Emotive belief is very different. Readiness to act as though some references were true is often involved, but the connections and circumstances in which this readiness remains are narrowly restricted. Similarly the extent of the action is ordinarily limited. Consider the acceptances involved in the understanding of a play, for example. They form a system any element of which is believed while the rest are believed and so long as the acceptance of the whole growing system leads to successful response. Some, however, are of the form ‘Given this then that would follow’, general beliefs, that is to say, of the kind which led Aristotle, in the passage quoted above, to describe Poetry as a more philosophical thing than history because chiefly conversant of universal truth. But if we look closely into most instances of such beliefs we see that they are entertained only in the special circumstances of the poetic experience. They are held as conditions for further effects, our attitudes and emotional responses, and not as we hold beliefs in laws of nature, which we expect to find verified on all occasions. If dramatic necessities were actually scientific laws we should know much more psychology than any reasonable person pretends that we do. That these beliefs as to “how any person of a certain character would speak or act, probably or necessarily”, upon which so much drama seems to depend, are not scientific, but are held only for the sake of their dramatic effect, is shown clearly by the ease with which we abandon them if the advantage lies the other way. The medical impossibility of Desdemona’s last speech is perhaps as good an example as any. The bulk of the beliefs involved in the arts are of this kind, provisional acceptances, holding only in special circumstances (in the state of mind which is the poem or work of art) acceptances made for the sake of the ‘imaginative experience’ which they make possible. The difference between these emotive beliefs and scientific beliefs is not one of degree but of kind. As feelings they are very similar, but as attitudes their difference in structure has widespread consequences.