Form and function in very early word learning

Learning a new word is, in part, acquiring an association between the word and the object or event to which it refers, but mature word learners have honed their expectations how words are related to objects. Specifically, they understand that words are symbols that are used by speakers to communicate about referents. A critical task in development, then, is learning about this relation between a word and its referent. A second critical task is determining which kinds of behaviors and signals serve this function. Although, in principle, the forms of words are arbitrary, within a language, there is a strong regularity in the form of words. As examples, in spoken languages words are units of speech, and in sign languages, they are manual patterns. The link between form and function here e.g., between spoken words and naming-is so obvious to adults that it is easy to assume that infants begin the language learning game with this correspondence in mind. However, consider that the infant must abstract this regularity in the context of learning many things from and about the behavior of other people, many of them with reference to objects (the sounds they make, their typical uses, etc.). It may take infants some time to sort out the forms and functions of names for things. One-year-olds seem to understand a range of relations between sounds and objects. For example, they may know that the sound of an airplane means that if they look up, they will see the airplane.They also seem to understand the relations between some words and their referents. According to one traditional model of language development, early on there is no distinction between these two sorts of learning, in that they are both based on contiguity between the sound or word and the item. The work of Baldwin (Baldwin, 1991; Baldwin, 1995) and Tomasello (Tomasello, 1995; TomaseUo & Barton, 1994), among others has provided compelling evidence that by the time babies are 20-24 months of age, this model of word learning will not work. In fact, these studies show that contiguity between name and referent is neither sufficient nor necessary for word learning in 20-24-month-olds. Instead, babies filter their word learning through their attention to and reasoning about the behavior of the person who uses the word. In particular, they seem to use behavioral cues to the speaker’s referential intent to inform their word learning (see Baldwin, 1995; Baldwin & Tomasello, in press; Tomasello, 1995 for reviews). Thus, by this age, babies seem to understand something about the referential and communicative function of words. So far, researchers have not made much progress at exploring this knowledge in babies younger than 2024 months. However, well before this, as young as 12 months, babies seem to understand many of the words they hear, and even produce a few of them (Woodward & Markman, 1997). A common assumption is that before the productive vocabulary spurt, at around 18 months, infants lack the insight that words are symbolic (the "nominal insight"). It is argued, then, that very young word learners acquire word-object relations via slow, laborious, associative processes (which accounts for their small productive vocabularies). In prior work, my collaborators and I showed this assumption to be wrong (Woodward, Markman, & Fitzsimmons, 1994), in that 13-montholds acquire new word-object mappings in comprehension based on brief training, as do 18-month-olds. However, this finding does not speak to the nature and specificity of these mappings. Do 13-month-olds, like older babies, understand that words are a special class of sigual, that relate to objects and events in a particular way? In recent work, I have taken on this question from two perspectives: (1) I have tested whether infants distinguish between words and other sounds in contexts that are typical for word use (communication about objects) and those that are not (operant conditioning), and (2) I have begun to assess the tent to which 13-month-olds draw on behavioral cues to speaker’s intent in acquiring new words.