I have been struggling for weeks to try to re-state Richard Swinburne’s argument concerning the coherence of the idea of there being a perfectly good person (from Chapter 11 of The Coherence of Theism, hereafter: COT). I think I can now at least point the way as to how to do this. The overarching argument goes like this:

1. The statement ‘There is a perfectly free and omniscient person’ is a coherent statement.

2. The statement ‘There is a perfectly free and omniscient person’ entails the statement ‘There is a perfectly good person.’

3. Whenever a coherent statement entails another statement, the entailed statement is also a coherent statement.

Therefore:

4. The statement ‘There is a perfectly good person’ is a coherent statement.

Swinburne argues for (1) in previous chapters of COT, and I will not challenge that premise here. Premise (3) seems right to me, and the logic of this argument is fine. As far as Chapter 11 is concerned, the question at issue is whether (2) is true. It is the argument for (2) that I have been struggling to re-state.

Swinburne does not spell out a definition of ‘perfectly good person’ but he states various necessary conditions, which, presumably, when taken together, amount to a sufficient condition:

A person P is a perfectly good person IF AND ONLY IF:

(a) P always does the morally best action, in circumstances where there is such an action, and

(b) P always does at least one equal best action, in circumstances where there are such actions, and

(c) P always does at least one morally good action, in circumstances where there is no morally best action and no equal best action, and

(d) P never does a morally bad action.

The entailment claim that Swinburne makes in (2) can be put in terms of a conditional statement:

2A. IF there is a perfectly free and omniscient person, THEN there is a perfectly good person.

If (2A) can be shown to be true, then (2) will be shown to be true, so long as we understand the conditional statement to mean that the antecedent entails the consequent. But because the concept of a ‘perfectly good’ person involves at least four different necessary conditions, a line of reasoning to show that (2A) is true would be rather complicated, too complicated for my taste and ability. So, I think the best approach is to use the basic strategy of mathematics and modern philosophy: analysis. We can deal with each of the necessary conditions one at a time.

Let’s focus on just the first necessary condition for being a perfectly good person:

(a) P always does the morally best action, in circumstances where there is such an action.

So, rather than try to prove (2A), I will make a more modest effort to prove the following conditional claim:

(2B) IF there is a perfectly free and omniscient person, THEN there is a person who always does the morally best action, in circumstances where there is such an action.

One can use conditional derivation to prove a conditional statement.  You start out by supposing the antecedent to be true, and then work at trying to prove the consequent to be true.

PFO1. There is a person P who is perfectly free and omniscient.  [supposition for conditional derivation]

PFO2. If there is a person P who is perfectly free and omniscient, then there is a perfectly free person P such that if P is in circumstance C and action A is the morally best action for P in C, then P knows that P is in C, P is in C, P knows that A is the morally best action for P in C, and P is able to do A in C. [this is a necessary truth that is based on the meaning of ‘omniscient’ and on the principle that ‘ought implies can’]

PFO3. There is a perfectly free person P such that if P is in circumstance C and action A is the morally best action for P in C, then P knows that P is in C, P is in C, P knows that A is the morally best action for P in C, and P is able to do A in C.  [deduced from (PFO1) and (PFO2)]

PFO4. There is a perfectly free person P such that if P is in circumstance C and action A is the morally best action for P in C, then P believes that P is in C, P is in C, P believes that A is the morally best action for P in C, and P is able to do A in C.  [deduced from  (PFO3) based on the meaning of ‘knows’]

PFO5.  If there is a perfectly free person P and P believes that P is in circumstance C, P is in C, P believes that A is the morally best action for P in C, and P is able to do A in C, then P will do A in C.   [a necessary truth based on the meaning of ‘perfectly free’ peson]

 PFO6. There is a perfectly free person P such that if P is in circumstance C, and A is the morally best action for P in C, then P will do A in C.  [deduced from (PFO4) and (PFO5)]

PFO7.  There is a person P who always does the morally best action, in circumstances where there is such an action.  [deduced from (PFO6)]

PFO8.  If there is a person P who is perfectly free and omniscient, then there is a person P who always does the morally best action, in circumstances where there is such an action. [ conditional derivation from (PFO1) through (PFO7)]

=====================

Proving Conditional Statements vs Proving Entailments

I believe my logic is OK, but my philosophy of logic is not, at least not here.

Proving a conditional statment by conditional derivation does not prove that the antecedent entails the consequent.

In conditional derivation, you may use a given (something known to be true) in order to derive the consequent from the antecedent.  But a given might only be a logically contingent factual claim.

Given: Q (a true factual claim)

Show that ‘IF P THEN Q’

1.  P    [supposition for a conditional derivation]

2.  Q  [ a given, a known fact]

3.  IF P THEN Q  [based on conditional derivation (1) through (2)]

Q could be a logically contingent factual claim, such as  ‘Obama was re-elected in 2012’.

P could be any statement you like, such as ‘The Space Needle is in Seattle”.  So, conditional derivation allows one to prove the following conditional statment:

4. IF the Space Needle is in Seattle, THEN Obama was re-elected in 2012.

Clearly, it is NOT the case that the antecedent of this statement entails the consequent. It is logically possible for the Space Needle to be in Seattle and for Obama to have failed to be re-elected in 2012.  The truth of the former claim does not guarantee the truth of the latter claim.

In my conditional derivation above, I did not use any logically contingent facts; I used only necessary truths as premises (other than the initial supposition).  However, that does not put my proof in the clear.  For one can prove a conditional statement using only necessary truths (other than the supposition of the antecedent) and still fail to establish an entailment.

Q might be a necessary truth, such as ‘All triangles have three sides’.

Show that ‘IF P, THEN Q’.

1a. P  [a supposition for conditional derivation]

2a. Q [a necessary truth]

3a. IF P, THEN Q.  [based on conditional derivation (1) through (2)]

P may, once again, be any statement you like, such as ‘Zebras have stripes’.

The above proof would thus show the following conditional statement to be true:

4a. IF zebras have stripes, THEN all triangles have three sides.

Clearly, the antecedent of (4a) does NOT entail the consequent of (4a).  The two statements are logically unrelated to each other.

I believe that the reasoning I gave above does (setting aside objections to the necessary truth claims) show that the statement ‘There is a person who is perfectly free and omniscient’  entails the statement ‘There is a perfectly good person’, but a conditional derivation of the related conditional statement is NOT sufficient to prove this, so there must be other constraints that I have observed here, but which I am not currently able to point out.