|Thus consider the law of gravitation. This says that, in any isolated physical system made up of bodies B1, . . ., Bn, each body Bk will be subject to the vector sum of the forces due to the other Bjs (j ≠ k) (namely, Gmkmj/r2jk— where mj is the mass of the other body Bj, rjk is the distance between Bj and Bk, and G the constant of universal gravitation). Now, this law qualifies as a microphysical law because it tells us what would happen in a very small localized system comprising a few tiny particles. But at the same time it is formulated in an entirely general way. So it also tells us what would happen to a large falling body near the surface of the earth, say. We don’t need any new principle to tell us what will happen to such a body. We simply apply the same gravitational law that applies to very small systems to the more complex set-up comprising the falling body and the earth.
Now, there seems no principled reason why all basic laws should be microphysical in this sense. Thus consider ‘emergent laws’ of the kind C.D. Broad (and other ‘British Emergentists’) envisaged. These are laws that (a) apply to specific large-scale physical initial conditions (b) don’t follow from microphysical laws and (c) are essential to the appearance of certain physical effects. For example, imagine that, when the molecules constituting animal cells are in the physical context characteristic of a developing embryo, they start behaving in ways that aren’t predictable given only the microphysical laws. Or, again, suppose that the molecules comprising neurotransmitters behave in a similarly unpredictable way when they are in the physical environment of a functioning brain.6
Emergent Broad-laws would thus violate (L). They would give us a kind of macroscopic law that is not metaphysically determined by microphysical laws and initial conditions. There could be two possible worlds that agreed in their microphysical laws and microphysical initial conditions yet differed in their large-scale emergent laws—for example, one might have a law about special molecular movements to be found in developing embryos, while another might lack any such law.
The question now is whether this kind of emergence would threaten Physicalism. Would Broad-style emergence transcend the physical realm and call into being something non-physical? Or would it merely be a violation of Physical Microscopism that transcended the microphysical but leaves Physicalism intact?
At first pass, there is no obvious reason why Broad-laws should be viewed as requiring anything non-physical. Broad-laws would mean that certain large-scale complexes enter into laws that don’t follow from basic microphysical laws and which make a real difference to the evolution of physical systems. But there would seem no immediate reason not to count both these large-scale complexes and the laws they enter into as physical. After all, nothing said so far requires these complexes to be anything more than large-scale arrangements of small physical parts. And nothing said so far requires the emergent laws to do anything except relate these physical initial complexes to physical results. (True, if ‘physical’ by definition required governance by microphysical laws, as in Pettit’s definition of ‘physical’, then the physical complexes entering into emergent laws would come out as ‘non-physical’. But they won’t if we adopt either the ‘resemblance’ or ‘inorganically identifiable’ conceptions of ‘physical’, as seems more natural in this context.)
What about the argumentative rationale for Physicalism? Would this survive the existence of emergent Broad-laws? Again, there seems no immediate reason why Broad-laws should stop us arguing for Physicalism. Maybe they would if the only argument for Physicalism somehow proceeded via a demonstration that all physical laws supervene on microphysical ones. However the causal argument for Physicalism sketched above makes no such assumption. Rather it hinges on the causal completeness of the physical realm, which says nothing about microphysics, but only that every physical effect has a fully sufficient physical cause. Broad-laws seem in perfectly good accord with this assumption. True, such laws would mean that some physical effects essentially result from macroscopic physical causes in ways unpredictable on the basis of microphysical laws alone. But for all that, they are still physical effects with sufficient (macro)physical causes. And so the causal argument will still tell us that any mental causes of those physical effects cannot be metaphysically distinct from those (macro)physical causes.
7 Special Fields
Despite the points made in the last section, there are further considerations that complicate the question of whether emergent Broad- laws are consistent with Physicalism.
Modern relativistic physics implies that causal influences exerted over spacetemporal distances must be mediated by the propagation of force fields. Relativity theory precludes any causal influences travelling faster than the speed of light. So there will be temporal gaps between any separated causes and effects. In typical cases this temporal interval will mean a violation of the conservation of energy. The standard solution is to suppose that the causes work locally, not at a distance, by propagating force fields which in turn produce the distant effects. These fields can then be viewed as embodying the relevant energy during the temporal delay between distal causes and effects. (Lange 2002 ch 5.)
This argues that any Broad-laws would be associated with the emergence of special fields generated by the specific macroscopic initial conditions appearing in those laws. It is not to be taken for granted that these fields will count as ‘physical’, even if the macroscopic initial conditions that generate them do. To the extent that they would, Physicalism will remain intact, and the special fields would at worst violate the within-physics supervenience required by Physical Microscopism. But if the extra fields were non-physical, then they would automatically invalidate Physicalism.
To see more clearly what is at issue here, return to the suggestion that organic molecules behave in a distinctive manner in a developing embryo, or that neurotransmitters do the same when in a functioning brain. These behaviours would give us reason to posit ‘vital’ and ‘mental’ force fields respectively. And these fields would be genuinely extra to basic physical force fields like gravitation and electromagnetism, given that Broad-style laws give rise to physical effects that cannot be accounted for by more basic force fields.
The question is now whether fields like these would count as ‘physical’ or not. This turns out to be a rather messy question. I earlier considered three ways of defining ‘physical’: (a) metaphysically supervenient on the microphysical (b) inorganically identifiable and (c) resembling currently recognized physical categories. At a first approximation, the last of these make special force fields come out as physical, the second argues that at least some are non-physical, while the first delivers no clear verdict.
Let me briefly run through these options. (a) ‘Physical’ = ‘supervenient on the microphysical’. At first sight it might seem as if special force fields won’t be ‘physical’ on this definition, because they aren’t supervenient on the aggregates of microphysical facts that generate them: after all, there are worlds containing those facts that lack the relevant Broad-laws and so the fields. But doesn’t necessarily decide the issue, for special force fields will still standardly supervene on the local values of the fields themselves: fix the field values at all spacetime points and you fix all the field facts. So special force fields will be microscopically determined. But does this mean they are microphysically determined? It depends on whether local values of special force fields count as physical or not. And this would seem to require a verdict from some other criterion of physicality, such as our second and third definitions. (b) ‘Physical’ = ‘inorganically identifiable’. On this definition, it matters what type of special fields are at issue. If they are a mental or vital force fields, then they will presumably count as non-physical. Referring to them as ‘mental’ or ‘vital’ force fields clearly doesn’t give us a way of referring to them directly in inorganic terms. Of course, we could always form new terms to name such fields. But these terms will arguably be ‘organic’ too, insofar as they refer specifically to entities that are found only in living bodies and never elsewhere. However, not all special fields associated with Broad-laws need be so exclusively attached to organic circumstances. There could be fields that arose specifically in certain complex inorganic chemical molecules, say. These fields would then come out as physical on the second definition. (c) ‘Physical’ = ‘resembles current physical categories’. As I suggested earlier, a natural way to fill this out is to require that putatively physical entities should display ‘mathematically simple and precise behaviour’. Any special force fields associated with Broad-style laws would be likely to satisfy this requirement. The principle of the conservation of energy is relevant here. Given this principle, any increases in kinetic energy occasioned by some force field must be compensated by a loss of potential energy with respect to that field, and vice versa. It is hard to see how this requirement could be satisfied if the evolution of any special fields were not governed by some definite mathematical principle that allowed us to define potential energy. To this extent, then, the third definition would count any Broad-style special fields as physical.
Overall, then, it looks as if special force fields associated with complex inorganic circumstances will come out as ‘physical’ on any definition, but that vital or mental force fields will only be ‘physical’ given the resemblance definition of ‘physical’, and not if ‘physical’ means inorganically identifiable. No doubt there is more to say on whether special force fields should count as ‘physical’. But I do not propose to pursue this issue any further. To the extent that special force fields do qualify as ‘physical’, the associated Broad-laws will illustrate my thesis that you can deny Microphysicalism without denying Physicalism: such laws will violate the Microscophysicalist thesis (L), yet will not take us beyond the physical realm. On the other hand, special force fields that count as ‘non-physical’ will be no good for my thesis, since their associated Broad-laws will not only violate Microphysicalism but Physicalism too.
Of course, this means that Physicalists must resist any force fields of the latter kind. But this presents no great difficulty. Whichever definition of ‘physical’ is in play, the only force fields that threaten physicalism are vital and mental fields. I take it that there is no good reason to believe in any such fields. Until the end of the nineteenth century, most scientists took vital and mental fields for granted, along with other special fields. But modern research has not supported their view. In particular, twentieth-century physiology has given no indication that there are any processes inside living bodies that cannot be fully accounted for in terms of more familiar physical forces. (Cf Papineau 2002 Appendix.)
8 Persisting Objects
I turn now from laws to another kind of fact that might fail to supervene on the microphysical facts, namely facts about persisting objects, like molecules, stones, brains, beetles and bicycles. These are objects that retain their identity through time: a stone at one time can be identical to a stone at another time. It will turn out that there is plenty of room for Physicalists to deny that facts about persisting objects are microphysically determined without compromising their Physicalism.
As with laws, a strong form of Microphysicalism about persisting objects would assert Humean Supervenience:
(O) All facts about persisting objects are metaphysically determined by the intrinsic properties and spatiotemporal relations of spacetime points.
Some contemporary philosophers endorse this claim. More specifically, they hold that facts about persisting objects depend on nothing but appropriate relations of spatiotemporal continutity among ‘time slices’ (and that facts about ‘time slices’ depend on nothing but the intrinsic properties and spatial relations of spatial points at the time in question). We can think of a time slice as conveying an instantaneous ‘snapshot’ of the putative object. The strong Microphysicalist view at issue is thus that a persisting stone, say, is determined by a sequence of stone-type ‘snapshots’ that over time trace a continuous stone-type ‘worm’ through space.
However, this strong Microphysicalist view is denied by at least as many contemporary philosophers as uphold it.7 In support, they standardly invoke Kripke’s ‘rotating disc’ argument.8 Consider a homogeneous disc made of completely smooth matter. A sequence of time slices will reveal where the disc is centred at each moment, but will not reveal whether it is rotating or not. In both cases, the times slices will simply be ‘frozen’ snapshots of homogeneous matter. So both a rotating disc and a non-rotating disc would display the same sequence of homogeneous time slices. Yet intuitively there is a difference between these two alternatives. It seems to follow that there are facts about the disc that are not fixed by relations of spatiotemporal continutuity among its time slices.
This then gives us one sense in which Physicalists might fail to be Microphysicalists about persisting objects without compromising their Physicalism. They can deny that persisting objects are sums of time slices. For it certainly doesn’t look as if this denial will somehow automatically undermine their Physicalism. After all, there seems no reason why Physicalists should withhold the term ‘physical’ from molecules or stones—or discs, for that matter—just because they think that these persisting objects fail to supervene on time slices. Persisting objects like these would seem to be the paradigm of physical objects, whether or not they supervene on time slices.
Perhaps there are few Microphysicalist philosophers who wish to uphold a strong Humean Supervenience thesis about persisting objects (just as few wish to uphold a strong Humeanism about laws). Still, the point I have just also applies also to various weaker Microphysicalist supervenience theses about persisting objects. There are in fact a range of possible weaker such Microphysicalisms, differentiated by what they add to time slices in search of an adequate supervenience base for persisting objects. Thus there are philosophers who hold that the way to stick the time slices together, so to speak, is to add instantaneous velocities to the supervenience base (Tooley 1988). Others favour the addition of primitive relations of singular causation (Zimmerman 1997).9 Yet others appeal to ‘non-supervenient relations’ between the time slices (Hawley 2001).
We need not dissect these strategies in any detail here. The important point for my purposes is simply that there seems plenty of room to dispute these weaker Microphyicalisms too, without thereby contradicting Physicalism. For a start, any supervenience thesis of the above form will be denied by ‘three-dimensionalists’, that is, those philosophers who deny that persisting objects have time-slices as temporal parts, and so a fortiori will reject any claim that persisting objects are time slices ‘glued together’ by such things as instantaneous velocities, singular causation or non-supervenenient relations. And even among ‘four-dimensionalists’, who do recognize time slices, none of these suggestions for gluing them together will have majority support. Yet, as before, there seems no reason why somebody denying any of these Microphysicalist theses should be deemed thereby to have compromised their Physicalism. As I said above, things like molecules and stones are paradigms of physical objects. We needn’t stop viewing them as such just because we deny one or more theses about how they are constituted out of temporal parts.
9 Brains, Beetles and Bicycles
Maybe molecules and stones are still paradigms of physical objects, even if they fail to supervene on time slices and relations between them. But what about other kinds of persisting objects, including organic entities like brains and beetles, and artefacts like bicycles? Here it is not so clear that their status as ‘physical’ will survive their failure to supervene on time slices plus ‘glue’. And, if their physical status doesn’t so survive, then this will argue that Physicalism about these entities does require some kind of four-dimensional Microphysicalism about persisting entities after all.
Let us suppose, for the sake of the argument, that three-dimensionalism is true, and that there is no way of ‘gluing together’ persisting objects out of time slices. Given this, it is by no means obvious that objects like brains, beetles and bicycles will still qualify as physical.
Recall how we earlier considered three different notions of ‘physical’: (a) microphysically determined, (b) resembling current physical categories and (c) inorganically identifiable. Under the hypothesis of three-dimensionalism, brains, beetles and bicycles clearly won’t qualify as physical because they are microphysically determined by time slices and their relations. Nor do they seem likely to qualify because they resemble current physical categories. As to the requirement of inorganic identifiablity, brains and beetles certainly won’t satisfy this; moreover, it’s not even clear that inanimate artefacts like bicycles will qualify, given that it is arguably essential to such artefacts that they are made by an intelligent designer.
This suggests that Physicalism isn’t compatible with three-dimensionalism after all, and that we need some doctrine of supervenience on time slices and relations to ensure that organic and artefactual persisting objects do not transcend the physical realm.
However, there is a further line of thought that promises to preserve the physical status of such objects even in the face of three-dimensionalism. For such objects might well supervene on their spatial parts even if they don’t supervene on temporal parts. And if those spatial parts are physical, then this will restore the physical status of brains, beetles and bicycles after all, even without any four-dimensional supervenience on time slices.
The thought here is that organic and artefactual objects will surely supervene on facts about atoms, molecules or other small material constituents, whatever view we take about temporal parts. Could you have two identical arrangements of molecules, and one constitute a beetle, or a bicycle, and the other not? It seems unlikely. And we have already argued, in the last section, that the physical status of paradigm physical objects like molecules will not be undermined by their failure to supervene on time slices. So this argues that beetles, brains and bicycles will retain their status as physical even if four-dimensionalist supervenience fails. All persisting physical objects, big and small, may fail to supervene on temporal parts, but as long as organic and artefactual objects supervene on small spatial parts, and those small spatial parts are physical, then organic and artefactual objects will count as physical too. (Note here how the recursive understanding of ‘physical’, flagged in section 3 above, matters here. Brains, beetles and bicycles may not qualify as physical in their own right, so to speak, but they will qualify derivatively, in virtue of their supervenience on their small spatial parts, plus the physicality of these parts.)
So the thought is that Physicalists can reject Microphysicalist four-dimensionalism and yet maintain their Physicalism by insisting that organic and artefactual persisting objects will still count as physical in virtue of the physicality of the spatial parts that they supervene on. A natural question to ask at this point is why there should be such supervenience on spatial parts, if there is a failure of supervenience on temporal parts. Does not my putative three-dimensionalist Physicalist owe us some argument for the claim that organic and artefactual persisting objects supervene on their spatial parts? However, such an argument is not hard to find. A version of the standard causal argument for Physicalism implies makes it very hard to see how organic and artefactual objects could fail to supervene on their spatial parts without generating an unacceptable species of systematic overdetermination.
To see how this would go, note that causes involving organic and artefactual objects characteristically have physical effects. (They dislodge stones, leave tracks, and so on.) At the same time those physical effects can surely be fully accounted for by causal processes involving only the small spatial parts of those objects. (The impacts of the molecules in those objects will fully account for the dislodging of the stones and the leaving of tracks.) So, if the organic and artefactual objects were metaphysically distinct from their molecular parts, in the sense of not supervening on them, we would two ontologically independent causes for the relevant effects, which would be absurd.10
So my putative three-dimensionalist Physicalists can offer a good argument in support of their crucial claim that organic and artefactual objects supervene on their small physical parts. At this point, however, we might well wonder why a similar argument won’t undermine their three-dimensionalism. If persisting objects can’t transcend their spatial parts without generating unacceptable overdetermination, then how come they can transcend their temporal parts? Why won’t this imply unacceptable overdetermination too, on the grounds that effects of causes involving the persisting object will already have full causes involving the temporal parts of that object?
However, I take it that somebody who is persuaded by the arguments for three-dimensionalism will deny the completeness premise assumed here. After all, they deny that persisting objects have temporal parts, and so a fortiori will not allow that there are already a full set of causes involving such temporal parts. Rather, they will insist that the only particular entities that feature in causes are persisting objects, like molecules and stones, or beetles and bicycles, not any supposed ‘time slices’ of those objects. So for them there will be no question of the effects of molecules and stones also being determined by facts involving temporal parts.
These last comments illustrate a general point. I have taken the canonical argument for physicalism to be the causal argument: putatively non-physical causes have physical effects; all physical effects have physical causes; so avoiding (strong) overdetermination requires the putatively non-physical causes to supervene on the physical ones. Now, if we could replace the second premise with a stronger claim that all physical effects in some sense have microphysical causes, then obviously the argument would deliver the conclusion that all putatively non-physical causes must supervene on causes which are microphysical in that sense. Correlatively, Physicalists who wish to deny that putatively non-physical causes are microphysical in some given sense must deny that all physical effects have microphysical causes in the relevant sense. The possibility of three-dimensionalist Physicalists illustrates the general point. It is specifically because they deny the relevant microphysical completeness thesis—that all physical effects have sufficient causes composed of time slices—that they are able to deny the metaphysical thesis that all physical causes must supervene on time slice facts.