this paper on these three: representation, action, and motivation. In particular, I will argue that the standard view of representation as some kind of correspondence, as an encoding, is wrong. I outline an alternative model of representation that emerges naturally in agents, biological or designed, that actually engage the world (Beer, 1990, 1995, in press; Beer, Chiel, Stirling, 1990; Bickhard, 1980, 1993; Bickhard & Terveen, 1995; Brooks, 1991a, 1991b, 1991c; Cherian & Troxell, in press; Malcolm, Smithers, Hallam, 1989; Smithers, 1994). One primary consequence of this alternative model of representation --- called interactivism --- is that functions that are standardly taken to reside in separate modules, such as representation, action, and motivation, are inherently integrated as separate functional aspects of one single underlying ontology. They are not inherently distinct modules. If standard models that permit such modularization are in error, then so are such modularizations per se. 2 Encoding Models of Representation.

The set of English modal verbs is widely recognized to communicate two broad clusters of meanings: epistemic and root modal meanings. A number of researchers have claimed that root meanings are acquired earlier than epistemic ones; this claim has subsequently been employed in the linguistics literature as an argument for the position that English modal verbs are polysemous (Sweetser, 1990). In this paper I offer an alternative explanation for the later emergence of epistemic interpretations by linking them to the development of the child's theory of mind (Wellman, 1990); if correct, this hypothesis might have important implications for the shape of the semantics of modal verbs.

How systems can represent and how systems can learn are two central problems in the study of cognition. Conventional contemporary approaches to these problems are vitiated by a shared error in their presuppositions about representation. Consequently, such approaches share further errors about the sorts of architectures that are required to support either representation or learning. We argue that the architectural requirements for genuine representing systems lead to architectural characteristics that are necessary (though not sufficient) for heuristic learning and development. These architectural constraints, in turn, explain properties of the functioning of the central nervous system that remain inexplicable for standard approaches. Topologies