Wednesday, August 13, 2008

Molinist evolutionary theory

Molinist evolutionary theory (MET) holds that evolutionary theory is correct and based on genuinely random processes. Nonetheless, according to MET, these processes are guided by God. For each random transition (e.g., a random mutation, recombination or selection event) has associated with it a subjunctive conditional of the form "if circumstances C were to occur, then transition T would occur". God non-trivially knows the truth values of all such conditionals, and created the world so as to ensure a sequence of circumstances C that, given the conditionals he knew, would result in a sequence of transitions that fits with his plan.

I have argued elsewhere (a version of this has appeared in Philosophia Christi) that this story undercuts the statistical explanations that evolution needs. Here I want to point out a second issue. We know the probabilities of outcomes of processes in nature essentially by looking at frequencies of outcomes[note 1]. But, almost surely[note 2], a Molinist God can get any sequence of outcomes he wants by tweaking the circumstances appropriately. If a coin is to be flipped a million times, a Molinist God can make them all come out heads not by intervening in the flips, but by ensuring that the conditions C in which the flips happen are such as to make true appropriate conditionals of the form "C→heads".

Given the existence of a Molinist God, one might expect, or so Mike Almeida has argued, observed frequencies that do not match the probabilities involved in the processes. In fact, this might even give rise to an interesting prediction: given a Molinist God, we might expect the more needy to be disproportionately represented among lottery winners, since it seems not unlikely that God would want to choose initial conditions to favor them. If this line of reasoning is right, then given the existence of a Molinist God, the frequencies we observe should not reflect the probabilities of the underlying physical processes. But if so, then our knowledge of the probabilities of the underlying physical processes is undercut. And this is surely a problem for MET, not because it falsifies evolutionary theory, but because it undercuts it epistemically, making it impossible for us to know the probabilistic claims on which evolutionary theory is based.

Suppose, on the other hand, our Molinist rejects the Almeida argument, and holds that even given a Molinist God, the observed frequencies will match the probabilities of the underlying physical processes, perhaps because God would want them to match in order to be a self-effacing creator, or to let us engage correctly in inductive reasoning. In that case, the following is still true. The observed frequencies are not directly evidence for the probabilities of the underlying physical processes. They are only indirectly evidence given some assumptions about how one expects God to act.

Here is another way to put this. On the Molinist view, there is a defeater to our knowledge of probabilities on the basis of frequencies: the frequencies come from God's decision as to the antecedents of conditionals. A controversial thesis about how God chooses to act, if substantiated, would provide a defeater for this defeater. This makes knowledge of probabilities of physical processes rather more roundabout than we think it is. Moreover, I am not clear whether on this view an atheist can know any claims about these probabilities, since God's contingent decision to make the frequencies match the probabilities seems to play a central role.

12 comments:

  1. The observed frequencies are not directly evidence for the probabilities of the underlying physical processes. They are only indirectly evidence given some assumptions about how one expects God to act.

    They are not evidence at all, are they? We really have no clue as the right propensities given that God is stacking the probabilistic deck. To know what the genuine propensities of what we call "fair coins" are we'd have to know whether this counterfactual is true,

    P. Had God not chosen the circumstances and coins to be instantiated, fair coins would have landed heads 1/2 the time anyway.

    Under Molinism, we have absolutely no clue whether that is true (setting aside it's being trivially true). So we don't have even indirect evidence for the genuine propensities of objects. I think this is really important and a major problem for Molnism when combined with rational choice.

    ReplyDelete
  2. There's a simpler way to put this. What we take to be, say, the half life of radon is based, as you say, on the observed frequency of decay. But if Molinism is true, then we could have no idea about the genuine half-life of radon. We have no idea since we have no reason to believe (one way or the other) that God is instantiating radon atoms in ways that reflect its genuine half-life. There's just no way to know that. And of course the problem generalizes to all undetermined events.

    ReplyDelete
  3. Mike,

    What would you say to the Molinist who would say that your 'genuine propensities' are just what is true over all possible orders of nature God could have chosen, and that therefore there is no difference between trying to find the genuine half-life of radon given that God selected one particular series of events and trying to find the genuine half-life of radon given that one particular series of events happened by chance -- i.e., that one might equally say that there's just no way to know the possible world corresponding to the actual world just happens to be one where the apparent half-life of radon diverges sharply from the genuine half-life of radon?

    ReplyDelete
  4. Hi Brandon,

    I think I'm not sure what you mean by saying...'genuine propensities' are just what is true over all possible orders of nature God could have chosen. On Molinist accounts, it is perfectly possible for God to make the propensity of coin C to land heads appear to be .5 or .9 or whatever, by choosing the circumtances in which C is tossed. Similarly for decay rates for radon, etc. So I think I'm misunderstanding you.

    On the other hand, Molinists do not deny that there is genuine chance in the world--among the chancy events we might even include free action (though not all would agree with that, I suspect). They simply add that God can manipulate the frequency of chancy events in accordance with his providence.

    ReplyDelete
  5. Perhaps I should have stated it differently. As you say, on Molinist accounts it is possible for God to make the propensity of coin C to land heads appear to be just about anything by choosing the circumstances in which C is tossed. But on any account of chance, it is possible, simply by chance, for the propensity of coin C to appear to be just about anything given the circumstances in which C is tossed -- a perfectly fair coin could have a run of five hundred heads in a row. This is extraordinarily improbable, but there are possible worlds where it happens. Thus it looks like the real propensity of anything is something that could only be known in terms of all possibilities. So I was wondering what the response would be to a Molinist who thinks the two cases analogous, just differing by precise details in causal explanation; and that the argument isn't built on any particularly Molinist principles. I'm primarily just trying to clarify what it is about Molinism in particular that is supposed to underlie the problem.

    ReplyDelete
  6. This is extraordinarily improbable, but there are possible worlds where it happens. Thus it looks like the real propensity of anything is something that could only be known in terms of all possibilities.

    Right, there are worlds in which this happens. But we can be very close to sure that, were such a sequence to happen, the propensity to land heads is not .5. We cannot be sure of that on the assumption of Molinism, since the probability of a multi-heads-landing world is not known to be improbable, no matter the propensity of the coin, given the Molinist God.

    ReplyDelete
  7. We cannot be sure of that on the assumption of Molinism, since the probability of a multi-heads-landing world is not known to be improbable, no matter the propensity of the coin, given the Molinist God.

    I guess I'm still not really sure why this would not also be true of any other case. As far as Molinism itself goes, the Molinist God serves the function, so to speak, of 'collapsing' the whole collection of orders of natures, or possible worlds (depending on the kind of Molinism), to an actual world. Thus your argument would seem to indict any non-determinist view, since in every such case the possible worlds still need to be 'collapsed' to an actual world -- something needs actually to happen -- and it seems one could then say that the probability is not known given whatever it is that makes things become actual. That is, it seems to be more of an argument that there can be no cause responsible for this world rather than some other world being the actual world, since any such cause would appear to serve a function analogous to that of God in Molinism, and would similarly highjack the probabilities.

    ReplyDelete
  8. I don't know what it means to collapse all worlds into the actual world. This suggests to me that there is really only one world. But nothing I said--I hope--entailed anything like that.

    In any event, at least one important difference between worlds in which a Molinist God exists and one in which there is no God at all is that the Molinist God is manipulating chance events. In non-God worlds, chance events are not being manipulated. In non-God worlds, for instance, nothing pulls the frequency of atomic decay away from the propensity for decay. I can learn directly from the observed frequencies what the propensities are. It is certainly true, however, that we might be wrong in what we believe are the decay rates. There is some minute present chance, as Lewis observes, that far more tritium atoms will exist in the future than have existed before and each will decay in only a few minutes. If this unlikely future were realized, it would complete a chancemaking pattern on which the half-life of tritium would be very much less than the actual 12 1/4 years. Could it happen, given the present chances? It could, in the sense that there's nonzero present chance of it. It couldn't, in the sense that its coming to pass contradicts the truth about present chances. That is, it is inconsistent with our probabilistic laws.

    But in Molinist worlds, we know that God is manipulating chance events for divine purposes, to realize divien goals, so we are simply not in an epistemic position to specify the probabilistic laws.

    ReplyDelete
  9. Mike, I know little of Molinism, but I'd like to add something to your comments on propensities, if I may... Observed frequencies do not allow you to learn directly about the underlying propensities. How unlikely it is, that we have observed deceptive frequencies, is something that we can only estimate once we have assumed some value for the underlying propensities, and those will not be (by hypothesis) as we think they are. The problem is that the range of such possibilities is inestimably vast. We would have to assign possible values to everything, e.g. we have observed that the speed of light is constant, but maybe that was just an unlikely fluke.

    Re your example of tritium atoms, there is a small chance of such an outcome given that our current estimate of their propensities is correct. If it happens, we would change our estimate, but we would be wrong to do so (by hypothesis). There is another small chance that we have observed deceptive frequencies, caused by propensities like the frequencies that will be observed. In that case, we will be correcting our estimate when we change it (but only partially unless we ignore the previous results). Most likely we would look to the more general physical laws, and see if the propensities are likely to have changed. If that seems unlikely then we might decide instead that our previous results were flawed by bad experimental techniques, and hypothesise about such flaws...

    ...more complicated than Lewis's picture, but not as bad as engineering, I think. In short, it strikes me that we could accomodate a Molinist God acting in a world of propensities much as we could accomodate any God acting sometimes in any world of natural laws. What is the difference between God setting up simple natural laws, binding on his creatures, and then violating them himself occasionally, and a God who sets up complex natural laws, and always acts through them?

    ReplyDelete
  10. Re your example of tritium atoms, there is a small chance of such an outcome given that our current estimate of their propensities is correct.

    I think your caution is healthy, Enigman. But I simply reject the skepticism that (I think) forms the basis for this objection. I reject the idea that we do not know the decay rate for tritium. We definitely know that. There is all sorts of evidence for this (much of it, if I'm recalling correctly, is astronomical--peering back to the big bang). But yes it could be all wrong and yes there might be tritium demons that love deceving us wrt the decay rates of these atoms. But I'm just going to resist walking down that road.

    ReplyDelete
  11. Since for various reasons I think it's logically incoherent for more than one possible world to correspond to the actual world, given what possible worlds are supposed to be, any suggestion of collapse to one would be on my part. But I don't think anything rides on it.

    In Molinism, as far as I can see, God is not really manipulating chance events, unless you also think someone picking a number from a table of random decimal numbers is manipulating chance events. Of course, people don't pick numbers from a table of random numbers unless they have reasons for doing so; and it's even possible that their reasons for doing so may constrain the number from the random number table that they choose -- for instance, any number may be acceptable as long as it has a two in the fortieth decimal place. But the numbers are still random except for the points of constraint, and no manipulation is going on. For the Molinist the parallel will be very close; where chance events are involved God will be choosing from possible world-lines exhibiting random variation, like the random numbers. So, for instance, the half-life of samples of radon will vary across world-lines in such a way as not to be constrained by prior events. And one of these world-lines will be chosen; we don't know all the constraints. Now, it is true that this means our certainty that our actual world-line may be an outlier for radon half-life. But even if it is so, in the mere fact of selecting the world-line God has not interfered with the world-line in any way, nor has He introduced any new feature to it, any more than the person picking the random number has interfered with the number. The world is already on the table of possible worlds (so to speak); so even if God were not selecting the world and it just happened to be selected in some other way by chance, it's a world that might come up anyway (just with low probability).

    ReplyDelete
  12. Mike, were the decay rate not precisely what we think it is, but in some finite range of slightly different values, we could have got the evidence we have for the rate that we think it is (and would probably be just about to find out that we were wrong). It is hard to resist that, given the theory we have on such evidence as we have.

    Even more unlikely, but still possible, is that our estimate is way off. And then, having made that first step, there are lots of possibilities, as we increase the parts of our theory that we question, including such demons. It is also hard to draw a principled line before such possibilities, having made the first step.

    I would agree that we know the decay rate, much as we know that we have hands etc. But our justification is not infallible, and so the problem as I see it is the huge range of the possible variables. We must be assuming some things without justification, and we should be careful deducing anything from those assumptions beyond their usual application.

    ReplyDelete