We can distinguish two kinds of skeptical perspectives on some domain of beliefs.
One kind of skeptic claims that the beliefs in question are false. Such a skeptic might do this by arguing that the beliefs are internally incoherent or contradictory, or by presenting independent arguments for a thesis which contradicts them. We have seen skeptical arguments of this kind from McTaggart and Bradley, who argued that our ordinary beliefs about, e.g., things existing in time or things standing in relations to each other are false. For lack of a better term, we can call this metaphysical skepticism.
A second kind of skeptic might not argue directly that the beliefs in question are false, but only argue that they are unjustified. For example, a skeptic of this kind about the existence of God might not argue that God does not exist, but only that we have no good reason to believe that God exists. Call this epistemic skepticism.
Moore’s target is skepticism about the external world. We have already, in McTaggart and Bradley, encountered several arguments for metaphysical skepticism about much of what we take ourselves to know about the external world. Following are two intuitively plausible arguments for epistemic skepticism about the external world.
Let REALITY be a world in which there are external objects; let DREAMWORLD be a world in which there are no external objects (other than you), but in which everything ‘seems the same’ to you as in REALITY. Then the skeptic can argue as follows:
| P1. | If an agent cannot distinguish between two situations, then that agent cannot know which of those two situations obtains. |
| P2. | We cannot distinguish between REALITY and DREAMWORLD. |
| C1. | Therefore, we cannot know whether REALITY or DREAMWORLD obtains. |
| C2. | Since there are external objects in REALITY but not in DREAMWORLD, we cannot know whether there are any external objects either. |
| P1. | For an agent to know that p is true, that agent must have evidence which rules out the possibility that p is false. |
| P2. | Our only evidence about external things is sensory evidence about how things appear to us. |
| P3. | Hallucinations -- cases in which we have sensory evidence about how things appear to us but in which there is no existing external thing presented -- are possible. |
| C1. | Therefore (by P2 and P3), we never have evidence which rules out the possibility that external things do not exist. |
| C2. | Therefore (by C1 and P1), we cannot know that external things exist. |
For now, it is important to note the form of these skeptical arguments: both begin from a certain premise about what is required for knowledge, and proceed to show that we fail, in our beliefs about the external world, to meet this requirement.
Moore begins the article by trying to get clear about the class of things whose existence is doubted by the skeptic; he does this by distinguishing between things presented in space and things to be met with in space.
These two classes of things are distinguished by the fact that things presented in space (but not to be met with in space) are such that (i) they can only be perceived by one person, and (ii) they exist only if they are perceived.
Examples of the two classes of things.
What the skeptic about the external world, or external things, doubts is the existence of things which can be met with in space.
(Moore discusses the category of things ‘external to our minds’, and takes the skeptic to be arguing that there are no things in this category. For our purposes, we can ignore this extra step. According to Moore’s definitions, if something is to be met with in space, then it is also external to our minds; so if we can give an anti-skeptical argument that there are some things to be met with in space, this will also be enough to show that there are some things external to our minds.)
Moore wants to go on to give an argument against skepticism about the external world; before we consider that argument, we should ask what is required of an argument for it to be a good argument against skepticism.
Moore gives us three criteria (see p. 146):
1. The premises must be different from the conclusion.
2. The conclusion must follow from the premises.
3. The premises must be known to be true.
Moore thinks that he can prove that the skeptic about the external world is wrong. His simple proof is as follows:
| P1. | Here (holding up one’s left hand) is one hand. |
| P2. | Here (holding up one’s right hand) is another. |
| C1. | Therefore, there are at least two hands. |
| C2. | Therefore, there are at least two things to be met with in space. |
How might a skeptic respond to Moore’s argument? By either calling into question Moore’s criteria for something’s being a proof, or by calling into question Moore’s claim that his proof meets those criteria.
It is not plausible to doubt either that the premises are different from the conclusion, or that the conclusion fails to follow from the premises. So, it might seem, the best bet for the skeptic is to argue that Moore does not in fact know the premises of his argument to be true. After all, if the skeptic is maintaining that we do not know that external things exist, shouldn’t the skeptic also maintain that we do not know that we have hands?
The example of the misprints (p. 147).
Moore discusses this reply on pp. 149-150. The status of ‘Nothing is known without a proof,’ and whether it can be used as part of an argument against Moore.
The skeptic may reply that Moore’s proof begs the question, since no one who disagreed with Moore’s conclusion would be convinced by his argument.
What ‘begs the question’ might mean here.
How this is related to the previous objection.
A Moorean reply: are you more sure of the fact that you have hands, or of certain abstract claims about knowledge?
A different kind of idealist skeptic might grant the conclusion that at least two hands exist, but doubt whether it follows from this that there are at least two things to be met with in space.
Can an argument akin to Moore’s original proof be constructed against this new skeptical position?
Why does he begin with the quote from Kant? Not to try to give the same kind of proof Kant tried to give; rather, to undercut the project of trying to give such proofs.
Philosophical methodology, and the question of whether the starting point of philosophical thought should be with commonsense particular claims that seem secure, or abstract philosophical claims which have intuitive support.