On vitalism

I came across this quote:

In our recent science the Aristotelian doctrine is not dead. For but little changed, though dressed in new garments, this Aristotelian entelechy, which so fascinated Leibnitz, enters into the Vitalism of Hans Driesch; and of those who believe with him, that far as physical laws may carry us, they do not take us to the end: that the limitations of induction forbid us to pass in thought and argument from chemistry to consciousness, or (as Spencer well knew) from Matter to Mind; that Life is not merely ‘an outstanding difficulty, but a veritable exception to the universal applicability of mechanical laws’; that not to be comprehended under the category of physical cause, but to be reckoned with apart, is the fundamental conception underlying Life and its Teleology.

[Thompson, D’Arcy Wentworth. 1913, p29]

It slightly amuses me that when people assert that biology is not reducible to physical laws, they are, in historical terms, asserting tradition vitalism.

Now, when you hear the word “vitalism” you immediately will think of some sort of mysterian occult spiritualism, and this is certainly true of Driesch’s version of it, and to an extent of Oken and others in the 19th century, but it is not true of those called “vitalists” before the 19th century, or even many of those up until Driesch. For example, Lamarck’s version was entirely physical. He thought that life was an organising principle, a force that drove the evolution (literally, the development) of progressive complexity. In that respect, Lamarck is not so different from the systems biologists of today. So, too, was Blumenbach, who formulated the Bildungstrieb (in Latin, Nisus formativus), according to which living things were organised and hence different to inorganic things. It is worth noting, as Tobias Cheung has shown, that “organism”, a term used occasionally in French, did not become widely used in English until the mid-2oth century, and the preferred term in all languages for living beings was “organised being”. The notion of organisation, not too far from “complexity” or “complex adaptive system” of the modern era, was the definiens of biology, and that was the principle used by the person who coined the term biology, Treviranus (just beating out Lamarck).

Now a good many people, including Ernst Mayr, have made the claim that the laws of biology are not reducible to those of physics (in contrast to those who argue that biology has no laws, a view I incline to). Ironically, these are often those who equally firmly reject vitalism (of Driesch’s variety); the irony being that according to the prior definitions, they are themselves vitalists of a kind. Not much of philosophical import can be taken away from this, as debates and positions shift over time, but it amuses me, as I said…


Cheung, Tobias. 2006. From the organism of a body to the body of an organism: occurrence and meaning of the word from the seventeenth to the nineteenth centuries. The British Journal for the History of Science 39 (03):319-339.

Thompson, D’Arcy Wentworth. 1913. On Aristotle as a biologist with a prooemion on Herbert Spencer: Being the Herbert Spencer lecture delivered before the University of Oxford, on February 14, 1913. Oxford: The Clarendon Press.

21 thoughts on “On vitalism

  1. I have to say, I’ve always been slightly confused by the fact that folks like Mayr and others seem so *terrified* by the prospect that biology might (only in principle, most definitely not in practice!) be reducible to physics. It seems there’s this fear that if biologists admit this, universities worldwide will shutter their biology departments and pour all that money into high-energy physics…

  2. Not just biology departments. Biology reduces to chemistry, which reduces to physics, so the chemists will all be out of a job too.

    Physics imperialism knows no bounds. When they’ve demonstrated that time is reversible in an ‘n’ dimensional spacetime and that emotional responses are merely electro-magnetic impulses induced by simple compounds reacting in the presence of visible light, then they’ll come for the Art Historians.

    Or not.

  3. There’s a general equivocation about the nature of reduction here. It seems to go:

    1. Everything is physics (physicalism, or sometimes materialism)

    2. But the laws of biology are not reducible to the laws of physics

    Ergo, not everything is physics.

    The equivocation is in the term “reduction”: initially it meant that the terms of a theory could (either in principle or in practice) be replaced by suitable terms of another theory (i.e., here, those of biology to those of physics). However, many people take it to mean that everything in one domain ultimately is constituted by things in another. The former sense is what you might call “theory-reduction”, and the latter is what you might call “ontological reduction”. Only if you think that all ontology is based on theory, and there are many that do, is there a problem.

    But it seems that certain terms of biology are neither reducible to physics nor eliminable. This leads people like Mayr to argue that biology is somehow an emergent domain, based on physics but not wholly explicable in terms of physics. Reductionists, or better, physicalists (because in biology, “reduction” also means something like “genetic determinism”, confusingly) like me think this is simply an artefact of the way we represent biological things.

    1. What strikes me about the debate is that even when phrased accurately, it still feels like a red herring to me. Even if theory reduction were (in principle) true (something I don’t have real strong opinions about, because I don’t know enough complex dynamics/systems theory/etc. to be able to be a very good judge of such things), I don’t know why anyone would be confused into thinking that (in practice) it wouldn’t be the case that biology is the best set of tools for answering some questions (why does this organism have the particular wing loading ratio it has?) and physics the best set for answering others (what’s the expected outcome of a collision between a pair of protons?). Perhaps I’m just too much of a pragmatist.

      Something in your comment raises another interesting thought, though. I agree with you that it seems that many people (myself included) want to base ontology on theory – or, at the very least, to consider theory as the best guide we’ve got to ontology. Are you obligated by that position, then, to read the ontology *at each level* of nature off of our best theory *for that level*? That is, are you obligated to read physics for your ontology of tiny and really huge things, chemistry for your ontology of slightly-less-tiny things, and biology for your ontology of medium-size things? Here, it seems, is where the problem would come in. Do you think that the view of Mayr and others comes from a metaphysical position like this? Or is it grounded in something else?

      Ultimately, I’m fairly sure I agree with you that the “weirdness” of biological ontology has to do with how we represent biological systems. Physics gives us our ontology, the rest is a conceptual crutch for us limited human types. But the sociology and underlying commitments of all sides in this debate have always been really interesting to me. Related extra credit problem: what the heck do people mean when they talk about “emergence”?

  4. Obligatory xkcd citation.

    The biology we know is the biology of our planet. It may be that every feature of our cells reflects our contingent history. We have no reason to assume that life forms on other planets would necessarily parallel our biochemistry, apart from the fact that ours works and the alternatives might not.

  5. Remember Hilary Putnam’s example about the “laws” of children’s toys not being reducible to physics: you can’t fit the square peg into the round hole, but if you tried to explain that in terms of fundamental particle physics, the explanation would be so complex (think of the number of particles involved making up the two pieces of wood!) that it would be incomprehensible, and so wouldn’t EXPLAIN the facts to us human beings.

    I’m happy with this; if it makes me an emergentist about kindergarten equipment, then I’m an emergentist.

    … Of course, that makes me something like an emergentist about the “laws” of familiar mathematical critters as well. Well before we get to the sixth line of a long formula as printed on the page, my eyes glaze over: a formal proof written out in the primitive notation of Zermelo-Fraenkel set theory will not convince (because it won’t be understood). Knowing the right higher-level concepts to use is the trick: these concepts are formally definable in terms of the primitives in mathematics, but I wouldn’t expect the appropriate concepts to use in thinking about biological processes would necessarily be definable, formally, in terms of the primitives of physics.

    Entelechies rule, o.k.? !

    1. I find your comment about entelechies in the context of Putnam rather amusing, since it looks to me like he’s resurrecting Aristotelian formal causes. I don’t think this is emergentism, by the way – but rather an imposed property (by whom or what? Well, us, of course).

  6. many people take [“reduction’] to mean that everything in one domain ultimately is constituted by things in another. . . what you might call “ontological reduction”

    Note that merely being composed of physical things (atoms, say), isn’t going to be enough to secure ontological reduction (or physicalism). The vitalists supposed that organisms were composed of physical stuff.

    The important question is whether physical laws completely govern the behavior of living things. They do, so physicalism is true.

  7. I don’t remember where, exactly, in the Putnamian corpus the example occurs, so I can’t check exactly what he said about it– the discussion has the level of detail and rigor typical of Putnam’s work outside mathematical logic. I’ll also confess that I don’t have a very good grasp of Aristotle’s notion of formal cause, but I THINK what Putnam has in mind is an efficient cause: what keeps the peg from going in is that the corners of its square cross-section bump into the edges of the circular hole. And I don’t see the relevant properties (squareness and roundness of particular sizes) as being particularly IMPOSED.
    But at least it IS an amusing example!

  8. Physicalist – You say, “The important question is whether physical laws completely govern the behavior of living things. They do, so physicalism is true.”

    Well, that depends on what you mean by ‘govern.’ Let’s grant that the behavior of living things does not violate any known physical laws. That might be one sense of ‘govern.’ But that’s not to say that the behavior in question can be predicted or explained on the basis of known physical laws; a different sense of ‘govern.’ And so far as I know, we don’t have any plausible theories about the logical and/or methodological relations between these two senses.

  9. I don’t remember where, exactly, in the Putnamian corpus the example occurs . . .

    Try “The Meaning of Meaning.”

  10. that’s not to say that the behavior in question can be predicted or explained on the basis of known physical laws

    Not predicted in practice, but predictability in principle amounts to basically the same thing as saying that the laws aren’t violated.

    (To be a little more careful, we can say that any event that is accurately predicted by a special science could in principle be predicted with even more accuracy by known physical theories.)

    Predictability-in-practice and questions of explanation are both somewhat beside the point when it comes to questions of ontology.

  11. “Not predicted in practice, but predictability in principle amounts to basically the same thing as saying that the laws aren’t violated. ”

    No, not “basically the same thing” — not even close.

  12. Physicalist, you assert that:
    “The important question is whether physical laws completely govern the behavior of living things. They do, so physicalism is true.”

    Some physicists disagree:
    Paul Davis disputes your contention arguing from information theory (see “Emergent biological principles and the computational properties of the universe”; arXiv:astro-ph/0408014).

    In addition, George Ellis in his interesting essay “On the applicability of quantum physics” (available from the Fqxi website -Fqxi essay 2009. Topic: What’s Ultimately Possible in Physics?) uses quantum mechanical arguments to conclude:

    “…The implication is that the ability of physics to comprehend the dynamics of complex systems, such as life, is strictly limited: physics underlies and strongly constrains what happens, but in the end does not determine the unique outcome that actually occurs. This is determined by autonomous emergent higher level dynamics, such as the processes of genetics, physiology, and neurology. Physics enables all of this, but is not sufficient to determine the outcomes, even if infinite computing power were available. The essential nature of these processes, as a matter of principle, lies outside the domain of applicability of physics.”

  13. bob koepp asserts:

    No, not “basically the same thing”

    Yes it is, as long as we agree on what principles we’re allowed when deciding what’s possible in principle (and what is to be relegated to limitations “in practice”).

    How do we decide whether a physical law has been violated? We take the initial state and evolve it forward in time using the physical dynamics and boundary conditions. If the resulting theoretical state differs from the actual state, then the law has been violated.

    We can (in principle) use the very same method for making predictions. Take the initial state, and evolve it forward. It’s basically the same thing.

    Now, I’m implicitly stipulating that questions of computability only address what is predictable “in practice” and not “in principle.” Some won’t be willing to grant me that premise, but I think questions of ontology are not illuminated by focusing on our calculational limitations. Laplace’s demon could make the prediction, and that’s good enough to say it’s predictable “in principle.”

    I’m also setting aside niceties concerning stochastic physical theories. Worries along these lines can (to some degree) be resolved as in my “more careful” parenthetical statement above: physics will yield the chances of some event more precisely than will any special-science theory. If the world is fundamentally stochastic, then of course not all actual events can be predicted (but the probabilities of their occurrence can be).

    It’s also worth noting that the challenges of stochastic prediction will be mirrored by problems deciding whether events violate or respect stochastic laws. So once again, it is basically the same thing.

    Skeptic Tim points out that

    Some physicists disagree:

    To the extent that they do, they’re wrong.

    Davies builds his argument on Landauer’s attempt to shift the divide between what’s possible “in principle” and “in practice.” I find this account thoroughly unconvincing: the computational resources considered by Davies/Landauer are irrelevant to the ontology of laws. We know a good deal about where the laws of (e.g.) quantum electrodynamics hold and where they don’t, and we know that they aren’t violated in (e.g.) cellular processes. The laws are insensitive to complexity.

    Ellis’s paper is built on a rather naive analysis of the quantum measurement problem. If you’re worried about the measurement problem, then you should adopt some consistent interpretation of QM (e.g. Bohm’s hidden variable interpretation, or a form of spontaneous collapse), and see if your worry holds up. Ellis instead uses a naive measurement-collapse interpretation (mistakenly thinking that it’s what Bohr and Heisenberg had in mind), so it’s no wonder he lands himself in a muddle. Once you look at Bohm or spontaneous collapse models, it’s clear that there’s no problem getting non-linear dynamics for complex systems. (Further, even if Ellis’s did succeed in showing a problem with linear QM accounting for such dynamics, he still wouldn’t have an argument for top-down causation).

  14. Physicalist – I must persist in my view that “predictability on the basis of a system L of laws” is not equivalent to “not violating a system L of laws.”

    Start with a system of just one law. Imagine some phenomenon P that isn’t even addressed by L. Since it’s not even addressed by L, P is obviously not in violation of L; by the same token, it’s not predictable on the basis of L. Is there any reason to think that things change if L contains a plurality of laws? Not that I know of. Of course, if you stipulate that L is complete, then there can’t be such phenomena as P; but on what basis could you claim that L is complete without begging the question?

  15. @bob koepp:

    Yes, I was taking the completeness of physics as a premise. I’m glad we agree that (given this premise) the basic equivalence holds.

    but on what basis could you claim that L is complete without begging the question?

    In short, we have extensive scientific evidence for the domains of applicability of our current physical theories. We know, for example, that at very high energies quantum electrodynamics will not offer an accurate description, and we’ll need to incorporate quantum chromodynamics. At even higher energies, we would need a (yet-to-be-discovered) theory of quantum gravity.

    But at low energies, we know that QED will be fantastically accurate. And we know that the energies in cells, brains, etc. are well within the domain of applicability of QED. So for all special sciences we worry about, we know that physics is causally closed.

    We could expand this answer by pointing out that physics is completely insensitive to complexity, and that the distinctive feature of higher level entities is complexity.

    So we have strong (non-question-begging) reasons for believing that physics is causally closed, and no reason to doubt this position. Ergo, we should all be physicalists.

  16. Physicalist – To say that physics is complete (in the relevant sense) is to say that every truth can be expressed in terms of physics. I don’t see how your comments about quantum dynamics bear on this issue. But there’s enough weirdness in quantum physics so I wouldn’t be too surprised to learn that under one or another interpretation, there could be a connection.

Leave a Reply