When must justifying a belief allow for new evidence? I suppose it depends on context. "Apodeictic" beliefs (which are necessarily true) are clearly established or beyond dispute, so don't need returning to except to make sense of them. Scientific experiments are rarely repeated. What about more everyday beliefs, either ones we take for granted (these are my real parents) or ones we may even doubt sometimes (I will get a pay rise soon): are they justified if we do not allow further reconsideration and evidence?

In the example I am thinking of, Smith is being lied about by some people, and Smith claims that all you have to do to prove his innocence is check the mail. Can anyone claim to know Smith is guilty, if no-one checks the mail (except Smith)?

  • 1
    As a comparator, you're relying on ___? Jul 15, 2023 at 11:40
  • 1
    You can claim whatever you like. There's a saying: Predicting the future is really easy, it's doing it accurately that's difficult.. People have all kinds of crazy beliefs, it's about the local norms of discourse whether others find them pursuasive. And the discourse that focuses on evidence, has contributed to debate that gets better answers than participants could alone, & the success of communities that pay attention to them
    – CriglCragl
    Jul 15, 2023 at 18:53
  • As a rule, it's best not to rewrite your question after answers have been submitted, especially when the rewrite of the question makes initial answers redundant. Jul 16, 2023 at 13:02
  • When must justifying a question allow for completely rewriting it? When you really really want an answer that justifies your position, I guess.
    – Scott Rowe
    Jul 16, 2023 at 13:44
  • no-one understood the original version @ScottRowe which is the same question at least if you have JTB
    – user66760
    Jul 16, 2023 at 17:19

4 Answers 4


Anything that can be known a priori will, by definition, not require experience. In the case of Smith, however, the answer is “no”. The concept of evidence implies relevance to the case.

Addendum: Now that the question has changed, the answer is “always”.


You cannot claim to know anything unless it’s logically necessary. So the answer is no.

Practically though, we usually use “know” to mean when it’s justified to believe something. For example, I know that the earth is a sphere. Is it possible that I’m being lied to and the earth is flat? I don’t think so. Can I disprove that the earth is flat and not some illusion in my head even if I went up to space and saw a ball? No. But I don’t think a disproof is required for reasonable belief.

EDIT: The person edited the question title and content so the answer is not fully relevant anymore. How is this allowed on SE?


It depends on the situation. Is this the tenth time Smith has pointed you at new evidence, and the other nine times you investigated the new evidence only to find it irrelevant or actually harmful to Smiths case? Investigation can be endless. If you suspend judgement until all possible evidence is in, then you might never be able to come to a conclusion.

Furthermore, some beliefs are more important than others and require greater investigation. I hear a particular tune playing outside. Given my experience of the neighborhood, I conclude that the ice cream truck is driving by outside. Now, it doesn't have to be the ice cream truck; other vehicles and devices can play that tune, but I'm justified in drawing a conclusion without going to look, in part because there is no reason to think it might be anything else, and partly because it's not important enough an issue to investigate thoroughly.

If someone were to tell me that they don't think it sounds like the ice cream truck, I might go and take a look out of curiosity. If someone's life depended on me being right about my conclusion, I'd go take a look.

So, there are at least three reasons not to gather additional evidence to draw a conclusion, even if additional evidence is available:

  1. You have reason to doubt that the additional evidence will be of use.
  2. You have practical constraints such as a deadline and have to conclude your investigation without covering all possible evidence.
  3. Given the circumstances, the matter is not important enough to warrant further investigation.
  • so does it depend on the reliability of smith, about his innocence, wrt the mail?
    – user66760
    Jul 15, 2023 at 18:10

I would imagine you can justify things without considering any more evidence to the contrary, but you may end up mischaracterising your supposed knowledge: as apodeictic or justice or even true.

In the case of someone professing their innocence, if they are not permitted to argue their case and produce evidence to the contrary, then that is a clear injustice and a violation of their human rights etc.. At least in a legal context. In a social context, such people are not your friends etc..

You must log in to answer this question.