Keelefilosoofia raamat
. . " what?
One obvious strategy is to invoke a domain of helpful entities, such as
propositions(!), and write truth rules for non-truth-functional expressions
in terms of quantification over that domain. (As we have seen, to deal with
some adverbs, Davidson introduced a domain of "events," and made adverbs
into adjectival predicates of events.) The main problem with that strategy
is that it strains syntax, since the transformations have to work harder to
transform the new, zany logical forms into familiar English; as Blackburn
points out (1984: 289), a Davidsonian treatment of a non-truth-functional
(and intensional) construction such as "because" or "believes that" requires
at least a "heavy commitment to concealed logical forms." (But, as before,
belief sentences in particular are already an ugly problem for any theory of
meaning.)
Objection 6
Truth-conditional semantics must fully disclose the general notion of truth it
is presupposing
annaabi.ee 
