Of all the things we could say, what determines what is worth saying? Greenfield's principle of informativeness states that, right from the onset of language, humans selectively comment on whatever they find unexpected. We quantify this tendency using information theoretic measures, and test the counterintuitive prediction that children will produce words that are low frequency given the context because these will be most informative. Using corpora of child directed speech, we identified adjectives that varied in how informative (i.e., unexpected) they were given the noun they modified. Three-year-olds (N=31, replication N=13) heard an experimenter use these adjectives to describe pictures. The children's task was then to describe the pictures to another person. As the information content of the experimenter's adjective increased, so did children's tendency to comment on the feature that adjective had encoded. Furthermore, our analyses suggest that children balance this informativeness with a competing drive to ease production.Keywords: Information theory, pragmatics, child language, language production. Greenfield's principle of informativeness (Greenfield, 1979;Greenfield & Smith, 1976), proposes that, right from the onset of language, infants choose to comment on things they find unexpected or uncertain and leave unmentioned whatever is constant or can be assumed. Greenfield suggested that this behaviour might be captured by the concept of information provided by the mathematical theory of communication (Shannon, 1948). In this case, a message provides information to the extent that it is unpredictable given what is already known. However, this early sketch of how to quantify informativeness was abandoned following a critique by Pea (1979). In this study, we demonstrate that an information theoretic approach is viable.
WHAT S WORTH TALKING ABOUTMoreover, adopting it brings to light a trade-off between informativeness (which requires using unlikely forms) and ease of production (which favours using likelyi.e., frequent -forms). We show that, even from 3 years of age, while children find it easier to produce frequent forms, they nonetheless make the effort to talk about the unexpected.
WHAT S WORTH TALKING ABOUT 4Greenfield's Principle of Informativeness Greenfield & Smith (1976) analysed naturalistic recordings of children at the one-word stage and observed cases where a child had the choice of saying one of two words to talk about an event (e.g., the words skate and on to talk about putting skates on). They explained why one or other word was said at any given moment using the concept of uncertainty. For example, if the object was out of the child's possession, then it became uncertain and the word referring to it (skate) was likely to be produced.If the child had the object, it became certain and they would express something else, such as a desired change of state (on). On the basis of these observations, Greenfield and Smith argued for 'certainty-uncertainty as the perceptual-cognitive basis for the...