Neurobiological Substrates of Illiteracy - Semantic Scholar

Brain Organization Volume 6, Number 6, 2000THE NEUROSCIENTIST

REVIEW

n

Neurobiological Substrates of Illiteracy ALEXANDRE CASTRO-CALDAS and ALEXANDRA REIS Lisbon Center for Neurosciences Lisbon, Portugal

Comparable subjects except for the knowledge of orthography and school attendance in the proper age provided the case material for a series of studies that are reviewed. The results suggest that the acquisition of or thographic skills provides a basis for changes in the pattern of activation of the brain. NEUROSCIENTIST 6(6):475–482, 2000 KEY WORDS Illiteracy, Language, Reading, Writing, Learning, Brain organization

The brain can be understood as an organ that adapts to several types of internal and external influences. The interaction of these complex concurrent stimuli along life shapes the highly differentiated biologic arrangement of the brain and its consequent physiology, at any given time. The past years have seen a significant development of the knowledge about the neural substrate of language in its oral and written forms. This is, in part, the result of understanding acquired oral and written language disorders (see, for example, the reviews of Neville and Bavelier [1], Patterson and Ralph [2], and Price [3]), of studies on the influence of sensory deprivation on the harmony of the so-called normal development (4), and of functional brain scanning of normal subjects while they perform tasks related to these skills (5, 6). This brief review is based on the assumption that the neurobiology of oral and written language is generally understood as the result of the interaction of the following structures and mechanisms: 1. auditory cortex and cortical and subcortical motor mechanism s responsible for m odul at i ng phonology; 2. auditory cortex and multisensory integrative cortex res ponsible for the lexicon-sem ant i c components; 3. visual cortex, and its connections to mechanisms 1 and 2, responsible for reading; 4. parietal cortex and the dorsal visual connection responsible for visually guided movement that is the origin of writing; and 5. connections among all of these mechanisms. Learning to Read There is considerable evidence that the mastering of the written component of language entails the action of many more areas of the brain than those simply responAddress correspondence to: Alexandre Castro-Caldas, University Department of Neurology, Hospital de Santa Maria, 1649-035 Lisboa Portugal (e-mail: [email protected]).

Volume 6, Number 6, 2000 Copyright © 2000 Sage Publications, Inc. ISSN 1073-8584

sible for the mechanisms of oral production (as represented in the cartoon in Fig. 1). Learning orthography can be considered as involving the adding at least of a visuo-spatial component to a previous existing audio-temporal mechanism. The adaptation to the new type of information concerns not only the recruitment of new mechanisms and therefore new regions of the brain but also a change in the proper oral-language-supporting mechanism. In the case of reading, bringing the implicit processing of phonology to a declarative conscious level is the main change of this acquisition (7, 8) because awareness of segmentation of word components is crucial for letterby-letter reading (2). In his pioneer work, Morais et al. (9) demonstrated that illiterate subjects failed in performing a number of tasks that require phonological awareness. This biofunctional model can thus predict the changes that are introduced in the system following the formal learning of orthography at school. The Illiterate Population In certain regions of the world, as was the case in Portugal in the past, illiteracy due to social reasons is a common finding. Indeed some 40 years ago, the first-born girl in a family was kept at home at the age of 6 instead of going to school. She was charged with being a caretaker for the other siblings, generally in large families. On the other hand, younger children, when reaching the age of 6 or 7, were sent to school because they were considered a nuisance for the normal functioning of the house. They spent all day at school and were fed there, which represented an economy for the meager budget of the family. We have studied a community that includes a significant number of illiterate people. In this community, mobility to other parts of the country is not common because the main source of economy derives from the sea and life revolves around fishing activities. Thirty percent of the population older than 55 is totally illiterate in Portugal due to socioeconomic reasons. This figure is higher in small rural or fishing communities such as the one where most of the subjects were selected. This is a small town, now composed of old and new

THE NEUROSCIENTIST

475

Fig. 1. According to our model, when one masters the skill of reading and writing, there are increased opportunities to make connections in the brain.

neighborhoods. This provides a setting in which literate and illiterate aged subjects live together, influenced by the same sociocultural background. We cannot say that literate subjects in the studies we will review are sophisticated readers; they know how to read as shown by their ability in reading and comprehending a short text. This knowledge represents the main social/behavioral difference between the groups, making them an interesting experimental population. Other studies have been performed with other groups comparing, for instance, extreme cultural differences (10, 11). In such cases, it seems difficult to know which cultural factor influences what. Behavioral Studies Our model predicted that the absence of the formal knowledge of phonology would prevent the performance of tasks requiring a declarative analysis of sublexical components. On the other hand, the absence of practicing tasks involving the word form would make most of language operations to be based on semantic knowledge. Exploration of these hypotheses revealed that there were important differences between literate

476

THE NEUROSCIENTIST

and illiterate subjects in the processing of tasks requiring the awareness of phonology, as in the case of repetition of pseudowords (sequences of phonemes that are according to the rules of phonology but that have no meaning) (12). To make the grapheme-phoneme matching operation, one has to understand explicitly that the words are composed of segments that can be isolated. If no such requirement is ever stimulated as an operative tool, operations that need the fine analysis of word components, such as the repetition of pseudowords, are difficult. Results obtained by two groups of subjects’—23 illiterate and 10 literate—repeating words and pseudowords are represented in Figure 2. Subjects were asked to repeat a list of 24 highly frequent words and a list of 24 pseudowords designed by changing the consonants of the real words and maintaining the vocal structure. Words and pseudowords were randomly mixed and presented as a single list. At the level of lexico-semantic processing, differences were also found in storing and retrieving verbal information cued by formal attributes (12). The same group of subjects as for the experiment reported above participated in this study. A variant of the word-pair association test of the Weschler Memory Scale was

Illiteracy and Brain Organization

Fig. 2. Repetition of words and pseudowords by literate and illiterate subjects (maximum score = 24). Two-way interaction: F(1, 56) = 16.22, P < 0.002 (13).

Fig. 4. Schematic representation of the experiment in which subjects were asked to manipulate the mouse to reach stimuli randomly presented on the computer screen. Performance using the right hand to reach the stimuli on the left side of the screen (crossed condition) was worse for illiterate subjects. Fig. 3. Performance of literate and illiterate subjects in a word-pair association test according to type of word analogy (maximum score = 30). Two-way interaction: F(1, 56) = 6.10, P < 0.0166 (13).

used. Two sequences of 10 pairs of words were prepared. In each sequence, 5 pairs were semantically related, and the other 5 were phonologically related. Figure 3 shows that the type of association matters for the performance. Parallel to these differences related to the central core of language processing, the exploration of visual competence also showed that the training of visual analysis and visual scanning was important when the subjects were asked to name drawings. In a first study (13), we were able to demonstrate that illiterate subjects had difficulties in naming a series of drawings of objects in comparison to their ability in naming the real objects. Naming the photographs of the objects was better than naming drawings but worse than naming the real objects. The comparison between groups revealed that illiterate subjects named real objects like their literate controls, but they were worse on naming line drawings. We recently revisited this topic with a better methodology (unpublished data) and measured the reaction time between stimuli presentation and the oral production of

Volume 6, Number 6, 2000

the name. Results on naming performances were reproduced with the same differences as before. We also examined visually guided hand motor behavior, which can be trained by writing (14). In a test paradigm in which literate and illiterate subjects were asked to direct the cursor toward a target in the screen of a computer using the mouse, illiterate subjects were slower then literate controls, in particular when the right hand had to move the cursor to the left side of the screen (see Fig. 4). In this study, all of the subjects were for the first time confronted with a computer. Apparently, the visually guided movement is slower in illiterate subjects, and there also is an effect considering the side of the screen where the stimuli are presented. Illiterate subjects are slower in reading stimuli presented on the left side of the screen. Biofunctional and Anatomic Findings Following up on the behavioral results mentioned above, where an important difference was found between literate and illiterate subjects’ repeating pseudowords, we asked whether there are differences in patterns of brain activation during a word and pseudoword repetition task in literate and illiterate subjects. To do this, we studied the brain activity of sub-

THE NEUROSCIENTIST

477

jects while performing these tasks (15) by means of Positron Emission Tomography (PET). Subjects involved in this study were selected from the usual target population in the South of Portugal and volunteered to travel to Stockholm to be scanned. Two groups of 6 women each (6 women completely illiterate and 6 with 4 years of schooling) matched for age and measures of social and intellectual proficiency traveled with us in groups of four to Sweden. Lists of words and lists of pseudowords were prepared and taped for use in an immediate repetition task. Subjects were not informed about the content of the lists and were asked to repeat whatever they heard. Subjects behaved according to the results obtained in the first study: repetition of pseudowords was significantly worse in the illiterate group compared to the literate one. Results of the activation images can be summarized as follows (see Fig. 5): there was a small difference between the groups while repeating real words—the left inferior parietal gyrus was more activated in the literate group—and there was an important difference between the groups while repeating pseudowords. The areas that were more activated in literate subjects, as compared to illiterate ones, were the right frontal operculum/anterior insula, left anterior cingulate, left putamen/pallidum, anterior thalamus/hypothalamus, pons and medial cerebellum (vermis). This was the first time that such an experiment was conducted. The interpretation of the differences between groups in terms of specific language processes, or increases/decreases of regional cerebral blood flow (rCBF) in a given state, was complicated by the lack of a language-neutral reference state in the experimental design. However, even in the absence of such a reference, these skills suggest that learning to read and write during childhood influences the functional organization of the adult human brain. One could argue that the absence of activation is the result of the poor performance. However, there was no correlation between performance and level of activation (16). The images found thus appeared to be the result of the declarative processing of phonology. In other words, in this particular task, illiterate subjects were unable to repeat pseudowords because they were unable to activate the proper areas of the brain to do it. A Network Analysis These results were recently reevaluated on the basis of a network analysis approach to the funct i onal neuroimaging data, which provide complementary information to the more common general linear approach. A model of connections between regions considered relevant for the performance was generated, and the covariance of the activation measures of each region was computed (16) based on structural equation modeling (17). Taking in consideration the literature on language processing, a model of interacting structures was constructed. These structures were organized in five subnetworks related to auditory input, phonological

478

THE NEUROSCIENTIST

loop, articulatory motor output, attention, and central executive. Regions of interest (ROIs) were selected and incorporated within the frame of the general model. Covariation matrix of the activity measured in each ROI was computed across subjects (literate and illiterate) and conditions (repetition of words and repetition of pseudowords). Results showed that the network interactions differ in the illiterate group while repeating words or pseudowords. This was not the case for literate subjects. On the other hand, network interactions differ between groups with respect to pseudoword but not word repetition. These findings may reflect differences in attentional modulation of the language network, executive aspects of verbal working memory, and the organization of verbal output, as well as differences in the coordination of the phonological loop. Left-Right Hemispherical Representation of Language A hypothesis raised a long time ago by some authors and apparently supported by empirical data of Cameron et al. (18), later denied by Damásio et al. (19, 20) and again suggested by Lecours et al. (21, 22) was that the functional balance between the two cerebral hemispheres, while processing oral language, could be modified by the knowledge of orthography. We had learned that a region on the left parietal lobe was more activated in PET of literate subjects compared to illiterate while repeating words, a task that was easily and well performed by both groups. The study of visually guided hand movements also suggested that the knowledge of writing could have implications in interhemispheric transference. Taking the PET data concerning repetition of words and pseudowords into consideration, we selected ROIs with particular focus on the posterior pariet al cor t ex ( BA 39,40 and BA 31) . L ef t / r i g h t hemispheric differences were compared between groups averaging each subject’s data over words, pseudowords and word plus pseudowords, respectively. The three different analyses indicate the same pattern of between-group differences, all relating to the ROIs in the posterior parietal cortex (superior part of BA40 on the border of BA7, inferior part of BA39 extending into inferior part of BA40, and BA31 part of precuneus). The differences between groups indicate a dissociation between the superior and inferior parts of the angular-supramarginal regions, that is, the superior parts being more active on the left than on the right in illiterate compared to literate subjects, whereas the reverse was the case for the inferior parts and the precuneus. Therefore, it can be suggested that language, which is known not to be a single function dependent on one hemisphere, has a different, balanced representation between the two hemispheres in each group. One lesson is that learning the visuo-spatial dimension of language (orthography) can have a modulatory influence in language representation between the two cerebral hemispheres (23).


Fig. 5. Maximum intensity projections of all significant activations thresholded at Z = 3.09 (omnibus significance P < 0.001) in the words-pseudowords contrast in the literate (A ) and illiterate (B ) groups. The reverse contrast (pseudowords-words) is shown for the literate (C ) and illiterate (D ) groups. (E ) Results of the interaction analysis Group × Words-Pseudowords masked with the word-pseudoword contrast in the literate group, and the interaction contrast (Group × Pseudowords-Words) (F ) masked with the pseudoword-word contrast in the literate group. For illustration, the threshold in the conjunction was set at Z = 2.33 (omnibus significance P < 0.01) (14). Reprinted from ref. 15 by permission of Oxford University Press.


THE NEUROSCIENTIST

479

Fig. 6. The corpus callosum of each subject was divided in 100 sections. The width of the sections were compared between groups. The region where parietal fibers are thought to cross is thinner in illiterate subjects (arrow).

The Corpus Callosum Information crosses between hemispheres through the corpus callosum. Therefore, the suggestion of an increased cross-talk made this region an interesting target of study. Variations in the anatomy of the corpus callosum have been reported in several conditions, for example, sex, gender, and handedness. However, reported differences have been inconsistent across these studies; see, for example (24–26). The study of monozygotic twins has also suggested that the size and shape of the corpus callosum may be influenced by genetic and nongenetic factors (27). Another piece of important evidence comes from studies that showed a significant growing of the corpus callosum until late in life (28–30), suggesting a prolonged interaction with environmental factors. Little is known about the factors that influence, or mechanisms that produce, the morphological changes. One possibility is that they may be related to the acquisition of new skills, that is, learning. Taking the potential variations in callosal shape and size, with sex, handedness, and age into consideration, we selected for study 41 right-handed upper-middle-aged women, 18 being illiterate and 23 literate (31). The

480

THE NEUROSCIENTIST

outline of each corpus callosum, as viewed in the midsagital section of MRI, was digitized by manually tracing over the film, using a modification of the method proposed by Denenberg et al. (32). In summary, the results of the comparison done between the two groups showed that there was a region of the corpus callosum that was smaller in illiterate subjects (see Fig. 6). This region corresponds to the area where interparietal fibres cross (33). The functional evidence that we summarized above, concerning differences between the activations of the hemispheres at the parietal level when orthography is mastered are in agreement with this morphological data. One possible conclusion is that stimulation of transfer of information between the two hemispheres may result in an increasing of the size in the specific regions where parietal fibers cross. It should be noted that the results of the study on visually guided motor behavior that we reported can be also understood on the basis of these findings. Coding and Retrieving Verbal Information Because of the previously mentioned results in which illiterate subjects were shown to be worse in retrieving


pair-associated words, a test paradigm was designed to be used while brain activity was measured by PET. In this paradigm, we studied the encoding period and the retrieval period of the task. The comparison between groups showed some slight differences that are still under evaluation (34). We also used this paradigm to ask whether the effective relational encoding correlates positively with the activity of the anterior medial temporal lobe (MTL) regions within the illiterate group (35). Specifically, we used the cued-recall success as a covariate in a general linear model. The results confirmed that the left inferior prefrontal cortex (PFC) and the MTLs were more active during effective encoding than during ineffective encoding. The results showed a positive correlation between cued-recall success and the regional cerebral blood flow of the left inferior PFC (BA47) and the MTLs. These effects were observed during encoding of both semantically and phonologically related word pairs, indicating that these effects are robust in the illiterate population studied, that is, it is reproducible within the group. These results generalize those of Brewer et al. (36) and Wagner et al. (37) to an upper-middle-aged or older illiterate population in showing that effective encoding (i.e., encoding of stimuli later recognized) activated the PFC and the MTL more than ineffective encoding (i.e., encoding stimuli not later recognized) using either visuospatial or verbal material.

Horizons and Opportunities The information reviewed in this article is part of a growing body of data on illiteracy. Notably, we have found that in carrying out certain tasks, the brains of illiterate subjects show patterns of activation that are different than those in literate subjects. We conclude that social and/or economic circumstances (lack of educational opportunity in this case) can be reflected in changes in the pattern of brain activation in humans and that these changes in brain actuation, in turn, can shape behavior. Other authors have addressed the same topic (for a general reference, see 38) in different populations and within different experimental and heuristic frameworks. Illiteracy, for social reasons, is disappearing in the world, but it is indeed a good model for understanding the adaptation of brain/behavior to a particular function. Thus, in addition to its social implications, illiteracy may serve to teach us about the brain and how it changes in response to the acquisition of new skills and knowledge. Hopefully, societies will eliminate illiteracy in the near future. Along the way, however, we may be able to learn from it.

References 1. 2. 3. 4. 5. 6. 7. 8.

9. 10. 11. 12. 13. 14. 15.

16.

17. 18. 19. 20. 21.

22. 23.

Acknowledgments

24.

The authors thank K. M. Petersson for the valuable suggestions in reviewing this article.

25.


Neville HJ, Bavelier D. Neural organization and plasticity of language. Curr Opin Neurobiol 1998;8:254–8. Patterson K, Ralph MAL. Selective disorders of reading? Curr Opin Neurobiol 1999;9:235–9. Price CJ. The functional anatomy of word comprehension and production. Trends in Cognitive Sciences 1998;2:281–8. Neville HJ. Developmental specificity in neurocognitive development in humans. In: Gazzaniga MS, editor. The cognitive neurosciences. Cambridge, MA: MIT Press; 1995. p. 219–34. Démonet J-F, Chollet F, Ramsay S, et al. The anatomy of phonological and semantic processing in normal subjects. Brain 1992;115:1753–68. Fiez JA, Petersen SE. Neuroimaging studies of word reading. Proc Natl Acad Sci U S A 1998;95:914–21. Bradley L, Bryant P. Categorizing sounds and learning to read—a causal connection. Nature 1983;301:419–21. Bradley L, Bryant P. Phonological skills before and after learning to read. In: Brady SA, Shankweiler DP, editors. Phonological processes in literacy. Hillsdale, NJ: Lawrence Erlbaum; 1991. p. 37–45. Morais J, Cary L, Alegria J, Bertelson P. Does awareness of speech as a sequence of phones arise spontaneously? Cognition 1979;7:323–31. Ardila A, Rosselli M, Rosas P. Neuropsychological assessement in illiterates: visuospatial and memory abilities. Brain and Cognition 1989;11:147–66. Rosselli M, Ardila A, Rosas P. Neuropsychological assessement in illiterates: II. Language and praxic abilities. Brain and Cognition 1990;12:281–96. Reis A, Castro-Caldas A. Illiteracy: a bias for cognitive development. J Int Neuropsychol Soc 1997;3:444–50. Reis A, Guerreiro M, Castro-Caldas A. Influence of educational level of non brain-damaged subjects on visual naming capacities. J Clin Exp Neuropsychol 1994;16:939–42. Reis A, Castro-Caldas A. Learning to read and write increases the efficacy of reaching a target in two dimensional space. J Int Neuropsychol Soc 1997;3:222. Castro-Caldas A, Petersson KM, Reis A, Stone-Elander S, Ingvar M. The illiterate brain: learning to read and write during childhood influences the functional organisation of the adult brain. Brain 1998;121:1053–63. Petersson KM, Reis A, Askelof S, Castro-Caldas A, Ingvar M. Language processing modulated by literacy: a network analysis of verbal repetition in literate and illiterate subjects. J Cogn Neurosci (in press). McIntosh AR, Gonzalez-Lima F. Structural equation modeling and its application to network analysis in functional brain imaging. Hum Brain Map 1994;2:2–22. Cameron RF, Currier RD, Haerer AF. Aphasia and literacy. Br J Disord Communication 1971;6:161–3. Damásio AR, Castro-Caldas A, Grosso JT, Ferro JM. Brain specialization for language does not depend on literacy. Arch Neurol 1976;33:300–1. Damásio AR, Hamsher K, Castro-Caldas A, Ferro JM, Grosso JT. Letter to the editor. Arch Neurol 1976;33:662. Lecours AR, Melher J, Parente MA, et al. Illiteracy and brain damage: 3. A contribution to the study of speech and language disorders in illiterates with unilateral brain damage (initial testing). Neuropsychologia 1988;26:575–89. Lecours AR, Melher J, Parente MA, et al., Illiteracy and brain damage: 1. Aphasia testing in culturally contrasted populations (control subjects). Neuropsychologia 1987;25:231–45. Petersson KM, Reis A, Askelöf S, Castro-Caldas A, Ingvar M. Differences in inter-hemispheric interaction between literate and illiterate subjects during verbal repetition. Neuroimage 1998;7: S217. Duara R, Kushch A, Gross-Glenn K, et al. Neuroanatomic differences between dyslexic and normal readers on magnetic resonance imaging scans. Arch Neurol 1991;48:410–6. Hynd GW, Hall J, Novey ES, et al. Dyslexia and corpus callosum morphology. Arch Neurol 1995;52:32–8.

THE NEUROSCIENTIST

481

26.

Larsen JP, Hoien T, Odegaard H. Magnetic resonance imaging of the corpus callosum in developmental dyslexia. Cognitive Neuropsychology 1992;9:123–34. 27. Oppenheim JS, Skerry JE, Tramo MJ, Gazzaniga MS. Magnetic resonance imaging morphology of the corpus callosum in monozygotic twins. Ann Neurol 1989;26:100–4. 28. Cowell PE, Allen LS, Zalatimo NS, Denenberg VH. A developmental study of sex and age interactions in the human corpus callosum. Developmental Brain Research 1992;66:187–92. 29. Pujol J, Vendrell P, Junque C, Mart¡-Vilalta JL, Capdevilla A. When does human brain development end? Evidence of the corp u s cal l o s um g r ow t h u p t o a d u l t h o o d . A nn N e ur ol 1993;34:71–5. 30. Witelson SF. Sex differences in neuroanatomical changes with aging. N Engl J Med 1991;325:211–2. 31. Castro-Caldas A, Miranda P, Carmo I, et al. Influence of learning to read and write on the morphology of the corpus callosum. European Journal of Neurology 1999;6:23–8. 32. Denenberg VH, Kertez A, Cowell E. A factor analysis of the human’s corpus callosum. Brain Research 1991;548:126–32.

482

THE NEUROSCIENTIST

33.

34.

35.

36. 37.

38.

Pandya DN, Seltzer B. The topography of commmisural fibers. In: Lepore F, Ptito M, Jaspar H, editors. Two hemispheres—one brain: functions of the corpus callossum. New York: Alan R Liss; 1986. p. 47–73. Reis A, Petersson KM, Rosenlund M, Castro-Caldas A, Ingvar M. The modulatory influence of literacy on the activation pattern during encoding and retrieval of auditory-verbal information. NeuroImage 1999;9:S1086. Petersson KM, Reis A, Castro-Caldas A, Ingvar M. Effective auditory-verbal encoding activates the left prefontal and the medial temporal lobes: a generalization to illiterate subjects. NeuroImage 1999;10:45–54. Brewer JB, Zhao Z, Desmond JE, Glover GH, Gabrieli, JDE. Making memories: brain activity that predicts how well visual experience will be remembered. Science 1998;281:1185–7. Wagner AD, Schacter DL, Rotte M, Koutsmal W, Maril A, Dale A, et al. Building memories: remembering and forgetting verbal expe r i e nc e s a s pr e di c t e d by br a i n a c t ivi t y. S c i e n c e 1998;281:1188–91. Coppens P, Lebrun Y, Basso A, editors. Aphasia in atypical populations. Hillsdale, NJ: Lawrence Erlbaum; 1998.
