Abstract

This article reviews the classical imagery-propositional debate, in which some theorists have argued that images are not really a separate type of representation but can be reduced to propositional representations, and some imagery researchers have refused this thesis. The authors present then two areas in cognitive science in which some central aspects of the debate seen to reappear: text comprehension and analogical thinking. While some researchers in these fields have proposed propositional models, some others have postulated diverse analogical representations to complement or replace propositions. Considering these more recent controversies, it is concluded that, against Pylyshyn´s reductionism, there is general agreement today that different representations are needed to characterize the richness of human cognition. However, propositions are still very useful: they seem necessary for representing structured knowledge, and they constitute a powerful instrument for the formulation of unified and mechanical theories of cognition. It is also indicated that, notwithstanding this usefulness, propositional models tend to underestimate the role of other forms of representations in human cognition. The authors suggest that the idea underlying the recurrent proposals of analogical constructs in cognitive science is that psychological meaning cannot arise solely from syntactic relations between amodal arbitrary symbols. Our minds afford us the possibility of having internal representations akin to the ones we construct when we interact with the external environment. From this perspective, the concreteness of the mediums we employ in cognition matters a great deal, and psychological meaning cannot be accounted for unless we seriously consider the corresponding non-abstract kinds of representations