Consider the sentence "It might be dark outside". Is this a proposition? What sorts of arguments could you make that this sentence isn't a proposition?

Is it that it is making -- potentially -- a claim about the states of affairs of other worlds (as opposed to the actual world) and hence can't be a claim about the actual world and hence can't be a proposition? (Genuine question here).

Basically, I'm trying to understand the view which states "not all sentences have propositional semantics; indeed, many of them just express properties of attitudes".

  • According to writers like Kratzer, propositions such as it might be dark outside are only interpretable according to the evidence that you have. If you're outside and it's daylight, then when I say "It might be dark outside" you'll say it's false - but if you have the same evidence as me, you'll say it's true, if you think it's a possibility. Kratzers take would be that it's true if it accords with the evidence a speaker has. If this holds then it undermines, I suppose, the idea that it might be dark outside is a proposition as opposed to just a reflection of a speaker's epistemic state. Commented Oct 13, 2014 at 2:11
  • But then couldn't we just take "It might be dark outside" to be equivalent to "It is consistent with such-and-such's evidence that it is dark outside". And wouldn't THAT be a proposition?
    – George
    Commented Oct 13, 2014 at 2:14
  • Erm, good point! Let me think for a second. Any thoughts, while I'm thinking about it, on my answer to your question here Grice I'll delete this comment in a little while! Commented Oct 13, 2014 at 2:24
  • Well, I suppose that's an argument, but it poses the following problem. If you're outside and it's broad daylight and I'm inside and for one reason or another darkness is compatible with my evidence, why is it you can disagree with my statement? If my statement actually consists of a proposition about my epistemic state then you shouldn't be able to say. That's not true!. Also stealing your useful computer thingie, suppose a computer only spits out "It might be/can't be/isn't dark outside" statements. It seems to me you could say they were true/false even though they're not evidence based .. Commented Oct 13, 2014 at 2:32
  • Then alter the proposition as follows: "It is consistent with an ideal knowledge source's evidence [defined in such-and-such a way] that it is dark outside". The definition of the ideal knowledge source could allow it to make sense to disagree with other's declarations. Then this, too, might just be a proposition.
    – George
    Commented Oct 13, 2014 at 17:23

3 Answers 3


Propositions are moves in a language game that intends to describe things. They have truth value, in the sense that what they propose is meant to be considered as an assertion describing reality. But there are other language-games, which are not primarily concerned with this task.

Statements of possibility are generally not informative. They are moves in the game of thinking itself, or in some sub-game about thinking, like auditing understanding. They inject a parameter over the range of possible worlds, and request that one's listener consider whether the parameter with possibility has been considered adequately. In doing so, they generally put you farther from certainty (correct or otherwise) rather than closer to it.

They are not even informative statements about alternative worlds. When would one say "It might be dark."? Not when one thinks this is information the listener does not have. If he did not know of this possibility, he would need a lot more information about the nature of light, in order to care. And any of that information would be more useful to contribute. Instead one says this when one thinks the plans so far do not adequately consider that dimension.

Since they do not cover descriptive ground, but instead open new ground, they do not contribute directly to understanding in the same way. So such statements do not play the role propositions are meant to fulfill. It is possible to interpret them as propositions, but doing so does not allow them to perform their intended function. If someone contributes "It might be dark." and your only response is to affirm the possibility, he has failed to communicate.

(Answering Pacerier's silly question, if Mary asks whether it is dark outside, and John replies that it might be, he is agreeing that he had not to that point considered whether or not it was, and whether it should have any effects. He might only be so agreeing in excessive politeness, if he is fairly sure it does not matter.)

  • Base on your answer, if Mary asks "Is it dark outside?" and John replies "It might be dark", what does John mean? How should Mary interpret the reply?
    – Pacerier
    Commented Oct 10, 2014 at 19:52
  • That is the very rare situation in which your answer would apply. He is hundreds of times more likely to use this phrase when it it not an answer to a question and hundreds of times more likely to answer "I don't know' than give this answer. You are acting as though information exchange is the only reason to say things. So you are framing this in a totally unnatural way. The more likely setup is when there is no question, and someone says the phrase, expecting you to check your logic.
    – user9166
    Commented Oct 10, 2014 at 20:27
  • It's not rare to ask someone the question "Is it dark outside?" Indeed, my example use case is no less common nor contrived than your example use case. And also, Meaning exchange is the only reason to say things to others. Every utterance sent to a receiver has an intended meaning by the sender.
    – Pacerier
    Commented Oct 10, 2014 at 21:36
  • It is rare AS I STATED for the phrase in question to be considered an answser. One is much more likely to simply admit one does not know, than to tell the listener something they are already perfectly aware of. After all, who would ask the question if they did not know that it might have either answer?
    – user9166
    Commented Oct 13, 2014 at 14:39
  • @Pacerier I prefer Wittgenstein over you. There are multiple games, and propositional information is only one. Co-ordination of actions with others is a separate use of language.
    – user9166
    Commented Oct 13, 2014 at 14:44

A proposition is-equal any message that has a "truth value". Consider this sentence:

"It might be dark outside".

Does it have a "truth value"? If yes, then it's a proposition. If no, then it's not.

The first step is to identify the meaning of John's message. It's important to understand that the exact same words can have different meanings depending on context.

Consider this scenario:

Mary asks:

Is it dark outside?

John replies:

It might be dark outside.

whereby John meant:

It might be dark outside. Maybe it is. Maybe it isn't. I don't know.

then John is not making a claim regarding the "darkness of outside". He is claiming that he has no knowledge of the "darkness of outside". Since that has a truth value —could either be true xor false—, the message is a proposition; not a proposition regarding the "darkness of outside", but a proposition regarding John's (own) knowledge.

However, if John meant:

It might be dark outside. There is a chance that it is dark outside. (There is above zero probability that it is dark outside).

then John's message is a proposition regarding the "darkness of outside". It has a "truth value":

  • It is true if there is above zero probability that it is dark outside.

  • It is false if there is exactly zero probability —no chance— that it is dark outside.

  • It must be either true or false.

If we tweak the scenario and change John's reply to:

It might be dark outside, but so what?

whereby John meant:

I don't care if it's dark outside. Why do I need to care?

then John's message is not proposition because it is a question and it can neither be true nor false.

  • Yes, but someone who is asserting that it might or might not be dark outside is not usually indicating that they do not know. They are usually indicating that you may have omitted this consideration from your prior deductions. They are injecting a parameter on the set of possible worlds into your frame of reference for consideration. So interpreting this as a proposition that the speaker is uncertain is not in keeping with the meaning. It is a move in a different kind of game: one of care-taking you or auditing your reliability. Not one describing the world.
    – user9166
    Commented Oct 9, 2014 at 22:53
  • 1
    @jobermark, As I've mentioned, the context determines everything. I've created 3 example scenarios to demonstrate that the exact same words can mean different things in different context. You can inject another example scenario and the same words will mean something else altogether as you've mentioned.
    – Pacerier
    Commented Oct 10, 2014 at 1:26
  • The point, I guess is that this is not generally an answer, but a spontaneous, helpful suggestion. So you are spending most of your answer on highly unlikely cases. Your last case touches on the most common case obliquely, but it dismisses it as a question, when it isn't one. It is in effect a polite request for action -- "please review your decision in light of this information (which I am not conveying, but already know that you know)".
    – user9166
    Commented Oct 10, 2014 at 17:33
  • @jobermark, I don't actually understand your comment. What exactly is this "common case" that you are talking about? Also, who is making the suggestion — John or Mary?
    – Pacerier
    Commented Oct 10, 2014 at 19:43
  • The common case is when you are in the middle of planning an outing, and someone says "It might be dark outside", so you take a flashlight. It is seldom given as the answer to a specific question. Having it be an answer is an unusual framing of the phrase. Come on, man, think about when you have heard a similar phrase, and pick the most common case!
    – user9166
    Commented Oct 10, 2014 at 20:23

I would say that the statement "It might be dark outside" is logically equivalent to the statement "It is either dark outside or it is not dark outside".

This is a tautological statement, i.e., a fact in the sense that it cannot be false.

Facts are not propositions.


Following virmaior's comments below, I will attempt to justify / correct my answer.

One might argue that the equivalent form "It is either dark outside or it is not dark outside" is an instantiation of the second law of thought - i.e., the law of the excluded middle. As such, it would not be considered a proposition. This is because our laws of thought are given, and therefore not subject to a change of status.

Some statements may fulfil the definition of tautology, but at the same time possess other attributes which override their status as a tautology. For example, an instance of a law of thought or an axiom in disguise. They are redundancies or unwanted baggage that result from our methods of formalization. One might say they are artefacts. We begin with what we are given, and we then generate our propositions. We do not generate our laws of thought.

The term proposition has a meaning in general philosophy beyond that of propositional logic, and the wiki entry for proposition states that according to general use of the term, facts are not propositions.

  • 1
    I'm confused. Are you stating that "It is either dark outside or it is not dark outside" cannot be a proposition? A tautology is one of the most basic types of propositions (en.wikipedia.org/wiki/Tautology_%28logic%29)
    – virmaior
    Commented Jun 7, 2015 at 4:21
  • @virmaior You're probably correct. The wiki entry on proposition states one view that " Propositions are the sharable objects of attitudes and the primary bearers of truth and falsity. This stipulation rules out certain candidates for propositions, including thought- and utterance-tokens which are not sharable, and concrete events or facts, which cannot be false." Since the OP was looking for a justification for ruling out the possibility of the statement being a proposition, I thought this might be one way of doing so.
    – nwr
    Commented Jun 7, 2015 at 4:52
  • 1
    The claim "facts are not propositions" could use some more explanation, if you ask me. If, by your definition, propositions are bearers of truth and falsity, how can something that is true (a fact) not be a proposition?
    – user2953
    Commented Jun 7, 2015 at 7:24
  • @virmaior I have edited my answer to include some justification for my response. It may sound a bit weaselly, but it is the best I can do currently.
    – nwr
    Commented Jun 7, 2015 at 17:49
  • @Keelan I have edited my answer in an attempt to provide some justification. I hope it is more satisfactory.
    – nwr
    Commented Jun 7, 2015 at 17:51

You must log in to answer this question.

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