changing a little bit the famous black swan example to:

not all swans are white

it seems to be unfalsifiable, but in the end (when we discovered australia), it turned out to be true.

  • 1
    Is not this statement unprovable but falsifiable, as demonstrated?
    – User65535
    Apr 6 at 15:18
  • If statement A is unfalsifiable, then "not A" is too. Then unless A is non sense ("dichotomy is more to the left than the color blue") either A or not A has to be true.
    – armand
    Apr 6 at 22:58
  • All black swans are black. Unfalsifiable. Also true. Analytic statements are not falsifiable, ie mathematical statements - accept premises, accept conclusions, QED.
    – CriglCragl
    Apr 6 at 23:16
  • tautologies cannot be shown to be false, yet are true
    – Dave
    Apr 7 at 18:17
  • thanks for your comments @CriglCragl and Dave, but I don't see why "not all swans are white" would be an analytic statement nor a tautology. could you please elaborate it?
    – csfb
    Apr 7 at 19:04

2 Answers 2


1.) The claim

not all swans are white

has been verified. Hence the claim is not falsifiable.

2.) The point of Popper's falsificationism is the falsification of the claim

all swans are white

The latter claim has been derived from induction. Popper used this counter-example to show that general statements cannot be derived from induction.

3.) Falsfication and verification are not symmetric: A statement of the form "There exists ..." can be verified, a statement of the form "For all ..." can be falsified. The statement from part 1) is equivalent to "There exists a non-white swan" and can verified - and has been verified indeed.

Changed part 1) after the OP simplified the question. I deleted also my corresponding comments.

  • 2
    "Not all swans are white" can be verified, but that has no bearing on whether it can be falsified, which is OP's point. There is no set of swans we could observe - even in a hypothetical universe in which "not all swans are white" is false - that would falsify the statement, "not all swans are white."
    – causative
    Apr 6 at 15:50
  • 1
    Well, that's debatable. Anyway, your comment here is agreeing with the OP and contradicting your own answer; you are agreeing that "not all swans are white" cannot be falsified. And yet it is true.
    – causative
    Apr 6 at 16:04
  • thanks for answering @JoWehler ... but why do you think both "not all swans are white" and "there exists at least one non-white swan" statements (which I agree to be equivalent) are falsifiable? to falsify them, you need to be able to demonstrate that all swans are white, right? but this doesn't look feasible. let's suppose you don't know the existence of black swans in australia, you go to the closest park and all swans are white. then I tell you, this doesn't demonstrate that all swans are white, go to the next neighborhood, and so on...
    – csfb
    Apr 6 at 18:16
  • 1
    @JoWehler Your own answer says, "Hence your example does not support the existence of 'unfalsifiable statements that, in the end, turn out to be true'." But you are claiming in the comment that the statement OP gave is unfalsifiable (because it is verifiable) and of course it is true. So that does make it an example of an unfalsifiable statement that turned out to be true.
    – causative
    Apr 6 at 18:36
  • 1
    “Hence the claim is not falsifiable.” But “not all swans are white “ is indeed falsifiable. The claim has been resolved, to be sure. But the original claim, as it is stated, remains falsifiable. Apr 7 at 4:28

This is where Ibn Sina's insight that logic should be temporalised would prove its value. As well as the epistemology that underlies it.

We should say, a certain statement is unfalsifiable now and in the forseeable future given what we now know.

  • To say something is falsifiable is to say it could be falsified at some point in the future. So, to say something is unfalsifiable "now" is to say it is unfalsifiable at every point in the future as well. Because, if it were falsifiable at some time t in the future, then it could be falsified at some time t2 in the future of t. But t2 is in the future of the current time, meaning it is falsifiable "now" as well.
    – causative
    May 8 at 17:13
  • @causative: Ibn Sina's temporalisation of logic is a lot more sophisticated than I have given above. I'm merely giving the gist of it so the idea can be understood. Constructivists, for example, would have no truck with how you define falsifiable as we do not know what we do not know. May 8 at 17:23
  • What does knowing have to do with it? Whether something is falsifiable is independent of whether we know it is falsifiable.
    – causative
    May 8 at 17:25
  • @causative: Really? If you falsify something, then you know how to falsify it. Look up modal logic. May 8 at 17:27
  • Yes, if you actually falsify it, you knew how. But it was falsifiable even before you knew specifically how. Falsifiable simply means "possible to be falsified." We are not aware of all possibilities, but that doesn't stop them from being possibilities.
    – causative
    May 8 at 18:11

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .