Contemporary Physics and the Little Leeway Thesis

Further support for the Little Leeway Thesis, this time from French and McKenzie's 2012 paper, "Thinking Outside the Toolbox":

"It is indisputably essential to Lewis’ reductive ambitions in general that all the perfectly natural properties are intrinsic, for otherwise they cannot be subject to the ‘principle of free recombination’ that lies at the heart of his modal system. Furthermore, the inclusion of the predicate of ‘perfectly natural’ as an ideological primitive is motivated by the theoretical benefits of doing so; since it is the suitability of these properties to analyze duplication that grounds the majority of these benefits, and since their ability to analyze duplication requires that they be intrinsic, unless these properties are intrinsic then the motivation for including this primitive is seriously undercut. Now, for Lewis (as for others) a property is intrinsic only if it could be had by a lone object (which of course is to say in the Lewisian framework that there is a world in which a lone object has it; cf. Langton and Lewis 1998). Furthermore, it is clear that the only ‘perfectly natural’ and hence fundamental properties that we – as this-worldly agents – may be said to have any epistemic acquaintance with are, by physicalism, the properties of fundamental physics. These Lewis lists as “the charges and masses of particles, also their so-called ‘spins’ and ‘colours’ and ‘flavours’, and maybe a few more that have yet to be discovered.” (1986, 60). Hence if we want to verify the crucial claim that all the perfectly natural properties are intrinsic - as all Lewisians surely should – then all we can reasonably hope to do is to check that these properties are intrinsic. But how are we to do this?

There seem to be just two places (to our knowledge) where Lewis considers the issue. In one, he writes that “On my analysis, all of the perfectly natural properties come out as intrinsic. That seems right.” (1983, 16). In the other, he asserts that “It can plausibly be said that all the perfectly natural properties are intrinsic.” (1986, 61). Unfortunately for Lewisians, however, it is not at all obvious that these properties are intrinsic: there are in fact good reasons to say that they are either simply not intrinsic or at best such that their intrinsicality must remain forever unbeknownst to us.

The most expedient argument that one can marshal against the claim that all the fundamental physics properties are intrinsic is perhaps that which exploits the fact that the (current best candidates for) the fundamental laws of physics are formulated as local gauge theories. The basic idea underpinning such theories is that the equations governing particle interactions should be generated from the interaction- free equations by demanding that those equations are invariant under a local gauge (or ‘local phase’) transformation. Thus, in order to generate the properties of particles through which they undergo fundamental interactions (such as the colours of quarks and the charges of electrons), one must apply the appropriate gauge transformation to their interaction-free equation (in both cases the Dirac Lagrangian, which describes the free motion of spin-1/2 particles). This is in fact now viewed as the fundamental guiding principle of particle physics (though the underlying reason for this is a matter of dispute). But the essential point for our purposes is that these local gauge transformations applied to the free-particle equations imply the existence of at least one new particle, since the implementation of the procedure inevitably introduces what is called a gauge boson. In the case of electrodynamics, for example, this particle is the photon; in the case of the strong interaction we introduce the gluons, and similarly in the case of the weak interaction we obtain the W and Z bosons. Thus if we understand the properties through which the fundamental constituents of matter interact in terms of gauge transformations, and these bring in their wake the appropriate gauge bosons, then it looks as if we have no choice but to say that the properties such as charge and colour are not the sort of properties that lone objects can have, and hence that these properties are not after all intrinsic.

If this conclusion is correct, it represents a very bad result for Lewisians. In light of it, we can envisage Lewisians defending themselves by means of one of the following two strategies. The first strategy is to accept that these properties are indeed extrinsic but to take this as a signal that they are not after all fundamental; rather, what is fundamental is a previously unacknowledged external relation. 14 The claim is thus that we should reconceive of charge, colour and other properties involved with gauge transformations in terms of relations that do not supervene on properties of their relata (presumably in this case principally fermions and gauge bosons). The details of this would certainly have to be worked out, though there are at least two worries that we have about this general approach. First of all, such a strategy sits uncomfortably with the supposedly ‘physicalistic’ claim of Lewis that it is up to physics to provide an inventory of the this-worldly fundamental properties and relations, given that physicists apparently do count these properties as fundamental and do not appear to ever make reference to the alleged external relation – whatever it is - that is (hypothetically) being appealed to here. And secondly, even if such an external relation can be cooked up, we do not see how any such relation could hope to be specified without making reference to gauge symmetry; since this symmetry is a feature of laws, presumably no such relation could be taken as a denizen of the ‘non-nomic base’ Lewisians take to determine laws.

A more plausible strategy would be that of holding that even though as far as actual physics is concerned these properties are conceptually entwined with the implementation of local gauge transformations, and even though these transformations bring in their wake the corresponding gauge bosons, these bosons are nevertheless only contingently associated with these properties. This would of course amount to a denial of a certain form of nomological essentialism that would need to be argued for in the specific case at hand, though we should note that Lewis’ general argument against nomological essentialism rests upon the principle of recombination - a principle whose validity turns on precisely that which is currently in question (see Lewis 1986, 162-3).

In any case, if the gauge-theoretic argument against intrinsicality is sound, then if Lewisians want to maintain that charge is nevertheless intrinsic they must establish that there is a possible world in which the laws are consistent with a lone charged particle – a lone proton, say. By the above argument, such a world of course cannot be a world in which the actual laws hold. In having to ascertain whether such a world is possible, we therefore find ourselves once again having to contemplate lone-proton worlds wholly bereft of the resources with which to analyze them, namely, the theories of protons and of electric charge that physicists have painstakingly constructed and provided us with. Without anything with which we can meaningfully establish that lone objects can have these properties, then, if we want to continue to claim that all the perfectly natural properties are intrinsic then we must 




















































either simply stipulate it or remain agnostic on the issue. But either way this is a bad result. Not only is intrinsicality required for their free recombination, the inclusion of perfectly natural properties within Lewis’ system is motivated principally by their theoretical fecundity and, as mentioned above, that fecundity is overwhelmingly dependent on their intrinsicality. Hence without any good reason to believe any longer that all the fundamental properties are intrinsic, we should be hesitant about continuing to appeal to them at all. And that is no less the case when it comes to appealing to them to solve the ‘simplicity’ problem that - if left unchecked - ruins the best system analysis.

Now, we are of course not pretending to have done a full survey of all the Humean analyses one could deploy in this situation, nor of all the possible get-out clauses that defenders of the one account we did look at might exploit. But we do hope to have shown that there is a significant lacuna in this most familiar of modal analyses from a naturalistic point of view. A defence of Lewis’ analysis, as with the rival anti-reductionist account, inevitably requires us to consider actual fundamental properties in worlds in which the actual laws of physics do not hold. But once again, since everything that we know about these properties is tied to our theories and hence to the laws that actually hold, we find in each case that there is simply nothing useful that we can say in such scenarios."

"...consider...the lone proton world. The only theory we know of that (we think) correctly describes the proton is the Standard Model of particle physics; but that the Standard Model’s laws could apply in this world is (to say the least) far from clear. Think of the questions that a physicist would have to address in an attempt to ascertain whether this was indeed the case. Could the lone-world proton – defined by a certain set of determinate fundamental properties – have (all the) mass that it is actually taken to have, in the absence of the Higgs? Could the proton be properly said to be charged in the absence of photons that mediate the electromagnetic interaction? Could the symmetries of the actual laws – say matter - anti-matter symmetry – be said to hold in a world permanently devoid of antimatter? Similarly, given that actual protons are related to other types of hadrons via global SU(3) symmetry, could this symmetry be said to hold in a world in which there are no tokens of these other hadron types? And is it possible for a world to contain just a single quantum particle throughout its entire history given that quantum mechanics gives a finite probability for all particles to decay to a particle of another type (although here the proton is perhaps a special case)? It can quickly be seen that layer upon layer of questions – questions with non-trivial answers - must be addressed if we are to progress with this issue. We once again see that although the assumed categorical basis is certainly ‘simple’, the defense of the claim that that basis represents a solution of actual laws is certainly not; and furthermore, if the defense of the claim that the lone-proton world represents a solution of the actual laws is highly non-trivial (if indeed it can be defended at all), what hope have we of defending the idea that it is a solution of realistic natural laws – i.e. laws of similar complexity to those of actual physics – when we don’t even have any idea what those laws are supposed to look like?"