186 Aristotle and his school
which are caused by actual perceptions – not, such as in dreams, by lingering
sense-movements which derive frompreviousperceptions in the waking
state – are not dreams, because we do not have these experiences ‘in so far
as’ (h ̄ei) we are asleep, but in so far as we are, in a sense, already awake.
This typically Aristotelian usage of the qualifierh ̄eialso provides us with
an answer to the other question I raised earlier in this chapter, namely
why Aristotle does not explicitly address the possibility of other mental
experiences during sleep such as thinking and recollection. The answer
seems to be that thoughts, beliefs, perceptions, hallucinations, recollections
and indeed ‘waking acts’ (egr ̄egorikai praxeis) performed while sleepwalking
( 456 a 25 – 6 ) are not characteristic of the sleeping state: they do not happen
to the sleeper ‘in so far as’ (h ̄ei) (s)he is sleeping. They do not form part of
the dream, but they exist ‘over and above’ the dream (para to enhupnion). We
cannot say, in Aristotle’s theory, that we ‘think in our dream’, although we
can say that we think in oursleep. The role of thought in sleep is apparently
not essentially different from that in the waking state,^27 although there is
nothing in what Aristotle says here to suggest that we might have clearer,
‘purer’ thoughts in sleep than in the waking state.
The definition of dreams that Aristotle presents here is, as I said, in
accordance with his views inOn Sleep and Waking. Dreams are not actual
perceptions, rather they are, as it were, reactivated perceptions which we
received during the waking state; they are ‘movements of sense-effects’
(
) *
# )). Aristotle, again, differentiates between
various kinds of experience in sleep. And although the recognition that we
may after all have perceptions in sleep constitutes an important qualification
of Aristotle’s initial, and repeatedly reiterated view that there is no sense-
perception in sleep, Aristotle avoids contradicting himself by saying that
although we have these perceptions while we are asleep, we do not have
them in so far as we are asleep. The point of the specificationskuri ̄os kai
hapl ̄osinSomn. vig. 454 b 13 andtropon tinain 454 b 26 has now become
clear, and they are answered here by the specificationsp ̄eiandh ̄ei.
4 on divination in sleep
When we turn to the third treatise,On Divination in Sleep,however,itis
becoming increasingly less clear how Aristotle manages to accommodate
the phenomena he recognises within his theory without getting involved
in contradictions. In this work, he does accept that we sometimes foresee
the future in sleep; but his theory of dreams as expounded so far does not
give much in the way of help to explain how this can happen.
(^27) But see no. 20 – 1 above.