The Factivity Constraint

Abstract

Memory implies truth. In the case of fact memory this is obvious. “S remembers that Fido was on the sofa” implies that Fido was the sofa. And generally, “S remembers that p” entails that p is the case. Object-, property-, and event memory are also factive since I cannot remember Fido’s flap-ear and his sitting on the sofa without Fido having a flap-ear and without it being the case that he sat on the sofa. “I remember Fido” implies that there is something (Fido) that I remember. Yet in what follows I will focus on the factivity of fact memory (propositional memory). Sections 8.1–8.3 motivate and explain the thesis that propositional remembering is factive. Many philosophers hold that one of the consequences of the factivity constraint on memory is that memory is supposed to work like a photocopier producing duplicates of past contents. On the xerox model of memory the recalled thought must be type-identical with the original thought. The content of a reproductive memory (and the embedded content of a meta-representational memory content, respectively) must be a token of the same type of content as the original thoughts from which it causally derives. Section 8.4, characterizes the xerox model and gives examples of its employment. Section 8.5 argues that the xerox model flies in the face of our actual practice of attributing memory. Frequently, our memory not only stores but also processes the encoded information. Three such information processes are discussed: cognitive dynamics, condensation and schematic processing. Finally section 8.6 argues that the fact that our memory not only retains but also processes the incoming information should not be regarded as an abnormal lapse of an otherwise reliable cognitive faculty, but as part of the very function of memory.