Contrast Arguments

One of the weapons in the arsenal of Ordinary Language and other philosophers is the contrast argument. Such arguments are used to show the meaninglessness of certain terms, typically, the terms we metaphysicians like to bandy about. One type of contrast argument has the form:

1. If a term T is meaningful, then there are items to which T does not apply.
2. There are no items to which T does not apply.
Ergo
3. T is not meaningful.

The form of this argument is Modus Ponens, so there is no question of its validity. The only question is whether (1) is true or has some such lesser epistemic virtue as plausibility or rational supportability. I am assuming that T is nonvacuous: it does apply to something.

The idea behind contrast arguments is plausible: if a term applies to everything, or everything in a specified domain, then there is a 'failure of contrast' that drains the term of all meaning. But I will argue that contrast arguments are not probative.

Suppose a Heideggerian strolls up to you in a shopping mall and intones: Das Seiende ist! Aber was ist das Sein des Seienden? (Beings are. But what is the Being of beings?) He is asking about that which makes beings be, that in virtue of which that-which-exists exists.

Since you are a ground-hugging anti-metaphysical type, and want to blow this portentous Schwarzwaldontologe, Black Forest ontologist, out of the water, you reach into your O.L. bag and pull out a contrast argument:

1. If a term T is meaningful, then there are items to which T does not apply.
4. There is nothing to which 'Being' does not apply.
Ergo
5. 'Being' is not meaningful.

From this you infer that inquiries into Being and existence are predicated on a mistake. 'Being' has meaning only if there are items that lack Being; but nothing lacks Being, hence 'Being' is meaningless. (Sidney Hook championed a version of this argument back in the '60s in a book entitled The Quest for Being.)

Although you may like the conclusion of the foregoing argument, try running the argument with 'physical object' substituted for 'Being.' The conclusion would be

6. 'Physical object' is not a meaningful term.

But this ought to strike you as highly counterintuitive. If you maintain that everything is a physical object, are you maintaining something meaningless? This version of materialism may be false, but is it meaningless? Or try 'Either a physical object or supervenient upon a physical object.' Or 'natural' in the sense of not supernatural. A naturalist who maintains that everything is natural and nothing is supernatural is presumably not maintaining something meaningless. False perhaps, but not meaningless.

One sort of nominalist maintains that everything that exists is a particular, an unrepeatable entity, and that there are no universals, repeatable entities.  Surely it would be feeble to argue that this brand of nominalism is false because of a failure of contrast.  "There must be nonparticulars because if there weren't any then the claim that every entity is a particular would be meaningless!"

According to Ayn Rand, "It is Aristotle who identified the fact that only concretes exist." (Introduction to Objectivist Epistemology, 2nd ed., p. 52)  One could not plausibly  argue that it is not that case that every entity is a concretum by urging that the existence of abtracta is necessary for the meaningfulness of the claim being forwarded.

Everything is self-identical and nothing is self-diverse.  And this is necessarily so.  Hence necessarily there is nothing to which 'self-identical' does not apply.  This fact, however, does not render 'self-identical' meaningless.  Since it is true that everything is self-identical, it follow that it is meaningful.

Harry Binswanger gives the following contrast argument: "But to identify something as 'an experience in my mind,' I have to contrast my mental experiences with something else. Without the contrast between the internal and the external, 'internal' loses its meaning."  That is, 'internal to consciousness' has  meaning only by contrast with 'external to consciousness,' so that if nothing is external to consciousness, then the claim that every object is internal to consciousness is meaningless.

But this contrast argument is no more compelling than the other ones cited above.  My point, of course, is not that there is nothing transcendent of consciousness, but that Binswanger's attempt to prove that there is is feeble and uncompelling.

My interim conclusion: contrast arguments are unsound. Does anyone care to defend them?

Comments

[Eric]

Hello Bill

I wonder if the problem might be with two ways of thinking about how terms apply to everything. We can say that a term applies to everything no matter how we concieve everything (which seems to me to be the sense in which the first premise of the contrast argument is to be understood), or we can say that if everything is so and so, then so and so applies to everything (which seems to be the case with your use of the first premise of the contrast argument in each of your counterexamples).

[Boram Lee]

Bill, I'd like to put 3 suggestions on the table.

[1]
Here's a modally modified version of the contrast argument:

1. If a term T is meaningful, then it is POSSIBLE that there are items to which T does not apply.
2. It is IMPOSSIBLE that there are items to which T does not apply.
Ergo
3. T is not meaningful.

As to the grade of modality, I suggest metaphysical possibility (rather than physical or logical). Traditionally philosophers have assumed an inferrable connection between conceivability and metaphysical possibility, and between inconceivability and metaphysical impossibility. So let's adopt that assumption.

What about the meaningfulness in question? Let's distinguish, as Frege has done, between sense (sinn) and reference (bedeutung). The sense of T is the descriptive content that allows us to pick out the reference or extension of T. Let's say, then, that the meaning in question is sense, not reference.

So the modally modified contrast argument would assume this: the sense of a term T is given by the set of metaphysically possible worlds in which term T applies to some objects and not others. If the set is null, T is senseless.

So even if the extension of T in the actual world includes everything, T could still have meaning (i.e., sense).

This modified version of the contrast argument wouldn't help Binswanger, of course. Even if the actual world turns out to be one where everything is internal to consciousness, there are metaphysically possible worlds where some things lie outside of consciousness, and these give sense to the claim that everything is internal to consciousness. The modified argument would also allow the claim "everything is physical or supervenient on the physical" to have sense, in a similar way.

[2]
As for "being", well, it seems that there are no metaphysically possible worlds where some items have being and some not. So it seems that "There is a table" would be senseless even according to the modified argument. But here one may dispute the Quinean claim that "Everything exists" is trivially true. The trick is to distinguish quantification claims from existence claims. If they are distinguished, it is possible to conceive that some things exist and some things don't. E.g., there is one talking lion called Aslan in C. S. Lewis's novels, and it doesn't exist.

[3]
'self-identical' seems to be a predicate that is true of everything. But some philosophers have disputed that, e.g., Wittgenstein (Tractatus 5.53ff).

I admit though that the case of being and self-identity poses the greatest challenges to the contrast argument. One would have to modify classical logic to accommodate the contrast argument, and that might be too heavy a price to pay.

[Matt Hart]

My problem is that I just don't see the appeal of (1) in either Bill's or Boram's version.

Bill writes, "if a term applies to everything, or everything in a specified domain, then there is a 'failure of contrast' that drains the term of all meaning."

But I just don't find this intuitive. I mean, if a term applied to anything and everything because it was somehow contentless or something like that, then sure, that's a dud. But if it applies to everything and is nevertheless informative, I just don't see the problem.

Matt.

[Bill Vallicella]

Gentlemen,

I hope to respond this evening. Matt, I hope you gathered that I am arguing against contrast arguments.

[bobkoepp]

I hope it will be excused if I make a point that might be tangential to questions about the probative value of contrast arguments. I'm more familiar with the notion that explanations must involve a contrast; i.e., must explain why x rather than non-x occurs or is the case. While explanations typically do have the form of arguments, the "contrast class" functions somewhat differently than in the sort of argument Bill is critiquing. There's a difference between a predicate applying to every element in a domain of discourse and an explanation that "explains" every possible occurrence in a domain.

[Matt Hart]

Bill,

Yes I did gather that! But the way you approached the question seemed to suggest that, absent counterexamples such as your own (which I agree are forceful), (1) would get some intuitive, and therefore probative, support.

That's what I don't see, for the reasons suggested in my above post.

Matt.

[Bill Vallicella]

Matt,

I am in the somewhat difficult position of having to explain why a worthless argument-strategy is worth refuting. It is worth refuting because distinguished philosophers have employed it. Sidney Hook is one example, J L Austin is another. I will given an Austinian example in a separate post.

Eric,

I am not sure what you mean.

Bob,

Contrastive explanations are indeed a separate topic.

[Bill Vallicella]

Boram,

Excellent comments! We should indeed distinguish nonmodal and modal contrast arguments. Your modal contrast argument is stronger than the nonmodal one I cited:
1. If a term T is meaningful, then it is POSSIBLE that there are items to which T does not apply.
2. It is IMPOSSIBLE that there are items to which T does not apply.
Ergo
3. T is not meaningful.

If Meinong is right and some objects are nonexistent objects, then existence is classificatory: it divides objects into the existent ones and the nonexistent ones. But then contrast arguments are irrelevant.

But we get a clear counterexample if we let T = 'has properties.' It is a broadly-logical (metaphysical) necessity that, for any x, x has properties. And it is clearly meaningful to say of any x that it has properties. If the above argument form were valid, however, 'x has properties' would not be meaningful. So the argument form is invalid.

It cannot be denied that there is the predicate ' . . . is self-identical.' Whether it expresses a property is a further question. Wittgenstein can be interpreted as saying that it is unnecessary to say of a thing that it is self-identical. But what it is unnecessary to say and what it is meaningless to say are two different things.

My conclusion: both nonmodal and modal contrast arguments are defective.

[Harry Binswanger]

Bill, et al.,

I don't think every concept has to have a contrast. You've put your finger on those special concepts that don't: "existence" and "identity."

And, no doubt to your chagrin, Rand made that very point:

Since axiomatic concepts are not formed by differentiating one group of existents from others, but represent an integration of all existents, they have no Conceptual Common Denominator with anything else. They have no contraries, no alternatives. The contrary of the concept "table"--a non-table--is every other kind of existent. The contrary of the concept "man"--a non-man--is every other kind of existent. "Existence," "identity" and "consciousness" have no contraries--only a void.

(I will post later on whether this passage implies I'm departing from Rand in the argument that the proposition "Everything is inside consciousness" is incoherent on these grounds.)

But every non-axiomatic concept does make a differentiation, and if you deprive it of that differentiation, it becomes meaningless (in both senses of meaningless).

BTW, it is not the proposition but the concept that has to have a contrast.

"Physical object" has several contrasts--things in the other Aristotelian categories (non-ousiai) and mental objects. So I don't see why "physical object" was even used as an example.

Ayn Rand identified the "fallacy of the 'stolen concept'" in the early 50s, and it is a little wider and stronger than the contrast argument. It is based on the idea that concepts are not "givens" not just "there" but are products of conscious activity. And concepts are not a separable product, like a car that comes out of the car factory, but an outgrowth--like a hand--that remains dependent on the activities of consciousness required to sustain it. For concepts to remain concepts, rather than meaningless sounds, one can't take the Wittgensteinian approach of pulling up the ladder from under oneself. The "stolen concept" fallacy occurs when one "makes off" with a concept, taking it away from its rightful origin, cutting it off from the mental content and mental processes that, so to speak, give it life.

So a concept is "stolen" when one attempts to use it while simultaneously negating or ignoring the prior concepts on which it depends for its meaning.

Do you agree that that's a fallacy? Do you agree that the following are examples?

1. Consciousness is a myth. (Steals "myth")

2. All life is a dream. (Steals "dream")

3. Logic is a Western prejudice. (Steals "prejudice")

4, I can prove that the world exists. (Steals "proof")

5. Property is theft. (Steals "theft")

6. The universe is moving. (Steals "motion")

I think it's clear that this is something wider and deeper than the contrast argument. "I can prove that the world exists" is not in any clear way lacking contrast, but the concept "proof" arises because of, and depends upon, recognizing the distinction between a belief vs. a fact of reality, and thus the need for a method of determining when an idea does correspond to fact.

A contrast is one thing (normally) required for concept-formation, but is not the only thing. And I agree that the issue, for contrast, is the possible--even the imaginable--not just the actual.

[Daniel Barnes]

Interestingly, Objectivists also use "stolen concepts".

For example:
7. Induction is valid. (Steals "valid" from deductive logic)

However, I do agree with him that there are mental (ie non-physical, or abstract) objects as well as physical objects, ie that we are dealing with a dualism at least. Rand herself is unclear on the issue, and that lack of clarity pervades Objectivist debate. Mr Binswanger is to be applauded for taking a much less equivocal line here, as he has done for some time.

[Harry Binswanger]

Bill wrote:

"If Meinong is right and some objects are nonexistent objects, then existence is classificatory . . ."

Meinong's idea is a direct contradiction: objects that exist as nonexistent. Objects are, that are not.

Perhaps what you intend is: imaginary objects. Imaginary things, like God, exist as mental content projections, but do not exist apart from the mind(s) that imagine them. So it is not "existence" that is classificatory, but "external existence."

(Yes, I can answer Quine's challenge, in "On What There Is", to this move. He leans on a claim "this mental entity is not what people are talking about when they deny Pegasus." But the idea of Pegasus is manifestly what people are talking about when they say "Pegasus does not exist"--unless you make a fetish about the difference between the subject of that statement being the idea vs. the subject being the idea's relation to logic and to fact. I assume that Quine's materialism is what leads him to oppose the reference to ideas.)

[Bill Vallicella]

Harry writes, >>"If Meinong is right and some objects are nonexistent objects, then existence is classificatory . . ."

Meinong's idea is a direct contradiction: objects that exist as nonexistent. Objects are, that are not.<<

You are just wrong about that, Harry. There is no formal-logical contradiction in 'Some objects do not exist' unless you assume that only existent objects are objects. Read Meinong, Routley, Butchvarov, Castaneda, T. Parsons. They have dealt adequately with your objection. In any case, I am no Meionongian. For me, everything exists: there are no nonexistent items. This is one of the few things we agree on.