Does answering “I don’t know” actually answer a direct question?
Recently in discord, yes I was actually hanging out in discord, I was talking to someone who answered “I don’t know” to many questions I had and it got me thinking about when is ‘I don’t know” an acceptable answer to a direct question and when perhaps it is not.
Let’s assume arguendo a closed question to mean a question really only requires a simple “yes” or “no” answer as opposed to an “open question” which can essentially have an infinite number of possible replies. (Note: A closed question could also include a limited amount of choices as a response like multiple choice). Let’s also assume that “answer” means a proper response that fulfills the requirement of being a suitable answer to the question being asked, as opposed to a “reply” which is just any response given to the question posed. An answer here being a subset of a reply.
Given the closed question of “Does God exist?” it would seem to me as a closed question there are still only two possible closed answers. Yes or No. However, there of course people who just can not answer that question with any type of conviction as they do not hold a belief that God exists nor do they hold a belief that God does not exist. In this case it seems apparent that “I don’t know” would be a suitable answer to the question, even though it isn’t a closed answer. But what is being asserted here when someone answers “I don’t know.”? It seems to me to be merely an expression of ordinary epistemic doubt versus the subject expressing “S does not know p”. “I don’t know” and “I do not know p” relate two vastly different things. The first being S that is merely saying that they are unable to answer in the affirmative or in the negative. If however someone is unable to answer and in order to justify the answer of “I don’t know” it further seems to me that would require some type of justification as to why the S was unable to evaluate the question. The second being S claiming they do not assert that they know p is true or false if p is the proposition of the question. If the question is “Does God exist?” the proposition being asked to be evaluated is “God exists”. Where “yes” holds p=”God exists” as being true and “no” being where someone holds p=”God exists” as false (or ~p=”God does not exist”).
This often tends to lead to epistemic confusion. As people hear or say “I don’t know” and mistakenly think it has to do with “knowledge” (knowledge being the tripartite condition of p being true, S believing P, and S being justified to believe p. Where S is the “Subject” and p is the proposition) when it doesn’t. It merely is an expression of ordinary epistemic doubt relaying that S was unable to answer the closed question with a closed answer.
However, given the closed question of “Do you believe God exists?” answering “I don’t know” seems to me more of a reply than an answer. This is not an ontological question about God existing, it is an epistemic question about if S holds the belief that God exist or not…even though the epistemic question does relate to ontology it just seems to me more of an epistemic question about one’s belief than it is asking what is the belief of S about God existing. Here you once again have two possible direct answers: Yes or No. If “yes” then S does have a belief God exists (theist) and if “no” then S is saying that they do not have a belief that God exists (Not theist/nontheist). While “yes’ indicates a positive epistemic status, “no” does not indicate a negative one. It merely indicates that they do not hold the belief God exists, not that they are explicitly saying they believe that God does not exist…but if one does explicitly hold the belief God does not exist they would of course answer “No” as well. This indicates that an answer of “No” could either be because S merely has no belief either way (agnostic on p) or S believes p is false. In either case S does not believe that God exists.
Does it then really make sense to say “I don’t know” when asked about if one holds a belief or not? Personally I don’t think so. I think someone would be in a state of epistemic confusion if they are actively aware if they hold a particular belief or not since in classical logic it is binary. It is either the case S believes p, or it is not the case that S believes p. If S is unable to answer “Yes” to the question of “Do you believe if God exists” then it implies to me that they are not convinced and therefore do not have a belief that God exists and thus do not have the belief by the principle of bivalence (A declarative sentence expressing a proposition has only one truth value: True or False): It is either true that S believes p or it is not true (false) that S believes p. Thus if someone answers “I don’t know” to this question it is to me a reply, but is not a rational answer to the question.