Organisms and Machines

In the “The Disunity of Reason” thread, Mung suggested that “the typical non-theist will insist that organisms are machines, including humans.” And there is a long tradition of mechanistic metaphysics in Western anti-theism (La Mettrie is probably the most well-known example). However, I pointed that I disagree with the claim that organisms are machines. I’m reposting my thoughts from there for our continued conversation.

A machine is a system with components or parts that can be partially isolated from the rest of the system and made to vary independently of the system in which they are embedded, but which has no causal loops that allow it to minimize the entropy produced by the system. It will generate as much or as little heat as it is designed to do, and will accumulate heat until the materials lose the properties necessary for implementing their specific functions. In other words, machines can break.

What makes organisms qualitatively different from machines is that organisms are self-regulating, far-from-equilibrium thermodynamic systems. Whereas machines are nearly always in thermodynamic equilibrium with the surrounding system, organisms are nearly always far from thermodynamic equilibrium — and they stay there. An organism at thermodynamic equilibrium with its environment is, pretty much by definition, dead.

The difference, then, is that machines require some agent to manipulate them in order to push them away from thermodynamic equilibrium. Organisms temporarily sustain themselves at far-from-equilibrium attractors in phase space — though entropy catches up with all of us in the end.

It is true that some parts of an organism can break — a bone, for example. But I worry that to produce a concept general enough that both breaking and dying are subsumed under it, one can lost sight of the specific difference that one is trying to explain.

Indeed, that’s the exact problem with Intelligent Design theory — the ID theorist says, “organisms and machines are exactly the same, except for all the differences”. Which is why the ID theorist then concludes that organisms are just really special machines — the kind of machines that only a supremely intelligent being could have made. As Fuller nicely puts it, according to ID “biology is divine technology”.

43 thoughts on “Organisms and Machines

  1. I think you are trying to cast a spectrum as a dichotomy, and I don’t think this works very well. Organisms come in a wide variety of equilibria with their environment. Tardigrades in their tun state can last almost indefinitely, at an amazing range of temperatures and other extreme conditions.

    Conversely, machines can be devised to include a wide variety of feedback loops allowing for self-repair and adaptation.

    You also seem to be hinting at modularity, with the notion that machines are constructed of independent modules. But this is not an inherent characteristic of a machine, and machines have been made as integrated non-modular wholes without components or parts that can be effectively isolated. Today we think of assembly lines assembling machines from interchangeable parts, but this is relatively recent. Earlier machines were more like sculptures than accumulations of lego blocks.

    Machines (as we’re using the term) can also be made to construct other machines. The (still theoretical) singularity in AI is for the AI to start reprogramming itself, thereby augmenting its intelligence geometrically, soon exceeding any human capability to understand it. This day may come.

    I personally have no problem regarding biological organisms as extremely sophisticated “machines”. Given a few billion years, the physical and chemical wherewithal for self-construction, and a reward system for doing it incrementally better all the time, the result can surely be regarded as “organic technology.” I just don’t see anything divine about the process.

    (and I think we are on the verge of genetic manipulation, “engineering” characteristics into organisms to meet specifications that selective breeding can’t achieve. At the same time, genetic algorithms are using evolutionary principles to solve problems intractable to brute force calculation. The line between organism and machine, if it ever existed, is getting very blurry.)

  2. So all the biology textbooks have to be rewritten? They all refer to the molecular machinery inside of living organisms. Ever study kinesiology? More machine talk wrt our limbs and the way they work.

    A living organism “is a system with components or parts that can be partially isolated from the rest of the system and made to vary independently of the system in which they are embedded.”

  3. Yes, I agree that organisms are not machines, while computers are machines.

    In the earlier thread, automobiles were brought up. My view is that the combustion is not mechanical, though everything else is mechanical (or near enough to mechanical).

    Organisms do have mechanical parts, such as bones. Well, those are mostly mechanical, but not entirely mechanical.

    I also agree that ID is a mechanistic program. It is the claim that a designer used mechanism to project the designer’s purposes.

  4. It will generate as much or as little heat as it is designed to do, and will accumulate heat until the materials lose the properties necessary for implementing their specific functions. In other words, machines can break.

    Humans die from overheating. Rather fragile materially, compared with many machines. Organisms “break,” too, we just call it death when they do.

    What makes organisms qualitatively different from machines is that organisms are self-regulating, far-from-equilibrium thermodynamic systems.

    They’re self-regulating, of course, as they must be to live as evolving organisms. Increasingly, machines are self-regulating, but have some way to get to organism level. And what does “far-from-equilibrium” really mean?

    I found this without looking for long. How right or wrong they might be I don’t really know, probably in part because “far-from-equilibrium” really is a pretty nebulous concept. I know that we’re thermodynamically unstable, being composed of reduced carbon, hydrogen, etc., but isn’t steel much the same? Or bronze? Or carbon fibers?

    Isn’t that one of the big faults of ID that is often ignored, too? Life can’t deal with high heat, yet there should be nothing fatally wrong with life making metal parts. It just would be almost impossible to evolve, so we’re stuck with squishy wet stuff, with which life began.

    Whereas machines are nearly always in thermodynamic equilibrium with the surrounding system

    Well, how is that so? Aside from heat engines, aren’t lights, TVs, and computers operating rather far from thermodynamic equilibrium? I mean, even apart from the materials, which generally are rather above equilibrium (rust is, more or less, equilibrium for iron).

    organisms are nearly always far from thermodynamic equilibrium — and they stay there.

    Well, if they turn off, they die (I don’t think an inert tardigrade or anything else is truly close to equilibrium, just meta-stable). Which isn’t bad for most organisms, who mostly need to get on with reproducing, etc. Machines usually are meant to stop at times because otherwise they’d uselessly expend resources when we don’t need them running.

    The difference, then, is that machines require some agent to manipulate them in order to push them away from thermodynamic equilibrium.

    If you want, you can buy a generator that starts when the power goes off, keeping your security system functioning, along with heating/cooling, as well as other systems. No agent needed to turn on. How many gadgets do we have by now that turn on “automatically”?

    It is true that some parts of an organism can break — a bone, for example. But I worry that to produce a concept general enough that both breaking and dying are subsumed under it, one can lost sight of the specific difference that one is trying to explain.

    And vice versa. I wouldn’t want to lose sight of the fact that an organism “dies” because it gets too far from good living conditions that it can no longer sustain its processes. As would be the case of the failure of a space station or some such thing. We do often use similar terms with respect to “breakage” of both machines and bodies, such as “damage,” “failure,” “break,” and “shut down.” Because the similarities are considerable, as are the differences.

    Indeed, that’s the exact problem with Intelligent Design theory — the ID theorist says, “organisms and machines are exactly the same, except for all the differences”.

    Certainly the differences are huge and important. For instance, life has all of the limitations expected of evolution, with a number of inefficiencies that humans would have fixed, especially over time. But it’s one thing to realize the differences, it’s another to think that there’s some especial difference between organisms and machines, when it’s physics for both of them. I don’t think well of vitalism, as if there is some category difference between organisms and machines.

    Which is why the ID theorist then concludes that organisms are just really special machines — the kind of machines that only a supremely intelligent being could have made.

    But they’re not. That’s one reason I’d not try to claim too great a difference between machines and organisms, something that smacks of vitalism.

    One should also note that it’s a kind of machine designed by the Designer, and yet it’s the magical soul for mental abilities. They’re never really content with humans being really special machines, rather they split between magic doing all of the pretty lights and things and a kind of Frankenstein’s monster for the body.

    As Fuller nicely puts it, according to ID “biology is divine technology”.

    I hesitate to call living beings “machines,” especially since we’re nowhere near to being able to make machines of that kind of complexity–and we don’t want to make machines with too much autonomy (an issue as we get closer to truly impressive robots). Then there are differences in that life isn’t made for a purpose, and it isn’t designed. However, “machine” doesn’t seem to rest on evolution, purpose, or anything else that exists in life and doesn’t exist in human-made machines, so I also don’t feel like life can be truly held not to comport to the meaning of “machine.” I don’t think it would be felicitous to call humans machines, then, but I’m not going to deny that they can fit the meaning, either.

    Glen Davidson

  5. From the OP:

    A machine is a system with components or parts that can be partially isolated from the rest of the system and made to vary independently of the system in which they are embedded, but which has no causal loops that allow it to minimize the entropy produced by the system.

    Don’t go all Neil on us, KN. 🙂

    That’s a highly non-standard definition of “machine”. Why not use a more generally accepted one?

  6. GlenDavidson:I don’t think it would be felicitous to call humans machines, then, but I’m not going to deny that they can fit the meaning, either.

    To paraphrase Clarke, a sufficiently advanced machine is indistinguishable from life.

  7. I’d like to know where the idea that intelligent design theorists think organisms are machines comes from. Does that come from Paley’s watch or something more recent?

    I doubt they think humans are machines but do think that humans are organisms.

  8. I was initially tempted to argue that with a machine there is no concept of self, and that this is different for organism.

    But then I came across this:

    An automaton is a self-operating machine.

    Dare we say that organisms are self-aware and maintain that machines are not?

  9. Thanks for clarifying your understanding of a machine, KN. I take it you apply this same understanding to a “mechanistic” approach to science and that this might influence how you think of “mechanisms” in the philosophy of science which emphasizes their role in scientific explanation.

    Based on this OP and your view that “I don’t think that physics and chemistry are mechanistic” from the Disunity thread, I am guessing a machine for you might be characterized as something that involves Newtonian mechanics only.

    I personally don’t think the word “machine” is that helpful. I prefer the three qualities that you raised earlier: being alive, being sentient, being sapient. I’d see all three as continuums and not simple dichotomies.

    For life, I think the paper you cited earlier What Makes Biological Organisms Teleological (pdf) is helpful, both for life and for agency. Being a self-organizing dissipative system seems necessary but not sufficient, for there are simple dissipative systems which are organized to maintain their own existence like Bénard cells, flames, hurricanes. To eliminate such systems, the authors require in addition a complexity in internal causal structures. That is vague and so in line with the idea that we are talking about a continuum.

    On sentience, which I take as phenomenal consciousness, I think we need to have a better theory of what kind of structures support it. Perhaps that theory can be mathematical, as Tononi’s IIT, or perhaps it will be expressed in the form of mechanisms, like Baars GWT. The philosopher Tye thinks that sentience can be recognized purely by behavior. But in any event it seems reasonable that there will be vagueness for some situations.

    On sapience: Here I understand that you rely on language, or perhaps the capacity to have language since you say in the Disunity thread that it requires an entity which “not only infers and reasons, but it reasons and infers with others”. Is using a language a vague concept? I don’t know.

  10. petrushka:
    When machines can evolve they will more nearly fit the definition of life.

    So cars have remained the same throughout their history? Really?? Computers haven’t changed over then years? Really??

  11. As Fuller nicely puts it, according to ID “biology is divine technology”.

    With the exception that ID does not require the divine…

  12. I don’t believe organisms are machines, but how does one make the case one way or another?

    The most obvious difference to me is that organisms are alive and machines are not alive, but that smacks of vitalism, and for that reason I think others may reject that distinction.

    It might be easier to make the case that organisms are machines, that they are constructed of machines that have been cobbled together by evolution (e.g., the flagellum) and are nothing but cobbled together assemblages of such cobbled together assemblages, and wholly mechanical. Given a naturalistic view of life and the cosmos, how could it be otherwise?

    And how are humans exempt?

  13. Frankie: With the exception that ID does not require the divine…

    What else do you call that which designs and creates the universe then? Which is what you claim, remember? ID from the smallest to the largest, it’s ID at all levels remember? So your particular version of ID certainly is IDC – Intelligent Design Creationism.

  14. BruceS:

    Based on this OP and your view that “I don’t think that physics and chemistry are mechanistic” from the Disunity thread,I am guessing a machine for you might be characterized as something that involves Newtonian mechanics only.

    I suppose I just don’t know what to day about quantum computers!

    I personally don’t think the word “machine” is that helpful .I prefer the three qualities that you raised earlier:being alive, being sentient, being sapient.I’d see all three as continuums and not simple dichotomies.

    Of course they are continuous. (I certainly hope I didn’t come across as suggesting otherwise!) A distinction only becomes a dichotomy when the items are conceptualized in such a way that it is impossible to see how they are related to one another.

    The basic distinctions urged here — life, consciousness, and rationality — correspond precisely to the three types of “soul” (psyche) described by Aristotle in De Anima (I recommend this this translation). Aristotle distinguishes between “the vegetative soul”, “the sensitive soul” and “the intellectual soul”. I think that’s pretty much right. I’m much more sympathetic to Aristotle’s moderately teleological psycho-biology than I am to the rest of his project.

    For life, I think the paper you cited earlier What Makes Biological Organisms Teleological (pdf) is helpful, both for life and for agency. Being a self-organizing dissipative system seems necessary but not sufficient, for there are simple dissipative systems which are organized to maintain their own existence like Bénard cells, flames, hurricanes.To eliminate such systems, the authors require in addition a complexity in internal causal structures.That is vague and so in line with the idea that we are talking about a continuum.

    All that seems right. I would also think that a semi-permeable membrane is necessary so that there’s a distinction between the internal and the external, wherein the dynamics of the internal system can vary somewhat independently of the larger system in which it is embedded. And the membrane must be semi-permeable to enable both organizational closure and thermodynamic openness.

    On sentience, which I take as phenomenal consciousness, I think we need to have a better theory of what kind of structures support it. Perhaps that theory can be mathematical, as Tononi’s IIT, or perhaps it will be expressed in the form of mechanisms, like Baars GWT. The philosopher Tye thinks that sentience can be recognized purely by behavior. But in any event it seems reasonable that there will be vagueness for some situations.

    I would say that phenomenal consciousness is one aspect of sentience. The other major aspect is animal intentionality. Some philosophers think that these are necessarily fused — whatever has one, has the other. I dislike strong necessity claims like that, but it nevertheless strikes me as highly plausible that intentionality and consciousness are somehow fused properties. At present I don’t have a complete theory of animal intentionality; that’s the very thing I’m working on right now, and will be working on for most of the summer.

    But sure, there will be marginal cases. It’s quite easy to say that mammals and birds have intentionality + consciousness. But does anyone have reliable intuitions about spiders or dust-mites?

    On sapience: Here I understand that you rely on language, or perhaps the capacity to have language since you say in the Disunity thread that it requires an entity which “not only infers and reasons, but it reasons and infers with others”. Is using a language a vague concept? I don’t know.

    I think language is a vague concept, actually. I say that because multiple lines of evidence suggest that language evolved from some sort of “proto-language.” If one looks at the ape language experiments, it seems that there are some grammars that are more easily acquired by apes than others, that they do understand some primitive semantic concepts, but they don’t transmit their language skills to others. To the best of my knowledge there has never been a case of two or more language-competent apes having a conversation in sign or icons.

    That said, I would also want to take language in a really broad sense here — when I’m talking about “sapience” or “sapient intentionality”. Joseph Margolis puts the point this way:

    “Intentionality (in an enlanguaged world) ranges over a great deal more than mental states: it ranges over everything that is a cultural artifact or to which we rightly attribute meaning or significance or signification (as expressive or representational or symbolic or geistlich: language, traditions, institutions, practices, products, and actions most particularly, all of which are actual and objective … I name this sort of intentionality, ‘Intentionality’” (p. 143).

    So when I talk about “sapient intentionality”, I’m not just talking about discursively articulable mental states — I’m talking about all stuff that Hegel, Cassirer, Gadamer etc talked about, too (“traditions, institutions, practices”) which is inseparable from any fully adequate examination of the discursively articulable mental states emphasized by analytic treatments of intentionality.

    Another way of putting it is to say that I’m interested in the evolution of culture or Geist (“spirit”) from nature, and how to naturalize spirit without reducing spirit to nature.

  15. Patrick:
    Guano’d a comment.From the Rules:“accusing others of ignorance . . . is off topic”.

    The shoe fit, even though you may disagree.

  16. Mung: I’d like to know where the idea that intelligent design theorists think organisms are machines comes from. Does that come from Paley’s watch or something more recent?

    It’s definitely in Paley, but there are older references as well. I would imagine that it has a long history. Once you assimilate organisms to machines (as Descartes was the first to do, 150 years before Paley), and then recognize that machines are extrinsically purposive — they have purposes because of the functions given to them by their makers — the modern ID argument falls into place.*

    In fact, come to think of it, we know that people were making this argument before Paley because Hume takes this argument seriously in Enquiry Concerning Human Understanding and Dialogues on Natural Religion, and both of those were published many years before Paley’s Natural Theology.

    * as distinct from the classical ID argument that goes back to the Stoics or even Socrates, if Sedley is right.

  17. Mung: I’d like to know where the idea that intelligent design theorists think organisms are machines comes from. Does that come from Paley’s watch or something more recent?

    The picture of the flagellum on the masthead of UD comes to mind. KF’s fishing reel for another

  18. Kantian Naturalist: I would also think that a semi-permeable membrane is necessary so that there’s a distinction between the internal and the external…

    I absolutely agree with you here.

  19. Kantian Naturalist: Once you assimilate organisms to machines (as Descartes was the first to do, 150 years before Paley), and then recognize that machines are extrinsically purposive — they have purposes because of the functions given to them by their makers — the modern ID argument falls into place.

    I think the point about Descartes is valid, but it applies just as well to non-theists, who look for “the ghost in the machine,” and it’s just as easy to argue that the modern ID argument is drawn from modern science and it’s mechanistic philosophy (and so argues Feser).

  20. Mung,

    I think the point about Descartes is valid, but it applies just as well to non-theists, who look for “the ghost in the machine”…

    Why do you think that non-theists look for a “ghost in the machine”?

    Many non-theists are physicalists who don’t believe that such a “ghost” exists.

  21. KN, to Bruce:

    A distinction only becomes a dichotomy when the items are conceptualized in such a way that it is impossible to see how they are related to one another.

    KN,

    X/not X is a dichotomy, and it’s easy to see that the relation is one of negation.

  22. keiths: Why do you think that non-theists look for a “ghost in the machine”?

    Many non-theists are physicalists who don’t believe that such a “ghost” exists.

    But they do believe that the machine exists, just not the ghost.

    Now if we could just find someone “skeptical” to support the claim that organisms in general, and humans in particular, are in fact machines. Because, you know, that’s what skeptics do. They support their claims or they retract them.

    Skeptical of the ghost but not skeptical of the machine. Because there is “objective empirical evidence” that organisms (including humans) are machines. Right?

  23. Mung: the modern ID argument is drawn from modern science and it’s mechanistic philosophy (and so argues Feser).

    I agree with Feser on that particular point.

  24. Kantian Naturalist: I suppose I just don’t know what to day about quantum computers!

    Perhaps we should ask the Canadian PM? Scott Grades Trudeau’s Description of Quantum Computing. Also see Scott’s 10 minute answer:

    Of course they are continuous. (I certainly hope I didn’t come across as suggesting otherwise!)

    Nah, I was just being pedantic.

    I would say that phenomenal consciousness is one aspect of sentience. The other major aspect is animal intentionality. Some philosophers think that these are necessarily fused — whatever has one, has the other.

    As I understand the paper linked in my previous post, the authors grant animal intentionality to any (biological alive) entity meeting the criteria they state in the paper, not just sentient life.

    An interesting NR article on the sentience of octopuses. Some good links, especially this short short SF story on the issue by a philosopher.

    From another post:

    Mung: the modern ID argument is drawn from modern science and it’s mechanistic philosophy (and so argues Feser).

    KN: I agree with Feser on that particular point.

    But then you say in the Disunity thread that “I don’t think that physics and chemistry are mechanistic”. That seems to be a contradiction. Are you saying that scientists are stuck in a “mechanistic” Newtonian-era view of science but if you look at modern science itself, it is not consistent with that view?

    For me, this is more evidence that it is unhelpful to use the term “machine” or the words “mechanistic” or “mechanism” if they are meant to reflect that Newtonian view of machines.

  25. BruceS: But then you say in the Disunity thread that “I don’t think that physics and chemistry are mechanistic”. That seems to be a contradiction. Are you saying that scientists are stuck in a “mechanistic” Newtonian-era view of science but if you look at modern science itself, it is not consistent with that view?

    Yes, that’s what I’m saying.

    I’m saying that ID people — both the ‘theorists’ and also the rank-and-file members of the ‘movement’ — mistakenly believe that (a) Darwinian biology is mechanistic and (b) physics and chemistry are mechanistic.

    It’s only because they believe (b) that they think that some kind of intelligent being is necessary to get from mechanistic physics/chemistry to non-mechanistic biology.

    Whereas on my view, none of the natural sciences are “mechanistic” in any robust sense — not even quantum mechanics, which is really a theory of fields or structures than a theory of “things”.

    Once one rejects mechanistic metaphysics all the way down, one has undermined the conceptual basis of ID, because one has eliminated the problem that ID is supposed to solve.

  26. BruceS: For me, this is more evidence that it is unhelpful to use the term “machine” or the words “mechanistic” or “mechanism” if they are meant to reflect that Newtonian view of machines.

    But quantum mechanics is no less mechanistic than Newtonian mechanics.

  27. Mung: But quantum mechanics is no less mechanistic than Newtonian mechanics.

    This is way above my level of meager competence, but my limited comprehension is that quantum mechanics actually is less mechanistic than classical mechanics. Classical mechanics is a theory of external forces acting on rigid bodies Quantum mechanics is a theory of fields, in which the forces/bodies distinction does not hold. I think that makes for a big difference. But I’m open to correction here from others who know more!

  28. Mung: But quantum mechanics is no less mechanistic than Newtonian mechanics.

    What did I say about using that word, “mechanistic”? So I am not going to get into a bun fight about what is and isn’t so.

    Besides, does anyone know what QM actually IS? I mean, is there any scientific or philosophical consensus on what QM says exists?

    Hint: No.

  29. Kantian Naturalist: Yes, that’s what I’m saying.

    I’m saying that ID people — both the ‘theorists’ and also the rank-and-file members of the ‘movement’ — mistakenly believe that (a) Darwinian biology is mechanistic and (b) physics and chemistry are mechanistic.

    I did not know you meant just ID people.

  30. BruceS: I did not know you meant just ID people.

    Well, it’s not only them.

    One sees elsewhere the argument that organisms must be mechanistic because non-living systems must be mechanistic and there’s no ontological divide between living and non-living systems, because any ontological divide would have be cashed out in terms of differences in underlying composition.

    That’s pretty much Jacques Monod’s explicit line of reasoning in Chance and Necessity, where his commitment to mechanism is dressed up as “the postulate of objectivity”, and one sees it also in Dawkins’s announcement that life merely has “the appearance of design.”

    I have three different objections here.

    Firstly, the science of dissipative systems at far-from-equilibrium thermodynamic relations seems like a very promising way of naturalizing teleology, which means that we don’t need to throw teleology under the bus.

    Secondly, the very idea that the sciences are unified, such that any smooth reduction can proceed, has been seriously contested by philosophers of science like John Dupre and Nancy Cartwright.

    Thirdly, if Ladyman and Ross are right in what they say in Every Thing Must Go, it’s just not the case that general relativity and quantum field theory are theories about things or objects that are passively acted upon by external forces. They are theories about spatiotemporal geometries (in GR) or about sub-spatio-temporal geometries (QFT). (Assuming I’m getting this right.)

    While I’m less than enthusiastic about ontic structural realism, I’m inclined to think that what there is, in the scientific image, is best characterized (relative to our most highly-confirmed models circa early 21st century) in terms of relatively stable processes that are themselves characterized by a plurality of stances or attitudes.

    The mechanistic stance is one such stance, and it works quite well when it comes to building an engine or setting a bone. The teleological stance works perfectly well in many other cases. We shouldn’t confuse either the stances with each other or confuse the stance itself with the processes characterized from that stance.

  31. Kantian Naturalist,

    I’m saying that ID people — both the ‘theorists’ and also the rank-and-file members of the ‘movement’ — mistakenly believe that (a) Darwinian biology is mechanistic and (b) physics and chemistry are mechanistic.

    Strange that no references were provided. It’s as if KN can’t support its claims and has to erect strawmen to make points.

    Evolutionary biologists say that Darwinian biology is mechanistic. That is because drift and NS are mechanisms of change.

  32. newton: So the the parts of an organisms are machines, but not the organism.

    And from this you get the idea that intelligent design theorists think organisms are machines?

Leave a Reply