Language and Evolution: Homepage Robin Allott


"As I understand it, minimalism currently is an expression of a move back to lexicon and away from syntax in isolation as the dominant aspect of language."

1. Minimalism:

In Hale, Kenneth and Samuel J. Keyser eds. 1993. The View from Building 20:

Chapter 1 Chomsky, Noam 1993. A Minimalist Program for Linguistic Theory.

approach, assumed here, takes language to be a part of the natural world Another recurrent theme has been the role of 'principles of economy' ... Another standard assumption is that a language consists of two components, a lexicon and a computational system. The lexicon specifies the items that enter into the computational system, with their idiosyncratic properties. We are led to dispense with the level of D-structure ... S-structure, another level that has only theory-internal motivation ... the implications of the theory that syntax is projected from the lexicon

From Hornstein, Norbert 1995. Logical Form: From GB to Minimalism: One truism is that sentences are pairings of sounds with meanings ... Since the earliest days of generative grammar, a central concern of linguistic theory has been to elucidate how it is that a natural language sentence expresses its meaning. Broadly speaking, Minimalism differs from GB theories in not according DS or SS any grammatical significance. Chomsky (1993) offers a substantial revision of the principles of Universal Grammar (UG). The general aim is to develop a theory of grammar based exclusively on natural concepts, as is required by 'virtual conceptual necessity'. The idea is to develop as conceptually economical a theory as possible ... Natural language sentences are pairings of sound and meaning. PF and LF are the theoretical reflexes A central aim of the Minimalist program is to show that the grammatical levels DS [Deep Structure] and SS [Surface Structure] do not exist.

The theoretical direction outlined by Chomsky (11993) requires a wholesale revision of previous GB approaches to LF if consistently pursued. It is clear (to me) that if Minimalism is roughly correct then the earlier GB analyses will have to be abandoned and radically rethought. The fundamental notions, constructs and principles are different. If the analyses above are correct, there is no place in Minimalism for the (antecedent government part of the) ECP, A'movement, or theories based solely on output constraints. This is a fundamental reorientation of grammatical theory. [ See Hornstein's list of 71 technical aspects of GB theory]

Extracts from: Chomsky, Noam. 2000. New Horizons in the Study of Language and Mind. Cambridge University Press.

Publisher's Blurb: "This book is an outstanding contribution to the philosophical study of language and mind, by one of the most influential thinkers of out time. In a series of penetrating essays, Noam Chomsky cuts through the confusion and prejudice which has infected the study of language and mind, bringing new solutions to traditional philosophical puzzles and fresh perspectives on issue of general interest, ranging from the mind-body problem to the unification of science.

Using a range of imaginative and deceptively simple linguistic analyses, Chomsky argues that there is no coherent notion of "language" external to the human mind, and that the study of language should take as its focus the mental construct which constitutes our knowledge of language. Human language is therefore a psychological, ultimately a "biological object", and should be analysed using the methodology of the natural sciences. His examples and analyses come together in this book to give a unique and compelling perspective on language and the mind.

Foreword by Neil Smith:

A number of consequences follow from his naturalistic thesis: there is no justification for the common assumption that natural languages ought to be treated like the invented formal languages of logic or mathematics Chomsky has cut the Gordian knot by emphasizing a more fundamental difficulty: the mind-body problem cannot even be formulated. Trying to reduce linguistics to neurology in the current state of our understanding is then unlikely to be productive. The theory of Principles and Parameters which has been developed over the past two decades is probably the first really novel approach to language over the last two and a half thousand years ... What is impressive about Chomsky's writing is not just its awesome breadth and considerable scope, but that after half a century he still has the power to surprise: Everything combines to give a unique and compelling view of language and mind.

[The book is actually a collection of previously published articles - 1997, 1992, 1992, 1993, 1994, 1995. The last chapter is the only new one]


One domain in which there has been substantial progress is the study of language, particularly in the past 20 years. . . . Some have plausibly termed it "biolinguistics". . . . Biolinguistic inquiry seeks unification with other approaches to the properties of the brain, in the hope that some day the slash "/" in the phrase "mind/brain" will gain more substantive content. . . . this "naturalistic" approach.

Chapter 1 New Horizons in the study of language [January 1997]

About 15 years ago . . . a much more radical departure from the tradition than earlier generative grammar had been. This "Principles and Parameters " approach . . . . rejected the concept of rule and grammatical construction entirely; . . . The familiar grammatical constructions are taken to be taxonomic artifacts, useful for informal description perhaps but with no theoretical standing. What has come to be called "the Minimalist Program" is an effort to explore these questions . . . The minimalist program requires that we subject conventional assumptions to careful scrutiny. the most venerable of these is that language has sound and meaning. . . . . Are there levels "internal" to language, particular, the levels of deep and surface structure that have been postulated in modern work. The minimalist program seeks to show that everything that has been accounted for in terms of these levels has been misdescribed and is as well or better understood in terms of legibility conditions at the interface; for those of you who know the technical literature, that means the projection principle, binding theory, Case theory, the chain condition and so on. The computational system that generates expressions has two basic operations: one assembles features into lexical items, the second forms larger syntactic objects out of those already constructed beginning with lexical items. We can think of the first operation as essentially a list of lexical items. In traditional terms, this list - called the lexicon - is the list of "exceptions", arbitrary associations of sound and meaning and particular choices among the inflectional properties made available by the faculty of language that determine how we indicate that nouns and verbs are plural or singular, that nouns have nominative or accusative case, and so on. These inflectional features turn out to play a central role in computation. Optimal design would introduce no new features in the course of the computation. There should be no indices or phrasal units and no bar levels (hence, no phrase-structure rules or X-bar theory. . . We exclude government, binding relations internal to the derivation of expressions, and a variety of other relations and interactions. A core assumption of the work within the Parameters-and Principles framework . . . is that everything I have just proposed is false. Languages plainly differ . . . in inflectional systems: case systems, for example. . . . it seems that much of the variety of language can be reduced to properties of inflectional systems. If this is correct, then language variation is located in a narrow part of the lexicon. There is a fair amount of hand-waving in this brief description [words] "on the meaning side there are profound disagreements. . . . referentially dependent elements . . . observe some distinctions and ignore others, in ways that vary for different types of words in curious ways. . . . What we discover is surprisingly intricate and . . . known in advance of any evidence. . One theory [minimalism] holds that . . . the properties of an expression that enters into language are completely drawn from the lexicon; the computation organises these in very restricted ways, but adds no further features

Extracts from: SILENT CHILDREN, NEW LANGUAGE Interview with Professor Noam Chomsky MIT:

INTERVIEWER: What are the features of language that must be, at least partly, innate? CHOMSKY: Every feature of language from the articulatory gestures to the meanings of words, to the ways sentences are constructed.

INTERVIEWER: What you're saying is that in no case is the child's input anywhere near rich enough for them to infer the language. How does this relate to the children in Nicaragua and their experience. CHOMSKY Here's a case where a language like system was created on the basis of, it seems, no input just by interaction. The Nicaraguan case appears to be a very rich example, the richest yet known, of a natural experiment in which a language like system, maybe an actual human language, was developed on the basis of no external input as far as we know and that's intriguing. In the case of humans, one fundamental part of their nature is a language organ and they plainly are designed, somehow, to develop this organ.

Comments on Chomsky's Minimalism:

[a comment on the Internet] On the subject of GB and Minimalism, does anybody understand better than I do what the relationship between them, if any, is supposed to be? And if there is no relationship, in what sense is GB still the "standard" theory of syntax, if it has been abandoned by Chomsky?

Franks: Paradigms in generative syntax don't die--they collapse under their own weight. They start as simple structures, designed to provide the basic needs of syntactic description. Slowly, inexorably, they become adumbrated with special fixtures for special purposes. Eventually,sporting gargoyles, they transcend the baroque. By the time a paradigm has evolved to such a degree that the original simple structure is so obscured by rococo ornamentation that it cannot be discerned, it is usually time to dismantle and begin anew. In the history of generative grammar we have cycled through this process several times. In the early 1980's the revised extended standard theory was refreshingly replaced by government-binding theory. Within ten years GB too had become encumbered by many otiose frills and decorations, culminating in its replacement with the "minimalist" approach to grammar.The leading idea of minimalism is the rejection of all devices and constructs except those that are absolutely necessary on conceptual grounds.

[See Working Minimalism S.D. Epstein and N. Hornstein (eds.) For similar over-elaboration of Minimalism] and also:

Extracts from: 1997 O'Neil III J.H. Means of Control: Deriving the Properties of PRO in the Minimalist Program.

The Minimalist Program grew out of the efforts of researchers in the Principles and Parameters framework. Earlier theories in the Principles and Parameters framework, such as Government and Binding Theory as presented in Chomsky (1981, 1982, 1986a, 1986b), include a rich set of principles from which it is possible to deduce logically the grammaticality of an utterance. In the Minimalist Program, as presented in Chomsky (1993), it was attempted to simplify the theory of the syntax of natural language to the greatest possible extent.

Chomsky (1993) assumes that syntactic structures are built iteratively, by taking smaller structures and combining them by a computational procedure, starting with a multiset of items drawn from the lexicon. This computational procedure takes lexical items — atomic syntactic structures — in addition to more complex structures and combines them into larger structures.

The GB arguments concerning the distribution and interpretation of PRO depend on theoretical tools which in the Minimalist Program are no longer available, including the Projection Principle, D-structure and S-structure, the notion of government, and the Binding Theory itself (bringing the PRO theorem into even greater doubt).]

Alternatives to Minimalism:

Dependency grammars: the approach to the explanation of language and its structure where the accent is on hierarchical relations between words. [For example: 1997 Melc'uk, Igor. 1997. Vers une linguistique Sens-Texte. 1998 Polguère, A. La théorie Sens-Texte. ]

[From Richard Hudson:

There are a dozen or so theories of language in general, and of grammar in particular (and even more particularly, of syntax with or without semantics). Dependency and cognitive grammars include:

Autolexical Grammar Cognitive Grammar Construction Grammar Functional Grammar Categorial Grammar Head-driven Phrase Structure Grammar Integrational Linguistics Lexical Functional Grammar Link Grammar Systemic Functional Grammar Transformational Grammar (was Government-Binding or Principles-and-Parameters, now the Minimalist Program) Tree Adjoining Grammar

[Richard Hudson: The main ideas of Word Grammar [Language as a cognitive network]:

It is monostratal - only one structure per sentence, no transformations. It uses word-word dependencies - e.g. a noun is the subject of a verb. It does not use phrase structure - e.g. it does not recognise a noun phrase as the subject of a clause, though these phrases are implicit in the dependency structure. It views concepts as prototypes rather than `classical' categories that can be defined by necessary and sufficient conditions. It presents language as a network of knowledge In this network there are no clear boundaries between different areas of knowledge - e.g. between `lexicon' and `grammar'

Optimality Theory

"Optimality Theory has revolutionized phonological theory, and its insights are now being applied to other central aspects of language. With contributors that include the leading researchers in the field, this book presents the first fruits of such research as applied to syntax and to language acquisition, as well as considering the main lines of attack on OT by rule-based grammarians." (OUP 1999)

2. Lexicon in the Brain

"In the brain different categories of words, content words, function words, vision words, action/motor words, are associated with topographically different patterns of excitation; the brain seems to be categorising the perception of words in ways very similar to the standard analyses of the lexicon."

Function words and content words

The class of function words (including subwords eg. morphemes such as terminations of abstract nouns, verb inflections etc) is of special importance for syntax. The class is closed; function words and subwords in any language can be enumerated in a way that is not possible for nouns or verbs; the number of function words and subwords is quite small. Evidence of a neurological distinction between function words and other words comes from clinical treatment of the aphasias over a long period; many aphasiologists have noted that in certain types of aphasia, content and function words are differentially affected.

Psycholinguistic theories propose that words of the 2 major vocabulary classes, content (open-class) and function (closed-class) words, are computationally distinct and have different neuronal generators. This predicts distinct EEG patterns elicited by words of the 2 classes.(Pulvermuller 1995) Evidence from studies of sentence production and comprehension points to a distinction between the open and closed class vocabularies. Data from the neurologically impaired suggest the left hemisphere selectively supports this distinction. (Shapiro & Jensen 1986)

In Luria's 'efferent aphasias' the selection of content words is unimpaired, but the combination of words and their serial order may be severely disturbed. Speech may degenerate to a succession of telegraph- style utterances. Grammatical words such as conjunctions, prepositions, pronouns and articles disappear; roots of words are preserved better than grammatical endings, tense or gender. The nominative may be the only case that survives. (Sommerhoff 1974).

One of the more remarkable observations was where patients were asked to read a list of homophone-pairs where one member of the pair was a content word and the other a function word. "H.T. was able to read 'four' but not 'for'; V.S. read 'sum' but not 'some'; and J.D. read 'for' and 'some' but not 'four' or 'sum'. (Marin et al. 1976)

In languages with well-developed inflectional systems, agrammatism is particularly striking. Nouns tend to appear only in the nominative case, verbs in the infinitive, auxiliary verbs and other words from closed classes are omitted. (Howard 1985)

There are also where electrically induced changes are confined to closed class words. At a few sites, only conjunctions, prepositions and verb endings were altered during stimulation. (Ojemann 1983)

Content words correspond to neuronal assemblies equally distributed over both hemispheres, while assemblies corresponding to function words are strongly lateralized to the left hemisphere and primarily located in the perisylvian region. (Pulvermuller , Lutzenberger , Birbaumer 1995)

Electrophysiological and metabolic imaging studies provide evidence that not only the language cortices in the left hemispheres, but additional cortical areas outside the left perisylvian areas play a role in word processing. indicate that both hemispheres are strongly involved in processing concrete content words, whereas predominantly left-lateralized activity in or close to perisylvian regions appear to be related to processing of highly abstract function words.(Pulvermuller 1999)

The child quickly acquires a closed class vocabulary, a relatively small set of frequently used function words (Levelt, Roelofs & Meyer 1999)

Exchange of stuttering from function words to content words with age. For the 2-6-year-old speakers that stutter, there was a higher percentage of dysfluencies on initial function words than content words. (Howell, Au-Yeung, Sackin 1999)

Nouns and Verbs

Nouns and Verbs, with the dependent categories of Adjectives and Adverbs, together form the open class of content words.

Selective impairment of word categories such as nouns vs verbs has suggested a regional representation of lexical knowledge in the human brain. (Dehaene 1995)

Performance of two brain-damaged subjects with modality-specific deficits restricted principally (H.W.) or virtually only (S.J.D) to verbs in oral and written production, respectively. (Caramazza and Hillis 1991)

Sparing of verbs and preserved, but ineffectual reading in a patient with impaired word production. This pattern of deficits suggests two functional lesions, one affecting the connections between the semantic store and the phonological lexicon and the other damaging the sublexical route that converts sound to sound and sound to print. It also implies that words are independently organized in the phonological lexicon, based on their grammatical class and have discrete connections with the semantic store. However, CT scan evidence does not support the hypothesis that this functional dissociation finds its anatomical correlate in the specialization of the frontal premotor cortex for verbs and the antero-medial temporal cortex for nouns.(De Renzi , di Pellegrino 1995 )

The neural correlates of verb and noun processing. A PET study. The hypothesis that categorical information, distinguishing among word classes, such as nouns, verbs, etc., is an organizational principle of lexical knowledge in the brain, is supported by the observation of aphasic subjects who are selectively impaired in the processing of nouns and verbs. These findings are compatible with the view that lexical-semantic processing of words is mediated by an extensive, predominantly left hemispheric network of brain structures. Additional brain activations appear to be related to specific semantic content, or, in the case of verbs, may be associated with the automatic access of syntactic information. (Perani et al.1999)

Nouns and verbs are retrieved with differently distributed neural systems. In agreement with recent evidence from brain-damaged subjects, these results provide evidence that (1) nouns and verbs have distinct neural generators and that (2) these generators involve areas outside the classical language regions of the brain. (Damasio and Tranel 1993)

Brain processing of grammatical word class was studied analyzing event-related potential (ERP) brain fields. The results indicate that different neural populations represent different grammatical word classes in language processing, in agreement with clinical observations. This word class differentiation as revealed by the spatial-temporal organization of neural activity occurred at a time after word input compatible with speed of reading. (Perani et al.1999)

Nouns and verbs in the intact brain: evidence from event-related potentials and high-frequency cortical responses. Lesion evidence indicates that words from different lexical categories, such as nouns and verbs, may have different cortical counterparts. As soon as approximately 200 ms after stimulus onset, event-related potentials disclosed electrocortical differences between nouns and verbs over widespread cortical areas. Our results are consistent with a neurobiological model of language representation postulating cell assemblies with distinct cortical topographies as biological counterparts of words. (Pulvermuller, Lutzenberger, Preissl 1999)

Grammatical differences alone, e.g. between two lexical classes such as action verbs and action-related nouns, are not sufficient for eliciting differential brain responses. In contrast, semantic differences between items from the same lexical category can be sufficient for changing the topography of cortical processes induced by word stimuli. (Pulvermuller, Mohr, Schleichert 1999)

Category-specific ERP differences began to appear around 260 ms. There was a left temporo-parietal negativity for animal names and verbs, a left inferior temporal negativity for proper names, and a bilateral positivity for numerals. These results provide a bilateral parietal positivity evidence for timing and coarse localization of category-specific word processing in the normal human brain. (Dehaene 1995)

Lexical processing of visually and auditorily presented nouns and verbs: evidence from reaction time and N400 priming data. We found that the temporal patterns of primed RTs and N400 latencies differed for nouns and verbs indicating a functional difference in processing. However, the absence of topographic differences in N400 between nouns and verbs did not support anatomically distinct representations of these word classes. (Gomes et al. 1997)

Concrete and abstract words

Neural pathways involved in the processing of concrete and abstract words. A direct comparison between the abstract and concrete stimuli epochs yielded a significant area of activation in the right anterior temporal cortex. The results are consistent with recent positron emission tomography work showing right hemisphere activation during processing of abstract representations of language. The results are interpreted as support for a right hemisphere neural pathway in the processing of abstract word representations. (Kiehl et al.1999)

Abstract content words: One may argue that the postulated difference in semantic meaning between content and function words does not apply for all members of these vocabulary classes. Rather, it appears that there is a continuum of meaning complexity between the "simple" concrete content words that have clearly defined entities they can refer to (so-called referents), more abstract items that may or may not be used to refer to objects and actions, and function words that cannot be used to refer to objects. If semantic criteria are crucial for intra-cortical representation, the suggested gradual differences in the correlation between word form and meaning-related stimuli or actions should be reflected in gradual differences in cortical lateralization and distributedness of assemblies. An abstract content word, such as "philosophy", may therefore have an assembly somewhat in-between typical content and function word assemblies: It may exhibit an intermediate degree of laterality mainly consisting of perisylvian neurons but including a few neurons clusters outside perisylvian areas. Among the abstract content words are words referring to emotional states, for example "anger" and "joy". For these words, it is not difficult to find characteristic visual stimuli related to their meaning - for example angry or a joyful faces. In addition, there are characteristic meaning-related patterns of muscle activity - namely the contraction of the respective face muscles - and autonomic nervous system activity . Although these words do not refer to objects and actions in the sense in which the word "house" refers to an object, the likely co-occurrence of body movements and visual stimuli and patterns of muscle contractions with the word forms may nevertheless lead to the formation of widely distributed cortical cell assemblies representing these words. These considerations should make it clear that the degree of abstractness of an item is not the only factor influencing assembly topographies. According to the present proposal, the important criterion is the strength of the correlation of the occurrences of a given word form and a class of non-linguistic stimuli or actions. In the clear cases, this likelihood is related to abstractness, but there are exceptions. (Pulvermuller 1999)

Animate and inanimate words

Word and picture matching: a PET study of semantic category effects: We report two positron emission tomography (PET) studies of cerebral activation during picture and word matching tasks, in which we compared directly the processing of stimuli belonging to different semantic categories (animate and inanimate) in the visual (pictures) and verbal (words) modality. These results are compatible with different brain networks subserving the identification of living and non-living entities; in particular, they indicate a crucial role of the left fusiform gyrus in the processing of animate entities and of the left middle temporal gyrus for tools, both from words and pictures. The activation of other areas, such as the dorsolateral frontal cortex, appears to be specific for the semantic access of tools only from pictures. These results are compatible with different brain networks subserving the identification of living and non-living entities; in particular, they indicate a crucial role of the left fusiform gyrus in the processing of animate entities and of the left middle temporal gyrus for tools, both from words and pictures. The activation of other areas, such as the dorsolateral frontal cortex, appears to be specific for the semantic access of tools only from pictures.(Perani et al. 1999)

Patients with cerebral lesions offer a unique opportunity to investigate the organization of meaning systems in the brain. Clinical neurologists have long been aware that knowledge of particular classes or categories of information may be selectively impaired in some cases and selectively spared in others. For example knowledge of letters, colours, objects, or people may be lost as a consequence of damage to the left hemisphere of the brain. (McCarthy and Warrington 1988)

The analysis of neuropsychological disorders of lexical processing has provided important clues about the general organization of the lexical system and the internal structure of the processing components. Reports of patients with selective dysfunction of specific semantic categories such as abstract versus concrete words, living things versus inanimate objects, animals, fruits and vegetables, proper names and so forth, support the hypothesis that the neural organization of the semantic processing component is organized in these categories. (Caramazza and Hillis 1991)

The evidence for knowledge of living things as compared with inanimate objects is particularly striking. Such observations have suggested that our semantic knowledge base is categorical in its organization.(McCarthy and Warrington 1988)

Evidence for selective impairments in knowledge of living and nonliving things. Beginning in the 1980's, Warrington and her colleagues began to report the existence of patients with selective impairments in knowledge of either living or nonliving things. Warrington and Shallice (1984) described four patients who were much worse at identifying living things (animals, plants) than nonliving things (inanimate objects). All four of these patients had recovered from Herpes encephalitis, and all had sustained bilateral temporal lobe damage. Two of the patients were studied in detail, and showed a selective impairment for living things across a range of tasks, both visual and verbal. One of these subjects was also tested with a completely nonverbal matching task, in which different-looking depictions of objects or animals were to be matched to one another in an array, and showed the same selective preservation of knowledge of animals relative to inanimate objects. (Farah 1994)

However, Warrington and colleagues have suggested an alternative interpretation, according to which semantic memory is fundamentally modality-specific. They argue that selective deficits in knowledge of living and nonliving things may reflect the differential weighting of information from different sensorimotor channels in representing knowledge about these two categories. They have pointed out that living things are distinguished primarily by their sensory attributes, whereas nonliving things are distinguished primarily by their functional attributes. For example, our knowledge of an animal such as a leopard, by which we distinguish it from other similar creatures, is predominantly visual. In contrast, our knowledge of a desk, by which we distinguish it from other furniture, is predominantly functional (i.e., what it is used for.) Thus, the distinctions between impaired and preserved knowledge in the cases reviewed earlier may not be "living/nonliving" distinctions per se, but "sensory/functional" distinctions. The modality-specific hypothesis seems preferable to a strict semantic hypothesis for two reasons. First, it is more consistent with what is already known about brain organization. It is well known that different brain areas are dedicated to representing information from specific sensory and motor channels. Functional knowledge could conceivably be tied to the motor system. (Farah 1994)

Vision and action words

Words may be encoded in areas of the brain that correspond to their meaning, say scientists in Germany. Neuroscientists have suggested half a dozen possible sites in the brain where semantic processing occurs. But Friedemann Pulvermüller and his colleagues at the University of Konstanz thought there may be more than one region. Words referring to movement could be coded in the motor cortex, while hearing-related words may be coded in the auditory cortex. They recruited nine volunteers whose right motor cortex had been damaged by a stroke, leaving their left arm paralysed. The patients were asked to quickly decide whether certain words were real or not. Some words, such as "write", were associated with movement, while others, such as "tiger", provoked images. None of the volunteers had obvious language disorders. But with action words, they made more mistakes than people with normal brains. On visual words they did as well as healthy people. Pulvermüller thinks the difficulties occur because action words are stored in the motor cortex. He believes other brain regions that deal with different stimuli--from hearing to emotions--may also deal with relevant words. "All cortical areas can be relevant," he says. He believes this is not surprising since we learn words by association. For instance, we often remember words after seeing, hearing or doing what the words describe. (New Scientist 1998)

Based on the studies summarized, it appears likely that motor, premotor and/or prefrontal cortices and possibly additional areas in middle temporal gyrus contribute to the processing of action words, whereas inferior temporal and/or occipital areas close to the primary visual cortex can be involved in processing vision words. The observed physiological double dissociations provide additional support for the idea that semantic characteristics of words determine the loci of their cortical processing. (Pulvermuller 1999)

Content words are used to refer to odors, tastes, somatic sensations, sounds, visual perceptions, and motor activities. During language learning, word forms are frequently produced when stimuli the words refer to are perceived or actions they refer to are carried out by the infant. If the cortex is an associative memory, the modalities and processing channels through which meaning-related information is being transmitted must be important for formation of cortical assemblies. This has inspired recent models of word processing in the brain postulating distinct cortical representations for word classes that can be distinguished based on semantic criteria. If the modality through which meaning-related information is transmitted determines the cortical distribution of cell assemblies, a fundamental distinction between action and perception words can be made. Action words would refer to movements of the own body and would, thus, be frequently used when such actions are being performed. In this case, a perisylvian assembly representing the word form would become linked to neurons in motor, premotor and prefrontal cortices related to motor programs. Perception words whose meaning can best be explained using prototypical stimuli would consist of a perisylvian assembly plus neurons in posterior cortex. In many cases, visual stimuli are involved and the respective word category may therefore be labelled vision words. The argument made above for action and vision words can be extended to words referring to stimuli perceived through other modalities. For those, additional word categories - odor, taste, pain, touch, and sound words - can be postulated. Members of these word classes should be represented in assemblies with specific cortical topographies. Neuroimaging studies were used to evaluate the hypothesis that words are cortically represented in distributed cell assemblies with defined topographies that vary with semantic word properties. Electrophysiological and metabolic imaging studies provide evidence that not only the language cortices in the left hemispheres, but additional cortical areas outside the left perisylvian areas play a role in word processing. Comparison of ERP responses to content and function words indicate that both hemispheres are strongly involved in processing concrete content words, whereas predominantly left-lateralized activity in or close to perisylvian regions appear to be related to processing of highly abstract function words. PET, fMRI and ERP studies revealed that cortical areas devoted to motor programming or visual perception are activated when words with strong motor or visual associations are being processed.(Pulvermuller 1999)

Thus grammatical classes arise which have a probabilistic relation to semantic classes rather than an exact correlation. Of course this example is an oversimplification of the actual processes involved; in particular it might be asked: where has the class labelled verb come from? The answer is that it arises by a similar process from the semantic class of action words. If the where-in-the-brain question can be answered in the case of content, function, action and vision words, the Hebbian approach also provides a tentative answer to the question of why these different localizations develop and why they involve the cortical lobes and gyri they probably do. Perception words whose meaning can best be explained using prototypical stimuli would consist of a perisylvian assembly plus neurons in posterior cortex. In many cases, visual stimuli are involved and the respective word category may therefore be labelled vision words. Assemblies representing words of this category would be distributed over perisylvian and visual cortices in parietal, temporal and/or occipital lobes.

Examples of words whose meaning is related to the visual modality are concrete nouns with well-imaginable referents, such as animal names. The best examples of action words are in the category of action verbs. While verbs referring to body movements are likely to be action words, and many concrete nouns (such as animal names) are almost certainly vision words, other word groups - for example nouns referring to tools - probably lead to both visual and motor associations. When evaluating the present ideas about word class-differences related to word meaning in neuroscientific experiments, it is, therefore, most important to quantitatively assess semantic associations elicited by word stimuli. The only way to do this is by asking study participants. Based on the studies summarized, it appears likely that motor, premotor and/or prefrontal cortices and possibly additional areas in middle temporal gyrus contribute to the processing of action words, whereas inferior temporal and/or occipital areas close to the primary visual cortex can be involved in processing vision words. The observed physiological double dissociations provide additional support for the idea that semantic characteristics of words determine the loci of their cortical processing. It is not yet clear, however, to which degree primary, secondary and higher-order association cortices, respectively, participate in word processing. The results (ERPs and PET) summarized here would suggest that both lower- and higher-order sensory and motor cortices, as well as multimodal association areas can play a role. A further open question concerns the contribution of the right hemisphere to word class-specific processes. Most ERP and fMRI investigations advocate such right-hemispheric contributions whereas most PET studies available at present do not.(Pulvermuller 1999)

In earlier studies, distinct neural systems were found to be required for the retrieval of words denoting actions versus those denoting objects. (Damasio & Tranel 1993). A double dissociation was found where some patients with lesions in one area of the brain could not access action words but had no problem with objects, and other patients with lesions in non-overlapping areas showed the reverse problem. Such converging evidence argue in support of a lexicon in which words are organized or connected according to semantic categories, but not necessarily limited to such an organization, since we have other evidence for phonologically connected words, and for syntactic categories, as well as for a separation, as shown above, of lexical and grammatical categories.(Fromkin 1997)

[A comment on the Internet] "But isn't it interesting that a person can lose the ability to name round fruits (apples, oranges, plums) when shown them (though he can call them "fruit"), but has no trouble with bananas? And, that the same individual has problems with many other categories of nouns (can't name any birds, members of the cat family, or of the horse family--nor raccoons; can name elephants and giraffes)"

3. Syntax in the brain

From Bates 1994:

More recent studies of language breakdown in aphasia have forced investigators to abandon the idea of a "grammar box", i.e. neural tissue that is devoted exclusively to grammar, and contains the representations that are necessary for grammatical processing. There are (1) numerous studies showing that so-called agrammatic aphasics can make remarkably fine-grained judgments of grammaticality and (2) a host of cross-linguistic studies showing differences in the symptoms displayed by agrammatic patients in different language communities -- differences that can only be explained if we acknowledge that the patient still retains detailed knowledge of his/her grammar. It begins to look as though linguistic knowledge is broadly represented in the adult brain -- a conclusion that is also supported by studies of brain activity during normal language use. So what is localized? The classic sensorimotor view of Broca's and Wernicke's aphasia has fallen by the wayside, and now the grammar/semantics view has fallen as well. But their successor is still unnamed. Some investigators have argued that left frontal regions are specialized for the rapid processes required for fluent use of grammar, while posterior regions play a more important role in controlled, strategic choice of words and sentence frames. These ideas are still distressingly vague. In contrast with the best-known examples of innate and domain-specific brain systems, the systems that support languages also show an extraordinary and perhaps unprecedented degree of neural plasticity. Research on the long-term effects of early focal brain injury suggests that children with large lesions to the classic language zones go on, more often than not, to attain levels of language ability that are indistinguishable from normal. Although it is difficult to compare apples and oranges, it looks as though there may be more plasticity for language than we observe in other perceptual and cognitive systems. Arbitrariness of form-meaning mapping. . A defining characteristic of language (indeed, one of its few universals) is the arbitrariness of the relationship between sound and meaning. ). The words "dog", "chien", "perro", "cane", "Hund", etc. do not in any way resemble the fuzzy four-legged creatures that they signify. The same is true for the relationship between grammatical forms and the communicative work that those forms carry out. For example, depending on the language that one speaks, basic information about "who did what to whom" can be signalled through word order (as it is in English), case inflections on nouns (e.g. Latin, Russian, Hungarian), agreement marking between subject and verb (a major source of information in Italian, but only a minor source in English), and a range of other cues. (Bates 1994)

From Grodzinsky 2000:

A new view of the functional role of left anterior cortex in language use is proposed. The experimental record indicates that most human linguistic abilities are not localized in this region. In particular, most of syntax (long thought to be there) is not located in Broca's area and its vicinity. Linguistic theory evolves, at times even rapidly, and terms tend to change with theoretical perspective. Certain generalizations remain stable, however, because they deal with basic syntactic phenomena. The relationship between traces and their antecedents in movement operations plays a central role in a large class of theories. In the main, then, the presentation here is compatible with most current theoretical frameworks, including the Minimalist Program. The current neuroimaging literature makes only tenuous connections to the lesion-based body of knowledge, and focuses, for the most part, on semantic and phonological aspects of the mental lexicon. Fewer works concentrate on the most salient aspect of language - its combinatory nature. Lesions to Broca's area and its vicinity do not affect semantic abilities, nor do they disrupt basic syntactic abilities. Most notably, Broca's aphasics combine lexical meaning into propositions, create and analyze sentences of considerably complex structure, and are also able to synthesize and analyze words morphophonologically. It thus follows that most human linguistic abilities, including most syntax, are not localized in the anterior language areas - Broca's area and deeper white matter, operculum, and anterior insula.

From Deacon 1997:

Neither grammar, nor syntax, nor articulate sound production, nor a huge vocabulary have kept other species from evolving languages. Just the simple problem of figuring out how combinations of words refer to things. The structure of syntax often only vaguely conceals its pragmatic roots in pointing gestures. The theory of deep structures paints itself into an evolutionary corner ... Chomsky's abandonment of Darwinian explanations for innate language knowledge is at least consistent. But what is left? Only certain structural features of language could have become internalised as part of a 'language instinct', and these turn out not to be those that are most often cited as the core of a Universal Grammar. Language areas [Broca Wernicke] are simply not the repositories of linguistic skill and knowledge that they were once thought to be. ... do not support any simple localisation of language functions.

From Fromkin 1997:

A more telling case can be made if there are individuals who have acquired the highly complex system which we call grammar, without parallel abilities of equal complexity. There are now a number of such studies of children who have few cognitive skills and virtually no ability to utilize language in sustained meaningful communication and yet have extensive mastery of linguistic structure. Yamada (1990) reports on one severely retarded young woman, named Laura with a nonverbal IQ of 41-44, lacking almost all number concepts including basic counting principles, drawing at a preschool level, and processing an auditory memory span limited to three units, who at the age of 16 produced syntactically complex sentences like 'She does paintings, this really good friend of the kids who I went to school with last year and really loved.' Although Laura produces sentences with multiple embeddings, can conjoin verb phrases, produce passives, inflect verbs for number and person to agree with the grammatical subject, and forms past tenses when the time adverbial structurally refers to a previous time, she can not add 2 + 2, read nor write nor tell time. She does not know who the president of the US is or what country she lives in and does not know her own age. Her drawing of humans resemble potatoes with stick arms and legs. Yet, in a sentence imitation task she both detected and corrected surface syntactic and morphological errors, but is unable to tie her shoes. One of their cases, Anthony, at the age of 5 yrs 2 months is reported to have had a Leiter IQ of 50 and a mental age of 2 years 9 months. His non-linguistic cognitive tests scored below all the norms/and or below the 2-year old level, contrasting sharply with his high scores on language tests and his spontaneous speech in which he used 61 of the 68 different elements and structures analyzed, including infinitival and sentential complements, relative clauses and other subordinate clauses. They conclude: "His ability to use a wide range of syntactic encode his limited and confused thoughts, illustrates the discrepancy between Anthony's grammatical knowledge and his conceptual/cognitive knowledge." [Another subject]She has been unable to learn to read and write in her late teenage years and cannot add or handle money, yet D.H. performs almost without error on grammaticality judgments. A similar and perhaps even more dramatic case is reported on by Smith and Tsimpli (1995) of Christopher, a 34 year old man, who is institutionalized because he is unable to take care of himself. As Smith and Tsimpli report, he finds the tasks of buttoning a shirt, cutting his finger-nails or vacuuming the carpet too difficult. Yet when given written texts in some 15 or 16 languages he translates them immediately into English. The languages include Germanic languages like Danish, Dutch, and German, Romance languages like French, Italian, Portuguese, Spanish, as well as Polish, Finnish, Greek, Hindi, Turkish, and Welsh.

In 1980, John Marshall stated: "Biologists...have accumulated a vast body of knowledge concerning the gross anatomy of those parts of the central and peripheral nervous system which seem to be implicated in the acquisition and exercise of linguistic abilities. Some knowledge is even available about the slightly less gross physiology of the relevant brain areas....Psycholinguists...have amassed alarming amounts of data of the progression from the birth cry to the multiply embedded relative clause. The problem is...found in the simple fact that no one .. has the slightest idea how to relate these two domains of inquiry to each other....We have so far failed ... to construct ... models ...that could mediate between noun phrases and neurons." We have come a long way in the sixteen years since this statement appeared in print. We still are unable to connect the noun phrase to a neuron, but we have already some idea of the neural architecture underlying different lexical categories and the neural modules underying different cognitive systems.

From Levelt 1999:

Clearly driven by a genetic endowment, children restructure their system of lexical concepts by a process of syntactization. Lexical concepts acquire syntactic category and subcategorization features, verbs acquire specifications of how their semantic arguments (such as agent or recipient) are to be mapped onto syntactic relations (such as subject or object), nouns may acquire properties for the regulation of syntactic agreement, such as gender, etc. More technically speaking, the child develops a system of lemmas[1], packages of syntactic information, one for each lexical concept. At the same time, the child quickly acquires a closed class vocabulary, a relatively small set of frequently used function words. These words mostly fulfill syntactic functions; they have elaborate lemmas but lean lexical concepts. This system of lemmas is largely up and running by the age of four. From then on, producing a word always involves the selection of the appropriate lemma.

From Pulvermuller 1990:

A word on syntax: For many language scientists, the question of how words are represented and processed in the brain is only a very basic one. Even substantial improvements in our understanding of the mechanisms underlying word processing may therefore be acclaimed only if the theoretical framework they are based on can offer perspectives on the solution of more complex problems known to be crucial for language. The question of how serial order is achieved in syntactic word strings has long been considered to aim at the heart of our language faculty, and, from a theoretical point of view, it appears important to make clear whether a neurobiological model of word processing can offer perspectives on the biological reality of grammar. Although it is not possible to discuss syntactic issues in more detail here, these remarks on center-embedding may suffice to show that neurobiological models of language are not necessarily restricted to the single word level. In fact, they can offer perspectives on the problem of serial order in behavior that meet linguists' claims that language mechanisms cannot be understood without considering brain mechanisms.