Aristotle on sleep and dreams 203
an argumentative, ‘dialectic’ or perhaps even didactic strategy (we should
not forget that Aristotle’s extant works derive from the teaching practice,
and that they are very likely to have been supplemented by additional
oral elucidation). Alternatively, it may be a matter of intellectual temper-
ament or style. However this may be, it is undeniable that Aristotle in his
works on sleep and dreams, as in his biological works at large, sometimes
shows himself an improviser ofad hocexplanations, constantly prepared to
adapt his theories to what the phenomena suggest. This inevitably means
a lower degree of systematicity than we would perhaps regard as desirable;
on the other hand, the elasticity of his explanations, and his readiness to
accommodate new empirical observations, are things for which he is to be
commended.
Lack of systematicity is, to a varying extent, characteristic of many Aris-
totelian works and can also be observed in other parts of theParva naturalia,
both within and between the individual treatises that make up the series.
But it seems to obtain particularly toOn Divination in Sleep,^55 which is
in general a less technical treatise whose degree of accuracy, both in scien-
tific terminology^56 and in the description of psycho-physiological details,
is rather low in comparison with the other two works. Instead, it shows
what could be called a more ‘dialectical’ character. Aristotle approaches the
problem of divination in sleep from different perspectives, but he offers
neither a definition nor a comprehensive explanatory account. The text
has a strongly polemical tone and is for a substantial part devoted to an
assessment of current views on the subject, such as the view (referred to and
criticised three times) that dreams are sent by the gods, or the view held by
the ‘distinguished doctors’, or the theory of Democritus.
(^55) As for systematicity, it is of course true that a discussion of the topic of prophecy in sleep is announced,
as we have seen above, in the preface toOn Sleep and Waking( 453 b 22 – 4 ), and thatOn Divination
in Sleeprefers, at one point ( 464 b 9 – 10 ), back toOn Dreams( 461 a 14 ff.). Yet not too much weight
should be attached to these cross-references, as they may easily have been added at a later, editorial
stage; besides, the preface toOn Sleep and Wakingpresents a programme of questions that is somewhat
different from what is actually being offered in what follows, and this also applies toOn Dreams.
Thus the beginning ofOn Sleep and Wakingannounces a discussion of the question ‘why people who
sleep sometimes dream and sometimes do not dream, or, alternatively, if they always dream, why
they cannot always remember their dreams’ ( 453 b 18 – 20 ); but these questions can hardly be regarded
as central toOn Dreams, where they are addressed only in passing (in 461 a 13 ) and incompletely (in
462 a 31 –b 11 , a passage that itself, too, shows signs of a hastily added appendix). Such discrepancies
between programme and execution need not, however, be due to later editorial additions, for it is,
again, not uncharacteristic of Aristotle’s works for there to be discrepancies between programme and
execution.
(^56) E.g. the use ofaisthanesthaiin the wide sense of ‘notice’, ‘be aware of ’ ( 464 a 10 , 15 , 17 ), or the
reference to ‘perception arriving at the dreaming souls’ (A #
"
3 1 :1 1
=-)in 464 a 10.