Labouring over master arguments

Wed 12 Mar 2008 05:37 PM

I have been teaching Berkeley in my 17th&18th c. philosophy course. It is always a bit of trip, because students never come to the good bishop's defense. That leaves me in the role of trying to make the view seem as plausible as possible. I won't convince any of them, of course, but I can perhaps engage them enough that they can think through the view's consequences without scoffing dismissively.

At the beginning of Part I of the Principles, Berkeley gives what I reconstruct as the following argument:

P1 What you perceive are ideas.

P2 Ideas cannot exist outside minds.

C1 Therefore, what you perceive can't exist outside of some mind.

P3 You perceive (eg) the screen in front of you.

C2 Therefore, the screen in front of you cannot exist outside some mind.

The conclusion is a naked consequence of idealism, the kind of claim that makes people scoff and kick stones.

It is interesting to work through the argument backwards from there: P3 is required if we are not going to be sceptics, and C2 is a deductive consequence of C1 and P3. So any principled resistance will have to come earlier than that.

C1 is similarly a deductive consequence of P1 and P2. So the two options for avoiding the conclusion of the argument are denying P1 or denying P2.

P2 is motivated by our introspection into the nature of ideas and also by Berkeley's attack on abstraction in the Introduction. I had one student this time who was tempted to deny P2, but most are willing to accept it. (Moreover, denying P2 would lead to a wacky metaphysics; possibly even wackier than Berkeley's idealism.)

So that leaves P1. It is a seductive position. Having starting with Descartes' first Meditation, we are invited to consider the play of sensations as a kind of flickering movie in our consciousness. This 'veil of perception' stands between us and the world, and the task of epistemology is to trace out the inferences that lead to belief in an external world behind the theater of sensation. Arguably, this premise frames the whole period from Descartes through the late-18th century. Kant accepted it, after a fashion, but put the empirical world on our side of the screen. Reid rejected the premise, which he called the theory of ideas.* Berkeley brilliantly distills this element of the tradition. He states P1 at the outset, in Sec 1 of Part 1 of the Principles.

That's is where the action is, and Berkeley knew it.

This is fun to teach because it is valid. If we accept P1 and P2, we are committed to idealism: Students always want to know about the consequences of idealism, how it can be reconciled with science and religion, and so on. But it is too late to worry about its consequences once you've accepted the first two premises. However strange they are, you are committed to them.

The argument cleverly, deductively draws the wildest metaphysics out of what seem like plausible premises. The first time I taught the Bishop of Cloyne, back at Bowdoin College, I had a student who stared intently at the argument on the chalkboard after class had ended. He stayed fixed on it even as he got up to leave. "This can't be right," he said. "I don't know what's wrong with it, but something has to be."

So in preparing lecture notes, I jotted down "Berkeley's master argument" above my reconstruction of it. However, I noticed today that Jonathan Dancy identifies a different argument as GB's "master argument" (in the "Analytical Contents" Dancey prepared for the OUP edition of the Principles that he edited.) Roughly, it goes like this:

P1 Necessarily, everything of which you can conceive is conceived by some mind.

C1 Therefore, everything is necessarily conceived by some mind.

P2 You can conceive (eg) the screen in front of you.

C2 Therefore, the screen in front of you is necessarily conceived by some mind; that is, it cannot exist except as conceived.

Admittedly, Berkeley does make this argument in the course of the Principles. Perhaps I have reconstructed it uncharitably, but greater care would not do it any favors. It equivocates like a mongoose eating a garden hose. It slides around modal operators like a hockey player who mistook necessity for a puck. It does things which are too unspeakable to be captured by cheeky metaphors.

So, when I teach the Berkeley, I either skip over this argument entirely or mention it only in passing. There is certainly a tradeoff here. If I discussed the argument in more detail, I could discuss why it is so many kinds of terrible. Although this might teach students something, it would not help them take our man GB seriously. The first argument is much better at that, because it can be pretty seductive.

Unsound, yes. But seductive.


from: Pete Murray

Sun 31 Jan 2010 07:00 AM

You're skeptical that the "Master Argument" (so-called by Andre Gallois) can be made out in any persuasive way. I think it fails, but still I think it can be understood in way that makes it not so awful.

The argument doesn't rely on modal equivocation in the following manner of parsing it out:

P1 Anytime I conceive of something existing, I conceive of that thing.

C1 It is impossible to conceive of anything existing unconceived.

C2 The following idea, 'a conception of an object existing absolutely, in itself,' is either contradictory or meaningless.

Note that this doesn't get us your C1, that everything that exists necessarily exists as perceived by some mind. What it does do is show us (if it worked, which it doesn't) that we can have no idea of an object existing unperceived, and thus no ground whatsoever to posit such an existence. Philonous says to Hylas, after giving the Dialogues version of this argument, "And yet you will earnestly contend for the truth of that which you cannot so much as conceive." I think this is the take-away: after arguing that we do not immediately perceive mind-independent objects, Berkeley attempts to demolish the claim that we have reason to think that we mediately perceive mind-independent objects via our ideas (a premise, I take it, that he takes both Locke and Descartes to hold).

Now, I think the right response to this argument is that it confuses the representation with the represented. I may not be conceive of a representation without thinking of it, but this doesn't imply that what is represented must be imagined in the same way. I think both Russell and Pitcher, among others, have made this kind of argument. I don't think this way of understanding the argument makes it look silly, though, because this is exactly the kind of distinction that is being worked out by these guys, that we now take for granted. Berkeley deals with this sort of issue explicitly, though not necessarily satisfactorily, in the first of the Dialogues when he talks about a picture of Caesar representing Caesar (it turns out it is just repeated conjunction: I'm led to one set of ideas from another that have been previously connected in past experience, and this is all the representation relation here amounts to for Berkeley, I believe).