Can one know whether one has hands?

Initially one might argue that one cannot know one has hands, because any past experience of hands could merely be a dream about hands. After all, and as Descartes suggested, dreams often feel real until one wakes up, and so there seems no fool proof way of testing whether one’s experience of hands is an accurate representation of reality. However, as Moore said, we can prove we have hands without having to prove that we are not dreaming. More specifically, Moore said that the standard of proof is not as high as Descartes suggests; Moore said that all we need for a successful proof is an argument where the initial premises are known, and they support the conclusion. Thus one can know whether one has hands simply because one can argue that: 1)    I have a right hand 2)    I have a left hand 3)    Therefore, I have hands. It has been common to respond to Moore by claiming that he assumes too much: we cannot know we have a right or left hand, precisely because of the possibility that we are dreaming about the hands. In defence of Moore, I think that work on different accounts of knowledge can be brought to bear on this discussion. Goldman claims that one can know something without realising it; knowledge can be defined as true belief, discovered by a process that is reliably true. Thus, one can be said to know that one has a right hand in the case where this is a true belief, and the process of perception is reliably true. Perception is reliably true only in the case where we are not dreaming, so one can know one has a right hand only when one is not dreaming. And so, on my account, one can use Moore’s argument to prove one has hands only when one is not dreaming. This is an improvement on Descartes’ original position as the possibility of dreaming is not enough to undermine knowledge of hands – one can know one has hands even if one (wrongly) thinks one could be dreaming.