What Is Life? Part III: What Might an Organism Be, If Not a Machine?

This is the third and final part of a series of columns exploring the conceptual foundations of biology.

In “Part I: The Problem of Agency,” I argued that there are good reasons for taking the teleological character of life at face value, and for viewing normative agency as an objectively real feature of all living things.

In “Part II: The Poverty of Darwinism,” I reviewed various aspects of neo-Darwinian theory and argued that, conceived of as a theoretical framework for reducing teleology and normativity to mechanism, Darwinism is faced with a dilemma. On the one hand, if we assume organisms are machines (with no inherent power of compensatory self-adjustment—normative agency), then evolution becomes inexplicable. On the other hand, if we relax the mechanistic assumption, then the theory of natural selection in effect smuggles in normative agency by the back door and therefore begs the question at issue.

As a matter of logic, demonstrating that Darwinism fails to provide an adequate explanation for the inherent normative agency of life in no way depends upon having a better theory to take its place. However, as a matter of rhetorical strategy, my claims are unlikely to persuade absent some indication of how normative agency is to be understood, from a physical perspective. This is the problem I turn to now.

* * *

The problem of life is fantastically difficult. Living systems have radically different properties from nonliving systems. The main difference between the two, phenomenologically speaking, is that living things act spontaneously so as to maintain themselves in existence.

In a phrase, broken bones heal; broken stones don’t.

Of course, this is hardly a novel observation. Already more than two centuries ago, the great physiologist, Xavier Bichat, the “father of histology” who first enunciated the concept of anatomical tissues, wrote:

La vie est l’ensemble des fonctions qui résistent à la mort.(1)

This power of resistance, if taken at face value, presupposes a capacity on the part of every living thing that is hard to describe in a non-tendentious way, but which might be termed “non-indifference” towards its own non-existence (“valuing” or “caring”); a capacity to perform work in a goal-directed manner, with the aim of keeping non-existence at bay (“acting” or “striving”); and a capacity to do so in an effective way—to do what is necessary and appropriate to achieve the goal of postponing death (“intelligence”). If we are honest with ourselves, all of this remains deeply mysterious from a physical point of view. How is it possible for such systems to exist?

To be sure, it has been widely supposed for some time now that we ought not to take the phenomenology of life at face value, because now we know better. Skeptics will say that this “vitalistic” characterization of life may have been plausible before the development of cybernetic control theory and before the rise of molecular biology—both in the wake of World War II—but that by now we know very well what accounts for these seemingly mysterious properties: Cells are machines composed of other machines within machines—all endowed with manifold, interacting, negative feedback controls. And it is this elaborate system of controls that gives rise to the false appearance of caring, striving, and intelligence in living systems.

Actually, the downfall of “vitalism” was heralded long before that. Since as long ago as Friedrich Wöhler’s inorganic synthesis of urea in 1828, scientists have been proclaiming the death of “vitalism.” So, let me take a moment to explain my position on this.

There are two different ways of interpreting the charge of “vitalism.” According to the first way, one is a vitalist if one believes there is some nonphysical substance or principle that is superadded to the matter composing living systems. According to this definition, I am not a vitalist, as will be apparent in a moment.

The other way of looking at “vitalism” is as the claim that living matter is categorically distinct from nonliving matter. This is what is often denied by scientist who point out that organisms are just composed of ordinary matter—CHNOPS.(2)

According to this second definition, I clearly am a “vitalist,” because I am indeed claiming that all forms of living matter have distinctive properties that all forms of nonliving matter lack. But, of course, the mere fact that living systems are composed of ordinary matter does not prove that the systems are not distinctive, because the way in which the ordinary matter is organized in living matter may give it such properties.

In the same way that water vapor and liquid water and ice are all made of the same matter—hydrogen and oxygen molecules—and yet have very diffferent properties, so too is it possible that macromolecules as huge a proteins may have quite distinctive properties—unlike anything else in nature—even though they are composed mainly of carbon, hydrogen, nitrogen, and oxygen. So, a fortiori, may the immense mass of densely crowded macromolecules that make up cytoplasm have distinctive properties. I will return to this important point shortly.

But before that, there is a pair of objections I need to consider. The first is the claim that, by all appearances, it is a simple observationsl fact that cells are machines. Why? Well, not to put too fine a point on it, because they look like machines.

Yesterday, in “Cells Sure Look Like Machines,” I posted a couple of video animations of knowledge we now have of the functioning of two cellular subsystems: the ribosome and the RNAi system. And, yes, it sure does look like what we have there are machines. So, this is an important objection to my position that I must adddress.

The second objection is a little hard to state precisely. It has been raised by different authors, using different terminology. But it boils down to the fact that the functional interactions within cells, though consisting individually of physical and chemical interactions, nevertheless are not directly determined by the laws of nature, in the sense that a given interaction can either occur or not occur, depending on the local circumstances. This means that both possibilities exist—occurrence and non-occurrence of any given interaction—so far as the general laws of nature are concerned.

Either possibility may be actualized, but which possibility is in fact actualized in a given instance will depend entirely on local circumstances, and not on any consideration of first principles of physics or chemistry. In other words, you cannot explain the physical interactions occurring in living things through the direct application of physical laws alone.

A concrete example that is often cited in this connection is the DNA molecule. There is no law of physics which mandates the particular arrangement of the four perpendicular nitrogenous bases, ACGT, along the sugar-phosphate backbone. On the contrary, it is precisely the fact that any of the four bases can occur in any position along the backbone, so far as the underlying laws of physics are concerned, upon which the function of the DNA molecule depends. This independence of the specific physical interactions in living systems from the general laws of physics is a deep and mysterious aspect of life, and one that absolutely distinguishes life from nonlife.

The Nobel Prize–winning molecular biologist, Jacques Monod, called this property of living systems “gratuité” [gratuitousness].(3) The philosopher of science, Ernest Nagel, introduced the term “orthogonality” into the discussion, meaning that the local funtional interactions in living systems are not conditioned upon (are “orthogonal to”) the laws of physics.(4) This is the term I will use, for the most part.

Most recently, David L. Abel has discussed this singular property of life in greater depth and with more persuasivness than anyone before him.(5) His way of describing the property at issue is in terms of the distinction between physical “constraints” and functional “controls.” He shows in exacting detail that the control aspect of life can never come into being as a direct result of mere physical constraints. Therefore, controls must not be identified with constraints. Rather, life uses physical constraints in order to achieve control. Controls are constraints transmuted into functions.

None of these ways of describing the phenomena is exactly perspicuous, but I will stick with Nagel’s term, “orthogonality,” as being less question-begging than “control,” while acknowledging that Abel is quite correct to stress the fact that the laws of physics alone—as currently understood—can never explain the phenomenon of orthogonality.(6)

The upshot of all of this is that I seem to be fighting an uphill battle if I wish to claim that organisms are not machines. How shall I respond to these two seemingly crushing objections?

As to the first point, that cells and their component parts look like machines: These videos do not constitute observational evidence in their own right, but rather are based on the theoretical interpretations of molecular biologists, who use the machine metaphor to help make sense of the bewildering experimental data. The videos do not depict anything anyone has directly observed. Rather, they are based on models—theoretical constructions—that are heavily influenced by the machine metaphor to begin with. Of course, this would not matter, if the machinelike nature of the models rested on direct, irrefutable experimental evidence. But does it?

There is growing experimental evidence that the machine metaphor for macromolecules, organelles, and cells may not be nearly as well grounded empirically as we have believed. I only have space here for a couple of examples, but those interested should see my dissertation, Teleological Realism in Biology, for other examples and for a much fuller discussion.

My first example is muscle contraction. Striated muscle tissue consists of many bundles called myofibrils. Each myofibril, in turn, consists of many tubes consisting of two types of paired filaments, thick and thin, the former made of the protein myosin, the latter of actin. Connecting across the filaments at regular intervals are myosin “cross bridges.”

A standard model has it that muscle contraction is explained by the myosin cross bridges’ exerting force on the actin thin filament, causing the thin filament to slide in relation to the thick filament, while both filaments remain constant in length. According to this model, the small “heads” of the myosin cross bridges act as motors, pulling the much larger thin filaments along. Calling the myosin heads “motors” emphasizes the machineline nature of the model.

It is not so well known, however, that there are problems with this model, and that other types of models have been proposed. The main problem is that it makes little physical sense to suppose that the tiny myosin heads can exert enough mechanical force to move the much larger thin filaments. Moreover, there is little direct experimental evidence that the myosin heads are actually displaced in the way called for by the mainstream model. Gerald H. Pollack, a professor of bioengineering at the University of Washington, has advanced a completely different model that appears to be in closer agreement with the empirical evidence. Accordng to the Pollack model, the actin filaments contract—technically, “reptate”—and the physical forces that cause the reptation derive from propagating phase transitions along the gel-like thin filaments themselves.(7)

My second example comes from the work of Alexei Kurakin, a pathologist at Harvard Medical School. In a series of seminal papers, Kurakin has emphasized the many ways in which what we know about cellular structures deviates from the machine model. Above all, he emphasizes the transience of many organelles, which seem to coalesce out of the cytoplasm as needed, only to dissipate again soon afterwards. But it is best for me to let him speak for himself:

To summarize, the newly revealed and unexpected properties—such as steady-state character, transient self-organization on demand, stochastic dynamics and interconnectedness—that characterize cellular structures and molecular machines believed to exist as pre-assembled complexes designed for certain functions according to programs and blueprints, clearly suggest the inadequacy of expectations and assumptions based on the mechanistic intuition.(8)

One objection to this might run as follows:

Fine. “Transient self-organization on demand” certainly doesn’t sound like any kind of machinery that human beings are capable of building. But what about the orthogonality problem? And doesn’t Abel’s critique of conflating constraints with controls show that ”self-organization” cannot be the answer we are seeking? Because the sort of phase transitions and self-organization that Pollack and Kurakin are invoking are just ordinary physics at the end of the day—exergonic, “downhill” processes driven by well-known thermodynamic forces. In a phrase, free-energy minimization. So, doesn’t Abel’s work show that Pollack’s and Kurakin’s ideas cannot be right?

Well, I agree that Abel’s work shows that Pollack’s and Kurakin’s work cannot be the whole story. Something more is needed. Nevertheless, there is good reason to believe that the currently mainstream machine models of cellular function leave a lot to be desired. And even if the specifics of Pollack’s and Kurakin’s proposals do not stand the test of tiem, the very fact that they offer such a drastically different interpretation of biological functions in itself shows that the machine model is not inevitable—that it is not true that no alternative to the machine model of life is conceivable. Many alternative, non-mechanistic physical models of biological phenomena are already on the table, and more are being proposed all the time.

But what about Abel? At the end of the day, I agree that his work shows that we must go beyond current concepts of self-organization. What does that mean?

One obvious proposal would be to look to look upon the disciplines of “self-organization”—principally nonlinear dynamics and nonequilibrium thermodynamics—as a necessary but not sufficient step towards understanding life. Nonlinear dynamics is especially promising, because the notion of a nonlinear attractor is an obvious candidate for the sort of virtual state that is needed if we are to avoid the “backwards causation” problem in our efforts to model teleological action.

But if attractors, phase transitions, and the rest of it are not sufficient to explain the orthogonality of biological interactions, then what would be?

I am not a scientist. Nor am I aware of any existing body of theoretical work that gives an adequate response to this question. But one obvious suggestion would be to take the fact that cytoplasm is a highly crowded phase of condensed matter with gel-like and liquid-crystal-like properties much more seriously than we have done up until now.(9)

If we do, then we can perhaps begin to catch at least a glimmer of what a real solution to the problem of normative agency might look like—namely, a hitherto unknown conservation law, and corresponding variation principle, associated with the macroscopic quantum-field properties of cytoplasm, that would correspond to a global constraint on biological action equivalent to something like “viability.” A physical but non-mechanical constraint on action that might amount to control.(10)

Of course, this suggestion is just a shot in the dark. Only time will tell if a genuine alternative to the machine model of life is really in the offing. But at least such speculations show that the thought that organisms are not machines—but rather genuinely normative agents with inherent capacities for acting in accordance with purposes and values—cannot simply be dismissed out of hand.
________________________
(1) “Life is the totality of those functions which resist death.” Xavier Bichat, Recherches physiologiques sur la vie et la mort (GF-Flammarion, 1994); p. 57. Reprint of 4th ed., 1822; 1st ed., 1800.
(2) Carbon, hydrogen, nitrogen, oxygen, phosphorus, sulfur.
(3) Jacques Monod, Le hasard et la nécessité (Seuil, 1970). Translated as Chance and Necessity (Vintage Books, 1972).
(4) Ernest Nagel, “Teleology Revisited,” Journal of Philosophy, 1977, 74: 261–301. Reprinted in idem, Teleology Revisited and Other Essays in the Philosophy and History of Science (Columbia UP, 1979).
(5) David L. Abel, The First Gene (LongView Press, 2011).
(6) Another concept closely related to the orthogonality problem is “information.” Originally, I intended to discuss the concept of information in this post, but space limitations force me to postpone that discussion to another occasion. Let me just say here that information, properly speaking, is always meaningful. In other words, information is always information-for-an-agent. Moreover, meaning is a normative concept. Therefore, we way in which cells use information is itself an important part of what we are trying to explain.
(7) Gerald H. Pollack, Cells, Gels, and the Engines of Life (Ebner and Sons, 2001); pp. 225–247.
(8) Alexei Kurakin, “Self-Organization vs. Watchmaker: Stochastic Dynamics of Cellular Organization,” Biological Chemistry, 2005, 386: 247–254; p. 250.
(9) See Gilbert Chauvet, The Mathematical Nature of the Living World (World Scientific, 2004); Mae-Wan Ho, The Rainbow and the Worm: The Physics of Organisms, 3rd ed. (World Scientific, 2008); Alwyn C. Scott, The Nonlinear Universe (Springer, 2007); Giuseppe Vitiello, My Double Unveiled (John Benjamins Publishing Co., 2001); and F. Eugene Yates, “Homeokinetics/Homeodynamics: A Physical Heuristic for Life and Complexity,” Ecological Psychology, 2008, 20: 148–179.
(10) This proposal has another virtue—it suggests a way of understanding the unification of science by means of the integration of ontologically equivalent phenomena across all hierarchical levels, as opposed to the reduction of ontologically deficient higher-level phenomena to a single, lowest level that alone is considered to be really real. On this issue, see Robert B. Laughlin, et al., “The Middle Way,” Proceedings of the National Academy of Sciences, 2000, 97: 32–37. It also raises the question of the best way to understand the controversial concept of “emergence,” a vast subject I have no space to go into here. However, see Margaret Morrison, “Emergence, Reduction, and Theoretical Principles: Rethinking Fundamentalism,” Philosophy of Science, 2006, 73: 876-887.