I've recently been critiquing some fluffy answers on another stack exchange and I been encountering this type of fluffy evasive answer:

The question asks about the future; all future is unpredictable; ergo the answer is unknowable; ergo the question is a pineapple.

But I can't quite put my finger on what fallacy or branch of philosophy they are using to make this evasive statement.

Edit: I've tightened up the question by changing proposition "the future is unpredictable" to "all future is unpredictable" as on further reflection almost all the non sequiturs I've encountered are evasively implying the question can't be answered on principle than on question specifics.

  • 1
    Can you be a bit more detailed about what the specific context here might be? What exactly are you looking for someone here to explain to you?
    – Joseph Weissman
    Commented Oct 21, 2013 at 23:55
  • @JosephWeissman Unfortunately as the answers themselves are evasive context-free non-sequiturs, I can't provide useful extra context. In attempting to add extra context I'd move further away from the class of answers I'm seeing. What I'm looking for a description of the fallacy, rhetorical trick or branch of philosophy being used to prop-up these answers. Commented Oct 22, 2013 at 0:00
  • I was directed to place the question here instead of in the Skeptics stack exchange, although that stack exchange would certainly have experienced this phenomena also. Commented Oct 22, 2013 at 0:03
  • It's a bit difficult, or at least would require some degree of speculation, to identify a fallacy without a more concrete sense of the argument involved...
    – Joseph Weissman
    Commented Oct 22, 2013 at 0:31
  • I mean, the objection that a question is unknowable might be valid (especially on SE, which places an emphasis on answerability.) It doesn't necessarily mean there's some fallacy here; or at least I feel like we'd need to understand a bit more about the argument itself.
    – Joseph Weissman
    Commented Oct 22, 2013 at 0:33

4 Answers 4


It's fairly clear in this case that the problem is simply a false premise. "All future is unknowable" is simply not correct in practise. This is what I would call a naive and extreme variety of good old-fashioned skepticism.

Much of our knowledge of the future relies on induction, which is to say extrapolating from experience of the past. As such we cannot be absolutely secure in it, because inductive reasoning can fail. However, anyone who claims that knowledge of the future is impossible probably does not fear that anything they eat will spontaneously turn into sulphuric acid, or a palm tree, or turn out to have been their own severed foot. The world is much less capricious than we are capable of imagining, and in this sense practical knowledge is possible. All science, and indeed all ways of living, not to mention the existence of enduring life itself, depend on this relative tameness of nature: that it is possible to adapt and react meaningfully to it.

Given that the future is predictable — but perhaps not perfectly, and certainly not to absolute precision by any means that we can thus far imagine — we might reasonably ask what "knowledge" means in this case. How much certainty, how much precision, do we require before we say that we know how some future event will turn out? This is a nuanced question which probably does not have a simple binary distinction (or even a linear gradient) between "knowing" and "not knowing". But the point is moot: there is a meaningful sense in which we definitely know more than nothing about the future.

Thus I would say the premise that 'all future is unknowable' is for practical purposes false, and that absurd conclusions may follow from it as a result. (This possibility is one which I know from experience, after all.)


It's a non-sequitur, which means that the conclusion "doesn't follow" from the premise.


Suppose someone is about to flip a coin, but beforehand asks you guess heads or tails. The question "What side will the coin land on?" is a question about the future, and it's essentially impossible to predict the answer for a fair coin. Ergo, what side lands up is unknowable.

HOWEVER, "what side lands up?" is not unknowable. That is, just because a question's answer is unknowable doesn't mean that the question itself is unknowable.

Here's the fallacious argument again: The question is "What side lands up?"; what side lands up is unknowable; ergo, the question is unknowable.

What the argument is apparently trying to do is apply the transitive property of identity (if A is B and B is C, then A is C), a mistake which might be easy to make if you just hear the argument spoken verbally, because B can sound just like "B?". Written down, though, the syntactical difference is unambiguous.

The above fallacy might be called a use-mention equivocation fallacy.

  • Sorry. I just realised after reading your excellent answer, that I made a rather nasty typo in the question. +1 for providing an answer for the original mistyped question. Commented Oct 22, 2013 at 10:41
  • I'll need to boot up IE. My security-hardened Firefox never works with stack exchange chat for some reason. Done. In main chat Commented Oct 22, 2013 at 10:50

The reasoning does seem to have some obvious issues, but rather than trying to correct them, I think it's worth thinking of what it is at root trying to say in terms of the philosophical school of Logical Positivism.

Very much out of fashion in analytical philosophy as a methodology for conceptual analysis, the basic idea of Logical Positivism/Empiricism is that the meaning of a statement should be strictly given in terms of logical conditions under which the statement would be verified or falsified. If there are some statements under which there are no strictly logical conditions under which they would be verified or falsified, those statements are treated as strictly nonsensical. And since we cannot tell nonsensical statements apart (because they have the exact same verification/falsification conditions, i.e. none at all), nonsense is nonsense is nonsense.

So if we take it that the future really is factively undecided, such that there is no way for us to even in principle validate what will happen in the future from our current standpoint, then a logical empiricist might denounce all future case statements as devoid of meaning. And so by a simple meaning-theoretic substitution principle, they can plug any old nonsense statement they like in there.

If we say that this kind of "verification conditionality" is all there is to meaning, and that truth conditions must be strictly explained in terms of finite proof or empirical procedures, then many statements that we would naturally think to have some kind of simple understanding simply fall out as non-cognate. It's widely recognised, however, that this kind of basic form of logical empiricism is not a good fit for the practice of classical mathematics, and hence for the wider part of science as is in fact practised.

Moreover, it is all too easy for questions about verification and proof to be dominated by subjective considerations that we might reasonably pose have no business being imposed upon objective scientific analysis. Logical Empiricists have a tendency to dismiss an awful lot of things as nonsensical simply on the grounds that they do not accept certain premises. It seems a mistake to think that the actual falsity of a premise makes it logically impossible. Following work by Kripke in the 70's, metaphysicists have widely reintroduced the idea of using simple set-theoretic models to ground talk about alternative possibilities than the brutely first-order factive notion of what can be strictly proven from presently observed and reified facts. This is particularly so if we want to be able to talk in terms of an objectively probabilistic methodology in science; we would like to be able to model counterfactual situations in order to establish a kind of state space framework for probability-based scientific models.

Nonetheless, certain projects might be interested in trying to squeeze out the methodology as far as it goes, and I think we see this a lot in certain "New Atheist" philosophy inspired by evolutionary biology. Strictly grounding mathematics in neuropsychology, for instance, seems to be one way to go about it, and it's premature to think they can't get something interesting out of a "finitistic" conception of theoretical resources by going down that road.

You must log in to answer this question.

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