J. L. Austin, in a footnote to p. 49 of Sense and Sensibilia (Oxford, 1962), writes of ". . . the absurdity of Descartes' toying with the notion that the whole of our experience might be a dream." In the main text, there is a sort of argument for this alleged absurdity. The argument may be set forth as follows:
2. If the phrase 'dream-like quality' were "applicable to everything," then "the phrase would be perfectly meaningless." (49)
Therefore
3. If dreams were not qualitatively different from waking experiences, then the phrase 'dream-like quality' would be perfectly meaningless. (From 1 and 2 by Hypothetical Syllogism)
4. The phrase 'dream-like quality' is not perfectly meaningless.
Therefore
5. Dreams are qualitatively different from waking experiences, and "the notion that the whole of our experience might be a dream is an "absurdity." (From 3 and 4 by Modus Tollens)
This is what is called a Contrast Argument. The idea is that a term cannot be meaningful unless there are items to which it does not apply. The idea has some plausibility: if a term applies to everything, or everything in a specified domain, then there is a 'failure of contrast' that might seem to drain the term of all meaning. So if every experience were dream-like, then it could seem meaningless to say of any experience that it is dream-like.
But I see no reason to accept premise (2) above and the contrast principle on which it rests. The principle is
CP. If a term T applies to everything, then T is meaningless.
(CP) readily succumbs to counterexamples. 'Self-identical' applies to everything without prejudice to its meaningfulness. The fact that nothing is self-diverse does not make it meaningless to say that everything is self-identical. Or consider a nominalist who claims that there are no universals, that everything that exists is a particular. Is our nominalist saying something meaningless? Clearly not.
Or if a Buddhist maintains that all is impermanent, is he maintaining something meaningless? Must there be permanent entities for it to be meaningful to say of anything that it is impermanent? I say no. It would be a cheap and feeble response to the Buddhist to allege that the very sense of 'All is impermanent' requires the existence of one or more permanent entities. It is not even required that it be possible that there be permanent entities. If it is necessarily the case that all is impermanent, then there cannot be any permanent entities. Nevertheless, it is meaningful to make this strong impermanence claim. One understands what is being said.
There are philosophers who hold that every being is a contingent being. Surely this cannot be refuted by claiming that the very sense of the thesis that all beings are contingent requires the existence of at least one necessary being. It is true that 'contingent being' has sense only by contrast with the sense of 'necessary being': one cannot understand the one term without understanding the other. But it does not follow that 'contingent being' has sense only if there are necessary beings.
The senses of 'good' and 'evil' are mutually implicative. But it does not follow that there cannot be good without evil.
Returning to the Cartesian Dream Argument, could the whole of my experience be dream-like? This supposition may at the end of the day be absurd; but it cannot be shown to be absurd by Austin's contrast argument.
I agree with your comments here (and elsewhere on the blog) about the Contrast Argument, but there is also a simple problem with Austin's specific argument: the use of the term "dream-like" does not require that all dreams share a certain form, which we call dream-like. I cannot speak for others, but some dreams I have are very dream-like (in part because I am aware that I am in a dream), and others are very life-like [lucid] (as I am not aware, until later, that it is a dream).
In other words, "dream-like" can be meaningful AND some dreams can be qualitatively the same as waking experiences.
Given that during a lucid dream I am not aware that it is a dream, it is at least possible that my entire experience is of this type. Not so much an issue with "dream-like" life, but with "life-like" dreaming.
Posted by: Andy | Wednesday, February 18, 2009 at 07:21 PM
I have no problem with contrast arguments as a species. The problem I have with Austin's argument is that it is not a proper contrast argument. The Cartesian argument is that some experiences can be shown not to represent anything (dreams), whereas other experiences do represent something (imagine we are looking both at someone and their reflection in the mirror. So we are able successfully to contrast represention-of-something with not-representing-something.
Descartes now points out that what I take to be "my wife" is really itself an image or sense-datum, and that the image of my wife in the mirror is really an image of an image. So, is it conceivable that there is no reality at all corresponding to my wife-sense-datum?
I think this is conceivable even if (for other reasons) it is impossible.
In summary, we do have an available contrast between 2nd-order images which have first order images (reflections) and 2nd-order images which have no corresponding 1st order images (dreams). Given this contrast, we can then legitimately ask whether the 1st order images (sense-experience) are really images of anything, or whether they are ilusory.
So Austin's argument fails for this reason.
Hope that makes sense.
Posted by: ocham | Thursday, February 19, 2009 at 06:10 AM