Imploding the Demon: Crispin Wright’s Response to Scepticism, Part 1
As promised, I’m going to have a look at some recent responses to philosophical scepticism, beginning with Crispin Wright’s “Scepticism and Dreaming: Imploding the Demon”.
I was initially struck by its intricacy, and flicking through I did wonder if perhaps I ought to postpone reading it till I had absorbed some elementary analytical philosophy. But I persevered, and in the end it was good fun. Sometimes we talk about works of philosophy being “difficult,” but there are different kinds of difficult. There is the difficulty that comes from ambiguity, obscurity and unconventional usage; but this paper is of the other kind: densely packed, rather like mathematics, but mostly very clear. There may be a case for saying that this kind of philosophy is too technical and specialized, but it is too early for me to form an opinion on that.
Wright believes that scepticism “can be made to succumb to a head-on, rational response”. This exciting claim contrasts with another common way of dealing with it, which is to say “in the end we just don’t have a satisfactory answer, and here’s why we shouldn’t worry…”. Barry Stroud, in particular, has written a whole book about the subject which begins with an admission that, on its own terms, “the problem has no solution.” Wright wants to show that scepticism can be answered directly, not just by side-stepping it.
Somewhat interesting is that he treats as equivalent the two hypotheses of Descartes, that we are dreaming and that we are being deceived by a demon (along with the brain in a vat hypothesis). This is indicated by the odd title, and nowhere does Wright draw a distinction between the two. I think this is justified, because while the dream argument is weaker on the face of it, when correctly formulated it is just as devastating. It seems to me that Descartes’ two hypotheses are proposed in the process of building one essential argument built on the premise that we cannot tell the difference between perceiving the external world and seeming to perceive it.
So in the arguments that Wright formulates he explicitly uses the example of dreaming, but any other similar hypothesis could be substituted for it, each of which posits an
...undetectable but cognitively disabling state – dreaming, pervasive hallucination, victimisation by the Cartesian demon, Brain-in-a-vathood, etc.
First, he lays out a framework of three constraints. These serve to define the problem and the criteria that a good response must satisfy.
Mild scepticism: our grounds for belief do not prove the beliefs conclusively.
Radical scepticism: our grounds for belief do not justify them at all. This is what Wright calls the sceptical paradox.
A common kind of response to scepticism is that exemplified by Russell and Strawson, to the effect that we can be satisfied with the probability and utility of our beliefs even though we would not, when pressed, claim that they attain to knowledge, strictly speaking. Wright thinks that this response does not get us off the hook, because radical scepticism brings our justifications into doubt.
Bertrand Russell, in The Problems of Philosophy, says:
In one sense it must be admitted that we can never prove the existence of things other than ourselves and our experiences … There is no logical impossibility in the supposition that the whole of life is a dream, in which we ourselves create all the objects that come before us. But although this is not logically impossible, there is no reason whatever to suppose that it is true; and it is, in fact, a less simple hypothesis, viewed as a means of accounting for the facts of our own life, than the common-sense hypothesis that there really are objects independent of us, whose action on us causes our sensations.
This is an example of what Wright calls the Russellian Retreat, the admission that we cannot prove the existence of an external world, but that our common sense explanations are more probable (simplicity supporting higher probability in Russell’s argument above) than any of the sceptical hypotheses, such as that we are dreaming, that we are being deceived by a malicious demon, or that we are a brain in a vat being deceived by a malicious scientist.
But the Russellian Retreat is a response to mild scepticism. The first constraint is that we need to respond to radical scepticism, which deprives us, not only of knowledge, but of sufficient grounds for belief, and asserts that we cannot tell the difference between grounded and ungrounded beliefs; that our beliefs are never actually justified at all. If we succeed in answering this, then there is no need to worry about knowledge or proof.
Wright emphasizes that the sceptical arguments are paradoxes, and must therefore be fully undermined and diagnosed. Although we will tackle scepticism “head-on”, we cannot be content with attacking the arguments: we must get to the root of the problem.
The second constraint is motivated by the thought that sceptical arguments are not properly rebutted by considerations whose force depends on the assumption of an adversarial stance: a scenario in which the object is to defeat a real philosophical opponent, the Sceptic, in rational debate. There are no real such opponents.
The real demon is the sceptic within.
The Sceptic does not need to win but only to draw.
The sceptic only has to show that we have no justification for rejecting the conclusion that none of our beliefs is justified. The sceptic does not have to justify his premisses, only show that we have no justification for denying that they are all true, ie. even if the sceptic’s premisses are not justified, if we have no justification for denying their conjunction, then we have lost.
In other words, we not only have to tackle the first-order sceptic, by revealing each of the premises to be unsupported; but we must deal with the second-order sceptic who comes back with “ah, but you still have no basis for an outright denial that my premisses are all true”.
Descartes via Stroud
We have to cope with Descartes’ observation that
...there are no conclusive indications by which waking life can be distinguished from sleep…
From here Wright sets out to construct the strongest sceptical argument in its general form, beginning with Stroud’s formulation.
Descartes’ Principle: To know P I must know that any condition I know to be necessary for knowing P is satisfied. For example, if I know that only if I’m not dreaming am I writing, then to know that I am writing I must know that I am not dreaming.
Given that I know that K(writing) → (not dreaming), then by Descartes’ Principle we get K(writing) → K(not dreaming), which can serve as the first premise (where K(x) stands for “I know that I am x”):
K(writing) → K(not dreaming)
(by modus tollens)
That is to say, I do not know I am writing.
But this is not a good candidate for the ultimate sceptical argument that Wright is looking for. Fundamentally, there are problems with the second premise. It can be seen two ways: if it is claiming that we never know we are not dreaming, then this is simply an unjustified claim; and anyway, we can just counter it by saying “actually, I am often justifiably certain that I am not dreaming, though I can’t prove it” i.e., with the Russellian Retreat. Stroud’s argument, concerning knowledge as it does, is mild scepticism.
Having established that the ultimate sceptical argument cannot be formulated in terms of knowledge, we have to decide on what to use instead. We are essentially looking for an appropriate definition of a belief’s grounds. How about reasonable belief? That won’t do, because we do not feel that our beliefs lose their justification if only reasonable beliefs are called into question. I have believed things in the past that turned out to be wrong, but I still regard those beliefs to have been reasonable. I can reasonably believe that I am not dreaming and yet be wrong.
We must find something a bit stronger, something whose possession, as a property, by a belief that I had in the past which turned out to be wrong, I would then be compelled to deny – but not because I had incontrovertable proof, i.e., not going so far as knowledge itself. Like knowledge it must satisfy non-internalist standards of justification, but unlike knowledge it must not allow the Russellian Retreat: it must be more dear to us than that.
Wright’s candidate, warrant, depends on the notion of pedigree.
Pedigree: the pedigree of a belief encompasses its history, circumstances, and the holder’s grounds
Warrant: a belief is warranted if there is sufficient reason to hold it and its pedigree is good
More explicitly, a belief held by x is warranted if
(a) x has sufficient reason to hold it; and
(b) its pedigree, if one knew it in detail and in full, would compel one rationally to view the belief as more probably true by virtue of its being held by x (independently of whatever one knows about the facts)
Why is warrant more challenging to an attempt to satisfy the first constraint than knowledge is? Because it is unacceptable to give up our claim to have warranted beliefs, while, like Russell, we can live with giving up the claim to knowledge. If our beliefs are unwarranted, then either we do not have sufficient reason to hold them, or else, objectively, the belief is no more likely to be true owing to our holding it for certain reasons than if we had just guessed. Either way, it is unacceptable. Warrant, then, looks like a good candidate for the radical sceptical argument.
The Ultimate Sceptical Paradox: First Attempt
With this groundwork in place Wright begins to build the sceptical argument. But first, some definitions are in order.
Iterativity: If I know P, then I know that I know P (also known as the KK thesis)
Transmission: if I know P, and I know that P entails Q, then I know Q on that basis
In looking into transmission I came across something that initially looked like just another name for the same thing: the principle of epistemic closure. However, what makes transmission different from closure is the key phrase in the above definition, “on that basis”. The closure principle alone is silent on the basis of knowing Q. Right now I don’t know why this distinction is important – I’ll have to come back to this.
Rx[P] = x has available warrant to believe P at time t.
P = a proposition that requires a warrant from perception if it is to be warranted at all.
Dx = x is dreaming at t.
here is the argument:
(P1) ¬Rx [¬Dx]
(P2) Rx [Rx[P] → ¬Dx]
Rx[P], assumed for reductio ad absurdum
Rx[Rx[P]], by iterativity
Rx[¬Dx], by transmission from P2
This is contrary to P1, so
The conclusion says that, at any time, there is no warrant available to anyone for believing any propositions that require perception to be warranted. In other words, at any time, if the warrant required to believe a proposition depends on perception, then you cannot get it. This means that if my beliefs about the external world “I have acquired either from the senses or through the senses,” then all of my basic beliefs about the external world are unwarranted.
A rather unpalatable conclusion. But Wright is only just getting going. He proceeds to strengthen the sceptical argument even further before attempting the head-on attack.
Wright, C. 1991: “Scepticism and Dreaming: Imploding the Demon”, Mind, vol. 100, pp87-116
Descartes, R. 1996: Meditations on First Philosophy, Cambridge University Press
Stroud, B. 1984: The Significance of Philosophical Scepticism, Oxford University Press
Russell, B. 1912: The Problems of Philosophy, Oxford University Press
Strawson, P.F. 1985: Skepticism and Naturalism: Some Varieties, Routledge