Young Toddlers’ Word Comprehension Is Flexible and Efficient

Citation: Bergelson E, Swingley D ( Young Toddlers' Word Comprehension Is Flexible and Efficient Elika Bergelson 0 Daniel Swingley 0 Antoni Rodriguez-Fornells, University of Barcelona, Spain 0 1 Department of Psychology, University of Pennsylvania, Philadelphia, Pennsylvania, United States of America, 2 Institute for Research in Cognitive Science, University of Pennsylvania , Philadelphia, Pennsylvania , United States of America Much of what is known about word recognition in toddlers comes from eyetracking studies. Here we show that the speed and facility with which children recognize words, as revealed in such studies, cannot be attributed to a task-specific, closedset strategy; rather, children's gaze to referents of spoken nouns reflects successful search of the lexicon. Toddlers' spoken word comprehension was examined in the context of pictures that had two possible names (such as a cup of juice which could be called ''cup'' or ''juice'') and pictures that had only one likely name for toddlers (such as ''apple''), using a visual world eye-tracking task and a picture-labeling task (n = 77, mean age, 21 months). Toddlers were just as fast and accurate in fixating named pictures with two likely names as pictures with one. If toddlers do name pictures to themselves, the name provides no apparent benefit in word recognition, because there is no cost to understanding an alternative lexical construal of the picture. In toddlers, as in adults, spoken words rapidly evoke their referents. - Funding: This work was funded by an NSF-Graduate Research Fellowship and an NSF-IGERT Fellowship to EB, as well as by an NICDH grant, R01-HD049681, to DS (www.nsf.gov and www.nichd.nih.gov). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript. Competing Interests: The authors have declared that no competing interests exist. No sooner do we hear the words of a familiar language pronounced in our ears, but the ideas corresponding thereto present themselves to our minds: in the very same instant that sound and the meaning enter the understanding: so closely are they united that it is not in our power to keep out the one, except we exclude the other also. We even act in all respects as if we heard the very thoughts themselves. G. Berkeley, An Essay Towards a New Theory of Vision, Dublin, 1709. One goal of language science is to understand the interpretation of speech: how the spoken words we hear become ideas in our minds. On current psycholinguistic accounts, listeners integrate the speech signal, their knowledge of the language, and their understanding of the speakers likely conversational goals all at once, to arrive at an interpretation of utterances even as the spoken words unfold [1,2]. As Bishop Berkeley [3] suggested, our facility in understanding language is remarkable, and has given rise to a substantial experimental literature characterizing the cognitive mechanisms at work [4,5,6]. The development of spoken language comprehension in children is less well understood, but researchers have claimed that in toddlers, as in adults, interpretation of the speech signal is incremental (children attempt to interpret words while the words are being spoken) and, by 24 months, rapid (understanding of familiar words in two-year-olds is only a fraction of a second slower than in collegiate adults). These claims rely substantially on procedures in which pictures of potential referents or topics are displayed, and then spoken sentences are presented that either refer to the displayed picture(s) or do not (e.g. [7,8]). When a spoken word matches the picture a child is looking at, he or she generally continues to look at it; when a word does not match, he or she tends to look away quickly, and in ERP measurements, may manifest the N400 response reflecting a measure of word understanding [7]. Such results are usually interpreted as revealing childrens ability to understand language in general, at least in simple sentences, and not only language in this constrained experimental situation. If this is so, then asking a child about dogs while she is petting the family Weimaraner is, from a speech processing perspective, not fundamentally different from asking her the same question while the dog is out of sight chasing deer. Yet there has been in recent years a series of empirical reports showing the influence of picture presentation on subsequent language processing, raising the possibility that word recognition is strongly affected by the local context. For example, Mani and Plunkett, Experiment 2 [9] found that prior picture presentation could under certain conditions prevent altogether 24-month-olds understanding of subsequent words. In the adult literature, Glaser and Glaser [10] found that a related context picture slowed down subsequent target naming (though see [11]). More generally, numerous studies across the cognitive sciences point to perception under uncertainty as being fundamentally integrative, with interpretive decisions being a product of strikingly diverse sources of information (e.g. [12,13,14]). Thus a priori there is nothing anomalous or controversial about the possibility that prior picture presentation could affect toddlers word recognition, perhaps substantially. The specific hypothesis we test here is whether recognition of words in object-fixation procedures differs from recognition under more ordinary circumstances by depending primarily on a comparison between the spoken word and the phonological form provoked by viewing an image. We call this hypothesis the phonological pre-activation hypothesis. The idea is that children looking at a duck, for example, name it to themselves: / dk/; and look away from the duck upon hearing a sentence naming anything other than/d k/. Clearly this is quite different from ordinary language understanding, in which words give rise to concepts rather than the other way around. If the phonological preactivation hypothesis is correct, picture fixation procedures as now (widely) implemented would fail to provide appropriate characterizations of word recognition in the wild because they rely on a word comprehension mechanism often unavailable in ordinary discourse. The alternative is that toddlers behavior in looking procedures depends upon a semantic comparison. Children hear a word (book,/bUk/), leading them to think about books. Then they decide whether the image they are considering is a book or not; if its not, they turn their gaze elsewhere. We refer to this hypothesis as the semantic interpretation hypothesis [15]. This hypothesis does not say that children or adults are unable to implicitly (or explicitly) name objects they see, but rather that this is not the typical course of word comprehension, and is not necessary for the rapid word understanding shown throughout the psycholinguistic literature and in particular in the large section of this literature th (...truncated)