Theories out of the mouths of babes

Margaret Deuchar and Suzanne Quay warn that theirs is not a work for a wide audience: "The main aim of this book is to explore the implications for linguistic theory of a case study of bilingual acquisition." The subject of their study is Deuchar's daughter, referred to as "M", who was brought up in a bilingual environment, exposed to Spanish at home and to English outside. Although Deuchar's records of her daughter's progress span seven years, this study concentrates on just one year, from the child's acquisition of her first word.

This is a slim book, more than a third of which is occupied by an index, a bibliography and exhaustive appendices, including a chronological listing of the child's lexicon, indexed and cross-referenced, so that, for instance, we know that her first word, "book", was acquired at age ten months and seven days, pronounced bu , and that its Spanish equivalent, libro , was not acquired until 21 months and five days of age, when she pronounced it liwo . Is this work, then, for an inner sanctum of narrow specialists?

Surprisingly, no. Although obviously written for a restricted audience, it is accessible to the general reader and a delight to the curious. The methodological approach, free of jargon and abstruse theorising, builds a relationship of intellectual trust with the reader.

One illustration of the trust-building approach is how the authors tackle the question of word identification. Over the period covered, the overwhelming majority of M's utterances consisted of single words, most of them again reduced to a single, open syllable, so that for instance pa might stand for daddy, shoe, panda bear, bird, duck, etc ( papá, zapato, panda, pájaro, pato ). Sceptical readers will entertain serious doubts about the reliability of such interpretations. Deuchar and Quay are aware of the problem and they list six criteria for word identification, but instead of bringing theoretical justifications for these criteria, they demonstrate their application through a few practical examples drawn from their experience. Readers are convinced - the proof of the pudding is in the eating - and they emerge from the demonstration feeling as if they had been invited to participate in the study. An example of the authors' thought-provoking approach is the wording of one of their conclusions: "We find that the voicing distinction can be acquired on the basis of acoustic differences rather than phonological features, as is generally assumed" (my emphasis).

By the end of the section on the acquisition of voicing, even those with no linguistic background will understand that those "acoustic differences" are physical phenomena, observable and measurable, whereas the "phonological features" are theoretical constructs; they may then wonder about the careful wording, and how theoretical constructs ever came to be considered the basis on which infants learn to distinguish voiced sounds ( b, d, g, z ) from voiceless sounds ( p, t, k, s ).

Our understanding of first-language acquisition remains inadequate, and Deuchar and Quay demonstrate this, again very simply. In their introduction, they present a survey of the literature on the subject over the past two decades. They note only one book-length study of a child's acquisition of Spanish and English, under markedly different circumstances: this child was not exposed to English until he was two, and his first English word was acquired at 30 months, whereas Deuchar's daughter learnt both languages contemporaneously.

That is as if, of the 6,000 or so languages of the world, we knew of barely a few hundred, some from grammatical sketches, others from disparate word lists, the rest from fleeting mentions in travellers' diaries - hardly a basis on which to build theories of language. Consider a child's acquisition of the phonological system. It is generally assumed that labials are acquired first. The first consonants uttered by M were  p and b . But what of those languages with two contrasting series of labials, as are found in Malekula (Vanuatu) where bilabials contrast with interlabials (uttered with the tongue held between the lips)? We do not know.

Nor do we know how Aboriginal children acquire the five-way contrast common in many Australian languages: interdental - dental - alveolar - palatal - retroflexed. Our understanding of children's acquisition of morphology and syntax is likewise restricted to the most unrepresentative sample, which is our familiar languages. But take Nasioi (Papua New Guinea), in which verbs are inflected by prefixation, and the reduplication or non-reduplication of a syllable in the string of verbal prefixes is conditioned by the number of syllables in the resulting, inflected verb, depending on whether it is odd or even. How do Nasioi children acquire this puzzling rule?

The scope of Deuchar and Quay's study is, of course, too narrow to advance our understanding of how children acquire language. The interest lies not so much in the authors' observations as in their methodology.

Jacques B. M. Guy, who is bilingual in French and English, received a PhD for his thesis on Sakao, a language of Vanuatu, from the Australian National University.