spoke not of proof and signs but of 'confirmation' and 'disconfirmation': 'if judgements are not confirmed
or are disconfirmed, falsity arises; if they are confirmed or not disconfirmed, truth'. A later author
illustrates the notion: 'when Plato is approaching from far off, I conjecture and judge because of the
distance that it is Plato; after he has approached and the distance is reduced, it is testified and confirmed
by direct evidence that it is Plato'. The ideas of confirmation and disconfirmation are relatively plain, but
non-confirmation and non-disconfirmation produce puzzles: those puzzles may have spurred Epicurus'
successors to develop their own theory of signs.
What, next, of the 'criterion of truth' by which we judge what is 'evident'? In general, 'the old philosophers
say that there are two kinds of evident things, those which are discerned by one of the senses. .. and those
which strike the mind with a primary and non-demonstrable impression'. According to Epicurus, all
objects are continuously emitting effluences of various kinds. When the effluences strike an appropriate
part of a sensitive body - the ears or the nose, say - their parent object is perceived. The effluences which
affect the eyes and cause the parent object to be seen are simulacra or replicas - thin skins which preserve
the contours of the object as they speed through space. Thought is analogous: Lucretius states that 'the
mind is moved by replicas of lions and of everything else it grasps in just the same way as the eyes are -
except that it perceives more flimsy replicas'.
Such effluences supply us with the concepts we employ and form the basis of our knowledge. Moreover,
the impressions they make on us are in some way infallible: 'whatever impression we grasp directly by the
mind or by the sense-organs, whether of shape or of other properties - that is the shape of the concrete
body ... Falsity and error always depend on a superadded opinion.'
The Stoics agreed that our conceptual resources and our knowledge are founded upon direct impressions
from external objects. They differed over the physics and physiology of cognition. But their most
important difference from the Epicureans lies in their assessment of the impressions themselves. For the
Stoics held that our knowledge rests upon one special type of impression. These 'apprehensive
impressions' were defined as impressions 'deriving from an existing object, signed and sealed on the mind
in conformity with the existing object itself, of a sort which could not derive from anything other than the
existing object.' The impressions of a madman may be mere figments, deriving from no existent object; an
impression may derive from a real object and yet misrepresent it; an impression may represent an object
correctly, yet fail to record its peculiar individuality. No such impression is apprehensive. 'For they
require that apprehensive impressions genuinely grasp their objects and imprint exactly all their individual
characteristics.'
Epicurean impressions are all true: falsity enters only when we misinterpret our data. The Stoics allow
that the data themselves are often distorted: in order to avoid falsity we must scrutinize and select. But
both schools insisted that we do obtain knowledge from our impressions, and each faced challenge from
the sceptics of the Academy.
The Epicurean dogma appeared easy to attack: there are familiar perceptual illusions, and the senses
sometimes offer conflicting testimony - how, then, can all impressions be true? The Epicureans were well
aware of the phenomenon of illusion, for which they offered physical explanations. Lucretius knows that