Surprise! Naturalistic metaphysics undermines naive determinism, part II

James Ladyman

by Massimo Pigliucci

[The first part of this essay was published here.]

So, we have seen that Ladyman and Ross, in their book Every Thing Must Go (ETMG) are claiming that the only metaphysics worth doing is one that takes onboard all of the sciences, both physics and the so-called special sciences (from biology to sociology). We have also seen that their discussion is framed within the background of two broad issues: the critique of non-naturalistic metaphysics (which traces back at least to Hume and his “fork,” and of course to the logical positivists of the early 20th century), and the debate in philosophy of science between realists and anti-realists (remember: realism here means that scientific theories actually track truth about the world, while antirealism is the position that scientific theorizing has the more modest aim of being empirically adequate).

We are now ready to get to the heart of the matter, which consists of two major points: the implications of physics for metaphysics (including notions such as determinism), and the “third way” out of the realism-antirealism debate. Let me start with the latter, it will soon enough nicely connect to the former in an intellectually satisfying package (for me, at least, though I’m sure I will hear otherwise from some readers).

The view that attempts to steer a way between the Scylla of realism and the Charybdis of antirealism is often referred to as structural realism. The Stanford Encyclopedia article on the subject claims that this is “considered by many realists and antirealists alike as the most defensible form of scientific realism,” and it comes in two varieties: epistemic and ontic. Let’s get our initial look at structural realism by way of a summary by John Worrall, the guy who introduced the idea into the debate back in 1989. Worrall is talking about the shift — in 19th century optics — between Fresnel’s and Maxwell’s theories:

There was an important element of continuity in the shift from Fresnel to Maxwell — and this was much more than a simple question of carrying over the successful empirical content into the new theory. At the same time it was rather less than a carrying over of the full theoretical content or full theoretical mechanisms (even in “approximate” form) … There was continuity or accumulation in the shift, but the continuity is one of form or structure, not of content.

To put it in other words, realists are correct in the broad picture, but they are wrong about what carries from one successful theory to the other: new theories do not (necessarily) retain older theories’ description of unobservables (like ether), but rather their mathematical or “structural” content.

I must admit that the first time I heard about this I immediately thought about “obvious” counterexamples: seriously, the Copernican theory is structurally similar to the Ptolemaic one? Well, it turns out that it is, and S. Saunders has actually done the proper work to demonstrate it (besides, the example is likely not fair, since Ptolemaic astronomy was pretty much proto-scientific).

There are other prima facie problems with structural realism, for instance the fact that many theories in the special sciences are simply not framed mathematically (much of the theory of evolution, for instance), and it is therefore difficult to analyze their structure. Still, the work done by structural realists with physical theories quickly becomes pretty darn compelling, once one starts looking into it.

Ladyman himself, in work previous to ETMG, has helped distinguish between epistemic and ontic structural realism (note to the non-philosophically inclined: I know, it’s getting heavy; bear with me, the payoff at the end may be handsome...). The weaker form is the first one: here structural realism is interpreted as metaphysically neutral (i.e., concerned only with epistemology): it tells us that scientific theories describe the structure of the world, but it does not commit itself to the content of that world. You may remember the “shut up and calculate” brand of physicists who study quantum mechanics: they may be interpreted as antirealists, but they would also be at home within epistemic structural realism, insofar as they think that the equations describe the way the world is, but do not commit themselves to any metaphysical interpretation of what the world is made of.

The second form of structural realism, the ontic variety, is the one endorsed by Ladyman and Ross in ETMG, and it begins with the (reasonable, I think) point that one simply cannot divorce metaphysics from epistemology, and that whenever there is a conflict between the two, it is the latter that gets to occupy the driver’s seat. Ladyman and Ross explain their position by way of a Peirce-style (i.e., pragmatic) form of verificationism (distinct from the one that failed the logical positivists, though remember that L&R label their position “neo-positivism”). As they put it:

This verificationism consists in two claims. First, no hypothesis that the approximately consensual current scientific picture declares to be beyond our capacity to investigate should be taken seriously. Second, any metaphysical hypothesis that is to be taken seriously should have some identifiable bearing on the relationship between at least two relatively specific hypotheses that are either regarded as confirmed by institutionally bona fide current science or are regarded as motivated and in principle confirmable by such science.

Chew on that for a moment, but it basically means that the metaphysician doesn’t get to run wild without a way for the rest of us to square what he says with the best empirical knowledge we have about the world. No metaphysics without adequate epistemology. Seems sensible enough to me.

And we now get to ontic structural realism, the position endorsed by Ladyman and Ross, and which is beginning to convince me (with some reservations here and there). This is how they themselves put it:

Ontic Structual Realism (OSR) is the view that the world has an objective modal structure that is ontologically fundamental ... According to OSR, even the identity and individuality of objects depends on the relational structure of the world. ... There are no things. Structure is all there is.

Hence the title of the book: Every Thing Must Go! Now, before you go all New Agey or Buddhist on me, please note that Ladyman and Ross derive their metaphysics from the best physics available. The details are fascinating, and in themselves make the book a must read, but essentially their claim is that all currently viable theories in fundamental physics — including quantum mechanics, string theory, M-theory and their rivals — have in common principles like non-locality, entanglement and such, which point toward the surprising conclusion that “at bottom” there are no “things,” only structure.

Yes, I know, you are going to ask “structure of what?” “relations among what?” and so on. And the answer appears to be that those are the wrong questions to ask. Fundamental physics seems to do away with objects, and indeed, it does away with yet another old chestnut of metaphysical speculation: causality!

Causality has been a troubled concept since Hume’s famous deflating analysis of it, but quantum mechanics — and, again, all the other currently viable candidate physical theories — simply tell us that at the lowest level of analysis the concept breaks down, it doesn’t do any work for the physicist. Philosophers have noted for a while now that fundamental physicists talk about laws and mathematical descriptions, but they don’t talk about causes very much, if at all. And modern physics explains why: at bottom, there are no causes.

But wait a minute! Are Ladyman and Ross telling us that causes and objects are illusory? Is this yet another instance of people claiming that things that we think exist and play a crucial role in our understanding of the world do not actually exist? Are we to do away with tables and people, just like some pundits these days want to argue that free will, consciousness, morality and so on, are illusions, because none of them have a place in fundamental physics? Are Harris, Rosenberg and other modern nihilists right after all??

Nope, they are not. (Here begins the payoff of all the hard work we’ve done so far.) Let’s take causality first. According to Ladyman and Ross it is a concept that is eliminated in fundamental physics, but needs to be retained by the special sciences (from biology to economics). That’s because causality makes sense only in systems for which there is temporal asymmetry (a before and an after), and that — while not being the case for physics — is very much the case for the special sciences. L&R do not treat the concept of causality as an “illusion” to be dispelled once the special sciences are reduced to physics, because no such reduction is in the cards.

But why not? For the same reason that “things” must go at the fundamental level but need to be retained at the level(s) of analysis of the special sciences. Let’s take the standard example of a table. It is fashionable these days among the scientifically literate to shock us by saying that the “table” right in front of our noses doesn’t “really” exist (and of course, neither does our nose, or ourselves), because physics tells us that the apparently solid object is actually made of things like protons, neutrons and electrons (or quarks, or strings, you pick, it doesn’t matter). But this, according to ontic structural realism, is still a pretty limited way of looking at the issue. At bottom there are no things, and hence not even protons, quarks or strings, there are only structures. These structures generate patterns, and science is in the business of describing such patterns. At one level, the pattern can best be captured by talk of protons and electrons; at another level (i.e., for material science, and of course for our everyday experience) they are captured by objects like tables. Tables, then, are not illusions at all, at least no more than protons and electrons are illusions; rather, they are the most appropriate way to describe a certain stable pattern.

The same goes for causality: when historians, economists, biologists and so on talk about “X causing Y” they are simply deploying a concept that is useful for capturing patterns that are affected by time asymmetry, and that are no more or less illusory than patterns at any other level of analysis of reality. The only difference between physics and the special sciences, according to Ladyman and Ross, is that the former is concerned with patterns that have for all effective purposes a very very large domain of stability (both in space and time). Biologists, instead, are concerned with patterns that have local stability both in space (earth-bound, for now) and time (the duration of the life of an individual, or of a species).

The surprising upshot of all of this is that physicalist reductionism — the idea that all the special sciences and their objects of study will eventually reduce to physics and its objects of study — is out of the question. And it is out of the question because of a metaphysics (ontic structural realism) that is based on the best physics available! If you are not blown away by this you may not have caught the thing in its entirety and may want to go back and re-read this post (or, if your philosophical and physical chops are adequate, ETMG).

This has all sorts of implication for those increasingly popular (and, I think, annoying) statements about determinism and reductionism that we keep hearing. Turns out that they are based on bad physics and worse metaphysics. There is no fundamental determinism for the simple reason that there is no fundamental causality, and that “cause” is a conceptual tool deployed by the special sciences that has no counterpart in fundamental physics, and so it cannot be reduced to or eliminated by the latter.

This doesn’t mean that all is fine and clear in ETMG or with ontic structural realism in general. There are still plenty of open questions to be worked out (I hinted at one above: what are we to make of scientific theories that do not deploy math? Which structures are conserved there?). But the satisfying picture emerging from all of the above is this: a) metaphysics has to be based on epistemology, and it cannot do without taking physics very very seriously; b) the special sciences — while obviously compatible with physics (which sets their universal boundaries) — retain an enormous amount of independence from it and cannot be reduced to it; c) we still have a lot of work to do, both within philosophy and within the special sciences, to make sense of the world.

One last parting shot, about a topic that the astute reader may have noticed I have bypassed so far: if every thing is gone and we only have mathematical structures and relations, what is the ontological status of mathematical objects themselves? Here are the only relevant quotes from Ladyman and Ross that I could find:

OSR as we develop it is in principle friendly to a naturalized version of Platonism. ... One distinct, and very interesting, possibility is that as we become truly used to thinking of the stuff of the physical universe as being patterns rather than little things, the traditional gulf between Platonistic realism about mathematics and naturalistic realism about physics will shrink or even vanish. ... [Bertrand Russell] was first and foremost a Platonist. But as we pointed out there are versions of Platonism that are compatible with naturalism; and Russell’s Platonism was motivated by facts about mathematics and its relationship to science, so was PNC [Principle of Naturalistic Closure] -compatible. 

Wild stuff, no? Now I don’t feel too badly about having written in sympathetic terms about mathematical Platonism...