Verbs and Nouns 1 Running Head: VERBS AND NOUNS 24-Month-Old Infants’ Interpretations of Novel Verbs and Nouns in Dynamic Scenes Sandra R. Waxman Northwestern University Jeffrey L. Lidz University of Maryland Irena E. Braun Northwestern University Tracy Lavin Canadian Council on Learning Vancouver, British Columbia, CanadaVerbs and Nouns 2 Abstract Recent evidence reveals that infants share with mature language users an expectation that different kinds of words (e.g., nouns, adjectives) refer to different aspects of a scene (e.g., object categories, object properties). Infants’ expectations for nouns and adjectives have been demonstrated, but the developmental trajectory underlying infants’ discovery of verbs has been more elusive. Although infants begin to produce verbs spontaneously by roughly 24 months, they show surprising verb-learning difficulties in experimental tasks, many of which persist well into the preschool years. The goal of these two experiments was to clarify the breadth and precision of 24-month-old infants’ mappings of novel verbs and nouns. We presented dynamic scenes (e.g., a man waving a balloon), and asked a) whether infants could construe these scenes flexibly, noticing the consistent action (e.g., waving) as well as the consistent object (e.g., the balloon) and b) whether their construals would vary with the grammatical form of the novel word used to describe the scene. Results revealed that infants successfully mapped novel nouns to object categories and novel verbs to event categories. Moreover, infants’ representations of word meaning were sufficiently abstract to permit them to extend novel verbs and nouns appropriately beyond the precise scenes on which they had been taught. Keywords: language acquisition, word learning, concept developmentVerbs and Nouns 3 24-Month-Old Infants’ Interpretations of Novel Verbs and Nouns in Dynamic Scenes How do infants discover that there are distinct kinds of words (e.g., nouns, verbs, adjectives) and that each brings with it a distinct set of semantic consequences? Speaking broadly, they must somehow discover that nouns refer to categories of objects, adjectives to properties of objects and verbs to categories of events and relations. At the onset of word learning, infants do not yet concern themselves with these distinctions. Instead, they enter the process with a broad initial expectation that links words in general (independent of their grammatical form) to a wide range of candidate meanings (including, for example, individual objects (e.g., Mommy, Fido), categories of objects (e.g., bottles, balls), properties of objects (e.g., soft, hot), and relations or actions in which they are engaged (e.g., holding, running) (Waxman & Booth, 2003; Waxman & Lidz, in press). From the perspective of the learner, this initially broad expectation offers an important developmental advantage: because different languages make use of different sets of grammatical categories and because they use these to carve up semantic space in slightly different ways (Baker, 2001; Croft, 1991; Frawley, 1992; Gentner & Goldin-Meadow (Eds.), 2003; Imai & Haryu, 2004; Lucy & Gaskins, 2001), it is to the learner’s advantage to begin with the most general of expectations, one that highlights a range of commonalities and enables the learner to break into the system of word learning in the first place. But it is also to the learner’s advantage to fine tune these general expectations in accordance with their native language. The evidence suggests that by 14 months of age, this fine tuning has begun, and that at this developmental point, infants have begun to tease apart the nouns from other grammatical forms (e.g., adjectives) and map them specifically to object categories and not their surface properties (e.g., color, texture) (Booth & Waxman, 2003; Waxman, 1999; Waxman & Booth, 2003; Waxman & Braun,Verbs and Nouns 4 2005; Waxman & Markow, 1995). However, at this same point, infants maintain a broader expectation for words from other grammatical categories. For example, 14-month-old infants interpret novel adjectives quite broadly, mapping them to either object categories or object properties (Booth & Waxman, 2003; Waxman, 1999). By roughly 21 months, infants have successfully carved out a more precise expectation for adjectives, linking them specifically to properties (and not categories) of objects (Mintz, 2005; Mintz & Gleitman, 2002; Waxman & Markow, 1998). Thus, over the second year of life, infants’ initially broad expectation linking words to meaning becomes increasingly refined as infants discover which grammatical forms are represented in their native language and how each is recruited to convey meaning. The evidence for this developmental progression is now robust. However, the research to date has focused on a limited range of grammatical forms. Although considerable attention has been devoted to documenting the evolution of infants’ expectations for nouns and adjectives, the developmental trajectory underlying infants’ discovery of another fundamental grammatical form – verb – has been more elusive (e.g., Bloom, Tinker, & Margulis, 1993; Bloom, 2000; Booth & Waxman, 2003; Gasser & Smith, 1998; Gleitman, Cassidy, Nappa, Papafragou, & Trueswell, 2005; Golinkoff et al., 2000; Hall & Belanger, 2005; Hall & Lavin, 2004; Hall & Waxman, 2004; Hollich, Hirsch-Pasek, & Golinkoff, 2000; Mintz & Gleitman, 2002; Waxman & Booth, 2001; Waxman & Braun, 2005; Waxman & Lidz, in press; Woodward & Markman, 1998; Xu, 2002). As a result, it is still unclear whether infants’ specific expectations for verb meaning emerge early (as is the case for nouns) or evolve more slowly (as is the case for adjectives). Because verbs are universally represented across languages, and because they play a pivotal role in grammatical knowledge, it is often assumed that infants’ expectations for verbs emerges early, as is the case for nouns.Verbs and Nouns 5 However, a review of the theoretical literature suggests that this might not be the case. First, despite the universality of the category verb, languages differ substantially in the particular aspects of events that are lexicalized within the verb system (Berman & Slobin, 1994; Navarro & Nicoladis, 2005; Talmy, 1985, 2000). Cross-linguistic differences like these are directly relevant to acquisition, for if learners are to establish appropriate links between the verbs in their