Robert Larmer has raised some interesting points in his response (149–159 above), but I think he is missing the main thrust of my argument. Perhaps that is my fault, for not emphasizing it more clearly. As he notes, I do grant that his account of miracles does work just fine – given that one rejects the physicalist causal closure principle (the idea that all systems are ultimately physically isolated.) But my paper is clearly oriented towards saving some sort of physicalist intuitions (and then seeing if, regardless, one can still have miracles in any interesting sense.) If read with that in mind, most of Larmer's criticisms dissolve.
So, given the luxury of a response to Larmer's commentary, I will focus on three main issues in this reply: a little more discussion of the motivation and approach I intended to take in the paper, a brief discussion of the ‘repository-transportation’ miracle, and then some more specific responses to Larmer's triple-pronged criticisms of each type of miracle.
Law violations and physicalism
My main discomfort with the ‘open-systems’ and ‘exemption’ approaches (indeed, with any approach that makes the claim that natural laws are inapplicable because, in essence, the relevant activity isn't natural) is that they commit, unwittingly, to an extremely strong metaphysical claim. What claim is that? That the natural is never sufficient to determine the natural. (The physicalist side is as extreme: the natural is always sufficient to determine the natural.) In the interests of conciseness, I will just say that Larmer's quotation from my e-mail, expressing my doubts about causal disposition theory, encapsulate my complaints quite nicely. The question I was attempting to address was, ‘Is it possible to affirm the determinative sufficiency of the natural in substantive respects, while still preserving the possibility of divine intervention?’.
Obviously, my particular closure principle (the CPP) doesn't count as ‘real’ physicalism. I do think it captures some substantial physicalist intuitions, which is why I thought it appropriate to borrow the label. My criticisms of his miracle-mechanism are all, I agree, quite incoherent or ineffective if one resists any metaphysical sufficiency conditions on laws.
The ‘repository’ mischaracterization
Larmer addresses a mischaracterization of his miracle mechanism I commit in the paper; the mistake is there, yet irrelevant. Let me explain: As I assert on 127–129 of the section ‘Physicalism and causal closure’, CPP by itself does not forbid a Larmerian ‘open-systems’ approach – either Larmer's ‘true-creation’ miracle, or my ‘repository-transfer’ variant. This is because CPP by itself does not ensure closure in the relevant sense: dynamical laws do not govern creation or annihilation events, they only require that dynamical effects must not be inconsistent with their physical causes. Noether's theorem with the CPP, however, requires that creation and annihilation effects be consistent with the physical causes that govern them – including causes strictly (via CPP) regulated by the conservation laws. Whether the effects are brought about by supernatural creation, or by supernatural transportation from a repository, is irrelevant: assuming CPP gives primacy (i.e. determinative sufficiency) to the operating conservation or dynamical law, forbidding such events in either variant.
Criticisms of the mechanisms
In his response, Larmer also raises many doubts about my proposed miracle-mechanisms: three for each mechanism. I shall briefly treat each in turn.
Determinism 1: Miracles as interventions
I agree that the deterministic mechanism strongly favours deism. (This conclusion, however, is contingent upon many other assumptions one makes about the temporality of God's epistemology and subsequent action, and our own liberty.) I take this to be an important conclusion, implied but not expressed in my paper: miraculous activity of a theistic kind is incompatible with these physicalist intuitions, in a deterministic universe.
Determinism 2: Impossibility of ‘invisible’ atypicality
A few examples (as I tried to allude to in the paper) can indeed demonstrate that ‘the degree to which the microstates of a system would have to be atypical in order to produce a miracle is consistent with the system appearing to be typical macroscopically’ (Larmer reply, 155) [my emphasis]. Real-life physical systems (under human control!) demonstrate this, including the spin-echo effect (Hahn (Reference Hahn1953), especially 8, and Sklar (Reference Sklar1993), 219–221). I fail to see the disanalogy in his note 15; playing cards have causal histories, too – especially stacked decks of them. The key here is his recognition that the Red Sea's parting is infinitesimally unlikely. Exactly. That non-zero measure of likelihood is a representation of the possible existence of a microstate that would appear (in future) macroscopically miraculous – no ‘major intervening causal factor’ required.
As for the chaotic mechanism, his comments are well taken – and I said such in the paper already (136).
Determinism 3: Incompatibility with libertarianism
Larmer is correct in criticizing this mechanism's incompatibility with libertarian free will. The claim of superiority, with regards to this mechanism, is in reference only to its ability to reconcile important physicalist intuitions with miracles. I included it for those who think that the evidence supports determinism, those who couldn't reject determinism as a possibility, and those who are compatibilists. Under other criteria, it obviously won't fare so well.
Indeterminism 1: Ad Copenhagen ignorantiam
The interpretation of quantum theory is a huge issue in itself, but this is a red herring: I'm not sketching out a definitive account of divine quantum interactionism here, just establishing its possible reconciliation with important physicalist intuitions. As with propensity-type probabilities, if instrumental interpretations like Copenhagen aren't compatible with this programme, then my reaction is simple: Copenhagen's got to go. (If such an interpretation becomes consensus, so much the worse for my programme.) There are many, many other interpretations to select from – see Albert (Reference Albert1994), chapters 5–7 for discussions of three other classes of interpretations more suited to my task.
Indeterminism 2: Impossibility of amplification
I'm not sure how ‘far from clear’ it is that these micro-interventions could affect macroscopic events, but it is clearly possible – and (again) that's all I'm concerned to prove. Not only is it possible, it is actual (without chaos!): besides the fictional ‘Schrodinger's cat’, other examples of macroscopic quantum effects (suitable for divine exploitation) are Bose-Einstein condensates, lasers, SQUIDs (superconducting quantum interference devices) and the millions of field-effect transistors in your nearby computer. Quantum effects are far from impotent in the macro-world.
Indeterminism 3: Interventions as changes
This complaint rests on a definition of miracle that presupposes determinism to some substantial degree: if we are to make room for miracles in an indeterministic world, then we mustn't shackle ourselves with a definition that requires a contrast with ‘what would otherwise occur naturally’ (Larmer, 157). Note that ‘what would, in the future, occur’ depends on the assumption that there exists a definite prescription of future events – something indeterminism expressly rejects. We must be careful not to define indeterministic miracles out of existence by assuming determinism unreflectively.
Conclusion
While Robert Larmer's objections to my paper often raise interesting and important questions, the main thrust of my paper remains: it does seem quite possible that miraculous activities are possible, for a ‘moderate’ physicalism. Given the costs associated with this view, some may reject it; but I hope their rejection became more informed by my efforts.
