Matching behaviours Numerous insightful descriptions of mechanisms that could be responsible for generating particular examples of social learning and related phenomena in ethology and psychology have been described (see e.g. Zentall and B. G. Galef, 1988; Heyes and Galef, 1996; Tomasello and Call, 1997; Byrne and Russon, 1998; Byrne, 1999; Heyes and Ray, 2000; Dautenhahn and Nehaniv, 2002). For robotics and software, particular general methods for solving such problems have been proposed by Nehaniv and Dautenhahn (2001) and Alissandrakis et al . (2002). Despite variations in embodiments and what they afford agents (biological, software or robots), this raises the possibility for harnessing sociality and of an artificial basis for cultural transmission of skills in societies of artifacts, which might also learn from and interact with humans (Dautenhahn, 1995; Billard and Dautenhahn, 1998; Alissandrakis et al ., 2003a, b; Chapter 12, this volume). To formulate problems of matched behaviour in general, we use the following variant of Mitchell's (1987) definition of imitation for what we will call matching behaviour : A behaviour C is produced by an organism and/or machine, where C is similar to another behaviour M , Registration of M is necessary for the production of C , and C is designed to be similar to M . Note that novelty and learning are not explicitly required here, and whether or not the entire behaviour, or its application, or its components or some combination of them is novel or being learned is left open (see Whiten, 2002b; Dautenhahn and Nehaniv, 2002).
Stored only when both papers are in the corpus — 14 citations exist upstream.