Fast Mapping and Exclusion (Emergent Matching) in Developmental Language, Behavior Analysis, and Animal Cognition Research

Article excerpt

For many years, psycholinguists, behavior analysts, and animal cognition researchers have investigated processes by which their respective participant populations learn new arbitrary relations without explicit training. Psycholinguists have studied young children's "fast mapping" of new words, a phenomenon that likely underlies the virtual explosion of word learning that occurs in the preschool years (e.g., Carey & Bartlett, 1978; Golinkoff, Mervis, & Hirsh-Pasek, 1994). Behavior analysts have studied so-called "exclusion" performances, in which participants immediately display arbitrary matching relations without explicit differential reinforcement (Dixon, 1977; McIlvane, Kledaras, Lowry, & Stoddard, 1992). Animal cognition researchers have used similar methods to examine the behavior of higher primates, dolphins, and sea lions (e.g., Premack, 1976; Schusterman & Kreiger, 1984). Despite well-known differences in philosophy and terminology, the programs have developed similar procedures and reported similar behavioral outcomes (Huntley & Ghezzi, 1993; Wilkinson, Dube, & McIlvane, 1996). These convergences suggest that the programs are studying the same phenomena. If so, we suggest that all areas of research may benefit from considering and adapting the best ideas and methods currently available, regardless of source.

To promote interdisciplinary discourse, this paper will (1) review the development of the research in each discipline and the converging outcomes reported, and (2) outline several research questions in which the same issues have been of interest to each. These analyses, particularly the latter, will provide a foundation for illustrating how interdisciplinary communication might benefit research in each of the respective disciplines.

Exclusion and Fast Mapping Procedures

Exclusion or fast mapping typically occurs in the context of a well-learned matching-to-sample (MTS) baseline. For example, an experimenter may display an array of familiar items and speak the name of one; the participant indicates which one was named. The top two rows in Figure 1 show examples of MTS trials that would be appropriate for a young child, and which are considered to represent tests of receptive language (e.g., Peabody Picture Vocabulary Test-Revised; Dunn & Dunn, 1981) or the repertoire of the listener (Skinner, 1957). With appropriate training procedures, most humans and many nonhuman animals demonstrate MTS performances (e.g., Carter & Werner, 1978).

Emergent matching. Given a set of defined relations in a MTS baseline, the experimenter, without any special instruction, may display an unfamiliar item in the array and give an unfamiliar name. The examples in the upper portion of Figure 2 illustrate such a task, where the participants might be asked "Which one is the symu?" or "Which one is the breel?". Virtually all published research thus far has reported that participants select the unfamiliar item in this situation. This performance has been called exclusion by behavior analysts (Dixon, 1977) and nonhuman animal cognition/language researchers (Schusterman, Gisiner, Grimm, & Hanggi, 1993) and, often, the disambiguation effect in psycholinguistics (Merriman & Bowman, 1989). For our present purposes, we shall henceforth use another, more neutral term - emergent matching (EM) - which is descriptive, makes no assumptions about the basis for the participant's selection, and is not associated with any particularly theoretical position. The term "emergent symbolic mapping" has also been used (Wilkinson et al., 1996) to emphasize that exposure to EM trials may establish mapping relations (i.e., equivalence relations) between undefined sample and comparison stimuli.

EM has been documented in 2- and 3-year-old children (Dollaghan, 1985; Golinkoff, Hirsh-Pasek, Bailey, & Wenger, 1992; Kagan, 1981; Markman, 1989; Merriman & Bowman, 1989), children with specific language impairments (Dollaghan, 1987), people with severe mental retardation (McIlvane et al. …