Dialectics Workshop, Harvard University, 18 November 1979, Contributed paper

Christopher Caudwell and the Second Law of Thermodynamics

Shaun Lovejoy (physics)
McGill University

One of the many paradoxes of Christopher Caudwell was that he considered himself mainly a poet yet he also attained a solid grasp of the contradictions in modern physical theory. Marxist insight accounted for this wide‑ranging comprehension.

Leaving school at 15, Caudwell worked at journalism, edited an aeronautics magazine, and wrote detective stories, apparently uninterested in politics. Then, at age 27, he became a Marxist, joined the local branch of the British Communist Party, and set for himself no less a program than that outlined by Lenin: "Communism becomes an empty phrase, a mere facade, and the Communist a mere bluffer, if he has not worked over in his consciousness the whole inheritance of human knowledge." In two short years, Caudwell accomplished an incredible amount of this program. His achievements became known only after lie had volunteered for Spain and had died soon hereafter (February 1937) in battle against fascism. The manuscripts left behind showed an unusual breadth of knowledge of realms of aesthetics, literature, anthropology, psychology, philosophy and natural science.

One unfinished manuscript titled "Crisis in Physics" (1939) has received relatively little critical attention from scientists though J. D. Bernal (1954) praised Caudwell for "hitting the nail on the head so many times." The book deals interestingly with important and still largely unresolved questions of physics including the wave‑particle contradiction, quantum mechanics, and general relativity. Here I will examine Caudwell's views on the second law of thermodynamics and the apparent contradiction between entropy and evolution. Caudwell's philosophical analysis of this subject concentrated largely on the role of ideology in physics. I propose here to contrast Caudwell's views with recent and relatively unspeculative explanations of this paradox: the work on non‑equilibrium thermodynamics by Ilya Prigogine. Gregoire Nicolis, Paul Glansdorf and others; and the work of astrophysicist David Layser on the evolution of the universe—explanations that appear to elucidate the principle mechanism that enable physical systems, and indeed the universe as a whole, to produce both information and entropy simultaneously. My purpose is to pose obvious questions concerning the limits that philosophy can impose on scientific research, emphasizing the importance of exposing such ideological influences elsewhere in physics.

Caudwell's views on thermodynamics appear in his discussion of physical reality in all its ramifications. Unfortunately, this part of the book is really in the form of notes which are often repetitive and difficult to follow. Another problem was presented by his language and style. Caudwell is at once intense, fresh and vital. Seething with metaphors. As one critic put it: "Caudwell's style is Caudwell's way of seeing" (E. P. Thompson 1977). Though very original and perhaps truly "dialectical," this style often says or suggests more than it should. In what follows, I have attempted to bring out the more important of his arguments and the more specific of his conclusions on this particular topic.

Caudwell viewed reality as a system of interacting, continually developing and newly emerging domains, domains in the process of becoming, in the dialectical processes of transformation of quantity to quality and vice versa. He regarded physical law as "a feature of natural domain—or more correctly, the specification of a natural domain." As such, physical laws could only follow the transformation of quality into quantity, or the “ingathering of likeness," since the transformation of quantity into quality brought forth a new level, a new domain, and thus contained elements which, by definition, were outside the purview of previously existing laws and gave rise to the existence of new laws. He thus concluded that the part of reality to which physics is applicable is that dealing with the production of likeness, the increase in quantity, or the increase in disorder. In short, the second law of thermodynamics defines the realm of physics, being the most general law of the transformation of quality into quantity, and thereby "it is a physical evolutionary transformation law, and as such, is the foundation of all higher evolutionary processes . . . It explains that, taking the universe as a whole, becoming is a certain universal characteristic which is what we mean by Time as immediately experienced by us in the passage of past, present and future. This universal characteristic is that the present can in no circumstances become the past. Time flows. Newness emerges. All is becoming . . ."

For Caudwell, there was no possible grounds other than ideological for the interpretation of the second law  as the law of the "running down of the universe" or of the "heat death." Quality, newness, order were all created simultaneously but on a new level or domain. To account for the newness, new facts and new laws had to be found, thereby enlarging the scope of physics. But if order appears simultaneously with disorder, through processes which can be incorporated into physics only "after the fact," this new order will provide the basis for producing a new entropy: "Every increase in complexity makes possible an increase in disorder—a well furnished room can be more untidy than a monastery cell. Hence the disorder of entropy is artificially created . . ."

Having sidestepped so far the question of specific mechanisms for producing order, what can Caudwell say about them from the general considerations outlined above? "Energy . . . is the most generalized component of quality. Energy, the quantum, quantity, is the likeness in all quality . . ." Thus, it is not surprising that, at thermodynamic equilibrium, where the production of entropy is at a maximum, the available energy is minimized, corresponding to a minimum production of order.

Similarly, with processes which increase order, we may expect a large flow of energy between new and old domains: "Thus the continual decrease of available energy between particles is matched by an increase in the available energy between systems or domains." Later we can examine Caudwell's anticipation of dissipative structures. Let us first see how close he came to finding specific mechanisms for the production of order.

The law of entropy, he says, is statistical, i.e., it is the law of a large number of particles, stripped of all attributes that could distinguish them. The particles are specified as devoid of individuality, abstracted from the domains to which they belong. However, "this reduction of particles to units excludes their small difference, and this difference may always add and emerge in the effect as a big difference, as accident. This accident is however an aspect of necessity, and means that in spite of physics, as it were, a new domain has been generated." Thus accident, chance, or, as we now conceive it, fluctuation plays a decisive role in the generation of new domains.

Not surprisingly, this is as close as Caudwell came to explaining evolution. His solution is formally correct, demonstrating fairly well that mechanistic materialism, by stripping matter of all qualities except those of numbers and by the reduction of all physical relations to particles with no individual attributes, leads to the notion of a universe running down and to the necessity of regarding biological systems as exceptions to the second law of thermodynamics. Since such mechanistic conclusions are widely used for inappropriate and ideological purposes, it is important to see how Caudwell's concepts have been largely vindicated by the recent work of Prigogine and Layzer.

David Layzer (1975) starts with essentially the same picture of physical reality as Caudwell: an interacting hierarchy of approximately closed systems, each with a certain autonomy but, due to interactions with other domains, not completely self‑determining. Starting at the fundamental level of elementary particles, Layzer seeks to determine the origin of the arrow of time, i.e., the origin of order and disorder. He argues that the arrow cannot come from the microlevel because there the laws of physics appear to be reversible. It must therefore come from special boundary conditions applying to an ensemble of particles whose regularities are in turn determined by higher laws, and so forth. For example, when an open bottle of perfume is placed in a room, the molecules of perfume disperse to fill the available space and do not return to the bottle. Of all the possible arrangements for perfume molecules in the room, that of all molecules being in a small bottle is exceedingly unlikely and may be regarded as a special boundary condition. The direction of time implied by this sequence is called the thermodynamic arrow, in contrast to the historic arrow implied by evolution.

Processes that display the thermodynamic arrow of time convert macroscopic information into microscopic information, e.g., when the perfume molecules disperse in the room, the information of their original confinement to the bottles is lost though, if we had followed the path of each molecule, it would have been converted into microscopic information. The special boundary conditions (the bottle) may be regarded as macroinformation as seen from a subsystem though it is microinformation for a higher domain in which various sets of boundary conditions are possible. The essence of Layzer's idea is this: to account for what's happening on the small scale you have to know what's happening on the larger scale.

Applying these concepts in a regression to the conditions at the beginning of the universe, he argues that the conditions then obtaining were a lack of detailed microscopic or macroscopic information—meaning that the universe was in thermodynamic equilibrium for the first microsecond, according to his model. He can thus refute Harlikar and Hoyle who start with a universe that is in a state of thermodynamic disequilibrium but continually approaches equilibrium and "heat death". Layzer's view is the exact opposite, that the universe started at thermodynamic equilibrium but, because of a Big Bang explosion, if you like, the dynamics pulled the thing apart; the rate of change was so great that thermodynamic equilibrium could not be maintained beyond the first microsecond. In contrast to some more slowly expanding cosmologies, thermal equilibrium between the matter and the radiation fields is never re‑established. A universe at uniform temperature of "heat death" thus never occurs. In this way, the dynamics generate order and information; the universe can't get to the state of maximum entropy because of these dynamics. Thus it is possible for galaxies to form with a certain order initially and the whole process cascades down, leading to conditions of thermal non‑equilibrium such as the Earth with its solar energy gradient. I won't say that this concept is completely accepted but I think it must be basically correct. It is important to note that Layzer has essentially the same idea as Caudwell concerning many different domains and levels of domains.

Of more interest, I think, for understanding the production of quality and order is the Prigogine (1978) discovery of a class of systems which exhibit two radically different kinds of behaviour. In one type of situation, close to thermodynamic equilibrium, there is a tendency to evolve toward a state of maximum disorder. In another situation, far from thermodynamic equilibrium, with a state maintained by a large entropy flow to the environment, the system exhibits what Prigogine terms coherent behaviour. In order to reach a state of lower disorder than it started with, such a coherent system must expel entropy in the environment. In real physical systems, such as chicken eggs, this is observed as a high rate of heat production and dissipation where the opposite would be expected if the system started near thermodynamic equilibrium. Such systems have come to be known as dissipative structures. This dissipation of energy is in accord with Caudwell's philosophical argument that the available energy should be large for evolving systems, in contrast to the minimum available energy state for systems producing disorder at thermodynamic equilibrium.

The analysis of systems having two such different kinds of behaviour, according to Prigogine, requires nonlinear equations of the type that characterize, for example, certain kinetic laws and hydrodynamic systems. These are horrible things to deal with, even on a computer. There is no general theory. Though the Rene Thom (1975) work on catastrophe theory is looking into the kinds of discontinuities that arise in nonlinear equations it is still not a general theory. So far, not enough is known.

Prigogine's basic point is that, as the system gets further from thermodynamic equilibrium and comes near to the coherent domain, it reaches a point where molecular fluctuations can take it across the boundary (or bifurcation point, or discontinuity, or whatever). Near the point of crossover from the realm of thermodynamic disorder to the domain of coherence, molecular fluctuations play a critical role in the behaviour of the system. And these fluctuations are essentially random; they follow different laws than the macroscopic laws obtaining before. As the system goes over into coherent behaviour, it also becomes very dependent on the large scale structures. Most systems near thermodynamic equilibrium can be described pretty well by the normal macroscopic measures such as temperature, volume and pressure. But these are not enough when the system gets near to production of order or quality. Large things such as the size and shape of the vessel, that could be ignored before, now become critically important. There is a dialectical interaction between the higher and lower domains that was previously not significant.

Thus the system can evolve deterministically under control of macroscopic variables between bifurcation points but once near a bifurcation, random molecular fluctuations and large scale boundary conditions are decisive in determining its state. New laws are now required that describe the nature of the boundary conditions and of the fluctuations. Often, as in the case of hydrodynamic theory, it is impossible to determine the evolution of the system in any detail; only the statistical properties of the states can be estimated. The dialectics of chance, i.e., the fluctuations, thus play an essential role in the necessity of producing order.

We see that Prigogine is in close agreement with Caudwell's dialectical interpretation on most important points. First, the strong dissipation of energy by systems producing order or quality contrasts with the minimum dissipation of energy and production of disorder by systems near thermodynamic equilibrium. Caudwell predicted this on the basis that energy is the most general component of quality and of producing newness. Second, a system or domain can be relatively self‑determined in the production of disorder, i.e., transforming quality into quantity. In the production of order, however, the behaviour of the system is critically influenced by the properties of the larger‑scale domain and of the sub‑scale domain, the latter depending in a fundamental way on events that are random from the point of view of the system itself. Caudwell described this as the effect of now‑no‑longer negligible differences between particles, which add up to a decisive role though appearing as accident in the larger domain. Thus, the new quality emerges on a new domain, giving rise to qualitatively new laws not reducible to the old laws. Caudwell stressed the need for physics to accommodate new laws in order to be able to describe the evolution of the new domain. Finally, since order can be created far from equilibrium via dissipative structures which expel entropy into the environment, order is produced simultaneously with disorder, in agreement with rather than in violation of the second law. This dialectical relationship between order and disorder, quantity and quality, was fundamental to Caudwell's entire way of thinking.

To conclude, I think that Caudwell provides a very useful model for the use of philosophy by the radical scientist in combatting professional obscurantism, false interpretations and reactionary ideologization of science. He shows us that, even when the detailed physical mechanism underlying a physical law or phenomenon remains unknown, it is possible to defend the materialist viewpoint successfully. There is no excuse for refraining from the battle, nor from mastering Marxism as a tool for the struggles within science.

I hope I have also shown that Caudwell's work in science is somewhat more interesting and significant than has been generally recognized and thereby have helped stimulate new appraisals of his work on other contemporary issues of physics.


Bernal, J.D. 1954 Science in History. Watts, London. More easily available in second edition MIT Press 1971.

Caudwell, Christopher 1939 The Crisis in Physics. Dodd, Mead, New York. The first part has been reprinted in The Concept of Freedom. Lawrence & Wishart, London 1965.

Layzer, David 1975 Scientific American December issue.

Prigogine, Ilya 1978 Time, Structure, and Fluctuations. Science 201: 777-785. His Nobel Lecture.

Thom, Rene 1975 Structural Stability and Morphogenesis. Benjamin, Reading.

Thompson, E. P. 1977 Socialist Register, MR Books.


Joseph Alper. (U MASS Boston) The question of how living things on earth can create order out of chaos has a simple answer: we don't live in an isolated system; we're maintained by the sun and the sun has a huge amount of energy. The questions about heat death are totally academic because the sun is going to last another five billion years.

Either entropy or energy can be considered the fundamental variable [of a system]. It's interesting that energy is always treated as the fundamental variable . . . because energy is what's basic to the capitalist system.

On the other hand, for living systems, for structure an organization, entropy is always the interesting quantity. But we never talk about it. Entropy is this mysterious philosophical thing you get in studying the origin of the universe, the Big Bang. But entropy has nothing to do with that. The principles of the increase of entropy in biochemical reactions, in life, are quite well understood.

They're not far from equilibrium. The temperatures and other gradients are very small. Thermodynamic fluctuations have essentially little to do with this. I think it's all needless sophistication.

Lovejoy. I agree with certain of your points. But it's not enough to say simply that the earth is not in isolation, not in thermodynamic equilibrium. For example, the principle of minimum available energy can explain how crystals form through a conflict between energy and order. The point is to explain how you get biology. Without qualitatively new laws to explain it, you have zero probability.

Alper. It's been brought up here that by using dialectics you can have a materialist explanation for the origin of life. That's what Oparin did. He showed it without introducing new laws, just using ordinary chemistry and realizing that conditions before life are not the same as conditions after life [has come into existence].

Jonathan King. (MIT) The physicists leave out the fact that the origin, of life is an event in the history of the physical universe. Trying to understand what's going on in the universe, the physicists also leave out the fact that society exists and that society transforms nature. People come up with new ideas . . . for tapping the rotational energy by a planetary pipe, for moving planetary life to another place. We have to understand that spreading human society to other planets will transform the universe in ways not foreseen by conventional physicists who separate life from the rest of the universe (as some biologists also do).

David Schwartzman. (Howard University) I think the question of the origin of life is still open. Not that it's unknown; Oparin showed how chemical evolution could proceed under plausible conditions for the emergence of living systems on earth. But I think it's wrong to say that irreversible thermodynamics could not contribute to our further understanding.

Lester Talkington. (Science and Nature) The law of thermodynamics is certainly valid for a closed system, but we don't know the whole open system of the universe. Any kind of formulation that predicts a specific end to the universe has to be based largely on ignorance. We need to go ahead investigating the mechanisms so we can know more.

Lovejoy. Yes, and, in the meantime it's comforting to know that there is specific evidence as well as philosophical conclusions to show that the heat death prediction is wrong.

Science as a Social and Historical Process

It should be noted that there is a difference between universal labour and co‑operative labour . . . Universal labour is scientific labour, such as discoveries and inventions. This labour is conditioned on the co‑operation of living fellow‑beings and on the labours of those who have gone before. Co-operative labour, on the other hand, is a direct co‑operation of living individuals. — Karl Marx, Capital, III, 124.

SOURCE: Lovejoy, Shaun. "Christopher Caudwell and the Second Law of Thermodynamics," Science and Nature, no. 2 (1979), pp. 24-30.

Science and Nature, Table of Contents, issues #1-10 (1978-1989)

Christopher Caudwell: Selected Bibliography

Positivism vs Life Philosophy (Lebensphilosophie) Study Guide

Marx and Marxism Web Guide

Home Page | Site Map | What's New | Coming Attractions | Book News
Bibliography | Mini-Bibliographies | Study Guides | Special Sections
My Writings | Other Authors' Texts | Philosophical Quotations
Blogs | Images & Sounds | External Links

CONTACT Ralph Dumain

Uploaded 9 March 2007

Site ©1999-2018 Ralph Dumain