University of Virginia Library

III

The content of skeptical teaching may be discussed
under three heads: (1) the arguments used by the
skeptics to refute the dogmatists; (2) the formulation
of the skeptical position in terms of phrases and tropes;
and (3) the defense of skepticism.

1. One of Pyrrho's achievements, according to
Timon (cf. DL IX.65) was to break the chains of false
opinion; what opinions he attacked, and by what
means, we do not know. Timon himself denounces
rather than refutes his adversaries; but the fragments
of his works indicate that he discussed appearance as
the limit of certainty (DL IX.105), the method of
hypothesis (AM III.2), and the divisibility of time (AM
VI.66, X.197). He also held, apparently, that nothing
is by nature good (AM XI.140). Our information, how-
ever, is too slight to permit the identification of his
opponents or the reconstruction of his arguments.

With Arcesilaus the picture is clearer but still far
from complete. He undertook to argue on either side
of any question (DL IV.28; cf. IX.51, where Protagoras
is said to have held that for every argument there is
a counterargument) and to refute whatever opinion
anyone expressed (Ac. I.45; ND I.11; Fin. II.2; Or.
III.67; PE XIV.7.15). The opinion that chiefly interested
him was the Stoic view that some appearances can of
themselves be apprehended as certainly true. In rebut-
tal he defended the thesis that no appearances can be
so apprehended. This thesis was known as akatalepsia
(“nonapprehension”) (cf. AM VII.153-55; PE XIV.7.4).
It was sometimes taken to be a statement of the skepti-
cal position, along with epochē, “suspension of judg-
ment” (Ac. II.59; Plutarch, Moralia 1121F-1122A). A
more cautious interpretation would be that the
polemical postures assumed by Arcesilaus in order to
refute the Stoics are not necessarily positions to which
he himself subscribed; otherwise he becomes liable to
the charge of holding “dogmatically” that nothing can
be known. Such an accusation was in fact made against
the Academy (cf. Ac. II.28-29; NA XI.5.8; PH I.226);
but Cicero says quite explicitly that the Academic
Skeptics suspended judgment on every question, as the
arguments on both sides were of equal weight (Ac.
I.45), and Sextus absolves Arcesilaus, at least, from the
charge of dogmatism (PH I.232).

A second view that Arcesilaus is said to have
attacked is the doctrine that pleasure is the highest
good (cf. Fin. II.2). This was the view of two contem-
porary schools, the Cyrenaic and the Epicurean. Here
Arcesilaus might well have drawn on Plato; but the
evidence is lacking.

With Carneades our knowledge of the nature and
range of skeptical polemic is greatly increased.
Carneades' discussion of the problem of knowledge was
broadened to refute not only the Stoics (who were still
the primary target) but all others who claimed to have
found an infallible test of truth. Some idea of his argu-
ments can be got from Cicero, Ac. II.79-98 and from
Sextus, AM VII.159-65. Similarly in his discussion of
ethical theory he gave an exhaustive enumeration of
possible views of the summum bonum, or “highest
good” (Fin. V.16-20). His practice of arguing both sides
of a question is illustrated by his two speeches on
justice delivered at Rome on the occasion of the em-
bassy from Athens in 155 B.C. His arguments against
natural justice can be recovered in part from the
fragmentary remains of Cicero, De republica III, where
Philus is his spokesman.

Carneades also developed arguments against philo-
sophical theology. The best-known are (1) that the
powers and activities assigned to divine beings are not
consistent with their being changeless and eternal (AM
IX.137-81; ND III.29-34); (2) that the evils in the
universe are not consistent with divine providence
(Plutarch, Moralia, frag. 193 ed. Sandbach, from
Porphyrius, De abstinentia III.20; Ac. II.120); (3) that
the occurrence of accidental designs, for example, a
rock that has the form of a head, invalidates the argu-
ment that a design implies a designer (Div. I.23); and
(4) that no clear boundary can be drawn between what
is divine and not divine (ND III.43-50; AM
VII.182-90). This last argument is an example of the
sorites, or “heap,” a device for obscuring boundaries
by pointing to continuous gradations. It was also used,
presumably by Carneades (cf. Ac. II.49 and 92-95), to
obscure the distinction between illusions and veridical
sense perceptions.

Carneades' methods exhibit certain tendencies that
became increasingly strong in later skepticism. One
such tendency is toward the schematic formulation of
alternatives, as exhibited in the discussion of the
summum bonum. Compare also his attack on divina-
tion (Div. II.9-12), where the possible objects of
divination are systematically enumerated and rejected.

It is tempting to assign also to Carneades the series
of propositions (PH III.10-11) on the power and provi-
dence of god, a scheme which Lactantius assigns to
Epicurus—rather improbably, inasmuch as it entails
the rejection of Epicurean theology; see Lactantius,
De ira Dei, 13.20-21 and De Lacy, Transactions of the
American Philological Association, 79 (1948), 18-19.
The tendency toward schematic analysis is seen also
in the argument on fate reported by Cicero, De fato
31; and it is commonly supposed that Carneades
formulated the “four heads” (Ac. II.83) from which it
follows that nothing can be apprehended through sense
perception. Carneades anticipates a later trend also in
his examination of the notion of the divine. Compara-
ble examinations were subsequently made of cause (AM
IX.195-266), body (ibid. 359-440), time (AM X.169-
247), and the like, the aim being in each case to show
that no consistent account of these concepts can be
given.

The development of characteristically skeptical
analyses and arguments led to two levels of refutation,
one level employing arguments that dogmatists use
against each other, the other the distinctively skeptical
arguments. There is a hint of this already in Carneades,
who opposes to the Stoic doctrine of fate not only his
own dialectical refutation but also the Epicurean re-
jection of fatalism (Cicero, De fato 21-23). The mutual
support of Epicureans and skeptics against Stoics or
Platonists appears again in Div. II.51 and in Sextus'
attacks on the teachers of the arts and sciences (AM
I.1-7).

A fairly good example of skeptical polemics in the
Carneadean tradition is the speech of Cicero in Ac.
II.64-146. It includes a historical sketch, arguments to
discredit both sense perception and reasoning as
sources of certain knowledge, a defense of Carneades'
doctrine of probability (see below, III, 3), and an ac-
count of the disagreements of the dogmatists in physics,
ethics, and logic.

At some time the scope of skeptical attack was
broadened to include among its targets the theoretical
arts and sciences. Medicine was among the first to be
involved in this controversy. The split between the
theoretical and empirical approaches to medicine is
evident already in the Hippocratic corpus (fifth century
B.C.); and an Empirical School of medicine, with strong
tendencies toward skepticism, was founded by Philinus
in the third century B.C. Members of this school in later
times praised Pyrrho for having followed appearances
in everyday activities and having suspended judgment
about all else (cf. Galen, Subfiguratio empirica, in
Deichgräber, p. 82). The statement that “Appearance
prevails wherever it goes” is indeed found in the frag-
ments of Timon (cf. AM VII.30), and it provides a basis
for the view that the arts and sciences which limit
themselves to the use of appearances are legitimate,
whereas those that claim to say something about the
real nature of things are not. This is the view that
Sextus adopts in AM I-VI. There are a few indications
that the Academic Skeptics also discussed the arts and
sciences. In Ac. II.122 Cicero mentions the empirical
physicians' conviction that the nature of the body
cannot be discovered by dissection, as the concealed
organs may be altered by the mere act of laying them
bare. Cicero also rejects the argument that the skeptic's
attack on knowledge is an attack also on the arts; some
arts, he says, admit that they use conjecture more than
knowledge, and others such as painting and sculpture
are guided by what appears rather than by what is
(Ac. II.22, 107, 146). There is no evidence, however,
that the Academic Skeptics attacked mathematics.

Aenesidemus is best known for the ten tropes (see
below, III, 2). Underlying the tropes is a formulation
of the epistemological problem in terms of signs: if
from the apparent we obtain knowledge of the non-
apparent, then the apparent serves as a sign, the
nonapparent as a thing signified (see Photius, III, 121).
But, Aenesidemus argues (cf. AM VIII.215-35), if signs
were apparent they would appear the same to all who
are in a similar state, that is, there would be no dis-
agreement about what they signify. But there is
disagreement; therefore signs are not apparent. Sextus
gives a much fuller account of this doctrine of signs.
He divides the nonapparent into three kinds: (1) the
absolutely nonapparent, e.g., the number of grains of
sand in Libya; (2) the nonapparent by nature, e.g., the
invisible pores in the skin, or the void outside the
universe; and (3) the nonapparent at the moment, e.g.,
the city of Athens. Things absolutely nonapparent may
be left out of consideration. Things by nature non-
apparent can be known only if there are appearances
which point to them unambiguously (indicative signs),
as the movements of the body are said to be signs of
the soul or motion a sign of void. Things nonapparent
at the moment can be signified by present appearances
that remind us of them (admonitive signs), as smoke
is the sign of fire, a scab is the sign of a wound (AM
VIII.141-55). Sextus does not challenge the possibility
of admonitive signs; they presuppose no necessary
connection between sign and thing signified, and they
are adequate to account for the connections that we
establish between things in everyday activities (AM
VIII.155-58). Reminding is, in fact, the skeptics' sub-
stitute for proof; see for example, PH III.20; De Lacy,
Phronesis, 3 (1958), 71. But about the indicative sign,
which is the invention of dogmatic philosophers and
theorizing physicians, the skeptic withholds judgment,
as he finds the arguments against it as strong as those

for it (AM VIII.159-298). Sextus includes the argument
of Aenesidemus mentioned above as a part of his attack
on the indicative sign.

The very length of Sextus' discussion of signs testifies
to their importance in skeptical polemic (cf. also DL
IX.96-97). His reference to theoretical medicine and
his use of medical examples suggest that the terms in
which he presents the problem may have been current
in medical controversy. A more widespread practice
was to distinguish between common signs, which are
ambiguous in their reference, and particular signs,
which signify one thing only. This distinction is found
in medical writers, who use it in the identification of
symptoms; see for example Galen, Commentary on
Hippocrates' De Officina Medici, I.1 (ed. Kühn, XVIII,
2, 643-45). It appears also in rhetorical theory (cf.
Cicero, Partitiones oratoriae, 34) and in philosophical
controversy of the Hellenistic period (see Philodemus,
De signis, cols. 1 and 14). There is reason to suppose
that Carneades' attack on Stoic epistemology was at
some time stated in terms of this distinction: the Stoics
regarded some sense perceptions as particular signs,
others as common signs. Carneades challenged them
to show that any appearance is ever a particular sign
(Ac. II.33-34, 84, 103).

Another matter of major concern to Aenesidemus
was the notion of cause. Sextus reports an argument
that he used to show that one thing cannot cause, i.e.,
generate, another (AM IX.218-26). In addition, he
formulated eight ways of attacking dogmatic theories
of causation (the eight tropes; cf. PH I.180-84; Photius,
III, 122). Photius reports that Aenesidemus also dis-
cussed truth (cf. AM VIII.40-47), motion, the universe
and the gods, the objects of choice and avoidance, the
virtues and the summum bonum (cf. AM XI.42).

Sextus' treatises are by far the most extensive of the
ancient skeptical writings that have survived; and
although Sextus incorporates many items derived from
earlier skeptics, his presentation and elaboration seem
to be his own. For example, no model has been found
for his six books on the special disciplines (AM I-VI)
or for his discussion of ethics (AM XI, PH III.168-279).
He also gives a long and detailed treatment of cause
(AM IX.195-266; cf. DL IX.97-99). Of the many other
matters that he takes up, perhaps the most important
is his attack on Stoic and Peripatetic logic (PH
II.134-203; AM VIII.300-481). Sextus had a com-
mendable familiarity with early Greek philosophy, and
he ranks as an important source of information about
the pre-Socratics and others whose works have been
lost. He has, besides, an obvious enthusiasm for his
subject, which sometimes turns to playfulness (e.g., PH
I.62-63). It is perhaps because his writings were more
than mere compilations that they survived.

2. A feature of ancient skepticism throughout its
history was a fondness for catchwords. The most widely
current was epochē, “suspension of judgment,” used by
both Pyrrhonists and Academics (see Ac. II.59;
Plutarch, Moralia 1122A; PH I.8, 10, 196; DL IX.61,
62). There is some question, however, whether it goes
back to Pyrrho himself, and it does not appear in the
fragments of Timon. Often coupled with epochē is
akatalepsia, i.e., “nonapprehension,” which probably
derives from the Academic attack on the Stoic theory
of apprehension (katalepsis; cf. Ac. II.17-18, 31). Sextus
(PH I.1-3) assigns it to the Academy rather than to
his own school (see also Photius, III, 119-20). Diogenes
Laërtius, however (IX.61), uses it along with epochē
to characterize Pyrrho's teaching. More certainly
Pyrrhonic is the phrase ou mallon (“no more this than
that”), which appears in the fragments of Timon (cf.
DL IX.76; PE XIV.18.3), in Aenesidemus (Photius, III,
119), and in Sextus (cf. PH I.187-91). It also had a place
in pre-Pyrrhonic philosophy; see De Lacy, Phronesis,
3 (1958), 59-71. The refusal to incline this way or that
was expressed by arrhepsia (PH I.190; DL IX.74), the
refusal to make assertions by aphasia (Timon in PE
XIV.18.4, 19; PH I.192-93; cf. Plutarch, Moralia
1123C), the avoidance of distinctions by ouden horizein
(Timon in DL IX.76; cf. 71 and 74; PH I.197), the
avoidance of rashness in assent by aproptosia (DL
IX.74; cf. PH I.20, 177, 186; II.21; Ac. I.45), the equal
balance of arguments for and against any thesis by
isostheneia (DL IX.73, 76, 101; PH I.8; AM IX.207;
cf. Ac. I.45: paria momenta).

Such terms as these gave the skeptics a kind of
identity and served as substitutes for positive doctrine.
The same may be said of their many schematisms, some
of which have already been mentioned. Among the
most important in later skepticism were Aenesidemus'
ten tropes in support of the view that although it is
possible for me to describe each thing as it appears
to me, I must suspend judgment as to what sort of thing
it is in itself.

The tropes are listed, with minor differences, by both
Sextus (PH I.36-163) and Diogenes (IX.79-88). As
Sextus presents them, the same thing appears different
(1) to different species of animals; (2) to different in-
dividuals, by virtue of their differences in mind and
body; (3) to different senses, as a painting appears flat
to the touch but to the eyes seems to have depth; (4) to
the same sense in different states, e.g., in sickness or
health, in youth or old age; (5) because of differences
in position, distance, or place, as the square tower
appears round at a distance, and the oar appears bent
where it enters the water; (6) by virtue of differences
in the things in whose company it appears, as an object
heavy in air appears light in water; (7) because of

differences in quantity and situation, as grains of sand
when scattered appear rough but in a heap appear soft;
(8) in different relations, everything being in some
sense relative; (9) insofar as it is encountered continu-
ously or rarely, as an earthquake is more frightening
to those who experience it for the first time than to
persons accustomed to earthquakes; (10) according to
different ways of life, customs, and beliefs, as the
Taurians sacrifice strangers to Artemis, but the Greeks
forbid human sacrifice. Diogenes gives to the list a less
arbitrary sequence; his order is 1-4, 10, 6, 5, 7, 9, 8.
Aristocles (PE XIV.18.11) gives the number of tropes
as nine; no satisfactory explanation of the discrepancy
has yet been found.

The five tropes of Agrippa are broader in scope,
dealing not only with appearances but also with proof.
They are given at some length by Sextus (PH I.164-77)
and summarized by Diogenes (IX.88-89). The first is
that as we are not able to resolve the disagreements
and conflicts in life and in philosophy, we end up by
suspending judgment. The second is that as anything
submitted in support of a proposition must itself be
supported, an infinite regress results. The third is that
a perceived object appears to be of such and such a
description relative to that which makes the judgment
and to the things perceived along with it, but we
suspend judgment about its true nature. The fourth is
that dogmatists, in order to escape infinite regress, take
unproved assumptions as their starting point. The fifth
is that when that which ought to establish some con-
clusion can only be proved from the conclusion, since
we can use neither for the proof of the other, we
suspend judgment.

Finally Sextus (PH I.178-79) assigns to certain
unnamed skeptics two tropes that aim at a formula
of universal application. Certainty about a thing is got
either (1) from the thing itself, or (2) from some other
thing. The first alternative is refuted by the unresolved
disputes of the natural philosophers; the second leads
to infinite regress.

3. There were two main attacks on skepticism in
antiquity: (1) it makes action impossible; and (2) it is
self-contradictory. The first rests on the observation
that action presupposes decision, and decision involves
a choice between alternatives. Anyone who says of
things that they are “no more this than that” thereby
destroys the ground for practical decisions and so
makes action impossible. As Aristotle said (Metaphysics
1008b 26-27), “All men make unqualified judgments,
if not about all things, still about what is better and
worse” (trans. Ross). To this charge the skeptics gave
a number of answers. First of all, they pointed out that
practical decisions are made in terms of appearances.
Men naturally seek what appears good and avoid what
appears bad; in this sense they follow nature as their
guide. Skepticism does not challenge appearances, but
only the dogmatists' claim to have certain knowledge
about the nonapparent (cf. Ac. II.103; PH I.19-20).
There is therefore no conflict between skepticism and
practical decisions. The view that we live by appear-
ances is attributed by Aenesidemus to Pyrrho himself
(cf. DL IX.106), and it is implied by the fragment of
Timon already quoted, “Appearance prevails wherever
it goes.” It receives explicit statement also in the Aca-
demic tradition (cf. Plutarch, Moralia 1122) and is
accepted by Sextus (AM VII.30; PH I.21-24). Indeed,
Sextus regards the observation of appearances as the
basis of the practical arts (AM V.1-2) and of the
admonitive sign (PH II.100; AM VIII.152, 156-57).
One is reminded of Plato's description of the skill in
observing, remembering, and predicting that was
held in honor by the inhabitants of the cave (Republic
516C-D).

The skeptics also accepted tradition and custom as
a guide to action (PH I.17, 231; DL IX.61, 108; PE
XIV.18.20). Custom may be observed on the level of
appearance and followed without intellectual commit-
ment. It is on this basis, for instance, that the skeptic
performs acts of piety and avoids impiety (cf. PH
I.23-24, III.2; AM IX.49). Thus Cotta, the Academic
spokesman in Cicero's De natura deorum, insists that
he may be a philosophical skeptic and still participate
in the traditional Roman religion (ND III.5, 9). It is
not unlikely that skepticism helped to strengthen the
trend toward traditionalism in the Greco-Roman world.

Arcesilaus advanced still another guide to practical
action. According to Sextus (AM VII.158), he held that
happiness is secured through practical wisdom, practi-
cal wisdom consists in right action, and right action
is action for which a reasonable defense can be given.
The reasonable (eulogon) is thus the guide. There is
no mention of this view in the extant portions of
Cicero's Academica, but perhaps it is not mere coinci-
dence that several letters which Cicero wrote during
the period when he was working on the Academica
(June-July, 45 B.C.) contain the Greek word eulogon
or eulogia (Letters to Atticus, XIII.5, 6, 7, 22). In the
last of these (22) the reference is to a decision that
Cicero must himself make. Another letter to Atticus
(XIV.22), written a year later, uses eulogon in the
context of making conjectures about the future. Atticus
would surely have seen in these letters an allusion to
the skeptical criterion.

Carneades formulated a three-step procedure for
determining the probability of an appearance. The first
step is to limit oneself to persuasive appearances, that
is, to those which appear to be in accord with the
objects from which they come, and from among these

appearances to select the ones that are not dim or
distant or in any way indistinct. The second step is
to inspect the persuasive appearance in the context
of the chain of apperances that accompany it. For
example, the appearances of a man brings with it ap-
pearances of his color, size, form, movement, speech,
clothing, and also of the air, light, day, sky, earth,
friends, and so forth. If none of these concomitant
appearances exerts a contrary pull by appearing false,
our confidence is increased. Sextus compares this
inspection to that of physicians who do not judge that
a man has a fever from one symptom alone but from
a concurrence (syndrome) of symptoms, pulse, temper-
ature, color, and so forth. So the Academic looks for
a concurrence of appearances, none of which exerts
a contrary pull. When this condition is met, the initial
appearance may be said to be persuasive, with no pull
to the contrary. The third step is to examine closely
all of the concomitant appearances in order to assure
ourselves in each case that our vision is not dulled,
the distance is not too great or the object too small,
the duration of the appearance is not too short, etc.

When all these conditions are satisfied by all the
appearances, then the appearance with which we
began may be described as persuasive (pithane), having
no pull to the contrary (aperispastos), and examined
from all sides (periodeumene). The fullest account of
the Carneadean criterion is in Sextus (AM VII.166-84);
there is a shorter account in PH I.227-29. Cicero
alludes to the three stages but does not explain them
in Ac. II.33, 36; further, in II.105-10 he defends
Carneadean probability as an adequate guide in prac-
tical matters and in the arts. The later Pyrrhonists
rejected Carneadean probability as a departure from
true skepticism (cf. Photius, III, 119-20; PH I.229-30);
and indeed in this they had the support of Galen, who
took Carneadean probability to be equivalent to Stoic
apprehension (katalepsis) and to the formula, which
Galen himself preferred, that whatever appears clearly
to mind or senses is true (ed. Kühn, V, 778).

The charge that skepticism is self-contradictory ap-
pears most often in the following form: a person who
says that nothing can be known must admit that he
cannot know whether nothing can be known and
therefore must admit that perhaps something can be
known. Alternatively, if he claims to know that nothing
can be known he thereby admits that at least one thing
can be known and so contradicts his principle.
Metrodorus of Chios (quoted above, II) was no doubt
attempting to escape the second alternative when he
said, “We know nothing, not even whether we know
or do not know.” Similarly, according to Cicero (Ac.
I.45), Arcesilaus denied that anything could be known,
not even what Socrates left for himself (he knew that
he knew nothing; cf. Ac. II.74). Cicero also reports (Ac.
II.28) an exchange between Carneades and the Stoic
Antipater on this point. Antipater suggested that the
skeptic might make his position consistent by saying
that nothing can be comprehended except this one
thing, that nothing can be comprehended. Carneades,
however, insisted that there be no exceptions; the
person who states that nothing can be comprehended
must include this statement among the things that
cannot be comprehended. Lucretius probably echoes
the Carneadean view when he says, “If anyone thinks
that nothing is known, he also does not know whether
this can be known, since he confesses that he knows
nothing” (RN IV.469-70). The Academic Skeptics did
not escape the ambiguity of their presentation; they
were sometimes accused of affirming that nothing can
be comprehended (see above, III, 1). Sextus was more
cautious; he carefully avoided saying anything that
might seem to commit him to such an affirmation.

It was possible to state the charge of inconsistency
in terms of the ou mallon formula. Aristotle anticipated
the skeptics' dilemma when he said, in discussing
Heraclitus' supposed denial of the law of contradiction
(Metaphysics 1062b 2-9), that from Heraclitus' position
it would follow that just as when contradictory state-
ments are taken separately the affirmation is no more
true than the negation, so when the two together are
taken as a single affirmation, the entire affirmation will
be no more true than its negation. The answer of the
earlier skeptics to this criticism is not known; but
Sextus at least recognized its force. He saw that ou
mallon as a principle includes itself: it is no more true
than false and is therefore not a tenable position (PH
I.14). It was probably in response to this difficulty that
some skeptics compared ou mallon and other such
formulas to a purgative that eliminates itself along with
the arguments of the dogmatists (cf. DL IX.76; PE
XIV.18.21). Sextus offers another way out. The skepti-
cal ou mallon, he says, is not to be taken as an affirma-
tion or a negation but rather as a report of the skeptic's
inability to decide between conflicting statements. It
is a description of his state of mind and is as much
a question as a statement (PH I.15, 191-93, 200).