Friday, August 17, 2007

Why There Can't Be Infallible Foreknowledge of Libertarian Free Choices

I say that God cannot have infallible foreknowledge of creaturely libertarian free choices. By 'infallible' I mean the impossibility of error. By 'foreknowledge' I mean knowledge of an event that is temporally prior to that event. Thus, "S foreknows that event E will occur" means that at some time t S knows that E will occur at some time subsequent to t. By 'knowledge' here I shall mean merely non-accidentally true belief. By 'true belief' I mean one the propositional content of which corresponds to how things actually are. By 'libertarian free choices' I mean choices in which the agent has an unconditional power to choose otherwise immediately prior to the time of the choice.

Now, for knowledge to obtain requires that there be a proper sort of relation between knower and known. Since knowledge entails truth, and truth consists in correspondence with reality, knowledge requires that there be a correspondence between mind and world, between what is believed to be the case and what is the case. Furthermore, since knowledge entails that true belief be non-accidental, there must be something that grounds or secures the mind-world correspondence. The possibilities for this seem to be limited to the following:
  1. World determines mind (e.g., God knows what will happen because it does happen).
  2. Mind determines world (e.g., God knows what will happen because he decrees that it happen and ensures that that decree is fulfilled).
  3. There is a probabilistic, non-determining relation between mind and world.
  4. There is a brute, ungrounded correlation between mind and world.
If (1) is the case, then God can have infallible knowledge of creaturely libertarian free choices, but it cannot, strictly speaking, be "fore"-knowledge. Rather, since God's knowledge on this scenario is grounded in the actual occurrences of the events themselves it is more aptly described as "post"-knowledge, for the events must already have happened before God can know about them. Proponents of divine timelessness will resist that way of putting it, but they too deny that God literally has "fore"-knowledge.

If (2) is the case, then God can have infallible knowledge of creaturely choices, but these choices cannot be free in a libertarian sense. If God knows that a creaturely choice will be made because he determines that it is made, then the creature cannot have unconditional power to choose otherwise. Instead, the creature will only have a conditional power of contrary choice - it is only IF God had decreed otherwise, that one could have chosen otherwise.

If (3) is the case, then God can have "knowledge" in the sense of non-accidentally true beliefs about what creaturely libertarian free choices will be made, but this knowledge cannot be infallible. Infallible knowledge requires not merely non-accidental truth, but also the impossibility of falsity.

Finally, if (4) is the case, then God can have true beliefs about what creatures with libertarian freedom will do, but not knowledge. The reason is that, on this option, the non-accidentality criterion for knowledge is not satisfied. That God's beliefs about what will happen sync up with what does happen turns out to be sheer luck.

So something has to give. We have to reject either (a) creaturely libertarian freedom, or (b) exhaustive, infallible divine foreknowledge.

Sunday, August 05, 2007

Eliminativism and Reductionism

One of Ian's comments on an earlier post of mine got me thinking about the distinction between theoretical elimination and theoretical reduction. This is a familiar distinction in the philosophy of mind, but it comes up in lots of different contexts.

In general, eliminativism (E) and reductionism (R) are reactions to a thesis (T) to the effect that "Things of kind X exist", where X is a description of the kind in question. Both E and R are denials of T, but they differ in an important respect. The eliminativist not only denies the existence of X-type things, but also rejects the propriety of talking as if X-type things existed. The reductionist, on the other hand, continues to affirm the propriety of talking as if the thesis were true while denying that there are any X-type things in the sense countenanced by the thesis. She accomplishes this by identifying a non-T substitute as the "real" referent of X. Here are some examples:
  • Thesis: God (understood in a broadly classical sense) exists.
  • Eliminativist: Such a being does not exist, nor should we say "God exists" because it cannot be affirmed without perpetuating falsehood and confusion.
  • Reductionist: God as understood in the broadly classical sense does not exist, but "God exists" can properly be said because the descriptor "God" really refers to something else (e.g., the Cosmos, energy, the idea of a broadly classical God, etc.).
  • Thesis: Cartesian souls exist - mental terms (e.g., "beliefs", "desires", etc.) refer to states that are strictly nonphysical.
  • Eliminativist: There are no Cartesian souls or nonphysical states. We should not continue to talk as if there were mental states.
  • Reductionist: There are no Cartesian souls, but it is okay to continue talking of mental states - we just have to realize that mental states are really just physical states.
The observation I want to make is that the eliminativist and reductionism agree on matters of substance. They concur that the thesis is false. Where they disagree is on matters of language. They disagree about the ethics of terminology, the propriety of speaking in certain ways. The eliminativist prefers to use language in much the same sense as the proponent of the thesis does. Hence, along with his denial of the thesis the eliminativist rejects the language used to formulate the thesis as false and misleading. The reductionist, on the other hand, doesn't want to let proponents of the thesis define how the key terms are to be used, and so she substitutes her own definitions. This allows her to preserve the verbal formula used in the affirmation of the thesis while gutting it of what she sees as objectionable commitments.

The debate between eliminativists and reductionists would seem to turn on prevailing or established usage. For example, with respect to the meaning of "God", the reductionist substitutes have little or no claim to capture the force of that word as it has actually been used in the Western theological tradition. This is just atheism in denial, not a redefinition of "theism" that is still worthy of the name. In contrast, the theoretical reduction of "heat" from traditional understandings of it as a manifestation of an element (fire) or a type of fluid (caloric fluid) to "average kinetic energy" marked a useful theoretical advance. The justification for continuing to use the word "heat" while changing its meaning lies in the broad commonality of the observational data that the respective theories were invoked to explain.

To some extent, whether a position should be called 'eliminativist' or 'reductionist' may be a matter of perspective. For example, in my ongoing dialogue with Ian, I affirm as an A-theorist that there are tensed facts. Ian, a B-theorist, also affirms the existence of tensed facts, but he defines them in accordance with his metaphysics. So he claims to be a reductionist and not an eliminativist with respect to tensed facts. From my perspective, however, Ian's "tensed facts" are trivial and uninteresting and the real metaphysical issue is whether there are tensed facts in the A-theorist's sense. So what he calls reductionism, I call eliminativism in denial.