THINGS AND THEIR ASPECTS

Attitude contexts pose well-known problems for truth conditional semantics, in particular rendering problematic inferences involving the substitution of coreferential proper names. But other, apparently extensional constructions, as Jenny Saul (1997a,b, 1999) has noticed, exhibit a similar behavior. (a) Superman always gets more dates than Clark Kent does. (b) ?? Superman always gets more dates than Superman does. (a) Clark Kent went into the phone booth, but Superman came out. (b) Clark Kent went into the phone booth, but Clark Kent came out. (a) Chris hit Clark Kent, but he never hit Superman. (b)?? Chris hit Clark Kent, but he never hit Clark Kent. (a) Lois slept with Superman before she slept with Clark Kent. (b) Lois slept with Superman before she slept with Superman. The (a) versions of the above examples all seem satisfiable, perhaps even true, and to convey nontrivial information. They differ intuitively from their substitutional variants in (b), which seem necessarily false. These intuitions about the truth conditional status of the (a) examples versus the (b) examples imply a semantic rather than pragmatic difference; that is, there is intuitivelya difference in the truth conditional status of the (a) and (b) variants, which should follow from the semantic interpretation of their logical forms.\footnote{There may be in addition differences in implicatures or felicity and infelicity conditions that would naturally demand a pragmatic explanation, but I won't consider these here.} A Fregean, semantic approach according to which the (a) versions are true and informative but their (b) counterparts are not appears promising. But it is challenging to apply the standard Fregean machinery to these simple sentences. On a Fregean semantic analysis, it is the presence of an attitude or modal operator that forces expressions to take their sense as their semantic value rather than their ordinary, ''extensional'' semantic value (their reference). But what operator or construction triggers the Fregean machinery of non customary semantic values in the examples above? We don't want to stipulate that all apparently extensional constructions are in fact intensional! To answer that question, consider the following, natural paraphrases of the (a) sentences using 'as' phrases: Superman as Superman always gets more dates than Superman as Clark Kent does. Clark Kent as Clark Kent went into the phone booth, but Clark Kent as Superman came out. Chris hit Superman as Clark Kent, but he never hit Superman as Superman. Lois slept with Superman as Superman before she …