W. Tecumseh Fitch

THE EVOLUTION OF LANGUAGE

The first half is described as introductory and covers evolutionary theory, linguistics, animal cognition, animal communication, and human evolutionary history. The second half looks at current hypotheses about language evolution and assesses their strengths and weaknesses. The intended audience seems to be academics, but there is plenty to interest the non-specialist reader as well. There are useful summaries at the ends of many of the chapters, which allow readers who are not interested in some of the finer details to indulge in judicious skipping without losing the thread.

The first chapter looks at language from a biological perspective and summarises present views of evolution, including some of the controversies. Fitch then "switches gear" and goes on, in Chapter 2, to consider language in its own terms—that is, writing as a linguist. Some of this is fairly technical and includes some discussion of formal language theory, including syntax, semantics, pragmatics and so on. A glossary is provided for those to whom this material is unfamiliar.

We then go on to look at animal cognition and communication. Fitch comes down firmly on the side of those who regard animals as capable of thinking; language is not needed to make thought possible. But animals have quite limited ability to communicate their thoughts to others, and humans are unique in being able to do this. Anything that can be thought can be spoken about.

A major shift in human evolution occurred with the evolution of Homo erectus. By this time meat eating, tool use, pair bonding and male parental care were probably in place, but it is difficult to find evidence for the existence of language. Although there have been many attempts to relate speaking ability to fossil anatomy, Fitch is sceptical about their validity.

Much attention is often directed to the vocal tract. The low position of the human larynx ("descended larynx") after infancy is thought to be an adaptation for language, so hominids who supposedly lacked this feature, including possibly even the Neanderthals, are said also to lack the ability to speak much if at all.

Fitch doesn't think it is possible to reconstruct the anatomy of the vocal tract from fossils with any degree of accuracy. In any case, he cites a lot of evidence to show that the position of the larynx is irrelevant anyway. In recent animal studies it has been found that the larynx descends considerably in dogs when they bark, and in others (deer, big cats) the larynx is permanently descended. Even in chimpanzees there is some descent as they mature. Moreover, the big cats have a "free-floating" hyoid just as we do. So far as anatomy is concerned, therefore, there is no reason why a lion shouldn't be able to speak. So why do some animals have a descended larynx if it's not for speech? Probably, Fitch thinks, to make the possessor sound bigger and more imposing.

About the only anatomical feature of modern humans that Fitch does accept as probably significant is the expansion of the vertebral canal in the thoracic region, which implies an improved ability to regulate breathing—important for speaking. Unfortunately the vertebrae seldom fossilise, but probably thoracic expansion occurred some time in the million years following Homo ergaster but before the Neanderthals, which would imply that speech was possible for later Homo erectus.

The reason dogs and lions don't speak, Fitch believes, is not any deficiency in their vocal tract anatomy but because they lack the necessary neural wiring. This was pointed out by Darwin, who Fitch thinks has been unjustifiably ignored by later language theorists who have reproduced his ideas, usually without acknowledgement. Humans, unlike most other mammals including primates, have direct connections from the lateral neocortex to the motor neurons controlling phonation, which seem to be necessary for fine control of the vocal apparatus.

The ability to imitate sounds is important for speech. Flying and swimming creatures are the main exponents of this (birds, seals, whales); we seem to be the exception among land-dwelling mammals. The FOXP2 gene plays some part in speech in humans and is also involved in the vocal learning of birdsong. The human version of this gene has been introduced experimentally in mice; unfortunately, they didn't start speaking but there were subtle differences in ultrasonic vocalisations and exploratory behaviour.

Speech and language are different things, and the later chapters consider the evolution of language in detail. For Fitch, a critical phase in the process was what he calls "protolanguage". There have been many suggestions for the form this might have taken. Intuitively, one could suppose that words came first, though initially with no complex syntax to link them together. It is quite difficult to account for the development of syntax in this case. Another suggestion is that language began as "gesture". Sign languages can be fully as complex as spoken languages, and gestures persist today alongside speech. But it is not clear why a gesturing language would be replaced by a spoken one. Various hypotheses based on these ideas are discussed at length.

A third possibility, which seems to be the one that Fitch favours, is a musical protolanguage. This idea, again, was put forward by Darwin and has been developed in different ways by later scholars. Music does not make propositional statements in the way that language does but Fitch provides an evolutionary scheme to show how complex languages, permitting analytical thought, could develop from singing. The hypothesis of a musical origin for language generates testable predictions, he tells us.

This book covers a very large amount of territory and would be important reading for anyone interested in contemporary ideas about how language arose. There were a couple of topics I hoped to find discussed but didn't. Although Deacon's views are cited quite extensively there is no mention of hisfascinating and illuminating suggestion of a co-evolution of language and the human brain—language as a brain "parasite". And there is no discussion of metaphor, which some linguists find to be central to the development of language. As Guy Deutscher puts it in The Unfolding of Language, all language is metaphorical, being made up of "a reef of dead metaphors". But even without these items there is plenty to be going on with here.

See also Terrence Deacon, The Symbolic Species.

24 May 2010