tly clarified, and what is genuine in
them would be more easily distinguished from what is artificial, if we
could gather together again the original sources for the belief in
separate minds and compare these sources with those we have already
assigned to the conception of nature. But speculative problems are not
alone concerned, for in all social life we envisage fellow-creatures
conceived to share the same thoughts and passions and to be similarly
affected by events. What is the basis of this conviction? What are the
forms it takes, and in what sense is it a part or an expression of
reason?

This question is difficult, and in broaching it we cannot expect much
aid from what philosophers have hitherto said on the subject. For the
most part, indeed, they have said nothing, as by nature's kindly
disposition most questions which it is beyond a man's power to answer do
not occur to him at all. The suggestions which have actually been made
in the matter may be reduced to two: first, that we conceive other men's
minds by projecting into their bodies those feelings which we
immediately perceive to accompany similar operations in ourselves, that
is, we infer alien minds by analogy; and second, that we are immediately
aware of them and feel them to be friendly or hostile counterparts of
our own thinking and effort, that is, we evoke them by dramatic
imagination.

[Sidenote: Two usual accounts of this conception criticised:]

[Sidenote: analogy between bodies,]

The first suggestion has the advantage that it escapes solipsism by a
reasonable argument, provided the existence of the material world has
already been granted. But if the material world is called back into the
private mind, it is evident that every soul supposed to inhabit it or to
be expressed in it must follow it thither, as inevitably as the
characters and forces in an imagined story must remain with it in the
inventor's imagination. When, on the contrary, nature is left standing,
it is reasonable to suppose that animals having a similar origin and
similar physical powers should have similar minds, if any of them was to
have a mind at all. The theory, however, is not satisfactory on other
grounds. We do not in reality associate our own grimaces with the
feelings that accompany them and subsequently, on recognising similar
grimaces in another, proceed to attribute emotions to him like those we
formerly experienced. Our own grimaces are not easily perceived, and