Saturday, April 15, 2006

Joe Markus on Michael Martin, again

Hello Professor Reppert,

Happy Easter! I hope you are having a pleasant Easter weekend.

If you don't mind indulging me, I'd like to make one more argument regarding Michael Martin's fall-back position in his book Atheism.

As you know, Martin argues, based on a reformulated verificationist criterion of meaning, that theism is likely factually meaningless--that is, neither true nor false. But in an effort to cover all his bases, he argues later in the book that IF it should turn out that theism is factually meaningful then it is likely that God does not exist. He attemps to support the latter claim by presenting several atheistic teleological arguments, several arguments that the concept of God is incoherent, and some arguments based upon evil.

In his own words:

"My position is something like this. Yes, "God exists" is probably meaningless. But I am not completely sure. If I am wrong, there is still no reason to believe in God (negative atheism) and good reason not to (positive atheism). So my argumentative strategy reflects to some extent my actual belief. I would believe that God does not exist if "God exists" is not meaningless (which it may not be)."

My new argument against the fall-back position is really fairly short and simple. If theism is PROBABLY neither true nor false then it is probably not false (and probably not true). If it is probably not false then any argument which suggests that it is false must be unsound or in some way less than cogent. Now Martin's position may not be strictly inconsistent. But I do submit that there is a tension here. I believe that his arguments for the factual meaninglessness of theism tend to counter his arguments that it is false. Insofar as he believes his argument for the factual meaninglessness of theism is cogent, he must believe his arguments for the falsity of theism are less than cogent.

Consider an example. Suppose I offer an argument that the activity of photons are probably neither moral nor immoral. Then later I offer an argument that their activity is immoral. I could say that my position is not inconsistent because I have only claimed that it is PROBABLE that moral concepts don't apply to photons. To allow myself a fall-back position, I might offer an argument that IF moral concepts do happen to apply to photons then their activity is immoral---offering an argument for the latter claim. Doesn't my argument that moral concepts do not apply to electrons tend to cast doubt on the cogency of my argument that electrons are immoral?

It looks like the core of my argument involves something like this:

"It is probable that X is neither A nor B" ENTAILS "It is probable that X is not B."

That principle seems true. For example, if it's probable that (A) "Joe is neither short nor bald" then it is probable that (B) "Joe is not bald".

If someone argues that A is probably true, how can they later argue that B is probably false without giving up their argument for A?

It seems to work with this sort of example but I'm not sure it would work in the case of Martin's argument. Would it work with claims about the application of the concepts of truth and falsity? Maybe.

Anyway, that's the gist of my argument.

I'd appreciate any comments. Thanks for your time.

Regards,
Joe

15 comments:

Hiero5ant said...

"If someone argues that A is probably true, how can they later argue that B is probably false without giving up their argument for A?"

This one is trivially easy.

Martin is A)90% confident that theological noncognitivism is true, B)9% confident that atheism is true, and C)1% confident that theism is true.

There's no "tension" in his argument, unless you're prepared to claim that there's a "tension" in the fact that 9 is a smaller number than 90 but a larger number than 1.

Mr. Grouchypants said...

If Martin is 90% certain that theism is neither true nor false, then wouldn't he be 90% certain that you cannot argue that theism is false?

Hiero5ant said...

Tautologically, yes.

Mr. Grouchypants said...

If you were 90% certain that you couldn't argue that theism is false, why would you present an argument that theism is false?

Joe Markus said...

You'd think he'd be committed to believing that there is a 90 percent probability that any arguments for the falsity of theism aren't cogent.

Or at least he'd have to believe that there is only a 9 percent chance that his arguments for the falsity of theism are cogent. That's kinda weird since he seems to express quite a bit of confidence in his arguments from evil.

Anyway, note that the confidence he has in the factual meaninglessness of theism isn't anywhere near 90 percent. He implies that he just leans that way.

Hiero5ant said...

If you were 90% certain that you couldn't argue that theism is false, why would you present an argument that theism is false?

In case your first argument turns out to be wrong? Really, this is elementary logical disjunction; there's no "gotcha" here.


Or at least he'd have to believe that there is only a 9 percent chance that his arguments for the falsity of theism are cogent. That's kinda weird since he seems to express quite a bit of confidence in his arguments from evil.


Given that noncognitivism is false, then his confidence is 90%. This isn't a particularly arcane point.

Of course, you're right that these numbers are chosen just for purposes of illustration. I have no idea what the exact confidence he places in each one is, just their relative plausibility. Note that his meaninglessness thesis needn't even be 51% plus; his position would be unaffected by a 45/30/25 breakdown too. Note also that as a matter of logical possibility, one could argue that it's 99% plausible that god-talk is meaningless, but that if it's meaningful then theism is 99% likely to be true. Martin just so happens to evaluate its plausibility differently.

Joe Markus said...

hiero5ant said:

[b]his position would be unaffected by a 45/30/25 breakdown too[/b]

This wouldn't represent his actual position. Martin himself states that he leans toward the belief that theism is factually meaningless. That wouldn't be the case with your 45/30/25 breakdown--unless a person can believe something which they think is less probable than not. Which wouldn't make much sense.

The question is: what does Martin believe on the basis of his arguments? There are two sets of arguments in tension. If he really believes in the cogency of his arguments for the factual meaninglessness of theism then he can't believe in the cogency of his arguments that theism is false.

One set of arguments purportedly supports the conclusion that theism ISN'T false and the other set purportedly supports the conclusion that it IS false.

He can rationally accept only one set of arguments as cogent. He should pick one and stick with it.

Adding the conditional statement that "if it's factually meaningful then it's false" doesn't solve the problem. If he accepts the cogency of his arguments that theism is factually meaningless then he must believe the antecedent of the conditional is unlikely and thus must believe that it's unlikely that theism is false. Again, that would mean that he can't accept the cogency of his arguments from evil, incoherence, etc.

Hiero5ant said...

I'm really at a loss as to explain any more plainly how the simple English conjunction "if" operates.

I am quite confident that I am capable of driving across town without getting into a headon collision. Who on earth would say that "If you accept the cogency of your arguments that you are a safe driver then you must believe the antecedent of the conditional is unlikely and thus must believe that it's unlikely that wearing a seatbelt is necessary"?

He believes the arguments for noncognitivism are cogent, for atheism less cogent, and for theism still less cogent. If you're going to dig in your heels on this and assert that there's a "tension" in saying that A is more plausible than B, but B is more plausible than C, then I can't help you.

Joe Markus said...

Hiero5ant,

You said:

He believes the arguments for noncognitivism are cogent, for atheism less cogent, and for theism still less cogent

Exactly. Since he believes that it is more probable than not that noncognitivism is true he must believe it is less probable than not that atheism is true. So he must believe it's not likely that his arguments for atheism yield their conclusion.

Throwing the word "if" around doesn't change anything as my previous post points out.

Joe Markus said...

One possible way to salvage Martin's argument is to claim that though HE doesn't think his arguments for positive atheism yield their conclusion, the theist should believe they do.

He would be saying something like this: "I personally believe these arguments don't work. However, other people might find that they have force and yield their conclusion." In other words, he'd be saying that if you think theism can be true or false then you should believe it is false on the basis of his arguments for positive atheism. Though he can't think his arguments yield their conclusion.

He suggests something like this when he says:

it is not clear to me why I could not have a strong belief that sentences about God are meaningless and yet for the sake of the argument suppose they are meaningful. We can believe strongly that a statement is false yet, for the sake of the argument, suppose it is true. It is unclear why a similar hypothetical position cannot be entertained with respect to meaningless sentences. Suppose we read in a faculty handbook that all members of the faculty will wear academic gowns at graduation. One might believe strongly that this is a command and reject it as being inappropriate. However, since commands lack truth value they are meaningless in the sense of lacking truth value. But one might also suppose, for the sake of the argument that what one has read is a prediction and, on this supposition, reject it as being false.

But his graduation example isn't parallel to his noncognitivism. In the case of the graduation, you have a statement which CAN be interpreted as BOTH factually meaningful and factually meaningless. But in the case of theism he can't believe that there is an interpretation of the statement "God exists." that is factually meaningful. That would give away his entire argument.

To be parallel, he would have to say that the statement "all members of the faculty will wear academic gowns at graduation" doesn't have an interpretation that is factually meaningful. If that's the case he couldn't treat it as hypothetically true.

It's difficult to see how I could formulate arguments against a statement that I believe couldn't be true or false. I can only formulate arguments that support the statement AS A PREDICTION because it is possible to interpret it that way..ie. there is a sense in which it IS factually meaningful.

By arguing so passionately against theism it seems that he implicitly admits that it has factual meaning.

Jason said...

(note: still recovering from lastest bout of pain-effect; will consider posting more extensive comments later, elsewhere, hopefully.)


Joe's most recent comment seems to be a restatement of his criticism from his original letter to Victor (some time ago); and it's the criticism I still stand with myself. (Fwiw, I _think_ I'm prepared to agree with Hiero5ant's counter-criticism of Joe's more recent attempt, though.)

Put shortly, one cannot make arguments about the falsity of a proposition without recognizing a meaning to the proposition to make arguments _about_. And once one does that, it is inconsistent to try to claim elsewhere that the proposition is (even probably) meaningless.


I'll illustrate with an example having nothing specifically to do with theism or atheism.

Consider the following:

yuqowç hfu yihhur mua,
lfuru paw I ko, I favu woh asruaçí kowu



Let us suppose I proceed to argue that we ought to consider the meaning of the claim represented by this to be false; for instance, because the way I have used the word 'I' involves a contradiction between the meaning implied by its relation to the phrase "lfuru paw I ko", and the meaning required by its relation to the phrase "I favu woh". Specifically, when I say "I favu woh" ('I' meaning the plural of cat), I am refuting the claim being at least implied when I say "lfuru paw I ko"; which meaning cannot merely be optional (even though arguably unstated in the phrase), because the phrase "yuqowç hfu yihhur mua" sets up a condition that can only be satisfied by reading "lfuru paw I ko" in x-particular way instead of perhaps in another (even though otherwise that might technically be possible, I allow.)

Let us suppose I also add, that I am xish percent sure that what is represented by the words up there in bold font, is meaningless.


Why would I bother to make this additional claim about the material in bold up there at all?

And how could it possibly improve matters if I _started_ with my contention of probable meaninglessness and _then_ went on to draw a conclusion (even if only provisionally) which has to be based on recognizing and applying specific meanings?


If there's a legitimate resolution between the two claims (probably meaningless, but otherwise false because of these meanings), it'll have to come from Martin's meaning when he claims 'meaninglessness'.


This would be a good time, if it hasn't been done already, for someone on one or another side of the debate to go into some _thorough_ detail about what Martin says he means by meaninglessness (in this particular case and/or generally.) If this has been done already and I simply missed it (which is frankly very possible; still not up to speed yet {s}), the discussion should be focusing on his meaning for "meaningless".


Joe's most recent comment provides some data for this (if not thoroughly so); in this case it appears that Martin's "strong belief" about the factual meaninglessness of "all members of the faculty will wear academic gowns at graduation", involves an extremely limited notion of meaninglessness: if a command, then not a truth claim; if not a truth claim then neither true nor false; if neither true nor false, then factually meaningless.

How Martin supposes that this analogy can be properly illustrative of a strong belief in the meaninglessness of "sentences about God" per se (which one would suppose would include even little short sentences like "God exists", which is a brief version of the claim of theism), seems sloppy to me, at best. The whole point to claiming factual meaninglessness about sentences about God is to nerf discussion a priori about truth claims (such as "God exists") relevant to theism vs. atheism. I suppose it's true that if "God exists" can somehow be read to mean a command, _then_ it would (as such) count as being factually meaningless in line with his analogy; but is _this_ the sort of thing he is really claiming in regard to "sentences about God" across the board??


Anyway, that's all for now; hope to be back later,

Jason

(PS: btw, the sentence does actually have a meaning--it's part of the language I invented for the Dedications of my novels, and is the first phrase in the Dedication for _Cry of Justice_, which is currently slated to street mid-summer or possibly early fall. And it's a question, not a statement. {s})

Joe Markus said...

Jason,

Thanks for the comments.

You said:

Joe's most recent comment seems to be a restatement of his criticism from his original letter to Victor (some time ago); and it's the criticism I still stand with myself.

Actually, I consider this argument to be significantly different from my previous criticism. This time around I'm suggesting that Martin has two sets of arguments which tend to counter each other. My previous argument was based on the notion that if you believe a claim is merely likely then you must admit that its denial is possible and that if theism is possibly factually meaningful then this could only be if there is in fact a statement of it which is meaningful.

You also said:

Put shortly, one cannot make arguments about the falsity of a proposition without recognizing a meaning to the proposition to make arguments _about_.

I think this is basically correct. This is my point at the close of my last post. When a person formulates an argument they are suggesting that some group of statements support another statement or statements. It seems that in order to do this one must implicitly accept that all the statements have some factual meaning---including the conclusion. Can a person formulate an argument to support a command?

Jason said...

Joe,

Sorry for the confusion; by "last comment" I meant your previous post in this comment thread: yes, I understand the new criticism in the letter to Victor recently is a somewhat different attempt. {s} I would hardly have said I'm more-or-less prepared to agree with Hiero's counter-criticism on it, so far anyway, if I didn't recognize the distinction.

To clarify myself a bit further on what I mean by agreeing (more or less) with Hiero: I think he's right that under _some_ (perhaps even many) circumstances, it's okay to have provisional arguments in case a primary argument happens not to obtain after all. And, if Martin (by his further argument(s)) intends to include (even if only implicitly) that any further argument of his own is just as probably (not certainly) meaningless as any other sentence (or claim or argument or whatever) about God (even if it happens not to _feel_ that meaningless to him or other people); then I'm prepared to agree that's consistent (as far as it goes)--and I suspect this was part of what Hiero was meaning in his various replies as well.

Of course, I also agree with _you_ that any argument for the factual meaninglessness of theism tends (in proportion) to counter any further arguments of his own that theism is false. (But then, MM and Hiero might just as easily agree with that, and for all I know they do.) But since MM wasn't making a claim of certainty of meaninglessness, then his _probability_ (per se) of meaninglessness can't (of itself) be called in against him when he goes on to attempt some arguments which obviously require him to have nominal meanings in mind. (Which, I imagine, is probably why he thought he could legitimately do it in the first place.)

That being said, I also (obviously) agree with you that there's still a serious tension between the two claims; and I think it comes down to what you originally criticised him on. And by and large, I think what power there is in your new criticism is being borrowed from your prior criticism.

Jason

Joe Markus said...

Jason,

You said:

To clarify myself a bit further on what I mean by agreeing (more or less) with Hiero: I think he's right that under _some_ (perhaps even many) circumstances, it's okay to have provisional arguments in case a primary argument happens not to obtain after all.

I agree that this is a sound approach IN GENERAL. I don't see a problem with offering an argument for a conclusion and then offering several more arguments just in case your primary argument doesn't work. But your arguments shouldn't clash.

The problem with Martin's strategy is that his primary argument tends to undermine his later arguments. He can't believe in the cogency of both sets of arguments. If he accepts the cogency of his initial argument then he can't reasonably believe that his later arguments yield their conclusion.

If someone gives an argument that "Joe Markus is neither short nor bald." and really believes that the argument works, then he can't believe in the cogency of any argument that suggests "Joe Markus is bald." Sure, he doesn't contradict himself if he's dealing with probabilities. But his arguments do undermine each other.

As Schopenhauer once said, "premises are not taxi cabs: you cannot ride them only as far as you want to go and then get off". The premise in this case is the claim that theism is factually meaningless.

Joe Markus said...

One more point that should highlight the tension in Martin's position.

Consider his graduation example again--- (P) All members of the faculty will wear academic gowns at graduation.

If Martin's fallback position is reasonable then he could offer an argument that P is factually meaningless AND THEN offer additional arguments that P is false IF it should turn out that it is factually meaningful. But note that one could only offer arguments that it is false if one already believes there is a sense in which IT IS FACTUALLY MEANINGFUL----ie. it's a prediction.

How could you possibly formulate an argument if you didn't think there is at least SOME SENSE in which it is factually meaningful?

What would connect the premises to the conclusion if there is no sense in which the conclusion is factually meaningful?

Consider another example. I think it's likely that "Twas brillig in the slithy toves" is factually meaningless. But I might be wrong. And just to cover the possibility that it is factually meaningful, I'm going to argue that it's false.

(P1) It was overcast that day.
(P2) It was the middle of summer.
(P3) There were no bears around.
(C) Twas brillig in the slithy toves.

How can I argue that the premises support the conclusion if I don't believe there is at least SOME sense in which the conclusion is factually meaningful?