Picture Naming and Lexical Access in Italian Children and Adults

Normative data are described and compared for 34 Italian-speaking children (5-6 years of age) and 50 Italian-speaking adults in a timed picture-naming task, with 250 pictures (simple line drawings). Dependent measures include overall nameability, percent agreement on the most frequent name (target), number of alternative names provided, overall reaction time and latency to produce the target name. Independent measures (characteristics of target words and pictures that might affect naming) include frequency (from both adult and child norms), age of acquisition (an objective measure from early lexical development norms, and a subjective measure based on adult ratings), length (in syllables and characters), animacy, semantic category, various word structure and grammatical category measures specific to Italian, and an objective measure of picture complexity. Although children were substantially slower and less accurate than adults, child and adult performance was highly correlated, and similar correlations were obtained for children and adults between lexical predictors and naming times. However, word complexity had effects on adults that were not seen in children, and grammatical gender had effects on children that were not seen in adults. Adult ratings of age of acquisition had strong effects on both children and adults (and reduced or eliminated effects of frequency in regression analyses), but an objective measure of age of acquisition only affected children (and did not eliminate frequency effects in regression analyses). Differences were also observed in the semantic categories that were easiest for children vs. adults. Naming is a fundamental aspect of human language use (Brown, 1958; Terrace, 1985), and it is one of the first linguistic functions mastered by small children (Bates, Camaioni, & Volterra, 1975). There is a long tradition of research aimed at understanding how people retrieve and produce names for things (Cattell, 1886), and how children achieve this competence (DeLaguna, 1927; Dromi, 1987; Greenfield & Smith, 1976; Leonard, 1998). In research with adults, timed picturenaming tasks are often used to investigate the naming process (Levelt, 1989; Levelt, Roelofs, & Meyer, 1999), including comparisons between picture naming and word reading (Paivio, Clark, Digdon, & Bons, 1989; Potter, Kroll, Yachzel, Carpenter, & Sherman, 1986), effects of sentential and grammatical structure on word retrieval (Bentrovato, Devescovi, D’Amico, & Bates, 1999; Jacobsen, 1999; Wicha, Bates, Hernandez, Reyes, & Gavaldón de Barreto, 1997), and the effects of congruent or incongruent word distracters on picturenaming times (Glaser, 1992). Both pictures and words are thought of as symbols, standing for referents that may not be physically present when the symbols are used. Because pictures bear a transparent iconic relationship to their referents, it is generally assumed that they should be readily interpretable by children. “Recognizing pictures does not require particular steps of learning or development beyond learning to know the represented objects... [whereas] the relation between a noun and the corresponding class of objects is determined during the centuries of evolution of a language” (Glaser, 1992). In fact, pictorial stimuli have been used with considerable success in studies of word production and comprehension in young children (e.g. Bates, Bretherton, & Snyder, 1988). Indeed, looking at picture books is a normal activity in young children's daily life, even in early infancy. Many authors report that children from the first year of their life are interested in this activity, and start to name two-dimensional representations of well-known objects as early as 12 months of age, at the same time that they begin to name those objects in real life (Bates, Benigni, Bretherton, Camaioni, & Volterra, 1979; Ninio & Bruner, 1988). Presumably, the naming process in young children involves many (perhaps all) of the basic processes that have been studied with adults using timed picturenaming tasks, but for a variety of reasons (most of them practical), reaction times studies of picture naming in children are relatively rare (Berman, Friedman, Hamberger, & Snodgrass, 1989; Cycowicz, Friedman, & Rothstein, 1997; Johnson, 1992; Roe et al., in press). The underlying cognitive process of naming a picture has been articulated by Johnson, Paivio, and Clark (1996) in three broad stages. The first step includes the identification of the object as a member of a particular class of objects; the second consists in name activation of the object from among thousands of words known by users; and finally, in the last step, articulatory commands for a specific response must be prepared and executed. These sophisticated operations must occur rapidly and efficiently in fluent speech (Johnson et al., 1996). One approach to the study of this process has been to vary individual characteristics of both the picture and the word, and to observe which characteristics affect the choice of a particular name as a target, and the time taken to do this (Cattell, 1886), in an effort to tease apart the stage in picture naming at which each variable has its effect. A large body of research conducted with this intent has shown that picture naming by adults is affected by frequency, familiarity, the age at which the word was learned (called Age of Acquisition, or AoA, measured various ways), length (in syllables, characters or phonemes), imageability (although by definition all pictures can be imaged to some degree), and degrees of abstractness or concreteness. In many of these studies,