How our ancestors learnt to say: 'There's a bear outside'

Evolution was a dirty word in linguistics for a long time. However, for about a decade and a half evolutionary linguistics have been sexy again. Having its finger on the pulse, Oxford University Press started its "Studies in the Evolution of Language" series seven years ago. Bernd Heine and Tania Kuteva's The Genesis of Grammar: A Reconstruction is the ninth book to appear in this series. It is also the best.

As the title suggests, Heine and Kuteva's focus is not on the emergence of language per se but on how it acquired grammar. Their approach is based on their considerable knowledge of "grammaticalisation", that is the development of grammatical structures out of lexical material (such as the rise of the preposition "beside" out of the expression "by the side of").

Grammaticalisation theory has thus far been used only to explain changes in fully fledged languages, called "modern languages" by the authors. But when our ancestors started to talk, it is extremely unlikely that their language (which the authors call "early language") had anything near a fully fledged grammar. Instead, this must have evolved gradually.

The authors use our understanding of grammaticalisation in modern languages to reconstruct the gradual development of grammar in early language.

For example, since prepositions often arise out of nouns (recall that "side" in "by the side of" changed into "beside") we may assume that nouns emerged in early language before prepositions. In fact, Heine and Kuteva suggest that nouns were the very first "layer" of grammar. Verbs came later, and more grammatical elements like prepositions later still.

The final layers to be added included even more grammatical items such as relative clause, case and agreement markers. Thus, our ancestors moved from grunts lacking any specific meaning to single-noun utterances such as "Bear!", to expressions involving verbs and other, more grammatical elements, such as "There's a bear beside your house!"

Interdisciplinarity is very much in vogue these days. Because of the absence of direct linguistic evidence related to early language, many scholars take an interdisciplinary approach. There are some good examples around, but much of the work is more "inter" than "disciplinary": researchers often lack a solid or broad enough knowledge of one or all of the disciplines they attempt to draw on. Heine and Kuteva's book is different, in that it is based almost exclusively on the linguistic theory of grammaticalisation. Ironically, this unfashionably narrow perspective comes as a breath of fresh air, and it shows that there is a place for good old-fashioned disciplinarity even in areas where one might least expect it.

Who is it for? Anyone who wants to know how apes came to produce complex phrases like "anyone who wants to know".

Presentation - Solid yet readable argumentation; however, the author index and references section should have been proofread more carefully.

Would you recommend it? A classic in the making.