Reproduction of this essay in whole or in part for noncommercial purposes, with attribution and with notice of copyright, is permitted; reproduction of this essay in whole or in part for commercial purposes, or without attribution, or without notice of copyright, without the author's express permission, is not permitted.

On Fatalism

When we think about past events, we normally think of them as "written in metaphysical stone." Nothing we could do could make an event E that has already happened at time t fail to happen at time t, and nothing we could do could make an event E that has already failed to happen at time t happen at time t. I express this by saying that E has an occurrence-value O(E). Even if one countenances time travel to the past, it seems as though one cannot make an event E which has already happened unhappen, or make an event E which has already failed to occur happen; all one can do is to create a second time-stream in which E's occurrence-value is different than it was in the first time-stream. E's occurrence or non-occurrence at t--O(E)--is fixed for any particular time-stream. And most of us do not countenance time travel to the past, and therefore take O(E) to be fixed, for any past event E.

By contrast, we tend to think of the future as open. A future event E might happen or might not happen in this time-stream; its occurrence-value O(E) is not fixed, even in this time-stream. We take there to be a fundamental disanalogy, in this regard, between past events and future events. The fatalist argument attempts to undercut this disanalogy. Reiterating that "O(E)" means "E has an occurrence-value," and taking p to be the proposition "E occurs at time t," I give a simple version here:

1. p v ~p
2. p-->O(E)
3. ~p-->O(E)
4. O(E)
 (Premiss: Law of the Excluded Middle)
(Premiss: If it is true that E occurs at time t, then E has an occurrence-value)
(Premiss: If it is true that E fails to occur at time t, then E has an occurrence-value)
(1, 2, 3, Hypothetical Syllogism)

Note that this argument seems to hold both for past events and for future events. Also note that it says nothing about human freedom. If p is now true, it may very well be via human choice that E occurs, and that the present truth of p is attributable to a freely made choice; and similarly if p is now false. The argument ascribes to E a present occurrence-value, but it does not say why E has that present occurrence-value. Even if one fully accepts the fatalist argument, therefore, one cannot conclude that there is no free will. If E's occurrence is inevitable, it may be so only because of a person's free choice; were he to choose otherwise, then E's nonoccurrence would be inevitable instead. The reason for E's (or not-E's) inevitability is not specified by the fatalist argument.

However....

The argument rests on the Law of the Excluded Middle, so one may deny that the Law of the Excluded Middle applies to future events; this seems to be Steven Cahn's suggestion in his book Fate, Logic, and Time. One might claim that for future events, neither p nor ~p is true. And this seems reasonable to me. I have written premisses (2) and (3) as conditionals, but the biconditionals p<-->O(E) and ~p<-->O(E) really hold, if one holds any sort of correspondence theory of truth. To say that the proposition "E occurs at time t" is true is just to say that E occurs at time t, and to say that E occurs at time t is to say that the proposition "E occurs at time t" is true. Thus, to say p v ~p is to say that either E occurs at time t or else E does not occur at time t, and to take p v ~p as saying that either "E will occur at time t" is now true or else "E will not occur at time t" is now true is to take it as saying that either E's occurrence at time t is now the case or else E's nonoccurrence at time t is now the case. But this, it can reasonably be claimed, is to assume the truth of the fatalistic doctrine itself, for what the non-fatalist wants to say is that neither E's occurrence at future time t nor E's nonoccurrence at future time t is "written in metaphysical stone"--that future event E's occurrence or nonoccurrence won't be decided, not only epistemically but also metaphysically, until future time t, and therefore that it is not now true that E will happen and it is also not now true that E will not happen.

The non-fatalist would want to read the argument, for future events, as

1. p will be true or p will be false
2. If p will be true, then E will have occurrence-value.
3. If p will be false, then E will have occurrence-value.
4. Therefore, E will have occurrence-value.

Such a reading, taking p to have a future truth-value but not a present truth-value, doesn't imply that E already has an occurrence-value; only that it will have one. Alternatively, the non-fatalist could construct an anti-fatalist argument, for future events:

1. p is not now true and p is not now false.
2. E now has occurrence-value if and only if either p is now true or p is now false.
3. It is not the case that either p is now true or p is now false.
4. Therefore, it is not the case that E now has occurrence-value.
 (Premiss: Anti-fatalism)
(Premiss: Truth-content correlation)
(1, de Morgan's Law)
(2, 3, Modus tollens)

Of course, the anti-fatalist argument assumes its conclusion just as much as the fatalist argument assumes its conclusion. Where the fatalist argument assumes that propositions about the occurrence of future events already have truth values and therefore, via the correspondence of a proposition's truth and the content expressed by the proposition's being the case, assumes that future events have occurrence-values, the anti-fatalist argument assumes that propositions about the occurrence of future events now lack truth-values and therefore, again via the correspondence between truth and content, assumes that future events lack occurrence-values. Both the fatalist and the anti-fatalist argument really fail to be arguments at all, but are simply choices of how to view the metaphysical nature of future events. And that may very well depend on one’s view of time. One who models spacetime according to Einstein’s “block universe,” for example, might very well make the fatalist choice. But it looks as though the difference between the fatalist and the anti-fatalist is simply one of differing intuitions.


(© 2008 by Keith Brian Johnson)

Philosophy Home
Site Home