Wordtrade.com
Psychology

Mental States

Mental States: Volume 1: Evolution, function, nature; Volume 2: Language and cognitive structure (set) by Andrea C. Schalley and Drew Khlentzos (Studies in Language Companion Series: John Benjamins Publishing) Collecting the work of linguists, psychologists, neuroscientists, archaeologists, artificial intelligence researchers and philosophers this volume presents a richly varied picture of the nature and function of mental states. Starting from questions about the cognitive capacities of the early hominin homo floresiensis, the essays proceed to the role mental representations play in guiding the behaviour of simple organisms and robots, thence to the question of which features of its environment the human brain represents and the extent to which complex cognitive skills such as language acquisition and comprehension are impaired when the brain lacks certain important neural structures. Other papers explore topics ranging from nativism to the presumed constancy of categorization across signed and spoken languages, from the formal representation of metaphor, actions and vague language to philosophical questions about conceptual schemes and colours. Anyone interested in mental states will find much to reward them in this fine volume.

This set of a two-volume collection on mental states. The contributions to this volume focus on evolutionary and functional aspects of certain mental states in an effort to understand their nature, whereas the second volume is concerned with the question what language and language use reveals about cognitive structure and under­lying cognitive categories.

Questions that are addressed in this volume include: (i) how early did cognitive states of a sort rich enough to support communication and planning appear in the evolutionary history of hominids?; (ii) is it possible to infer the existence of sophisti­cated cognitive states from evidence of tool use?; (iii) how do mental states represent situations or events or actions?; (iv) how can we theoretically model mental states?; (v) how can we simulate mental states and their functions?; and (vi) what insights can conceptual categorisation — both linguistic and non-linguistic — give us into the or­ganisation and structure of the mind and hence of mental states?

The contributions to the second volume focus on what language and language use reveals about cognitive structure and underlying cognitive categories. Wide-ranging and thought-provoking essays from linguists and psychologists within this volume investigate the insights conceptual categorization can give into the organization and structure of the mind and specific mental states. Topics and linguistic phenomena discussed include narratives and story telling, language development, figurative language, linguistic categorization, linguistic relativity, and the linguistic coding of mental states such as perceptions and beliefs. With contributions at the forefront of current debate, this book will appeal to anyone with an interest in language and the cognitive structures that support it.

The second in the two-volume collection on mental states. The contri­butions to this volume focus on the question what language and language use re­veals about cognitive structure and underlying cognitive categories, whereas the first volume is concerned with evolutionary and functional aspects of certain men­tal states in an effort to understand their nature.

The contributions to this volume address the question what insights concep­tual categorisation can give us into the organisation and structure of the mind and thus of mental states. Topics and linguistic phenomena investigated under this view include narratives and story telling, language development, figurative lan­guage, questions of linguistic categorisation, linguistic relativity, and more gener­ally the linguistic coding of mental states (such as perceptions and attitudes).

The set contains contributions from psychologists, linguists, artificial intelli­gence researchers, neuroscientists, archaeologists and philosophers, bringing together scholars from the diverse fields of cognitive science, or more specifically, the study of language and cognition. This reflects the provenance of the chapters, most of which were presented at the International Language and Cognition Conference, held in Sep­tember 2004 at Pacific Bay Resort in Coffs Harbour, Australia.

Mental States: Volume 1: Evolution, function, nature by Andrea C. Schalley and Drew Khlentzos (Studies in Language Companion Series: John Benjamins Publishing) presents a rich diversity of views from researchers in cognitive sci­ence and associated disciplines - archaeology, artificial intelligence, linguistics, neuroscience, philosophy, and psychology - on the nature, function and evolution of "mental" or "cognitive" states. A quick glance at the titles of the contributions and/or the disciplinary backgrounds of the contributors might lead one to suspect little commonality in theoretical interests. However, this would be a mistake. Al­though the contributions differ markedly in approach and methodology, common questions about mind and cognition unite them.

One question, addressed in the first three chapters of this volume, concerns the evolution of cognitive states: How early did cognitive states of a sort rich enough to support communication and planning appear in the evolutionary history of homin­ids? What types of evidence could be marshalled to decide this question - is it pos­sible to infer the existence of sophisticated cognitive states from the evidence of tool use, for example?

A different question concerning mental states is addressed in the four chapters that follow. This question concerns the function of cognitive states: How do mental states represent situations or events or actions? What types of representation must the brain process to achieve this feat? It is extraordinarily difficult to break free of the Cartesian intuition that the mind is wholly distinct from anything physical, the brain included. Descartes notoriously thought that animals were mere automata, not possessed of mentality at all. Even though contemporary work on animal in­telligence makes this view seem implausible, Descartes' challenges have become no easier to answer: What exactly is a mind? Which creatures have one?

The remaining six chapters deal, in one way or another, with the insights that conceptual categorisation (both linguistic and non-linguistic) can give into the organisation and structure of the mind. Although approaches range from the purely theoretical to the experimental, all of these chapters seek some unifying theoretical account of the nature of categorisation, and thus the nature of mind.

With these general observations behind us, let us turn to specific issues raised in the individual chapters. Since the dramatic discovery in September 2003 of Homo floresiensis on the island of Flores by Michael Morwood and his team, de­bate has raged over whether this tiny hominin, nick-named "the hobbit", was a new species or not. How did a creature half the size of an adult human with a brain smaller than a chimpanzee's manage to traverse the deep waters separating the is­land of Flores from the mainland? If this tiny hominin is a new species, from which hominin did it evolve? Could these diminutive hominins really have hunted mas­sive prey such as Stegodon or were they merely opportunistic scavengers? How could they have planned such hunts without sophisticated forms of communica­tion only a language makes possible?

From the start a vigorous sceptical view has maintained Horno floresiensis is something of a misnomer: The skeletal remains found at Liang Bua were those of a microcephalic form of modern human. However, this theory is hard-put to explain the latest findings of stone tools, remarkably similar to the ones found in 2003 to­gether with the "hobbit" skeletal remains - the tools date to 880,000 years ago, more than 600,000 years before Homo sapiens appeared on the evolutionary scene.

The first three chapters - presenting conflicting assessments and therefore pro­viding a stimulating basis for future discussions - focus on Homo floresiensis' cogni­tive abilities as inferred from the design of their stone tools. Mark Moore in his chapter "Lithic design space modelling and cognition in Homo floresiensis" develops a sophisticated theoretical model that permits one to infer cognitive abilities from stone flaking. On the assumption that increases in the complexity of the flake unit are reliable indicators of advancing cognitive ability, Moore argues that Homo floresiensis most probably did not possess a high level of cognitive ability. The rea­son? Homo floresiensis simply chained together the simplest types of flakes without any attempt to stack them hierarchically as can be expected from a creature capable of advanced thinking and planning.

The question of Homo floresiensis' small brain size in comparison to that of other hominins and the inferences it sustains to cognitive ability is directly ad­dressed by Iain Davidson in his chapter "`As large as you need and as small as you can': Implications of the brain size of Homo floresiensis". Davidson argues that the evolutionary pattern of changes in hominin brain size is not a simple "onwards and upwards" trend. Rather, it is a punctuated pattern explained by the interaction between body size and brain size. These two factors in turn need to be given a se­lectionist explanation, Davidson contends. As large brains are expensive to main­tain there is a definite selectionist pressure against their emergence. The most plausible explanation for the disappearance of hominins with small brains is that they could not compete with the cognitive advantages conferred upon their rivals by their larger brains, Davidson argues. However, in the isolated environment of Flores, no selectional pressures emerged to counteract the default choice of small­er body and brain size. What are the implications for the cognitive abilities of Homo floresiensis? Davidson thinks this can only be answered by looking to the behaviour of Homo floresiensis, in particular, their tool use, their use of fire and their hunting.

In their contribution "Homo on Flores: Some early implications for the evolu­tion of language and cognition" Michael Morwood, the discoverer of Homo floresiensis, and Dorothea Cogill-Koez investigate the implications of the Flores find for the evolution of language and cognition more broadly, drawing on diverse evidence from different disciplines. They integrate a meta-theoretical discussion on models in evolutionary theory and stress the risky nature of any inferences to cognitive capacity from the fossil record of bones and stone implements. Nonethe­less they are hopeful that the mass of diverse evidence from "animal communica­tion to zoogeography" will impose enough specific constraints to filter out in the end all but the uniquely correct explanatory hypothesis.

Putting the special case of Homo floresiensis to one side we are left with some unresolved general questions, in particular about the correlation between brain size and cognitive ability: Are bigger brains invariably better cognitively? What minimum brain size is necessary to sustain higher cognitive processing? These are questions about the function and nature of mental states, and it is these issues that are explored in very different ways in the next four chapters.

In "Evolving artificial minds and brains" Pete Mandik, Mike Collins and Alex Vereschagin argue for the need to posit mental representations in order to explain intelligent behaviour in very simple creatures. The creature they choose is the nem­atode worm and the behaviour in question is chemotaxis. Many philosophers think that a creature's brain state or neural state cannot count as genuinely mental if the creature lacks any awareness of it. Relatedly, they think that only behaviour the creature is conscious of can be genuinely intelligent behaviour. When the standards for mentality and intelligence are set so high, very few creatures turn out to be ca­pable of enjoying mental states or exhibiting intelligent behaviour. Yet the more we learn about sophisticated cognitive behaviour in apparently simple organisms the more tenuous the connection between mentality and consciousness looks.

If there is a danger in setting the standards for mentality and intelligence too high, there is equally a danger in setting them too low, however. Many cognitive scientists would baulk at the suggestion that an organism as simple as a nematode worm could harbour mental representations or behave intelligently. Yet Mandik, Collins and Vereschagin argue that the worm's directed movement in response to chemical stimuli does demand explanation in terms of certain mental representa­tions. By "mental representations" they mean reliable forms of information about the creature's (chemical) environment that are encoded and used by the organism in systematic ways to direct its behaviour.

To test the need for mental representations they construct neural networks that simulate positive chemotaxis in the nematode worm, comparing a variety of networks. Thus networks that incorporate both sensory input and a rudimentary form of memory in the form of recurrent connections between nodes are tested against networks without such memory and networks with no sensory input. The results are then compared with the observed behaviour of the nematode. Their finding is that the networks with both sensory input and the rudimentary form of memory have a distinct selectional advantage over those without both attributes.

Even if it is too much to require mental states to be conscious, there is still the sense that there is more to mentality than tracking and responding to environ­mental states. One worry is that there is simply not enough plasticity in the nema­tode worm's behaviour to justify the attribution of a mind. A more important worry is that the nematode does not plan - it is purely at the mercy of external forces pushing and pulling it in the direction of nutrients. In this regard, it is in­structive to compare the behaviour of the nematode worm with the foresighted behaviour of the jumping spider, Portia Labiato. Portia is able to perform some quite astonishing feats of tracking, deception and surprise attack in order to hunt and kill its (often larger) spider prey. Its ability to plot a path to its victim would tax the computational powers of a chimpanzee let alone a rat. It has the ability to plan a future attack even when the intended victim has long disappeared from its sight. Portia appears to experiment and recall information about the peculiar habits of different species of spiders, plucking their webs in ways designed to arouse their interest by simulating the movements of prey without provoking a full attack. Yet where the human brain has 100 billion brain cells and a honeybee's one million, Portia is estimated to have no more than 600,000 neurons!

How large must a brain then be to sustain mental states? Having looked at evo­lution theory and the construction of neural networks, approaches that promise to shed further light on this issue include those that simulate and model merely the functionality: Artificial intelligence researchers aver brains are not strictly necessary - all that is required is the relevant functional capacity of the brain. In their contri­bution "Multi-agent communication, planning and collaboration based on percep­tions, conceptions and simulations", Peter Gärdenfors and Mary-Anne Williams develop a conceptual framework for grounded communication between agents which they test in the domain of soccer-playing robots. These robots need to assess and then communicate to their fellows rapidly-changing information about the lo­cation and movement of the ball, opposition players and other team members.

The Gärdenfors-Williams framework makes use of the idea of conceptual spaces to model the changing conditions governing items in the robots' world. The concepts in these conceptual spaces categorise information geometrically by means of similarity to given exemplars. Similarity is a matter of the distance be­tween a given item and an exemplar in a multi-dimensional similarity space on this view. Concepts are represented by sets of convex regions in various domains. Gärdenfors and Williams distinguish concepts or mental representations that are cued to current environmental goings-on from those that are detached from such proceedings. Conceptual spaces then bridge sensorimotor information to de­tached representations of various sorts.

An interesting feature of their theoretical framework is the crucial role simula­tions play in it - and that they do so on different levels in their theoretical frame­work. Indeed, Gärdenfors and Williams hold that simulation underwrites all higher thought processes. In the case of their soccer-playing robots, each robot must form a world-model that has to be properly grounded to ensure success in collaboratively planned actions.

In the following chapter "The modal-logical interpretation of the causation of bodily actions" Hiroyuki Nishina provides a formal model of the various rotations and turns an agent's body makes when performing various actions. He builds on Ray Jackendoff's work and David Marr's 3D model. Nishina uses modal-logical formulae that are evaluated in relation to rotation/turn functions to describe the characteristic movements associated with various actions. A causal relation ties one set of movements to another in the performance of a typical action. Nishina then attempts to show that the conceptual structure underlying action verbs such as kick really does correlate with the modal-logical representation of the turning/rotation­al microstructure of such actions.

It would be an interesting exercise to see whether Nishina's modal-logical rep­resentations of the rotation of various joints could be incorporated into the simu­lations of Gärdenfors and Williams' soccer-playing robots to refine prediction of other robots' movements as well as control of their own.

Anna Borghi, Claudia Bonfiglioli, Paola Ricciardelli, Sandro Rubichi and Rob­erto Nicoletti in their contribution "Do we access object manipulability while we categorize?" set out to experimentally determine whether information on the af­fordances of objects (the types of interactions we can have with them) is indeed part of our conceptual representation of them. They designed two experiments to test subjects' ability to categorise things. They construe their results as evidence that the brain responds preferentially to information about object manipulability rather than to information about an object's function, and - given that their ex­periments are based on drawings or words - provide evidence that recognition of pictures or words denoting objects automatically activates motor representations associated with these objects.

In their chapter "Speaking without the cerebellum: Language skills in a young adult with near total cerebellar agenesis" Alessandro Tavano, Franco Fabbro and Renato Borgatti present the results of a case study of a person born with cerebellar agenesis who gradually acquired adequate communicative skills over a twenty year period. They point out that whilst the part played by the cerebellum and other sub-cortical structures such as the basal ganglia in the execution, coordination and timing of motor actions has long been appreciated, it is only recently that research­ers have realised these structures also have an important role in language-learning, and, indeed, other forms of higher cognition. Precisely what that role is, however, remains unclear. Perhaps it is restricted to the coordination and sequencing of those processes that go to make up language comprehension and production? If so, the primary function of these subcortical strued many efforts to examine issues of everyday memory scientifically. Examples of this line of research are studies of autobiographical memory, studies of the memory for past and prospective actions and events, and studies of the to language-processing. Tavano, Fabbro and Borgatti note there are two separate types of learning involved in language acquistion: procedural and declarative. De­clarative knowledge is explicit and propositional. Our knowledge of the lexicon is declarative. Knowledge of phonology and morphosyntax on the other hand is largely implicit and resistant to verbal formulation - knowledge acquired in the learning process, without reflection on the principles governing it. The aim of the study of the patient G. R. was to determine whether language-learning was possi­ble in the presence of cerebellar agenesis and to investigate the grammatical and lexical limitations on acquisition in the absence of the usual procedural-learning support provided by the cerebellum. They conclude that even though many of G. R.'s cerebellar functions seem to have been reassigned to cerebellar networks, this relocation is far from complete and that normal language acquisition is very far from being an exclusively neocortical concern.

A distinctive feature of human cognition is the effortless ability to construct analogies and interpret metaphors. How do we do this? As Helmar Gust, Kai-Uwe Kuhnberger and Ute Schmid point out in their chapter "Ontologies as a cue for the metaphorical meaning of technical concepts", metaphors have no general form, they can connect quite disparate domains and they are highly context-sensitive. Yet everyone immediately understands Romeo's utterance "Juliet is the sun".

Gust, Kuhnberger and Schmid focus on metaphors in the technical domain drawn from Information Technology - such as mouse, buffer, browser, virus etc. Classical accounts of analogical reasoning posit mappings between the source and target domains grounded on extensions of properties and relations. Gust, Kuhn­berger and Schmid demur - they suggest an analogical relation associates source and target and that this relation explains the meaning of metaphors. They propose a formal theory, Heuristic-Driven Theory Projection, HDTP, formulated in a many-sorted first-order language to model analogical reasoning. HDTP permits the computation of metaphorical meanings from the analogical relations it models using background knowledge of the ontologies of source and target domains to reduce the complexity of the computations.

The hope of cognitive science is to provide a wholly scientific or naturalistic account of the mind. If something as elusive as metaphorical meaning can prove amenable to a computational approach, that goal might seem a little closer. Yet there are certain sorts of mental states that seem highly resistant to any scientific reduction. These are states associated with immediate conscious experience. Thus our experience of pain and other bodily sensations together with our experience of external stimuli such as colours, sounds, tastes, touch are often taken to show that pace the aspirations of cognitive science, the mind is not part of the natural order but is distinct from it. This is because of the difficulty of finding a physical property with which to identify the colour red or the bitter taste of lemons, for example.

A famous argument designed to establish this conclusion goes by the name of the Knowledge Argument. This argument purports to prove that colour experi­ence puts us directly in touch with non-physical sensory qualities or qualia. As philosophical arguments go, it is disarmingly simple and direct: Since a colour-blind superscientist could master all the physical facts about colour experience without knowing how red things characteristically look to normal colour perceiv­ers, this phenomenal feature of colour experience could not be anything physi­cal. Drew Khlentzos argues in his contribution "Antirealist assumptions and chal­lenges in the philosophy of mind" that in spite of its apparent simplicity the argument rests on some dubious verificationist (or "antirealist") assumptions. When these are exposed and questioned the argument is nowhere near as compel­ling as it appears, he contends.

An important question about the human mind is whether our categorisations or conceptual schemes are the only ones possible. Edward Sapir and Benjamin Lee Whorf famously denied this. Far from being inevitable, our categorisations are mere artefacts of our language and culture, they averred. Moreover, they claimed different cultures nurtured very different conceptions of such central notions as time, causation, agency, the self and morality, conceptions which were reflected in their languages. This is the Sapir-Whorf hypothesis of linguistic relativity, a hy­pothesis that has been enormously influential in linguistics, anthropology and so­ciology. In his famous paper "On the very idea of a conceptual scheme" the Amer­ican philosopher Donald Davidson took issue with the idea that there could be radically different conceptual schemes.

Building on a thought-experiment of George Simmers, Arcady Blinov in his contribution "Vagueness, supertranslatability and conceptual schemes" argues that Davidson is mistaken. Blinov imagines beings who do not share our colour experiences but instead have direct phenomenal experience of the same electro­magnetic radiation that underwrites human colour experiences. He argues that their language, "Simmelian", is not translatable into any human language. None­theless it is "supertranslatable", he maintains - we can at least pair sets of Simme­lian sentences, if not the individual Simmelian sentences required for bona fide translation, with English sentences such as This surface is red. Blinov goes on to recommend supertranslation as a method for modelling the pervasive phenome­non of vagueness in natural language.

Dorothea Cogill-Koez provides a different perspective on the universality of cat­egorisation in her chapter "Visual representation in a natural communication sys­tem". The question she investigates is whether human categorisation is constant across signed and spoken languages. She argues against the tendency to assume that, on the one hand, verbal representations of spoken languages reflect mental catego­ries and, on the other hand, that the analogue structure of pictorial representation deployed by signed languages merely captures visual similarities between sign and referent. This is a false dichotomy, she contends, since the signs of signed languages are not mentally processed by means of matching the physical structure of their in­tended referents with any visual image of that referent and since the connection be­tween verbal representations and mental categories is not immediate.

Focusing on signs used to represent visual spatial situations and events, "classifier predicates", Cogill-Koez notes that whilst these signs do convey information about the visual-spatial properties of their referents without any conceptual preprocessing, nonetheless across the various signed languages classifier predicates assume standard discrete forms - standard handshapes and movements are deployed to depict ideal­ised prototypes plausibly associated with distinctive mental categories.

The theme of mental categorisation is further pursued in the final chapter of the volume "Hidden units in child language" by Stephen Crain, Takuya Goro and Utako Minai who present some intriguing linguistic evidence for a nativist expla­nation of childrens' understanding of a certain class of expressions common to all languages. This is the class of logical particles of the language: In English, these are words such as not, and, or, every, only. Crain, Goro and Minai focus on the interac­tions between negation not and disjunction or and between negation and conjunc­tion and. In particular, they wish to know whether or and and in natural language retain the usual Boolean interpretations they receive in classical logic. If they do, then embedding them within the scope of an overt negation operator such as not or a covert one such as only ought to produce some systematic results. The simplest of these go by the name of De Morgan's Laws of which there are two, one for ne­gated disjunctions:

(I) -.(X v Y):: -.X A

and one for negated conjunctions:

(II) -,(X n Y):: iX v -'Y

In these two laws, is the sign for negation; 'v' is the sign for disjunction, 'A' is the sign for conjunction and is the sign for logical equivalence; 'X' and 'Y' are variables ranging over propositions. Crain, Goro and Minai call one direction of the first law the "Conjunctive Entailment" of disjunction in the scope of negation and one direction of the second law the "Disjunctive Entailment" of conjunction in the scope of negation.

Semantic Nativism predicts the existence of a default mental setting for the meanings of the logical particles prior to the acquisition of any specific natural language. To an empiricist Semantic Nativism might sound implausible but for logical expressions the view has considerable appeal - prelinguistic children ap­pear to reason and if so they must be able to perform logical transitions of some sort. Moreover, there are some good evolutionary reasons for Nature to set the constraints governing the interpretation of logical particles not, and and or at their Boolean readings - for the Boolean versions of these operators correspond to sim­ple set-theoretic operations of complementation, intersection and union, respec­tively. Hence, if there were languages with overtly non-Boolean interpretations of or and and for which young children assumed the default Boolean readings in the learning process, Semantic Nativism would have received some impressive em­pirical support.

Grain's team's finding is that Japanese and Chinese children in the process of learning their native languages do indeed appear to adopt a Boolean interpretation of their disjunctive operators ka and huozh, respectively. This is in spite of the fact that both ka and houzh are "positive polarity items" that cannot be embedded within the scope of local negation. Thus, whilst the English sentence the pig did not eat the pepper or the carrot licenses the Conjunctive Entailments (i) the pig did not eat the pepper and (ii) the pig did not eat the carrot, the Japanese equivalent Butasan-wa ninjin ka pi' iman-wo tabe-nakat-ta licenses neither of these entail­ments. For although this sentence literally says 'the pig did not eat the pepper or the carrot' what it actually means is 'the pig did not eat the pepper or the pig did not eat the carrot. That is, the Japanese disjunction ka takes wider scope than ne­gation nakat irrespective of its surface position in the sentence. The unequivocal finding of Grain's team is that both Japanese and Chinese children interpreted ne­gated disjunctions as licensing Conjunctive Entailments even though their adult counterparts did not. Contextual and scope ambiguities conspire to unsettle our confidence in the interpretation of even the simplest sentences containing nested logical operators. Crain, Goro and Minai's chapter contains a subtle and engross­ing discussion of these issues and a persuasive case for Semantic Nativism. In so doing, it advances our understanding of the structure of the human mind.

 

Mental States: Volume 2: Language and cognitive structure by Andrea C. Schalley and Drew Khlentzos (Studies in Language Companion Series: John Benjamins Publishing) Is the way we conceive of the mind an artefact of the culture in which we happen to live? A recurrent question in the study of languages and cultures concerns the extent to which language and culture shape thought. No one doubts that impor­tant conceptual categories are derived from our native tongue but how deep is the impression language makes on thought? Is the way we classify the various phe­nomena we experience language-and-culture relative in some strong sense?

Consider the following thesis: But for the language we speak and the culture in which we live we would not conceive of the world in just the way we do. This is pre­sumably true. However thus stated it is also a very weak claim. For it amounts to little more than the truism above - that many of our conceptual categories are derived from our specific linguistic-cultural context. To be sure, the thesis be­comes more interesting once we discover that certain classifications such as those of colour are not the universals we might have expected them to be. Yet even then the thesis falls well short in logical strength of the famous Sapir-Whorf hypothesis that the language we speak determines what we perceive. That hypothesis implies that the XYXYs do not see that the ripe tomato is red, the weaker hypothesis only entails that they do not remark upon it.

The question is not whether different cultures conceptually classify the world in different ways (they surely do), it is whether the specific ways in which each culture classifies the world are accessible to those from other cultures, whether their languages are genuinely inter-translatable.

A major concern of the authors of this volume is the nature of conceptual clas­sifications of mental states and, in particular, the relation of any such classification to the particular language in which it is framed. Suppose it could be shown that some languages lacked any means for verbally identifying certain mental states that English speakers regarded as fundamental. What should we infer from this?

According to supporters of the Natural Semantic Metalanguage (NSM) pro­gram in linguistics we should infer that these English words are "ethnocentric" and that the mental states they denote are thereby unfit to discharge any serious explanatory role in cognitive science. Thus Cliff Goddard contends in his essay "A culture-neutral metalanguage for mental state concepts":

[Many key terms in the discourse of mainstream Western psychology, cognitive science and philosophy are language-specific, including ... fear, sadness and sur­prise ... belief and the construct of mind itself.

This is a rather dramatic-sounding claim - could there be a culture where no one is ever sad or fearful or surprised, one where no one ever believes anything or, most striking of all, where the idea of mind had gained no traction at all within folk consciousness?

However this is not what is meant by the NSM ethnocentricity thesis. It is not that no one is ever sad or ever believes anything (as eliminative materialists in cognitive science aver), it is just that the English-language concepts of sadness or belief are highly parochial ones that English speakers mistakenly presume to be universal and primitive. Whence, if our goal is to describe basic human psychol­ogy rather than Anglocentric accretions to it, we should not invoke culturally-distilled descriptions of mental states such as "belief" and "sadness" or even "mind" as theoretical primitives suited to explain the workings of human cognition or emotion. Belief and sadness are (culturally) derivative concepts, not primitive ones, according to NSM.

Following Goddard's NSM essay and one by Anna Wierzbicka, the linguist who first propounded NSM, the next three chapters in this volume apply the tech­niques of NSM to the meanings of words for mental states in various languages. Anna Gladkova compares Russian and English propositional attitude verbs, Kyung-Joo Yoon focuses on the meanings of three Korean words kiekna- 'memory comes, remember, kiekha- 'remember, and chwuekha- 'reminisce, Zhengdao Ye discusses the meanings of Chinese `taste'-related words that describe various mental states.

NSM is a bold and ambitious program in semantics. Its claim to have discov­ered sixty or so universal human concepts is, if true, of the utmost significance in understanding human categorisation. The crucial NSM thesis that an acceptable semantic theory must effect a conceptual analysis of words can be challenged. In defending the conceptual analysis requirement, Wierzbicka in her contribution "Shape and colour in language and thought" takes her lead from Leibniz who stressed that there can be no understanding anything without conceptual decom­position into parts "which can be understood in themselves". Formal semantic theories are of no use according to Wierzbicka because their technical concepts are not accessible to ordinary speakers, yet "ordinary English" is of no use either since it contains too many hidden ethnocentric assumptions. The solution is to develop an informal metalanguage built from semantic primes common to all languages.

In her study, she presents a case for Leibniz's thesis that language is the mirror of the mind. Where Leibniz merely dreamt of a future wherein, as he put it, "a precise analysis of the signification of words would tell us more than anything else about the operations of the understanding", through NSM his dream has today become a reality, according to Wierzbicka. For NSM's semantic analyses of differ­ent languages has resulted in the identification of "a rich set of empirical linguistic universals" that represent "a set of sixty or so universal human concepts - concep­tual primes, which correspond to Leibniz's notion of 'the alphabet of human thoughts', and a universal grammar, which corresponds, to some extent, to his no­tion of a universal 'philosophical grammar":

Until quite recently Eleanor Rosch's conjecture that the eleven basic colour categories of English might represent universal and innate human categories seemed empirically plausible. But recent ethnological studies of colour classifica­tion in various languages has cast doubt upon it. Thus neither the Berinmo people who inhabit dense tropical rain forest in Papua New Guinea nor the nomadic Himba who wander the arid land on the border of Angola and Namibia are sensi­tive to the English green/blue boundary. Wierzbicka observes from her study of Burarra-speaking Australian Aboriginals in Arnhem land that it can be extremely difficult to even fit some indigenous colour classifications into the English colour framework. Even more striking is her claim that basic metrical and shape predi­cates such as 'long' and 'round' are not the common linguistic currency of speakers of other languages such as Polish!

We should be careful to distinguish the question of whether people from a given culture have a certain concept from the question of whether the language they speak has a specific term that expresses that concept. Polish speakers clearly can perceptually discriminate round things from non-round things. To the extent that perceptual discrimination requires concepts they have the concept of round­ness. Whether the concept of roundness for Polish speakers is primitive or com­plex is a matter for further investigation. The mere absence of a dedicated term denoting that concept does not decide this issue.

In her contribution "Universal and language-specific aspects of 'propositional attitudes': Russian vs. English", Anna Gladkova argues that English is not the only language to promulgate ethnocentricity under the mantle of cognitive science. Just as for the English word belief the Russian word scitat is touted as denoting a primi­tive mental state where it is nothing of the sort, Gladkova contends. scitat represents a considered opinion, one arrived at through a process of weighing evidence or re­flecting on one's store of information and knowledge. To that extent it has some af­finity with the English belief and is often rendered as 'believe' in translation.

Gladkova compares the NSM approach to linguistic meaning with that of the Moscow School of Semantics (MSS). Both are committed to the existence of semantic primitives. Both seek to explicate the semantics of natural languages by means of a metalanguage containing such primitives or primes. However ;where MSS allows that the metalanguage might be a formal one, it is important to NSM that the metalanguage be a natural language or at least be constructed from those natural language expressions identified as semantic primes.

In her "Mental states reflected in cognitive lexemes related to memory: A case in Korean", Kyung-Joo Yoon promotes NSM as a semantic metalanguage that is ideally poised to avoid ethnocentrism by consisting of words denoting universal conceptual primes. Yoon sees ethnocentrism at work in the way English speakers and scientists conceive of memory, leading them to expect every language to have a word denoting memory. Korean doesn't. Yoon uses the resources of NSM to offer neutral semantic explications of the key Korean terms associated with recollec­tion: kiekna-, kiekha-, and chwuekha-.

Finally, in "Taste as a gateway to Chinese cognition", Zhengdao Ye outlines the central role taste plays in the Chinese conceptual scheme arguing that the Western philosophical tradition that shapes Western ways of thinking sees the sense of taste as inferior to vision and hearing. Indeed, Ye offers the intriguing proposal that the mind/body problem that so bedevils Western philosophy has no counterpart in "the peculiarly Chinese 'embodied' way of thinking, knowing and feeling". Ye's chapter comprises a contribution to the NSM program by providing NSM explica­tions of Chinese words associated with taste.

We have been considering a semantic theory that makes interesting predic­tions about the cognitive structure of competent speakers of a language. We might wonder whether markedly atypical linguistic performance is a reliable indicator of specific cognitive deficits? In the chapter "'Then I'll huff and I'll puff or I'll go on the rail' thinks the wolf: Spontaneous written narratives by a child with autism", Lesley Stirling and Graham Barrington examine the written and oral re-tellings of familiar fairy tales by a high-functioning seven-year old boy with autism, Lincoln.

Stirling and Barrington focus on the aspects of perspective marking and epi­sodic macrostructure in story-telling. While approximately half of all autistic chil­dren never acquire functional language, "higher-functioning" individuals are those of normal intelligence who possess functional language. Story-telling in­volves the understanding and occupation of a variety of differing perspectives -those of the narrator, the audience and the characters within the story. It is thus ideally suited to test some of the competing theories about the nature of autism.

Stirling and Barrington consider three major theories of autism: (1) mind-blindness, (2) executive control, and (3) weak central coherence. According to the "mind-blindness" account, autistic individuals lack a theory of the minds of others. They therefore have severe difficulties comprehending another person's perspective. Executive control theory conjectures that autistic children are deficient in skills of planning and inhibition of immediate desires subserving long­term goals. These executive control functions are somehow impaired. The central coherence theory posits an inability to identify the local detail that captures the gist of a higher-level meaning as the key cognitive anomaly of the autistic mind.

Stories are hierarchically arranged into episodes. Stirling and Barrington re­port that apart from being shorter and containing fewer episodes than usual, Lin­coln's re-tellings revealed he had a very clear idea of what it was to tell a story and a keen grasp of episodic macrostructure. He represents the characters' mental states using predicates such as think and worry, even imputing to characters men­tal states not explicitly ascribed in the original story. Lincoln is clearly not "mind-blind" So does this exercise reveal any cognitive deficits in Lincoln?

Stirling and Barrington identify Lincoln's difficulty with knowledge state man­agement as the most serious one. Lincoln has the mother pig tell the three little pigs of the imminent threat of the wolf. He has the wolf divulge to the little pigs his plan to climb onto the roof. The high-functioning autistic individual's problem is not that of comprehending the very idea of mental states other than one's own but instead appears to be a difficulty with grasping how nested mental states are to be represented and how they interact, Stirling and Barrington aver.

Heather Winskel also examines narratives in an effort to understand the rela­tion between language and cognition in normal language development in her chapter "Interaction between language and cognition in language development". Focusing on temporal vocabulary, Winskel investigates Thai childrens' narratives. Early use of temporal connectives to express temporal relations declines in adult speech with relative clauses and causal connectives supplanting the connectives. Aspectual marking, optional in Thai, is easily accessible to the child and as a result invariably appears in the early language-user's temporal vocabulary.

Winskel's conclusion is that language clearly influences the child's cognitive development during the early phase of acquiring a language in order to express thoughts (thinking-for-speaking). In this phase "it can be seen that early caretak­er-child interactions, and the input that the child is exposed to, is critical in the formation of language-specific semantic categories", Winskel maintains.

Herbert L. Colston's contribution "What figurative language development re­veals about the mind" is the last chapter in this volume addressing the develop­ment of mental states and abilities. Colston empirically studies the naturalistic production of figurative language in children and adults - specifically of hyper­bole, which often functions as a means of complaining. His findings suggest that children develop the ability to produce hyperbole much earlier than generally thought so far, and that their production of hyperbole is remarkably similar to the hyperbole production of adults. This poses a challenge to the "late development" view, according to which children do not show adult competency at comprehending or producing figurative language until they are roughly 7 or 8 years old. However, the study presented here provides support for the claim that at least simple figura­tive language cognition is in place early.

The author provides several alternative possible explanations for why the pro­duction of the figurative form of hyperbole appears at or prior to comprehension of other figurative forms, but especially argues that a Theory of Mind - in contrast to claims by the proponents of the late development view - strongly supports that children are fully capable of using figurative language at near adult capacities much earlier than generally thought. He presents arguments that Theory of Mind is in­deed a requirement of all language functioning, based on the view of "Language as Sensation/Perception of Others' Minds" (LASPOM): In order to communicate successfully, speakers have to be able to mutually "sense" or "perceive" each other's mind, and this reflexive property is in Colston's view Theory of Mind and makes language cognition possible. The fact that very young children communicate - and that they, for instance, communicate with people but do not try to engage inani­mate objects in conversation - is a strong indication that they do have knowledge that other people have minds. Therefore, the conclusion is, they have to have Theory of Mind from very early on, and it is not surprising that they also produce hyperbole as an instance of figurative language much earlier than generally thought. Theory of Mind and hence mental capabilities are presented as prerequi­sites for language development.

This contribution thus highlights and explicitly argues for the relevance of the existence of particular cognitive states for a person to achieve linguistic compe­tence. Cognition is seen as a necessity for language development. This might ap­pear as contradicting those contributions that see language as heavily influencing categorisation and thought; however, we have to distinguish language develop­ment as an instance of ontogenesis from the development of a language's lexicon, which corresponds to phylogenetic development.

Leaving the question of narrative and figurative language production, the vol­ume turns to the theoretical question of how children learn to assign words to syntactic or grammatical categories. It is argued that children need to learn this ability to categorise words in order to master the grammatical rules that govern word combinations, which in turn is imperative for any full-fledged language ac­quisition. Joanne Arciuli and Linda Cupples' contribution "Would you rather 'em­bert a cudsert' or 'cudsert an embert'? How spelling patterns at the beginning of English disyllables can cue grammatical category" addresses this issue of how speakers categorise words into grammatical categories and thence how their mind processes linguistic structure and meta-linguistic information.

In contrast to the natural hypothesis that the representation and processing of grammatical category information is syntactic, previous research has shown that the classification surfaces at different levels of representation. Arciuli and Cupples give an overview of which cues have been confirmed as facilitating this process of categorisation so far: mainly semantic and phonological ones. They pose the obvi­ous question whether another cue might be found in orthography. Having studied the relevance of word-endings, they embark on the empirical investigation of word-beginnings in English disyllables. The study comprises two stages: a diction­ary analysis and an experiment involving native speakers. The dictionary analysis was carried out to confirm that word-beginnings can be associated in a statisti­cally significant way to grammatical categories. Having successfully demonstrated this association, in order to address the motivational question outlined above, the authors then set out to test whether skilled native English readers are sensitive to this correlation. Their experiments - including tasks with non-words - show that speakers are indeed sensitive to this correlation and thus that there appear to be orthographical cues for the assignment of grammatical categories, which in turn might cue lexical stress. The results lend support to multi-layered approaches to language and cognition and show that cues on different levels of language repre­sentation exist for the relationship between language and cognitive structure, that is, for the present study, between orthography as a representational system for language and mental categories linked to meta-linguistic information.

The next contribution, Brett Baker's "Ethnobiological classification and the environment in Northern Australia", focuses on the cognitive structures underly­ing ethnobiological classification systems of two Australian Aboriginal languages, Ngalakgan and Wubuy (both spoken in Arnhem Land in the Northern Territory). These structures - the folk taxonomic systems - are extracted via an analysis of names and lexemes for flora and fauna. Baker highlights two characteristics as typical of Indigenous Australian folk taxonomies: (i) the majority of botanical names refer to species, with genera remaining unnamed, and (ii) most names for biological taxa are "monomial" or monomorphemic. In addition, the linguistic structure of the discussed languages does not allow the creation of compound names for biological taxa.

Both (i) and (ii) are counter to the hitherto widely accepted claims about ethno­biological classifications systems by Brent Berlin. In particular, the importance of species names and unimportance of genera in the presented systems conflicts with Berlin's position that generics are a universal category (and correspond to Rosch's basic level categories), and that they form the starting points of taxonomic classifications. Baker argues that this non-conformance with Berlin's system is rooted in the unique structure of the Australian flora, where species belong to large, internally diversified genera, i.e. because the biological discontinuities are perceived at the level of species rather than - as Berlin's universals predict - at the level of genera.

This indicates that we are dealing with an example where human perception of the biological environment heavily impacts on cognitive structuring - on the folk taxonomy or more broadly folk ontology - and that these structures show up in linguistic coding in a way which makes the coding structurally different from the coding of other classificatory systems that arose in other environments - due to the differences in mental categorisation based on perceptual experience. It might be worth noting that the Australian environment has also affected the taxonomy of Eucalypts as coded by Australian English, as Baker elaborates, which is a further argument in favour of his claims. Last but not least, another feature of the Aborig­inal classification systems should not remain unmentioned: The superordinate terms that do exist are almost always related to function or use rather than (biological) taxonomic classification, and therefore functional classificatory terms in­termingle with taxonomic terms.

Perception and common function also play a central role in Ruth Singer's con­tribution "Events masquerading as entities: Pseudorelative perception verb com­plements in Mawng (Australian) and Romance languages". Singer in particular discusses one verb complement construction of perception verbs - the pseudo-relative - and uncovers an unexpected convergence in the expression of perception in the Australian Aboriginal language Mawng and some Romance languages. This is rather unusual given that Mawng and the Romance languages are not re­lated in any way - why, then, does such an unusual structure as the pseudorelative occur in unrelated languages?

The answer, she asserts, can be found in the common function of the construc­tion in discourse - which, she argues, has led to a convergence in form and func­tion in the different languages. More specifically, the pseudorelative's function is to introduce a new referent into the discourse while at the same time making a pred­ication of it, which in principle violates Lambrecht's "Principle of the separation of reference and relation". This principle posits that there is a cognitive constraint on introducing new referents and simultaneously making new predications of them. Universal cognitive tendencies or constraints and their deliberate circumvention can thus lead to the development of rather unusual grammatical structures in un­related languages. In addition, the mere fact that this rare construction occurs in unrelated languages is itself an indication of how cognitive structure impacts on linguistic coding and language.

In their essay "Word and construction as units of categorization: The case of change predicates in Estonian", Renate Pajusalu and Ilona Tragel describe the relation between some common Estonian words describing change and the concep­tual categories they represent. The four words on which they focus, jääma, saama, minema and tulema, are all highly polysemous. Choice of each verb depends upon the type of change the speaker has in mind. Moreover, a specific meaning of each predicate is tightly linked to the particular construction in which it occurs, and whilst it is possible to give "change-of-state" meanings for all four verbs, these are not the primary meanings for them, Pajusalu and Tragel observe. The change-of­state meanings for jaama, saama, minema and tulema are, respectively, 'remain, `get' 'go, and 'come.

Pajusalu and Tragel note the complexity of the principles governing selection of the appropriate change verb - which of the four verbs to use depends on wheth­er the change in question is conceived of as a negative or positive one quantita­tively as well as whether it is positive or negative from the evaluative perspective of the one experiencing the change (where this is relevant). If the change is a positive one, a further distinction is drawn between active and passive change.

In "Categories and concepts in phonology: Theory and practice", Helen Fraser argues against what she regards as a naive realism that equates the meaning of a word with the thing to which it refers. As against naive realism, words refer not to things but concepts, she contends. This being so, cognitive science requires a the­ory of concepts and the theory she favours is a phenomenological one that brack­ets the external reality behind concepts (what the concepts are concepts of) and studies concepts directly.

Concepts are abstract entities as are words and it is very easy to confuse the two especially when the object of study is words. Fraser sees this bimodal confu­sion of words with concepts and concepts with objects in the world vitiating phon­ology both in theory and also in practice. Literacy merely requires the apprecia­tion that letters represent sounds. However naive realism (the "Natural Attitude") impels ordinary speakers to think that the sounds in question are physical acoustic phenomena. In fact they represent abstractions from physical sounds - that is con­cepts of sounds - Fraser contends.

Let us conclude this introduction as we began - with the question of concep­tual relativity. Roger Wales in his essay "You can run, but: Another look at linguis­tic relativity" looks at the psychological processing involved in linguistic comprehension by comparing descriptions of verbs of motion made by English and Spanish speakers. He argues that linguistic relativity is primarily concerned with conceptual processes rather than the products of those processes in mental repre­sentations together with the words expressing them that linguists have studied. To study processing in using verbs of motion requires taking into account a speaker's perception, memory and task construal, according to Wales.

Path-dominant verbs of motion such as enter and ascend are manner-inde­pendent. Such verbs predominate in Spanish and other Romance languages. Eng­lish contains many motion verbs where manner is a determinant of application together with path which is a separate determinant - verbs such as ran to. By ask­ing people to act out verbs of motion, recording their movements with point-light displays, participating observers were asked to make forced choice decisions as to the verb best describing the acted movement.

The target set of English verbs included hop, walk, stroll, shuffle, shamble, run, jog and sprint. 'Walk' and 'run' sets of verbs are distinguished by two factors: rela­tive speed of movement and whether the agent's feet are typically on the ground at the same time or not. Participants' judgments clustered around two sets - the `walk' verbs and the 'run' verbs in the case of English-speaking participants.

Wales' finding is that this result is robust across the two languages of English and Spanish in spite of the inclusion of path information and suppression of man­ner information in the Spanish verbs of motion. He infers that English and Span­ish perceivers of action see the relevant actions in precisely the same way. Although this is exactly the result we would expect from a biological perspective it does not sit well with the more extreme versions of linguistic relativity promulgated by Sapir and Whorf.

 

Headline 3

insert content here

WT Main | About WT | Review Links | Contact | Review Sources | Search