Evidentialism and Reformed Epistemology

Read 6033 times


  • ****
  • 6304 Posts
    • View Profile
Jamin Hubner's blogging on logic
« on: December 27, 2010, 10:52:04 am »
Jamin Hubner has recently been posting some blogs, a series which he dubs Lessons in Logic and Argumentation.  I've noticed lots of mistakes in them, and though I tried to talk to him about it directly, he seemed completely disinterested at the time in finding and correcting any errors he had made, perhaps because he was busy at the time.  So I'm posting my criticisms here, in order that he may come read them at his convenience, and respond if he likes.

For now I will restrict my attention to his Dec 20 post on types of arguments.   There he delineates four distinct kinds of inference:  abduction, induction, deduction and transcendence.  Now, I should note that abduction is by no means universally regarded as a distinct form of inference from induction and deduction, as we find denied by Hans Reichenbach, Harry G. Frankfurt, Justus Buchler and Richard A. Fumerton, to name just a few scholars.  However, some scholars do indeed accept abduction as distinct, so Mr. Hubner's classification of abduction has some precedent, even though I personally disagree with it.  I just want to point out that this is a somewhat controversial matter, and in my opinion deserves mentioning.

His differentiation of transcendental arguments from other forms of argument is considerably less founded.  I admit that I am by no means an expert on Kant or transcendental arguments, but I can find not a single scholar who refers to transcendental argumentation as anything but a special kind of deduction and/or rhetoric.  Kant himself referred to his method as (translated into English) "the transcendental deduction." It seems at best extraordinarily misleading, then, to claim that there is a fourth distinct form of argument, and to give it the label 'transcendental.'

To be fair, many of Hubner's errors are actually inherited from Old Testament researcher Don Collett, who wrote an online paper on transcendental arguments which seems to be the source of much of Mr. Hubner's misinformation on the subject.  In this way, Hubner has repeated a number of rather bad mistakes originally made by Collett.  In fact, I have recently posted my own blog entry on Collett's paper, where I go over Collett's two most significant (in my opinion) errors.

Hubner attempts to show what is his idea of a transcendental argument by drawing from Bas van Fraassen's presuppositional semantics from his 1968 paper on the subject.  Quoting van Fraassen without a citation, he writes:

A presupposes B if and only if:
1. if A is true, then B is true.
2. if ~A is true, then B is true.

Unfortunately, Hubner doesn't give any background on the subject, and so there is no way for us to make sense of this characterization unless we happen to have read van Fraassen's paper (which, again, was not cited).  Interpreted classically, this account of 'A presupposes B' is equivalent to, simply, 'B'.  This is easily seen:  for if 1. and 2. are both true, then by the contrapositive of each we have that ~B is true only if A and ~A are both true; but A and ~A are never both true, so again by contrapositive we have that B is true.  The converse, of course, follows immediately, and indeed 'A presupposes B' is equivalent to 'B'.

Hubner gives the following example of a "transcendental" argument:

1. The existence of creation presupposes that God (the Creator) exists.
2. Creation exists.
3. Therefore, God exists.

Recall that by the classical interpretation outlined above, 'creation exists presupposes God exists' is equivalent to 'God exists'.  So we can rewrite this argument as:

1. God exists.
2. Creation exists.
3. Therefore, God exists.

Obviously, we have three immediate problems with this:  First, the argument begs the question, since it has its conclusion as a premise.  Second, it is needlessly lengthy, since we can dispense with 2. and still have our conclusion follow from the premises.  Third, it is deductively valid, which means it is not distinct from deduction.

But there is a fourth problem which is apparent to those who have read van Fraassen:  the characterization of the presupposition relation which Hubner quotes is not intended to be interpreted classically!  For van Fraassen rejected the law of bivalence in his logical system, and so we cannot obtain, for instance, the contrapositive '~B is true only if ~A is true' from 'A is true only if B is true.'  For on van Fraassen's semantics, it could be that A is neither true nor false.

Recall Hubner's "transcendental" argument:

1. The existence of creation presupposes that God (the Creator) exists.
2. Creation exists.
3. Therefore, God exists.

By property 1. of the presupposition relation, premise 1. of the argument gives us that 'creation exists' is true only if 'God exists' is true.  Premise 2. gives us that 'creation exists' is true.  By modus ponens, 'God exists' is true, which is to say that we conclude God exists.  In this way, Hubner's argument is deductively valid according to the logical system afforded by van Fraassen's presuppositional semantics.

Unfortunately, Hubner has attempted to show that transcendental arguments are NOT deductive arguments.  Yet since we have seen that his transcendental argument example is deductive, he must have made a mistake somewhere.  Either he has given us an argument which is not transcendental, or else transcendental arguments are sometimes deductive (here I refer to his understanding of transcendental arguments, which appears to me dramatically different from scholarly treatments).  In either case, he needs to correct his error.

Finally, Hubner claims that, of the four types of arguments he discusses, we can order them from weakest to strongest:  abductive, inductive, deductive and transcendental.  I have no idea what he intends to communicate, here, and I suspect he doesn't have a very good idea of it either.  Abductive arguments are by no means universally weaker than non-abductive arguments, for example.  For another, inductive arguments can be (and in my experience usually are) a lot more persuasive than deductive arguments.  And far from being the strongest, transcendental arguments, at least in their scholarly treatment, are often highly controversial and dubious.  So, if he really does have some well-thought-out idea in mind in ordering argument types from weakest to strongest, he needs to clarify what it is.

I could go on, but this post is getting to be rather lengthy.  What I have outlined above are, I think, the most serious of his errors.  I wish him luck in making his corrections!