Skip to content

Some loose ends – Reductionism and Phylocode

Last updated on 18 Sep 2017

I’ve been asked in the comments to cover two topics, neither of which I want to discuss at length because they are not easy to cover, and because they aren’t the focus of my rather intense monomania right now. They are: Reductionism and Phylocode.

Reduction is a relation between theories, according to the book by Ernest Nagel back in 1961 (incidentally, Nagel’s title, The Structure of Science, found an echo in Thomas Kuhn’s book the next year).

Nagel proposed that the task of science was to “reduce” all theories to physics; and reduction was where all the parts, processes, causes and other formalisations of the one theory could be accounted for by the other (“reducing”) theory without any remainder. In short, and very roughly, one theory explains everything the other theory explains. Consider chemistry: chemistry reduces to physics just insofar as all the facts chemistry explains can be given a physical explanation (say, in terms of subatomic particles and fields).

Now in biology, “reductionism” usually means that the reducing theory is genetics, but it need not be between genetics and other fields (such as animal behaviour), but between different versions of genetics. The late David Hull proposed that molecular genetics did not reduce to population genetics, because there was a many-to-many relation between them. A reduction usually is expected to either have a many-to-one (the “one” being an item in the reducing theory) or a one-to-one relation between theories.

The alternative view is that two theories do not reduce, and each stands as a whole on its own (a view known as “holism”). However, in my view there are several problems with holism: one is that it presumes that there exists something, a property or process or entity, that is not physical. I have discussed this before here. It is my view that physicalism, the view that everything is physical, entails an ontological reductionism. That is, we must believe that nothing exists which cannot be given a physical explanation, at least in theory (we do not, of course, have the One True Physical Theory yet, and even if we had, it might be too Hard to give the full explanation). So I reject emergent properties as properties of the world. But one might be a methodological holist and assert that given our limitations, we must always address phenomena at their “level” as self-standing wholes, and hope that one day we may find a path to reduce them to physics. I think this is wrong too, because it soon appears that any phenomenon is at least partially going to appeal to physical facts right off the bat; what happens is that we increase the fraction of facts that are simply physical over time.

Another problem with holism is that it assumes reduction is “layercake”: that psychology reduces to biology, which reduces to chemistry, which reduces to physics. But this cannot be true either, in my view: if something is reducible and is a physical process, then we must appeal to physical facts at every point. Ultimately (and this is an ontological claim), everything reduces immediately to physics. A physicalist like me cannot presume the layercake has anything but historical importance. You have a brain state that is you being sad? That’s physical. You are a reproducing organism? That’s physical, and so on. There is no traction in thinking that the reduction must go through intermediary “layers”.

Emergence and holism are facts about us, about what we find hard to integrate or which surprises us given what else we know. It’s a computational point about bandwidth, processing time, and algorithmic parsimony. Okay, enough from me: there’s a very good entry at Stanford I have linked before, which calls my version epistemic reductionism. Suffice it to say I am what my friend Paul Griffiths calls a “clubfooted reductionist”, and proud of it.

Now, onto the Phylocode. A lot of things have been asserted about the Phylocode, which is an attempt to make classification of organisms a matter of phylogenetic relations, and not of the traditional Linnaean scheme. It has been said that it removes essentialism, that it is historical not conceptual, that it will remove unnatural groups in taxonomy, and that it is modern, Darwinian, or something like that. Critics say that it is costly (it involves reclassifying everything we have classified already, and teaching this, changing the reference works, and the specimens in museums and repositories), that Linnaean taxonomy was already removing unnatural (i.e., nonmonophyletic) groups, and that the new scheme will be wildly unstable as new phylogenies are published.

I have no dog in this hunt. It seems to me that it is a solution to a badly needed problem, and that in no way is it any less or more “essentialistic” than its target, the Linnaean scheme (which has been pretty adaptable). If we agree that ranks are arbitrary and conventional, then we can stop using Linnaean ranks as inference tickets, and there remains, so far as I can tell, no other objection to it. But if people think that the Phylocode is better as a convention, then fine. This is not about natural classification, and most certainly not about philosophies of classification, so I don’t care. Many who I know do care though. I’ll watch from over here with the pop corn and scorecard.

21 Comments

  1. Perplexed in Peoria Perplexed in Peoria

    A physicalist like me cannot presume the layercake has anything but historical importance.

    I would think that would make yours a rather odd career choice, then. Critiquing the craft used in the construction of ephemera.

    • John S. Wilkins John S. Wilkins

      Who said history was unimportant? Except for metaphysics, that is?

  2. Brian Brian

    John, what do you think of Searle’s argument in his book ‘mind’ that consciousness is causally reducible to biological function, but not ontologically reducible. That is consciousness if a product of biology, but is not just biology. I guess the analogy would be with Searle’s argument that money exists, it’s ontology was social. It’s substrate is reducible to paper, metal or bits, but none of them would be money if the money didn’t exist as a social construction. I’ve probably made a mess of all that, but I’m sure you know what Searle meant.

    • Brian Brian

      I take it from your silence that you have used the revered simian tradition of flinging poo?

    • John S. Wilkins John S. Wilkins

      Brian, this came while I was travelling and I didn’t see it until just now.

      I am unsure what “consciousness” refers to, or “mind”. So instead I would rather refer to definite and more or less verifiable aspects of the brain’s function in a social context. So I tend to reject the idea that “mind” exists except as a trashcan category under which we lump a range of disparate processes, features, facts and mythological properties.

      There’s also a problem of order here, which the money example highlights nicely. Money is an abstraction of an abstraction of an abstraction of concrete entities. It’s a faux metric for dispositions to treat a range of things that have no real connection, apart from their being the sorts of things we have a disposition to behave about, which we represent in particular ways. So the object levels are:

      Stuff in the physical world;
      Our representations of stuff in the physical world;
      Our dispositions to behave towards our representations of stuff in the physical world;
      Our metrification of our dispositions to behave towards our representations of stuff in the physical world.

      As you can see, “money” does not exist except as a third order abstraction.

      Now, repeat this ad nauseum for all “social” objects. Call the resulting aggregate “mind”. So I disagree with Searle.

  3. David Williams David Williams

    But isn’t the Phylocode dead?

    • John S. Wilkins John S. Wilkins

      That depends on who you are talking to, I think…

      • David Williams David Williams

        Three meetings of the International Phylogenetic Nomenclature have taken place, the 3rd (the last one) in 2008. None of the proceedings have been published, in spite of various promises. The International Society for Phylogenetic Nomenclature fora seem deathly quiet. The code itself gets tinkered with every now and then but, like the Norwegian Blue, it looks pretty much dead to me.

        • John S. Wilkins John S. Wilkins

          That looks dead. Now I have all this popcorn. What am I going to do with it?

          Oh well; I’m sure I’ll find a philosophical barney going on somewhere…

    • It’s not dead; the companion volume is just taking longer than initially expected.

      Reports have been published for each of the three meetings. See: http://www.ohio.edu/phylocode/events.html

      There’s a bug in the forum’s software; it should probably be taken down.

      • John S. Wilkins John S. Wilkins

        Do we have any indication how many taxonomists, as a proportion of those actively doing it, are likely to adopt the Phylocode? I have an impressionistic view that they are in the minority by a fair amount.

      • John Harshman John Harshman

        Responding to Wilkins: There is of course no real data (or are, if you prefer), but that doesn’t stop me. I think the PhyloCode will be quickly adopted except by the older crowd and (oddly) the Hennig Society. However, it will be adopted unevenly. The ICZN stakes no claim on any group above family-level, and this will be at first the main province of the PhyloCode. I expect that ICZN will continue to dominate the species level, where PhyloCode makes no major claims. And in between, we’ll initially see PhyloCode definitions fit into ICZN forms and endings. Or at least that’s how I intend to go about it. I expect that if a genus is found to be not monophyletic, specific epithets will indeed be placed in new combinations with generic names, for example. And I’m not sure I’m going to italicize everything either. But phylogenetic definitions have definite advantages in a molecular era; a lot of clades just have no clear diagnoses otherwise.

        • John S. Wilkins John S. Wilkins

          I’m reminded of the point made by Toulmin, that the uniformitarians and catastrophists eventually met in the middle, with only words separating the two schools…

      • I basically agree with John Harshman’s take. I’ll also add that I think different disciplines will vary wildly. The ISPN’s membership is overwhelmingly botanists and vertebrate zoologists, with the odd malacologist, protistologist, etc. Many fields are not represented at all; for example, I don’t think there is a single entomologist in the society. So, for example, we will probably see a relatively high adoption rate within vertebrate zoology (especially vertebrate paleontology) and a relatively low (or perhaps nonexistent) adoption rate within entomology.

        Hard to predict, though.

        I made some further speculations here.

  4. To a partisan like me, it’s heartening to know that reductionism of some flavor remains a live option in biology. I recently read Superconductivity: A Very Short Introduction, and the author takes the history of superconductivity research as a refutation of reductionism itself.

    I would have brought out a quote, but my computer is being slow/stupid.

    • John S. Wilkins John S. Wilkins

      Much of this is going to depend on the definitions of terms. For me, “reduction” means the intertranslatibility of one theory into another without remainder. That the properties of the theory being reduced require the arrangements of the parts as well as the properties of them is besides the point. If superconductivity involves properties of the ensemble, that is not a refutation of reductionism.

  5. bob koepp bob koepp

    I’m sympathetic to reductionism, in the sense that I reject all a priori arguments purporting to demonstrate that it is a fatally flawed program. Hull’s argument about many-many relations is unpersuasive. While ‘reduction’ is a slippery notion, if we accept Nagel’s exemplar of reduction (i.e., temperature reduces to mean kinetic energy), then this sort of reduction is perfectly compatible with many-many relations — a particular temperature for an ensemble is compatible with many distinct molecular configurations; and a particular molecule having a particular kinetic energy is compatible with many temperatures for the whole gaseous ensemble of molecules.

  6. Jim Thomerson Jim Thomerson

    The proper museum will continue using whatever obsolete system of nomenclature it started out with, and will never change. As a result those of us who work on particular groups know where the group is located on the selves. Just a quick look will tell us whether fishes are located according to Regan or to Berg.

  7. John Harshman John Harshman

    There is only one proper test of whether the PhyloCode is living or dead. Do systematists use it? I’ve seen a number of recent publications that do, so by that criterion it’s alive. Here’s one that I happen to recall right now:
    McGuire, J. A., C. C. Witt, J. V. Remsen, Jr., R. Dudley, and D. L. Altshuler. 2009. A higher-level taxonomy for hummingbirds. Journal of Ornithology 150:155-165.

  8. Adam Adam

    If you’ll forgive my presumptuousness, here’s my Cliff-Notes argument against reductionism. I am not familiar with the the philosophical debates, so I can’t address them (I’ll get around to reading that encyclopedia article). I can only present my opinion as an anecdote of what biologists believe about biology. This may reflect my own concern with science as a practical matter (allowing us to predict the future) as opposed to any concern with The Truth.

    As I understand physics, every particle affects every other particle. Therefore, reductionism becomes a holism in itself — the universe exists and does what it does, no particle can be examined in isolation. As a corollary of Occam’s Razor the behavior of the universe cannot be calculated using any system smaller than the universe. Even the behavior a moderately complex system, such as billiard balls, cannot be predicted with much accuracy without accounting for the state of a huge portion of the universe.

    With that background, each holistic/emergent “layer” of knowledge is focused on a class of phenomena that can be understood/predicted reasonably well without reference to the universe as a whole.

    So maybe emergence and holism are about us, but it is “us” as a part of the universe. Not us as Homo sapiens. Not us as scientists in 2010. It is about how the universe is structured that allows parts of it to be examined and understood in isolation from the rest of the universe.

  9. John S. Wilkins John S. Wilkins

    It’s not about the independence of any physical objects from any other; that is not required. It’s about the one domain (say, the biological) being fully and totally explained by the properties and processes and parts of another (say, the physical), with no remainder.

Comments are closed.

Optimized by Optimole