linguistlist.org>
Editor for this issue: Andrew Carnie <carnielinguistlist.org>
linguistlist.org
phys.ufl.edu>Hirsh-Pasek K. & Golinkoff, R. M. (1996). The Origins of Grammar: Evidence from Early Language Comprehension. MIT Press. 201 pages. Reviewed by Dina Belyayeva, CUNY Synopsis: This book provides a thorough description of a new methodological tool, the Intermodal Preferential Looking Paradigm, and outlines a Coalition Model of Language Comprehension. Chapter 1. Introduction. The introduction draws attention to the overwhelming complexity of a task that children undertake to learn a language. The authors present a very brief overview of presumptions held by prior theories of language acquisition in order to set a premise for further discussion. They contend that only a "biased learner" who relies on varied coalitions of input sources can cope with the task of such enormity. They also introduce the reader to the structure of the book and the goals they pursue. Chapter 2. Theories of Language Acquisition. In this chapter the authors provide a taxonomy of pre-1996 language acquisition theories. They use extreme nativist and empiricist positions as the opposite ends of the three continua that represent the source of initial language structure (innate vs. constructed), the type of input (purely linguistic vs. environmentally constructed categories) and the mechanisms employed in the acquisition process (domain-specific vs. domain-general). The authors caution against the literal interpretation of the proposed taxonomy, since the primary purpose of their effort was not to enter into the alliance with any of the reviewed positions but rather to use "previously undiscovered consensus among the theories" in order to establish the types of input that help a child to take the language learning process off the ground. Chapter 3. The Intermodal Preferential Looking Paradigm. The chapter is devoted to the detailed description of the Intermodal Preferential Looking Paradigm introduced in this book. The paradigm represents a new methodology for assessing early language comprehension. The beginning of this chapter provides a review of methods used to assess language production and comprehension in nascent learners. The primarily focus is given to methods for studying language comprehension. The authors outline general advantages of using these methods in language development research. They contend that these methods provide more accurate account of the child's emerging language system, capture the time-window when processing of a particular structure begins before the child attempts to produce it, and can be used as a means to control the child's ability to comprehend certain grammatical structures. In addition to the above advantages, the Intermodal Preferential Paradigm offers two other benefits that make this method especially attractive to anyone working on early language assessment. The method requires "no overt action on the part of the child" and can be used to assess the knowledge of grammatical relationships via presentation of "dynamic" stimuli. The method is based on the procedure that makes use of the child's visual fixation responses; that is, whether the child looks more at one stimulus than at the other. It was adapted from the work of Spelker (1979) who studied intermodal perception in infants. Here is a much simplified outline of the paradigm. During an experimental session an infant is seated on a parent's lap between two television screens. A concealed speaker plays a linguistic stimulus that matches an action going on on one of the screens. An experimenter's task is to register whether the child allocates more attention to the action that matches the auditory stimulus. The rest of the chapter provides a detailed account of the general experimental procedure, apparatus, materials, experimental variables, participant solicitation, and the criteria used for discarding data. Chapters 4, 5 & 6 present a series of studies with the two general goals: (1) to determine the time-window when infants become sensitive to meaning differences in various linguistic structures, and (2) the type of input that they chose to rely on at every given time-window. The purpose of the study described in Chapter 4 was to identify whether infants who have only single-word utterances in their repertoire can isolate sentence constituents (e.g., verb phrase). To test this processing ability, the infants were presented with two similar-looking video events. For example, on one screen they saw a woman kissed a ball while moving a set of keys on the foreground, whereas on the other screen a different woman kissed the keys while moving the ball in the foreground. The auditory stimulus "She is kissing the keys" was played to test whether infants can identify the verb phase by directing more of their attention to the matching action on one of the screens. A series of experimental conditions were introduced to control for the factors such as name familiarity, action typicality, sentence- final position. The results revealed that 13- to 15-month-old infants were able to process the verb phrase without reliance on their extralinguistic knowledge that could be be manifested in better performance on the sentences describing more typical events. Chapter 5 describes a series of experiments that tested the ability single-word speakers to comprehend word order in active reversible sentences. (e.g. "Look! Big Bird is feeding Cookie Monster!") Although the results demonstrated infants' comprehension of word order, it was not entirely clear whether the infants carried out syntactic (subject-verb-object) or semantic (agent-action-patient) analysis. Chapter 6 presents a series of experiments that were conducted with the purpose to distinguish between the two types of analyses. Children were presented with sentences hat contrasted transitive and intransitive frames of the same action verbs. (E.g., "Big Bird is bending Cookie Monster!" vs "Big Bird and Cookie Monster are bending!"/"Big Bird is bending with Cookie Monster!") The authors used the cross- sectional comparison of three age groups, 18-, 24- and 29-month-old children. The results revealed that only the two older groups were able to override the "agent- action-patient" interpretation of word order by demonstrating their preference for the intransitive event. This ability was lost when intransitive frames were presented in the bare sentence condition, without the grammatical markers such as the preposition "with" and the plural form of the auxiliary verb "are". (E.g., "Watch Big Bird and Cookie Monster bending!") The developmental picture that emerges from the study suggests that older children are able to switch from greater reliance on semantic processing to more prominent reliance on syntactic processing when additional structural information was made available to them. Chapter 7. A Coalition Model of Language Comprehension. In this chapter the authors recapitulate the experimental finding within the new theoretical framework. The framework was proposed with the purpose to answer the question that is considered to be the source of major disagreement between the language acquisition theories. What inputs drive the language-learning system? The framework uses language comprehension as its major functional element. Along the developmental continuum, comprehension evolves from internalization to interpretation. Internalization involves extraction of "acoustic packages" from the speech stream. At this point of development (phase I) infants form associations between perceptual correlates that provide building blocks for future interpretations. Phase II marks transition from internalization to interpretation when children use their ability to parse the speech sequence to "carve the observed world into events and sequences." At this phase they rely on redundant cues from the coalition of environmental, contextual, social, prosodic, semantic and syntactic cues with a bias toward semantic analysis. Phase III represents an advanced stage in language development. Children become less dependent on correlated cues in the input and gradually switch to relatively independent syntactic analysis. Although the cues are available at all times, they are differentially weighted in the tree phases. In Phase I children are biased to focus on prosodic cues, whereas in later phases (II & III) they are biased to rely on semantic and syntactic cues, respectively. In conclusion, the authors contend that the proposed coalition of cues resolves the controversy about the as the driving force that helps to take the language learning process off the ground (e.g., the controversy between the semantic and syntactic bootstrapping). By achieving that they hope to take the discussion "beyond the traditional nativist- empiricist dichotomy that permeates much of the field". Evaluation: One of the goals that the authors set in the Introduction was "to make the domain of language acquisition accessible to psychologist through clear and relatively jargon-free exposition." The authors achieved that goal. In fact they were so successful in their effort that the book can be recommended for a much wider audience. The manner in which the book is composed and written makes the experimental methodology and theoretical considerations behind it highly accessible to anyone interested in the issues related to language acquisition. The book in its entirety presents a useful supplemental reading for the graduate- level course. Parts of it can be used to introduce undergraduate students to the variety of theoretical positions permeating the field. The overview of language acquisition theories in Chapter 2 offers a concise yet thorough reference guide to an array of theoretical issues in language acquisition. Exceptional clarity of research objectives and painstaking thoroughness in the description of the experimental paradigm can be used as a handbook for anyone learning how to conduct experimental research in the field of language development. The authors were also successful in attaining their primary goals. The new method, the Intermodal Preferential Looking Paradigm, was proven to offer a tool that would enable a researcher to investigate the nascent knowledge of language structure, such as sensitivity to constituent structure & verb subcategorization frames. Among other advantages mentioned in the book the method offers a way to measure duration of infants' attentional states without the need to use cumbersome eye tracking equipment. The intermediate position that the authors adopted in respect to their theoretical affiliation allowed them to take their investigation beyond the plateau of the nature-nurture debate. There are a few points that I found open for critique. One of them is the taxonomy offered in Chapter 2. Although the authors explicitly warn against the literal interpretation of their hyperbolic dichotomies, I find it difficult to ignore the fact that innate and constructed structures are portrayed as the opposite ends of the same continuum. Even the most medial position (in respect to the position on the nature-nurture spectrum of theories), the Emergentist Approach (MacWhinney 1999) would not accept the continuum as a way to define its position on the "grammar organ." In Chapter 6 the liberal use of the term 'active' (p. 147) entails that sentences with intransitive frames are not active sentences. In their discussion of the role language comprehension plays in the overall cognitive development, the authors claim that mental models are constructed via the processes involved in language comprehension. Although a serious attempt was made to define the notion of a mental model, little ground was provided to substantiate that claim. The stronger version of this claim can be rooted in the assumption that preverbal infants do not have mental models of their own, which is hard to prove. The weaker version may be rooted in what constitutes the anatomical basis of language capacity. Caplan (1987) in his review of Geschwind's work points out that humans are the only species able to form direct connections between visual, auditory and somesthetic (i.e., sensory) association areas. According to such neural organization, "acoustic packaging", the comprehension process of phase I, constitutes one type of the stimuli involved in formation of cross-modal associations that contribute to construction of mental models. The mechanisms proposed to motivate a transition within and between phases (the guided distributional analysis and Bloom's Principles of Discrepancy and Elaboration) do not differ significantly in respect to their major driving force. In fact, they are essentially the same mechanism that has different manifestations at every developmental phase due to different cognitive and communicative demands at a given stage of language development. The above critical remarks do not diminish the significance of the proposed model. The reader should bear in mind that the authors consider their model as a work in progress. This work could only develop out of the new experimental methodology presented in the book. Bibliography: Caplan, D. (1987). Neurolinguistics and Linguistic Aphasiology. Cambridge University Press. MacWhinney, B. (Ed.). (1999). The Emergence of Language. Lawrence Erlbaum Associates, Publishers. Mahwah, NJ. Spelke, E. (1979). Perceiving bimodally specified events in infancy. Developmental Psychology 15, 626-636. About the reviewer Dina Belyayeva is a Research Associate at the City University of New York. She earned her Ph. D. in Lingusitics at the University of Florida. In her doctoral dissertation she proposed a model of the bilingual memory that has implications for many language acquisition phenomena. Language Acquisition and Bilingualism are currently the major areas of her research interests. Other areas of interest include semantic memory disorders and models of language production and comprehension.Mail to author|Respond to list|Read more issues|LINGUIST home page|Top of issue