ABSTRACT
The idea that there is no representational distinction between the interpretation of language and the interpretation of the world about us is compatible with the approach to cognition termed embodied cognition. According to this view, cognition is rooted in the interactions between the cognitive agent and its environment and, hence, in perception and action (cognition is thus constrained by the sensory modalities through which we perceive and act on the world). The job of cognition is not to reconstruct, mentally, the external world, but rather, to enable interaction with that world. Cognition is therefore embodied in representations that are based on such interaction and is thus rooted in motoric and sensory representation (for recent reviews see Barsalou, 1999; Barsalou et al., 2003; Glenberg, 1997; Lakoff & Johnson, 1999; see also Spivey et al., this volume)3. However, the idea that the representations that result from language interpretation are in some sense equivalent to those that would result if the event being described by that language were being observed directly is not as straightforward as it may at first glance appear. There must be some distinction between the mental representation of the world (constructed through prior linguistic or visual input) and the real time perception of that world-if only because (a) the sensory component is lacking when we refer to an event that is not taking place before us, and (b) the temporal properties of the events that we can observe in the real world cannot be directly “mirrored” in the mental version of those events. An event that we observe to take place over a period of some minutes is not encoded in a form that itself takes a period of some minutes to unfold. And, of course, the temporal properties of events dynamically change as time passes: an event that was observed two days ago was not perceived, at the time of that event’s unfolding, as taking place in the past, and its temporal encoding now (two days after) is different from the encoding it will have two months hence. To put it simply, the
temporal properties of the perceived event and the corresponding temporal encoding within the mental representation of that event cannot be fully isomorphic (but see, for example, Zwaan, 1996, and van der Meer, Beyer, Heinze, & Badel, 2002, for data suggesting some, perhaps limited, isomorphism). Thus, the distinction between the memory of an event and the perception of that event (in real time) does have implications for a representational distinction between the two, at least with regard to the representation of the temporal properties of that event and the sensory components that accompany real time perception.
