This book "supplies a vocabulary of English words and idiomatic phrases 'arranged … according to the ideas which they express'. The thesaurus, continually expanded and updated, has always remained in print, but this reissued first edition shows the impressive breadth of Roget's own knowledge and interests."
Carstairs-McCarthy, Andrew (2000) The Origins of Complex Language: An Inquiry into the Evolutionary Beginnings of Sentences, Syllables, and Truth. New York: Oxford University Press, paperback, 260 pp.
Asunci�n �lvarez, Universidad Complutense de Madrid.
It comes as quite a surprise that Andrew Carstairs-McCarthy, the author of the highly specialized, exhaustive Current Morphology, should tackle a topic as apparently unrelated to morphology as evolution. However, it was precisely his former work that led him to evolutionary linguistics. As he puts it, "Meaning is the senior partner [in the sound-meaning pair], we tend to feel, in that spoken words exist in order to express meanings. Yet my work on inflectional morphology led me to wonder whether, in some real sense, things may be the other way round: meanings exist in order to provide something for spoken words to express". This provocative idea is the motor of the book, which is an attempt to explain it in evolutionary terms and to explore its philosophical implications (Carstairs-McCarthy trained as a philosopher and Classicist before turning to Linguistics stricto sensu).
Chapter 1: Introduction Chapter 1 consists of a brief introduction to the book's main themes. Foremost among these is the universality of the syntactic distinction between sentences and noun phrases (NPs). Former inquiries into the nature of language have consistently taken this distinction as a necessary given. However, Carstairs-McCarthy argues, this is due to their being tainted by the fact that they were necessarily carried out by speakers of human languages. Carstairs-McCarthy holds that human language as we know it was not the only possible outcome of evolution, and that seeing the sentence-NP opposition as a necessary foundation for the possibility of language, far from clarifying the question, is merely a display of anthropocentric teleology.
Chapter 2: The Peculiarities of Language In Chapter 2 Carstairs-McCarthy expounds what he takes to be the three main peculiarities of human language: vocabulary size, duality of patterning, and the distinction between sentences and NPs. Carstairs-McCarthy sees the vocabulary size in languages as an oddity. From the viewpoint of adaptive specialization, he argues, the human being, with his capacity to learn many different specialized vocabularies, would resemble "a mammal with the aquatic habits of a seal but whose forelimbs were equipped not only for swimming but also for terrestrial food gathering [...] [and which] also had an elongated digit subtending a web of membrane attached to its flank, like a bat, equipping it for a mode of locomotion that it had never used". Natural selection encourages specialization - then "why", wonders Carstairs-McCarthy, "has not natural selection also encouraged a form of language in humans that is specialized [...] - that is, in which the capacity to learn and remember words is just sufficient to accommodate the social and economic needs of small bands of individuals living by hunting and gathering, without scope for technical and cultural elaboration or versatility?" Duality of patterning (a.k.a. as double articulation) is the characteristic of language whereby linguistic expressions are analysable on two levels: a meaningless level, in which mere arbitrary sounds (phones) are arranged into abstract units made up of bundles of distinctive features (phonemes); and a level in which phonemes are arranged into larger, meaningful units (morphemes). Some animal systems of communication can be analysed as involving combinable units, but none of them involve a dual structure of meaningfulness-meaninglessness. Carstairs-McCarthy argues that duality of patterning is not necessary for vocabulary expansion, and indeed that there exist linguistic forms which work in a non-dual way, such as sign languages of the deaf and dumb and some writing systems. I find this is a misrepresentation. Sign languages and the Chinese writing language, asserts Carstairs-McCarthy, lack minimal distinctive units which can be rearranged into larger units, such as phonemes. Thus Carstairs-McCarthy defines the double articulation as a dual structure in which both levels are linked by a causal relationship - the 'higher', meaningful level being componentially made up of units belonging to the 'lower', meaningless level. However, duality of patterning need not involve a causal, componential relationship between the two levels. The 'higher', meaningful level does indeed have its material support in the 'lower' level of physical sounds (or ink stains, or whatever), but the laws for the self-regulation of the levels need not be the same for both of them. Besides the interface between meaninglessness and meaning, each level has its own internal laws, which may or may not resemble those of the other one. The point of the notion of duality of patterning, I believe, is precisely the distinctiveness of the levels. Saying that meaning is determined exclusively by the combination of the meaningless units which underlie it amounts to establishing a continuum between both levels, thus blurring the radical difference between them. Finally, Carstairs-McCarthy states that the distinction between sentences and noun phrases, however essential to language we believe it to be, is in fact quite arbitrary. He then proceeds, rather amusingly, to illustrate this point with several invented languages lacking altogether this distinction: 'Asyntatic' (which has no syntax as such), 'Spatiotemporal' (characterized by the compulsory specification of spatial and temporal coordinates),'Monocategoric' (lacking different syntactic categories - there are only simple and complex expressions), and 'Nominalized English' (which has no verbs and uncannily resembles Officialese). Carstairs-McCarthy ascertains that such systems work correctly, in terms of communication. Thus he concludes that there is no necessary reason for the privileged status of the sentence-NP distinction - indeed, he argues, there are many similar linguistic distinctions which a grammar might express but generally neglects, such as those between token and type, or between activity and achievement.
Chapter 3: Truth and Reference Chapter 3 deals with the philosophical implications of questioning the sentence-NP distinction. Indeed, this distinction practically amounts to that between the classic philosophical notions of predication - the function by which statements are made and deemed to be true or false - and reference - the function by which 'things' (entities, both abstract and concrete, in the world or in possible worlds) are named. Pursuing his inquiry into the nature of the sentence-NP distinction, Carstairs-McCarthy seeks an extralinguistic foundation for it. To that end, he examines the views of three main philosophers on this question: Gottlob Frege, Ludwig Wittgenstein, and Peter Strawson. The conclusions he draws from this examination confirm him in his assertion that there are no extralinguistic reasons for the sentence-NP (or truth-reference) divide.
Chapter 4: Attempts to solve the problems This chapter is an overview of the main explanations that have been proposed for the three aspects of human language the book deals with. Insofar as vocabulary size is concerned, Carstairs-McCarthy has little to oppose to, as few theories dealing with this feature of human language have been advanced. Thus he merely points out the paucity of synonyms and the abundance of homonyms which seem to prevail in language. As regards duality of patterning, Carstairs-McCarthy discusses the theories of call blending, sound symbolism, and phonological self-organization. Finally, and concerning the sentence-NP distinction, he examines Chomsky's universal grammar, Pinker and Bloom's study of auxiliaries, cognitive grammar, pragmatics, Bickerton's ideas about protolanguage, Sampson's hierarchical syntax, gesture and sign language. Carstairs-McCarthy finds none of the explanations he examines satisfactory. He therefore presents his own account of the three features in question, which is advanced in the next chapter.
Chapter 5: A Different Solution Chapter 5 presents Carstairs-McCarthy's own theory concerning vocabulary size, duality of patterning, and the sentence-NP distinction in human language. This theory is based on evidence from two quite disparate areas of language research: vocabulary acquisition in children and the anatomy of the vocal tract. Basing himself on arguments drawn from these two areas (and which are too elaborate to summarize here), Carstairs-McCarthy arrives at the conclusion that syntax is motivated by phonology, which in turn is conditioned by the descent of the larynx in our hominid ancestors, caused by bipedalism. This physiological modification would have led to what Carstairs-McCarthy charmingly terms later in the book (chapter 7) "the binary strait-jacket". The bipolar continuum of sonority made possible by the descent of the larynx would have produced a syntactic mould which in turn would have generated an analogous bipolarity in semantics (thing vs. event, static vs. dynamic, noun vs. verb, NP vs. sentence).
Chapter 6: Apes, Anthropology, and the Brain In this chapter Carstairs-McCarthy attempts to buttress the scenario presented in chapter 5 with extralinguistic evidence drawn from biological anthropology, brain neurophysiology, and studies of the linguistic abilities of the great apes. The research he draws on includes Stringer and McKies' ideas on human evolution, Calvin and Ojemann's work on neurobiology, Deacon's inquiry into language and brain evolution, and Lewis and Savage-Rumbaugh's study of the omnipresent chimp Kanzi.
Chapter 7: Just How Unique Are We? This chapter is a sum-up of the book. It states its conclusions quite simply and concisely: "Human language turns out (if I am right) to be a by-product of a change in the anatomy of the vocal tract brought about by our ancestors' taking to the ground and walking on two legs". It also serves as a disclaimer against possible attacks, denying any pretence to explanatory exclusiveness - his argument, Carstairs-McCarthy pleads, is quite compatible with cognitive claims: "The issue becomes not how cognitive developments caused our large vocabularies or our sentence-NP syntax, but how they interacted with them." However, this chapter does imply an attack on propositional thought, which Carstairs-McCarthy considers the consequence of a "parochial linguistic phenomenon, scarcely more relevant to philosophy than (say) the distinction between strong and weak noun declensions in German".
The Origins of Complex Language is a well-written, eminently readable book, despite the occasional convolution of the author's argument. There are however several theoretical points which I find fuzzy. One is Carstairs-McCarthy's notion of double articulation, which was mentioned above. Another one is the universality of the sentence-NP distinction: I wonder whether the notion of configurationality as defined by Hale (1983) does not suppose a challenge to this assumption. Also, I find that the lowering of the larynx does not explain in itself the appearance of the dual structure of language. While providing its physical support, it does not account for the twofold leap to abstraction (the coding of a coding, one would say) which the double articulation supposes. And finally, the book's title: I fail to grasp what The Origins of COMPLEX language means. Does the author imply that a simple (as opposed to complex) language is possible? As for the book's merits, I believe that the foremost one is its pointing out the peculiarity of the double articulation and the sentence-NP distinction (vocabulary size seems to me to be a derived feature). Although I find Carstairs-McCarthy's account ultimately unconvincing, I certainly agree with him that they are key concepts which must be clarified if we are to achieve a greater understanding of the nature of human language.
Hale, K. (1983): "Warlpiri and the grammar of nonconfigurational languages" in Language and Linguistic Theory,1, pgs. 5-47.
Asunci�n �lvarez is a Linguistics graduate. Her research interests include philosophy of language and mind, writing systems, and the relationship between linguistic and psychoanalytic thought.