Botanical guide to the production of speech

A language system is more like a plant than an animal," claims William Croft, professor of linguistics at the University of Manchester, in this novel approach to language change. The book outlines a framework for understanding change as a fundamentally evolutionary phenomenon. In the course of it, Croft weaves together multiple areas of linguistics: pragmatics, discourse analysis, semantics, language acquisition and historical linguistics. These he entwines with evolutionary biology and the philosophy of science.

Croft's starting point is the observation that "languages don't change, people change language through their actions". The study of language is therefore about "empirically real entities", not idealised abstract systems. He rightly points out the need to distinguish innovation as a synchronic phenomenon from propagation as a diachronic one. In language change, multiple options spring up, of which only a few get adopted and propagated, a scenario now widely accepted.

Croft espouses the "invisible hand" theory of change, which was introduced into linguistics by Rudi Keller (1990) but dates back to the 18th-century economist and philosopher Adam Smith who, speaking of wealth, suggested that every individual intends only his own gain, but is "led by an invisible hand to promote an end which was no part of his intention" (1776). Keller labelled language change "a phenomenon of the third kind". This is the creation of a novel variant by an individual speaker that then spreads, in contrast to phenomena of the first kind (purely natural phenomena), and phenomena of the second kind (those that are artifactual and intentional).

Language as an evolutionary phenomenon is a hot topic, and, as often happens with key areas of contemporary interest, numerous researchers have similar insights, all of them using slightly different terminology. Social imitation has been widely talked about in recent years. Animals copy behaviours to which they are naturally predisposed, sometimes known as "innately guided learning", and they hand this on to future generations. Birds can be trained to utter alarm calls to supposed new predators, for example, and fear of specific birds is then passed on to their offspring.

Richard Dawkins labelled these units of imitation "memes", parallel to "genes". Croft re-labels language imitation units "linguemes". He himself is particularly attracted to the work of the biologist-evolutionary philosopher David Hull. Hull redefines concepts as historical individuals, and treats them as the basic components of socio-cultural and scientific change. He argues that the concept, as a spatio-temporally bounded individual, is the "replicator", the social equivalent of the gene.

But why does Croft claim language is like a plant, an idea that was prevalent in the last century? This is explained in the last chapter. Language is like a plant because of the high receptivity of language to change by contact with other languages, and so, like a plant, it is prone to hybridisation (mixing). Croft argues, "any lingueme can be borrowed without destroying the communicative power of language". In his view, "a language is a loosely coordinated set of linguemes that is relatively simple compared with a truly finely balanced and complex system". Croft overlooks the fact that language is like both an animal and a plant: he fails to mention the important role of genes, implicated in the adaptions to language of the human lungs, vocal organs and brain, which must have taken millennia to evolve, and which interleave with the faster-moving linguemes with which he is primarily concerned.

Overall, this worthwhile book is one of signposts, rather than a definitive account of language change. In my view, most of the signposts are pointing in the right direction.

Jean Aitchison is professor of language and communication, University of Oxford.