Skepticism’s alleged self-refutation, refuted

One common (and understandable) argument often encountered online and in real life is the one that skepticism is self-refuting, posed in the form of sayings like “How do you know that you know nothing?” , “Do you doubt your doubt?” , ect. While it is easy to understand why one might think that, it ultimately fails, as I shall show in this article.

  1. An overview of the problem with the argument

I’ll just quote British philosopher Alex Malpass’s response to Christian apologist Jeff Durbin’s use of this argument (from: :

“It is not a stretch to say that Pyrrho’s position is that ‘we cannot know anything’. If so, then Pyrrhonian Scepticism is the position that Durbin’s argument was attacking. His idea is that this sort of extreme scepticism refutes itself. And this line of attack certainly has some appeal to it. We derived a contradiction from the assertion that ‘I know that I know nothing’. So imagine that I were to go around saying ‘My view is that Pyrrhonian Scepticism scepticism is true’. This would make me vulnerable to Durbin’s line of attack, as he could ask me whether I know that Pyrrhonian scepticism is true. If I said that I did know it was true, then I would be contradicting the main claim of Pyrrhonian scepticism; but if I said that I didn’t know it, then I would be tacitly conceding that it might not be true after all.

It’s not clear that even this extreme position is vulnerable to Durbin’s attack though. I could say that when I affirm Pyrrhonian scepticism, I am not making a knowledge claim at all; the content of the claim that Pyrrhonian Scepticism is true could plausibly be taken to be: ‘I believe that Pyrrhonian scepticism is true’. If so, I would be saying that ‘I believe that (I can’t know anything)’. If Durbin asked his gotcha question, ‘But do you know that?’, I could reply ‘No, I do not’, quite without self-contradiction. If Durbin asked ‘But do you believe that?’, I could reply ‘Yes, I do’, also quite without contradiction. So a Pyrrhonian sceptic[Note 1]can construe the statement of their own doctrine as merely a belief claim rather than a knowledge claim, and thus avoid Durbin’s accusation of self contradiction.

However, even if we forget this nuance, and insist that anyone who claims to be a Pyrrhonian sceptic is making a knowledge claim, we still have not refuted Pyrrhonian scepticism with this argument. The most we would have demonstrated is an inconsistency between the sceptic’s behaviour and the content of Pyrrhonian scepticism. It doesn’t prove that Pyrrhonian scepticism is wrong; it just shows that the person making the claim isn’t acting like a good Pyrrhonian sceptic. A good Pyrrhonian sceptic should not make knowledge claims. But criticising a claim on the grounds that the person making the claim’s actions are inconsistent with it, is to commit the tu quoque fallacy. So, even if we pretend that Durbin had caught us doing something which was inconsistent with Pyrrhonian Scepticism (like making a knowledge claim), and if he implied that this showed that Pyrrhonian Scepticism was false, then he would have committed the tu quoque fallacy.

But, even if we overlook this informal fallacy, Durbin is still in trouble. Even if Durbin had completely scored his point, and established that it is not possible to say anything as a Pyrrhonian sceptic without immediately contradicting yourself, this still leaves an escape route; you can be a Pyrrhonian sceptic and not make any claims at all. In fact, this response is the one advocated by followers of Pyrrho. When confronted with an argument, the best you can do is wag your finger at it, like Cratylus. If you don’t say anything at all, you cannot contradict yourself!

Presumably, this position would be open to ridicule by Durbin. Being able to avoid the problem only by retreating to complete silence would seem like a capitulation rather than a victory. Despite initial appearances though, this response might actually be thought to have something going for it. The view recommends a sort of spiritual, monk-like silence, which Pyrrhonians thought could be a pathway to enlightenment and happiness:

“…the result for those who are so disposed [to Pyrrhonian scepticism] will be first speechlessness, but then freedom from worry; and Aenesidemus says pleasure.” (ibid.)

Indeed, even if this supposed benefit were not there, a theory is not deemed false merely because it has been arrived at by retreat. Even if the silent Pyrrhonian monk only took his vow of silence reluctantly and after he conceded in a debate that there was no other way to be consistent with Pyrrhonian scepticism, this doesn’t make Pyrrhonian scepticism false. It could still be true for all that. Therefore, even here, when we have been as generous to Durbin as we possibly could, we have not found a refutation of Pyrrhonian scepticism.

So perhaps Durbin has a point against the Pyrrhonian sceptic who goes around making explicit knowledge claims, which is that he is not a ‘good’ Pyrrhonian sceptic. But Durbin does not have a point against one who makes more nuanced belief claims, and certainly not against one who remains in a peaceful silence. So, the argument only even slightly works if you straw-man it so that the opponent has to be an inconsistent Pyrrhonian sceptic (one that makes explicit knowledge claims). “

[Note 1: Technically, whether the Pyrrhonian can even claim this is doubtful, possibly except in the broadest sense, as their “position” is one of complete neutrality other than about personal seemings, e.g. “It seems to me like it is daylight” as opposed to “It actually, objectively is daylight”. Of course, the more moderate (e.g. Academic) skeptic who allows probabilistic degrees of opinion can. Some may respond by still demanding a reason to support the skeptic. It can, however, still be somewhat

reasonable to have an opinion which is still not strong enough to be knowledge. Still, Durbin’s argument is weak against even the Pyrrhonian as lack of an opinion doesn’t preclude arguing for a position, as shall be shown below.]

I also would have added that it commits the Argument from ignorance fallacy by implying that something (that we can know) must be true just because we cannot know it to be false.

Even the Pyrrhonian can still plausibly assume their position ex hypothesei in favor of suspense of judgment. Here is a debunking of, or at least a reasonable response to, the more general notion that arguing for something requires a truth claim:

2. The “doubt your doubt” formulation

One form of the argument that has been put to me is “If you doubt everything then you must doubt your doubt implying that something is sure, therefore you don’t do so.” Or, as Descartes (if I remember right) put it, “You can’t doubt your doubt without actually doubting”.

This argument could be phrased more formally as something like:

  1. If I am doubtful or skeptical about everything (which I am), then I am skeptical about my doubt or skepticism.
  2. If I am doubtful about that (which I am), then I am skeptical.
  3. If I am skeptical (which I am), then there is something of which I can be sure.
  4. Therefore, there is something of which I can be sure and so I don’t doubt everything.

The problem with this argument is that it’s really a reformulation of the original one as applying to any skepticism rather than “doubting everything”. It, especially 3, presupposes that the “escape routes” (belief rather than a knowledge claim, assumption ex hypothesei) mentioned above cannot apply. We can show this by assuming that 1 and 2 and the claim of doubt in 3 are, say, just probabilistic opinions. Therefore, this argument seems to just be circular or question-begging against the skeptic.

3. The argument from meaning of concepts

One claim that has been made is that even the statement “There are no certainties, only probabilities” is self-refuting for presupposing the concepts and being able to refute them. The problem with this is that it, again, presupposes that the “escape routes” (probabilistic opinion, assumption ex hypothesei) are not acceptable.

Perhaps the best argument that can be made against this is the thesis that probability presupposes certainty, as per the “all probability reduces to nothing” argument which is as follows:

  1. A proposition is X% sure, and X is less than 100
  2. The reasonability of the estimate is itself uncertain and has a probability of Y%, hence adding uncertainty and making the original proposition less sure.
  3. Same goes for that too, adding still more doubt.
  4. And so on and so forth, making the original probability drop and eventually vanish.

A counterexample:

  1. The original proposition is 90% sure.

2. That estimate is 95%.

3. That estimate is itself 97.5%.

4. The original estimate does not vanish, but asymptotically approaches 80%.

Some argue that probability presupposes the probability calculus (The axioms and tools that define, constrain and determine probability), but this must still contend with the “escape routes” mentioned above.

4. The infinite regress objection

Some respond to the above by saying something like “How do you know that you don’t know if but believe that you don’t know?”, attempting to imply that the response just “kicks the can down the road”, shifting the problem rather than solving it. This, however presupposes that that can’t itself be just a probabilistic opinion, or an assumption ex hypothesei. So, it still has not been shown why a (potentially) infinite regress of uncertain probabilistic estimates (as opposed to knowledge claims) or, for the Pyrrhonian, assumptions ex hypothesei or “non-standard” arguments as described above in “Stefan Molyneux and UPB” is unacceptable. Of course, if you do think that probability presupposes certainty or that arguing requires a truth claim then you might have an argument. The one about probability presupposing certainty might be supported by Hume’s all probability reduces to nothing argument which, however, it seems is mathematically unsound. It’s possible for it to converge to a reasonable degree. The notion that arguing requires a truth claim is poorly founded as per the article, but that’s probably a topic for another post. Either way, the argument still has the charges of the argument from ignorance and tu quoque fallacies, as mentioned above, to contend with.

5. Is the dogmatist really much better off?

The open-minded skeptic could respond to the opponent of such by saying “How do you know that you DO know, or can dismiss, something?” or “Is it not possible that everything and anything IS possible?”

No, I don’t = The skeptic might be right.

Yes, I do = presupposition that that can be known, presupposing that something can be known or dismissed, hence circular or question-begging or self-presupposing.

So, if anything, the skeptic’s position might actually be, instead of self-defeating, self-reinforcing.

6. Conclusion

The argument that skepticism is self-refuting commits the following fallacies or mistakes:

•Ignoring the possibility of a moderate opinion, for the moderate, as opposed to a knowledge claim

•Ignoring the possibility of an assumption ex hypothesei for the radical skeptic

•Argument from ignorance

•Tu quoque

•Assuming that the anti-skeptic doesn’t also have a problem of presupposing that the skeptic is wrong.


5 thoughts on “Skepticism’s alleged self-refutation, refuted

    1. The “all probability reduces to nothing” argument: P has some uncertain value, and so does that estimate, and that too, and on making the original P ever more doubtful, leading to it being indeterminate.

      A counterexample:

      P is 90%
      The estimate of that estimate: 95%
      The estimate of that: 97.5%
      And so on.
      This converges to ~80+%


  1. Excellent post! I wonder, if a pyrrhonian skeptic can affirm pyrrhonian skepticism as a belief, rather than a claim to knowledge, does that mean that a pyrrhonian skeptic can have beliefs other than their belief that pyrrhonian skepticism is true? Can a pyrrhonian skeptic believe in the existence of God (or gods) and still be a pyrrhonian skeptic or must the pyrrhonian skeptic have no belief either way?


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s