202 Aristotle and his school
at night and because in sleep we perceive these slight movements more
clearly than in the waking state. Aristotle does not say to which category
the dreams discussed here belong, but it seems that, if the category of
‘coincidence’ (sumpt ̄oma) is eliminated, these dreams stand to the events
they predict in a relationship of signs (s ̄emeia), and that both the event and
the dream go back to a common cause.
It is difficult, however, to see how the experiences described here can be
accommodated within Aristotle’s theory of sleep and dreams. They clearly
do not fulfil the requirements for dreams as posited inOn Dreams; nor
do they seem to belong to the category of borderline experiences, because,
again, Aristotle stipulates that they appear to us stronger than in the waking
state. Unless we were to assume that Aristotle is contradicting himself, we
might prefer to accept that in addition to dreams and to the borderline
experiences of hearing faint sounds and suchlike, he recognises yet another
kind of experience during sleep and that, by calling these experiencesen-
hupnia, he uses the term in a less specific, more general sense than the strict
sense in which it was used inOn Dreams. After all, as I have said, the word
enhupnionbasically means ‘something in sleep’, and this could be used both
at a more general and at a more specific level. But in that case, very little is
left of Aristotle’s initial,a prioriassumption that sleep is an incapacitation
of the sensitive part of the soul, for it turns out that we are perfectly well
capable of perceiving these movements while asleep, provided that the at-
mospheric conditions are favourable. Nor is it open here to Aristotle to say
that these movements originating from remote places such as the Pillars of
Heracles are perceived by us not ‘in so far as’ we are asleep but in so far
as we are, in a certain way, already awake: in fact, Aristotle explicitly says
that we receive these stimuli ‘because’ we are asleep – indeed, they ‘cause
perception because of sleep’ (A #
$
1 3 2), which
seems in blatant contradiction to everything he has said inOn Sleep.
A different approach to this problem is to seek an explanation for these
apparent inconsistencies in what Charles Kahn has called ‘the progressive
nature of the exposition’ in Aristotle’s argument.^54 In the course of his ar-
gument, Aristotle sometimes arrives at explanations or conclusions which
implicitly modify or qualify things he has said earlier on without recognis-
ing this explicitly or revising his earlier formulations. Instead, he simply
goes on, eager to explain as much as he can and carried away by the sub-
tlety and explanatory power of his theories, but without bothering to tell us
how these explanations fit in with what he has said earlier on. This may be
(^54) Kahn ( 1966 ).