Apr 7, 2013 - Downing and Locke (1992:379), will put it, âmodality is said to express a ...... Vanessa be.able.to sing song for goose.pimples pour 2SG.
LINGUISTIC STRATEGIES IN THE EXPRESSION OF MODALITY IN AKAN
BY
AUGUSTINA OWUSU POKUA (10396178)
THIS THESIS IS SUBMITTED TO THE UNIVERSITY OF GHANA, LEGON IN PARTIAL FULFILLMENT OF THE REQUIREMENT FOR THE AWARD OF M.PHIL LINGUISTICS DEGREE.
JULY 2014
DECLARATION I declare that except for references to works that have been duly cited, this thesis is a result of my original research, under the supervision of Professor Nana Aba A. Amfo and Dr. Evershed Kwasi Amuzu, and that, it has neither in whole nor in part been presented for another degree elsewhere.
CANDIDATE .............................. AUGUSTINA OWUSU POKUA
14/02/2015 ....................................
DATE
SUPERVISOR …............................ PROF. NANA ABA A. AMFO
.…................................... DATE
SUPERVISOR .............................. DR. EVERSHED KWASI AMUZU
!
…....................................... DATE
ii
DEDICATION This thesis is dedicated to my parents, WOI Samuel Owusu and Mrs. Joyce Owusu Thank you for supporting me this far, I love you.
iii
ACKNOWLEDGEMENT I am very grateful to all those who in diverse ways assisted toward the success of this thesis. My ultimate gratitude goes to God; for it is His mercies and favour that have brought me this far. I am also very grateful to my supervisors, Professor Nana Aba A. Amfo and Dr. Evershed Kwasi Amuzu for bringing out the best in me. God richly bless you in all your endeavors. To all the lecturers of the Linguistics Department, especially Prof. Kofi Agyekum, Prof. K. Saah, Dr. George Akanlig-Pare and Dr. Clement K. Appah, thank you for your advice and help. I will like to say a big thank you to Mr. Kofi Busia Abrefa of the Linguistics Department, University of Cape Coast. I also thank my family: my parents WOI Owusu and Mrs. Joyce Owusu, my siblings, Maame Yaa and Nana Yaw and my niece Nana Yaa for the financial, moral and spiritual support. I love you guys. I appreciate the help of all my colleagues and friend especially Miss Monica Apenteng, Miss Rachel Thompson, Miss Abigail Nkansah, Miss Emelia Amfo-Okyere and Mr, Bernard Okpattah for helping with the editing. Mention must be made of my colleague and friend, Mr. Obed Nii Broohm, who also helped with proofreading and printing of this work. A special thanks goes to members of my focus group, you have been of great help.
Thank you.
iv
ABSTRACT This thesis explores the concept of modality in Akan, a Kwa language spoken in Ghana. The main focus of the thesis is the linguistic strategies employed in the expression of modality in Akan. In addition to these strategies, other issues that are linked to modality such as modal concord and the use of some modal expressions as politeness strategies are also discussed. The linguistic strategies for expressing modality in Akan are group into three main categories: lexical strategy, modals in the verbal group and periphrastic modal constructions. The lexical strategies include the use of factive nim ‘to know’ and attitudinal verbs, susu ‘to suppose’ or ‘to suggest’, and dwene ‘to think’ and epistemic modal adverbs such as nokware ‘truly/honestly’ and ampa ‘truly/honestly’, ebia ‘maybe’, gyama ‘perhaps’, anhwe a ‘probably’ sɛsɛɛ ‘by this time’ and annyɛbiara ‘probaby’. I use verbal group to refer to the strategy that involves the use of the auxiliary verb tumi ‘be able to’ and the modal prefix bɛ-. Following Ameka (2008), I refer to the combination of words that does not qualify as either phrases or sentences that are used to express modality in Akan as periphrastic modal constructions. These constructions come in three forms: 3SG + verb+COMPL clause, 3SG + bɛ+ verb+ COMPL clause and NP + wͻ + NP + COMPL clause. The verb slot in these constructions can be filled by only some selected verbs in Akan. These verbs are wͻ ‘to be located at’ or ‘to have’, sɛ ‘to be fit’, fata ‘to be deserving’ and yɛ ‘to be good’ can fill the verb position. Only three nouns can occupy the second NP position in the third construction; awerehyerem ‘hope’, anidaso ‘hope’ and gyedi ‘belief’. Modal concord is possible in Akan though there are no modal
v
auxiliaries in the language. All the modal strategies could be involve in a concord interpretation. This disapproves Zeijlstra’s (2007) claim that modal concord is necessary between a modal verb and an adverb. Modal markers in Akan are procedural markers that point the hearer to a specific interpretation path during communication.
vi
TABLE OF CONTENTS
DECLARARTION ............................... ERROR! BOOKMARK NOT DEFINED. DEDICATION ......................................................................................................... II ACKNOWLEDGEMENT ...................................................................................... III ABSTRACT ............................................................................................................IV TABLE OF CONTENTS ........................................................................................VI LIST OF ABBREVIATIONS .................................................................................IX CHAPTER ONE ....................................................................................................... 1 INTRODUCTION ..................................................................................................... 1 1.0 OVERVIEW ...................................................................................................... 1 1.1 BACKGROUND TO THE STUDY ........................................................................... 2 1.1.2 Modal strategies ......................................................................................... 5 1.2 PROBLEM STATEMENT .................................................................................... 8 1.3 RESEARCH QUESTIONS ................................................................................... 9 1.4 OBJECTIVES OF THE RESEARCH....................................................................... 9 1.5. SIGNIFICANCE OF THE STUDY ........................................................................ 9 1.6 ORGANIZATION OF THE THESIS ..................................................................... 10 CHAPTER TWO..................................................................................................... 11 LITERATURE REVIEW, THEORETICAL FRAMEWORK AND METHODOLOGY .................................................................................................. 11 2. 0 OVERVIEW ..................................................................................................... 11 2.1 REVIEW OF RELATED LITERATURE................................................................... 11 2.1.1 Modal auxiliary ....................................................................................... 12 2.1.2 The future marker and modality .............................................................. 15 2.1.3
Modal adverbs, adjectives and other ways of expressing modality .... 18
2.1.4
Modal concord .................................................................................... 22
vii
2.1.5
Studies on modal systems of languages .............................................. 24
2.1.6
Non-Modal meanings .......................................................................... 27
2.2
THEORETICAL FRAMEWORK: RELEVANCE THEORY .................................... 29
2.2.1 Procedural versus Conceptual meaning ................................................... 31 2.3 METHODS ...................................................................................................... 33 2.3.1
Sources and Methods of Data Collection ............................................ 33
2.3.2
Method of data analysis ...................................................................... 35
2.4
CHAPTER SUMMARY ................................................................................... 36
CHAPTER THREE ................................................................................................. 37 THE LEXICAL MEANS OF EXPRESSING MODALITY IN AKAN ................. 37 3.0
OVERVIEW .................................................................................................. 37
3.2
VERBS THAT EXPRESS MODALITY IN AKAN ................................................. 38
3.2.1
Nim ‘to know’ ..................................................................................... 39
3.2.2
Gye di ‘to believe’ ............................................................................... 42
3.2.3 Susu ‘to assume’ or ‘to suggest’ ............................................................. 43 3.2.4
Dwene ‘to think’ ................................................................................. 46
3.3.5
Summary of discussion on mental state verbs..................................... 49
3.3
ADVERBS .................................................................................................... 49
3.3.1
Certainty adverbs: Ampa and Nokware .............................................. 51
3.3.2 Uncertainty adverbs: ebia, gyama, anhwɛ a, sɛsɛɛ and annyɛbiara. ...... 54 3.4
CHAPTER SUMMARY ................................................................................... 60
4.0 OVERVIEW .................................................................................................... 61 4.1
TUMI ‘BE ABLE TO’ AS A MODAL MARKER ................................................... 62
4.1.1
Tumi indicating dynamic modality ..................................................... 64
4.1.2
Tumi indicating epistemic modality .................................................... 66
4.1.3
Tumi indicating deontic modality ....................................................... 66
4.1.4
Conclusion on tumi ............................................................................. 67
4.2
THE STATUS OF FUTURE MARKERS .............................................................. 69
4.2.1
The future marker bɛ- .......................................................................... 70
4.2.2
Bɛ- as a modal marker in Akan ........................................................... 71
4.2
THE CURRENT STATUS OF BƐ- IN AKAN ....................................................... 79
viii
4.4
CHAPTER SUMMARY ................................................................................... 81
5. 0
OVERVIEW .................................................................................................. 82
5. 1
CONSTRUCTIONS THAT ARE USED TO EXPRESS MODALITY IN AKAN ............ 82
5.1.1
3SG + verb + Sɛ Clause ...................................................................... 86
5.1.2.
3SG + bɛ + verb + Sɛ Clause .............................................................. 95
5.1.3
NP + wͻ + NP + Sɛ Clause ................................................................. 96
5.2
MODAL CONCORD IN AKAN ........................................................................ 97
5.3
MODAL MARKERS IN AKAN AS POLITENESS MARKERS. ............................. 101
5.3.1 Politeness ............................................................................................. 101 5.4
CHAPTER SUMMARY ................................................................................. 108
CHAPTER SIX ..................................................................................................... 110 CONCLUSION ..................................................................................................... 110 6.0
OVERVIEW ................................................................................................ 110
6.1
SUMMARY ................................................................................................. 110
6.2
FINDINGS .................................................................................................. 112
6.2.1 Lexical categories that express modality ............................................... 112 6.2.2 Modals in the verbal group ................................................................... 114 6.2.3 The Periphrastic Modal Constructions ................................................. 115 6.2.4
Modal Concord.................................................................................. 116
6.2.5
Modal markers as politeness markers. .............................................. 117
6.3. Recommendations ................................................................................... 118 REFERENCES ...................................................................................................... 120
ix
LIST OF ABBREVIATIONS
1SG
Ist Person Pronoun
PST
Past Tense
COMP
Complementizer
PROG
Progressive marker
PERF
Perfective aspect
POSS
Possesive marker
TM
Topic Marker
SUBJ
Subject
1PL
Ist Person Plural
3PL
3rd Person Plural
NEG
Negative marker
COND
Conditional
IMP-SUBJ
Impersonal subject
CFM
Clause Final Marker
FOC
Focus Marker
DEF
Definite Determiner
CONT
Continuative Aspect
MOD
Modal marker
COMPL
Completive aspect
Singular
10
CONJ
Conjunction
MOT
Motional Prefix
DET
Determiner
REL
Relativiser
QM
Question Marker
COP
Copular Verb
CONJ
Conjunction
COND.
Conditional Marker
QUANT
Quantifier
EMP.PART
Emphatic Particle
1
CHAPTER ONE INTRODUCTION
1.0 Overview This thesis explores the concept of modality in Akan, a Kwa language spoken in Ghana. The focus of the thesis is to discuss the linguistic strategies that are employed to express modality in the language. The discussion is divided into three parts. The first section focuses on the use of lexical words (verbs, adverbs) to express modality. The second part investigates the use of the auxiliary verb1 tumi and future affix2 bɛ to express modality. The third section speaks to three issues: periphrastic modal constructions, modal concord and the use of modals as politeness markers in Akan. This chapter provides an introduction to the study. The chapter begins with a brief background to the concept of modality in language; paying attention to the linguistic devices that encode the concept. Then the need for the research is stated; emphasis is placed on the gap in knowledge that this research s ee ks t o fill. The main questions upon which the research thrives are laid out in the research questions. The objectives and significance of the study are also stated. This first chapter further provides an overview of the thesis.
1
This is the first time the verb is discussed as an auxiliary verb. This thesis argues elsewhere that bɛ is not a future marker in Akan; it use here is therefore consistency with earlier literature. 2
2
1.1 Background to the study The concept of modality is illusive and thus there is no unitary definition of modality that is agreed upon by scholars who work in the field. Nuyts (2006:2) comments that “there is also no unanimity among scholars as to how the set of modal categories should be characterized …..and also a difference in opinion about which features motivate the grouping of categories under the modal umbrella”. For the reason stated above, different scholars provide different definitions to describe the concept of modality. Kratzer (1977, 1981 and 1991), one of the early scholars on modality, working in the modal logic framework, defines modals as propositional operators that quantify over possible worlds (PWs) and relate them to the proposition under question. Kratzer treats modals as operators like and and or. The only difference between these operators is that while and and or quantify over sets, modals quantify over possible worlds. Possible worlds refer to alternative worlds including the real one where a proposition in a modalized utterance could occur. Scholars who work in the modal logic framework usually adopt Kratzer’s definition. Similar to Kratzer’s definition is Fasold and ConnorLinton (2006:153) definition of modality. They define modality as “aspects of meaning that cause sentences to be about the non-factual that is, about the alternative possibilities for how things could be". They also seem to posit the idea of possible worlds.
3
Palmer (1986, 1990, 2003) defines modality as the attitude a speaker expresses towards the proposition in the sentence he utters. Similar to Palmer’s definition, Downing and Locke (1992:379) add that “the attitude modals express can be grouped into two; assessing the truth of the proposition or the potential occurrence of the event and intervening in the speech event, by laying down obligations and giving permission”. What all these definitions have in common is that they all express the fact that modals do not state a fact nor express what is real. As Downing and Locke (1992:379), will put it, “modality is said to express a relation to reality whereas an unmodalized declarative treats the process as reality”. Cook (1978) identifies two types of modals, epistemic modals and root modals. Differentiating between the two, he explains that “epistemic modality modifies a sentence and deals with the truth value of that sentence; root modality relates . . . a subject to an activity and deals with permission, obligation and ability” (1978:6). He further explains that root modality can be of different flavors such as dynamic, bouletic or boulomaic, teleological and deontic modality. Dynamic modality 'is concerned with the disposition of certain empirical circumstances with regard to the occurrence of some event' (Perkins 1983: 34); Perkins (1983: 11) classifies boulomaic modality, as a type of dynamic modality where there is a disposition from a human source. Deontic modality describes possibilities in which certain rules, regulations, laws are upheld and teleological modality describes possibilities in which certain goals or preferences are achieved. Other authors like Bybee et al. (1995), differentiate between three types of
4
modality; agent-oriented modality, speaker-oriented modality and epistemic modality. Agent-oriented
modality
applies
to
all
modalities
in
which
conditions are predicated on an agent (ability, desire, and obligation) as opposed to speaker-oriented modality which aims at getting something done (imperatives, permissives, and proposition
optatives).
Epistemic
modality
applies
to
a
whole
and communicates a speakers’ stance concerning its truth (1995:5).
The agent/speaker- oriented modalities proposed by Bybee et al. (1995) are not different flavors of Cook’s root modality: deontic and dynamic modalities. In this thesis, we focus on only three types of modalities, and these are epistemic modality which is central in all modality discussions and the two flavors of root modality; dynamic and deontic modality. Deontic modality according to Papafragou (1998) is concerned with the necessity or possibility of acts performed by morally responsible agents and thus with obligation and permission. Dynamic modality also is concerned with the ability or volition of the subject of the sentence (Palmer 1990:36). Epistemic modality on the other hand according to Palmer (2003:7) is concerned with the speaker’s “judgments about the factual status of the proposition”. Following Carrerota (1992), expressions that do not indicate doubt to the linguistic codability of the knowledge the speaker (S) wants to transmit will be considered as epistemic in this study. She notes that expressions of this sort weaken S’s commitment with respect to the truth of the proposition indicating imprecision of S’s knowledge. An example of such an expression is about in the sentence ‘Mary’s about six feet tall.’
5
1.1.2 Modal strategies Modality i s expressed by various morphological, syntactic, and lexical categories in language. Most often than not a language has more than one strategy of expressing modality. Palmer (2001) distinguishes two major grammatical devices for expressing modality, namely modal systems and mood
3
(indicative and
subjunctive), (Palmer 2001: 104). Narog (2009) identifies adverbs and mental state predicates
as non-grammatical devices
for expressing modality.
In
English, devices for expressing modality are mostly lexical categories, except for the modal auxiliaries. In other languages, especially African
languages, the
grammatical devices may include affixes and also constructions. In the subsequent sections, we describe the devices that are used to express modality crosslinguistically
1.1.2.1 Modal auxiliary verbs The most studied linguistic item for expressing modality is the modal auxiliary; this is because most research on modality is on Indo-European, which has modal auxiliaries. Modal auxiliaries are hard to define cross-linguistically; some languages are argued not to have any. In English they are differentiated from other verbs by what is called the ‘NICE’ properties;(i) Negation without do (isn’t, mustn’t vs doesn’t go); (ii) Inversion without do (is he? must he? vs does he go?), (iii) Clitics (isn’t, mustn’t vs doesn’t go), (iv) Ellipsis without do (often called 'code': but I don’t think Peter is /can vs but I don’t think Peter did. English modal auxiliaries include
3
Most authors use mood and modality synonymously
6
can, may, might, must, will, and should. They are usually used to specify likelihood, ability, permission, and obligation. More often than not modal auxiliaries express more than one modal type; this ambiguity can only be resolved in context. For instance the sentence below has two possible interpretations depending on the interpretation must possess in the context. 1) Obed must be asleep by now The first possible interpretation is that looking at the evidence available, probably there is no light in his room and the door is also locked, t h e r e i s a h i g h c h a n c e Obed is asleep. This is the epistemic interpretation. The second possible interpretation is that there is higher authority that is ordering Obed to sleep at the time of speaking; this has a deontic modality interpretation.
1.1.2.2
Mood
Ferdinand (2006:33) defines mood as “a morphological verbal category, w h i c h expresses the modal value of the sentence”. He goes on to explain that mood is “therefore the grammaticalized expression of modality, just as, say, tense is the grammaticalized expression of time”. For Bybee & Fleischman (1995: 2) the term mood is used for a category of meanings that are 'expressed inflectionally, generally in distinct sets of verbal paradigms, e.g. indicative, subjunctive, optative, imperative, conditional etc. Mood can thus be explained as the use of verbal inflections that allow speakers to express their attitude towards what they are saying. The distinction made between mood and modal expressions as one being inflection on the verb and the other
7
being a free morpheme is quite arbitrary. This is because languages encode these concepts differently; a modal expression may be bound while a mood marker is a free morpheme. Examples of moods include, indicative, interrogative, imperative, emphatic, subjunctive, injunctive, optative, and potential.
1.1.2.3 Modal adverbs and adjectives These are adverbs and adjectives that are used to express modality in language. Modal adverbs in English include probably, frankly, possibly and certaintly. In Akan, modal adverbs include ebia ‘maybe’, gyama ‘probably’ and sesɛɛ ‘perhaps’. Adjectival modals in English include, probable, certain and sure. Akan adjectives comprise pefee ‘clear’, nokware ‘true’ and ampa ‘true’. Adverbial and adjectival modals usually have epistemic modal meanings. They are mostly used to show the level of commitment of the speaker to a proposition expressed. They can also be 'harmonic' (reinforce the modality of the auxiliary), or can modify the modality in a different direction (perhaps we could…).
1.1.2.4 Mental state verbs Apart from the modal auxiliary verbs, verbs that relate to the mental state of the speaker can also be used to express modality. These verbs can be group into two: factive verbs and attitudinal verbs. The use of factive verbs presupposes the truth of an embedded sentence that serves as complement. Factive verbs include know, learned, realize etc. Attitudinal verbs on the other hand refer to mental or emotional states and feeling or opinion, they include verbs like believe, think, guess etc. The
8
complements of these verbs are usually complement clauses. They are used to express epistemic modality, which is the expression of speaker’s confidence towards the truth of the proposition. Sometimes, they can also occur clause finally, and this is what Thompson and Mulac (1991) call “modal tags” examples are; I think, I guess and I know. These tags occur at clause final position, and they express a speaker’s attitude toward the whole proposition. For example 2) I think his car was parked at the car park, when I climbed up 3) His car was parked at the car park, when I climbed up, I think.
1.2 Problem Statement Cross linguistically, research on modality has concentrated on modal auxiliaries, leaving the other strategies less explored. In Akan, research on modality often targets: single markers of modality such as Amfo’s (2005) treatment of anka as a modal marker, or one modal strategy for example Sakyi’s (2013) discussion of the expression of epistemic modality by predicative adjectives. Modals are also briefly talked about in works on tense, mood and aspect, for instance Osam (2008) talks about modality in his paper Akan as an Aspectual Language and Boadi (2008) also discusses modality briefly in his article Tense, Aspect and Mood in Akan. There is how e v er no holistic description of the modal system of Akan. This thesis therefore seeks to fill this gap in the literature by providing a holistic study of the modal system of Akan.
9
1.3
Research Questions
The main questions on which this study thrives are: What lexical strategies are used in the expression of modality in Akan? What syntactic strategies are used to indicate modality? What types of modal meanings are encoded by these strategies? What other meanings do modal markers have in discourse?
1.4
Objectives of the Research
The main objective of the study is to provide a comprehensive description of modal expressions of Akan. The specific objectives of the study are as follows: s seems to hold To identify the lexical strategies used to mark modality in Akan. To determine the syntactic structures used to mark modality in Akan. To ascertain the different modal meanings encoded by the different strategies. To investigate if modal markers have other uses in discourse.
1.5.
Significance of the Study
The study is a comprehensive work on modality in Akan. It is a useful addition to existing literature in linguistics in general, and the area of the study of modality precisely. Moreover, it serves as a reference document for future studies on modality.
10
1.6
Organization of the Thesis
The main objective of this thesis is to describe the linguistics strategies that are used to express modality in Akan. Chapter One provides an introduction to the study by explaining the concept of modality, and the linguistic items used to express modality. It also covers the statement of the problem, research questions, and the objectives, significances of the study and an organization of the study. Chapter two discusses some relevant literature on modality and also the theoretical framework in which the study is advanced (Sperber and Wilson’s Relevance Theory). In Chapter two, the methods employed in the collection of data for the study is also discussed. Chapters three, four and five present analyses of data collected. Chapter six which is the final chapter presents a summary of the thesis, findings, conclusion and recommendations.
11
CHAPTER TWO
LITERATURE REVIEW, THEORETICAL FRAMEWORK AND METHODOLOGY
2. 0 Overview This chapter builds the necessary contextual and conceptual background needed for the thesis. The chapter is divided into three sections. The first section reviews relevant literature on modality. The second section discusses Sperber and Wilson’s (1988, 1995) theoretical framework ‘Relevance Theory’, the framework in which the study is situated. It specifically focuses on the ‘procedural versus conceptual meaning’ distinction made by Blakemore (1987) within the relevance theoretic framework. The third section discusses the methods that are employed in the collection of data for this study. The chapter ends with a summary of the issues that have been discussed.
2.1 Review of related literature Like most linguistic phenomena, the concept of modality has been studied from different perspectives and various aspects of it have been studied. Following Kratzer’s
seminal
work
on modality in 1981, most studies on modality,
especially in Indo European languages have focused on modal auxiliaries. Another issue that has gained prominence in most modal studies is the status of the future
12
marker in most languages. There are also studies on the use of adverbs, adjectives and other single lexical items to express modality. Another interesting aspect of modality that is studied is what Geurts and Huitink (2006) call ‘modal concord’. A few studies also focus on the description of all aspects of modal expressions in a specific language (Ameka 2008, Narrog 2009 and Uchechukwu 2008). The review of the literature will be done under these headings; modal auxiliaries, the interaction between modality, tense and aspect, the future and modality, modal adverbs, adjectives and other modal works, and other uses of modals. Literature on modal concord will also be reviewed.
2.1.1 Modal auxiliary The first thing that comes into mind when modality is mentioned is modal auxiliaries or modal verbs. This stems from the fact that the English modal auxiliaries have been the focal point of most studies on modality. In fact, most assertions about modality in general are based on the analysis of the modal auxiliaries. Works on modal auxiliaries either discuss a selected number of modals or present a detailed analysis of a particular modal auxiliary. Papafragou (2000) is one such work that focuses on a selected number of modals; the modals she used are must, may, can, should, and ought to. The main aim of her work w a s t o explore the interaction between lexical semantic information and pragmatic inference in relation to these modal verbs. Working in the relevance theoretical framework she claims, “the semantically encoded content of these verbs grossly underdetermines the interpretations which the verbs
13
receive
during
utterance
comprehension” (Papafragou 2000:7). She argues
that English modals have unitary semantic content, which in conjunction with different pragmatic considerations, give rise to an array of distinct contextual readings. She also explains that modals are semantically underdetermined which means that the information encoded by modals are the inputs to pragmatic mechanisms of utterance comprehension, which yield the contextual variability of modal interpretations (Papagragou 2000: 206). Thus, she concludes can, should and ought to are semantically general and require free enrichment while must and may are semantically incomplete and therefore require pragmatically induced saturation. Gisborne (2007) on the other hand does not focus on the meaning of modal verbs in themselves but how the meaning of can and will prove that there is no such thing as dynamic modality. He argues that the meaning encoded by can and will, which is labeled dynamic modality, is actually retention of some earlier non-modal meanings of these forms. His main argument is that dynamic modality does not fit any of the criteria for modality. These criteria are that: “it involves the context, the force-dynamic relations Initiator and Endpoint link to elements within the context such as the speaker rather than to elements within the sentence; it is performative and temporally located in the speech event” Gisborne (2007:50). He continues that it will be “inappropriate to subsume it under the category of root modality or to assert that it belongs together with deontic modality as a kind of agent-oriented modality” (Gisborne 2007:46). He concludes that, dynamic meaning is not a variety of modality and that it is wrongly labeled as such because it is regularly associated with will and can. The picture that emerges is more interesting: modal
14
verbs can also have non- modal meanings In a descriptive analysis of tense, aspect and modality in Saramaccan/Saam ka, van de Vate (2011) discusses the semantics of two Saam ka modal auxiliaries musu and sa. She categorizes musu as a necessity modal and sa as a possibility modal. “The necessity modal musu conveys a deontic obligation reading and a deductive epistemic reading, and the possibility modal sa conveys a dynamic ability reading, a deontic permissive reading and a speculative epistemic reading” van de Vate (2011:147) notes. She analysed her data with Kratzer’s (1977, 1991, 2002) analysis of modals using possible worlds. She asserts that when musu conveys a deontic obligation reading, it can have a circumstantial modal base and a deontic ordering source or a circumstantial modal base and bouletic ordering source expression (van de Vate 2011:148). These ordering sources are dependent on the context in which the modal is used. When musu has deductive epistemic reading, its ordering sources range from empty to stereotypical to informational. According to van de Vate (2011:154) “[i]n possible world semantics, the morpheme has a universal quantificational modal force”. Depending on the context, sa also conveys a deontic permissive reading, a dynamic ability reading or a speculative epistemic reading (2011:157). When used in a sentence, the modal force of sa is existential and the modal base is circumstantial. Although other ordering sources are allowed, the most common ordering source for the permissive is deontic. When sa conveys an ability reading, it could be a mental or physical abilitiy of an agent. In this instance, it has a circumstantial modal base and the ordering sources range from dynamic to teleological, and from bouletic to stereotypical. Sa as used with the speculative
15
epistemic modality reading has an epistemic modal base and the ordering source can be informational, stereotypical and/or empty. The modal base is either epistemic, expressing the speaker's knowledge, or circumstantial expressing the circumstances under consideration.
2.1.2 The future marker and modality One of the things that have always puzzled researchers who worked on modality is the status of the future marker. The main question has always been whether the linguistic expressions used to express future states of affairs are different from those used to refer to modality. This issue stems from the fact that reference to future states of affairs is inherently non-factual and non-factuality is also a main defining component of modality. Lyons (1968:275) adds that “the so-called future tense of many languages has modal as well as tense value…thought should be borne in mind that this is often as much a mood as a tense”. There is no census on what these linguistic expressions really encode cross-linguistically; they are analyzed differently in different languages. In some languages they have been analyzed as encoding only modal meanings, in others both modal meanings and time referent while in others they only have time referent. The analysis of the future marker in Ewe, Akan, English and Paraguayan Guaraní is presented below. In Ewe, a Kwa languae, Essegbey (2008) argues that the a-morpheme has erroneously been referred to as a future marker in previous works. He points out that a future analysis of the a-morpheme is not compatible with the analysis of tense in general in the language. He explains that like most Kwa languages, Ewe is a tenseless language; there is no explicit marker of tense. If the language does not
16
mark tense then it cannot have a future and non-future tense distinction. According to Essegbey (2008), the a-morpheme cannot be a future marker because it does not have temporal deixis as its primary function. According to him, the reason why the morpheme often receives future-tense analysis is because in some contexts it invites “yet-to occur” interpretations. He however quickly adds that such interpretations are merely preferred interpretations and, therefore, do not constitute the core meaning of the form. In support of Essegbey’s claim, Ameka (2008:139) affirms that the morpheme “has been erroneously characterized as the future tense marker in Ewe”. He also adds that, though there are some uses of the morpheme that can be said to have future meanings, there are many uses of the morpheme that are not related to future. Essegbey agrees that the morpheme is on its way to being grammaticalized into a future morpheme in the language but “It has not been fully grammaticalized yet and so for now it is a modal marker which has the meaning of potentiality” (Essegbey 2008:209). An extreme opposite to Essegbey’s position is Osam’s (2008) and Boadi’s (2008) evaluation of the so-called future morpheme in Akan, bɛ. According to Osam (2008:71), though Akan like Ewe is a tenseless language, “Akan makes a distinction between future and non-future tense”. He argues that the future is signaled by the prefix bɛ- while the non-future is unmarked. He adds that though the primary function of the prefix is to indicate future time reference, it also has some other semantic functions. These secondary functions include prediction, intention and sometimes ability. More importantly he argues that these secondary functions can only be realized when the morpheme is used with other modals or the verb tumi ‘be able to’. Boadi (2008:22) adds that “[a]lthough in most contexts, the affix under
17
consideration has a future time reference; there are sentences in which the future affix does not refer to future time at all”. He gives instances when the prefix is used to mark habitual aspect and also be used to express probability or likelihood. Like Osam, he points out that the probability or likelihood meaning is only possible when it is used with some specific verbs such as verbs like wͻ ‘be located at’, nim ‘know’; só ‘be big’; sùa ‘be small’, (Boadi 2008:23). The more work done on it, the fuzzier its analysis is; the English will has received many different analyses in the literature (see Walton 1988, Klinge 1993, Nicolle 1997, Berbeira- Gardón 2006). All these analyses have different conclusions and so the debate about will has still not been resolved. This study focuses on BerbeiraGardón’s (2006) analysis of will. Berbeira- Gardón (2006) presents the view that will does not encode time reference. In other words will is not a future marker. He based his argument on the fact that it is possible to have propositions expressed by utterances containing will having recent or past time reference. For BerbeiraGardón (2006: 447), will has the meaning “will: p is true, and the world type is potential”. He explains that, when one utters a sentence containing will, “the speaker is, on the one hand, communicating that the proposition expressed is potential, while on the other hand it is true” Berbeira- Gardón (2006: 448). He also asserts that will encodes four modal meanings: dynamic, deontic, epistemic and interrogative, depending on the context. His position is similar to Essegbey’s (2008) position on the status of the a-morpheme in Ewe. The analyses we have seen so far have taken extreme sides of the argument; Tonhauser (2011) on the other hand suggests that the future marker-ta in Paraguayan Guaraní, a language spoken in Paraguay is in the middle. She argues
18
that –ta entails future time reference but it is also compatible with the modal meanings of intention and prediction. As a future time marker, it is used to talk about a time in “the future of the utterance time” (Tonhauser 2011: 5). The future time referred to could be near or distant; this is evident in the types of adverbs it is compatible with, for instance ko’áĝa ‘now’, koéstedía ‘today’. She asserts that from the data she gathered, there were no instances where –ta does not express future time reference. This morpheme is thus different from the Ewe a-, Akan bɛ and the English will. However, she argues that there are modal meanings such as intention and prediction/expectation that are compatible with –ta. She asserts that this is possible
because
intention
and
prediction/expectation
are
“intrinsically future oriented”. What is intended is not already known and what is predicted is yet to happen. The studies by Essegbey (2008), Osam (2008), Boadi (2008), Berbeira- Gardón (2006), and Tonhauser (2011) give a glimpse of the different analysis of future markers cross linguistically. Scholars like Essegbey (ibid) and Berbeira- Gardón (ibid) argue that what is referred to as future markers in Ewe and English respectively are modal markers. Osam (2008) and Boadi (2008) also claim that Akan has a future marker which has certain modal meanings when is used with some verbs, while Tonhauser (2011) argues that -ta in Paraguayan Guaraní is a future marker which is compatible with some modal markers.
2.1.3
Modal adverbs, adjectives and other ways of expressing modality
Although most early studies on modality focused on modal auxiliaries, there are now studies on the use of adverbs, adjectives, mental state predicates and other
19
linguistic expressions to express modality. Most researches on modality in Ghanaian languages are based on these modal expressions since most Ghanaian languages do not have modal auxiliaries. One of such studies is Amfo’s (2005) discussion of the use of the discourse connective anka as a modal expression in Akan. According to Amfo (2005) anka has four different functions in discourse but at the core of all these functions is the expression of non-factuality or the alternative possibilities for how things could be. The first function of anka is as a politeness marker when making a request. It is as well used to communicate someone’s unfulfilled intentions and desires; it is also used to express the existence of a state of affairs that belongs to the past and has ceased to exist. Finally, it is used to communicate that what is described did not actually happen but almost happened. Anka is an epistemic modal; it is used to express judgments about the factual status of a proposition. Amfo (2005: 1013) explains “anka encodes the assumption that the state of affairs described by the propositional content is not factual at the time of the utterance”. Working in the relevance theoretic framework, Amfo (2005: 1011) asserts that “anka has the higher-level implication; in relation to a speaker (S) and a proposition (P), anka conveys that in S’s opinion P is not factual”. Another of such studies is Gbegble’s (2012) survey of epistemic modal adverbs in Ewe. She identified four epistemic adverbs; ɖewo(ma)h ‘perhaps’, ‘maybe’ and tanyoa ‘perhaps’, ‘maybe’, godoo ‘certainly’, kokoko ‘certainly’. Apart from their function as modals, the certainty adverbs are also used as strengtheners; they enforce the proposition expressed by a sentence. They are used as ‘strengtheners’ in clause initially or when they are used together with other modal expressions. For ɖewo(ma)h and tanyoa, she explains that they express epistemic possibility, and are
20
more or less equivalent to ‘maybe’ or ‘perhaps’ in English. Unlike the certainty adverbs, neither of them has any other (non-epistemic) meanings or uses. Godoo and kokoko can appear in clause initial, internal or final positions; at clause final position, they usually have epistemic meaning. Dewo(ma)hi and tanyoa on the other hand occur exclusively in clause initial position or immediately after a clause connector if there is one in the sentence. Sakyi (2013) discusses the use of predicative adjectives as modal expressions in Akan. In her work, she notes, “there are only a few basic adjectives that express epistemic and evidential meanings in Akan discourse” (Sakyi 2013:51). Some of the modal adjectives she identified in Akan include nokware ‘true’, ampa ‘true and the evidential marker pefee ‘clear’. In most languages there is distinction between certainty and uncertainty modal adjectives. There appears to be no such distinction in Akan; there are no uncertainty epistemic adjectives. Concepts that are encoded by epistemic adverbs and adjectives in English are translated by verbs and other expressions in Akan. For instance the English adverbs possibly, allegedly and evidently is expressed by the expression Ebe-tumi aba, wo-se and biribiara kyerε ‘evidently’ respectively. Attitudinal verbs in Akan typically occur with the interpretive marker sɛ. Amfo (2007) argues that there are two different interpretive markers with the form sɛ; they occur are two different syntactic positions. One occurs at sentence initial position and other introduce a subordinate clause (COMP sɛ). Generally, initial sɛ when used, is an expression of a speaker’s opinion and also a request for consent. Comp sɛ on the other hand occur with certain attitude verbs such as pɛ ‘like’ and gyedi ‘believe’ to express certain attitudes towards proposition
21
expressed. It is also used with certain factive verbs like nim ‘know’ and pene ‘agree’ to show a speaker’s commitment to a proposition. There is also a form yɛsɛ that Amfo (2007: 109) explains is as result of a grammaticalization process whereby an original impersonal pronoun plus the verb say is now being used as a pragmatic marker, indicating speaker distancing. Yɛsɛ has three functions; the first one is to signal that the information introduced is not first hand. Used this way, it signals that the speaker does not want to commit to the proposition expressed and is distancing herself from it. It can also be used to echo opinions of others that the speaker does not necessary approve of. According to Amfo (2007: 113), when it is used this way, the speaker shows his/her disapproval of the situation. However, there may be occasions where the speaker may not necessarily show disapproval. For instance using example 7 of Amfo (2007:112) 1) Yɛse m’ani
n- sͻ adeɛ
ITM POSS’-eye NEG-catch thing ‘It is said that I’m disrespectful’ A speaker might endorse this assertion by adding another clause, which states that the assertion is true. The third use of yɛsɛ shows disapproval of a situation; when a speaker does not support a certain situation, she might report the situation using yɛsɛ. In Akan, mood is signaled by verbal affixes. Boadi (2008) distinguishes two mood contrasts: Jussive-Imperative and the Indicative. He asserts that “the Jussive-Imperative is used to express wishes, requests and commands, and the Indicative expresses facts and assertions” (Boadi 2008:10). The imperative in Akan is marked by the bare form of the verb if it is a direct request or command
22
but when it is a command given through another person, the nasal prefix N– is used. This prefix varies in form depending on the place of articulation of the consonant it precedes (homorganic nasal assimilation). When the command is direct, it is referred to as the Jussive mood and when it is indirect, it is called the Optative mood. The indicative mood is divided into two, the finite and the non-finite affixes. The close relationship between the TAM markers in Akan is evident in the fact that the non-finite mood affix –a does not express aspect or time, while the finite affixes express aspect and tense.
2.1.4
Modal concord
Most often than not, we find two modal expressions being used in the same construction. Geurts & Huitink (2006) proposed the name Modal Concord (MC) for this phenomenon. Two modals are used in each of the sentences below; the modals in example two are probably a modal adverb and may a modal auxiliary and the modals in 3 are may and certainly. 2) Probably Mary may leave 3)
There may certainly have been weapons of mass destruction in Iraq.
According to Zeijlstra (2008: 1) “Modal concord, hereafter (MC) is a phenomenon where two modal expressions do not yield a cumulative reading, but yield only one modal operator at Logical Form”. In other words, we talk about modal concord when we have a situation where the modal meanings of two modal expressions in a sentence seem to converge; the modals do not maintain their individual meanings. Geurts & Huitink (2006) argue that though all types of modal expressions can cooccur in a sentence, only modal adverbs and modal auxiliaries have an MC
23
interpretation. Zeijlstra (2008:3) concurs with them and adds that since MC is a grammatical phenomenon, “an expression consisting of two non-auxiliary modal elements of the same type cannot yield a Modal Concord reading”. As a grammatical phenomenon, MC obeys syntactic locality constraints such as strong islands” (Zeijlstra 2008:9). He adds that, “MC may be established by a modal verb/ adverb in the matrix clause and a modal auxiliary in the complement clause; however MC cannot be realized if one of the two modal elements is located in an adjunct” (Zeijlstra 2008:9). The fact that a modal auxiliary and a modal adverb cooccur in a sentence is not enough to have MC; the two modal expressions should be of the same type and have more or less the same quantificational force. If we take a look at the sentences above again, this means that only example 2 is an instance of MC. The two modals both have epistemic interpretations in this context and they both express probabilities. Even though the two modals in example 3 both have epistemic interpretations, a concord interpretation is not available because may express probability while certainly express certainty. So far all the literature and therefore conclusions on modal concord are based on Indo-European languages. From the literature presented above one thing that is obvious is the importance of modal auxiliaries in MC. This raises several questions. a. Is modal auxiliary a cross-linguistic phenomenon? b. If not, does it mean MC is not possible in languages that do not have modal auxiliaries? c. Do languages that do not have modal auxiliaries employ other strategies to encode MC? In section 5.4 of this thesis, these questions are explored as MC in Akan is discussed
24
as Akan has been argued to have no modal auxiliaries. 2.1.5
Studies on modal systems of languages
Cross-linguistically, there is almost always more than one strategy that is used to express modality in a language. Works that are reviewed in this section are works on African languages that focus on describing all the tools used in a language to express modality; the earlier reviews focused on particular strategies or expressions. These studies include Aboh and Nauze (2008) exploration of how modality is grammatically expressed in Gungbe, Ameka (2008) investigation of ways of expressing modality in Ewe, and Uchechukwu’s (2008) survey o f modality in Igbo According to Aboh and Nauze (2008), in Gungbe, a Gbe language of the Kwa family, grammatical modality is expressed with the modal auxiliaries sigan ‘can/may’ and ɖona ‘must/have to’ and mood marker ni. The mood marker ni is used to mark subjunctive, condition or time adjunct clause. It also has the meaning of weak deontic modality. The modal auxiliaries sigan and ɖona have both root and epistemic meanings. sigan is used to express participant-internal ability and epistemic possibility. It is used to signal that the participant has the ability to perform an activity or there is the likelihood that the proposition expressed will take place. ɖona on the other hand, expresses deontic modality; goal oriented necessity and also epistemic necessity. We now turn to Ameka’s (2008) analysis of the modal system of Ewe, another Gbe language. There are three main strategies that are employed to express modality in Ewe: the use of preverbal markers in the verbal cluster, periphrastic modal constructions and
25
utterance particles, (Ameka 2008). There are a number of markers that are attached to the verb to mark aspect or modality in Ewe, these markers together with the verb are what Ameka (2008) calls verbal cluster. In the Ewe verbal cluster, “there are three modal positions, two directional positions and a position for aspect; these elements occur in a linear order” (Ameka 2008: 136). The modal positions include modals like the potential –la, often written as –a, the certainty marker nyá, the xa ‘bother’ etc. There are also other modals like the counter-expectation marker kpᴐ ‘to see’, the ability and possibility marker té ŋú, ‘press body’, and the dare modal katse. Modality is also expressed by constructions of the form É- ‘3SG Expletive’ Verb bé ‘COMP’ clause, this is what he refers to as periphrastic modals. This modal strategy is used to express obligation, necessity and evaluation. About this construction Ameka (2008: 147) explains: The construction is made up of the expletive (or cataphoric) 3SG pronoun as subject followed by a verb (chosen from a semantic class) followed by the quotative or complement clause
introducer,
bé followed by a clause expressing the
proposition that must happen or is being evaluated. There is a similar structure in Akan to express similar modal meanings. In Akan, the schema can be represented as ɛ -‘3SG Expletive’ Verb sɛ ‘COMP’ clause. The last strategy he discusses is the use of utterance particles like the ignorative marker à and agreement seeking marker ɖé to express propositional and speaker attitudes. Quoting (Ansre 1966, Duthie 1996:51), Ameka (2008:148) explain further that these markers “are used to signal, attenuate or boost the illocutionary forces of the utterances in which they occur. Thus they convey various attitudes of the speaker
26
towards what is being communicated, i.e. the propositional content, or towards the elements in the communicative context such as towards the addressee or towards other observers.” Though Akan does not have an elaborate verbal cluster with modal positions like Ewe, there are some similarities that are shared by both languages. This makes Ameka’s study a vital background for the present study. There are three morphosemantic ways of expressing modality in Igbo: “(1) modal verbs, (2) modal affixes (mood markers) and (3) peripheral modals” (Uchechukwu 2008:276). According to Uchechukwu (2008), of all these, mood is the most studied aspect of modality in Igbo. In his study therefore, he focuses on the use of verb morphology related structures as single root verbs like –tosi ‘should/ought to’, verb+NP structures like -nwe ike ‘can’, and verbal suffixes likenwu ‘able’ and -ri.ri ‘compulsion’ to express modality. He looks at the modal flavor and the marker used to express it. The modal meaning of compulsion and possibility are expressed by the verb -ga and the suffix -ri.ri respectively. Uchechukwu (2008:253) states “morpho-tonologically –ga has two forms, the implicit form and the explicit form”. The implicit form is am bi guous between a deonti c m eaning of (±) weak compulsion and a epistemic meanings of possibility and speculative prediction. The explicit form only receives modal interpretation when it is combined with the affixes -sị´ and -siri, -nátá and -rí.rí. or -lílí. “These suffixes have the deontic meaning of compulsion and strong obligation that has the force of law, and the epistemic meaning of possibility and deductive or speculative prediction” Uchechukwu (2008:254). The verbs -nwé íké ‘can’ and -pụ` ‘can, be able’, and the suffixes -nwu ‘able’ and –lí ‘able’ are used to express ability and possibility.
There are also the modal meanings
27
of appropriateness and possibility and these are expressed by the verbs -kwesi ‘be appropriate/suitable’ and -tosi ‘be appropriate/suitable’. Both verbs can be used to express the deontic meanings of obligation as well as the epistemic meaning of possibility. Uchechukwu (2008) does not however describe adverbs, mood markers and other periphrastic modal expressions because they have been discussed in earlier works.
2.1.6
Non-Modal meanings
Most of the literature reviewed from the beginning of the chapter has focused on modal strategies languages employ to express modality and also the inherent meanings these expressions have. In most languages modal expressions have more than one function. The studies in this section look at the other uses of modal expressions. One of the discussed non-core modality usage of modal expressions is its use as a politeness marker. Modal expressions are used to mitigate certain utterances a speaker makes. Lapeyre (1992) observes instances where the use of modal auxiliaries like may, will and must are only due to politeness and not the expression of lack of absolute knowledge. She notes that when used this way,
these
modal communicate a wide array of assumptions, therefore
conveying vagueness in utterances in which they occur. Boicu (2007) observes the role of modal verbs in one of Ashley Mote’s (a British independent nationalist politician) political speeches. The analysis focuses on the two main semantic values of modal verbs in English, the epistemic and the deontic. Modal verbs are used to mitigate or aggravate the illocutionary forces released by the speech acts they belong
28
to. As an extremist nationalist ideologist, with a more radical and illocutionarily forceful discourse, there more occurrences of absolute modals on both semantic and pragmatic scales: deontic and epistemic. The unambiguous mitigating or aggravating force of the modal verbs in Mote’s speech is noticed in the use of the absolute prescriptive must, the strongest on the deontic scale, when used to express obligation or prohibition. He also uses softer mitigating modals like should and need to, with deontic semantic values and uses these more than other modals. He enhances the mitigation to which the modal verbs contribute by the use of personal deixis. Mote uses plural forms as grammar subjects, mainly ‘inclusive we’, a typical legitimization strategy in political discourse, or impersonal phrases, but, in this case, corelated with soft modal verbs, it erodes his self-image. Vázquez and Giner (2008) using academic research works as their data also argue that some modal expressions are used as hedging devices in English. This implies that there is a relationship between modality and hedging. They assert that the meaning of hedging and modality seems to overlap in certain instances. They argue that in English, epistemic modals are used as hedging devices. For instance epistemic modals like may, probably and perhaps are used to hedge in most academic works. The example below is from their paper illustrating the use of modal expressions as hedging devices. 4) By analogy, it may be possible to walk from one point in hilly country to another by a path, which is
always level or uphill, and yet a
straight line between the points would cross a valley. These works provide a background to the current research because one of the objectives of the research is to investigate if Akan modal markers have other uses
29
in discourse.
2.2
Theoretical Framework: Relevance Theory
The study employs Sperber and Wilson’s theory of communication, ‘Relevance Theory’, as its theoretical framework, focusing on the conceptual/procedural distinction made by Blakemore (1987). Relevance Theory has developed in several stages. A detailed version is published in Relevance: Communication and Cognition (Sperber & Wilson 1986a; 1987a, b) and updated in Sperber and Wilson 1995, 1998a, 2002; Wilson and Sperber 2002. Relevance theory is a theory of pragmatics that argues that inference is critical to virtually all instances of communication. As a theory of communication, relevance theory contrasts with the classical code model where information is given as a set of codes, w h i c h have to be decoded by another party who has another copy of the code. It is preceded by Grice’s (1957, 1968) inferential model of communication. According to Grice, during communication speakers try to achieve certain principles and hearers evaluate the speaker’s intentions based on these principles. Grice conventionalize the ‘co-operative principle’ and ‘maxims of conversation’. These maxims are, Quality (truthfulness), Quantity (infromatives), Relation (relevance) and Manner (clarity) which speakers are expected to observe during communication. Relevance theorists take a different position from Grice and claim that during communication speakers do not try to adhere to certain principles; they rather tend to pay attention to the most relevant phenomena available. This claim is the main argument of the theory and it is summarize in the cognitive principle of relevance
30
“human cognition tends to be geared to the maximization of relevance” (Sperber & Wilson 1995: 260). The theory thus expounds that communication is a process of sorting through the available inputs in communication to find the one that is most relevant. Scholars of Relevance theory argue that what causes an input to stand out from others is its relevance to the receiver. The central claim of relevance theory is that the expectations of relevance raised by an utterance are precise enough, and predictable enough, to guide the hearer towards the speaker’s meaning. The main aim of the theory is to explain in cognitively realistic terms what these expectations of relevance amount to, and how they might contribute to an empirically plausible account of comprehension. According to relevance-theorists, an input is relevant to an individual when its processing in a context of available assumptions yields a positive cognitive effect. They explain a positive cognitive effect as a worthwhile difference to the individual’s representation of the world. In relevance-theoretic terms, other things being equal, the greater the positive cognitive effects achieved by processing an input, the greater its relevance will be.
Also, other things being equal, the greater the
processing effort required, the less relevant the input will be. Thus, relevance may be evaluated in terms of cognitive effects and processing effort. In other words, when one makes an utterance, he is asserting that what he has said is worth listening to, which means that the ‘cognitive effects’ it provides are worth the processing effort required to find the meaning. According to relevance theorists, it is not enough for a stimulus to attract the attention of the hearer; the hearer will have to recognize that speaker has this intention. This they claim is done by the use of an ostensive-stimulus, designed to
31
attract an audience’s attention and focus on the communicator’s meaning. The use of an ostensive stimulus may create precise and predictable expectations of relevance not raised by other stimuli. This is what leads to the ‘communicative principle of relevance’, which states that “Every ostensive stimulus conveys a presumption of its own optimal relevance” (Sperber and Wilson, 1995:260, Wilson & Sperber, 2004:612). Sperber and Wilson (2004) further claim that information is relevant to you if it interacts in a certain way with your existing assumptions about the world. To them, information can be relevant in three ways:
a. New information is relevant in any context in which it has contextual implications, and the more contextual implications it has, the more relevant it will be. b. New information is relevant in any context in which it strengthens an existing assumption; and the more assumptions it strengthens, the greater its relevance will be. c. New information can achieve relevance by contradicting and eliminating an existing assumption. 2.2.1 Procedural versus Conceptual meaning There are certain words and expressions that provide information as to how, or in which direction, the interpretation of an utterance is expected to proceed. Blakemore (1987) claims these expressions are said to
encode procedural
information. According to Blakemore, procedural expressions “do not encode a constituent of a conceptual representation (or even indicate a concept), but guide
32
the comprehension process so that the hearer ends up with a conceptual representation” (Blakemore 2002: 90-91). Procedural encodings guide the audience
towards
an
inferential
route
that
results
in
a
conceptual
representation. Wilson (2012: 2) distinguishes between the two as “Conceptual items encode stable concepts, i.e. constituents of a ‘language of thought’, on the other hand, Procedural items do not encode concepts because they (a) never affect truth conditions (eg. but), or (b) affect truth conditions only indirectly (eg. he) or (c) are purely expressive (egs. alas, intonation)”. Procedural information involves the use of certain linguistic items to constrain the context in a way that aids the interlocutor in his inferential processing. Words and expressions which are said to encode procedural information contribute to the interpretation process by encouraging the interlocutor to look for certain types of inference relations which further give rise to contextual effects. In other words, these linguistic expressions indicate to the interlocutor how the linguistically encoded concepts of an utterance are expected to be interpreted such that maximum contextual effects are yielded for minimum processing effort. As Carston (2002) puts it, words and expressions of this nature “indicate, guide, constrain or direct the inferential phase of comprehension”. Procedural markers include pronouns, connectives and adverbials. This study analyzes modal expressions in Akan as procedural markers that are used to constrain the context in a way that aids the interlocutor in his inferential processing. These procedural markers are used to direct an addressee to the attitude a speaker has towards a certain proposition. For instance when we look at the sentences below, the proposition of the lower clauses is all the same,
33
however the interpretation received is different. 5) Me-nim sɛ Kofi pɛ m-maa 1SG know COMP K like PL-woman ‘I know Kofi likes girls’ 6) Me-dwene sɛ Kofi pɛ m-maa 1SG think COMP K like PL-woman ‘I think Kofi likes girls’ The verbs in the higher clauses are the source of the different interpretation; they suggest different speaker attitudes to the proposition being expressed in the lower clause. In example (5), the speaker expresses certainty typically based on the fact that he has enough facts to support his claim. In example (6), the proposition in the lower clause is presented as a thought of the speaker. This allows the speaker to distance him/herself from the proposition. This supports Wilson’s (2012:2) argument that “some procedural markers encode procedural constraints on speech-act, propositional-attitude or affective-attitude information (analyzed in terms of higher-order explicatures rather than implicatures)”. 2.3 Methods An integral part of every research is the data collection process. In this section of the thesis, the methods that were used in the collection and analysis of data are discussed. 2.3.1
Sources and Methods of Data Collection
Data for the study were collected from both primary and secondary sources. This study draws data primarily from two programs on two local radio stations; Peace FM and Adom FM. Peace FM is an Accra-based radio station, which broadcast most of its programs in Akan. It was established on May 25 1999 and i s the first
34
Akan-speaking radio station in Accra. It is located at Abeka, a suburb of Accra. It operates on the frequency 104.3. Data for the thesis was collected from their morning show Kokrokoo. Kokrokoo is an onomatopoeic word; it is the sound of a cockerel crowing. In the villages, the crowing of the cockerel is seen as the start of the day, therefore the morning show signifies the start of the day. The program starts as early as 6:30 am and ends at 10:00 am. The host of the program is Kwame Sefa Kayi, also affectionately called ‘Chairman General’ by the participants of the show. The show starts with a newspaper review, sports and then a panel discussion of some of the topical issues in the newspapers and other issues affecting Ghanaians. During these morning shows, the general public is also invited to call in and give their comments. The second radio station Adom Fm is also a privately owned radio station based in Tema, a municipality in the Greater Accra region. It is owned and run by the Multimedia Group Limited. The morning show of Adom FM is Dwaso Nsem ‘Matters in the market’. It is also a platform where panelists are invited to the studio to comment on national issues. During these morning shows, the general public is also invited to call in and give their comments. Adom FM also broadcast primarily in Akan. All the recordings were done during the live broadcasts of these events. Later, the data were transcribed manually. I listen carefully
to
the
recordings, purposively sample them and transcribe them accurately. Secondary sources of data collection came from literary texts and other research works in Akan. The literary texts were two prose, Forosie by Edwin Efa, and Bediako by Victor Amarteifio. These literary works are important sources of data because they mirror actual language usage in the Akan community. The data
35
from the secondary sources were taken together within context so as to allow for a correct interpretation in the analyses.
2.3.2
Method of data analysis
After the data had been transcribed, the data was purposively sampled for the analysis. Through purposive sampling, data that best answer the research questions are selected. The main aim of purposive sampling as a method of data analysis, according to Cresswell (1994:148), is to “purposefully select respondents or documents that will best answer the research question posed”. I therefore purposively sampled the recordings of Krokrokoo on Peace FM, and Dwaso Nsɛm on Adom that are most useful to the study. Afterwards a focus group was set up to help identify the modal expressions used in the data and also help identify the modal meanings encoded by these markers. Two focus groups were formed. For the first group, since the area of research is rather technical, people who had some knowledge about linguistics were used. This group thus was made up of eight level 400 students and two national service personnel in the Ghanaian Languages and Linguistics Department of the University of Cape Coast, Cape Coast. The second group consisted of ten native speakers who were semi literates. These two groups became necessary when it was realized that English and previous knowledge was interfering with the judgment of the first group. The first group therefore mostly helped in the identification of the modals.
36
2.4
Chapter Summary
In this chapter, I have discussed some related literature to the study, looking at modal auxiliaries, modality, tense and aspect, other modal strategies, the future and modality and the pragmatics of modals. I have also discussed Sperber and Wilson’s Relevance Theory, which is the theoretical framework in which the study is couched. I explained the theory by using some of the data gathered. A detailed account of the methods employed in the collection of data for this study has also been discussed
37
CHAPTER THREE THE LEXICAL MEANS OF EXPRESSING MODALITY IN AKAN
3.0
Overview
In this chapter, lexical items that are employed to express modality in Akan are discussed. According to Carnie (2007:43) “lexical parts of speech provide the “content” of the sentence and these are nouns, verbs, adjectives and adverbs”. This is contrasted with functional parts of speech, which he argues, “provide the grammatical information”. Payne (1997:37) explains, “a lexical category is a syntactic category for elements that are part of the lexicon of a language”. Carnie (2006:43) adds that they are the words that are left in a “telegraphic speech”. Words in the lexical category can be divided into major and minor categories. The major categories include verbs, nouns, adverbs and adjectives. Minor category words include adpositions, pronouns and conjunctions. In this section, we will explore how Akans use two of the lexical categories: verbs and adverbs, to express modality. The discussion on verbs will start with the mental state verbs nim ‘to know’, susu ‘to suppose’ or ‘to suggest’, dwene ‘to think’, gye di ‘to believe’. The discussion will then turn to the use of modal adverbs in expressing modality. As already discussed in section 2.2.3, this study analyzes modal expressions in Akan as procedural markers that are used to constrain the context in a way that aids
38
the interlocutor in his inferential processing. According
to Blakemore,
procedural expressions “do not encode a constituent of a conceptual representation (or even indicate a concept), but guide the comprehension process so that the hearer ends up with a conceptual representation” (Blakemore 2002: 90-91). The modal markers, as procedural markers, are used to direct an addressee to the attitude a speaker has towards a certain proposition expressed. The attitudes range from total commitment to the proposition (certainty) to distancing from the proposition (uncertainty). 3.2
Verbs that express modality in Akan
One of the devices for expressing modality is the use of mental state predicates (Narog 2009). Mental state predicates are verbs that have meanings related to understanding, discovering, planning or deciding. They usually refer to cognitive states that are generally unavailable for outside evaluation. Mental state verbs in Akan include the factive nim ‘to know’ verb and attitude verbs, susu ‘to assume’ or ‘to suggest’, gye di ‘to believe’ and dwene ‘to think. These verbs are used to express attitudes towards proposition based on how much evidence is available to the speaker at the time of speaking. As asserted by Amfo (2007), some mental state verbs in Akan when combined with the interpretive use marker sɛ can be used to show the attitude of a speaker towards a proposition expressed. The mental state verbs are also used in what Thompson and Mulac (1991) call modal tags. As verbs, they occur in both simple sentences and complex sentences. It is only when they occur in complex sentences with the complementizer or interpretative marker sɛ that they have modalized interpretations.
39
This study is thus only concerned with the use of mental-state verbs used in the main clause of the complex sentence as in examples (2) and (3). 1. Kofi nim me K know me ‘Kofi knows me.’ 2. Kofi nim sɛ mehu- u no. Kofi know COMPS 1SG-SUBJ see- PST 3SG-OBJ ‘Kofi knows I saw him.’ 3. Kofi nim sɛ menim no. Kofi know COMPS 1SG-SUBJ know 3SG-OBJ ‘Kofi knows I know him.’ The main reason as asserted by Amfo (2007) is that when mental state verbs occur in the higher clause they have higher-level implicature, which is associated with the expression of an attitude towards a proposition. When these verbs occur in the lower clause, they lead to base level implicatures. 3.2.1
Nim ‘to know’
Nim is a factive verb and thus communicates to the addressee that the speaker has enough information to make the statement or assertion he is making. It has the epistemic meaning of strong certainty. The interpretation of nim is not as strong and the English gloss know. The source of evidence for a sentence containing nim need not be real; the evidence may be assumed or false. What is important is that the speaker believes that what he/she is saying is true at the time of speaking. The speaker thus commits himself/herself to the truth condition of the proposition being expressed. When nim is used in the main clause, the verbs that occur in the complement clause can take any of the tense/aspect markers in Akan. In the examples below, the various tense/aspect markers are attached to the verbs in the
40
lower clause. The past tense -a marker is attached to da ‘sleep’ in example (4), in example (5) the perfective aspect marker a- is attached to ba ‘to come’. The future marker bɛ and the progressive re are attached to the verb ba ‘to come’ in examples (7) and (8) respectively. In example (9), there is no affix attached to the verb to ‘to sing’, to mark the habitual aspect. This is because the habitual is marked by tone in Akan. 4. Me- nim sɛ Kofi da -a hͻ 1SG- know COMPS K sleep - PST there ‘I know Kofi slept there.’ 5. Me- nim sɛ Kofi aba fie. 1SG- know COMPS K PERF- come home ‘I know Kofi has come home.’ 6. Me- nim sɛ Kofi 1SG- know COMPS K ‘I know Kofi is sitting there.’
te hͻ. sit.stative there
7. Me- nim sɛ Kofi bɛ- ba fie. 1SG- know COMPS K will- come home ‘I know Kofi will come home.’ 8. Me- nim sɛ Kofi reba fie. 1SG- know COMPS K PROG- come home ‘I know Kofi is coming home.’ 9. Me- nim sɛ Kofi 1SG- know COMPS K ‘I know Kofi sings.’
to dwom. sing.habitual song
10. *Me-re- nim sɛ Kofi bɛ- ba fie 1SG- prog-know COMPS K will- come home 11. *Me- a- nim sɛ Kofi bɛ- ba fie 1SG- perf.-know COMPS K will- come home
41
As a stative verb nim does not inflect for tense/aspect. Sentences 10 and 11 are thus not felicitous in Akan. One thing worth noting is that though the perfective aspect is used to talk about actions that have already been completed, in some instances it can be used to describe actions, which are yet to occur. In (12) below, the football match that is being talked about is yet to take place. 12. Menim sɛ Ghana ahyɛ Congo a- wie 1SG-subj- know COMPS Ghana PERF-score Congo PERF-finish ‘I know Ghana has already beaten Congo’. When nim and the perfective aspect is combined in this way, the speaker communicates that he believes beyond reasonable doubt that the action described in the lower clause will take place. 13. Menim sɛ me dimafo te ase. 1SG-subj- know COMPS 1SG-POSS redeemer stay under ‘I know my redeemer lives.’ 14. NDCfoͻ deɛ, yɛn nyinaa nim sɛ wͻmmo NDCpeople TM, 1PL-SUBJ all know COMPS 3PL-SUBJ n- nwene Ghanafoͻ ho. NEG- think Ghanaians skin ‘As for the NDC, we all know they do not think about Ghanaians.’ To interpret the sentences above, the addressee assumes that the speaker has enough evidence to support his claim. In (13), the higher level implicature Me-nim expresses certainty to the proposition in the lower clause containing the base level implicature me dimafo te ase. In example (14), the base level implicature wͻmmo nnwene Ghanafoͻ ho is submerged in the higher level implicature yen nyinaa nim. The speaker expresses an attitude of certainty to the proposition expresses in the higher clause
42
3.2.2
Gye di ‘to believe’
The concept of believing something or believing in something is expressed in Akan by two verbs, gye ‘collect’ di ‘eat’. It is a serial verb construction; specifically an example of what Osam (2003) calls Integrated Serial Verb Construction (ISVC). This means that the events encoded by the verbs are tightly integrated and thus cannot be separated into constituent parts. Gye di communicates to the addressee that the speaker has enough information to make the statement or assertion he is making. It has the epistemic meaning of certainty. When a speaker uses gye di, s/he commits him/herself to the truth condition of the proposition expressed in the complement clause. However, the use of gye di posits a lower or weaker commitment than nim. 15. Megye di sɛ yɛka 1SG-subj- receive eat COMPS 1PL-SUBJ- touch a ɛyɛ. COND IMP-SUBJ- good ‘I believe that it is good for us to unite.’
bo-m hit in
16. Yɛ nyinaa gye di sɛ nnea ͻreka 1PL-subj all receive eat COMPS what 3SG- PROG. say no wͻ mu. CFM to be located in ‘We all believe that what he is saying is true.’ In sentence (15) and (16) above, the base level implicature is submerged under the higher- level implicature expressed by gye di. By communicating sentence (15), the speaker signals that he is committed to the proposition sɛ yɛkabom a ɛyɛ. The choice of gye di over nim indicates strongly that the speaker intend a weaker commitment towards the proposition he expresses.
43
3.2.3 Susu ‘to assume’ or ‘to suggest’ The verb susu is ambiguous between the English gloss ‘to assume’ and ‘to suggest’. Below are examples of some sentences in which they are used. The intended meaning is usually made prominent by the context of use. 17. Me- susu sɛ yɛn- kͻ fie. 1SG- suggest COMPS 1PL- go home ‘I suggest we go home.’ 18. Me- susu sɛ Kofi ada. 1SG- assume COMPS K PERF- sleep ‘I assume Kofi is asleep.’ 19. Kwame susu sɛ yɛn- pa ne kyew. K suggest COMPS 1PL- take.off 3SG-POSS hat ‘Kwame suggests we beg him.’
In example (17), the verb means ‘to suggest’, here the referent of the pronoun me suggests to the addressee that they should leave. In (18), the referent of the pronoun me assumes that the proposition in the lower clause is true or has taken place. When susu is used as to mean ‘to suggest’, the speaker implies that the interlocutor could perform the action described in the lower clause. The verb communicates that whatever is in the lower clause is only a suggestion the speaker is putting forward. The addressee is therefore not in any way obligated to take it. In example (18) above, the speaker implies that the interlocutor should go home with him. In example (19), the proposition in the lower clause is seen as a suggestion put forward by Kwame. Thus the referents of yen are not obligated to accept the proposition put forward. The speakers do not impose their opinion on the
44
addressee but leave the final decision making to the addressee. When susu ‘to suggest’ is used in the main clause, the verbs that occur in the complement clause can only be in the imperative, as seen in examples (20) to (23) below. The verbs duru ‘to reach’ and pa kyew ‘to beg’ are in the imperative. But as shown in examples (21), (22) and (23), the verb susu ‘to suggest’ can inflect for past, future and progressive.
20. Me deɛ me- susu sɛ ɛ- duru mmere a, 1SG Top. 1SG- suggest COMP 3SG- reach time REL. tumi gyae saa nneɛma yi. be. able.to stop these things this. ‘As for me, I suggest it is high time we stop such things.’ (Peace FM 15/04/14)
yɛ 2PL
21. Kwame susu -u sɛ yɛn- pa ne kyew. K suggest- PST COMP 1PL- take.off 3SG-POSS hat ‘Kwame suggested that we beg him.’ 22. Kwame bɛ- susu sɛ yɛn- pa ne kyew. K will.- suggest COMP 1PL- take.off 3SG-POSS hat ‘Kwame will suggest we beg him.’ 23. Kwame resusu sɛ yɛn- pa ne kyew. K PROG.- suggest COMP 1PL- take.off 3SG-POSS hat ‘Kwame is suggesting that we beg him.’ Susu ‘to assume’ on the other hand communicates to the addressee that the speaker may not have enough evidence to support his claim about the proposition in the lower clause. Susu communicates a non-committal attitude to the base level implicature. However, what is relevant is that the speaker believes there is a probability that the proposition might be true at the time of speaking.
45
24. Amanfoͻ susu sɛ light adumdum-adumdum no People assume COMP light off DET yɛ president fault. be president fault ‘People assume that the frequent power outage is the fault of the president.’ 25. Mesusu sɛ Kofi da dan 1SG-subj- suppose COMP K sleep room ‘I assume Kofi is sleeping in the room.’
no mu DET inside
In ( 25), the base level implicature expressed in the lower clause is submerged under the higher level implicature Mesusu. This means that the speaker believes to an extent that Kofi is sleeping in the room; however he does not wish to commit himself to the proposition. It leaves room for the speaker to be right or wrong about the proposition. When susu ‘to assume’ is used in the main clause, the verbs that occur in the complement clause can take any of the tense/aspect markers in Akan. In example (26) the past tense marker -a is attached to the verb da and in example (27) to (30), the perfect, future, progressive and habitual markers prefix the verb ba ‘to come’. 26. Me- susu sɛ Kofi da -a hͻ. 1SG- assume COMP K sleep - PST there ‘I assumed Kofi slept there.’ 27. Me- susu sɛ Kofi te hͻ. 1SG- assume COMP K sit. CONT there ‘I assumed Kofi is sitting there.’ 28. Me-susu sɛ Kofi aba fie. 1SG-assume COMP K PERF- come home ‘I assumed Kofi has come home.’
46
29. Me- susu sɛ Kofi bɛ- ba fie. 1SG- assume COMP K will- come home ‘I assume Kofi will come home.’ 30. Me- susu sɛ Kofi reba fie. 1SG-assume COMP K PROG- come home ‘I assume Kofi is coming home.’ 31. Me- susu sɛ 1SG- assume COMP ‘I assume Kofi sings.’
Kofi to dwom. K sing.habitual song
Unlike when it means ‘to suggest’, susu ‘to assume’ does not inflect for tense or aspect. As can deduce from the discussion above, it is the interpretation ‘to assume’ that is modal. It communicates modal uncertainty; the probability or possibility of a proposition being true.
3.2.4
Dwene ‘to think’
Another attitude verb is dwene. Dwene communicates to the interlocutor that the proposition in the lower clause is a thought that is being expressed by the speaker.
32. Daa nyinaa na Forosie dwene sɛ ͻno na daakye Day all PST F. think COMP 3SG FOC future ͻbɛ- di adeɛ 3SG-SUBJ – will inherit thing ‘All along Forosie thought he was going to be the next king.’ (Forosie pp10) 33. Na afei, Bediako dwene sɛ nipa biara ndͻ CONJ then,,B think COMP person all NEG- love no mpo. 3SG-OBJ even ‘Bediako thought that nobody liked him.’(Bediako pp89)
47
In example (32), ͻno na daakye ͻbɛdi adeɛ is presented as a thought of Forosie and in (33), nipa biara ndͻ no mpo is presented as a thought of Bediako. When dwene ‘to think’ is used in the main clause, the verbs that occur in the complement clause can take any of the tense/aspect markers in Akan. The verb in example (34) is in the past tense and the verb in (35) has the perfect aspect. The verb in (36) is in the stative aspect, the verb in (37) is prefixed by the socalled future tense marker and the verb in (38) is prefixed by the progressive aspect marker. 34. Me- dwene sɛ Kofi 1SG- think COMP K ‘I thought Kofi slept there.’
da -a hͻ. sleep - PST there
35. Me- dwene sɛ Kofi aba fie. 1SG- think COMP K PERF- come home ‘I thought Kofi has come home.’ 36. Me- dwene sɛ Kofi te hͻ 1SG- think COMP K sit.stative there ‘I thought Kofi is sitting there. 37. Me- dwene sɛ Kofi bɛ- ba fie. 1SG- think COMP K will-come home ‘I thought Kofi will come home.’ 38. Me- dwene sɛ Kofi reba fie. 1SG- think COMP K PROG- come home ‘I thought Kofi is coming home.’ The verb dwene ‘to think’ is compatible with the so-called future tense. When speakers want to indicate that their thoughts have changed, they begin the sentence with na. Example (40) also expresses the fact that the referent of the pronoun me no longer thinks that the clause Kofi daa hͻ is true. These are shown in the examples below
48
39. Kofi bɛ- dwene sɛ ɛhͻ na Kwame da. K will- think COMP there FOC Kwame sleep ‘Kofi will think that is where Kwame sleep’. 40. Na me- dwene sɛ Kofi da -a hͻ. PST 1SG- think COMP K sleep - PST there ‘I thought Kofi slept there.’ Another point to note about dwene is the fact that it communicates to the addressee that the proposition in the lower clause is not true at the time of speaking. Therefore if we take example (32), repeated here as (41), what is being communicated is that ͻno na daakye ͻbɛdi adeɛ is false at the time of speaking. Also in (42), Kofi kͻ fie is false at the time this assertion is made. The sentence therefore means that Kofi has not gone home yet. 41. Daa nyinaa na Forosie dwene sɛ ͻno na daakye Day all PST F. think COMP 3SG FOC future ͻbɛ- di adeɛ 3SG-SUBJ – will inherit thing ‘All along Forosie thought he was going to be the next king.’ (Forosie pp10) 42. Me- dwene sɛ Kofi kͻ fie. 1SG- think COMP K go home ‘I thought Kofi went home.’ Usually when a speaker wants to state that the proposition in the second clause is true, they will use susu or firi. Firi/fili is a borrowed a n d nativised rendition of the English verb feel. Only the cognitive or mental state sense of the English verb is borrowed and used this way. The tactile sense of the verb is borrowed as feeli. This sentence can also have an interpretation that is similar to the use of dwene. In that instance, the sentence will mean ‘I thought Kofi has one home’.
49
Thus example (43) means that Kofi has actually left the premises. 43. Me- firi sɛ Kofi kͻ fie. 1SG- think COMP K go home ‘I feel Kofi has gone home.’
3.3.5
Summary of discussion on mental state verbs
From the discussions above, we notice that the verbs when used in the higher clause of a complex sentence have higher level implicatures. They express certain attitudes towards the proposition expressed in the lower clause. These attitudes can be ranked on the basis of the speaker’s commitment to the proposition expressed, starting from verbs that show highest commitment. This is shown in table 1 below.
Verbs
3.3
Table 1. Mental Verbs Attitudes Hierarchy Speakers’ attitude
Nim
Certainty
Gye di
Certainty (weaker than Nim)
Susu ‘to assume’
Probability
Dwene
Lower probability
Susu ‘to suggest’
Suggestion
Adverbs
The second lexical item to be discussed is the class of adverbs. The class of adverbs, according to Givon (1993:71), is the lexical category “whose membership is least homogenous and the hardest to define”. Trask (1993:9) defines an adverb as “a
50
lexical category, or a member of this category, whose members are usually grammatical adjuncts of a verb and most typically express such semantic notions such as time, manner etc.” Carnie (2007) classifies words into word classes based on their morphology, which is whether they take derivational or inflectional affixes and their syntactic distribution. He defines an adverb in English as a word that takes the derivational suffix -ly and does not take inflectional affixes. Syntactical, they can appear almost anywhere else in the sentence except between a determiner and a noun (*the quietly boy). Adverbs are the class of words that modify verbs, adjectives and other adverbs. They are the head of the adverb phrase (AdvP). They can also stand alone as an AdvP as in, ‘quickly’ or be modified by another adverb (a degree word) as in ‘very quickly’. Syntactically, adverbs occur after the verbs they modify in Akan as seen in the examples below. 44. Kofi kasa kese. Kofi talk Adv. ‘Kofi talks loudly. 45. *Kofi keseԑ kasa Kofi loudly Adv. 46. ᴐ- nante ntԑmtԑmtԑm. 2SG-walks quickly. ‘He walks very quickly.’
Example (45) is ungrammatical because the verb comes after the adverb. Adverbs in Akan are reduplicated to show intensity. In example (46), the reduplication of the stem ntԑm ‘quickly’ shows the intensity in which he walked. Saah (2004) identifies six types of adverbs in Akan: manner, pace, temporal, frequency, epistemic/ speaker oriented and aspectual. Epistemic/ speaker oriented
51
adverbs are the focus of this study. In Akan, Saah (2004) following Givon (1993) and Cinque (1998) explains that epistemic adverbs “convey the truth, certainty or probability of the proposition, or they express the speakers’ degree of confidence about the truth proposition. Ernst (2000: 96) adds that epistemic adverbs have the meaning, “speaker judges likelihood that P is true, based on how the real-world situation or source of knowledge, affects the likelihood that
the
corresponding
event
occurred”. Epistemic adverbs include ampa
‘truly/honestly’, nokware ‘truly/honestly’, ebia ‘maybe’,
gyama
‘perhaps’,
anhwe a ‘maybe’, sɛsɛɛ ‘perhaps/probably’ and annyɛbiara/gyengyenbiara ‘perhaps’. In the discussion, these adverbs will be grouped into two, adverbs that mark certainty and adverbs that show uncertainty or probability’.
3.3.1
Certainty adverbs: Ampa and Nokware
There are two certainty adverbs in Akan, ampa ‘truly or honestly’ and nokware ‘truly or honestly’. The use of ampa and nokware suggests that the speaker is certain that the proposition, which is in the scope of the adverb, is true at the time of speaking. As words encoding procedural meaning, the adverb nokware or ampa signals to the addresser, the speaker is certain about the state of affairs described in the proposition stated. As shown in Sakyi (2013), examples 49 and 50, these two words can also be used as adjectives in Akan where they mean true. This is seen in the examples below. 47. Nokware, nsɛm a yɛ te no bi yɛ hu. Honestly, matter REL. 3PL hear DET. QUANT. COP. strange ‘Honestly, some of the things we hear are strange.’
52
48. Bediako ho kyere- e no ampa. Bediako body catch- PST 3SG-OBJ truly ‘Bediako become truly impoverished (Bediako pp 44)
49. ε- yε nokware sε ͻ- boro -o abofra no It be true that 3SG-beat- COMPL child DET ‘It is true that he beat the child.’ (Sakyi 2013:49) 50. ε-yε ampa sε wͻn nyinaa aba. It-be true that they all PERF-come It is true that all of them have come.’ (Sakyi 2013:49) As adverbs, they are optional elements that are used to modify a clause. When they occur in a clause they have scope over the whole clause. These two adverbs are synonymous; therefore, the choice to use one instead of the other seems to depend on a particular speaker’s preference. Apart from being attitude markers, the Akan certainty adverbs can also be used as strengtheners. As strengtheners, they are used to place emphasis on the proposition a sentence expressed. They are also used to express the fact that the idea expressed by the proposition is extraordinary or above expectations. Gbegble (2012) is also discusses characteristics the Ewe certainty modals.4 51. Ampa ara
mmarima no ko
kyini -i
bɛyɛ
4
Gbegble (2012) argues that the Ewe certainty modal góódó and kó k ó kó act as strengtheners when they occur in a clause initial position and with other modal markers. 1. góódó la, wo- na- a nu- ɖuɖu-m zi etɔ gbe ɖe-ka. certainly TOP, 3PL-give-HAB thing-RED. eat 1SG time three day one Surely/I assure you that they give me food three times a day.’ (AME 3)
53
Truly EMP.PART. men Def MOT. roam-PST about nnasa kosi -i sɛ wͻ- kͻ- hu –u three.days MOT. reach PST COMP 3PL- MOT.see-PST. ͻbarima bi a ahia no man certain COND. PERF- poverty.stricken 3SG-OBJ ampa truly
‘Truly, the men searched for three days until they found a man who was truly poor.’ (Forosie pp16) 52. Nokwa ͻdan mu ha yɛ fɛ. Truly house inside here COP. beautiful ‘This room is truly beautiful.’ (Bediako pp 45) In the sentences above, the adverbs act as strengtheners, strengthening the proposition made in the sentences. In example (54), for instance, the fact that the speaker had to walk for a long time before they chanced on a very poor person is strengthened by the adverb ampa. In (55), nokwa emphasizes the beauty of the room. Nokwa is a variant of nokware. The use of nokwa/ nokware in the sentence guides the addressee to interpret the clause (that is the scope of the adverb) as extraordinary or above expectations.
3.3.1.1 Distribution of the certainty adverbs in Akan The Akan epistemic adverbs are not as flexible in their distribution as the Ewe modal certainty adverbs. As described by Gbegble (2012) godoo and kokoko can occur at clause initial, medial and final positions. Ampa can occur at both clause initial and clause final positions but nokware occurs only at the clause initial position. This echoes Saah’s (2004:65) assertion that, “though the two adverbs are synonymous, they differ in their syntax”. Their distributions are exemplified in examples (53) t o ( 5 6 ) below. Example (57) is ungrammatical in Akan.
54
53. Nokware, nsɛm a yɛ te no bi yɛ hu. Honestly, matter REL. 3PL hear DET. QUANT. COP. strange ‘Honestly, some of the things we hear are strange.’ 54. Bediako ho kyere- e no ampa. Bediako body catch- PST 3SG-OBJ truly ‘Bediako become truly impoverished.’ (Bediako pp 44) 55. Ampa, Forosie bɛ- brɛ -ɛ wͻ n’ abrabͻ mu. Truly, Forosie MOT.- suffer- PST in 3SG-OBJ life inside ‘Forosie really suffered in life.’ (Forosie pp 59) 56. Sɛ ɛ- yɛ nokware ampa sɛ obi afa wo If 3SG COP. true true COMP. someone PERF. take 2SG Sika a dɛn nti na w’ anka money COND. why then CONJ. 2SG PERF.- NEG.- talk ankyerɛ polisifoͻ no? PERF- NEG. teach policemen DET? ‘If it is true that your money got stolen, why didn’t you tell the police about it?’ (Bediako pp 89) 57. * Nsɛm a yɛ te no bi yɛ hu, nokware. Matter REL. 3PL hear DET. QUANT. COP. strange, honestly Nokware as used in example (56) is a predicative adjective. For more on modal predicative adjectives in Akan see Sakyi (2013).
3.3.2 Uncertainty adverbs: ebia, gyama, anhwɛ a, sɛsɛɛ and annyɛbiara. In the data, five possibility modal adverbs were identified, ebia ‘maybe’, gyama ‘perhaps’, anhwɛ a ‘probably’ sɛsɛɛ ‘by this time’ and annyɛbiara ‘probably’. This shows that the number of possibility modals exceeds the number of certainty modals. We will discuss these adverbs separately by looking at their distribution and also the meaning they contribute to an utterance.
55
3.3.2.1 Ebia ‘maybe’ Syntactically, it occurs at clause initial position. In a complex sentence, it is found at the beginning of the second clause, just after the complementizer. In very few contexts, it occurs at clause final position. When it occurs in this position, it is interpreted as a reflection. This use is mostly found in spoken discourse and not in written texts. Sentences 58, 59, 60 and 61 show the distribution of ebia in Akan. 58. Ebia ͻn- nim hͻ. Maybe 3SG- NEG-know there ‘Probably, he does not know there.’ 59. ͻ- fa -a no sɛ ebia ɛyɛ ͻ3SG-take PST 3SG-OBJ COMP maybe 3SG- COP 3SG sɛ na obi a w’ - ahu no look.like FOC someone REL. 2SG- PERF.-see 3SG-OBJ Akwamu sɛ no. Akwamu look.like 3SG-obj ‘He took it that maybe he resembles someone from Akwamu.’ (Forosie pp 34) 60. ͻ- ayɛ din saa no, ebia na kͻm de no 3SG- PERF- do quiet like CFM, maybe CONJ hunger take 3SG ‘The way she is so quiet, she is probably hungry’. 61. ͻmmo n- nim nea ͻmmo de ͻmmo sika yɛ, ebia. 3PL NEG-know what 3PL take 3PL money do, maybe ‘Probably, they don’t know what to do with their wealth’
Semantically, ebia as a procedural marker guides the addressee in his interpretation process. It does this by directing him to access the information that the speaker is not certain about the state of affairs described in the proposition stated or the truth conditions of the proposition. For an addressee to get the conceptual meaning of an ebia clause, s/he is guided by ebia to include the attitude
56
of the speaker. This is the main difference that exists between (56) above and (60) below. 62. ͻnnim hͻ. 3SG- NEG- know there ‘He does not know there.’ Example (60) is a statement of fact; the speaker does not express any attitude or shows any degree of commitment to the proposition.
3.3.2.2 Gyama ‘perhaps’ In the data collected, gyama was found mostly in the written text. Syntactically, gyama behaves very much like ebia. It occurs mostly at clause initial position but could occur at clause final where it is interpreted as an afterthought. The only difference that exists between them is that gyama can be used to ask a question but ebia cannot be used that way. Examples 63 to 67 show the distribution of gyama. In examples (64) to (66), gyama is used in interrogatives. Example (65) is unacceptable in Akan because ebia cannot be used in interrogatives. 63. Gyama yɛ- kͻ- du -u Kodiabɛ- foͻ asase so na Probably 3PL-go- reach- PST Kodiabɛ-people land on. CONJ mmerante yi mu baako yaaree bͻͻ no. men DET. in one sickess play 3SG-OBJ ‘It was probably when we got to Kodiabɛ that one young man fell sick.’ (Forosie pp 45) 64. Gyama wo papa Probably 2SG-POSS father ‘Your father is in, right?’
aba? PERF- come?
65. Gyama wo ho m- fa Probably 2SG-POSS body NEG-take ‘It seems you are not well?’
wo? 2SG-OBJ
57
66. Wo anim ayɛ kusuu yi, gyama kͻm de wo? 2SG face PERF.- do cloudy DET., perhaps hunger take 2PL ‘The way you look moody, I guess you are hungry? 67. * Ebia Kofi a- hu Probably Kofi perf- see
me? 1SG
When gyama is used in an interrogative sentence, the main idea is not to solicit for information. The question is either rhetorical (does not need an answer) or the speaker is just seeking for confirmation of the proposition expressed. In 64 the question is meant to be rhetorical. A mother asked his son this question, when he suddenly stopped playing and sat down quietly. In example 65, the speaker has noticed that the addressee is in not her usual happy self and thus seeks to confirm what he feels is wrong with the addressee. When it is not used as a question marker, gyama like ebia directs the addressee on the inferential path to interpreting the utterance as containing a speaker’s attitude of uncertainty. Thus in 63, the speaker expresses the attitude that he is not too certain if it was Kodiabɛ they got before one young man became sick or any place.
3.3.2.3 Sɛsɛɛ ‘probably’ I concur with Sakyi’s (2013) assertion that sɛsɛɛ is both an epistemic adverb and an adverb of time. As a temporal adverb, it is seen as a truncated form of the expressions saa bere yi ‘by this time’ or sɛsɛ bere ha ‘by this time’. As a modal adverb it occurs only at clause initial position. However as a temporal adverb, it can occur both at clause initial and final positions. It means the marker is ambiguous between epistemic modal interpretation and a temporal interpretation in
58
clause initial position 68. Sɛsɛɛ ͻ- gyina bɛɛbi bͻ nkͻmmͻ. Probably 3SG-stand somewhere play conversation ‘Probably, she is somewhere chatting.’ 69. Sɛsɛɛ deɛ wo nkͻ- to no. By this time TM 2SG NEG.-MOT. meet 3SG-Obj ‘It is not likely you will meet him by this time.’ 70. Akwasi a ͻ- rekͻ afafa yi, sɛsɛɛ Akwasi REL. 3SG-PROG-go half-half DET., probably ͻ - abo 3SG-PERF.- drunk ‘The way Akwasi is going sideways, he is probably drunk.’ 71. ɛ n- tumi m- ma sɛ Kofi a- da 3SG-NEG- be.able.to NEG-come COMP K. PERF.-sleep sɛsɛɛ. by this time ‘It cannot happen that Kofi is asleep by now’ Sɛsɛɛ is used in example 68 as an epistemic modal that signals the possibility or probability of the proposition embedded in it being true or false. Example (67) on the other hand provides a situation where it is used as a temporal adverb. In example 71 sɛsɛɛ in word final position is a temporal adverb.
3.3.2.4 Anhwɛ a ‘maybe’ Saah (2004) does not list anhwɛ a as an epistemic modal adverb. However in this study, it will be treated as a modal adverb. Syntactically, it behaves as an epistemic adverb, occurring only at clause initial position. This adverb anhwɛ a, is a truncated form of the clause sɛ wo anhwɛ a literally ‘If you don’t look’.
59
72. Anhwɛ mpo a, na obi de deɛ Probably even COND.,PART. someone take one na ama wo. CONJ PERF. give 2SG ‘Someone might even give you a bad one.’
ɛ- n- yɛ 3SG NEG. COP.
73. Anhwɛ nso a wo m- mɛ- to me. Probably FM COND. 2SG NEG-MOT-meet 1SG ‘You will probably not meet me.’ When used in a sentence, anhwɛ a signals the attitude of possibility of the speaker to the addressee. In interpreting a sentence with anhwɛ a, the addressee is directed to understand the proposition expressed as being potentially true or false at the time of speaking. Like all the other uncertainty adverbs it is non-committal. 3.3.2.5 Annyɛbiara/gyengyenbiara ‘Perhaps’ This is one of the modal adverbs listed by Saah (1998). Syntactically, it occurs only at clause initial position. Like gyama, it can be used in interrogative sentences. When used in an interrogative sentence, it demands an answer that is either a confirmation or denial of the proposition. It can also be used in rhetorical questions. 74. Annyɛbiara afei me- tumi akͻ- da? Perhaps now 1SG- be.able.to PERF.- MOT.-sleep ‘Perhaps can I go and sleep now?’ 75. Annyɛbiara obi e- bisa wo asem? Perhaps someone 3SG ask 2SG matter ‘Has anyone asked you a question?’ 76. Me- kͻ akͻhwɛ sɛ annyɛbiara ͻa1SG- go CONS.-MOT.-see COMP perhaps 3SG PERFba a come CFM ‘I am going to check if perhaps he is in’
60
Whether or not an annyɛbiara question demands an answer or is rhetorical depends solely on context. The intonation with which the sentence is said, the facial expression of the speaker and other gestures are usually used to determine the difference. Such expressions are usually used as a reproach. If the question is soliciting for an answer, it is usually a ‘yes/no’ answer. When annyɛbiara is used in used in a statement, it directs the addressee on the inferential path to interpret the utterance as containing a speaker’s attitude of uncertainty.
3.4
Chapter Summary
In this chapter, the mental state verbs and modal epistemic adverbs employed in expressing modality in Akan were discussed. The mental state verbs discussed are t he factive verb nim ‘to know’, and the attitude verbs susu ‘to suppose’ or ‘to assume’, and dwene ‘think’ and gyedi ‘to believe’. These verbs occur as part of the matrix clause in a complex construction. They show a speaker’s attitude towards the proposition expressed in the embedded clause. The epistemic adverbs in Akan can be grouped into two, the certainty adverbs and the uncertainty adverbs. There are two certainty adverbs, nokware ‘truly/honestly’ and ampa ‘truly/honestly’. Though these adverbs are synonyms they differ in their syntactic positions. Nokware cannot occur at clause final position, while ampa occurs in clause intial position. There are five uncertainty epistemic adverbs; and these are ebia ‘maybe’, gyama ‘perhaps’, anhwe a ‘probably’ sɛsɛɛ ‘by this time’ and annyɛbiara ‘probaby’. They all occur at clause initial positon. These adverbs express epistemic modality showing a speakers commitment to a proposition expressed.
61
CHAPTER FOUR MODALITY IN THE VERBAL GROUP
4.0 Overview In this chapter, I discuss the expression of modality in the Akan verbal group. The verbal group, according to Downing and Locke (2006:317) is “the grammatical unit by means of which we must typically express our perception of events”. Thus the verbal group refers to a combination of all the grammatical devices we use to express actions, states, activities, etc. The verbal group consists of a lexical verb occurring alone or preceded by auxiliaries. In the verbal cluster, modality can be expressed by the finite element or the auxiliaries (lexical auxiliaries). Lexical auxiliaries, according to Downing and Locke (2006:319), “is a term used for a set of verbs with modal or aspectual meanings”. Though Downing and Locke (2006) present a description of the verbal group of English, most of their observations are also true about most languages. In almost all languages, the verbal can consist of just one verb or with other markers. However one major difference between English and a language like Akan (see Osam 2008) is that Akan does not have an auxiliary system. Thus the notions of aspect and modality in the verbal group are expressed by certain affixes or sometimes as in the case of the habitual aspect in Akan, unmarked. Since I have already discussed verbs that encode modal meanings in the preceding chapter, I would focus mainly on the auxiliary verb, tumi, and the
62
prefix bɛ in this chapter. Tumi ‘be able to’ as a modal marker
4.1
When Christaller (1875) noticed the phenomenon in Akan where there is a sequence of two verbs without a coordinator, she described them as a case of main verb and auxiliary verb sequence. Since that study, linguists like Lord (1993), Osam (1997), Ameka (2005a) etc. have characterized this phenomenon as a case of verb serialization and called the phenomenon serial verb constructions (SVC). Example of an SVC is the example below; there are two verbs fa ‘take’ and beat bɔ ‘beat’ which are not separated by a coordinator. 77.
Mary fa-a dua bɔ-ɔ Peter. Mary take-PST stick hit- PST Peter ‘Mary took a stick and hit Peter.’
Thus the sequence of two or more verbs in most African languages is no longer looked at as auxiliary and main verb sequence but as a chain of two or more main or independent verbs. The SVC analysis is also based on the fact that unlike in languages like English, all the verbs in a sequence can be negated and carry aspectual affix. (Muysken and Veenstra (1995) provide a comprehensive list on the features of an SVC.) One of the results of these analyses is that it has led scholars like Brown and Miller (1991) to assert that there are no auxiliary verbs in Akan. Though there is no other study about the language that supports this assertion, there is also none that argues against it. The first evidence of auxiliaries in Akan is Amfo (2010) treatment of the form san ‘to return’ or ‘turn back’ as an auxiliary verb in Akan. In a footnote she acknowledges, “[t]he issue of what qualifies to be auxiliary verbs in Akan and other
63
related languages is quite contentious. In fact there is no well-defined class of verbs in Akan which belong to this category.” She argues she analyzes san as an auxiliary because it lacks independent meaning; it has apparently reduced verbal morphology and functional distinctness. In this study I also argue that tumi like san primarily lacks independent ‘action-type’ or ‘state of being type’ meaning. Tumi typically appears in the company of fully lexical verbs. These main verbs carry the core (conceptual) meaning in the verb phrase, while tumi directs the addressee to the inference the subject’s ability or its characteristics. I n examples 78 and 79 the main verbs nua ‘cook’ and to ‘sing’ express the conceptual meaning needed to make the sentence meaningful, tumi give rise to the contextual assumption that the subjects have the ability to perform the acts described by the lexical verb. 78. Vanessa tumi to dwom ma awosen gu wo Vanessa be.able.to sing song for goose.pimples pour 2SG ‘When Vanessa sings I get goose bumps.’ 79. Ama tumi noa aduane. Ama be.able.to cook food ‘Ama is able to cook.’ Tumi can also be used as the only verb in the sentence as in examples 79 and 80 below. People who argue against its auxiliary status use these examples as evidence, though English auxiliaries are also sometimes used as main verbs in some instances. Also these examples have been used to support the serial verb construction status of sentences like examples 77 and 78; the two verbs are all independent verbs and thus can be used as main verbs in any sentence. 80. Ama tumi nneɛma no. Ama be.able.to things DET ‘Ama is able to carry the things.’
64
81. Yare no tumi –i papa no. Sickness DET. able to-PST man DET ‘The sickness overcame the man.’ I contend that in these instances, these examples are examples of sentences in which the main/lexical verb has been ellipted. The sentence remains acceptable and meaningful because of the context in which it is used or because it has become conventionalized. The sentences above are thus the ellipted forms of the sentences in 81 and 82. The verbs soa ‘to carry’ and di ‘overcome’ are ellipted in the sentences above. 82. Ama tumi soa nneɛma no. Ama be.able.to carry things DET ‘Ama is able to carry the things.’ 83. Yare no tumi di -i papa Sickness DET. be.able.to overcome PST man ‘The sickness overcame the man.’
no so. DET top.
In the rest of the thesis I consider tumi as a modal auxiliary. Tumi has similar interpretations as the English modal can; it is thus primarily a dynamic modal. However, in certain contexts it can be used to express epistemic and deontic. The succeeding sections examine the different modal interpretations of tumi and the contexts in which each is made prominent. The first to be discussed is its primary function, dynamic modality.
4.1.1 As
Tumi indicating dynamic modality discussed
earlier,
dynamic
modality is
concerned
with
describing
characteristics or abilities of the NP in subject position. Some literature (Bybee
65
1985, Bybee and Fleischman 1995) also refers to this type of modality as agentoriented modality since the focus is on the agent of the sentence. This means that when tumi is used this way, it directs the addressee or hearer to process the information contained in the VP as ability or characteristics of the NP in the subject position. 84. Mmra tumi hyɛ nipa ma no yɛ nnea Law be.able.to force person for 3SG-OBJ do things ͻ- m- pɛ. 3SG- NEG-like ‘The law can force people to behave in a way they do not like.’ (Peace FM 6/01/14) 85. Mmra hyɛ nipa ma no yɛ nnea ͻ- m- pɛ. Law force person for 3SG-OBJ do things 3SG NEG-like ‘The law forces people to behave in a way they do not like.’ The sentence in 83 is taken from a discussion on radio about whether the laws of the country actually work. The speaker who made this utterance argues that indeed they do. Therefore the sentence needs to be interpreted on the basis that the law has an ability or power to do something. The sentence therefore means that though the law has the ability to perform that function, it does not happen in all situations. The example that does not contain tumi gives a different interpretation. It means that the law actually forces people to act against their will. This is why though the addition of tumi in the first sentence gives the addressee more information to be processed, it remains relevant by generating more cognitive effects that the non-modalized one. Tumi thus performs a modal function in this instance.
66
4.1.2
Tumi indicating epistemic modality
When tumi is interpreted as an epistemic modal, it infers a speaker’s stance on the proposition under discussion. The speaker usually communicates the possibility or impossibility, if the verb is negated, of the situation stated in the proposition. 86. Me- hwɛ nea 1SG-look what
ɛ-
re-
keka kͻ
yi
a,
ͻman yi
mu
3SG- PROG.-say go DET, REL. nation DET in.
tumi kyɛ. be.able.to divide ‘Looking at what is going on, the nation can be divided.’ (Adom FM 04/02/13) 87. Nnea ͻmmo yɛ yi nyinaa tumi de ntͻkwa ba. what 3PL do DET all be.able.to take fight come ‘ What they are doing can result in a fight.’ (Adom FM 04/02/13) Examples 85 and 86 are comments by a panelist on the Newspaper review on Adom Fm. He was reacting to certain agitations regarding the country’s 2012 general election. The speaker expresses his fear of instability in the country but does not state emphatically that there is going to be a war. He states that looking at the tension and what people are doing, politicians inciting ethnic groups against each, there is enough evidence to show that there is the possibility of war. Omitting tumi in any of the above sentences will probably have lead the host asked the speaker to renounce his speech.
4.1.3
Tumi indicating deontic modality
In certain context, the verb is used to give or seek permission. In these contexts it is being used as deontic modal. When it is used to give permission the authority rests
67
in the speaker while when it is used to seek permission the authority lays in an outside force. Authority as used here does not imply dominant power or control but the having ability in a specific context to give an instruction. 88. Mmerɛ no a yɛ- ka- eɛ no adu, time DET REL 3PL-SUBJ-say- PST DEF PERF-reach metumi ba anaa? 1SG-SUBJ be able come QM ‘The time we talked about is up, can I come?’ 89. Mepa wo kyew afei metumi kasa? 1SG-subj- remove 2SG-poss hat now 1SG-subj- be.able.to speak ‘Please, can I speak now?’ 90. Wo- tumi pue 2SG be.able.to go.out ‘You may leave.’ Examples 87 and 88 show instances where tumi is used a speaker to seek permission. Sometimes when it is used this way the addressee may not even have any real authority and the speaker is only being sarcastic; example 88 can be interpreted this way. In 89, the speaker is the authority that grants permission to the addressee. In these cases, the speaker has to have some authority otherwise the sentence becomes infelicitous.
4.1.4
Conclusion on tumi
In certain contexts, a sentence containing tumi can be ambiguous between all these modals meanings associated with it: epistemic, deontic meaning and dynamic. When we take the sentence below for example, it can be interpreted in two ways. 91. Kofi tumi da. Kofi be.able.to sleep ‘Kofi can sleep.’
68
The three possible interpretations are: a. It is possible for Kofi to asleep. (It may so happen that Kofi is sleep) b. Kofi is being given permission to sleep. (Kofi may sleep) c. Kofi has the ability/ characteristic of sleeping. (Kofi can sleep)
The first reading gets an epistemic interpretation. The speaker infers that there is some evidence available to him/her at the time of speaker for him to come to a conclusion that Kofi is sleep. In the second reading, the speaker as a form of authority communicates that it is favourable or desirable to him (speaker) that Kofi sleeps at the time of speaking. This is the
deontic
interpretation of the
sentence. The last reading is based on the fact that Kofi has the characteristics or ability to sleep; Kofi does not have insomnia or any sleep disorder. The ambiguity is only resolved in the context of use. Modal expressions conveying more than one modal type are not new in the literature on modality. The English modal verbs must, may, will and can have all been analyzed as having more than one modal interpretation. Scholars of modality however do not agree on how these multiple meanings should be analyzed. Scholars like Coates (1983) and Palmer (1990) argue that the modals are polysemous. Haegeman (1989), Klinge (1993), Nicolle (1998) and Papafrogou (2000) on the other hand argue for a monosemantic approach. Among those who take the monosemantic approach, Klinge (1993) and Nicolle (1998) assert the meanings of the modals should be looked for outside of context. On the other hand scholars like Papafrogou (2000) who work within relevance theoretical framework
69
argue for the interpretation of the modals in context. Papafrogou (2000) argues that the meanings of these modals are undetermined and therefore need free enrichment form context to get contextually different interpretations. With regards to the interpretation of the Akan auxiliary verb tumi, this study supports the view of Papafrogou (2000). We suggest that tumi is undetermined and that the different interpretations it gets are as a result of the free enrichment from context.
4.2
The status of future markers
One of the issues that have received much attention from scholars working on modality is the interaction between future interpretation and modal interpretation. This issue stems from the fact that reference to future states of affairs is inherently non-factual and non-factuality is also the main defining component of modality. Looking at the expression of modality in English, Comrie (1985) asserts that though English has been extensively studied, the question of whether will is a future marker or modal marker has still not been sufficiently discussed. He adds that the auxiliary has a number of other uses in addition to the expression of future time reference. In particular, there are modal uses which do not necessarily have future time reference. Essegbey (2008) on the other hand argues for a non-future analysis of the morpheme which was previously termed the future morpheme in Ewe. In this section, we will examine the interpretation the morpheme bɛ- contributes to discourse. This morpheme is labeled as the future marker in Akan which is sometimes used with other modal expressions to express modality. The main focus of this section is to determine its status as a modal or a future marker.
70
4.2.1
The future marker bɛ-
In almost all literature on Akan syntax, the morpheme ‘bɛ-’ is treated as a future marker (see Dolphyne 1988, 1996, Boadi 2008, and Osam 2008). As a future marker, it is used to express the idea that the event described by the verb will occur later than the utterance time. According to Osam (2008:72), “[t]he main function of this prefix is to indicate future time reference”. Bɛ- is thus glossed as will in English. It presupposes that the primary function of the morpheme is to locate a state of affairs in a time posterior to the moment of speech. This should be its primary meaning because according to Comrie (1985) since other modals markers can also perform this function. The examples below taken from Osam (2008:71) were used as instantiations of the future interpretations of the morpheme. The (Fa) attached to the examples is used to indicate the examples are from Fante one of the dialects of Akan. 92.
a. Mboa bɛ- ba Help FUT-come ‘Help will come.’
(Fa)
b. Yɛ -bͻ5- kͻ Kumasi. (Fa) 1PL.SUB-FUT-go Kumasi ‘We will go to Kumasi.’ In the above examples, we notice that the action or event which ‘ comes’ and ‘ goes’ denote have not yet taken place. These actions will take place at a later time than the utterance time. These examples seem to confirm that the morpheme bɛis a future marker. However, it appears that there is more to the morpheme than this.
5
In Fante, vowel harmony refers to both tongue root assimilation and rounding assimilation. bͻ is thus the variant of bɛ after a verb containing a –ATR rounded vowel.
71
Osam (2008:14) adds that when the morpheme bɛ- co-occurs with other modals in
Akan,
the
morpheme
gets
a
prediction,
intention
or
possibility
interpretation. H e does not however explain what accounts for this sudden change in meaning. For instance when we take an example like 92 Osam (2008) argues that bɛ- in this instance has the meaning of possibility because of the presence of the modal adverb bi ana ‘maybe’. 93. Bi ana ͻbɛba(Fa) may be 3SG.SUB-FUT- come ‘May be he might come’
Osam (2008:15)
Boadi (2008) a l s o adds that in certain contexts, the morpheme bɛ- has probability or likelihood interpretations. He explains that this happens when the morpheme co- occur with verbs such as wͻ ‘to be there’, te ‘to stay’, so ‘to be big’. Therefore in a sentence like 93 below, bɛ- is said to have a potentiality interpretation because of the presence of the verb wͻ ‘to be there’.
94. Sika no bɛ- wͻ hͻ ara. Money DET FUT- be there EMP. ‘The money is most likely there.’
(Boadi 2008:23)
He also does not explain the sudden change in meaning of the morpheme.
4.2.2
Bɛ- as a modal marker in Akan
Though in the previous section and other studies there seems to be a consensus that bɛ- is a future marker, in this section, I draw attention to the modal interpretation of bɛ-. There are two things that challenge the analysis of bɛ- as a future tense morpheme in Akan. The first one is Osam’s (2008) assertion that
72
Akan is a purely aspectual language like Ewe. As an aspectual language, he argues that there is no past tense in the language rather a completive aspect. The future/non future distinction, which is tense and not aspect, is therefore not consistent with his own analysis. The second challenge is that the morpheme has other uses that are not related to future. In the previous section we notice Osam (2008) and Boadi (2008) point to the fact that bɛ- is more than a future morpheme. Apart from the contexts pointed out by these authors, there are other contexts in the language where the morpheme has no future reading at all. Examples are provided below: 95. Ataadeɛ no bɛ -kͻ no. Dress DET MOD.-go 3SG.obj ‘The dress will fit him/her.’ 96. Mpaboa no bɛ- yɛ me dea. Sandals DET MOD-do 1SG mine ‘The sandals might be mine.’ 97. ͻ- bɛnim fie hͻ 3SG-MOD.- know house there ‘S/he is likely to know the house.’ The focus of bɛ- in examples 94 to 96 above is not to draw the addressee’s attention to the time the action (described in the proposition) will take place; the speaker is more concerned with the possibility or probability of the proposition taking place. The ‘yet to occur’ interpretation is not present in this instance. In 94, the speaker does not communicate that in a time posterior to the utterance time, the dress will fit him/her. He rather communicates the possibility of the dress being his/her size. In 95 and 96 as well, there is hint for a future interpretation. The co- occurrence of bɛ- and nim in 96 is one further challenge against the
73
future meaning of the morpheme. Nim is a stative verb and stative verbs do not take tense or aspectual markers. Therefore there are two possible explanations of the co-occurrence of these two forms. First, we can argue that it is an exception; stative verbs can occur with future tense. The second rationalization, which is what I argue for, is that bɛ- is a modal marker and thus does not falter the rule. As noted by Essegbey (2008) in the discussion of the Ewe a-morpheme, the morpheme has been categorized as a future tense morpheme because it usually occurs in the context where the action is yet to occur. This morpheme prior to Essebgey’s analysis as a potential marker had been labeled a future tense morpheme. The Akan morpheme is also used in similar (where the action is yet to occur) contexts. This might also be a reason why the morpheme is labeled a future marker in Akan. Consider examples 97 and 98 below: 98. Kofi â- ɖu mͻlu le ɤe- trͻ me. Kofi POT-eat rice LOC sun-turn containing.region ‘Kofi “will” eat rice in the late afternoon.’ (Essegbey 2008:205) 99. Egypt bɛ- hyɛ Ghana. Egypt MOD-score Ghana ‘Egypt “will” beat Ghana.’ 6 In these examples, the actions that the propositions express have not taken place yet. The presence of such sentences is what leads to the tendency to interpret the morpheme as expressing future. However as he clearly states, there are instances where the action is taking place in the present or has already taken place (past). The English gloss will is in quoatation marks because the researcher does not accept that interpretation. I prefer ‘There is a possibility of Egypt beating Ghana in the game’. 6
74
This argument is in consonance with Berbeira- Gardón’s (2006) assertion that will in English is not entirely future. He adds that there are instances where it is used in past and present situations. Essegbey gives an instance in Ewe where the amorpheme describes an action that has already taken place. In example 99 below, the presence of the adverb already clearly indicates that the action being described happened in the past. 100.
Kofi â- yi Ga xóxó. Kofi POT-go Accra already ‘Kofi may have gone to Accra already.’ (Essegbey 2008:207)
In Akan, bɛ is used in similar contexts. The context for example 100 is A has lost her money and suspects Kofi of being the thief. He insinuates this by using by saying the sentence in 100. In 101 A has done something bad to B, B complains to C who is a mutual friend. The sentence in (101) is a comment C makes about A. 101. Kofi ara na ͻ- bɛfa sika no. Kofi Emp Part. FOC. 3SG MOD.- take money DET ‘Kofi is certainly the one who stole the money.’ 102. ͻno ara na ͻ- bɛ- di saa mmoa- sem yi. 3SG Emp part. foc. 3SG-MOD. eat this animal matter Det ‘S/he is the one who will behave that foolishly.’ What is common bet w een these sentences is that the actions denoted by the predicates have already taken place. Thus there is no way the morpheme can be interpreted to locate a state of affairs in a time posterior to the moment of speech; the actions are already in the past. In the examples below the actions described by the verbs are taking place in the present. 103. Ga lá â- sue ákpá. Money DEF POT-be.small too.much ‘The money may be too small.’
(Essegbey 2008:207)
75
104. Mary will be at the airport by now.
Berbeira- Gardón (2006:451)
105. Kwame bɛ- wͻ fie. Kwame MOD.-be.located home ‘Kofi is at home.’ The meaning that seems common to all instances where the morpheme bɛ- is used, is “the potential occurrence of the event” Downing and Locke’s (1992:379). The modal meaning of bɛ-is the most prominent meaning in most contexts. If it has been established that bɛ is a modal, then in the subsequent sections the modal meanings that associated with the morpheme will be discussed. The morpheme bɛ is compatible with all three modal types discussed previously: epistemic, deontic and dynamic. As noted for the auxiliary verb tumi, bɛ is also undetermined in isolation and thus need free enrichment from context to express a particular modal meaning. It is the different enrichment it gets in different context that leads to the three modal interpretations that will be discussed below. We will take these meanings one after the other in the sections below. 4.2.2.1 Bɛ- indicating epistemic modality As discussed in chapter one, epistemic modality is concerned with the speaker’s “judgments about the factual status of the proposition” Palmer (2003). Epistemic modality is usually on a range from certainty to uncertainty. In between these, there are the notions of possibility and probability. Epistemic modality concerns itself with the evidence (physical or psychological) available to a speaker during communication. In English the modals like must, may, possibly and certainly have epistemic meanings. The morpheme bɛ- can have an epistemic meaning in certain contexts. As an epistemic form, it could have meanings ranging from strong
76
certainty to weak certainty or probability. The particular meaning it acquires depends on the evidence available to the speaker at the time of speaking. To explain this further, we take a look at example 98 again. A radio newscaster at Peace Fm made this assertion when he was asked (unofficially) about the score line of an upcoming match between Ghana and Egypt. Here we can assume that the speaker uses the football history of Egypt and Ghana as evidence; Egypt had won more African Cups than Ghana and also had seasoned players. Based on this evidence, he concludes that there is a strong certainty that Egypt was going to beat Ghana. The meaning or interpretation this sentence receives in this context is similar to the non-modalized form in 105. In non-modalized form, the speaker makes an absolute claim using the perfective to describe an action that is yet to occur. 106. Egypt ahyɛ Ghana a- wie. Egypt PERF.-score Ghana perf-finish ‘Egypt has already beaten Ghana.’ 107. Ɛnnɛ yɛ- bɛ- tͻ adeɛ. Today IMP- MOD- buy thing ‘Today, the sales will be good.’
(Adom FM 16/10/2013)
A similar usage is seen in 106, a statement by a market woman who was interviewed after the Ghana-Egypt match. The presenter asked the woman if she thought the outcome of the match was going to influence sales. Market sales usually go up after the national team wins a match and so based on this the woman looks forward to higher sales. In both instances the morpheme express certainty on part of the speaker that the proposition he expresses will come to pass. Both speakers will be greatly disappointed if what is expressed does not come to pass. As an epistemic modal, bɛ- does not express possibility or probability. A possibility or probability
77
interpretation is however possible if combined with tumi as in the example below. 108. Egypt bɛtumi ahyɛ Ghana. Egypt MOD-be.able.to PERF.-score Ghana ‘There is a possibility Egypt will beat Ghana.’ From the above examples, one thing that is obvious is that all the actions are yet to occur; it is epistemic interpretation that is associated with the future interpretation. 4.2.2.3 Bɛ- indicating Deontic modality Deontic modality is concerned with the criterion by which speakers/writers decide which future events are necessary, possible, desirable, etc. Deontic modality relates to “obligation or permission”. It is usually based on authority, a force which is outside the addressee. In general, deontic modality is dependent on some kind of authority, often, the speakers’. A deontic modal is therefore used to affect a situation. Bɛ- usually gets a deontic interpretation when it is used in the imperative: 109. Wo- bɛ- kͻ sukuu! 2SG MOD-go school ‘You will go to school!’ 110. Obiara bɛ- tua saa sika yi bi Everyone MOD-pay this money this some ‘We are all going to pay this money.’ Example 108 is a command; it is usually used to stop an argument from proceeding further. It can be used in a case where a child who is refusing to go to school tries to give some excuses or to argue with his mother about his decision. In both sentences the authority is from the speaker.
78
4.2.2.3 Bɛ- indicating dynamic modality Dynamic modality does not express the speaker's opinion, it describes the abilities or properties or the power an entity has. According to Downing and Locke (2006:390), dynamic modality expresses properties or disposition of the subject referent. Unlike deontic modality, when a dynamic modal is used, the speaker does not affect the situation, he states it as it is. When bɛ- is used as a dynamic modal, it shows that the objects described have the ability, property or power to perform a certain task. The use of bɛ-directs the addressee to interpret the VP as a characteristic of the NP subject. This is similar to the interpretation Boadi (2008) calls the habitual usage of bɛ-. We will take a look at an example he cites as an example of habitual usage. 111. Wo- frɛ no a ͻ- bɛ- gyina hͻ aa na 2SG call 3SG-OBJ COND. 3SG-fut.-stand there still COMP ͻ- agye so. 3SG respond. ‘If you call him, he will stand still for a long time before he responds.’ 112. Annwa no bɛ- da wᴐ nsuo no ani Oil Det MOD-sleep be.located water Det top ‘Oil will float on water.’ The VP that bɛ is a part of describes an action which is asserted to be a characteristic of the referent NP, in this case, ͻ(no). This is not different from the idea expressed in 111 and 112. The VP bɛ-da wᴐ nsuo
no ani describes a
characteristic of the NP annwa ‘oil’ which is in the subject position. In example 112 the VP bɛ-tɛ nsuo no ani describes a characteristic of the NP bͻͻlo which is also in the subject position. We can infer from the examples above that the VP describes a characteristic that is inherent in the NP. In other words, it describes
79
some characteristics, which are part of the chemical composition of the entity encoded by the NPs; oil has some chemical compositions that make it float on water and so does balls. This is not always the case though; bɛ- in the VP can also be used to describe a persistent habit or character of a person as seen in example 113. 113. Ama ba class a, ͻbɛkasa saa. Ama come class COND. 3SG.SUBJ- MOD- talk a lot ‘When Ama comes to class she will talk a lot.’ The fact that Ama talks in class is not part of her composition or makeup as a human being; rather it is an attitude she continually puts up in class. After looking at the various discussions, I will attempt to spell out the status of the bɛ morpheme in Akan in the next section. 4.2
The current status of bɛ- in Akan
From the discussion so far, we have seen instances where bɛ is interpreted as a future time reference marker and in others as a marker expressing modality. If we take a closer look at examples 91a and 98 repeated here as (113) and (114), we notice that they are similar. 114. Mboa bɛ- ba (Fa) Help FUT-come ‘Help will come.’ 115. Egypt bɛhyɛ Ghana. Egypt MOD- score Ghana ‘Egypt will beat Ghana.’ These two sentences appear in contexts which readily lends itself to a yet to occur interpretation; and it appears this is the meaning Osam (2008) and Boadi (2008) assign to it. I have argued that these two sentences and others like them are all
80
instances of the epistemic use of the morpheme. The reason is that the speaker’s interest is on the likelihood or potential of the action happening. In most instances where bɛ- has epistemic readings, these readings are also compatible with the future readings or as Essegbey (2008) will call ‘yet-to-occur’ interpretations. We do not find this strange because the idea of prediction is inherently future; it is not possible to predict the past or the present. We must note that in some instances of the epistemic use of the morpheme a future reading is not possible. In 115 for instance, a future reading is not possible. 116. Ataade no bɛ- fata wo. Dress Det MOD-suit 2SG ‘The dress will suit you.’ This sentence does not have future connotations; the speaker does not refer to a point in time where the dress will be a perfect fit; it rather expresses the possibility that it can. The deontic meaning of bɛ- is also compatible with future readings. This is because when directives are given, it presupposes that the action is yet to occur. The analysis in the present study suggests that the morpheme bɛ- in Akan is a modal marker which has future time reference reading when it is used as an epistemic or deontic modal. This echoes Ferdinand (2006:50) assertion that, there is a connection between modality and future. This is because quite often, future morphemes develop from modal (deontic) forms. This analysis is also in consonance with Coates (1983:61) assertion that, “Future reference is subtly bound with modality, and is an essential component of requests, warnings etc.” This observation contrasts with Osam’s (2008) assertion that the morpheme bɛ- needs to
81
be collocated with other modals to express modality. In almost all the examples presented above, bɛ is used alone but the modal interpretation is still evident, and even, the most prominent. In this way bɛ- is similar to will in English as analyzed by Berbeira- Gardón (2006) and the a-morpheme in Ewe as discussed by Essegbey’s (2008). It however differs from Tonhauser’s (2011) analysis of Paraguayan Guaraní future marker –ta (see section 2.1.3).
4.4
Chapter summary
This chapter examined two markers that are used to express modality in the Akan verbal complex; the auxiliary verb marker tumi ‘be able to’ and the modal prefix bɛ-. Tumi acts as an auxiliary verb in Akan that is used to show that an NP has a characteristic or ability to do an action described by a lexical verb it occurs with. Its use is similar to the use of teŋu in Ewe (Ameka 2008) and can in English (Palmer 2003). As an ability modal, its primary modal meaning is dynamic modality but it can also convey epistemic and deontic modality in certain contexts. The other functional category that was discussed was the prefix bɛ- in Akan. From this chapter, we conclude that bɛ- like the a-morpheme in Ewe is a modal marker in Akan. The modal expresses three types of modal meanings, epistemic modality, deontic modality and dynamic modality. The only difference between the amorpheme in Ewe and bɛ- in Akan is that bɛ- is compatible with future interpretations in some contexts. In some instances where it expresses epistemic and deontic modal meanings, it is compatible with future interpretation.
82
CHAPTER FIVE CONSTRUCTIONAL MEANS OF EXPRESSING MODALITY
5. 0
Overview
In this chapter we will deal with three issues; periphrastic modal constructions in Akan, modal concord (Geurts & Huitink 2006) and the use of some modals to mark politeness in discourse. In the preceding chapters we focused on how lexical words (verbs and adverbs) are used to express modality and also some modal markers (the auxiliary verb tumi and the prefix bɛ-) in the Akan verbal group. In this chapter our main focus will be on the use of multiple words to express the concept of modality in Akan. For want of a better term to describe this phenomenon, we will follow Essegbey (2008) and refer to them as perisphrastic modal ‘constructions’. They are labeled ‘constructions’ because they do not qualify to be called clauses or phrases. An example of such a construction is ɛ-wͻsɛ, which is made up of a third personal ɛ- singular marker, the verb wͻ and the complementizer sɛ. This is followed by the clause which is being modalized. In this thesis, we will refer to that clause as the sɛ-clause. 5. 1
Constructions that are used to express modality in Akan
Morphemes, the smallest meaningful units of words, come together to form words (though some morphemes qualify to be considered as words); and words come together to form phrases (though some words may also qualify to be phrases), and phrases (specifically the noun phrase and the verb phrase) come together to
83
form clauses, and clause come together to form sentences; and sentences may also come together to form larger utterances like paragraphs, passages, etc. Our interest is in how to use a whole sentence to express modality in Akan. In view of that, it would be prudent to first look at the immediate constituents of a sentence (i.e. clauses). Sentences are of two types, simple and complex sentences. The simple sentence is made up of just one clause while the complex sentence is made up of more than one clause. A clause, according to Crystal (1997:62), is “the smallest grammatical unit that can express a proposition”. A clause mostly consists of a subject and a predicate. Crystal (1997:62) distinguishes primarily two types of clauses; main clause (matrix clause or independent clause) and subordinate clause (embedded clause or dependent clause). The main difference between these two is that a main clause can stand on its own to express a complete thought, but the subordinate clause depends on the main clause; otherwise the meaning it expresses would not be complete. We should also mention that the main clauses are grouped depending on its
function:
statements/declaratives,
questions/interrogatives
or
commands/imperatives. Each main clause can function as a simple sentence (in isolation). Likewise, the subordinate clauses are of three types depending on their function; reported, reasoning and adding/relative clauses. Let us look at some examples of the main clauses and the subordinate clauses below. 117. Kofi di fufuo. (Statement) K eat fufu ‘Kofi eats fufu.’
84
118. Bra ha! Come here ‘Come here!’
119. Ama ada A PERF sleep ‘Is Ama asleep?’
(Command)
(Question)
Sentences 117, 118 and 119 are simple sentences functioning as statement, question and command respectively. Two or more main clauses come together to form a compound sentence; a main clause and a subordinate clause (or a main clause and more than one subordinate clause) would form a complex sentence; and more than one main clause together with a subordinate clause (or more than one subordinate clause) would form a compound-complex sentence as the examples below depict. 120. Kofi yɛ kͻkͻͻ nanso ͻ- ndware yie. K be fair but 3SG-NEG- bath well ‘Kofi is fair but he does not bath well.’
(Compound sentence)
121. Kofi ka -a sɛ Ama ada. K say-PST COMP A PERF- sleep ‘Kofi said that Ama is asleep.’
(Complex sentence)
122. ͻ- kͻ-ͻ hͻ na ͻ- tena - a ase sɛ ͻ3SG-go-PST there CONJ 3SG- sit- PST down COMPS 3SGredidi. PROG-eat ‘S/he went there and sat down to eat.’ (Compound Complex sentence) Sentences 120, 121 and 122 are compound, complex and compound-complex respectively. The underlined parts of these sentences are the subordinate clauses and the un-underlined, the main clauses.
85
Unlike those we saw in chapter three, where the verb in the main clause expresses the modal meaning, for these types of sentences, the verb in the main clause alone does not express any modal meaning on its own; it has to combine with the subordinate clause, specifically the complementizer, to convey this meaning. Let us now turn to the structure of these types of constructions. These constructions are of three types: i.
3SG + verb + Sɛ-Clause
ii.
3SG + bɛ + verb + Sɛ-Clause
iii.
NP + wͻ + NP + Sɛ-Clause
Only a few Akan verbs can occur in the verb slot in these constructions. The first of such constructions we will consider is 3SG + verb + Sɛ Clause.
In these
structures, the third person singular pronoun is an expletive. This means that the pronoun performs a syntactic role but has no semantic connotation or referent. The function it plays here is similar to the role of it in the English sentence below: 123. It is about to rain. In 123 though no meaning can be assigned to it, it plays a significant syntactic role. This is because the sentence becomes ungrammatical without it. In most languages, sentences require subjects (either covertly or overtly). Therefore, where there is no semantic subject, an expletive or dummy subject is put there. This is the requirement of the Extended Projection Principle (EPP) (see Carnie 2007). The expletive is then followed by a verb in the construction. The complement clause begins with the complementizer sɛ. We will take a detail look at each of the structures.
86
5.1.1
3SG + verb + Sɛ Clause
As stated earlier, there are a limited number of verbs that can occur at the verb slot in this structure. These verbs include wͻ ‘to be located at/ to have’, in this context however, to have is more prominent, sɛ ‘to be fit’, hia ‘to need’, yɛ ‘to be good’ and fata ‘to deserve/ to be fit for something’. Ameka (2008: 147) describing this phenomenon in Ewe groups these verbs into two; necessity and obligation verbs, and emotional and attitudinal predicates. The necessity and obligation verbs will include wͻ ‘to be located at’ and sɛ ‘to be fit’ whiles the emotional and attitudinal predicates are hia ‘to need’, yɛ ‘to be good’ and fata ‘to deserve/to be fit for something or someone to do something’. The verb is then followed by the complement clause, usually preceded by the complementizer sɛ. The clause that follows the complementizer is the one that is modalized (see Ameka 2008: 147 for a similar analysis in Ewe)7. The examples below throw more light on the above assertions. We would begin with ‘ɛ-sɛ sɛ …’ constructions.
5.1.1.1 Necessity and Obligation verbs The Akan necessity and obligation verbs wͻ ‘to be located at’ and sɛ ‘to be fit’ in these constructions seem to be losing their semantic properties in this context. In fact, they are gradually going through the process of grammaticalization. It is very likely that they may lose their lexical status and later become more or less
The structure in Ewe is É- ‘3SG Expletive’ Verb bé ‘COMP’ clause. (Ameka 2008: 147) é- dze bé nà- vá kpͻ- m 3SG-fit COMP 2SG:SUBJ-VENT see- 1SG ‘You must come to see me’ 7
87
grammatical words in this context. The reason is that, in isolation, these words have lexical meanings: wͻ means ‘to be located at/ to have’ and sɛ ‘to be fit’; but when they occur in the context provided above, they lose those lexical meanings as seen in examples 124, 125and 127 below. 124. Ɛsɛ - sɛ kristoni deɛ woyɛ 3SG fit COMP Christian TM 2SG- COP ‘As a Christian you must be on fire.’
hye. hot
125. Ɛ- sɛ - sɛ barima tumi bre ne mogya ani. 3SG fit COMP man be.able.to suffer 3SG-OBJ blood top ‘A man must work hard.’ (Forosie pp 67) 126. Safoa no - deɛ ɛ- sɛ - sɛ ɛ- da pono wei so. Key DET TM 3SG fit COMP 3SG-sleep table DET on ‘As for the key, it should be on this key’ Compared to the examples below, it becomes evident the verb ‘sɛ’, loses its meaning in these contexts. 127. Ahennie no sɛ Kofi. Throne DET be.fit K ‘Kofi deserves the throne.’ 128. Abranteɛ yi deɛ, me ara na ͻsɛ Guy DEF TM 1SG EMP.Part FOC 3SG- be.fit ‘As for this guy, he befits me (alone)/I deserve this guy.’
me. 1SG
We get the actual sense of befitting or deserving in sentences 127 and 128, unlike those in 124 to 126. In 127, the speaker asserts that the throne befits Kofi while in 127; the speaker asserts that the gentleman is the right suitor for her. This may be due to some qualities she sees in the guy. On the other hand, in 124, 125 and 126 the verb sɛ alone has no meaning; it only gets its meaning when it combines with the CC. When its meaning then becomes equivalent to the English auxiliary must or should.
88
In the context where it is equivalent to must, ɛsɛsɛ has the meaning ‘it is necessarily/obligatorily the case that X is true” (Downing and Locke 1992: 382). It is compatible with both the deontic and epistemic interpretations of must. . In examples 124 and 125 above, the constructions express deontic modality; the speaker seeks to intervene in the speech act (get the addressee to perform an action). In example 124, the speaker communicates that ‘it is necessarily the case that a Christian be on fire for God’; this is from a pastor admonishing his members to worship God. In example 125, s/he communicates that ‘it is necessarily the case that a man works hard’. Example 126 on other hand expresses epistemic modality; s/he states that it is necessarily the case that ‘the key be on the table. Here, the speaker is more concerned about the truth of the proposition expressed by the main clause. The speaker, thus, communicates that, beyond reasonable doubt, the key is on the table. When it is used this way, the speaker cancels the possibility that the proposition is false. In this example, a family has misplaced their bunch of keys; the father is the last person who handled it. The only place he searches is the table. When he was advised to search other places he insists that he placed them on the table himself. When ɛsɛ sɛ … is used, the speaker leaves little room for the proposition to be denied or contradicted. The next construction we will look at is ɛwͻ sɛ ... The verb wͻ, in isolation, means ‘to be located at or to be at a particular place’ as in example 129 or ‘to have’ as in example 130. It is also sometimes used as a preposition as shown in example 131. 129. Kofi wͻ Kumase. K be. located Kumasi ‘Kofi is in Kumasi.’
89
130. Kofi wͻ sika. K to.have money Kofi has money. 131. Me- hunu-u safoa no wͻ pono no so ISG- see- PST key DET PREP table DET on ‘I saw the keys on the table.’ In 129, we can see clearly that wͻ is a verb that means to be at a location; the speaker declares the location of Kofi, which is Kumasi. In 130 the speaker states that Kofi has money in his possession or has sufficient money or is rich. In example 131 on the other hand, wͻ is no longer a verb but it is used as a preposition; though it still has the sense of being in a location. However, like the verb sɛ in ɛ-sɛ sɛ…, wͻ in the construction ɛwͻ sɛ ..., has lost loses its semantic property (Its location meaning) completely. It gets a new denotation when it combines with the sɛ-clause. This new meaning, depending on the context, is equivalent to the English auxiliary should or the semi-modal ought to. Though it also has a modal necessity meaning, this is weaker than the meaning of ɛsɛ sɛ. Ɛwͻ sɛ…, in certain contexts, has epistemic readings while in others, a deontic modal reading. The examples below throw more light on the discussions above. 132. Ɛwͻ sɛ Ministers no yɛ ready 3SG to.have COMP Ministers DET COP. ready to sacrifice nneɛma bi to sacrifice things some ‘The ministers should/ ought to be ready to sacrifice certain things.’ (Peace FM 20/11/13)
90
133. Adeɛ asa, ɛ- wͻ sɛ wo- kͻ fie. Thing- PERF-finish 3SG- to. have COMP 2SG- go home ‘It is late, you should go home.’
134. Abͻ nine, ɛ- wͻ sɛ ͻwͻ hͻ. It.is nine o’clock, 3SG to.have COMP 3SG be.located here ‘it is nine o’clock, s/he should be there.’
135. Na ɛ- wͻ sɛ me ne no kͻ fie. PST 3SG-to.have COMP 1SG CONJ 3SG-OBJ go home ‘I should have gone home with him.’ 136. Ɛ- nnisɛ wo- kasa neho. 3SG NEG. to.have COMP 2SG talk 3SG-POSS- body ‘You should not talk to him’ In example 132, the speaker puts across what someone should do as a minister. This may be based on the person’s knowledge o r expectations about the duties of a minister. In example 133, the speaker advises the addressee to go home because it is late. In 134 the speaker, based on his experience, asserts that the person they are talking about should be there at that time. In 135, the particle na indicates that the speaker is talking about a past event. He states that he should have gone home with someone (the person in question). Though in Akan, a negative marker ‘n/m’ is prefixed to a verb to negate the proposition without necessarily changing the shape of the verb as in kͻ ‘go’ and n-kͻ ‘don’t go’, su ‘cry’ and n-su ‘don’t cry’, etc, the shape of wͻ change to nni when the proposition is negated as seen in example 136. The verbs wͻ and sɛ cannot take the tense/aspect markers like future, past, progressive, etc. When a speaker uses ɛwͻ sɛ …, he implicitly communicates a/ an intermediate
91
degree of commitment to the proposition. The speaker therefore implicitly accepts the possibility that, the proposition is not true. When it is used as a deontic modal, it communicates what is reasonable to expect based on a set of laws (natural moral laws or laws in a particular jurisdiction) that are available. As an epistemic modal,
ɛwͻ sɛ … communicates what is reasonable to expect based on the facts that are available to the speaker. One thing to note is that whichever is prominent at a specific point in time solely depends on the context of use. For instance, example 131 was uttered on the premise that Ghanaian ministers of state are getting too many allowances and free things when the government is finding it difficult to pay the ordinary worker. The speaker, thus, communicates, in the light of the facts available (the suffering of the masses), that it is reasonable to expect ministers to sacrifice some of their benefits. This has an epistemic interpretation but in a different context, it could readily have a deontic interpretation. If the speaker had used ɛsɛ sɛ …, example 132 would have had different contextual implications. A statement with ɛsɛ sɛ … will have the contextual effect as in 1 below while an ɛwͻ sɛ statement will have the contextual effect in 2. 1. Ministers are being mandated to sacrifice some benefits. 2. Ministers are expected to sacrifice some benefits. This implies that in 1, it is a must that ministers do sacrifice some of their entitlements, but in 2, the proposition states that it would be good if they do it, or they are supposed make those sacrifices but no one can force them to do it since they are not necessarily obligated to do so. It should be noted that example 132 could also be interpreted deontically if there was a constitutional provision that
92
mandated ministers of state to make some of these sacrifices. 5.1.1.1 Emotional and Attitudinal verbs The next set of verbs that are discussed is referred to as emotional and attitudinal predicates by Ameka (2008). Verbs that can occur in this construction include hia ‘to need’, yɛ ‘to be good’ and fata ‘to deserve / to be fit for something or someone’. This set of constructions express deontic modality, and it is thus concerned with “the necessity of acts performed by morally responsible agents and thus with obligation and permission” Papafragou (1998). As is characteristic of deontic modality, these constructions are dependent on some kind of authority who desires to affect the situation. These constructions are therefore evaluative; that is, they evaluate the propositions embedded in the construction. One thing that needs to be made clear is that unlike the necessity and obligation verbs, these verbs contribute their independent meanings to the construction. On the other hand, they differ from the mental state verbs discussed in the previous section in that the mental state verbs can occur with any NP in subject position while these verbs have to appear in this fixed construction to have modal meaning. In the construction ɛhia sɛ…., the verb in the main clause hia means ‘to need’, ‘to be important’ or ‘to be necessary’. All these closely related meanings are prominent in the construction. The expletive 3SG can be replaced by the Akan word for body or self ‘ɛho’. 137. Ɛ- hia sɛ yɛn nyinaa bͻ yɛn ho mmͻden. 3SG need COMP 3PL all play ourselves effort ‘It is necessary that we all work hard.’ (Peace FM 20/11/13)
93
138. Ɛ- ho hia sɛ yɛ- hwɛ yɛn a- bͻdamfoͻ yie 3SG-body need COMP 3PL-look 3PL.POSS PL-mad.people well ‘It is necessary that we take care of our mental patients well.’ In example 137 the speaker expresses the notion that it is necessary for the addressees to work hard. He evaluates the proposition in the subordinate clause and then decides it is important or necessary that they achieve so. Apart from this, he also communicates in this context that ‘all of us working hard’ is desirable to him at the time of speaking. The addressees in the context of example 138 are supposed to interpret the proposition yɛhwɛ yɛn abͻdamfoͻ yie (i.e. taking good care of the mentally challenged as necessary and desirable to the speaker. The verb in ɛfata sɛ ... fata means ‘to deserve’. The deontic modal strength of this construction is a little lower than the meaning of ɛhia sɛ…. It shows that whatever proposition embedded in it deserves to be true. However it is not the case that the proposition is necessary as in the instance of ɛhia sɛ .... Rather, it states that it would be nice or good for the proposition embedded in it to be true or to take place. The examples below would throw more light on this assertion. 139. Ɛ- fata sɛ yɛ- kͻ kyea 3SG to.deserve COMP 3PL go greet ‘It would be nice if we go to greet them.’
wͻn. 3PL-OBJ.
140. Ɛ- fata sɛ yɛ – yi Nyame ayɛ. 3SG to.deserve COMPS 3PL- praise God adoration ‘It is good/deserving to praise God.’ In example 139 the speaker evaluates the proposition yɛkͻkyea wͻn and states that it is ‘deserving’ that they go and extend their greetings not that they are obligated to do that. In other words, if they refuse to do so, they are not going to be punished; they just want to go and demonstrate their kind gesture. Likewise, in
94
example 140 the speaker is expressing his views that God deserves to be praised. He does not say that it is the case that when we do not praise God we have sinned, rather we have to show our appreciation to let Him know that we appreciate what He has done for us. The propositions evaluated under these constructions are seen as things that need to be done by ‘morally upright’ people. The last construction of this type is ɛyɛ sɛ … . The verb in the main clause, yɛ, means ‘to be good’. On the scale of emotional and attitudinal constructions, ɛyɛ sɛ … is the lowest. Ɛyɛ sɛ… evaluates a proposition as being good at the time of speaking. The verb yɛ cannot co-occur with any other type of NP in subject position except the expletive. 141. Ɛ- yɛ sɛ yɛ- bɛ- tena asomdwoe mu. 3SG good COMP 3PL MOD- stay peace inside. ‘It is good we live in peace.’ (Forosie pp 23) 142. Abofra deɛ ɛyɛ sɛ wo- bɛ- tie Child TM 3SG- good COMP 2SG- MOD-listen wo awofo asɛm. 2SG parent matter ‘As a child, it is good to listen to your parents.’ (Bediako pp 45) In example 141, the speaker states that it is good that they live together in peace. The speaker in no way asserts that he demands that to happen. What he states is just his desire and the addressees might even disagree and more importantly it might not happen. By using this modal construction, the speaker communicates the additional meaning that there is another means of living together which may not to ideal to him but is also a possibility. Likewise, in 142, the speaker just indicates how good it is for a child to be obedient to his parents. It does not necessarily mean the child is going to be obedient to his parents.
95
5.1.2. 3SG + bɛ + verb + Sɛ Clause Only two verbs can occur in the verb slot of this modal construction. These are hia ‘to need’ and yɛ ‘to be good’. They both evaluate the propositions embedded in them. They therefore express deontic modality. Let us consider the examples below. The only difference between this construction and the one already discussed above is the presence of the modal prefix bɛ. 143. Ɛ- ho bɛ- hia sɛ yɛ- bͻ yɛn ho mmͻden. 3SG body MOD-need COMP 3PL-play ourselves effort ‘It would be necessary that we work hard.’ 144. Ɛ - bɛ- yɛ sɛ yɛ- bɛ- hu no. 3SG- MOD-be.good COMP 1PL- MOD-see 3SG OBJ ‘It would be good that we see him/her.’ 145. Kwame, e- be- ye na e- be- hia nso se Kwame, it-FUT-be.good CONJ it- FUT-necessary too COMP ye -be- nya abotre ahwehwe mu yie, we-FUT- receive patience PERF- search inside well, na` ye-de nokware paa e-wo mu no a- to dwa CONJ we-take truth INT it-be inside DEF PERF-put public.’ ‘Kwame, it would be good and necessary for us to be patient so that we can investigate (it) properly, and then we make available the real truth to the public.’ (RTR) (Amfo 2007:374) As discussed in the preceding chapter (see section 4.3.2), bɛ- in these constructions does not express future tense but rather modality. Again, these constructions express deontic modality since they are all pieces of advice given by the speakers at different point in time. For instance, in 144 the speaker encourages or advises his colleagues that they should work hard. In 144 the speaker suggests to his colleagues that it would be necessary for them to see a particular person, and in 145, the speaker advises the addressee that it is necessary that they take their time to do
96
a proper investigation in order to come out with the truth. In fact, the combination of the modal prefix, bɛ, and the verbs hia and yɛ makes the modal meaning more prominent.
5.1.3
NP + wͻ + NP + Sɛ Clause
The verb and the second NP positions in this modal construction are fixed. Only the wͻ ‘to have’ can fill the verb slot. The first NP can be filled with all +human nouns and pronouns for instance, Kofi, Ama, ͻ ‘3SG’ and me ‘1SG’. The second NP slot is filled by the nouns; anidasoͻ, awerehyɛmu, gyedie and ahotosoͻ. The CC contains the proposition that is being modalized. This construction expresses epistemic modality as they show a speaker’s commitment to the proposition expressed in the CC. 146. Ghanafoͻ nnyina wͻ awerehyiamu sɛ Black Stars Ghanians all to.have hope COMP Black Stars bɛ- yɛ adeɛ World Cup yi mu. MOD-do thing World Cup this in ‘All Ghanaians are hopeful that the Black Stars will perform well in this World Cup.’ 147. Kristofoͻ wͻ gyedie sɛ Nyankopͻn yɛ Nyame. Christians to.have believe COMP God be God ‘Christians have the belief that God is really God.’ 148. Me- wͻ ahotosoͻ sɛ me asetena mu bɛ- sesa. 1SG- to.have hope COMP 1SG living in MOD-change ‘I am hopeful that my living conditions will change’
97
5.2
Modal Concord in Akan
The main focus of this section is to explore the concept of modal concord in Akan. As discussed in section 2.1.5 of chapter two, modal concord is concerned with how the meanings of two modal markers occurring in the same clause converge. Though the phenomenon has been discussed in earlier works, it was Geurts & Huitink (2006) who coined the term ‘Modal Concord (MC)’. They explained that there are two main constraints on MC. These are as follows: 1. two expressions can only participate in a concord construction if they are of the same modal type (i.e. if they are both deontic, epistemic, etc). 2. the modals involved have to have the same, or at least similar, quantificational force (i.e. the modals should both express certainty or possibility and not a situation where one expresses certainty and another possibility) (Geurts & Huitink 2006:3-4). We speak of modal concord if the combination of a modal adverb with a modal auxiliary seems to be interpreted as if it was just a single modal operator expressed (Zeijlstra 2007). With this in mind, we will look at some examples in Akan in order to find out if there is the possibility for more than one modal marker to occur in an utterance. Secondly, we would like to ascertain whether when this occurs there is meaning convergence or a cumulative reading; paying particular attention to the fact there are no modal auxiliaries in Akan. The last thing we would investigate is which modal strategies are involved when modal meanings converge. The examples below would help us to answer these questions. 149. Kwame Amponsa bɛ - tumi adi saa dwuma Kwame Amponsa MOD-be able CONS- do this work. no ama wͻn.
98
DET for 2PL ‘Kwame Amposa can do that work for them.’
(Bediako pp 33)
150. Sɛ ɛ- yɛ nokware ampa sɛ obi afa wo sika If 3SG be true true COMP. someone PERF-take 2SG money a, dɛn nti na w’ anka anCOND., why so CONJ 2SG PERF.-NEG talk PERF. NEG. kyerɛ polisifoͻ no. tell policemen DET. ‘If it is really true that someone has stolen your money, why didn’t you inform the police about it?’
(Bediako pp12)
151. Ɛ- wͻsɛ wo- tumi twɛn Nyame anim kakra. 3SG-have.to COMP 2SG- be able to.wait God face small ‘You should be able to wait upon the Lord.’ 152. Gyama ͻ- yare ampa. Probably 3SG to.be .sick truly ‘He might truly be sick.’ 153. Ebia na ͻ - bɛtumi adi aduane no nyinaa. Maybe PST 2SG MOD.- be.able.to CONS.-eat food DET all ‘Maybe, he can eat all the food.’ In examples 149 to 153, there are two or three modal markers in each construction. In 149, there is the modal prefix bɛ- and the modal auxiliary tumi. In 150, the modals are the adjective nokware and the adverb ampa. Example 151 contains the modal construction ɛ-wͻ sɛ… and the modal auxiliary tumi. The two modals present in example 152 are the adverbs, gyama and ampa. In example 153, there are three modal markers; ebia, bɛ- and tumi. These examples show that it is possible for more than one modal marker to co-occur in a sentence in Akan, and so our first question is answered.
99
We now turn our attention to whether or not there is MC in each of these constructions. Knowing what we know about bɛ- and tumi now, they are compatible with epistemic, deontic and dynamic meanings depending on the context in which they are used. If we therefore would assume a dynamic modality reading for both bɛ-and tumi in 149, the interpretation will be that the subject has the ability to perform the action denoted in the VP as in 1 below. 1.
Kwame Amponsah has the ability to do the work for you.
We notice the modals do not maintain separate meanings but their markers converge and hence we have a case of MC. But we interpret bɛ as an epistemic modal and tumi as a dynamic modal; the interpretation we get is the one in 2. 2.
Kwame Amponsah possibly has the ability to do the work for you.
This interpretation is not a case of MC reading because it violates Geurts & Huitink’s (2006) first constraint on MC; the modals should be of the same modal type. If the two meanings do not converge and we get a cumulative meaning; each modal contributes its own meaning to the proposition. The reading sound somewhat as weird as what we have in 3. 3. Kwame Amponsah has the ability, ability to do this work for you. Though the adverbs in 150 are both epistemic modals thus they not have the same quantificational force and are they are both certainty modals, they do not have a concord reading. A careful look at ampa in that utterance shows that it functions as a strengthener and not as a modal, and so does not contribute any modal meaning to the sentence. In example 151 gyama and ampa are of different quantificational force; ampa is a certainty modal while gyama is a possibility modal.
100
With regards to the modal markers that are involved in MC, Geurts and Huitink (2006) assert that MC is possible only between modal auxiliaries and adverbs. Zeijlstra (2007) adds that MC is also possible between modal auxiliaries and affixes. Zeijlstra (2008) also argues that MC is a grammatical phenomenon and so we expect that to have an MC, one of the modal markers should be in the verbal group. This presupposes that the presence of two adverbs that satisfy the two constraints specified earlier on would still not have an MC reading. Following their argument, examples 149 above, 154 155 and 156 below are typical examples modal concord in Akan. 154. Ebia, ͻbɛba. Maybe 3SG MOD come ‘Maybe, he might come.’ 155. Ebia, ͻ- bɛtumi aba. Maybe 3SG MOD.- be.able to CONS.- come ‘Maybe, he might be able to come.’ 156. Gyama ͻ- bɛ- ba. Perhaps 3SG-MOD come ‘Perhaps, he might come. As we have already seen, bɛ has three modal meanings, and one of them is compatible with ebia ‘maybe’ or gyama ‘perhaps’; this is the epistemic meaning. This example according to Geurts & Huitink (2006) and Zeijlstra (2007)’s classification will be a perfect instance of MC in Akan. In the first instance, the modals are of the same type; they are both epistemic modal markers and they also have the same quantificational force; that is, they both communicate possibility. More importantly, it involves a modal adverb and a functional affix. So far we have been able to throw light on the strategies Akan, which do not
101
have an extensive system of modal auxiliaries, employ to express MC. We notice that because Akan does not have an elaborate system of modal auxiliaries, the modal prefix bɛ- and the ability modal tumi are mostly used in structures that show MC. The affix bɛ- can also co-occur with some modal adverbs to express MC. Also the modal auxiliary, tumi, when used together with a modal adverb can also have an MC interpretation. In all these strategies, one of the modal has to be in the verbal cluster, emphasizing the fact that modal concord is grammatical. 5.3
Modal markers in Akan as politeness markers.
In this part of the thesis, we look at some of the other roles Akan modal expressions perform apart from their primary roles. We will thus explore the use of modals as politeness markers. As Lapeyre (1992) points out in English modal auxiliaries, there are instances of the use of some modal markers in Akan, which are only due to politeness, and not the expression of lack of absolute knowledge. Downing and Locke (1992: 382) also add that even modal certainty is diminished certainty because the unmodalized constitutes a stronger statement of fact. They argue that a modalized statement might therefore be used for reasons of politeness. In this section we adopt Brown and Levinson’s (1987) politeness model as a theoretical framework.
5.3.1 Politeness Politeness is the practical application of good manners and etiquette. It is a culturally defined phenomenon and thus what might be polite in one society may be impolite in another. Penelope Brown and Stephen Levison formulated the
102
politeness theory. In their theory, they adopt Goffman’s (1955) concept of face. Brown and Levinson (1987: 66), define face as a person’s public self-image, “something that is emotionally invested, and that can be lost, maintained, or enhanced, and must be constantly attended to in interaction”. Mills (2003:6) thus defines politeness as “the expression of the speaker’s intention to mitigate face threats carried by certain face threatening acts toward another”. Brown and Levinson (1987) differentiates between two types of face: the negative face - the person’s desire to preserve her/his freedom of acting (‘territory and self determination’), and the positive face - the person’s desire to see her/his self-image respected by others. According to these authors, during conversations, face maintenance and redress is more important than information exchanges. Politeness strategies have the function of avoiding Face Threatening Acts (FTAs) or of diminishing their consequences. FTAs are acts that are inherently meant to damage the face of the addressee or speaker by acting in opposition of the wants or desires of the other. In their framework, they outlined four strategies that are used to show politeness. These strategies are the bald on-record, negative politeness, positive politeness and off-record (see Brown and Levinson 1987). In this section of the thesis, we shall discuss the use of modals in two strategies; negative politeness and positive politeness. Positive politeness strategies aim at minimizing the threat to the hearer’s positive face; they do this by making the hearer feel good about himself/herself. Negative politeness strategies are oriented towards the speaker’s negative face and it emphasizes avoidance of imposition on the hearer. These two politeness strategies also have sub strategies. We will look at the sub strategies and the modals that are used to achieve them.
103
5.3.1.1 Negative politeness strategies a.
Be conventionally indirect; requesting an addressee to perform an action is a
threat to the hearers negative face, to mitigate this threat the request can be made indirectly. This indirect means sort of gives the addressee the option to perform the action or not. This strategy mainly involves making one’s request in the form of questions. In English, these questions are usually preceded by the modal auxiliaries, can, will and would. 157. Can you pass the salt to me? Two main modals are used to achieve negative politeness in Akan, the use of anka and the combination of the modals bɛ and tumi. As elucidated by Amfo (2005), anka is used in marking polite requests. She asserts that “that Akan anka should be considered as such a conventional indirect request marking device; anka is a sort of hedge which tones down an otherwise direct request…anka is a semantic device used to mitigate the directness of the request”. The combination of the modal affix bɛ and the ability modal tumi also tones down the threat to the negative face of the addressee during requests. Power is put into the addressee’s hands to perform or not perform the action by the fact that the speaker ask if he is willing or has the ability to do it. 158. Wo- bɛ- tumi ama wo mu 2SG- MOD-be. able.to PERF- rise 2SG inside ‘Could you get up for me?’
so ama me? up PERF. give 1SG
159. Sɛ ɛ- bɛtumi a, anka yɛ- n- nyae. If 3SG-MOD- be.able.to COND, MOD. PL-OPT- stop ‘If possible, let us stop’
104
160. Wo- tumi ka wo ano ato mu 2SG- be.able.to touch 2SG.POSS mouth PERF- close- in anka ɛ- bɛboa yɛn MOD. 3SG-MOD- help 3PL.OBJ ‘If you keep quiet, it will help’
b. Don’t presume or assume: The speaker usually gives a directive but disguises it in such a way that it does not impose on the addressee. With this strategy, the speaker uses a more indirect approach than the first one. He presents the directive as a thought or suggestion, or does not show absolute knowledge of the proposition. 161. Me- dwene sɛ wo som bra saa suban yi. 1SG- think COMP 2SG religion ban this character DET ‘I thought your religion is against this character.’ 162. Me nim sɛ wo maame se gyae. 1SG know COMP 2SG mother say stop ‘I thought your mother said you should stop.’ 163. Ebia ɛnnɛ ne ɛda a edi kan a w’ aba ha. Maybe today is day cond. eat first cond, 2SG Perf.- come here ‘Maybe, today is your first time here.’ In example 161, what the speaker wants to communicate is simply ‘stop this habit’ however he presents the decision not as his own but as a requirement of religion. Also, he chooses to use a non-committal modal ‘dwene’ distancing him from the proposition. In 162, the speaker communicates ‘stop what you are doing’ but presents it as a directive from the addressee’s mother. In this sentence the speaker uses ‘know’ which shows a high level of commitment to the proposition. The choice of nim ‘know’ depends on how sensitive the issue being discussed is. In example 161 for instance, the speaker was addressing a Muslim
105
who drinks excessively while the speaker in 162 is addressing a child who is misbehaving. In example 163, the speaker expresses lack of absolute knowledge using the uncertainty modal ebia. A church secretary made this statement as he invites people who are visiting the church for the first time to come in front of the church. In this context the speaker is aware that it is the first time for anybody who gets up but because he does not want to impose on the addressee, he uses the modal ebia. c.
Avoid disagreement: this strategy is usually used when the addressee is of
higher status. Givon (1990:823) explains, “in communicating to an interlocutor of higher status, one downgrades one’s own subjective certainty”. This means that even if the speaker is certain about the proposition expressed, he presents the information as if he is not. The examples below explain it further. 164. Ebia wo na w’ anhu Maybe 2SG FOC. 2SG PERF.-NEG- see ‘Probably you didn’t see her.’ 165. Me- dwene sɛ ebia wo- hu –u me enti na 1SG-think COMP maybe 2SG see-PST 1SG so CONJ wo- sͻre -e . 2SG- get.up -PST ‘I thought you got up becuase you saw me.’ For context in which 164 w a s m a d e w a s , an older person is complaining about a worker who he claims does not come to work regularly, citing the previous day as an example. The speaker is a colleague of this person who knows the guy was at work the previous day. What the speaker wants to communicate is ‘the person came to work, it is you who did not see him’. However, a statement like this will be considered impolite and thus he tones it down by using the
106
uncertainty modal adverb ebia. In example 165, the speaker uses two modal expressions that show non-comittal and uncertainty. The simultaneous use of these two expressions tones down the speaker’s certainty to its minimal level. Even if the speaker knows for a fact that the addressee got up because of him, he does not presuppose it.
5.3.1.2 Positive Politeness strategies Of the strategies that are used to show positive politeness, only one of them appears to include the use of a modal in Akan. a.
Be vague: the speaker gives information as if he is not sure about the
proposition expressed. This involves the use of hedging strategies like ‘almost’, ‘appears’ and ‘seems’ in English. In Akan speakers usually used the seemly similar modal expressions ɛbeyɛ sɛ and bɛyɛ. When a speaker uses the expression ɛbeyɛ sɛ as in example 166 he communicates that he might not have enough evidence to make the assertion he is making. Or they could just be distancing themselves form the proposition; the truth could be true or false, do not hold me accountable for it. Speakers usually use this strategy when they communicate sensitive issues or something negative about others. 166. It appears the student didn’t take the exam. 167. Ɛbɛ- yɛ sɛ neho m- fa no. 3SG- MOD do COMP 3SG.POSS- self NEG- take 3SG-OBJ ‘It seems he is not feeling well.’ 168. Wo m- ma me bɛyɛ mmienu bi 2SG NEG-give 1SG about two some ‘Won’t you give me about two’
107
Bɛyɛ ‘about’ on the other hand is used with concepts that deal with numbers for example time, age and count nouns. When a speaker uses bɛyɛ, he implicitly states that what he actually wants is more than the number quoted. The main function of this marker is to mitigate the face threat of a request by reducing the imposition on the addressee. In being vague, the speaker gives the addressee the choice to decide on the quantity he wants to give out. However, it must be made clear that the speaker expects more than the number quoted. Any addressee who gives exactly that number is considered to be either communicatively incompetent or a miser. Sometimes, speakers intentionally flaunt all these politeness strategies and use abusive language. The use of abusive languages is seen as impolite. Impoliteness according to Culpeper et al. (2003:1546) is a “communicative strategy designed to attack face, and thereby cause social disharmony and conflict”. When an interlocutor chooses to be impolite, he disregard his own face and also attack the face of another. Even in such situations, some speakers still try to mitigate certain clearly impolite utterances. In the examples below, the speakers use some modals to tone down the effect of certain abusive utterances. 169. Gyama w’ adwene mu ka wo kakra? Perhaps 2SG mind in itch 2SG small ‘Perhaps you are insane.’ 170. Anyebiara wo- yɛ papa deɛ? Perhaps 2SG-do good TM ‘Are you sure you are alright?’ 171. Ebia yɛ- wo- o no no, ne ti bͻ-ͻ fam Maybe 3PL-born-PST DET CFM,3SG-POSS head hit- pst floor ‘Maybe she hit her head on the ground when she was born.’ (i.e. she is
108
crazy)’ 172. Me dwene sɛ wo nim nyansa 1SG think COMP 2SG know sense ‘I thought you were sensible.’ In examples 169 to 172, the speakers use the uncertainty modal adverbs to tone down the invectives they express in the sentences. Though examples 168 and 169 are presented in interrogative form they do not solicit for information but rather used to make assertions. As already stated in the discussion, when dwene it is used it presupposes that the proposition embedded in it is false at the time of speaking. The sentence in 171 therefore clearly implies ‘you are not sensible’, but this utterance is blunt and thus a serious face threat, therefore the speaker avoids it.
5.4
Chapter Summary
Three main topics were discussed in this chapter. In the first section, we looked at the periphrastic modal in Akan. We discussed three such constructions: 3SG + verb + Complement Clause, 3SG + bɛ + verb + CC and NP + wͻ + NP + CC. In the second section, we discuss the notion of modal concord (Geurts & Huitink 2006) in Akan. The function category modal bɛ appears to be compatible with most of the other modals. It co-occurs with tumi than all the other modals. When they co-occur, their meanings converge, demonstrating instances of modal concord. Bɛ also co-occurs with the modal adverbs. In such instances, we may get a modal concord interpretation or a cumulative reading depending on the context. In the third section, we discussed the use of some modal markers as politeness strategies.
109
We notice that in certain contexts some modals are only used to mitigate face threatening acts. We discussed Akan modals as politeness markers in two of Brown and Levinson (1987) politeness theory: Positive politeness and negative politenes.
110
CHAPTER SIX CONCLUSION
6.0
Overview
This thesis contributes to the cross linguistic study of modality. The study explored how modality is expressed in Akan, a Kwa language spoken in Ghana. In the thesis, the linguistics strategies employed to encode modality were divided into three groups: lexical categories, modals in the verbal group and the periphrastic modal constructions. The thesis also includes a discussion on modal concord and the use of modals as politeness markers in Akan. In this chapter, we present a summary of the thesis, a discussion of the findings and recommendations for further research.
6.1
Summary
The concept of modality in language was explained in the background of the study in chapter one. Attention was drawn to the fact that the concept is very difficult to describe. Chapter one also talked about the types of modality distinguished in the literature and also the linguistic devices that are used to encode this phenomenon cross linguistically. In addition, the gap in the literature, which this research seeks to fill, was presented in ‘the statement of the problem’. Again, the research questions that guided the study, the objectives of the research, the significance of the study to the academic community as well as the chapter synopsis were all presented in the first chapter. Also in the chapter, the significance
111
of the study to the academic community is presented. Lastly, the chapter presents how the rest of the thesis is organized. Chapter Two is divided into three parts; literature review, theoretical framework and methods. The literature review section discussed relevant studies that have been conducted on modality cross linguistically. This section was divided under the following headings: modal auxiliaries, the future marking and modality, modal adverbs, adjectives and studies on the modal system of languages. Literature on modal concord and other uses of modal expressions were also reviewed. The second section discussed Sperber and Wilson’s (1988, 1995) theoretical framework ‘Relevance Theory’ which is the framework in which the study is situated. It focused on the ‘Procedural verses Conceptual’ meaning distinction made by Blakemore (1987) within the relevance theoretic framework. The third section discussed the methods, which are employed in the collection and analysis of data for this study. The third, fourth and fifth chapters presented the analyses and discussion of data. The discussion in the third chapter was centered on the lexical categories that are used to express modality in Akan. Two word categories were surveyed, and these are verbs and adverbs. The fourth chapter explored modal markers in the Akan verbal group. The modal status of the auxiliary verb tumi and the prefix bε were examined in this chapter. In the fifth chapter, which is also the final analysis chapter, three main topics were discussed. The first one explored is the use of periphrastic modal constructions to express modality in Akan. The second topic is the concept of modal concord, a term coined by Geurts and Huitink (2006). It refers to the convergence of the meanings of two modal markers in an utterance.
112
The third section explored the use of some modal marker as politeness markers. 6.2
Findings
This section discusses the major findings of the study following the discussion of data in chapter three, four and five. The findings will be grouped under five main sub-headings: lexical categories that express modality, functional categories that express modality, periphrastic modal constructions, modal concord and modality and politeness.
6.2.1 Lexical categories that express modality Two of the lexical categories that can express modality in Akan were discussed. These are mental state verbs: factive and attitude verbs and modal epistemic adverbs. The following came to light in the course of the discussion: The verb nim communicates to the addressee that the speaker has enough information to make the statement or assertion he is making. It has the epistemic meaning of absolute certainty. Gye di on the other hand communicates to the addressee that, the speaker has j u s t enough information to make the statement or assertion he is making. Thus, it also has the epistemic meaning of certainty. The verb susu is ambiguous between two meanings, ‘to assume’ and ‘to suggest’. It is disambiguated in the context of use and also the tense/aspect markers it takes. When it means ‘to suggest’, the speaker implies that the interlocutor could perform the action described in the embedded clause. In that instance, it is presented as a suggestion only. The verb in the
113
embedded clause can only be in the imperative and susu can inflect for progressive, future and past. The other meaning of Susu, ‘to assume’, communicates a non-committal attitude to the proposition in the embedded clause. The verbs in the embedded clause can take any tense/aspect marker but susu cannot inflect for tense or aspect. Dwene communicates to the interlocutor that the proposition in the embedded clause is only a thought that is being expressed by the speaker. The table below shows the hierarchy of Akan mental state verbs. Verbs
Speakers’ attitude
Nim ‘to know’
Certainty
Gye di ‘to believe’
Certainty (weaker)
Susu ‘to assume’
Probability
Dwene ‘to think’
Lower probability
Susu ‘to suggest’
Suggestion
Epistemic adverbs that express modality are grouped into two, the certainty adverbs and the uncertainty adverbs. There are two certainty adverbs in Akan, nokware ‘truly/honestly’ and ampa ‘truly/honestly’. These adverbs are synonymous. The only difference between them is syntactic, nokware cannot occur at clause final position. The use of any of these two adverbs suggests that the speaker is certain, that the proposition, which is in the scope of the adverb, is true at the time
114
of speaking. Apart from being attitude markers, these adverbs are also used as strengtheners. As strengtheners they are used to place emphasis on the proposition a sentence expresses. Akan has at least five uncertainty epistemic adverbs which comprise ebia ‘maybe’, gyama ‘perhaps’, anhwe a ‘probably’ sɛsɛɛ ‘by this time’ and annyɛbiara ‘probaby’. All these adverbs occur at only clause initial position. The use of ebia shows that the speaker is not certain about the state of affairs described in the truth conditions of the proposition. Ebia and gyama are synonymous. However, gyama can be used to ask a question but ebia cannot be used that way. Sɛsɛɛ signals that there is a possibility or probability that the proposition embedded in it, is true. The use of anhwɛ a signals the addressee to interpret the proposition expressed as being potentially true or false at the time of speaking. Annyɛbiara expresses the potential that a proposition expressed is true at the time of speaking. It can also be used to form questions. 6.2.2 Modals in the verbal group In the Akan verbal group, as highlighted in chapter four, apart from the mood markers discussed by Boadi (2008), modality is marked by the auxiliary verb tumi ‘be able to’ and the modal prefix bɛ-. The following claims are true about the auxiliary, tumi and the verbal prefix, bɛ- as evidenced in the discussion in chapter four: When used in a sentence, tumi shows that an NP has a characteristic or ability to do an action described by a lexical verb it occurs with. It is used
115
in a way similar to the use of can in English as discussed by Palmer 2003. As an ability modal, tumi’s primary modal meaning is dynamic modality. However in some contexts, it also conveys epistemic and deontic modality. Also, though in previous studies in Akan, the verbal prefix bɛ- had been referred to as the future marker, the data revealed in this thesis that in most instances, when this prefix is used, the meaning it contributes is a modal meaning and not just the future tense. Even in instances where the future interpretation seems t o b e t h e m o s t prominent interpretation available, there is always an underlying modal interpretation. But there were instances where only a modal interpretation was deduced from the prefix. As a modal, bɛ-expresses three types of modal meanings, epistemic modality, deontic modality and dynamic modality. 6.2.3 The Periphrastic Modal Constructions It was again, evident from the data that one of the strategies used to express modality in Akan is the use of constructions as discussed in chapter five. The following findings were also made: It came up that these constructions are of three types; 3SG + verb+ Comps clause, 3SG + bɛ+ verb+ Comps clause and NP + wͻ + NP + Comps clause. The 3SG in the first two constructions is an expletive or the dummy it. Only a few verbs in Akan can occupy the verb position in these structures. They include hia, yɛ, sɛ, fata and wͻ. These verbs are grouped into two,
116
necessity and obligation verbs, and emotional and attitudinal predicates. The necessity and obligation verbs include wͻ ‘to be located at’ and sɛ ‘to be fit’ whiles the emotional and attitudinal predicates are hia ‘to need’, yɛ ‘to be good’ and fata ‘to deserve/to be fit for something or someone to do something’. The first construction performs most of the roles that the modal auxiliaries must, should, ought and might in English perform. They are compatible with all the modal meanings the English modal auxiliaries have. The verbs in these constructions are evaluation verbs so they evaluate the propositions embedded in them. They therefore express deontic modality.
In the third structure, the first NP could be any [+ human] noun but the second NP is subcategorized for only these nouns, awerehyɛmu, anidaso, ahotosoͻ and gyedie. This third construction expresses epistemic modality as they show a speaker’s commitment to the proposition expressed in the complement clause.
6.2.4
Modal Concord
Modal concord is concerned with how the meanings of two modal markers cooccurring in the same clause converge. As stated by Geurts and & Huitink (2006), modal concord is only possible if their modal flavor is the same. In relation to modal concord, the following findings were emerged from the data: Bɛ-and tumi usually co-occur in sentences. When they do, there is always a convergence of meaning. The dynamic meaning of bɛ-is merged with the dynamic meaning of tumi.
117
The epistemic modal adverbs can also co-occur with the affix bɛ-. When this happens, it is the epistemic meaning of bɛ- that is merged with the meaning of the adverb. Though it is possible for the modal adverbs, tumi and bɛ- to co-occur with the periphrastic modal construction, their meanings do not converge. The data also sheds light on strategies that languages which do not have an extensive system of modal auxiliaries employ to express MC in the language. It supports Zeijlstra (2007) and Geurts and Huitink’s (2006) assertion that the modals would have to be of the same flavor to have a concord interpretation. It however challenges Geurts and Huitink (2006) claim that modal concord is only possible between modal adverbs and modal auxiliaries.
6.2.5
Modal markers as politeness markers.
In certain contexts, modals are used only to mitigate face-threatening acts during conversation. In Brown and Levinson’s (1987) politeness theory, we identified two main strategies that employ the use of modals as politeness markers: positive politeness and negative politeness. Positive politeness strategies aim at minimizing the threat to the hearer’s positive face while negative politeness strategies are oriented towards the speaker’s negative face and emphasize avoidance of imposition on the hearer. Only one positive politeness strategy was
118
found that employed a modal in Akan. This strategy states that ‘be vague’. Being vague involves the use of hedging strategies like ‘almost’, ‘appears’ and ‘seems’ in English. In Akan, the modal that is used is ebeyɛ (sɛ). Three negative politeness strategies were found that involve the use of modals. These strategies and the modals used are: a. Be conventionally indirect; tumi alone or a combination of bɛ and tumi b. Don’t presume or assume: dwene, nim and ebia c. Avoid disagreement: ebia and dwene There are certain instances when a speaker uses a clearly impolite utterance but mitigates its force by the use of some uncertainty modal adverbs. For instance, invectives can be preceded by gyama or anyebiara.
6.3. Recommendations It has been revealed in this study that apart from the modal adverbs and modal adjectives that are used to express modality in Akan, there are other strategies such as the use of the auxiliary verb tumi, the verbal prefix bɛ, and some verbs such as nim, dwene and susu as well as some other constructions. Similar research could therefore be conducted in other Ghanaian Languages to bring to the fore the strategies that are used to express modality in those languages as well. Again, it came out in the discussion that the verbal prefix bɛ- has both tense and modal readings in Akan; the most prominent among the two is the modal reading. Therefore, scholars should consider glossing bɛ- as a modal marker rather than a future tense marker since whenever it has a future meaning, a modal interpretation is possible, but the reverse is not the case.
119
Similar research should also be conducted in other Ghanaian languages on the future marker in the language in order to come out with the actual status on the future marker in these languages. This is important for cross-linguistic studies, providing ample evidence for the characterization of these markers in African languages.
120
REFERENCES Aboh Enoch O. and Nauze Fabrice. 2008. Tense, mood, and aspect in Gungbe (Kwa). In Werner Abraham and Elisabeth Leiss (eds.), ModalityAspect Interfaces. Implications and typological solutions.Vol.79 Abraham, Werner. 2002. Modal verbs. Epistemics in German and English. In Barbiers, Sjef, Frits Beukema & Wim van der Wurff (eds.), Modality and its interaction
with
the
verbal
system,
pp.18–50.
Amsterdam:
John
Benjamins. Abraham, Werner. 2005b. Event arguments and modal verbs. In Maienborn, Claudia & Angelika Wollstein (eds.), Event arguments. Foundations and applications, pp.243–276. Tubingen: Niemeyer. Ameka, Felix K. 2005c. Multiverb constructions on the West African littoral: microvariation and areal typology. In M. Vulchanova and T. A. Åfarli (eds.), Grammar and beyond: Essays in honour of Lars Hellan, 15-42. Oslo: Novus Press. Ameka, Felix K. 2008. Aspect and modality in Ewe: a survey. In Kropp Dakubu and Ameka Felix K. (eds.) Aspect and Modality in Kwa Languages, pp. Amsterdam: John Benjamins Amfo, Nana Aba A. 2005. Modal marking in Akan: the case of ‘anka’. Journal of Pragmatics 37, 997-1013. Amfo, Nana Aba Appiah 2007. Clausal conjunction in Akan. Lingua 117,666–684 Amfo, Nana Aba Appiah 2007. Intrepretive Use Marking in Akan. In Randi A. A. Nilsen, Nana Aba Appiah Amfo and Kaja Borthens (eds.),
121
Interpreting utterances: Pragmatics and Its Interfaces, Essays in honour of Thorstein Fretheim. Oslo: Novus 108-126 Amfo, Nana Aba. Appiah. 2010. Recurrence marking in Akan. Pragmatics, 15(2). Berbeira-Gardón, José Luis 2006. The Semantics and Pragmantics of Will. Pragmaling istica 3-4, 223-240 Blakemore, Diane. 1987. Semantic constraints on relevance. Blackwell Publishin Blakemore, Diane 2002. Relevance and linguistic meaning. The semantics and pragmatics of discourse markers. Cambridge studies in Linguistics 99, Cambridge; Cambridge University Press. Boadi, Lawrence .A. 2008 Tense, Aspect and Mood in Akan. In Kropp Dakubu and Ameka Felix K. (eds.) Aspect and Modality in Kwa Languages, pp. 9-68 Amsterdam: John Benjamins Boicu, Ruxandra. 2007. Modal verbs and politeness strategies in political discourse University of Bucharest, Faculty of Journalism and Communication Studies Online at http://mpra.ub.uni-muenchen.de/45913/ MPRA Paper No. 45913, posted 7. April 2013 02:28 UTCBrandt, Soren. 1999. Modal verbs in Danish (Travaux du cercle linguistique de Copenhague 30).Copenhagen: C. A. Reitzel. Brunner, Gisela & Angelika Redder. 1983. Studien zur Verwendung der Modalverben. Tubingen: Gunter Narr. Butler, Johnny. 2003. A minimalist treatment of modality. Lingua 113, 967–996 Bybee, Joan. 1985. Morphology: A study of the relation between meaning
122
and form. Amsterdam; Benjamins. Bybee, Joan and Suzzane Fleischman (eds) 1995. Modality in Grammar and
Discourse.
Amsterdam/Philadelphia:
John
Benjamins
Publishing house. Bybee, Joan and Revere Perkins, and William Pagliuca.1994. The Evolution of Grammar: Tense, Aspect, and Modality in the Languages of the World. Chicago: University of Chicago Press. Carnie Andrew 2006. Syntax: A Generative Introduction. 2nd edition. Oxford: Wiley Blackwell Publishers. Carston, Robyn. 1998. The semantics/pragmatics distinction: a view from relevance theory.UCL Working Papers in Linguistics 10 Cinque, Guglielmo. 1998. Adverbs and the Functional Hierarchy of Functional projections. New York: Oxford University Press Cinque, Guglielmo. 1999. Adverbs and functional heads. Oxford: OUP. Comrie, Bernard. 1985. Tense. Cambridge Textbooks in Linguistics. Cambridge: Cambridge University Press. Cook, Walter A. 1978. Semantic Structure of English Modals. TESOL Quarterly 12:5-15 Cresswell, J. 1994. Research design: Qualitative and qualitative research. Newbury Park, CA: Sage Crystal, David. 1997. A Dictionary of linguistics and phonetics, fourth edition. Oxford Uk: Blackwel Publishing. Dolphyne, Florence. Abena.1988. The Akan (Twi-Fante) language: its
123
sound systems and tonal structure. Accra: Universities of Ghana Press Coates, Jennifer 1983. The Semantics of the Modal Auxiliaries. London: Croom Helm Downing, Angela and Locke, Philip. 2006. English Grammar: A University Course. 2nd ed. Abingdon and New York : Routledge Press Enҫ, Murvet. 1996. Tense and Modality. In Shalom, Lappin (ed.), The handbook of contemporary semantic theory, pp. 345–358. London: Blackwell. Ernst, Thomas. 2000. Semantic Features and the Distribution of Adverbs. In Fabricius-Hansen C., Lang E. and Maienborn C. (eds.), Approaching the Grammar of Adjuncts. ZAS Papers in Linguistics 17: 79-99. Essegbey, James. 2008. The potential morpheme in Ewe. In KroppDakubu and Ameka
Felix
K.
(eds.)
Aspect
and
Modality
in
Kwa
Languages. Amsterdam: John Benjamins. Fasold, W. Ralph. & Connor-Linton Jeff. 2006. An introduction to language and linguistics. Cambridge: Cambridge University Press. Ferdinand de Haan.2006. Typological approaches to modality. In Wolfgang Klein & Stephen Levinson (eds.), The Expression of Cognitive Categories, pp. 27-65. Berlin · New York: Mouton de Gruyter. Foley, William and Robert Van Valin. 1984. Functional syntax and universal grammar. Cambridge: Cambridge University Press.
124
García Gonzálvez, Francisco. 2000. Modulating Grammar through Modality: A Discourse Approach. ELIA I:119-133. Gbegble, Nada. 2012. Modality in Ewe: A Functional Exploration of Epistemic Adverbs. In Selected Proceedings of the 42nd Annual Conference on African Linguistics, ed. Michael R. Marlo et al., 231-239. Somerville, MA: Cascadilla Proceedings Project. www.lingref.com, document #2772. Geurts, Bart & Janneke Huitink. 2006. Modal concord. In Paul Dekker & Hedde Zeijlstra (eds.) Proceedings of the ESSLLI 2006 workshop Concord phenomena at the syntax semantics interface, pp15–20. University of M´alaga. Givon Talmy. 1993. English Grammar: A Function-Based Introduction. John Benjamins Publishing. Grice, H.P. 1957. ‘Meaning’. Philosophical Review 66: 377-88. Grice, H.P. 1968. ‘Utterer’s meaning, sentence meaning and word meaning. Foundations of Language 4: 225-42. Hengeveld, Kees. 1988. Illocution, mood and modality in a functional grammar of Spanish. Journal of Semantics 6: 227-269. Hoye, Leo. 1997. Adverbs and modality in English. London: Longman Huddleston, Rodney D., and Geoffrey K. Pullum. 2002. The Cambridge grammar of the English Language. Cambridge University Press. Klinge, Alex. 1993. The English modal auxiliaries: from lexical semantics to
125
utterance interpretation. Journal of Linguisics 29, 315-357 Kratzer, Angelika. 1977. What must and can must and can mean. Linguistics and Philosophy 1(3). 337–355. Kratzer, Angelika. 1981. The notional category of modality. In H. J. Eikmeyer & H. Rieser (eds.), Words, worlds, and contexts: New approaches in word semantics, 38–74. Berlin: de Gruyter.. Kratzer, Angelika. 1991. Modality. In Arnim von Stechow & Dieter Wunderlich (eds.), Semantics: An international handbook of contemporary research, pp. 639–650. Berlin: de Gruyter. Lapeyre, Carretero M. 1992. The role of epistemic modality in English politeness strategies. Miscelánea 13, 17-35 Lord, Carol. 1993. Historical change in serial verb constructions. Amsterdam: John Benjamins. Lyons, John 1968. Introduction to Theoretical Linguistics. London: Cambridge University Press Lyons, John. 1977. Semantics, Vol. II. Cambridge: C.U.P. Matthewson, Lisa. 2006. ‘Temporal semantics in a superfially tenseless language’. Linguistics and Philosophy 29: 673–713. Muysken, Pieter and Tonjes Veenstra (1995) “Serial Verbs.” In Jacques Arends, Pieter Muysken, and Norval Smith, eds., Pidgins and Creoles: An Introduction. Philadelphia: Benjamins.
126
Nuyts, Jan. 2006. Modality: Overview and linguistic issues. In Wolfgang Klein & Stephen Levinson (eds.), The Expression of Cognitive Categories, pp. 173-205. Berlin · New York: Mouton de Gruyter Narrog, Heiko. 2009. Modality in Japanese; The layered structure of the clause and hierarchies of functional categories. John Benjamins Nikola, Gisborne. 2007. In SKASE Journal of Theoretical Linguistics [online]. Vol. 4, no.2 43-61. Osam, E. Kweku. 1997. Serial verbs and grammatical relations in Akan. In Givón, Talmy (ed.), Grammatical relations: a functionalist perspective, 253-279. Amsterdam: John Benjamins. Osam E. Kweku. 2003. An introduction to the verbal and multi-verbal System of Akan. Proceedings of the workshop on Multi-Verb Constructions, Trondheim Summer School 2003. Dorothee Beermann and Lars Hellan (eds)
Osam E. Kweku. 2008. Akan as an Aspectual Languge.In Kropp Dakubu and Ameka
Felix
K.(eds.)
Aspect
and
Modality in Kwa
Languages. Amsterdam: John Benjamins Palmer, Frank. R. 1986. Mood and modality (1st Edition). Cambridge: CambridgeUniversity Press Palmer, Frank, R. 1990. Modality and the English Modals. Longman, London
127
Palmer, Frank. R. 2001. Mood and modality (2nd Edition). Cambridge: CambridgeUniversity Press Palmer, Frank. R. 2003. Modality in English: Theoretical, Descriptive, and Typological Issues. In Facchinetti, R., M. Krug, and F. R. Palmer (Eds.), Modality in Contemporary English, pp.1-17. Berlin: Mouton de Gruyter Papafragou, Anna. 1998. Inference and Word meaning: The case of modal auxiliaries. Lingua 105,1-47 Papafragou, Anna. 2000. Modality: Issues in the Semantics-Pragmatics Interface.Amsterdam: Elsevier. Payne, Thomas Edward.1997. Describing Morphosyntax: A guide for field Linguistics. Cambridge: Cambridge University Press. Saah, Kofi. 2004. A survey Adverbs and Adverbials. Journal of West African Languages 31.47-72 Sakyi, Joana Portia. 2013. Predicative Modal Adjectives in Akan Language. At the University of Essex (LangUE) 2012 Proceedings, 43-52 Silva-Corvalán, C. 1995. Contextual Conditions for the Interpretation of “Poder” and “Deber” in Spanish in J. Bybee& S. Fleischman (eds.) Sperber, Dan and Wilson, Deirdre. 1986. Relevance: Communication and Cognition. Blackwell, Oxford and Harvard University Press, Cambridge M A (Second edition 1995. Blackwell, Oxford) Sperber, Dan and Wilson, Deirdre. 1986b. On defining relevance. In R.
128
Grandy & R. Warner (eds) Philosophical Grounds of Rationality: Intentions, Categories, Ends: 143-158. Oxford University Press, Oxford. Sperber, Dan and Wilson, Deirdre. 1987a. Pracis of Relevance. Behavioral and B Brain Sciences 10.4: 697-710. Reprinted in Asa Kasher (ed.) 1998, vol. V: 82-115. Sperber, Dan and Wilson, Deirdre. 1987b. Presumptions of relevance. Behavioral and Brain Sciences 10, 736-53. Sperber, Dan and Wilson, Deirdre. 1995. Postface to the second edition of Relevance: Communication and Cognition. Blackwell, Oxford. Sperber, Dan and Wilson, Deirdre. 1998a. The mapping between the mental and the public lexicon. In P. Carruthers & J. Boucher (eds.), Language and Thought: Interdisciplinary Themes. Pp. 184-20. Cambridge University Press, Cambridge. Sperber, Dan and Wilson, Deirdre. 2002. Pragmatics, modularity and mind-reading. Mind & Language 17: 3-23. Special Issue on Pragmatics and Cognitive Science. Sweetser, Eve. 1990. From Etymology to Pragmatics: Metaphorical and Cultural Aspects of Semantic Structure. Cambridge University Press. Talmy, Leonard. 1988. Force dynamics in language and cognition. Cogntive Science 2, 49–61.
129
Tavangar Manoochehr, and Mohammad Amouzadeh. 2009. Subjective modality and tense in Persian. Language Sciences 31 (6), 853-873. Thompson, Sandra A. and Anthony Mulac. 1991. A quantitative perspective on the grammaticization of epistemic parentheticals in English. In Traugott E. and Heine B.(eds.), Approaches
to Grammaticalization,
pp.313–329. Amsterdam: John Benjamins. Tonhauser Judith. 2011. The Paraguayan Guarani future marker-ta: Formal Semantics and cross-linguiistic comparison. In, Rathert, Monika and Renate Musan (eds.), Tense Across Languages, pp. 207231. Tubingen: Niemeyer. Trask, Robert Larence 1993. A Dictionary of Grammatical terms in Linguistics. New York: Routeledge. Traugott, Elizabeth. 1989. On the Rise of Epistemic Meanings in English: An Example of Subjectification in Semantic Change. Language 65 (1), 31–55. Traugott,
Elizabeth Closs. 2003.
Approaching
modality
from
the
perspective of Relevance Theory. Language Sciences 2 (5), 657–669 Uchechukwu Chinedu. 2008. The Modal System of the Igbo language. In Werner Abraham and Elisabeth Leiss (eds.) Modality-Aspect Interfaces. Implications and typological solutions.Vol.79 Van der Auwera, Johan and Vladimir A. Plungian. 1998. Modality’s semantic map. Linguistic Typology 2, 79–124. Van Valin, Robert and Randy LaPolla. 1997. Syntax: Structure, meaning,
130
and function. Cambridge: Cambridge University Press. Vázquez Ignacio and Giner Diana. 2008. Beyond Mood and Modality: Epistemic Modality Markers as Hedges in Research Articles. A Cross-Disciplinary Study. Revista Alicantina de Estudios Ingleses 21, 171-190. Verstraete, Jean-Christophe. 2001. Subjective vs objective modality: Interpersonal and ideational functions in the English modal auxiliary system. Journal of Pragmatics 33 1505-1528 Wilson, Deirdre (2012). Semantics, pragmatics and the conceptual procedural distinction.
Meaning
and
Relavance.
University of Gent Wilson, Dierdre and Sperber, Dan., 2004. Relevance theory. In Horn, L., Ward, G. (eds.), The Handbook of Pragmatics. Blackwell Publishing, pp. 607–632. Wunderlich, Dieter. 1981. Modalverben im Diskurs und im System. In Rosengren, Inger (ed.), Sprache und Pragmatik. Lunder Symposium 1980, 13–53. Lund: CWK Gleerup. Wunderlich 1981 Ziegeler, Debra. 2006. Interfaces with English aspect. Diachronic and empirical studies. Amsterdam: John Benjamins. Zeijlstra, Hedde. 2008. Modal concord is syntactic agreement. Proceedings Of SALT 17, CLC Publications, Ithaca, NY http://ling.auf.net/lingBuzz/000494.