Harvard Educational Review
  1. Winter 2000 Issue »

    Editor's Review: How Language Comes to Children: From Birth to Two Years by Benedicte de Boysson-Bardies and How Children Learn the Meanings of Words by by Paul Bloom

    Leslie Nabors Olah
    How Language Comes to Children: From Birth to Two Years
    by Bénédicte de Boysson-Bardies.
    Cambridge, MA: MIT Press, 1999. 274 pp. $27.50.

    How Children Learn the Meanings of Words

    by Paul Bloom.
    Cambridge, MA: MIT Press, 2000. 300 pp. $39.95.

    Over the past thirty-odd years, many volumes have been published that have attempted to describe how young children acquire language. This modern inquiry into language development began with Roger Brown’s work, especially the 1973 publication of A First Language, and has continued with such recent works as Steven Pinker’s The Language Instinct (1994). The field of language development has progressed far beyond offering mere description of this remarkable process and is now attempting to account for several different facets of language acquisition, such as sensitive periods, pragmatic factors, and crosslinguistic differences, just to name a few. In fact, it seems as though our field has progressed so far so rapidly that we no longer read monographs on “language development” per se, but rather volumes dedicated to specific subdomains and perspectives on language acquisition.

    By far the most crucial challenge facing language-acquisition researchers today is accounting for the mechanism that drives language learning. In the days when behavioral psychology was at its peak, the issue of mechanism was nonexistent; language was seen as a learned behavior like any other. This view of language began to change, however, in 1959, when Noam Chomsky’s review of B. F. Skinner’s Verbal Behavior refuted the notion that language was similar to other observable behaviors, calling this idea “as simple as it is empty” (p. 31). Instead, Chomsky maintained that by studying the formal properties of a grammar, we could come to understand how “human beings are specially designed” (p. 57) to learn language. He posited a “built-in structure” that helps humans acquire the ability to form new, creative sentences in a short amount of time, using relatively impoverished linguistic input.

    While Chomsky’s theory is appealing, it has become unsatisfying for many developmentalists. The fact that Chomsky’s account of language development rested not on observations but on theories of how language could be acquired unsettled empirical researchers. Furthermore, while Chomsky succeeded in accounting for initial and final states of language acquisition, the theory was not particularly developmental. In other words, it lacked a mechanism. While neo-Chomskyites such as Steven Pinker have gone a long way toward relating some empirical findings to the tenets of Chomsky’s innatist account (see Tomasello, 1995), the mechanism is still missing.

    Two more recent perspectives on language acquisition have given more attention to the actual development of linguistic abilities. The constructivist approach views the child as an active participant in her own language learning. The young child interacts with her environment and builds language by communicating with other speakers. Critics of this view, however, note that the psychological mechanism is often underspecified and is sublimated to the social context. The connectionist approach, on the other hand, returns to basic biological mechanisms in order to account for language acquisition. According to this perspective, language acquisition is probabilistic learning. The child’s brain seeks out patterns in the language input and infers permissible sequences of output. Constructivists note that connectionism, while remaining true to human neurobiology, has neglected important aspects of language use such as meaning. It has even been suggested that connectionism is a return to behaviorist stimulus-and-response theories, albeit at the neuronal level.

    Clearly, language-acquisition researchers must continue to address the nature and function of the mechanism that drives language learning. Two recently published volumes offer much-needed ways to continue to approach this monumental task. Through reviewing and analyzing crucial empirical studies from the past forty years, these works offer at least three specific possibilities for further investigating the language-learning mechanism. I contend that the most important messages contained in these volumes are as follows: first, a reintegration of perception into the field of language development; second, an awareness of crosslinguistic data that support or contradict currently held theories about the mechanism; and third, a recognition of the complexity that meaning brings to the debate.


    Perception, generally defined as the selection and organization of sensory information by the central nervous system, has been neglected by many language-acquisition researchers for at least two reasons. First, it has been seen as lying within the domain of experimental psychologists; therefore, those interested in the learning of language in social contexts have tended to view perception as tangential to their work. Second, when one talks of perception in language acquisition, one must necessarily include phonetics and phonology, which is frightening for some and merely tolerable for others. Phonetics and phonology courses are underenrolled and are the least desirable topics for professors to address in their survey language classes. Native English-speaking students find English phonology difficult because they never had to study it, and non-native English-speaking students find it difficult because they had to. As a result, a crucial element of the mechanism of language learning has been overlooked by many in this field.

    How Language Comes to Children by Bénédicte de Boysson-Bardies addresses the importance of perceptual factors throughout most of her book. In spite of its rather general and passive title, this publication is really an incredibly detailed account of developmental phonology in the first two years of life. De Boysson-Bardies has made her career by exploring the earliest language acquisition — the phonology and composition of babbling by infants seven to twelve months old. She believes that “children’s processing of language is initially more acoustic than linguistic” (p. 129). At this early age, a main question “has to do with the reality of innate mechanisms” (p. 9). De Boysson-Bardies modifies the Chomskian notion of innate knowledge by using Darwin’s term instinctive tendency to describe the “program of acquisition that develops on the basis of potentials inscribed in the genetic code of the child” (p. 7) at this very early stage in development. Once the child can segment speech sounds, however, the author ascribes more agency to the young language learner. This development in perceptive abilities, in de Boysson-Bardies’s mind, lays the earliest foundation for language development. The discrimination of vowels and consonants, for example, is used by the ten- to twelve-month-old to register differences in word meaning, such as the difference in English between pat and bat. Further exposure to language-specific sounds in turn helps the child eventually become proficient in language production.

    In How Children Learn the Meanings of Words, Paul Bloom offers another way that perception is crucial to the mechanisms that foster language development. For Bloom, the lexical acquisition process relies on “different cognitive capacities working together in an elegant fashion” (p. 2). In his view, it is not necessary to postulate a mechanism specific to language learning; rather, the learning of words is “a paradigm case of inductive learning” (p. 4). As such, contemporary research on nonlinguistic perception and representation provides us with clues as to how children acquire words. Specifically, Bloom argues that research on children’s object recognition can tell us a lot about how they learn the names of such objects. It is known that visual principles such as continuity, solidity, and cohesion help guide the recognition of objects in children as young as four months old (see Spelke, 1994); likewise, these principles help constrain the language input that the child receives (i.e., input is more likely to refer to objects that are continuous, solid, and cohesive than not). Bloom’s strong claim is that children are able to put these two together; in other words, they know that objects are likely candidates for naming (see pp. 97–98).

    The crucial difference between the perceptive abilities described by de Boysson-Bardies and those described by Bloom is that, while recognition and segmentation of phonemes are by necessity linguistic skills, visual perception is not. This, Bloom would argue, is the point. Whereas some domains of language acquisition (e.g., phonological development) certainly rely on language-specific mechanisms, and other domains may rely on such mechanisms (e.g., syntax), lexical development is unique in that “words are learned through abilities that exist for other purposes” (p. 10).

    Crosslinguistic Data

    The second lesson to be learned from these two works is that crosslinguistic data is indispensable not only to judging the relative contributions of innate tendencies and environmental input, but also to understanding the variation in language learning that must be accounted for by language-learning mechanisms. By crosslinguistic data, I do not mean anecdotes of “the Eskimos-have-fourteen-words-for-snow” type, but, rather, theoretically grounded questions addressed by methodologically valid empirical studies. Such studies need not be conducted in the laboratory; in fact, one might argue that young children should be studied in the environment in which they learn language, be that in the fields, at the favela, or around the dining-room table.

    De Boysson-Bardies is particularly talented in relating the importance of crosslinguistic data. As her book was first published in France, readers in other parts of the world will benefit from her perspective. Readers in the United States, for example, will be interested to read about the “vocabulary illusion” among many U.S. middle-class parents — the tendency to attribute meaning to early, underarticulated utterances. French mothers, on the other hand, see the child as having plenty of time to develop language and are more concerned that they will learn to “speak nicely” (pp. 179–188). Exposure to different languages and to different pragmatic conventions leads even very young children to “specialize” in learning their own language. Chapters two through four provide a particularly informative review of the crosslinguistic data on speech segmentation, language input, and the rarely mentioned subject of prosody. The author’s own work contrasting French and Japanese five- to seven-month-olds’ terminal contours in babbling is almost enough to convince one that babies must be born with innate capacities for analyzing speech. Still, even if one does not subscribe to the “instinctive tendency” perspective, de Boysson-Bardies provides enough crosslinguistic data to inform any view on the necessary language mechanisms needed to account for development of not only “language,” as a proposed cognitive domain, but also “all languages,” as they are acquired by real children.

    Bloom notes the importance of crosslinguistic data in reexamining the notion of semantic bootstrapping, or using syntactic categories to infer lexical meaning (see Pinker, 1994). Semantic bootstrapping necessitates an a priori structure from which children can tell, for example, that a word ending in ing is a verb. Such a mechanism, according to Bloom, would be nearly impossible, given the variation of syntactic structures across the world’s languages. Not only do syntactic cues vary from language to language, but the assignment of cues is random. In other words, there is nothing inherently “verb” about the suffix ing. While de Boysson-Bardies uses crosslinguistic data to explore possible innate tendencies in the youngest infants, Bloom uses it to respond to many universalist notions about language itself.


    As mentioned above, one way that connectionist theories of language development have been critiqued by empirical researchers is that they fail to account for the role of meaning in the language-learning process. Yet the language-learning mechanism must somehow process meaning; to stipulate otherwise would run contrary to observations of researchers and parents alike. The strength of How Children Learn the Meanings of Words is that it does not shirk the challenges that meaning causes for a language mechanism. While it may be economical to postulate a statistical learning mechanism, Bloom argues that it would be incomplete, for “we would still need to know what higher-level capacity . . . this neurological change gives rise to” (p. 44).

    In order to learn words, the child must have access to the form, possess the concept, and be able to connect the form to the concept. While de Boysson-Bardies’s work is primarily concerned with explaining the first of these three requirements, Bloom’s work focuses on the latter two. In reviewing his own work, Bloom maintains that young children are extremely sophisticated at inferring intentions when asked to perform many tasks, such as picture naming. This non–language-specific ability to infer the intentions of others (known as theory of mind) develops concurrently with language and is a crucial component of Bloom’s mechanism. The meanings of many closed-class words (e.g., pronouns, prepositions, articles), for example, cannot be learned through association, but rather are learned as a result of perceiving the behavior and inferring the goals of the child’s interlocutor. Only by recognizing the intention of his mother, for example, does a child learn that his name is not you and that his mother’s name is not me (see Bloom, ch. 8).

    Bloom’s exploration of meaning takes an intriguing yet, in my view, gratuitously philosophical turn in chapter six when he discusses John Locke’s and J. S. Mill’s views on concepts and essences. He returns in chapter ten to the relationship between language and thought, the companions to form and meaning. “Words,” concludes Bloom, “are not necessary for thought. . . . Words are important because they are the building blocks of language, and language allows us to express our thoughts and understand the thoughts of others” (p. 259). De Boysson-Bardies, from her perspective as a researcher in phonological development, concurs: “The gift of speech permits us to create a mental world that enriches communication with others” (p. 2). It would benefit the field of language development greatly to heed these words and recommit to the idea that meaning is what is communicated — language is merely the instrument.

    An oversight may have been committed by focusing on just these three topics to the exclusion of many other, perhaps more important, ones presented in How Language Comes to Children and in How Children Learn the Meanings of Words. De Boysson-Bardies’s section on prenatal linguistic development is not only a superb review of the research, but it is also a simply fascinating read. Bloom’s assertion that the “word spurt” is a myth is such a controversial and potentially paradigm-shattering claim that it deserves a book of its own. However, the three subjects mentioned above — perception, crosslinguistic data, and meaning — are so crucial to the definition of a language-acquisition mechanism that they deserve special attention. Without a complete theory of how children progress from initial state to final linguistic proficiency, the field has much more work ahead.

  2. Share

    Winter 2000 Issue


    Do Teacher Unions Hinder Educational Performance?
    Lessons Learned from State SAT and ACT Scores
    Robert M. Carini, Brian Powell, Lala Carr Steelman
    American Indian Geographies of Identity and Power
    At the Crossroads of Indígena and Mestizaje
    Sandy Marie Anglas Grande
    “Good Enough” Methods for Ethnographic Research
    Wendy Luttrell

    Book Notes

    Call 1-800-513-0763 to order this issue.