Originally posted by DeepThought
Thanks. The book I read on logic was EJ Lemmon (on loan and given back so I don't have access to it) so I'll follow his notation but give the dependencies on the right as typesetting on RHP forums isn't the easiest. Also he doesn't distinguish between assumptions and givens which I like to do.
M.P. = modus ponens
&I = conditional inclusion ( ...[text shortened]... milar going on with the problem I have with conditional proof, which is what I'm confused about.
Great, let’s take the first case. From P>Q we want to derive (P&R)>Q. To the right of each line, in parentheses, I’ve included their justifications and line dependencies.
1. P>Q (Premise, 1)
2. P&R (Assumption, 2)
3. P (Conjunction Elimination, 2)
4. Q (Modus Ponens, 1,2)
5. (P&R)>Q (Conditional Proof, 1)
Just after stating our premise the proof begins, as conditional proofs invariably do, with the assumption of the antecedent of the conditional we aim to derive. When we get Q on line (4), we’ve done so using both lines (1) and (2). But when we “conditionalize out” in our conclusion, we thereby discharge the assumption in line (2). So, our conclusion rests solely on line (1). So, our conclusion follows from our sole given premise.
Contrast this with the attempt to derive (PvR)>Q :
1. P>Q (Premise, 1)
2. PvR (Assumption, 2)
3. P (Assumption, 2,3)
4. Q (Modus Ponens, 1,3)
5. R (Assumption, 2,4)
6. ????
As above, after stating our premise we begin with the assumption of the antecedent of the conditional we aim to derive. In this case, we’re assuming the disjunction (PvR). Disjunctions are true just in case at least one of their disjuncts is true. So, to derive something from a disjunction we have to show, in effect, that given our premise set our desired conclusion follows independently from each disjunct. In this case, that means we have to show
both that Q follows from P>Q together with the assumption P,
and also that Q follows from P>Q together with the assumption R. The problem, though, is that our premise allows us to derive Q from the assumption P, but neither it nor anything else allows us to derive Q from the assumption R.
In your proof, line (5) is invalid. You haven’t derived Q from (PvR), so you can’t “conditionalize out” and discharge the assumption upon which (PvR) is based. Line (5) is still predicated on the assumption in (2) and hence doesn’t follow merely from our premise.