Frede-Stoics and Skepticson Clear and Distinct Impressions

9
Stoics and Skeptics on Clear and

Distinct Impressions
The history of Hellenistic philosophy is dominated by the rivalry between Stoics

and skeptics, first Academic skeptics and later Pyrrhonian skeptics who tried to
revive a more radical form of skepticism when in the second and first centuries
B.C. Academic skeptics seemed to have softened their stand to a degree that
made it difficult to distinguish them from their Stoic rivals. The debate between
Stoics and skeptics primarily concerned the nature and possibility of knowledge.
If the skeptics also tried to attack the Stoic position on all other questions, the
point of this, at least originally, was in good part to show that the Stoics them-
selves had failed to attain the knowledge they claimed to be attainable.
Both Stoics and skeptics saw themselves as followers of Socrates, but they
took a different view as to the moral to be drawn from Socrates' experience. Soc-
rates by his dialectical practice had shown that, in spite of claims to the contrary,
nobody actually possessses the kind of knowledge which would guarantee a ra-
tional and happy life, and that, if he himself had any claim to wisdom, it rested
only on his ready recognition that he was no less ignorant than anybody else.
But Socrates had not resigned himself to his ignorance. And the Stoics seem to
have assumed that the reason for this was that Socrates thought that the special
kind of knowledge which he had shown people to lack is in fact attainable. They
assumed that nature must have constructed human beings in such a way as to
make it possible for them to lead a rational and good life. And if this, as Socrates
was thought to have shown, is a matter of being wise, nature must also have
provided us with the means to gain the kind of knowledge which constitutes wis-
dom. The skeptics, on the other hand, thought that it remained an open question
whether such knowledge could be attained and that hence all one could do mean-
while was go on looking for the truth and subject all claims to the kind of dialec-
tical scrutiny Socrates had subjected them to. Since the Stoa was rapidly de-
veloping into the most influential school, it was only natural that the skeptics
would turn their dialectical skill in particular against the Stoics who claimed to
151
152 CLEAR AND DISTINCT IMPRESSIONS
be on the way to the kind of knowledge Socrates had searched for in vain.
Now when the Stoics claimed that such knowledge is attainable, they also
thought that they had to construct an epistemology in terms of which they could
show that and how such knowledge is to be gained. On this account nature has
provided us with a firm basis for knowledge by providing us with clear and dis-
tinct impressions, the so-called kataleptic or cognitive impressions, which by
their very nature cannot be false and hence constitute an unfailing guide to the
truths one has to know in order to have the wisdom that guarantees the good life.
Thus the Stoic theory of knowledge is based on a doctrine of clear and distinct
impressions. Given that the skeptics not only were not persuaded that such
knowledge had been attained, but even questioned whether such knowledge was
attainable, they naturally focused their attack on the Stoic theory of knowledge
and in particular on the Stoic doctrine of clear and distinct impressions by means
of which we are supposed to be able to acquire the knowledge in question. As
a result a lively debate ensued which lasted for more than two centuries and
which attracted the best philosophers of the time.
Tradition, though, has developed a view of the Stoic position which makes
it so vulnerable to skeptical attacks that it becomes very difficult to understand
how the Stoics, through centuries, were able to sustain the criticism without hav-
ing to concede defeat. If the Stoics had defended the position that tends to be
ascribed to them, their school should have been deserted in no time. That instead
it was defended by men of the ingenuity of a Chrysippus should encourage us
to take a fresh look at the Stoic position to see whether it might not be more at-
tractive or at least easier to defend than tradition would make us believe.
The Stoic Position
Impressions
Animals and human beings are constructed in such a way that their survival
and well-being depends essentially on the adequacy of their cognitions. They
have to be able to recognize and to shun what is bad for them, and they have
to be in a position to realize and seek out what is conducive to their preservation
and well-being. For this purpose they are equipped with a sensory apparatus and
a soul which, via the senses, receives impressions of the outside world, and thus
provides them with some kind of awareness of the world around them. There
is a crucial difference, though, between the impressions of rational beings and
the impressions of animals. The impressions of rational beings are called "ratio-
nal impressions" (D.L. VII 51). Rational impressions have a prepositional con-
tent, they are impressions to the effect that something is the case very much in
the sense in which we might say ordinarily, "the impression which one gets, if
one looks at the evidence, is that. . . . " Thus rational impressions are thoughts
CLEAR AND DISTINCT IMPRESSIONS 153
(D.L. VII 51; Ps.-Gal. Def. med. XIX 381 K.) which present themselves to the
mind and which the mind either accepts or refuses to accept. To accept or give
assent to a thought or impression is to have the belief that the proposition which
forms the content of the impression is true, to refuse to accept a thought is to
suspend judgment. Thoughts may present themselves to the mind in all sorts of
ways. They may come to mind when one considers the evidence concerning a
question in doubt. But many of them are brought about by the causal agency of
an external object which, through the sense organs, gives rise to an impression
in us. Thus to see something, on this view, is to have a certain kind of thought
generated in a certain way. But thoughts may also be generated in all sorts of
other ways.
Now the Stoics follow Socrates, Plato, and Aristotle in the view that it is a
mark of moral knowledge that one never has a mistaken view in moral matters.
The Stoics even take the stronger view that the wise man will never have any
false beliefs at all (Stob. Ed. II 111, 18 W.), because, for reasons we shall see
later, any false belief might stand in the way of one's acquiring the kind of
knowledge we are after. One way in which nature could construct a mind which
has the ability to avoid any false beliefs whatsoever would be to endow the mind
with the ability unfailingly to sort the true impressions from the false ones. But
such a mind would be superhuman; nothing like the human physiology would
be able to support such a powerful mind. Instead nature provided human beings
with the ability unfailingly to distinguish true impressions of a certain kind—
namely, clear and distinct impressions-from all other impressions whether they
are true or not. In this way human beings are not in a position to know all truths
but only those whose truth is guaranteed by clear and distinct impressions. But
then we do not need to know all truths to lead a good life, and the clear and dis-
tinct impressions we receive in the ordinary course of events provide an ample
basis for what we need to know. If our ability to know is restricted this way,
our ability to avoid false belief is unlimited: all we need to do is not to accept
as true any impression that is not guaranteed to be true by clear and distinct im-
pressions. Thus there will be many true impressions which we nevertheless will
not give assent to, but there will be no false impressions which we accept as true.
All this presupposes that there is a class of impressions which by their very
nature cannot be false and that the mind can discriminate between these and
other impressions. Our main task in the following will be to explain how the
Stoics could make these assumptions. To understand this, we first have to have
a closer look at the Stoic doctrine of rational impressions quite generally.
On the one hand, rational impressions are not mere sensory affections. This
distinguishes them from the impressions of irrational animals. There are several
passages according to which the Stoics distinguish rational impressions from
mere sense-impressions (cf. Cic. Acad. II 21; SE M VII 345). Even the most
primitive rational impression, like the impression that this is white, already in-
volves the representation of the object by means of a concept, in this case the
concept "white." It is in this way that they require a rational mind and manage
to be thoughts and to have a definite prepositional content. Sometimes commen-
tators talk as if we applied concepts to objects on the basis of impressions which
in themselves are preconceptual. But this cannot be the way the Stoics think of
rational impressions. For given an impression which does not yet involve the
conceptualization of the object, we could have any number of beliefs about the
objects on the basis of such an impression. Hence there would not be any one
definite proposition that forms the content of the impression, and assent to the
impression would not constitute a definite belief. It may be objected that impres-
sions are supposed to be passive affections of the mind, whereas the mind's con-
ceptualization of an object would be an active contribution of the mind to the
impression. But it has to be kept in mind that the Stoics characterize an impres-
sion as a passive affection of the mind to contrast it with the act of assent and
not to deny that the mind has any part in the formation of a thought. As we shall
see, the Stoics think that the kind of impression which we have very much de-
pends on whether our mind is in normal working order; and part of what an ob-
ject does, when it gives rise to an impression in a rational mind that is in working
order, is that it makes the mind conceptualize the object in a certain way. In this
sense the rational impression is a passive affection of the mind, though it does
involve the operation of the mind. It will also be objected that impressions only
give rise to concepts and hence cannot themselves already presuppose concepts.
This objection overlooks the developmental character of the Stoic account. Hu-
man beings, according to the Stoics, start out as irrational animals. As such they
have the kind of sense-impressions which animals have. But in the case of human
beings these impressions give rise to concepts of very simple perceptual features
like colors, shapes, tastes, and the like, and thus reason slowly starts to grow.
Once we have these simple concepts, we can have corresponding rational im-
pressions and, what is more important, corresponding cognitive impressions.
These will naturally give rise to more complex concepts, like that of a man or
a tree, which in turn will enable us to have more complex rational and in particu-
lar cognitive impressions (cf. Cic. Acad. II21). Thus these common notions that
arise in us naturally on the basis of more primitive impressions turn out to be
truly anticipations (Cic. nd I 44; prolepseis); for they are needed to form the
impressions that afford us a grasp on things (katalepsis); it is in terms of them
that the mind has a grasp on things. Thus rational impressions and in particular
cognitive impressions do presuppose concepts, but these arise from more primi-
tive impressions that do not presuppose these concepts, and ultimately from
sense-impressions that do not presuppose any concepts whatsoever but that are
not rational either. Given this developmental account, it is easy to see how the
Stoics can claim that concepts only arise from the appropriate impressions and
nevertheless maintain that a rational impression involves the conceptualization

of the object.
On the other hand, there is more to a rational impression than just the
prepositional content. We cannot identify an impression by just specifying the
proposition it is a thought of. To have a rational impression is to think a certain
proposition in a certain way. The kind of impression we have depends not only
on the prepositional content, but also on the way in which this content is thought.
For the same proposition may be thought in any number of ways, and depending
on the way it is thought we get different kinds of impressions. One way they
differ is the way in which the subject of the proposition—that is, the object of
the thought-is represented in the impression. The thought that this (a book in
front of me) is green which I have when I look at the book differs considerably
from the thought that this (the very same book) is green which I have when I
close my eyes and touch the book, though the propositional content, at least in
Stoic logic, is exactly the same. The thought that John's cat is gray is quite differ-
ent depending on whether I see the cat or whether I am just told that John bought
a gray cat, though, again, the propositional content may be exactly the same.
But thoughts may also differ in the way in which the feature that is attributed
to the object is represented. I may be in the habit of thinking of death as some-
thing bad and dreadful, in which case it would be a pain for me to accept the
thought that I am dying. If, on the other hand, death is matter of indifference
to me, the thought that I am dying would be a rather different kind of thought,
whose acceptance would not be a pain. In fact, the Stoics seem to think that all
emotions and passions are a matter of accepting thoughts thought in a certain
way, and that the way these thoughts are thought is entirely a matter of certain
further beliefs we have—in particular, beliefs about what is good and what is
bad—which we draw on to represent the object of the impression and the feature
attributed to it in the thought. Thus all contents of the mind turn out to be
thoughts. And it becomes even more apparent why the Stoics should be so con-
cerned with our ability to distinguish between true and false impressions; for on
this view even our feelings and desires turn out to be nothing but a matter of
accepting true or false thoughts of a certain kind.
For our purposes one difference in the way objects may be represented in our
thoughts deserves special emphasis. If one perceives an object, it tends, at least
under normal conditions, to be represented in one's thought in such a way that
just on the basis of this very representation one could go on to say lots of things
about the object in addition to what one thinks about it, and these things that one
could say about it may or may not be things one antecedently believed to be true
of the object. In cases in which one neither is perceiving the object nor even has
perceived it, the object will be represented in one's thought entirely in terms of
what one antecedently believed to be true of it. And thus it will be represented
in terms of general concepts each of which might equally apply to other objects.
But if I see the object and think that it is green, the object may not be represented
by general concepts at all, except for the concept "green," though it will be
represented in such a way that, just on the basis of the impression, we could go
on to represent it in terms of a host of concepts.
From what has been said it should be clear that there is some sense in which
impressions have parts corresponding to the various features that are rep-
resented in the impression—more particularly, a part or parts corresponding to
the features in terms of which the object of the thought or the subject of the
proposition is represented, and a part or parts corresponding to the feature or
features the object is represented as having—that is, a part or parts that corres-
pond to the predicate of the proposition the impression is a thought of. The
Stoics seem to be willing to call such parts of impressions "impressions," too.
For they call general notions "impressions" (SE M VII 246; Plut. Comm. not.
1084F; cf. Cic. Acad. II 21). But this seems to be misleading, since parts of
impressions are not true or false in the way impressions properly speaking are.
Hence it might be better to call such parts of impressions "ideas" and to
distinguish the way ideas have a prepositional content and are true or false from
the way impressions properly speaking are prepositional and true or false. The
Stoics also seem to distinguish between generic, or abstract, and specific, or
concrete, ideas (cf. SE M VII 246). The idea of man in general, for example,
is abstract, whereas the idea of Socrates and the idea of his complection may
be specific, or concrete. The fact that we represent an object in an impression
by means of a general concept is reflected by the fact that the corresponding
part of the impression is an abstract idea. Moreover, we have to assume that
the parts of rational impressions are ordered so that their combination in the
appropriate order amounts to the thought of a proposition, whereas their com-
bination in a different order might amount to the thought of a different proposi-
tion or to no thought at all.
To sum up: impressions are impressions of an object; in the case of rational
impressions this impression consists in a thought concerning the object; such a
thought involves the conceptualization of the object, but it need not be, and in
the case of perception is not, entirely conceptual; nevertheless, the thought is
the thought of a proposition; but it is characterized not only by the proposition
it is a thought of, but also by the way this proposition is thought; the way a
proposition is thought depends on the way the constituents of the proposition are
represented in the thought; this representation does not have to be entirely con-
ceptual—that is, it does not have to consist entirely of abstract ideas—in order
to represent a constituent of a proposition and in order to be constitutive of a
thought; in the case of perception the thought is partly nonconceptual; it
nevertheless is a thought, because it does involve the conceptualization of the
object, and in particular because it minimally involves the kind of conceptualiza-
tion of its object which gives it a prepositional content that is true or false, as
a result of which the thought itself can be said to be true or false.
Cognitive Impressions
How could there be impressions that cannot fail to be true, not for the trivial
reason that they are true or correspond to the facts, but because of some other
feature that is logically independent of their truth? It seems that there could be
such a feature, namely the property of having a certain kind of causal history,
and that the Stoics are relying on this feature.
Impressions have a certain causal history. In the course of this history all
sorts of things can go wrong. The mind, for example, may be defective and
hence produce the wrong impression. In the case of vision the light may be
wrong, the distance too big or too small, the sensory apparatus malfunctioning,
and as a result we may get a false impression. On the other hand, it stands to
reason that nature has constructed things in such a way that under normal condi-
tions the impression we receive is true. If under normal conditions something
appears to be red or appears to be a human being, then it is red or is a human
being. Thus impressions with the right kind of history cannot fail to be true,
though the fact that they have this kind of history is logically independent of their
truth. Let us call such impressions "normal."
There are different kinds of normal impressions. In particular it seems useful
to distinguish two kinds. If, for example, I have the impression that 2 + 2 =
4 because I have a proof for the proposition that 2 + 2 = 4, my impression will
have the right kind of causal history that will guarantee its truth. But it is not
a causal history that links the object of the impression, say the number 4, with
my impression; the impression, though produced in an appropriate, normal way,
is not produced or caused by the object of the impression itself. It is, at least
according to the Stoics, only in cases of perception that the normal impression
is caused by the object itself. Hence it will be useful to treat normal impressions
of this particular kind as a separate class and to call them "perceptual im-
pressions."
That the Stoics think of cognitive impressions as normal is suggested by the
following. Sextus Empiricus (M VII 247) characterized noncognitive impres-
sions quite generally as those one comes to have because of some abnormal con-
dition (pathos). "Abnormal condition" here can hardly refer just to abnormal
states of mind; for even in a normal state of mind one will have noncognitive
impressions—for example, if one is seeing something from too far away. Hence
"abnormal conditions" here has to be understood as referring to a whole set of
normal conditions. And in SE M VII424 we are in fact given such a set of condi-
tions for the case of vision. Five conditions have to be met for a visual impres-
sion to be cognitive: conditions on the sense organ, on the object of vision, on
how the object is placed, on how the impression comes about, and on the state
of mind. And though this is not said explicitly, it is strongly suggested that if
these conditions are met, the impression will be cognitive. Similarly Cicero
(Acad. II19) refers to such a set of sufficient and necessary conditions for cogni-
tive impressions.
Moreover, though this is a matter of considerable controversy, it also seems
that the Stoics think of cognitive impressions as perceptual. Aetius (Plac. IV
8.1) explicitly says that cognitive impressions come about through a sense or-
gan. Cicero talks of cognitive impressions as if they originated in the senses
(Acad. II 83). And the way the Stoics define cognitive impressions (they are sup-
posed to arise from an object) and what they have to say about the clearness and
distinctness of impressions make straightforward sense only for perceptual im-
pressions.
What seems to stand in the way of this assumption is the following. The
Stoics clearly assume that there are nonperceptual cognitions, namely, in those
cases where we have a proof of a theorem (DL VII 52). But it is also the case
that according to the Stoics even nonperceptual cognitions involve impressions.
(SE M VII370). Hence, it seems natural to assume that the impressions involved
in cognitions, whether they are perceptual or not, are cognitive. Moreover, there
are texts which claim that a cognition consists in the assent to a cognitive impres-
sion (SE M VII 151; VIII 397). Hence, if there are nonperceptual cognitions,
there should be nonperceptual cognitive impressions. Finally, cognitive impres-
sions are supposed to be the criterion of truth. Whatever else this may mean,
it must mean that the truth of cognitive impressions is the guarantee of the truth
of whatever impressions the wise man accepts as true. But if we restrict cogni-
tive impressions to perceptual impressions, it is difficult to see how their truth
would suffice as a basis to guarantee the truth of all other impressions the wise
man will accept as true.
To deal with the last point first, we have to take into account that the Stoics
seem to think that all features of objects—that is, of sensible bodies—are per-
ceptible. Thus they think that we can even learn to see that something or some-
body is beautiful, good, or virtuous (Plut. Comm. not. 1062C; Stoic, rep.
1042E-F; Cic. ND II 145), just as we have to learn to see that something is a
man or a horse (Cic. Acad. II 21). If this at first sight seems strange, we have
to remember that according to the Stoics, qualities of bodily objects like virtue
are bodies themselves that form a mixture with the bodies they are the qualities
of and hence cannot fail to affect our perception of the objects, given that our
perception, at least if trained, is extremely discriminatory; a virtuous body must
look quite different from a vicious body to a trained eye. Thus perception, as
the Stoics understand it, provides a much broader basis than we would assume.
And it will also turn out, when we consider the doctrine of the criterion, that
the Stoics do in fact think that all other impressions can be accepted as true to
the extent that their truth is guaranteed by the truth of perceptual impressions.
Thus Cicero (Acad. II 21-22) points out that at some time in our development
we come to have the (nonperceptual) cognition that if something is a man, it is
a mortal rational animal. But when he explains why this cannot but be true, he
does not say that the corresponding impression is cognitive; instead he says that
it cannot be false because it is due to impressions that cannot be false, namely,
cognitive impressions that are perceptual.
Once we realize that all truths available to us are supposed to be certified by
the truth of perceptual impressions, it seems fairly clear that our problem about
the scope of cognitive impressions is not so much a problem concerning Stoic
doctrine but rather a problem concerning terminology. In fact, it is rather simi-
lar, and materially related, to the problem which we have about the scope of
"clear" or "evident" and which it seems best to solve by distinguishing between
self-evident impressions and impressions whose evidence depends on the evi-
dence of other impressions. Similarly, it seems that the Stoics take the view that
only perceptual impressions are cognitive in their own right. Thus other impres-
sions can be called cognitive only to the extent that they have a cognitive content
which depends on the cognitive content of impressions which are cognitive in
their own right. Thus we may distinguish between self-evident impressions
which are cognitive in a narrow sense, and evident impressions which are cogni-
tive in a wider sense. And if we do so, we can say with Sextus Empiricus that a
cognition consists in the assent to a cognitive impression, and we can also say that
any cognition, whether perceptual or not, involves a cognitive impression, and
nevertheless assume that cognitive impressions, strictly speaking, are perceptual.
Perceptual impressions, in addition to being normal and hence true, have cer-
tain other features that are of interest for our purposes. In the case of perceptual
impressions, the impression represents the object the way it does because the ob-
ject is this way—that is, all representational features of the impression are due
to the object and not to some abnormal condition that would cause the mind to
produce an impression different from the one it would produce normally. Thus
a perceptual impression in no way misrepresents its object. But considering the
purpose for which we have been endowed with cognition, it also stands to reason
that nature has constructed things in such a way that under normal conditions
we not only have an impression which does not misrepresent things but have one
which represents them clearly, that is, affords us a clear answer as to what kinds
of objects we are facing. And under normal conditions we do in fact have a clear
view of an object we are confronted with, and we can tell without difficulty what
its visual features are. Let us call such an impression "clear" or "evident."
The term "evident" has been used, misused, and misunderstood in many
ways. To guard against such misunderstanding of the Stoic position some re-
marks may be in order. The adjective "evident" (enarges) can be used in ordi-
nary Greek to qualify a term "A" to refer to something as being obviously an
A; thus an evident robber is somebody who quite obviously is a robber (Soph.

O. T. 535). But even in ordinary Greek the term can be used in cases in which
appearances are deceptive; the evident ox may not be an ox at all, but Zeus in
disguise (cf. Soph. Tr. 11). Things also can be said to evidently appear to be
a certain way. And hence it is easy for philosophers to move on to talk of evident
appearances or evident impressions, though by this they obviously do not mean
to suggest that some of our impressions are such that it is evident they are im-
pressions. This move must have been facilitated by the fact that even in ordinary
Greek, dream images can be said to be evident (Aesch. Pers. 179). Given the
ordinary use of the term, evidence suggests but does not guarantee truth. Thus
Platonists (cf. SE M VII 143) and, of course, Academics (cf. Cic. Acad. II 34)
do not take evidence to be a criterion of truth. Theophrastus, on the other hand,
seems to have been the first philosopher to assume that it does guarantee truth
(cf. SE M VII 218), and in this he was followed by the Epicureans and the
Stoics. Since they cannot rely on ordinary usage for this assumption, we have
to look for some argument that would justify this restricted use of the term "evi-
dent" or the assumption that even given the ordinary use it turns out that only
true impressions are evident. The Stoics may have argued along the following
lines: we can learn to see whether something is an ox or a robber; and under
normal conditions, if nothing impedes our seeing things clearly, we do see
whether something is an ox or only an ox in disguise; for the only things that
can really look and move like oxen are oxen; thus something cannot be an evi-
dent ox without being a real ox. For it could appear to be an ox without being
one only if we had not yet learned to see oxen properly or if our view was some-
how impeded because one of the normal conditions was not met; but in this case
the ox would not be evident. Evidence is an objective feature of impressions
which is not to be confused with a subjective feeling of conviction or certainty,
however strong that feeling may be, just as having a clear view of something
is a matter of objective fact and not of subjective feeling. How we know that
an impression is evident is a different matter, to which we will turn later; for
this, our "feeling" may very well be relevant, but it seems, even in optimal cir-
cumstances, to be no more than a symptom of the evidence of an impression.
To get clearer about the notion of evidence which is in question here, it may
be useful to consider the connection between truth and evidence. Impressions are
true, because their prepositional content is true, and not because of the way this
prepositional content is thought, that is, represented in the impression. The same
propositional content, as we have seen, can be thought in all sorts of different
ways, and correspondingly we get different kinds of impressions; but this differ-
ence between the impressions is of no relevance of their truth, which entirely
depends on the truth of the proposition. Evidence, on the other hand, is primar-
ily a feature of impressions which does depend on the way a proposition is repre-
sented by thought. Thus the same proposition that this is octagonal can be
thought by an evident thought when I see an octagonal tower under normal con-
ditions, and by a nonevident thought, if I just know from a book that the tower
is octagonal. Propositions only secondarily may be called evident, if there
should be any propositions that cannot be thought at all except by evident
thoughts. What makes a thought or an impression evident is that it is already
part of the representation of the subject of the proposition that the predicate
should be true of it and that the representation of the subject is entirely due to
the subject itself. Thus evidence is not what makes an impression or a proposi-
tion true, but an evident impression cannot but have a true proposition for its
content and hence be true itself.
So far it would seem that for the impression that S is P to be evident, the
representation of S already has to represent S as being P. But it seems that under
normal conditions, when we have a clear view of an object, more than one of
its features is clearly represented. And, in fact, Sextus (M VII 248, 250, 251)
talks as if a cognitive impression captured all the characteristics of the object
in precise manner. Cicero, on the other hand, explains that a cognitive impres-
sion does not pick up all the features of an object, but only all those features
which are appropriate for its kind, visual features in the case of vision, auditory
features in the case of hearing, etc. (Acad. I 42). Since even the weaker claim
is extraordinarily strong, it will be safer to follow Cicero. In this case a cognitive
impression will be evident in that it involves a representation of the object which
clearly represents all the features of the object that are appropriate for the kind
of impression it is; and since it represents all the features of the object in ques-
tion, it will also represent the particular feature which it represents it as having,
that is, the feature attributed to it in the proposition.
Cognitive impressions are not only clear, as opposed to obscure (amudros;
cf. Alex. Aphrod. De an. 71.5ff.), they also are distinct (ektupos; cf. DL VII
46), as opposed to confused (sugkechumenos; cf. SE M VII 171). To see what
their distinctness is supposed to consist in, it will be useful to refer to a doctrine
which is never explicitly attributed to the Stoics but which we do find in Hellen-
istic dogmatic medicine and of which we have some reason to believe that it is
in part of Stoic origin. According to this doctrine, the discriminatory power of
the senses far outruns the ability of the mind to conceptualize the object. Thus,
if under normal conditions we see an object clearly, its features are represented
in the impression in such detail that our concepts do not capture them in all their
detail. Hence, though a normal impression, as a rational impression, has a
propositional content, the way it represents the subject of the proposition cannot
be exhausted by any number of propositions (cf. Gal. De he. aff. VIII 86.12ff.,
87.4, 117.6 339.13, 355; Depraesag. ex puls. IX 366.10K; De sanit. tuenda,
CMC V 4.1, p. 185, 16). Now the Stoics assume that the properties of bodies
themselves are particular (Cic. Acad. II 56). Hence they are called "idibmata,"
that is, properties (SE M VII 248). And they seem to be particular not in the
sense that Socrates' wisdom is Socrates' wisdom rather than Plato's wisdom, but
in the sense that they are qualitatively different individuals. After all, on the
Stoic theory, Plato's wisdom and Socrates' wisdom quite literally are two partic-
ular bodies, which, by the law of the identity of indistinguishables which the
Stoics adhere to, should be internally distinct and not just differ in their relational
properties. A property, given its intimate connection and interdependence with
the whole body it is the property of, cannot but take a certain form reflecting
the idiosyncrasy of the object and hence be peculiar to it. Moreover, both Sextus
and Cicero emphasize the artistic precision with which the features of the object
are represented in a cognitive impression down to their last detail (SE M VII
248, 250-251; cf. "subtiliter impressa" in Cic. Acad. II 34). Hence a cognitive
impression of an object will involve a representation of this object which is so
articulate that the only object which will fit this representation is the very object
the impression has its origin in (cf. SE M VII 252). This feature of cognitive
impressions, that they represent their objects in such detail as to fit only them,
is their distinctness. Since the Stoics assume that clear impressions represent all
the relevant features of an object, cognitive impressions will be highly distinct.
Now normal impressions in general and perceptual impressions in particular
have been characterized in such a way that their normality or perceptuality is
a relational feature of these impressions, a feature which these impressions do
not have by themselves, but only in virtue of the fact that they stand in a certain
relation to the world. Hence it would seem that to determine whether an impres-
sion is cognitive or perceptual it will not suffice just to consider the impression
by itself; we also have to consider its relation to the world.
But the Stoics also seem to assume that cognitive impressions by themselves
differ from all other impressions, that there is some internal characteristic that
serves to mark them off from other kinds of impressions and allows the mind
to discriminate between cognitive and noncognitive impressions without having
to consider their relation to the world (Cic. Acad. I 41). Cognitive impressions
are supposed to differ from noncognitive impressions in the way in which horned
serpents differ from all other kinds of snakes, that is, by some internal differen-
tiating mark (SE M VII252). The reason the Stoics postulate such a mark is easy
to see. All the mind has to go by is its thoughts or impressions. If there is not
a privileged set of impressions which we can rely on to be true, we shall be
reduced to considerations of plausibility and coherence, to inferences to the best
available explanation for our impressions, to decide which of them to accept as
true and which to reject as false or to suspend judgment on. But even in the best
of all circumstances such considerations could not fail to occasionally produce
wrong conclusions, and there is nothing to guard us against the possibility that
they generate conclusions which are so much off the mark that they would dis-
rupt our life radically. But the Stoics want to argue that we are entirely responsi-
ble for our life and for that reason nature has put us into the position to avoid
any false beliefs at all. And the only way to do this, it seems, is to provide us
with impressions which cannot but be true and which we can discriminate.
Most of us will be thoroughly disinclined to believe that there is such a
qualitative difference between our impressions. But one has to keep in mind that
its postulation fits in with Stoic physics without any difficulty. Given that accord-
ing to Stoic physics all states of the world and all parts of a state of the world
are closely interdependent, any variation of the conditions under which an im-
pression arises should affect the impression itself. Thus the assumption that nor-
mal impressions have a distinctive character seems not to be ad hoc but to be
required by Stoic physics anyway. Even if this were not so, it would not be much
of a problem for an omniscient nature to ensure that only impressions which
have a normal history have a certain distinctive character which is the effect of
the kind of history they have. Moreover, we have to take into account that it
is part of the Stoic position that we are so corrupted that we tend to give assent
to and to act on cognitive and noncognitive impressions rather indiscriminately.
Hence our awareness of their difference is not just seriously retarded but also
very distorted. And in any case it does not follow from the fact that we have
such difficulties in telling whether an impression is cognitive or not, that there
is no clear difference between them. Finally it has to be kept in mind that Plato
and Aristotle had already made very strong claims regarding the power of the
knowledge they attributed to the wise man; the man of practical wisdom is al-
ways right in practical matters. The Stoics explicitly refer to this Aristotelian
doctrine (Pap. Here. 1020, col. 1 n., SVFII, p. 41, 25), and they just seem to
try to provide a theory that would explain how the wise man might manage to
invariably get things right. If one gives up this conception of the wise man, one
will, of course, not have the motivation the Stoics had to resort to such a strong
assumption. But this conception of wisdom was too firmly embedded to be given
up lightly in the face of epistemological difficulties.
Stoic Definitions of Cognitive Impressions

On the basis of what has been said, it should be relatively easy to understand
the force of the Stoic definitions of cognitive impressions. These come in basi-
cally two versions. In a shorter version, which we find in DL VII 46 and SE
M XI 183, cognitive impressions are defined by two clauses, whereas on the
other, more common, version a further clause is added to the two clauses of the
shorter version. There may be some truth in Cicero's claim (Acad. II 77) that
the shorter definition is the one Zeno originally gave, before he went on to add
a third clause to avoid an Academic objection, especially since this notice gets
some support from Sextus's remark (M VII 252) that the Stoics added the third
clause only to block an Academic objection based on an assumption which the
Stoics did not share.
Let us, then, first consider the definition in its shorter version. To follow the
formulation in DL VII46, an impression is cognitive exactly if (i) it comes about
from what is (apo huparchontos) and (ii) it is imprinted and impressed in exact
accordance with what is. Though this is by no means obvious from the formula-
tion of the second clause by itself, Sextus's comments on this clause in M VII
250-251 show that it is supposed to amount to the requirement that the impres-
sion be clear and distinct. And this interpretation is borne out by the characteri-
zation of noncognitive impressions which in DL follows immediately on the def-
inition of cognitive impressions. According to this definition an impression is
noncognitive if "it either is not from what is or, though it is from what is, is not
in exact accordance with what is; one which is not clear nor distinct." Here the
phrase "one which is not clear nor distinct" looks like a gloss on "is not in exact
accordance with what is," that is, the negative counterpart to the second clause
in the definition of cognitive impressions. And if this is correct, the second
clause of the definition of a cognitive impression should amount to the require-
ment that cognitive impressions be clear and distinct.
It is tempting to think that the first clause amounts to the requirement that a
cognitive impression have its origin in a real object rather than some disturbance
or affection of the mind, that the object the impression presents itself as an im-
pression of be a real object rather than a mere figment of the mind. And this
seems to be the way Sextus interprets the clause, as one can see from his com-
ments on the first part of the second clause in M VII 249. Nevertheless the in-
terpretation of the first clause has been the subject of considerable controversy,
which mainly turns around the force of the term "what is" (huparchori). It has
been pointed out that Cicero in this context again and again renders "huparchon"
by "what is true," that is, understands "what is" in the sense of "what is the case"
(cf. Acad. II42, 112), and that the Stoics do use "to huparchori" for a true propo-
sition. Against this it has to be remembered that there is a great number of exam-
ples in which Sextus talks of impressions that have their origin in something or
other, and that in his examples the something or other in question never is a
proposition but always a real or a fictional object.
Nevertheless, there is some reason to think that Cicero's rendering is not a
mere mistranslation or due to misinterpretation, but rests on the correct assump-
tion that the first clause was not meant to amount to the requirement that the im-
pression should have its origin in a real object, but to the stronger requirement
that it be altogether true. That this assumption may be correct is suggested by
M VII 402ff. There Sextus, following Carneades, argues that there are impres-
sions which have their origin in what is not, but which present themselves as
impressions which have their origin in what is just as much as purportedly cogni-
tive impressions do. And as an example of such an impression Sextus adduces
the case of Heracles who in his madness took his own children to be those of
Eurystheus. Heracles here is explicitly said to have an impression that has its
origin in his own children who are standing in front of him, that is, in a real
external object. And nevertheless this impression, too, is supposed to be an ex-
ample of an impression that has its origin in what is not. The reason for this must
be that the impression is false in that it represents Heracles' children as being
Eurystheus's children. In what sense could such an impression be said to have
its origin in what is not? The answer seems to be that the impression does not
as a whole have its origin in what is; part of it—namely, the part that represents
Heracles' children as being Eurystheus's children—is made up by the mind and
is not due to the object. We saw earlier that it is characteristic of perceptual im-
pressions that all their representational features are due to the object. In this
sense, only true impressions, and more particularly impressions that are true not
by accident, have their origin in what is. If we interpret the first clause in this
way, not only do we not have to assume that Cicero has misunderstood the Stoic
definition, it will also be easy to explain why the third clause of the longer ver-
sion of the definition—which runs, "it is such that it could not come about from
what is not"—is standardly interpreted as meaning that a cognitive impression
has to be such that it could not be false (cf. SE M VII 152, 252; Cic. Acad. II
42, 112). But even a confused and obscure impression may be entirely true and
true not by accident but because all of its representational features are due to
the object that has given rise to it. Hence, to single out cognitive impressions,
the second clause is added. So much about the shorter version of the Stoic defini-
tion of cognitive impressions.
Standardly, though, the Stoics define cognitive impressions by adding a third
clause. A cognitive impression is supposed to satisfy the further requirement that
it be "such that an impression of this character could not come about from what
is not" (SE M VII 248, 252; DL VII 50). This, as we noted above, is taken to
imply that an impression of this character could not be false (cf. SE M VII 152,
252). Given the strong reading of "has its origin in what is," it is easy to see
how the clause would have this implication. The main question concerning the
third clause is the identity of the character referred to. Is this a further charac-
teristic of cognitive impressions which is postulated, but not specified by the de-
finition, or is it the property of satisfying the first two conditions, or is it perhaps
just the property of being clear and distinct? The phrase "of this character" (hold)
is ambiguous in this respect.
Given what we said earlier about cognitive impressions, it seems most plausi-
ble to take this to refer to the distinctive inherent feature that cognitive impres-
sions are supposd to have. And this seems to be confirmed by remarks in Sextus
(M VII 252) and in Cicero (Acad. II 77) which suggest that the Stoics think that
any impression which satisfies the first two conditions will in fact also satisfy
the third condition, but that they add the third clause because this implication
is denied by the Academics, though both agree that cognitive impressions, in or-
der to play the role assigned to them by the Stoics, would have to satisfy the
third condition, too. And this dispute about the third clause turns out to be a dis-
pute about whether cognitive impressions have an internal differentiating feature
(cf. SE M VII252). Hence, it would seem that the third clause refers to this dis-
tinctive feature of cognitive impressions, which is postulated, but not specified.
The Criterion
To get a clearer notion of this feature and of the role it is supposed to play
in cognition, it will be useful to briefly consider in which way cognitive impres-
sions are supposed to constitute the criterion or canon of truth. We have already
seen that they are not a criterion of truth in the sense that they put us in a position
to determine the truth of any proposition whatsoever. There are lots of proposi-
tions that cannot be certified by them. Nor are they the criterion of truth in the
sense that whenever the truth of a proposition is in question, we at least consider
the corresponding impression and try to determine by introspection whether it
has the distinctive mark of a cognitive impression. There are several reasons
why this can hardly be the Stoic view of the matter.
First of all, cognitive impressions will directly guarantee only the truth of
their own prepositional content. And if it is true that cognitive impressions are
perceptual, the only propositions whose truth they can guarantee directly are
propositions that attribute a perceptual feature to a particular object. If they
nevertheless are called the criterion of truth, it is because in an indirect way they
also guarantee the truth of all other propositions that are known to be true by
human beings. And they do this in the following way. They give rise to general
ideas, the so-called common notions which the mind forms naturally on the basis
of cognitive impressions and which in turn allow us to have further cognitive
impressions. And since cognitive impressions do represent things as they are,
the common notions based on them will represent things as they are. Thus if the
common notion of a man represents a man as a biped rational animal, the propo-
sition that man is a biped rational animal will be certified not by an impression
that man is a biped rational animal, which is cognitive in its own right, but by
the common notion, and this in turn will be certified by the cognitive impressions
which give rise to it and which it gives rise to, and these will be cognitive in
their own right (cf. Cic. Acad. II 22). And the truth of propositions certified by
cognitive impressions and of propositions certified by common notions in turn
will guarantee the truth of further propositions derived by deductive inference
from the former propositions. It is for this reason that Chrysippus sometimes can
say that perceptions and common notions constitute the criterion (DL VII 54).
Cognitive impressions, then, are the criterion of truth in the sense that their truth
guarantees the truth of whatever can be known by human beings. It is only
through them that we have any knowledge of what is true and what is not true.
Second, we have to remember that there is no such thing as the impression
that corresponds to a given proposition, and, therefore, when the truth of a prop-
osition is in question, we may have to go through a number of impressions all

of which have the proposition in question as their propositional content till we
hit upon a cognitive impression. Thus we may not be certain of the color of an
object we see in the distance. As we move nearer we have a series of different
impressions which all may be impressions that the object is blue. Similarly in
the case of a theoretical problem our impression of the proposition in question
will change as we consider the matter. The impression we have when we have
a proof for a proposition is quite different from the impression that we had to
start with. Thus cognitive impressions cannot be the criterion in the sense that
we just have to look at our impressions to determine whether a proposition is
true. It is, rather, by considering the proposition that we may get a clearer and
clearer impression.
Most important, though, we have to avoid thinking of Stoic impressions as
pictures or images of the world which can be looked at introspectively, with the
mind's eye, as it were, to see whether they have this feature that guarantees their
truth. What we see and grasp, according to the Stoics, are objects in the world,
and not pictures or images of them, though grasping objects does involve the
awareness of their representations in the mind, just as it involves an awareness
of the mind itself. For we have to take into account that impressions for the
Stoics are mental states that are identified as highly complex physical states, as
we can see from the fact that originally they were conceived of quite literally
as imprints. When Chrysippus objected to this, it was because he thought that
they were much more complex than the term "imprint" suggested; in calling
them "alterations" or "modifications" (cf. SE M VII 229-230; VIII 400; PHII
70; DL VII 50) of the mind instead, he deliberately, it seems, left open what
their precise nature consists in. There is no suggestion that we could observe
them to find out exactly what they are like. It is, of course, true that the Stoics
think that impressions reveal themselves along with the object they are impres-
sions of (Aetius, Plac. IV 12.2). But all that this means is we can tell what our
impressions are; after all, they are our thoughts. But we do not know our
thoughts by introspection, nor is there any reason to believe that the Stoics think
so. Moreover, if the Stoics thought that we could see by introspection whether
an impression has the distinctive feature of a cognitive impression, we would
expect them to say, at least on occasion, that the criterion of truth is this feature.
But they never say anything of this sort. Also, if they had taken this view, they
would have opened themselves to the charge of an infinite regress. For we would
have to ask what is supposed to guarantee the truth of the impression that a given
impression has this distinctive feature. Quite generally, the criterion will fulfill
its role only if it does not require the judgment that an impression is of a certain
kind. For this will always raise the question how this judgment is to be certified.
The Stoic theory, I want to suggest, escapes this difficulty because it assumes
that the distinctive feature of cognitive impressions is a causal feature of impres-
sions such that cognitive impressions play their criterial role not through our
awareness of their distinctive feature, but through the causal effects they have
on our minds in virtue of this feature. The word "to discriminate" is ambiguous.
It is used in cases in which one recognizes things to be of different kind and,
in virtue of this awareness of the difference, treats them differently. But there
are also cases in which somebody reacts differently to things of a different kind
not in virtue of an awareness of their difference and perhaps even without know-
ing that there is such a kind of thing which he systematically reacts to in a dis-
tinctive manner; there is a causal link between a feature of the object and the
behavior of the person, but the awareness of the feature on the part of the person
is not an essential part of the causal chain; and nevertheless such a person can
be said to discriminate or to discern the feature. Many forms of discrimination
in the pejorative sense are of this kind. The suggestion, then, is that the distinc-
tive mark of cognitive impressions is a causal feature in that it makes the mind
react in a distinctive way and that it is in this sense that the mind can discriminate
cognitive and noncognitive impressions. It can also learn to tell whether an im-
pression is cognitive or not, but that is a different ability not at issue at this point
in our argument.
What reason do we have to think that this is the Stoic position? The Stoics
assume that cognitive impressions give rise to common notions. Common no-
tions have their privileged status exactly because the mind forms them naturally
on the basis of cognitive impressions. Nobody, so at least the Stoics think, can
help but end up with the notion of a tree and the notion of a human being and
the notion of the color green if he grows up normally in a normal environment.
This formation of common notions is not something we engage in deliberately
according to certain rules and precepts; if we did, we could make mistakes and
end up with the wrong notions. The Stoics clearly assume that the mind sorts
out cognitive impressions to form concepts on the basis of them without our be-
ing aware of this at all; we just find ourselves having certain concepts that we
did not have to start with. Thus the Stoics also must assume that the mind can
discriminate cognitive impressions without our being aware of it.
We also have to find some explanation of the fact that the mind gives assent
to some impressions but not to others. As soon as the mind has acquired all sorts
of beliefs, it is easy to see how it would accept or reject impressions against the
background of the beliefs it already has. But in the beginning, it would seem,
the mind has no more reason to accept than not to accept any given impression.
This problem would be solved if we assume that cognitive impressions cause the
mind to accept them. And there is some evidence, though by no means decisive,
that this is in fact the Stoic position (cf. SE M VII 405, 407; Cic. Acad. II 38;
Plut. Adv. Colot. 1121E, 1122C). This is perfectly compatible with the further
Stoic claim that we are responsible for our acts of assent, for it is explicitly not
part of the Stoic doctrine of responsibility that we are responsible only for those
things which we could have done otherwise. But it clearly cannot be the Stoic
view that we acquire our first beliefs by scanning our impressions and by being
caused to assent to those which we detect to have the distinctive remark of cogni-
tive impressions. It, rather, must be the case that the Stoics assume that the mind
does this without our being aware of it.
Moreover, the Stoics point out (SE M VII 258; Cic. Acad. II 19) that if we
do not have a clear impression we take the appropriate steps to receive an evi-
dent impression, in case the subject is of any importance to us; that is, not having
a clear impression naturally makes us consider the matter further till we have
a clear impression. The suggestion does not seem to be that we recognize that
our impression is confused and obscure and hence decide to get a clearer one,
but, rather, that there is a causal mechanism that sets us going and would natu-
rally make us stop once we had a clear impression. For these reasons, then, it
seems that the differentiating mark of cognitive impressions is a causal feature
rather than a phenomenological character to be detected by introspection.
But this is not to say that we cannot be aware of the fact that an impression
is cognitive or noncognitive, that we cannot learn to tell whether an impression
is clear and distinct or obscure and confused. In fact, the Stoic view seems to
be that this is a matter of practice and that in principle one can get so good at
it that one will never take a noncognitive impression to be cognitive. But to learn
this is not to acquire a mysterious sixth sense which, unlike the other senses,
is not subject to the possibility of abnormal conditions and hence unfailingly
gives us notice of an equally mysterious feature of cognitive impressions. Judg-
ments regarding the evidence of an impression are notoriously as fallible as any
other judgments, and there is no reason to saddle the Stoics with the assumption
that this is not so. But we can get better and better at seeing how variations in
the conditions under which our impressions arise, especially variations in our
mental state and the beliefs we have, do affect our impression.
Cognition, Knowledge, and the Wise Man

Whereas their predecessors had distinguished only between knowledge and
mere opinion, the Stoics distinguish between knowledge, cognition, and mere
belief (SE M VII 15Iff.). Cognition consists in the assent to, or acceptance of,
the appropriate kind of impression, that is, an impression that is at least cogni-
tive in the wider sense. A mere opinion, on the other hand, even if it is true,
may or may not involve the appropriate kind of impression; if it does, it is also
a cognition; but whether it does or not is not what one focuses on when one calls
it an opinion. Knowledge differs from cognition in that it involves not only the
appropriate kind of impression but also the appropriate kind of assent—namely,
the kind of firm assent that one cannot be persuaded to withdraw by any argu-
ment to the contrary. This presumably is one reason why we have to try to avoid
having any false beliefs whatsoever. For if we do accept a false premise we
might be led by a chain of reasoning to accept the contradictory of what we had

already believed to be true, even if we had accepted it on the basis of a cognitive
impression. And as long as one is susceptible to this, one's assent will not be
firm. On the other hand, once one has learned to accept true impressions only,
no amount of dialectical skill will suffice to make one withdraw one's assent from
impressions that are cognitive in the wider sense; and then one's assent will be
stable and firm or certain; in this sense of "certain" one will have certain
knowledge.
All cases of cognition are cases either of knowledge or of opinion. For though
they all involve the appropriate kind of impression, they will be a matter either
of opinion or of knowledge depending on whether or not they also involve the
appropriate kind of assent. Nevertheless, there is a point to the distinction. It
emphasizes the fact that the conditions on knowledge are so strong that only the
wise man will have knowledge (SE M VII 152, 432). In fact, his wisdom will
consist in this kind of knowledge. The ordinary person will have nothing but
mere beliefs, for he is not yet able to avoid any false belief and hence his assent
is not yet firm. But it is important that many of his beliefs are at least cognitive.
For they will afford him a basis to acquire the knowledge that constitutes
wisdom.
This view has one consequence that hardly seems to have been noticed, but
which is highly relevant to our topic. For the Stoics also assume that there are
no wise men or at least that not even the members of their own school have at-
tained the blissful state of wisdom (SE M VII 432-33). It immediately follows
that there is no knowledge or at least that the Stoics do not have any knowledge.
And once we realize this, all sorts of Stoic texts with a strong skeptical flavor
come to mind. Thus Seneca (De ben. IV 33.2) says: "We never expect com-
pletely certain cognition of things, since the exploration of truth is extremely
difficult; we follow where likelihood guides us." The Stoic claim is not that they
have attained the knowledge Socrates tried to find, but rather that the knowledge
Socrates was after is attainable by human beings.
The Skeptical Attack

It should be clear, then, that skepticism did not arise as a reaction to overly
confident claims to knowledge on the part of the Stoics. The Stoics were in no
mood to make such claims. But the Stoics did claim some expertise, and on the
authority of this expertise tried to put forth views on the nature and the material
content of the knowledge Socrates had been looking for in vain. Hence the cen-
tral role of the notion of a dogma and the charges of dogmatism in skeptical at-
tacks on Stoicism. Moreover, the view the Stoics did adopt turned out to be ex-
tremely revisionist and literally paradoxical. Thus it would easily occur to one
to subject the Stoic claims to exactly the kind of dialectic that Socrates had used
to test and expose unfounded claims to expertise. And this is precisely what the
skeptics did.
Now there are some crucial features of Socratic dialectic which it is worth-
while to recall if we want to understand the skeptical position. The Socratic
method allows one to test expertise in a subject without being oneself an expert
in this subject—in fact, without committing oneself to or even having any views
on the subject. All one has to do is to show that the person who claims expertise
or makes statements with the air of authority involves himself in contradictions
concerning the very subject he claims to be an expert in or that he is unable to
discard a thesis which is the contradictory of a thesis he has put forth with the
air of expertise. For if he were an expert, he should be able to defend his position
against theses to the contrary, and he certainly should not involve himself in con-
tradictions. Hence such dialectical arguments are not meant to establish the truth
or falsehood of some thesis. All they are meant to show is that the opponent is
no authority on the matters in question.
It is important to keep this in mind, because otherwise one might be misled
into thinking that the skeptics themselves accept either the premises or the con-
clusions of their arguments. Thus one might think that the ancient Academic
skeptic fits the prevailing modern notion of what a skeptic is, in that he believes
that all that is given to us are our impressions and that he tries to convince us
that since this is so, there is no way in which we ever can have certain knowledge
of what the world that gives rise to these impressions really is like. The skeptic
may argue this way, but if he does so, it is just another ad hominem argument
against those who believe that all that is given to us immediately are our impres-
sions. There is no reason why the skeptic himself should feel committed to this
very dogmatic, speculative, unskeptical assumption and the dualism between the
mental and the physical, the subject and the object which tends to go with it.
Thus it is not surprising that in other contexts the skeptic is quite willing to chal-
lenge the dogma of the impression as a given (Gal. De diff. puts. VIII710, 18ff.
K.; Depraenot. XIV 628, 14ff.). He is quite willing to say that some things evi-
dently appear to be the case, as we ordinarily do, but he does not think that this
commits him to the view that there are such entities as impressions, assents, and
evidence. Nor are the skeptics committed to the conclusions of their argu-
ments—for example, the conclusion that there is no knowledge or the conclusion
that nothing can be known, or the conclusion that the wise man will suspend
judgment on all questions. He is not even committed to the view that the conclu-
sions of his arguments follow from their premises. For, as he will emphasize,
he does not subscribe to the canons of logic worked out by his opponents, either
(cf. Cic. Acad. II 91ff.). He is just prepared, for the sake of argument, to meet
whatever standards of logic are met or required by his opponents. For otherwise
his arguments will not have the desired effect on them.
What is the envisaged effect of such arguments? Reporting his experience the
skeptic might say that they tend to leave one with the impression that the Stoics
have not successfully argued their case concerning the nature and the attainabil-
ity of knowledge. They may also tend to leave one with the impression that it
is doubtful whether such a case can be made at all. More generally, it will appear
doubtful whether the case for any revisionist conception of knowledge can be
made; we might just have to accept the fact that all that is available to us is the
kind of everyday knowledge the vulgar have. Even more generally, it may ap-
pear doubtful whether the case for any position can be made. On the other hand,
it would not be desirable, from the skeptic's point of view, if one was left with
the impression that the positions attacked by him are false, or that, even if they
are true, there is no way to definitively establish them as true. This would lead
to a dogmatic pseudoskepticism quite alien to true Academic or Pyrrhonean
skepticism (cf. SE PHI 200, 226, 236; Gal. De subf. emp. 84, 22 D.).
Given the central position of the doctrine of cognitive impressions in Stoic
epistemology, it is not surprising that the skeptics focused their attack on this
doctrine. And here the main point at issue was whether cognitive impressions
differ qualitatively from all other impressions. This, as we saw, is an assumption
so central to the Stoic position that Zeno already added it to his definition of cog-
nitive impressions. The skeptics were quite willing, at least for the sake of argu-
ment, to accept the first part of the definition and to grant that there may be im-
pressions that have their origin in what is and that represent their object faith-
fully and clearly (SE M VII402). But they took issue with the added assumption
that such an impression, just given its internal characteristics, could have no
other origin than the object it faithfully represented, that there could not be an im-
pression exactly like it which was nevertheless false. Already Arcesilaus attacked
the further assumption (Cic. Acad. II77; SE M VII154), Carneades pursued the
same line of attack (SE M VII 164, 402ff.), and it was to remain the main point
of contention throughout the debate (Cic. Acad. II 33, 78; SE M VII 252).
We have only a rather general idea of the form this debate took, since its de-
tails have not been studied with the care they deserve. Apparently, the skeptics
adopted the strategy of arguing for the more general thesis that for any true im-
pression there could be another impression exactly like it which is false (Cic.
Acad. II 40, 41, 42; 44, 84, 90; SE M VII 154, 164, 252, 402, 415, 428), or
at least one which differs so minimally from the true one that we cannot distin-
guish between it and the true one and which, nevertheless, is false (Cic. Acad.
II 40, 85). More particularly, they seem to have argued the matter for the vari-
ous kinds of true impressions, kind by kind (Cic. Acad. II 42). In the case of
cognitive impressions, they did so in at least two ways. To start with, they tried
to show that there are impressions which, as far as their representational features
are concerned, differ in no way, or at least in no discriminable way, from cogni-
tive impressions, though they themselves are not true. But then they also tried
to show that there are impressions which have all the supposed characteristics
of cognitive impressions—which, for example, are vivid or striking, or which
at least could not be distinguished from a cognitive impression by the person
who has the impression at the time he has them, and which nevertheless are false
(SE M VII 408).
Let us first turn to the impressions that are supposed to be exactly like, or
at least indistinguishable from, cognitive impressions in the way in which they
represent their object (cf. Cic. Acad. II 84ff.; SE M VII 408ff.). Suppose that
Socrates is standing in full view in front of one; in this case one may have the
cognitive impression that this is pale or that this is a man or even that this is
Socrates, if one has learned to grasp his Socraticity and has a corresponding idea
of Socrates. Now also suppose that Socrates has a twin brother, whom we do
not know anything about, but who is exactly like Socrates, or who at least looks
exactly like Socrates. In this case, the skeptic rightly claims, the impression one
would have of Socrates' twin brother under identical normal conditions would
be exactly like the cognitive impression one has of Socrates. Hence, he goes on
to argue, it is not the case, as the Stoics claim, that an impression which has
all the characteristics of a cognitive impression can have its origin only in the
object which gives rise to it and that there could not be another impression ex-
actly like it which does not have its origin in this object. Moreover, suppose (i)
that we first see Socrates and have the cognitive impression that this is Socrates
and (ii) that then Socrates disappears and his twin brother takes his place. We
would have an impression exactly like our first impression and on the basis of
it judge again that this is Socrates. But this impression and the corresponding
judgment would be false.
The Stoic answer to this relies on the assumption that no two objects are ex-
actly alike (cf. Cic. Acad. II 85). Thus Socrates and his twin brother will differ
from each other at least minimally. Hence, a cognitive impression of Socrates,
being by definition distinct, could not be exactly like an impression that had its
origin in his twin brother. If the impression one received of Socrates were ex-
actly like the one which one received of his twin brother, both impressions
would be confused and hence not cognitive. But the impressions we receive of
Socrates and his twin brother do not need to be indistinguishable and hence con-
fused. For, the Stoics assume, the two brothers do differ from each other at least
minimally, and by sufficient training we can learn to distinguish perceptually any
two perceptible objects (cf. Cic. Acad. II 20; 56; 57; 86). Thus we can learn
to distinguish Socrates and his twin brother however much they may look alike,
and only if we have learned this can we have the cognitive impression that this
is Socrates. Hence, it cannot happen that we first have a cognitive impression
of Socrates and then a false impression exactly like it that this (Socrates' twin
brother) is Socrates.
The crucial issue here is the metaphysical principle of the internal distinctness
of different objects or the identity of indistinguishables. Since this principle is

firmly embedded in Stoic metaphysics, their reliance on it here cannot be dis-
counted as an ad hoc move. And once this principle is granted, the claim that
for any object there could be another object so much like it that we could not
possibly discriminate the two is considerably weakened. For though the skeptics
can point to many cases in which we could find it exceedingly difficult, if not
impossible, to distinguish different objects from each other because of their
similarity, the Stoics point out, not without plausibility, that if we just put our
mind to it we would also learn to tell these objects apart (Cic. Acad. II 56, 57).
It also may be mentioned that according to Stoic logic the two impressions one
receives when one sees Socrates and his twin brother would differ in one crucial
respect even if the two brothers were exactly alike: if they are impressions that
this is Socrates, they would differ in prepositional content since the demonstra-
tive has a different reference.
Now one may think that the skeptic's case gets a good deal of its force from
the fact that it seems to show that even under normal conditions we do not know
whether our impression is cognitive, since we do not know whether it is an im-
pression of the object it presents itself as an impression of, or whether it is in
fact an impression of an object very much like it which we have not yet learned
to distinguish from it. But we have to keep in mind that the Stoics do not deny
that we can make the mistake of thinking that an impression is cognitive when
it is not. They are committed only to the view that under normal conditions we
shall have a cognitive impression of the object in view, that the mind can discrim-
inate the impression as cognitive, and that we could not have the cognitive im-
pression that this is Socrates without being able to distinguish Socrates from all
other objects. But this does not mean that we cannot have all sorts of other cogni-
tive impressions of Socrates without being able to distinguish him from all other
objects. Similarly, we shall have a cognitive impression of Socrates' twin brother
if we see him under normal conditions, even if we do not know him at all, let
alone are able to distinguish him from all other objects. But this impression,
whichever it is, will be quite different from the cognitive impression that this is
Socrates. There is also nothing to prevent us from having the impression, concern-
ing Socrates' twin brother, with him in full view, that this is Socrates. But this
impression will not be any of the cognitive impressions we have when we have
the brother in full view, though we may make the mistake of thinking that it is.
The other line of attack the skeptics choose seems more promising. They
point out that even the patently false impressions of dreamers, madmen, and
drunkards all seem to have the features supposed to be characteristic of cognitive
impressions, or that they at least seem to be indistinguishable from them for the
person who has them.
The first thing to notice is that these impressions are due to nonnormal or ab-
normal states of mind; and it does seem far from obvious that such states of mind
do not have an effect on the internal character of the impressions they produce;
in fact, often it seems obvious enough that an abnormal state of mind systemati-
cally changes the character of our impressions. And, for reasons indicated
above, Stoic physics would seem to require that the internal character of impres-
sions implies a certain state of mind. Second, it needs to be noticed that even
if it were the case that in certain abnormal states a person is not in a position
to tell whether his impressions are cognitive or not, because the noncognitive
ones seem to him to have all the features of cognitive ones, this would not show
that he does not have cognitive and noncognitive impressions which differ from
each other qualitatively and which his mind discriminates accordingly. And cor-
respondingly we do not find the Stoics arguing that even dreamers and madmen
can tell that their dreams and hallucinations are noncognitive, but that even
dreamers and madmen react differently to cognitive and noncognitive impres-
sions (cf. SE M VII247). And this seems true enough, if we consider the matter
in general. The Stoics are, of course, committed to the view that the mind in
each case manages to discriminate between cognitive and noncognitive impres-
sions, but their theory also seems to allow them to explain apparent counterex-
amples. It is exactly a sign of a severely abnormal state of mind, if the mind
treats cognitive and noncognitive impressions almost indiscriminately so that in
particular cases there may seem to be no difference in observable behavior.
Thus, it seems that the skeptics fail to show that cognitive and noncognitive
impressions do not differ from each other qualitatively and that, hence, the mind
cannot discriminate between them on the basis of their inherent difference. They
even fail to show that it is impossible to tell absolutely reliably whether one's
impression is cognitive or not. What they perhaps do show is that we, in our
present state, cannot invariably tell whether an impression is cognitive or not.
But then the Stoics would be the last to deny that.
Conclusion
Academic skepticism is not characterized by a certain philosophical position, by
a set of philosophical views Academics are expected to subscribe to, but by a
certain dialectical practice and the impression the purusit of this dialectical prac-
tice left on them. Now it seems that earlier Academic skeptics like Arcesilaus
and Carneades were left with the impression that they had no reason to accept
philosophical beliefs. Whatever reasons they may have had when they started
out had been neutralized by arguments to the contrary. Later Academic skeptics,
though, starting with Metrodorus and Philo, seem to have had the impression
that however much one argued on both sides of any philosophical or theoretical
question, one still may find in the end that, as a matter of fact, one is still inclined
toward one side of the matter, that there is no reason to think that this is just
due to the fact that one is lacking in dialectical skill or has not considered the
matter carefully enough, and that there is no reason not to report which view
one feels inclined to, at least as long as one is among one's peers and there is
no danger that one's report is mistaken for an authoritative statement, as it might
be, for example, by young students. As a result many Academic skeptics came
to articulate quite elaborate philosophical beliefs. And given the dominance of
Stoicism and the syncretism of the time, these often hardly differed from the
views of their Stoic rivals. And since the Stoics did not claim knowledge for
their views either, the two positions became more and more difficult to distin-
guish, as soon as one left the field of epistemology. But given that both sides
now tended to have more or less the same beliefs on the basis of the same con-
siderations anyway, the epistemological debate must have started to look some-
what academic and futile, especially since it seemed to have ended in a deadlock.
Galen (De dogm. Plat: et Hipp. 796, 8ff. M) could even claim the following:
the younger Academics say that everything should be judged by means of plausi-
ble, tested, incontrovertible impressions (the Carneadean "criterion"), Chrysip-
pus maintains that matters should be judged by cognitive impressions, and com-
mon sense tells us that it is all a matter of perception and evident thought; but
their disagreement is only verbal: if one considers the matter more closely, Ga-
len says, one will see that they all advocate the same epistemic practice.
Thus it is not surprising that some skeptics thought that the Academy had be-
come unfaithful to its skeptical tradition and that they tried to revive the radical
skepticism of the early Academics, but now under the name of "Pyrrhonism" to
distinguish themselves from their Academic contemporaries. But by this time,
it seems, the Stoics were no longer inclined to engage in a real debate on the
matter and to refine their position accordingly. And thus orthodox Stoicism itself
was soon a matter of the past, whose views only lived on in the more or less
distorted form in which they were assimilated into other systems. And in this
distorted form the Stoics' views on cognitive impressions and their clarity and
distinctness, in fact the whole Stoic epistemology, have exercised, through sur-
viving Greek and Latin authors like Cicero and Sextus Empiricus, an enormous
influence well into modern times.

Frede-Stoics and Skepticson Clear and Distinct Impressions

Uploaded by

Copyright:

Available Formats

Frede-Stoics and Skepticson Clear and Distinct Impressions

Uploaded by

Document Information

Original Description:

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Frede-Stoics and Skepticson Clear and Distinct Impressions

Uploaded by

Copyright:

Available Formats

9

Stoics and Skeptics on Clear and

The history of Hellenistic philosophy is dominated by the rivalry between Stoics

The Stoic Position

nevertheless maintain that a rational impression involves the conceptualization

A; thus an evident robber is somebody who quite obviously is a robber (Soph.

Stoic Definitions of Cognitive Impressions

osition is in question, we may have to go through a number of impressions all

Cognition, Knowledge, and the Wise Man

might be led by a chain of reasoning to accept the contradictory of what we had

The Skeptical Attack

of different objects or the identity of indistinguishables. Since this principle is

You might also like