I was in an online debate and had this statement presented to me.

I would note further that your apparent positivism rests on what is logically a faith claim - specifically, the unproveble claim that nothing counts as true without being supported by logic and empirically verifiable data.

Now granted, I suppose you could call this a faith claim but only because I have never considered an alternative, nor been taught an alternative. So... my question would be:

  1. Can something "count" as TRUE without being supported by Logic?
  2. Additionally, is empirically verifiable data a pre-requisite for allowing something to be "counted" as TRUE?
  • Can your percieving of a color in plain daylight be supported by logic? And is having an imagination of a unicorn empirically varifiable? Both might be true.
    – Philip Klöcking
    Commented Nov 24, 2015 at 19:34

4 Answers 4


This question needs quite a bit of unpacking. For a start, one must distinguish between truth and evidence. To a realist, something may be true without there being any evidence for it. I have no evidence as to whether it rained on Lands End in England at noon on May 1st of the year 1 CE, and I strongly suspect that no such evidence exists or will ever come to light. But I believe that proposition is either true or its negation is true. I have no evidence as to whether matter existed prior to the big bang, but I'm willing to believe that either it did, or it didn't.

This means that the primitive idea that only verifiable sentences are meaningful is simply false. Furthermore, sentences, and even whole theories, typically are not verifiable at all, but only falsifiable. Even if we accept a kind of inductive reasoning, such as that of bayesianism, we cannot speak of verifying hypotheses or theories in any absolute sense, but only in comparison to rival hypotheses. If I have two rival hypotheses A and B, and A gives a better account of the data than B, then I'm willing to say that A is confirmed relative to B, but I cannot rule out that an even better hypothesis C might come along in future and trump both of them.

I suspect that despite your question being framed in terms of truth, what matters more to you is whether one has evidence for a belief. This question might be phrased, should one believe anything without some evidence? Even this is a highly contentious claim. To begin with, you have to address Agrippa's Trilemma. The only thing that can provide evidence for a belief is apparently another belief, so where does evidence come from? Any evidential claim must either (1) be circular, (2) involve a regress, or (3) involve an appeal to a belief that is infallibly and indubitably true. None of these options is very appealing.

Even if you think you can surmount this hurdle, one must address conventionalist objections that what counts as evidence depends on all kinds of methodological and linguistic assumptions. So you and your opponent may be working with different criteria of what counts as evidence with respect to some domain. What counts as evidence of the existence of God, for example? (Wittgenstein, and the later Carnap, regarded theological language as a distinct language game with its own rules that allowed assertions to be made that didn't admit of falsification in the usual way.)

And even if, in straightforward scientific cases, all this can be resolved satisfactorily, does it generalise to other kinds of judgement, e.g. moral or aesthetic judgements? Is it true that one ought to do unto others as you would have them to do to you? Is it true that Adolf Hitler was a bad man and Albert Schweitzer was a good man? Is it true that wisdom is a virtue and greed is a vice? Is it true that Bach's music is beautiful and my tuneless whistling is not? One might reasonably have firm beliefs about any of these without being able to explain how it is supported by logic or empirical data.


There are many cases where human beings find it meaningful to count something as TRUE without a logical support. Terms like "gut instinct" and "flash of inspiration" appear all over our societal structures.

One of the paradoxical loops that arises with the term "empirically verifiable" is that one has to define that criterion. How does one determine that that is the true criterion? Empirical verification of course -- and the cycle continues. Either:

  • There has to be a root of that cycle, which is accepted without empirical verification.
  • You have to rely on non-standard set theories to describe the mathematical structures with which empirical verification occurs (these cycles cannot be described with a well-founded set theory)
  • Or you have to decide on another approach to truth.

Another thing to consider: perhaps there are other definitions of "true" and "false" which are more nuanced than those you use currently. Both are just words, four and five letters long respectively. It's the concepts behind those words that are so powerful. However, logic does not define True nor False. In fact, if you look at the fundamentals of mathematics, True and False are not constructed using predicate logic or anything like that. They are constructs that are defined implicitly by defining their behaviors with respect to operators (such as X OR False -> X)

One utilitarian point of view would be to declare that the version of "true" you are using can never possibly be attained via empirical evidence, because there are always questions of whether we perceive what we perceive. Such a truth becomes a mathematical oddity, creating a skeptic mindset. However, if we recognize that, we can repurpose "true" to also include things where determining their truth-ness is actually more expensive (from a utilitarian POV) than the gains which one can get from knowing the answer. Why not declare something to be true when one's value is increased by such a declaration, even if one is ontologically wrong.

This is certainly not the only definition of "truth" out there, but I provide it as a tool to kindle ideas for other meanings. For example, many religious individuals "count" something as true without any empirical evidence for it at all (or, in some cases, empirical evidence which cannot be described to the non-initiated). Thus one can say there are many many individuals who have other definitions of truth besides the one you are using, and you have the freedom to explore them all, if you so wish.


My short answer is: You are right, nothing counts as true without being supported by logic or empirically verifiable data.

In the second approximation I would like to make the following refinement: Truth is a property of propositions. In general, the truth of mathematical propositions can be proved - neglecting the issue of undecidable propositions. In the field of natural science general propositions are hypotheses. We cannot prove them, but we can either confirm or falsify them. Hence your term "support" has a different meaning in both cases.

Note. In my opinion, Corts examples concerning "gut instinct" and "flash of inspiration" from the beginning of his answer do not relate to the objective property of truth but to the subjective property of certainty.

  • Again, it is an objective, but not empirically or logically supportable truth that my totally subjective state is thinking of a pink unicorn right now.
    – Philip Klöcking
    Commented Nov 24, 2015 at 22:55
  • @Philip I consider your thinking of a pink unicorn a fact (German: Tatsache) not a truth. Because I consider truth or falsehood a property of propositions not a property of circumstances (German: Sachverhalt).
    – Jo Wehler
    Commented Nov 24, 2015 at 23:01
  • 1
    Then you should clarify that your answer is bound to Moore/Russel, logical atomism and so on. Going with Sellars by reducing knowledge to propositions in this sense, we buy in the problem of defending non-inferential knowledge without the Myth of the Given, which according to him is essentially impossible.
    – Philip Klöcking
    Commented Nov 24, 2015 at 23:26

Wittgenstein's Tautologies. A tautology is a proposition that is always true regardless the true-values of its constituent propositions.

For example: p v ~p

 p        p v ~p
 True     True
 False    True

As you can see, the truth-value of p v ~p does not depend on empirical facts, nor is it derived from any primitive propositions or axioms.

As Wittgenstein pointed out, all the asserted symbolic propositions in Whitehead & Russell's Principia Mathematica are tautologies.

Take Syll ✳2.05 for example:

p    q    r    q⊃r.⊃:p⊃q.⊃.p⊃r

F    F    F     T 
F    F    T     T
F    T    F     T
F    T    T     T
T    F    F     T
T    F    T     T
T    T    F     T
T    T    T     T 

You must log in to answer this question.

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