VazScep wrote:I should mention that truth-tables aren't in the formal semantics of propositional logic. They can't be, because in general, there can be infinitely many atoms, requiring infinitely large truth-tables.
Relatively speaking, a definition of the semantics for propositional logic requires little ingenuity, and as with everything in maths, there are many equivalent and perfectly sensible definitions. Off the top of my head:
A propositional alphabet is a (possibly infinite) set of symbols. Given an alphabet A, the propositional language L(A) consists of:
1) The elements of A.
2) ~p and p → q for any p and q in L(A).
An interpretation of L(A) is a function e from A to {0,1}, which is extended to a function e' from L(A) to {0,1} according to the rule that e'(~p) = 1 - e'(p) and e'(p → q) = 1 when p <= q and 0 otherwise.
We now talk stuff about the set of all functions e from A to {0,1} and their extensions e'.
What's your beef with this?
I have no beef with L(A). L(A) is fine all by itself. But where to next? Sets of sentences I don't think are part of L(A). Nor are arguments or validity or entailment. But it's all a choice I guess.