I exist now. But my nonexistence now is possible. ('Now' picks out the same time in both of its occurrences.) 'Possible' in my second sentence is not intended epistemically. Surely it would be absurd were I to say, 'My nonexistence now is possible for all I know' or 'My nonexistence now is not ruled out by what I now know or believe.' If I am certain of anything, I am certain that I exist, and that rules out my present nonexistence. So in the second sentence above 'possible' is to be taken non-epistemically. The metaphysical point is that I am a contingent being. But how put this into ordinary English?
Suppose I say to a competent English speaker who is a nonphilosopher: 'Although I now exist, it is possible that I not now exist.' It is unlikely that he will understand me. He may even consider me crazy. So my problem is this: How do I convey in ordinary correct English the compound proposition: BV now exists & possibly, BV does not exist? The issue at present is not whether the second conjunct of this compound proposition is true, or how one knows it to be true if it is true; the question concerns the correct ordinary language rendition of the proposition.
I have more than once in these pages used such locutions as 'I might not have existed,' 'I might never have existed,' 'I might not have existed now.' They sound correct to my ear. But a professional writer friend considers them bad English. His claim is that 'might' has only epistemic uses in correct English. Is my friend right?
Compare (a) 'I might not exist now,' (b) 'I might not exist tomorrow,' (c) 'I might not have existed now,' and (d) 'I might never have existed.' I claim that (a) and (b) feature epistemic uses of 'might' while (c) and (d) feature non-epistemic uses. The linguistic impropriety of (a) is due precisely to its inclusion of an epistemic use of 'might.' In almost all contexts (a) is either nonsense or evidence that the utterer is in a very strange mental state. (b), however, is perfectly in order. It is easy to imagine contexts in which its utterance makes sense.
With respect to (c), my claim is that it is a correct English sentence and that it makes sense. Because it does make sense, the 'might' it features expresses metaphysical possibility as opposed to epistemic possibility. The same goes for (d). It is an acceptable English sentence which is evidence that there are legitimate non-epistemic uses of 'might.'
But suppose my writer friend persists and starts whacking me upside the head with his badly dated copy of Fowler. Then I would challenge him to say what I want to say without using 'might.' How would he rewrite (c)?
Consider (e) 'I could not have existed now' and (f) 'I could not exist now' and (g) 'I can not exist now.' Of these three only (f) is serviceable, but it strikes me as distinctly inferior to (c).
My conclusion is that there are legitimate non-epistemic uses of 'might.' My friend also opined that 'could' is used correctly only in a non-epistemic way. But that too strikes me as false. Suppose I ask my wife, 'Does Mike Baumer still teach at Cleveland State?' She could with all due linguistic propriety reply, 'Could be!' She would not thereby be expressing any real possibility, but only an epistemic possibility. For all she knows, Mike is still there.
"But a professional writer friend considers them bad English. His claim is that 'might' has only epistemic uses in correct English. Is my friend right?"
--------------------------------------
I am not a professional philosopher, but I am not able to get my head around why anyone would think that "might" is an epistemic term. It simply seems obvious to me that "might" refers to contingency, which pertains to metaphysics much more than epistemology.
I fear that I am seriously missing something, as the case seems an obvious one. Nothing worthy of a blog post is that obvious, so what am I missing?
Posted by: Kevin Currie | Thursday, February 12, 2009 at 03:10 PM
Kevin,
You are missing something. Many if not most uses of 'might' are epistemic. Suppose a soldier is going into battle tonight. He has a lawyer draw up a quick will with the explanation 'I might not be around tomorrow.' 'Might' in this use expresses epistemic possibility. The soldier is expressing the idea that, for all he knows, he will not be returning from the battle alive.
'I wonder where my keys are; I might have left them in the office.' The idea here is that the proposition *I left my keys in the office* is possibly true for all the speaker knows.
Posted by: Bill Vallicella | Thursday, February 12, 2009 at 03:41 PM
Hi Bill. As neither a professional writer nor a professional philosopher, but an amateur in both disciplines, I find myself in principled agreement with your friend, and in practical agreement with you. I'd say that "might" is still most correctly used epistemically; however, English is by nature a flexible language which—regardless of the consternation of purists—is being constantly adapted to use existing words in new ways. The continual verbification of nouns is a great example.
I wouldn't immediately notice anything wrong with saying that you might not have existed. But on closer examination and reflection I would be inclined to say that, although I have no problem with you saying it that way, I myself would say either:
1. I could perhaps not have existed; or
2. I may not have existed.
I would favor (2) for its simplicity. And I would flick your friend's ear for trying to stringently grammatical in a language which defies stringent grammar. That's the second sentence I've started with a conjunction, for example, and I don't consider it wrong—aesthetics or emphasis can trump the ordinary rules of grammar just as much as philosophical clarity. My ending a sentence in a preposition for the sake of flow isn't even unheard of. Maybe your friend should learn French.
Regards,
Bnonn
Posted by: Dominic Bnonn Tennant | Thursday, February 12, 2009 at 03:58 PM
Ironically, looking at (2) now, I want to rephrase it again:
2*. I may have not existed.
This seems, to my mind, the perfect balance between grammar, aesthetics, and philosophical precision.
Posted by: Dominic Bnonn Tennant | Thursday, February 12, 2009 at 04:01 PM
Bill,
Thanks. I see what he is getting at now. "Might not be around," is taken to signify an indeterminacy in knowledge of whether he will be around.
But I still see this as a strange use of the word "might," and one that does not accord with the way most people seem to use it.
When I say that I might not be around tomorrow, I am not only referring to the epistemic indeterminacy, but more importantly, to the metaphysical indeterminacy. One might best express this by saying: "Tomorrow, the world may not have me in it." It is expressing a metaphysical contingency first and (I suspect) only an epistemic contingency ("I am not sure whether I will be around,") after.
So, if I say, "Bill might respond to this post but might not," I am first and foremost stating something about whether or not it is plausible that you may or may not respond. Only secondarily am I saying something about whether I think you may or may not respond.
I can see where your friend is going, but think he is taking a rather strangely limited view of "might"'s potential.
Posted by: Kevin Currie | Thursday, February 12, 2009 at 04:26 PM
Hi Bnonn,
Good comments. I was amused by "The continual verbification of nouns is a great example." That very sentence contains an example of what the sentence is about.
I consider myself a linguistic conservative, but not a hidebound linguistic conservative. Here is an example. The correct spelling of 'tranquillity' is just as I have spelled it. But what work does that second 'l' do? No work. It is there because of the Latin tranquillitas. So if the second 'l' were to fall into desuetude, I would not make too much of a stink, though as a conservative I am averse tochange for the mere sake of change. But when people fail to observe the distinction between the subjunctive and indicative moods, that is a more serious matter. For that is a logical distinction antecedent to natural language.
I am a prescriptivist, but my friend could be even more of one that I am. He could say to me: 'Yes, there are in fact non-epistemic uses of 'might,' but there ought not be. One issue is whether everything I want and need to say with the non-epistemic 'might' could be said with 'could' or some other word of ordinary language.
There is also the question of the source of the prescription if one is a prescriptivist. Can't be usage; must be logic and ontology. But I don't have the time to explain that. Consider 'she' used gender neutrally. That contravenes old-time usage and smacks of Political Correctness, which I despise; but, logically, if 'he' can be used gender-neutrally, why not 'she'?
You suggest: 2*. I may have not existed.
I'm sorry, but to my American ear that is so far from any idiom known to me that I think it would fail to communicate the proposition I wish to communicate. 'I could not have existed' is better, except that it suggests an ability to not exist, which is not the idea.
Posted by: Bill Vallicella | Thursday, February 12, 2009 at 05:06 PM
I tend to agree with you regarding not changing for the sake of change; but I do prefer to adopt the option which seems most sensible. For example, although "awesome" is a compound word, it seems to me that the "e" does no work in modern English, and so I habitually spell it by the fairly uncommon variant "awsome". I also prefer US spelling, despite being a South African living in New Zealand (both Commonwealth countries of course), since it doesn't contain any odd French accretions which are incongruent with the original Latin roots; American English is more simple and consistent.
Indeed. I don't have a problem, in principle, with "she" as a gender-neutral pronoun. What I do have a problem with is people using it just because they're feminists and think that "he" is somehow sexist. I myself have used "she" in the past to differentiate the gender-neutral Christian ("she") from God ("he"). This is an obvious case where "she" is employed for a good and useful purpose rather than to merely make a statement.I must confess that surprises me. I don't consider myself greatly familiar with American idioms, but certainly I speak to many Americans online routinely, read American blogs, and watch plenty of American television and movies. I'd never have suspected that "I may not have existed" would fail to relay the proposition I intended—I'd have accepted that it would sound perhaps odder to you than to me, but for it to completely fail surprises me. What would you take it to mean?
That said, your proposal of "I could not have existed" sounds to me not merely like it implies an ability to not have existed, but a counterfactual necessity of non-existence. For example, I might say that, if my parents had never been born, "I could not have existed." A strange animal, English.
Regards,
I'm a clever fellow, aren't I. Notice also that my sentence about ending sentences in prepositions itself ends in a preposition. Truly I am more than half a wit.Bnonn
Posted by: Dominic Bnonn Tennant | Thursday, February 12, 2009 at 07:51 PM
Bnonn writes, "Indeed. I don't have a problem, in principle, with 'she' as a gender-neutral pronoun. What I do have a problem with is people using it just because they're feminists and think that 'he' is somehow sexist."
That's exactly what I would say. There is nothing sexist about 'he' used gender-neutrally, and if there were, then it would be equally sexist to use 'she' gender-neutrally. And there is nothing sexist about 'man' used to refer to human beings. Ayn Rand uses it often. Standard English certainly did not keep her down!
On the other hand, I sympathize with women who feel that 'he' slights them in the way that 'coed' slights them, and so I have no objection to their use of alternative expressions.
Posted by: Bill Vallicella | Friday, February 13, 2009 at 05:13 AM
Bill, to my ear your (e) though (g) do not mean the same thing that (c) does. To my ear, (c) means "it is possible that I not exists" while (e)-(g) mean "it is not possible that I exist". I would propose instead: (e2) 'I could have not existed now' and (g2) 'I can have not existed now'. (although (g2) is arguably ungrammatical due to misuse of verb tense).
Posted by: Dave Gudeman | Sunday, February 15, 2009 at 01:06 PM