Profiles in skepticism: Carneades

The philosopher for whom belief comes in degrees

Marble head of Carneades by Osama Shukir Muhammed Amin, image from Wikimedia.

One of ancient skepticism’s towering figures was without a doubt Carneades of Cyrene (214-129 BCE), who became scholarch of the Academy little over a century after Arcesilaus turned Plato’s school in a skeptical direction. Like other skeptics associated with the Academy, Carneades’ driving idea was to go back to Socrates’ dialectical approach to philosophy and to his famous admission that he didn’t know much. Socrates, according to the Academics, was the original skeptic.

Arguably, Socrates had devoted himself to the pursuit of three aims: (i) making people realize that, despite their assumptions to the contrary, they believe a lot of things that are not true; (ii) subtly hinting at the most likely answers to the various philosophical problems being debated; while (iii) hiding his own convictions so not to unduly influence his interlocutors. Carneades followed this same three-pronged approach to philosophical dialectics, one that is practiced still today under the broad term of “Socratic method.”

Carneades, like Arcesilaus before him, focused his criticism on the dominant school of the time: Stoicism, and particularly on Stoic epistemology, the branch of philosophy that deals with how we know things. More precisely, he rejected the Stoic idea of “kataleptic” impressions, that is impressions that somehow bear a mark of truth that can be detected by a sage.

Recall than an “impression,” in Stoic-Skeptic lingo, is an immediate judgment that is prompted in us either by a sense experience or by an internal thought. For instance, “I see my brother across the street” is a sense-generated impression. Notice that I am not just stating a fact, but rather a fact+judgment. The fact is that I perceive something out there that looks like a person. The judgment is that I believe that person to be my brother.

Impressions can also be internally generated. If I think “the square root of nine is three” I am also having an impression, which is not triggered by a sensorial experience. Just like in the case of my brother, I may also turn out to be correct or incorrect in my judgment that the square root of nine is, indeed, three. If I agree with the initial judgment I am “assenting” to the impression. Otherwise I am denying assent to it.

Now, the Stoics thought that some impressions are special because they bear a mark of truth that definitely separates them from false impressions. Moreover, that mark of truth can be recognized by sages, that is by human beings who have achieved perfect reasoning ability. I know, it sounds far fetched, but bear with me just a bit longer.

These special impressions were named “kataleptic” by Zeno of Citium, the founder of Stoicism. Katalepsis means “to grasp,” indicating that we can grasp the truth of such impressions—again, if we are sages.

For instance, if I look out of my window right now I see a beautiful sunny sky over Trastevere, in central Rome. The impression that it is day and sunny is very forceful, so much so that I cannot really doubt it. If you wanted to convince me that it is actually midnight you’d have a hell of a difficult time doing so. Since I’m not a sage, though, I could, in theory, be mistaken. It is possible that what’s going on is that a film troupe is shooting a movie out there and their intense lighting system makes it look like it’s day while it is, in fact, midnight. A sage, by contrast, would be able to grasp the truthfulness of this kind of impression because of the above mentioned mark that allegedly separates it from untruthful impressions.

Note that many impressions are not kataleptic. For instance, if I say that the number of stars visible tonight in the heavens is even rather than odd, is that the case or not? It would be unwise to heavily bet one way or the other, because that sort of impression does not bear an unmistakable mark of truth and not even a sage would therefore be able to grasp it. Sages, after all, are human beings, not omniscient gods.

Why on earth should we care about any of the above, other than as a historical footnote about how the ancient Stoics thought? Because if Zeno & co. are right it is humanly possible—at least in some cases and for some people—to arrive at certain truths, at Knowledge with a capital “K.” That’s a fairly big deal, don’t you think?

Enter Carneades to shutter that dream of certainty in all possible cases, including those involving sages. Carneades and the other Academic Skeptics thought that there simply is no such thing as a kataleptic impression, ever. An impression may feel undeniable, but it is always possible to build an argument that something else caused the impression. Therefore, we cannot be sure whether the impression is true or false.

For example, am I really seeing my brother across the street, or is it someone who just looks like him? Even if I’m highly confident that it is my brother I can still imagine the possibility of a doppelgänger that is so close in appearance as to be indistinguishable from my brother. Indeed, perhaps my brother has an identical twin from whom he was separated at birth, so that the similarity between the two would extend down right to the level of DNA. How is any human being, sage or not, going to be able to distinguish the two with one hundred percent confidence? If Carneades is right then nobody, not even a sage, is capable of achieving certain knowledge. Some impressions may be more or less in doubt than others, but none bear the elusive mark of absolute truth the Stoics imagined.

Doppelgängers, image from thedailybeast.com.

In turn, this implies that if the sage doesn’t want to risk assenting to something that may turn out to be untrue she may have to suspend judgment about any given impression. It may be daylight, or maybe not. That may be my brother, or maybe not. The Stoic sage has now being turned into a skeptic!

The broader implication is that certainty is not a thing to which human beings are privy, which ought to instill in us more than a bit of epistemic modesty. We should be cautious whenever we say that this or that is “true,” because we can’t possibly know that for sure.

The Stoics, of course, had a response to this. It relied on their stock anti-skeptic argument: if we could never confidently tell truth from falsehood we could never act, which is absurd. Therefore, there must be kataleptic impressions.

This is a really bad argument, it must be said. First off, as we shall see in a moment, it is simply not true that inaction is the only path open to someone who admits she doesn’t have certainty. Second, even if inaction were the only possibility, wishing to avoid such an admittedly paralyzing outcome wouldn’t establish the reality of kataleptic impressions. It would be like saying that I don’t like the idea of dying, therefore immortality is a thing. No, it isn’t.

There are two, far more sensible paths available if we agree that there are no certain truths: we could simply accept an apparently persuasive opinion without committing ourselves to any statement about its possible truth value (radical skepticism); or we could accept an impression as provisionally or likely true, with the caveat that we could be wrong and might have to change our mind in the future (modest skepticism).

The first option, radical skepticism, is essentially the Pyrrhonist way, which we have explored in a recent post and to which we’ll return once more by the end of this series of essays. The second option, modest skepticism, is the way of Carneades and the Skeptical Academy more generally.

As we have seen more in detail in another essay Carneades went further than other Academic Skeptics and articulated the most sophisticated “criterion” for assenting to an impression: persuasiveness (pithanon), which Cicero translated into Latin as probability (probabilis).

The idea is that we always, or at least very often, have some bases to prefer one interpretation of an impression to another. Back to the example of my brother: yes, it is certainly conceivable that he has a stunningly look alike doppelgänger. It is even conceivable that he has an identical twin who was separated at birth. But, really, how persuasive, or likely, are such possibilities? Do I actually doubt that it is my brother I am talking to now? Probably not, though of course I have to acknowledge that I could, in fact, be wrong, as far fetched as that may seem.

The criterion, in other words, doesn’t have to be certainty. The Stoics thought it should, and they ended up limiting true knowledge only to sages. The Pyrrhonists also thought infallibility to be mandatory, rejected its possibility and went instead for suspending judgment on everything. Carneades struck a happy middle: we can assent to things on the basis of probability, so long as we are open to revise our judgment if a better case presents itself.

Most usefully, Carneades proposed that there are different degrees of persuasiveness of impressions. If the stakes are not too high, or we do not have the time to investigate further, we can go along with a given notion merely because it is persuasive on the face of it, even if barely so.

If the stakes are higher and we have more time, however, we want to assent to impressions that are strongly, not just apparently, persuasive.

Are the stakes even higher? Then we need to take the time to check whether the impression is coherent with a number of other impressions we already accept (in skeptical jargon, whether it is “undiverted”).

The highest possible degree of persuasiveness is achieved when an impression is not only strong and undiverted, by survives a sustained scrutiny based on arguments and/or empirical evidence.

In order of increasing confidence, then, we assent to impressions under one of these conditions:

Basic persuasiveness < Strong persuasiveness < Undivertedness < Researched thoroughness

As Cicero, a self-declared follower of Carneades, put it:

“The wise person will use whatever strikes him as persuasive … and the whole structure of his life will be governed in this way.” (Academica II.99)

That’s the best that anyone, sage or not, can humanly do. The underlying notion being put forth by Carneades is similar to a much more modern concept advanced by Karl Popper in the 20th century: “truth” is what survives refutation.

Notice also that Carneades and the Academic Skeptics do not say that there are no truths, only that we cannot be certain of any of them. Their statement is not a metaphysical one (truth does / does not exist) but rather an epistemological one (truth can / cannot be achieved by human beings).

Carneades was innovative and anticipatory of modern thought in another way as well. Just like impressions come in degrees of persuasiveness, so the assent we give to such impressions is also a matter of degree. This sounds to me very much like David Hume’s famous dictum (in Of Miracles) that a wise person proportions his beliefs (assent) to the evidence (strength of the impressions), a notion that astronomer Carl Sagan famously rendered as “extraordinary claims require extraordinary evidence” (which implies that ordinary evidence suffices for ordinary claims).

Carneades’ sensible treatment of impressions and assents also gives the lie to the Pyrrhonist, more rigid and unpersuasive, approach. For the Pyrrhonist all impressions have equal claim to be true, and it is always possible to come up with a counter-argument that is just as convincing as a given argument. But this is rather untenable, and few people actually think in that manner. Assume again that it is sunny right now outside my window. Is that impression truly just as likely as the impression that it is rainy, despite all the evidence pointing to sun rather than rain? If you wanted to articulate an argument to convince me that it is, in fact, raining, do you really think it would be as persuasive as an argument aimed at showing that it’s sunny? I think not.

Harald Thorsrud, author of a chapter on Carneades in Skepticism: From Antiquity to the Present, provides us with the upshot:

“[The Academic Skeptic] will proportion the strength of his assent to the persuasiveness of the impression, whether through systematic scrutiny of his perceptual impressions or through argument pro and con in the case of philosophical propositions, and he will remain perpetually open to being corrected by subsequent experience and additional argument. On those rare occasions in which a thoroughly examined persuasive impression turns out to be false, or dialectically defeated, the [Skeptic’s] opinion that it was probably true is not blameworthy.” (ch. 4, p. 179)

[Next in this series: Marcus Tullius Cicero. Previous installments: The Cyrenaics; Pyrrho; Arcesilaus]