Introduction

In Chapter 5 we considered the first level in Brandom's three-tiered ISA inferentialist semantic structure, that of inference. In this chapter, this inferentialist semantics will be extended to incorporate the other two subsentential tiers, substitution and anaphora. The reason for extending an inferentialist semantics in this manner stems, in part, from the need to respond to two challenges facing the account developed thus far. We shall call these: the challenge of token repeatability and the challenge of subsentential structure.

The first challenge can be illustrated by comparing the counters used in the game as described in Part I with the counters used in an actual language, such as English. In introducing the game, the impression given was that the counter-types, the bearers of semantic content, should be thought of as equivalence classes of token counters that are similar in some physically specifiable respect, such as looking and sounding the same. In the case of English, this impression is erroneous, as lexical cotypicality is neither necessary nor sufficient for two tokenings to be occurrences of the same semantically significant counter-type. On the one hand, two tokening performances of the sentence “Mandela is a lawyer” may have different semantic content if, for example, the sentence is uttered in different contexts (e.g. a context involving Nelson Mandela, the former president of South Africa, and a context involving Mandla Mandela, his grandson and current chief of the Mvezo traditional council).