Imitation as a learning mechanism and research tool: how does imitation interact with other cognitive functions?

© Blackwell Publishers Ltd. 2002 Want and Harris highlighted the lack of scrutiny that human imitation research has received regarding the specific mechanism of learning. Given the available data, they proposed a developmental progression from mimicry to imitation and imitation to emulation. Importantly, they recognized that the learning mechanism employed, however, also depends on the child’s general knowledge, task complexity and social motivational factors. An overriding problem described by the authors was the unknown interrelationships between different learning mechanisms and the social dynamic of the imitation situation. In addition, the authors were concerned that the developmental trajectory of tool-use learning in human childhood is relatively uncharted. They proposed that investigating imitation of tool use was the best way to uncover imitation mechanisms because imitation of ‘simple actions’ cannot address the question. Some clues to their concerns are already available in the literature of simple action tasks. As mentioned by the authors, the simple imitation tasks do a good job of ruling out local or stimulus enhancement by using adult manipulation controls (e.g. Meltzoff, 1985). Furthermore, one important safeguard that was not mentioned by Want and Harris, was that in some tasks used with infants the goal or outcome of the event is hidden, and therefore, is unlikely to be produced by emulation because the affordances are not visible to the infant (Barr, Dowden & Hayne, 1996; Meltzoff, 1985). As the authors pointed out, imitation is a highly adaptive mechanism in situations when it is not readily apparent how the demonstrator is solving the problem. While the tasks described above allow us to distinguish between imitation and emulation, they do not distinguish between mimicry and imitation. Some studies of sequencing may, however, have begun to address that issue by examining how including irrelevant components disrupts sequencing. In general, infants’ imitation of a sequence of actions that can only be performed in a specific order (enabling sequence – such as making a rattle by placing a ball in a container, putting a lid on it, and shaking it) is consistently superior to their imitation of a sequence of actions that can be performed in any order (arbitrary sequence; e.g. Bauer & Mandler, 1989; Bauer, 1992). Want and Harris reported that children tend to reproduce irrelevant actions when reproducing causally related sequences. The authors did not point out, however, that the irrelevant components are reliably omitted or displaced to the end of the sequence during re-enactment by children between the ages of 19 months and 7 years (Bauer, 1992; Bauer & Mandler, 1989; Hudson & Nelson, 1983; Price & Goodman, 1990). Importantly, these studies did include pretest baseline levels. Bauer and Mandler (1989), for example, added the irrelevant component of attaching a sticker during the demonstration of making a rattle (an enabling sequence). Both 19and 25-month-olds displaced attaching the sticker to the end of the sequence so as to maintain the enabling sequence order. The irrelevant action was not displaced in a systematic way for arbitrary sequences. The act of omission or displacement during the enabling sequence imitation reflects infants’ knowledge of the causal structure of the event and not mimicry of that event. Although Want and Harris suggest that studies of tool use may shed new light on the development of imitation, there are methodological problems with this approach. One of the problems of a focus on tool use might be that tool use necessarily requires more complex motor skills than other object manipulations and inability to imitate tool use may reflect motor incompetence, rather than imitative difficulty per se. It also requires a number of other cognitive abilities. Tool use is relational in nature. It potentially requires combinatorial, relational, causality and sequencing ability, abilities that develop concurrently with imitation. Reliance on tool use to differentiate mimicry, imitation and emulation from one another may be confounded by infants’ developmental ability to combine information. In a recent diary study of imitation by 12to 18-month-olds under naturalistic conditions, imitation of tool use such as painting the wall Imitation as a learning mechanism and research tool: how does imitation interact with other cognitive functions?