Innate or Resilient? How Children's Gesture Creation Uncovers the Ontological Origins of Language

The Resilience of Language: What Gesture Creation in Deaf Children Can Tell Us about How All Children Learn Language by Susan Goldin-Meadow. (New York: Psychology Press, 2005, 288 pp., $24.95, hardcover)

One of the fundamental questions about language development is, what are the unlearned or innate aspects of knowledge that allow language to emerge so quickly in children? In order to situate Goldin-Meadow's approach in The Resilience of Language, it is helpful to consider two examples of the myriad ways researchers have tried to address this question: by comparing perceptual categories with conceptual categories underlying language and by looking at universals across a number of unrelated languages.

The advent of methods that could be used to tap infants' perceptual categories has been a boon for researchers interested in children's early knowledge. For example, based on studies from a habituation paradigm, Spelke (1994) argues that infants have very early knowledge about the characteristics of objects. Similarly, Mandler (1996) argues that children have early knowledge about the path of movement of objects, including into and out of containers and on and off surfaces. This knowledge about the path of movement could lead children to focus on language about how objects move in the world. In fact, children's early language about motion is most often that which is most frequent in their input (e.g., Choi & Bowerman 1991; Sinha and Jensen de López 2000). These results suggest that there may be no simple mapping of early language onto the perceptual categories that can be discerned in infancy. In other words, the knowledge leading to early language use is certainly not exclusively based on earlier perceptual categories. Furthermore, these categories may not be the most important aspects of knowledge on which language use is based.

Another way in which researchers have been interested in uncovering innate or unlearned aspects of knowledge is by cross-linguistic comparisons. By looking at a variety of unrelated languages and showing what is common to all languages, researchers can posit what might be part of humans' genetic predisposition to language. For example, lexical categories such as nouns and verbs are universal across languages (Baker 2003). It is therefore possible that knowledge of lexical categories is either unlearned or at least in place by the time children start to produce word combinations (Wexler and Culicover 1980). This claim has been challenged by recent evidence showing that initially children most often use word combinations that they have heard before (Akhtar 1999; Tomasello 2000), suggesting that knowledge of lexical categories may not precede syntactic acquisition but may rather be the result of using a lot of different word combinations (see also Plunkett and Marchman 1993).

Characterizing just these two approaches to uncovering innate predispositions to acquire language is by no means exhaustive coverage (see, for example, Slobin 1982). I have characterized these two popular approaches in order to highlight what Goldin-Meadow does differently in The Resilience of Language. In this book she switches the question from "What is innate?" to "What is resilient?" By emphasizing the resilience of language rather than innateness, she focuses the question on those aspects of language that can be learned without a model. In doing so, she implies that it may be more appropriate to consider the ontological origins of language in terms of unfolding over a period of time rather than in terms of discrete prelinguistic knowledge (cf. Piaget 1948). While Goldin- Meadow spends little time justifying the switch in questions, it is certainly warranted by pressures in the field. Researchers in cognitive development and language acquisition have been gradually discovering that questions about innate concepts may be ill framed to answer in biological and therefore psychological terms. By focusing the question on resilience, we may be in a better position to understand how language develops over the lifespan.

To support her approach...