions between two-month-olds and their mothers. Social perception in infants, 177-197. Musolff, A. (2006). Metaphor scenarios in public discourse. Metaphor and symbol, 21(1), 23-38. Nadel, J., Carchon, I., Kervella, C., & Marcelli, D. (1999). Expectancies for social contingency in 2 month olds. Developmental Science, 2(2), 164-173. Nesi, H. (1995). A modern bestiary: a contrastive study of the figurative meanings of animal terms. ELT Journal, 49(3), 272. Newmark, P. (1988). Approaches to translation. Hemel Hempstead: Prentice Hall. Nietzsche, F. (1979). On truth and lies in a nonmoral sense. In Philosophy and Truth: Selections from Nietzsche's Notebooks of the Early 1870's (pp. 79-93). Atlantic Highlands (NJ): Humanities Press.

382


Niveleau, C. (2006). Le concept gibsonien d’affordance: entre filiation, rupture et reconstruction conceptuelle. Intellectica, 1(43), 159-199. Noë, A. (2004). Action in perception. Cambridge, MA: MIT Press. Norrick, N. R. (1986). Stock similes. Journal of Literary Semantics, 15(1), 39-52. Nöth, W. (2010). The Criterion of Habit in Peirce's Definitions of the Symbol. Transactions of the Charles S. Peirce Society: A Quarterly Journal in American Philosophy, 46(1), 82-93. Nunez, R. (1999). Could the future taste purple? Reclaiming mind, body and cognition. Journal of Consciousness Studies, 6, 11(12), 41-60. O'Donnell, P. E. (1990). Entre chien et loup: a study of French animal metaphors. French Review, 63(3), 514-523. O'Regan, J. K. (2010). Explaining what people say about sensory qualia. In N. Gangopadhyay, M. Madary, F. Spicer (Eds.) Perception, Action, and Consciousness: Sensorimotor Dynamics and Two Visual Systems. Oxford: Oxford University Press. O'Regan, J. K., & Noë, A. (2001). A sensorimotor account of vision and visual consciousness. Behavioral and brain sciences, 24(05), 939-973. O'Regan, J. K., Rensink, R. A., & Clark, J. J. (1999). Change-blindness as a result of ‘mudsplashes’. Nature, 398(6722), 34. Oakley, T., & Coulson, S. (2008). Connecting the dots: Mental spaces and metaphoric language in discourse. Pragmatics & beyond. New series, 170, 27-50. Olsen, K. (2010). Social interaction enhances the rate of individual visual perceptual learning. Presented at the British Association for Cognitive Neuroscience. Ortony, A. (1993). Metaphor and thought (2nd ed.). Cambridge & New York: Cambridge University Press. Oudeyer, P. Y. (2006). Self-organization in the evolution of speech. New York: Oxford University Press.

383


Palmatier, R. A. (1995). Speaking of animals: a dictionary of animal metaphors. Westport: Greenwood Pub Group. Paolo, F. (1998). La svolta semiotica. Roma-Bari: Laterza. Paolucci, C. (2004). Piegature della continuità: semiotica interpretativa e semiotica generativa. Versus, 97, 111-149. Paolucci, C. (2005). Anti-Logos. Bologna. Paolucci, C. (2006). Semantica referenziale, inferenziale, differenziale: per una semantica unificata. In A. Frigerio & S. Raynaud (Eds.), Significare e comprendere. Roma: Aracne. Paolucci, C. (2010). Strutturalismo e interpretazione. Milano: Bompiani. Paprotté, W., & Dirven, R. (1985). The ubiquity of metaphor: metaphor in language and thought. Amsterdam & Philadelphia: John Benjamins. Partington, A. (2003). The linguistics of political argument: the spin-doctor and the wolf-pack at the White House. Hove, Uk: Psychology Press. Partington, A. (2006). The linguistics of laughter: A corpus-assisted study of laughter-talk. Taylor & Francis. Pascal, B. (1670). Pensées. Pascual, E. (2007). Imaginary trialogues: Conceptual blending and fictive interaction in criminal courts. International Journal of Speech Language and the Law, 11(1), 169. Pawelec, A. (2007). A note on the formalism of cognitive linguistics. Studia Linguistica Universitatis Iagellonicae Cracoviensis, 124. Peirce, C. S. (1931-1958). Collected Papers of Charles Sanders Peirce, voll.I–VI, Hartshorne C., Weiss P. (eds.) 1931-1935, voll.VII–VIII Burks A.W. (ed.) 1958, Cambridge (MA): Belknap Press (CP). Peirce, C. S. (1976). The New Elements of Mathematics by Charles S. Peirce, edited by Carolyn Eisele, voll. I –IV, The Hague: Mouton (NEM).

384


Peirce, C. S. (1977). Semiotic and significs: the correspondence between Charles S. Peirce and Victoria Lady Welby. Indiana University Press (LW). Peirce, C. S. (1982- ). Writings of Charles S. Peirce, A Chronological Edition, Peirce Edition Project (eds.), Indiana University Press, Bloomington and Indianapolis (W). Peirce, C. S., & James, W. (2000). Che cos' è il pragmatismo. Jaca Book, Milano. Peterson, D. K., & Pitz, G. F. (1988). Confidence, uncertainty, and the use of information. Journal of Experimental Psychology: Learning, Memory and Cognition, 14, 85-92. Petitot, J. (1995). Morphodynamics and attractor syntax: constituency in visual perception and cognitive grammar. In R. F. Port & T. Van Gelder (Eds.), Mind as Motion (pp. 227-83). Cambridge, MA: MIT Press. Pfeifer, R., & Scheier, C. (1999). Understanding Intelligence. Cambridge, MA: MIT Press. Phillips, B. J. (2003). Understanding visual metaphor in advertising. In L. M. Scott & R. Batra (Eds.), Persuasive imagery: A consumer response perspective (pp. 297-310). London: Lawrence Erlbaum Associates. Piccinini, G., & Scarantino, A. (2010). Computation vs. information processing: Why their difference matters to cognitive science. Studies In History and Philosophy of Science Part A. Pickering, M. J., & Garrod, S. (2004). Toward a mechanistic psychology of dialogue. Behavioral and Brain Sciences, 27(02), 169-190. Pietarinen, A. V. (2006). Signs of logic: Peircean themes on the philosophy of language, games, and communication. Dordrecht: Kluwer Academic Publishers. Port, R. F., & Van Gelder, T. (1995). Mind as motion: Explorations in the dynamics of cognition. Cambridge, MA: MIT Press. Premack, D., & Woodruff, G. (1978). Does the chimpanzee have a theory of mind? Behavioral and Brain sciences, 1(04), 515-526.

385


Propp, V. I. A. (1968). Morphology of the Folktale. Austin: University of Texas Press. Pulford, B. D. (2002). The influence of speech confidence on choice and decision confidence. Proceedings of the British Psychological Society, 10. Pulvermüller, F. (2005). Brain mechanisms linking language and action. Nature Reviews Neuroscience, 6(7), 576-582. Raczaszek-Leonardi, J., & {Scott Kelso}, J. A. (2008). Reconciling symbolic and dynamic aspects of language: Toward a dynamic psycholinguistics. New Ideas in Psychology, 26(2), 193-207. Raczaszek-Leonardi, J., Shapiro, L. P., Tuller, B., & Kelso, J. A. S. (2008). Activating Basic Category Exemplars in Sentence Contexts: A Dynamical Account. Journal of psycholinguistic research, 37(2), 87-113. Radden, G., Köpcke, K., Berg, T., & Siemund, P. (2007). Aspects of meaning construction. Amsterdam & Philadelphia: John Benjamins. Rakova, M. (2002). The philosophy of embodied realism: A high price to pay? Cognitive Linguistics, 13(3), 215-244. Rakova, M. (2003). The extent of the literal: metaphor, polysemy and theories of concepts. Basingstoke: PalgraveMacmillan. Rastier, F. (1987). Sémantique interprétative. Paris: PUF. Rastier, F. (1999). Cognitive Semantics and Diachrony. In A. Blank & P. Koch (Eds.), Historical Semantics and Cognition (pp. 109-144). Berlin & New York: Mouton de Gruyter. Rastier, F. (2001). Sémantique et recherches cognitives. Paris: PUF. Rastier, F. (2002). L'accès aux banques textuelles - des genres à la doxa. In T. Cabré & C. Gelpi (Eds.), Lèxic, corpus i diccionaris. Barcelona: Université Pompeu Fabra.

386


Rastier, F. (2004). Doxa et lexique en corpus-pour une sémantique des “idéologies”. In J. Pauchard & F. Canon-Roger (Eds.), Actes des Journées Scientifiques en linguistique (Vol. 22). Reims: Presses Universitaires de Reims. Rastier, F. (2005). Enjeux épistémologiques de la linguistique de corpus. In G. Williams (Ed.), La linguistique de corpus (pp. 31-45). Rennes: PUR. Rastier, F. (2006). Formes sémantiques et textualité. Langages, 3, 99-114. Rastier, F., Cavazza, M., & Abeillé, A. (1994). Sémantique pour l'analyse:(de la linguistique à l'informatique). Paris: Masson. Reddy, M. J. (1979). The conduit metaphor: A case of frame conflict in our language about language. In A. Ortony (Ed.) Metaphor and Thought (2nd ed., pp. 164-201). Cambridge & New York: Cambridge University Press. Reddy, V. (2008). How infants know minds. Cambridge, MA: Harvard University Press. Refaie, E. E. (2003). Understanding visual metaphor: the example of newspaper cartoons. Visual Communication, 2(1), 75. Rensink, R. A., O'Regan, J. K., & Clark, J. J. (1997). To see or not to see: The need for attention to perceive changes in scenes. Psychological Science, 8(5), 368-373. Richards, I. A. (1936). The Philosophy of Rhetoric. 1936. New York: Oxford University Press. Richardson, M. J., Marsh, K. L., & Baron, R. M. (2007). Judging and actualizing intrapersonal and interpersonal affordances. Journal of Experimental Psychology, 33(4), 845-859. Riegler, R. (1907). Das Tier im Spiegel der Sprache, Neusprachliche Abhandlungen. Dresden-Leipzig. Rohrer, T. (2002). The cognitive science of metaphor from philosophy to neuroscience. Theoria et Historia Scientiarum, 6(1), 27. Rohrer, T. (2005). Image schemata in the brain. In B. Hampe (Ed.) From Perception to Meaning: Image Schemas in Cognitive Linguistics (pp. 165-196). Berlin & New York: Mouton de Gruyter.

387


Rohrer, T. (2007). The body in space: Embodiment, experientialism and linguistic conceptualization. In R. M. Frank, T. Dirven, T. Ziemke, & E. Bernárdez (Eds.), Body, Language and Mind, Vol. 2, Sociocultural Situatedness (Vol. 2). Berlin & New York: Mouton de Gruyter. Ronchi, R. (2003). Teoria critica della comunicazione. Pearson Paravia Bruno Mondad. Rosch, E., & Mervis, C. B. (1975). Family resemblances: Studies in the internal structure of categories. Cognitive psychology, 7(4), 573-605. Rosenthal, S. B. (1994). Charles Peirce's pragmatic pluralism. New York: State University of New York Press. Rosenthal, S. B. (2000). Time, continuity, and indeterminacy: A pragmatic engagement with contemporary perspectives. New York: State University of New York Press. Rosero, I., Lecusay, R., & Cole, M. (2010). Ambiguous coordination. Versus, 113-115. Rothenberg, A. (2008). Rembrandt's Creation of the Pictorial Metaphor of Self. Metaphor and Symbol, 23(2), 108-129. Rowlands, M. (1999). The body in mind: Understanding cognitive processes. Cambridge & New York: Cambridge University Press. Rowlands, M. (2009). Enactivism and the extended mind. Topoi, 28(1), 53-62. Rowlands, M. (2009). Extended cognition and the mark of the cognitive. Philosophical Psychology, 22(1), 1-19. Rumelhart, D. E., McClelland, J. L., & The PDP Research Group. (1986). Parallel distributed processing: explorations in the microstructure of cognition (Vol. 1&2). Cambridge, MA: MIT Press. Russell, J. (1996). Agency: Its role in mental development. Hove, Uk: Psychology Press. Salanskis, J. M., & Sinaceur, H. (1992). Le labyrinthe du continu: colloque de Cerisy. Paris: Springer. Sambre, P. (2010). Towards a new legacy of Merleau-Ponty. The Companion of Cognitive Semiotics, I.

388


Santa Ana, O. (1999). Like an animal I was treated': anti-immigrant metaphor in US public discourse. Discourse & Society, 10(2), 191. Schelling, T. C. (1980). The strategy of conflict. Cambridge, MA: Harvard University Press. Schnall, S., Haidt, J., Clore, G. L., & Jordan, A. H. (2008). Disgust as embodied moral judgment. Personality and Social Psychology Bulletin, 34(8), 1096. Schober, M. F. (1993). Spatial perspective-taking in conversation. Cognition, 47(1), 1-24. Schwartz, D. L. (1995). The emergence of abstract representations in dyad problem solving. Journal of the Learning Sciences, 4(3), 321-354. Sebanz, N., Bekkering, H., & Knoblich, G. (2006). Joint action: bodies and minds moving together. Trends in Cognitive Sciences, 10(2), 70-76. Segall, M. H., Campbell, D. T., Herskovits, M. J., Bender, D., & Boye, M. (1966). The influence of culture on visual perception. Indianapolis: Bobbs-Merrill. Semino, E. (2002). A sturdy baby or a derailing train? Metaphorical representations of the euro in British and Italian newspapers. Text, 22(1), 107-139. Semino, E. (2006). A corpus-based study of metaphors for speech activity in British English. Berlin: Walter de Gruyter. Semino, E., & Masci, M. (1996). Politics is football: metaphor in the discourse of Silvio Berlusconi in Italy. Discourse & Society, 7(2), 243. Shannon, C. E. (1948). A Mathematical Theory of Communication. Bell System Technical Journal, 27, 623-656. Shinohara, K., & Matsunaka, Y. (2009). Pictorial metaphors of emotion in Japanese comics. In C. J. Forceville & E. Urios-Aparisi (Eds.), Multimodal Metaphor. Berlin & New York: Mouton de Gruyter.

389


Shintel, H., & Keysar, B. (2009). Less is more: A minimalist account of joint action in communication. Topics in Cognitive Science, 1(2), 260-273. Shockley, K., Richardson, D. C., & Dale, R. (2009). Conversation and coordinative structures. Topics in Cognitive Science, 1(2), 305-319. Short, T. L. (2007). Peirce's theory of signs. Cambridge & New York: Cambridge University Press. Simons, D. J., & Chabris, C. F. (1999). Gorillas in our midst: Sustained inattentional blindness for dynamic events. Perception, 28, 1059-1074. Sinclair, J. (1991). Corpus, concordance, collocation. New York: Oxford University Press. Sinha, C. (1988). Language and representation: A socio-naturalistic approach to human development. New York: New York University Press. Sinha, C., & de Lopez, K. J. (2000). Language, Culture and the Embodiment of Spatial Cognition. Cognitive Linguistics, 11, 17-41. Sini, C. (2000). Presentazione. In C. S. Peirce & W. James, Che cos'è il pragmatismo. Milano: Jaca Book. Sini, C. (2004). La mente e il corpo. Filosofia e psicologia. Milano: Jaca Book. Skagestad, P. (1999). Peirce's Inkstand as an External Embodiment of Mind. Transactions of the Charles S. Peirce Society, XXXV(3), 551-561. Skorczynska, H. (2001). Metaphor in scientific business journals and business periodicals: an example of scientific discourse popularisation. Ibérica, 3, 43-60. Skorczynska, H., & Deignan, A. (2006). Readership and purpose in the choice of economics metaphors. Metaphor and Symbol, 21(2), 87-104. Smith, E. E., & Medin, D. L. (1981). Categories and concepts. Cambridge, MA: Harvard University Press. Smith, E. R., & Conrey, F. R. (2009). The social context of cognition. Cambridge & New York: Cambridge University Press.

390


Smolka, A. L., De Goes, M. C., & Pino, A. (1995). The constitution of the subject: A persistent question. In J. Wertsch, P. del Rio & A. Alvarez (Eds.) Sociocultural Studies of Mind (pp. 165184). Cambridge & New York: Cambridge University Press. Sommers, T., Spurrett, D., Kincaid, H., Ross, D., & Stephens, L. (2007). Distributed Cognition and the Will. Cambridge, MA: MIT Press. Sørensen, B., Thellefsen, T., & Moth, M. (2007). Metaphor and Cognition from a Peircean Perspective. Transactions of the Charles S. Peirce Society: A Quarterly Journal in American Philosophy, 43(3), 562-574. Soskice, J. M. (1987). Metaphor and religious language. New York: Oxford University Press. Speelman, D., Grondelaers, S., & Geeraerts, D. (2003). Profile-based linguistic uniformity as a generic method for comparing language varieties. Computers and the Humanities, 37(3), 317-337. Spivey, J. M. (2007). The continuity of mind. New York: Oxford University Press. Staton, A. Q., & Peeples, J. A. (2000). Educational reform discourse: President George Bush on “America 2000”. Communication Education, 49(4), 303-319. Stefanowitsch, A. (2003). A construction-based approach to indirect speech acts. In K.-U. Panther & L. Thornburg (Eds.), Metonymy and pragmatic inferencing (pp. 105-126). Amsterdam & Philadelphia: John Benjamins. Stefanowitsch, A. (2004). HAPPINESS in English and German: A metaphorical-pattern analysis. In Language, Culture, and Mind (pp. 137-149). Stanford: CSLI Publications. Stefanowitsch, A. (2007). Collocational overlap can guide metaphor interpretation. In G. Radden, K.M. Köpcke, T. Berg and P. Siemund (Eds.), Aspects of Meaning Construction in Lexicon and Grammar (pp. 143-167). Amsterdam and Philadelphia: Benjamins. Stefanowitsch, A. (2010a). Cognitive Linguistics as a Cognitive Science. In M. Callies, A. Lohöfer & W. Keller (Eds.), Bi-directionality in the Cognitive Sciences. Examining the Interdisciplinary Potential of

391


Cognitive Approaches in Linguistics and Literary Studies. Amsterdam & Philadelphia: John Benjamins. Stefanowitsch, A. (2010b). Empirical cognitive semantics: some thoughts. In G. Dylan & K. Fischer (Eds.), Quantitative Methods in Cognitive Semantics. Berlin & New York: Mouton de Gruyter. Stefanowitsch, A., & Gries, S. T. (2003). Collostructions: Investigating the interaction of words and constructions. International Journal of Corpus Linguistics, 8(2), 209-243. Stefanowitsch, A., & Gries, S. T. (2006). Corpus-based approaches to metaphor and metonymy. Berlin & New York: Mouton de Gruyter. Steffensen, S. V. (2009). Language, languaging, and the Extended Mind Hypothesis. Pragmatics & Cognition, 17(3), 677-697. Stern, J. J. (2000). Metaphor in context. Cambridge, MA: MIT Press. Stjernfelt, F. (2000). Diagrams as centerpiece of a Peircean epistemology. Transactions of the Charles S. Peirce Society, XXXVI(3), 357-392. Stjernfelt, F. (2003). Thermodynamic metaphors: A discussion of basic ideas in cognitive semantics exemplified in a hot topic. Semiotica, 146(1-4), 267-285. Stjernfelt, F. (2007). Diagrammatology: an investigation on the borderlines of phenomenology, ontology, and semiotics. New York: Springer. Stockwell, P. (2002). Cognitive poetics: an introduction. London: Routledge. Strong, J., Mathews, T., Sussex, R., New, F., Hoey, S., & Mitchell, G. (2009). Pain language and gender differences when describing a past pain event. Pain, 145, 86-95. Sutton,

J.

(2010).

Memory

(Stanford

Encyclopedia

of

Philosophy).

Retrieved

from

http://plato.stanford.edu/entries/memory/ Sutton, L. A. (2001). Bitches and skankly hobags. In K. Hall & M. Bucholtz (Eds.), Gender articulated (p. 279). New York: Routledge.

392


Sweetser, E. (1990). From etymology to pragmatics: the mind-body metaphor in semantic structure and semantic change. Cambridge & New York: Cambridge University Press. Talebinejad, M. R., & Dastjerdi, H. V. (2005). A cross-cultural study of animal metaphors: When owls are not wise! Metaphor and Symbol, 20(2), 133-150. Talmy, L. (1985). Force dynamics in language and thought (Vol. 21, pp. 293-337). Presented at the Papers from the Regional Meetings, Chicago Linguistic Society, Chicago. Talmy, L. (2000). Toward a Cognitive Semantics, Volume 1&2. Cambridge, MA: MIT Press. Tang, S. (2002). Chengyu Dadian. Taipei: Wunan Press. Taub, S. F. (2001). Language from the body: Iconicity and metaphor in American Sign Language. Cambridge & New York: Cambridge University Press. Teigen, K. H. (1988). The language of uncertainty. Acta Psychologica, 68, 27-38. Tendahl, M., & Gibbs, R. W. J. (2008). Complementary perspectives on metaphor: Cognitive linguistics and relevance theory. Journal of Pragmatics, 40(11), 1823-1864. Thelen, E., & Smith, L. B. (1996). A dynamic systems approach to the development of cognition and action. Cambridge, MA: MIT Press. Thibault, P. J. (1997). Re-reading Saussure: the dynamics of signs in social life. London: Routledge. Thibault, P. J. (2004). Agency and consciousness in discourse: self-other dynamics as a complex system. London & New York: Continuum. Thom, R. (1988). Esquisse d'une sémiophysique. Paris: InterEditions. Thompson, E. (2004). Life and mind: From autopoiesis to neurophenomenology. A tribute to Francisco Varela. Phenomenology and the cognitive Sciences, 3(4), 381-398. Thompson, E. (2007). Mind in life: Biology, phenomenology, and the sciences of mind. Cambridge, MA: Belknap Press.

393


Thompson, E., & Stapleton, M. (2009). Making sense of sense-making: Reflections on enactive and extended mind theories. Topoi, 28(1), 23-30. Thompson, E., & Varela, F. J. (2001). Radical embodiment: neural dynamics and consciousness. Trends in cognitive sciences, 5(10), 418-425. Thompson, R. K. R., Oden, D. L., & Boysen, S. T. (1997). Language-naive chimpanzees (Pan troglodytes) judge relations between relations in a conceptual matching-to-sample task. Journal of Experimental Psychology: Animal Behavior Processes, 23(1), 31. Tissari, H. (2003). LOVEscapes: Changes in prototypical senses and cognitive metaphors since 1500. Helsinki: Société néophilologique. Tolias, A. S., Sultan, F., Augath, M., Oeltermann, A., Tehovnik, E. J., Schiller, P. H., & Logothetis, N. K. (2005). Mapping cortical activity elicited with electrical microstimulation using FMRI in the macaque. Neuron, 48(6), 901-911. Tomasello, M. (1999). The cultural origins of human cognition. Cambridge, MA: Harvard University Press. Tomasello, M., Call, J., & Hare, B. (2003). Chimpanzees understand psychological states-the question is which ones and to what extent. Trends in Cognitive Sciences, 7(4), 153-156. Trevarthen, C. (1979). Communication and cooperation in early infancy: A description of primary intersubjectivity. Before speech: The beginning of interpersonal communication, 321-347. Trevarthen, C. (1993). The self born in intersubjectivity: The psychology of an infant communicating. The perceived self: Ecological and interpersonal sources of self-knowledge, 121-173. Trevarthen, C. (2010). Embodied Human Intersubjectivity Acting Imaginatively To Create and Share Meaning. The Companion of Cognitive Semiotics, I. Turk, D. C., & Melzack, R. (2001). Handbook of pain assessment. New York: Guilford Press. Turnbull, C. M. (1961). The forest people: A study of the Pygmies of the Congo. New York: Simon & Schuster.

394


Turner, M. (1987). Death is the mother of beauty: mind, metaphor, criticism. Chicago & London: University of Chicago Press. Turner, M. (1991). Reading minds. Princeton: Princeton University Press. Turner, M. (1996). The literary mind: The origins of language and thought. New York: Oxford University Press. Turner, M. (2001). Cognitive dimensions of social science. New York: Oxford University Press. Turner, M. (2003). Double-scope stories. In D. Herman (Ed.), Narrative Theory and the Cognitive Sciences (pp. 117-142). Stanford: CSLI Publications. Turner, M. (2006). Compression and representation. Language and literature, 15(1), 17. Turner, M. (2010). Blending Box Experiments, Build 1.0. Working paper available at http://www.ssrn.com/csn/index.html Turner, M., & Fauconnier, G. (1995). Blending and metaphor. In Y. Shen & A. Kasher (Eds.) Cognitive Aspects of Metaphor. London: Routledge. Turner, M., & Fauconnier, G. (1999). A mechanism of creativity. Poetics Today, 20(3), 397-418. Turner, M., & Fauconnier, G. (2000). Metaphor, metonymy, and binding. In R. Dirven & R. Porings (Eds.), Metaphor and metonymy in comparison and contrast (pp. 469-488). Berlin & New York: Mouton de Gruyter. Turner, M., & Fauconnier, G. (2003). Polysemy and conceptual blending. In B. Nerlich, Z. Todd, V. Herman & D. Clarke (Eds.), Polysemy: Flexible patterns of meaning in mind and language (pp. 79-94). Berlin & New York: Walter de Gruyter. Turner, M. E., & Pratkanis, A. R. (1998). Twenty-Five Years of Groupthink Theory and Research: Lessons from the Evaluation of a Theory. Organizational Behavior and Human Decision Process, 73, 105-15.

395


Tylén, K., Fusaroli, R., Østergaard, S., & Bundgaard, P. (2011). Making Sense Together: a dynamical account of linguistic meaning making. Semiotica. Tylén, K., Weed, E., Wallentin, M., Roepstorff, A., & Frith, C. D. (2010). Language as a Tool for Interacting Minds. Mind & Language, 25(1), 3-29. Urban, B. J., Keefe, F. J., & France, R. D. (1984). A study of psychophysical scaling in chronic pain patients. Pain, 20, 157-68. Van Gelder, T. (1995). What might cognition be, if not computation? The Journal of Philosophy, 92(7), 345-381. Varela, F. (1991). Whence the origin of perception? A cartography of current ideas. In J.-P. Dupuy & F. J. Varela (Eds.) Understanding origins (pp. 235-265). Boston: Kluwer Academic Publishers. Varela, F. J. (1979). Principles of biological autonomy. New York: North Holland. Varela, F. J. (1997). Patterns of life: Intertwining identity and cognition. Brain and Cognition, 34(1), 7287. Varela, F. J., & Thompson, E. (2003). Neural synchrony and the unity of mind: A neurophenomenological perspective. The Unity of Consciousness, 266-287. Varela, F. J., Thompson, E., & Rosch, E. (1991). The embodied mind: Cognitive science and human experience. Cambridge, MA: MIT press. Violi, P. (1997). Significato e esperienza. Milano: Bompiani. Violi, P. (2007). Beyond the body: towards a full embodied semiosis. In T. Ziemke, J. Zlatev, R. M. Frank (Eds.), Body, Language and Mind: Embodiment (pp. 53-76). Berlin & New York: Mouton de Gruyter. Violi, P. (2010). How our bodies become us. The Companion of Cognitive Semiotics, I. Visetti, Y. M., & Cadiot, P. (2006). Motifs et proverbes. Paris: PUF.

396


Vygotsky, L. S. (1962). The Genetic Roots of Thought and Speech. In Thought and Language. Cambridge, MA: MIT Press. Waddie, N. A. (1996). Language and pain expression. Journal of Advanced Nursing, 23, 868-72. Wang, M. (n.d.). Yongda Xinbian Chengyu Cidian. Taipei: Yongda Press. Weber, A., & Varela, F. J. (2002). Life after Kant: Natural purposes and the autopoietic foundations of biological individuality. Phenomenology and the Cognitive Sciences, 1(2), 97-125. Whaley, C. R., & Antonelli, G. (1983). The birds and the beasts: Woman as animal. Maledicta: International Journal of Verbal Aggression, 7, 220. Wheeler, M. (2004). Is language the ultimate artefact? Language sciences, 26(6), 693-715. Wheeler, M. (ms). Extended X: Recarving the Biological and Cognitive Joints of Nature. Available at

http://www.philosophy.stir.ac.uk/staff/m-wheeler/wheeler-page.php Wierzbicka, A. (1985). Oats and wheat: Mass nouns, iconicity and human categorization. In J. Haiman (Ed.), Iconicity in syntax. Amsterdam & Philadelphia: John Benjamins. Wierzbicka, A. (1996). Semantics: Primes and universals. New York: Oxford University Press. Wilcox, P. P. (1993). Metaphorical Mapping in American Sign Language. Ph.D Dissertation, University of New Mexico. Willerslev, R. (2007). Soul hunters: hunting, animism, and personhood among the Siberian Yukaghirs. Berkeley: University of California Press. Williams, A. C. D. C., Daview, H. T. O., & Chadury, Y. (2000). Simple pain rating scales hide complex idiosyncratic meanings. Pain, 85, 457-463. Williams, R. (2005). Material anchors and conceptual blends in time-telling. Ph.D Dissertation, Berkeley University. Wilson, M. (2002). Six views of embodied cognition. Psychonomic Bulletin & Review, 9(4), 625.

397


Wilson, R. A. (2004). Boundaries of the mind: the individual in the fragile sciences: cognition. Cambridge & New York: Cambridge University Press. Wilson, R. A., & Clark, A. (2009). How to situate cognition. In P. Robbins and M, Aydede (Eds.). The Cambridge handbook of situated cognition. Cambridge & New York: Cambridge University Press. Wilson, T. D. (2002). Strangers to ourselves. Cambridge, MA: Belknap Press. Winter, S. L. (2001). A Clearing in the Forest: Law, Life, and Mind. Chicago & London: University of Chicago Press. Wittgenstein, L. (1953). Philosophical investigations. Oxford: Blackwell Publishing. Wright, G. (2007). The anatomy of metaphor. Cambridge, UK: Clare College. Zahavi, D. (1996). Husserl's intersubjective transformation of transcendental philosophy. Journal of the British Society for Phenomenology, 27(3), 228-245. Zarnoth, P., & Sniezek, J. A. (1997). The Social Influence of Confidence in Group Decision Making. Journal of Experimental Social Psychology, 33, 345-66. Ziemke, T., Zlatev, J., & Frank, R. M. (2007). Body, Language and Mind: Embodiment. Berlin & New York: Mouton de Gruyter. Zinken, J., Hellsten, I., & Nerlich, B. (2008). Discourse metaphors. In R. M. Frank, T. Dirven, T. Ziemke, & E. Bernárdez (Eds.), Body, Language and Mind, Vol. 2, Sociocultural Situatedness (Vol. 2). Amsterdam & Philadelphia: Mouton de Gruyter. Zlatev, J. (1997). Situated embodiment: Studies in the emergence of spatial meaning. Lund: Gotab. Zlatev, J. (2005). What's in a schema? Bodily mimesis and the grounding of language. In B. Hampe (Ed.) From Perception to Meaning: Image Schemas in Cognitive Linguistics (p. 313-333). Berlin & New York: Mouton de Gruyter. Zlatev, J. (2007). Embodiment, language and mimesis. In T. Ziemke, J. Zlatev, R. M. Frank (Eds.), Body, Language and Mind: Embodiment (Vol. 1, pp. 297-337).

398


Zlatev, J. (2010). Phenomenology and cognitive linguistics. In Handbook of Phenomenology and Cognitive Science (pp. 415-443). Zwaan, R. A., & Kaschak, M. P. (2009). Language in the brain, body, and world. In P. Robbins and M, Aydede (Eds.). The Cambridge handbook of situated cognition. Cambridge & New York: Cambridge University Press. Zwaan, R. A., & Radvansky, G. A. (1998). Situation models in language comprehension and memory. Psychological

Bulletin,

123,

399

162-185.


Acknowledgments My thanks to those who crossed and sometimes coordinated their intellectual path with mine. My colleagues and collaborators in Bologna and Århus. The members of the Distributed Language Groups. The numerous guest lecturers and discussants. The contributors to the journal I edited. My supervisor. My students. All parts of that distributed process that lasted three exciting years, that led to this dissertation, that leads beyond it. I am a better researcher because of you. My thanks to those that went beyond that. Those who believed in me. Those who shared days and months of discussions and friendship and words that went deep. Those who sat around a beer, a bottle of wine or a cachaça. Or a rye bread. Those who got driven along the Pacific ocean and had to drive me back. Those who got stranded in a hotel in Turku. Those who are still waiting for me in Paris. Those whose mails still make me think and laugh and carry me through the day. My family. I am a better person because of you. My last most precious thank to the one who keeps reminding me that, at the end of the day, semiotics is not all that important. Tú ert eitt gott glas av vin sera sjáldsom við djúpum farvum. Eg elski teg.

400