To: Alan L.
From: Geoffrey Klempner
Subject: On possibility and possible worlds
Date: 3rd June 2011 12:20

Dear Alan,

Thank you for your email of 26 May, with your first essay for the Metaphysics program, entitled 'Possibility and Possible Worlds'.

I liked the way you went about this essay. Talking about what might have been the case is something we do quite naturally, but the question is, exactly what are we doing when we talk this way? What do we mean?

You offer the example of the number of crows on your lawn. There were four crows on your lawn this morning but there might have been three, or five, or some other number.

First of all, we need to distinguish this from the mental act of imagining that there are a particular number of crows on your lawn. We can imagine things which aren't possible, and which we admit aren't possible. For example, I imagine bringing it about through an act of sheer will, that I pass the exam when the letter telling me that I have failed the exam is in front of me. As I concentrate, the words in the letter change before my eyes. And all the other things needed to make me 'have passed' the exam change too.

With these powers, I don't need to pass exams. I can beat Superman any day. As Superman is about to slug me, I imagine him weak and puny. Or I imagine that Superman never existed, or grew up on some other planet in another solar system.

Very nice, but it isn't possible. Not only not possible given what we know about the laws of nature, but logically impossible. Exercising these powers would entail the truth of a logical contradiction. E.g. 'The candidate GK actually scored 21 per cent and the candidate GK actually scored 77 per cent.'

I might seem to be labouring this a bit, but the point is that statements about 'things that might have been' are true, or false. And they are not statements about the content of some person's mental state. If true, they are objective facts. But facts about what?

You distinguish, correctly, between possibility as 'what might have been' and possibility as what you term 'expectation', things that might yet come to be, or fail to come to be. The possibility of expectation is actually one species of a more general phenomenon, epistemic possibility. There are many things consistent with what I know. Some of these things refer to the future. But others refer to the past. Homer might have been a single author. Or 'he' might have been several authors whose works were collected together over time. We don't know. Either is 'possible'.

However, you also introduce a third idea: the idea that 'a possibility is an event, or state of affairs, which we include in our implicit view of the world and how the world works.' To see that this really is a third idea, not simply a variety of epistemic possibility, consider the things that it is possible *for* you to do. You can write a good metaphysics essay. You can't run a mile in under three and a half minutes. Powers and capacities, whether of people, or things, are in a very real sense 'in' the entity in question.

Perhaps we could stretch this to cover the case of the crows on your lawn. There might have been a thousand crows on your lawn, if your lawn was big enough. Otherwise not. So we are talking about a power or capacity of your lawn, and also, perhaps, a power or capacity of crows. What things can, or are able to do or not do. Once again, this seems to be about the actual world.

Maybe, with sufficient ingenuity, we could extend this further to any statement purporting to be about metaphysical possibility. There might have been only seven planets. I might have had two heads.

You also offer an account of how we 'construct' possible worlds, by taking a true description and changing the truth value of one or more of the statements in the description. Here, by contrast, there is no need to explain possibility in terms of powers or capacities because we are not aiming for explanation. Our only concern is logical consistency.

But still the question remains what makes statements about metaphysical possibility true, when they are true. Perhaps we can get away with saying that 'A is metaphysically possible' is equivalent to 'The description "A" is logically consistent.' Problem is, we also make conditional statements of the form, 'If A had been the case then B would have been the case.' Here the truth seems much more 'meaty'. David Lewis, one of the foremost proponents of strong realism about possible words, argues in his book 'Counterfactuals' that possible worlds are *needed* in order to account for the truth conditions of counterfactual statements. So you would either have to deny that counterfactuals have truth conditions or accept the existence of possible worlds.

You offer what looks like a knock-down refutation of Lewis's view. 'The actual world exists as the world of our experience, while possible worlds exist as abstract concepts of our own creation.' But Lewis has an answer to this. Who are 'we'? A possible world is actual if, and only if, it is a world which we inhabit. But other possible worlds are inhabited too. The inhabitants of such a world call the 'actual' world, the world of their experience. From their point of view, our world is 'an abstract concept' of their own creation.

All the best,

Geoffrey