Slaying systems disorders

Early on in the inquiry, I introduced the energy concept specifically as an entailment of considering physical phenomena in systems terms. In the course of doing that, I gave a brief introduction to basic systems ideas, and outlined what is entailed in considering any situation in such terms. We’re now at a point where some further background in systems will be particularly useful, as we start to look at the energy costs of our energy use under the broad theme of efficiency. In providing the earlier general introduction to thinking with systems as a basis for then introducing energy from the very outset explicitly as a property of physical systems, I drew on a set of ideas that could be considered foundational for the broad and heterogeneous field of systems thinking and practice, or simply systems.Note 1 The way in which I introduced those ideas is more-or-less entirely consistent with the ways that systems concepts are used in the physical sciences and engineering disciplines that deal directly with energy, especially the various branches of thermodynamics. In other words, in relation to those introductory systems ideas as set out here at Beyond this Brief Anomaly, the fields of systems and thermodynamics are in close agreement.

It would be nice to say that all questions of commensurability between these fields can therefore be taken as settled, and that it’s a straightforward matter to draw on the more general insights from systems to inform our inquiry into energy and society. Unfortunately, as these things sometimes go, the situation is a little more complicated. It will likely be apparent to readers who’ve been following for a while that I hold the science of thermodynamics, as foundational for understanding the physical behaviour of any situation, in high regard. What I haven’t stated previously is that I also understand this science, as originally set out by its founders, to be essentially consistent with the observed behaviour of all physical phenomena.  This extends from phenomena characterised by interactions that are principally physical in nature (e.g. mechanical, geological, hydrological and atmospheric processes, and also including chemical interactions associated with such processes) to the physical foundations of phenomena that we tend to think of as principally biological, ecological or social in nature. For readers trained in the physical or engineering sciences and without much in the way of prior exposure to the systems sciences, my position on this may seem rather obvious—perhaps more noteworthy for having bothered to highlight it than for the stance itself. That is, for most conventionally trained scientists and engineers, that the first and second laws of thermodynamics as set out by Rudolph Clausius and elaborated by William Thomson in the mid nineteenth century apply to all physical processes, including those associated with biological and social phenomena, is quite uncontroversial. If you place yourself in this camp, then you may be surprised to learn that very different understandings are not only prevalent in the systems field, but in some branches are foundational. As someone trained in engineering thermodynamics who initially formed a self-directed interest in considering engineering practice from a perspective that I subsequently learned was consistent with key insights from the systems field, the views that I encountered there in relation to thermodynamics were not only surprising, but initially also a source of some confusion.

What I found there was that the second law of thermodynamics as stated by Clausius in its classical form is often regarded as inapplicable to the types of systems that are used within the systems field to characterise living organisms. Furthermore—and this is where the confusion stemmed from in particular—I heard that a different form of thermodynamics, namely “irreversible” or “non-equilibrium” thermodynamics, had been developed and was necessary to account for such systems, and hence for living organisms as a category of phenomena. The incommensurability of this view with my existing understanding was, in a word, confronting. Most of my past experience and understanding seemed to not only fit, but in fact to make better sense within an encompassing systems outlook. When it came to this matter though, there was no way to reconcile what I’ll call the engineering view of thermodynamics (though in practice it’s much broader than that) with what I’ll refer to as the systems view (though in practice, it tends to be most prevalent in areas of systems associated with or influenced by Ludwig von Bertalanffy’s general system theory, first-order cybernetics and the complexity sciences).

The resolution of these apparently contradictory views turned out to be rather straightforward, though unfortunately for the branches of the systems field involved, not particularly edifying. The problem comes down to erroneous interpretations of thermodynamic ideas in both their classical (after Clausius and Thomson) and statistical (after Ludwig Boltzmann [2]) forms, that seem to have taken root early on in the development of the systems field and have perpetuated as apparent articles of faith ever since (despite efforts from within the field to address them e.g. see articles by Corning and Kline [3], [4]).

The engineering view

Before pointing out the three major respects in which the systems view goes astray, I’ll set out the engineering view, which is foundational for Beyond this Brief Anomaly. Firstly, the engineering view recognises three basic types of systems:

  1. Isolated: there is no exchange of either matter or energy with the environment;
  2. Closed: there is no exchange of matter with the environment, but energy in the form of work or heat can cross the boundary between system and environment; and
  3. Open: both matter and energy (work or heat) are exchanged between system and environment (momentum is also exchanged along with any matter that crosses the system boundary).

In engineering analysis, an open system is also known as a control volume—a volume of space enclosed by a defined control surface across which matter, heat and work (as well as momentum) can flow. While classical thermodynamics—the original statement of the first and second laws—was first formulated specifically in relation to closed systems, this formulation did not restrict its postulates to such systems. However, when Ludwig Boltzmann later developed the statistical approach to thermodynamic analysis, he was specifically interested in the kinetic behaviour of gases as collections of micro-scale particles, and the focus of his work was far more circumscribed. His statistical formulation of the second law was explicitly developed for systems of fixed mass i.e. for which there is no exchange of matter with the environment.

At this point, a few words about the distinction between these two approaches may help with untangling the problems in the systems view a little later on. Firstly, classical thermodynamics deals with matter in terms of behaviour at the aggregate level, and properties associated with this—it is ambivalent with respect to the micro-scale nature of matter, and indeed to the physical basis for the phenomenon of heat. The aggregate-level behaviour involves emergent phenomena that cannot be reduced to, or defined causally in terms of, the behaviour of a thermodynamic system’s constituent components considered in isolation from one another. If the atomic theory of matter and the kinetic theory of heat were ever overturned, the classical formulation of thermodynamics would remain largely unaffected. Statistical thermodynamics, on the other hand, deals with matter not in terms of behaviour at the aggregate level, but as a collection of particles considered at the micro-scale. The statistical approach allows for handling enormously large numbers of particles without the need to track each one individually. From this micro-scale perspective, there are certain behaviours that remain hidden from view, but that we know from experience occur in practice at the macro-scale or aggregate level. For example, from the micro-scale statistical viewpoint, it is in principle possible for a gas, having expanded to completely fill a container, to accumulate again at a later time in one part of the container. That this is never observed to occur in practice is not just because it is vanishingly improbable, as is often assumed when considering the gas’s thermodynamic behaviour using statistical methods, but because in practice this would entail an increase in the gas’s pressure that would oppose such a re-accumulation (see for instance Kline’s account of this [5]). The restoring effect that local pressure fluctuations have on any excursions away from the equilibrium condition in which the gas is evenly distributed throughout the entire container only comes into view at the macro-scale when the behaviour of the gas is considered as a whole.

While the original formulations of both classical (macro-scale) and statistical (micro-scale) thermodynamics have in common that they were set down in relation to closed systems, they diverge from one another in terms of their respective applicabilities beyond that original context. The view presented by classical thermodynamics is more general than the original closed system context, and the first and second laws of thermodynamics for closed systems are readily extended to open systems. In practice, this is achieved via the engineering method known as control volume analysis, the techniques for which were completed in the early 1950s. Control volume analysis has been part of the syllabus for thermodynamics courses now for several generations of engineers, and is routinely applied in the design of aircraft, as well as turbo machinery, such as compressors and jet engines. In fact, it is applicable to the analysis of any equipment that involves the continuous through-flow of a working fluid. It’s especially noteworthy that all such devices can be considered as operating far from thermodynamic equilibrium, being maintained in such a state by a continuous energy input in the form of work or heat, the latter very often by combustion of a fuel input within the control volume itself. In all respects, such equipment behaves in a manner that is entirely consistent with the first and second laws of thermodynamics as formulated for and applied to closed systems—or isolated systems for that matter—save for the fact that under stable operating conditions they tend towards a steady-state operation out of equilibrium with their surrounds. The most important point to note in relation to this for our present purposes is that all open systems, whether mechanical artefacts of human design, or biological in nature, are essentially equivalent in explicitly thermodynamic terms. As such, control volume techniques can be readily extended to the analysis of living organisms, as Corning and Kline have proposed [3], [4]. That these techniques have not to date been widely applied in biological or ecological contexts is more a reflection of the difficulties faced in communicating across disciplinary boundaries, than a result of any fundamental difference in the nature of the problems that the different disciplines are dealing with in seeking to understand phenomena particular to their own domains in thermodynamic terms.

The situation with respect to statistical thermodynamics is a little different. While the general principles on which the micro-scale thermodynamic view is based apply equally to isolated, closed and open systems, the statistical methods developed to actually carry out practical analysis are suited only to systems of fixed mass. Moreover, Boltzmann’s original work dealt specifically with the transition of such systems from initial states away from equilibrium, to the equilibrium state in which no further macroscopic changes can occur. Boltzmann characterised such systems in terms of their probabilities of being in a particular state. He described the tendency for a system’s energy to disperse as widely as possible—and hence for the extensive system property entropy to increase to a maximum—in terms of a change from a state with a low probability of occurrence to a state with a high probability of occurrence. He also introduced a particular metaphor—a conceptual device with which to communicate his ideas by invoking for the reader a shared domain of experience—that when restricted to the very specific context he was considering, and at that particular time in history, may well have been a worthwhile source of insight. Via its subsequent decoupling from that context in later accounts though, it has almost certainly made the second law less comprehensible to generations of students in thermodynamics as well as a good proportion of those students who’ve gone on to become experienced science and engineering practitioners. Perhaps more significantly though, it has contributed to a body of secondary literature that presents thermodynamic ideas in ways that bear only passing resemblance to the science purportedly represented, and that has contributed to pervasive distortion of the ways that many people understand the physical processes underpinning our very existence. The device that Boltzmann introduced was the metaphor of disorder—specifically the characterisation of the improbable initial state for a system of gas molecules at some remove from equilibrium as “ordered”, and its probable final state of equilibrium as “disordered”. More extensive exploration of the issue, including just what it was that Boltzmann was characterising in this way, is presented here, in a version of an article originally published in the journal Futures. As we’ll see a little later, this plays a central—though only partial—role in the erroneous thermodynamic views that are so prevalent in the systems sciences.

The systems view (and three disorders)

In presenting the systems view, I’ll draw specifically on general system theory (GST) as articulated by its originator, Ludwig von Bertalanffy [6]. While von Bertalanffy credits earlier influences for his treatment of thermodynamic ideas, I’m looking specifically at his work here due to the prominence given to the erroneous interpretations in GST, and the wide influence that the interpretations presented as part of GST have subsequently had on the systems field. While thermodynamics recognises three types of systems (isolated, closed and open), GST recognises two types: closed and open. GST’s closed systems are equivalent to isolated systems in thermodynamics; GST’s open systems are equivalent to open systems in thermodynamics. The relationships between the use of terminology in thermodynamics and GST are summarised in the table below. It’s also important to note that the convention presented in GST is typical for the systems field in general.

Thermodynamic system types Equivalent system types in GST and the systems field more broadly
Isolated Closed
Closed (No equivalent)
Open Open

the first error

The first interpretive error in GST is that the laws of thermodynamics apply only to what it calls closed systems (isolated systems in thermodynamics). The origin of the misinterpretation isn’t immediately apparent to me, as Clausius very clearly presented the insights behind the first and second laws of thermodynamics in relation to systems involving transfer of heat and work across the system boundary i.e. thermodynamically closed rather than isolated systems. Perhaps the problem arose due to establishing a definition for closed systems in GST, and then assuming without confirmation that prior use of similar terminology must mean the same thing. It strikes me as a little odd though that this confirmation wouldn’t have been carried out in formulating a theory intended as sufficiently general to accommodate existing disciplines such as thermodynamics. One other possibility is that in developing GST, an assumption was made that Boltzmann developed the statistical approach to thermodynamics (in isolated systems) in relation to systems of the same type for which Clausius had previously developed the classical statement of thermodynamics. These are speculations on my part though—answering them would require further research into the origins of GST (which in any case may well have already been carried out by others).

the second error

Beyond the confusion regarding the type of systems in relation to which the laws of thermodynamics were first established, the interpretation on which GST is based involves a more significant error: while Clausius’s statements of the first and second laws were developed specifically in the context of thermodynamically closed systems, this original formulation was not restricted to such systems. There is nothing in the original formulation of classical thermodynamics that suggests it should necessarily be limited in this way.

Just to be clear about von Bertalanffy’s views on this—and those on which GST is based—let’s look at exactly what he wrote:

Conventional physics deals only with closed systems, i.e., systems which are considered to be isolated from their environment…Thermodynamics expressly declares that its laws apply only to closed systems. [6] (p. 38)

classical thermodynamics, by definition, is only concerned with closed systems, which do not exchange matter with environment. In order to deal with open systems, an expansion and generalization was necessary which is known as irreversible thermodynamics.[6] (p. 167)

Here von Bertalanffy is using the term closed system expressly in its systems context i.e. as equivalent to the term isolated system in thermodynamics. The task of interpreting von Bertalanffy in relation to this is not easy though, as in defining what he means by open and closed systems, he refers explicitly only to material inputs and outputs:

We term a system ‘closed’ if no material enters or leaves it; it is called ‘open’ if there is import and export of material. [6] (p. 128)

To appreciate that by “closed system” von Bertalanffy does in fact imply that he has in mind systems for which there is no exchange of matter or energy with the environment, it’s necessary to look at how he discusses the thermodynamics of such systems as by definition involving processes that stop when a state of equilibrium is reached [6] (p. 38). Given that there is no necessity that closed systems in thermodynamic parlance reach equilibrium—heat transfer or work input can maintain them in states away from equilibrium with their surrounds—then the systems von Bertalanffy is discussing must by definition be equivalent to isolated systems in thermodynamics.

The third error

So far, we’ve looked at two specific errors: one relating to the type of thermodynamic systems in relation to which classical thermodynamics was originally formulated; and the other relating to the applicability of classical thermodynamics to thermodynamic contexts beyond those in relation to which it was originally formulated. Each of these errors involves in its own way an apparent misinterpretation of the contextual relevance of the central insights of classical thermodynamics. The third error is more significant again, in that it relates to a distortion of those central insights themselves, by misattributing to them claims that the originators did not make.

The origin of this third error appears to lie in Boltzmann’s use of the metaphor disorder to characterise the equilibrium state for a thermodynamic system. Boltzmann described as disordered the state of a system comprising a quantity of molecules, for which the constituents’ distribution of velocities is greatest i.e. the state amongst all possible system states, for which each molecule has available to it the greatest range of possible velocities. Such a state can be considered the most probable for the system on the basis that for all velocity distributions that the system’s molecules could occupy across every possible state from initial to final, the greatest numbers correspond to the “disordered” state, the state in which the system is in equilibrium and for which no further macro-scale change can occur. Following from this, there is a statistical bias for the system to change in the direction of, and eventually settle in, its “disordered” state, bearing in mind that the term disordered relates specifically to the velocity distribution for a system of molecules. In other words: given that the system must at any time occupy one of its available velocity distributions, the velocity distribution in which it is most likely to be found (the most probable velocity distribution) will correspond with the macro-scale system state that has associated with it the largest available number of such distributions.

Boltzmann’s velocity distribution concept was established specifically in relation to systems comprising collections of gas molecules, and for which the system’s total energy is the sum of the molecules’ kinetic energy. The concept is a precursor to the modern concept of energetic micro-states. An energetic micro-state can be thought of as the particular way in which energy is distributed across a system’s components. The more widely a system’s energy is dispersed across its components, the greater the number of energetic micro-states available to the components for the system’s corresponding macro-state. The key insight to take from this is that when Boltzmann used the terms “ordered” and “disordered”, he did so in relation to the energetic configuration of the systems in which he was interested, not in relation to their material structure. In using this terminology, he was not describing the degree of “orderliness” of such structure, or making observations about the complexity of a system’s material form.

This distinction has been widely overlooked, and appears to undergird the erroneous view in GST that the second law of thermodynamics describes a tendency towards breakdown of structural “order” in a system’s material configuration. That the behaviour of living systems entails increase in structural complexity is then taken as evidence that classical thermodynamics must therefore apply only to systems of a type that is inadequate for characterising self-organising phenomena. Such a view is, however, simply a result of interpreting the second law of thermodynamics as implying something that it does not, and extending that law to the description of phenomena to which it does not—and is not intended to—relate.

In von Bertalanffy’s account of GST we read, for instances, that:

the second principle of thermodynamics indicated destruction of order as the general direction of events. [6] (p. 46)


According to the second principle of thermodynamics, the general direction of physical events is towards states of maximum entropy, probability and molecular disorder, levelling down existing differentiations. In contrast and “violent contradiction” to the second principle (Adams, 1920), living organisms maintain themselves in a fantastically improbable state, preserve their order in spite of continuous irreversible processes and even proceed, in embryonic development and evolution, toward even higher differentiations. This apparent riddle disappears by the consideration that the classic second principle by definition pertains only to closed systems. In open systems with intake of matter rich in high energy, maintenance of a high degree of order and even advancement toward higher order is thermodynamically permitted. [6] (p. 167-8)

None of these attributions to the second law arise from the original literature on thermodynamics, in either its classical or statistical mechanics forms. That the second law relates to “destruction of order”, the apparent “violent contradiction” between the behaviour of living organisms and the second law, and the occurrence of an “apparent riddle” all result from erroneous interpretation of thermodynamic principles, rather than these problems originating within the science of thermodynamics itself.

It is entirely consistent with both the classical and statistical thermodynamic accounts of physical behaviour that work associated with the dispersal of energy from being more concentrated to more spread out enables maintenance, modification and fabrication of new physical structure. In fact, all maintenance, modification and fabrication of all physical structures requires the dispersal of energy from higher to lower concentration—and so the behaviour described by the second law of thermodynamics, without need for recourse to “irreversible” or “non-equilibrium” extensions,  is essential for enabling physical change, whether we consider that change to be either constructive or destructive. Constructive and destructive physical processes are all associated with such energy dispersal. Once again, for further details in relation to this, see the Futures article here.

So what?

Why go to such lengths to clear the air on these matters here? The misunderstandings that are common in the systems field don’t have any immediate consequences for Beyond this Brief Anomaly’s inquiry, so this may all seem rather tangential. I do have a very clear motivation in mind here, though. Taking an explicitly systemic or systems-based approach to inquiry in any area involves stepping beyond what, as a cultural norm, we tend to “do when we do what we do” (to borrow a way of framing such an approach from my colleague Ray Ison [1]). Given the novelty implicit in this systems-based approach for people not previously familiar with it, I’m very much aware that it can initially appear to be confusing, or simply confused, and that its claims for improving on more strictly reductive analysis can appear light on for justification in a conventional scientific sense. In fact, scepticism is especially justified for practitioners in scientific and technical disciplines who have direct experience of the power of reductionism for dealing with practical situations as solvable problems. Given my own background, I have a great deal of sympathy for such scepticism. I’m acutely aware of the implications that inaccuracies or misrepresentations of conventional science within the systems sciences can carry for the credibility of the systems approach. This is particularly so give that, as I understand it, such an approach is a way of thinking and acting that one can endogenously grow towards, or exogenously be invited to expand into, but that involves more than the acquisition of new knowledge within one’s established ways of thinking and acting. On this basis, encountering inaccuracies within the systems domain could be very reasonable justification for not taking up such an invitation. Given that Beyond this Brief Anomaly is in part a form of advocacy for the value of trying out a systemic approach to dealing with the dilemmas of our civilisational situation, the credibility of the inquiry demands that the limits of this approach also be recognised. A systemic approach is no less prone to error than any other. The benefits of taking this path don’t include an inherent reduction in the likelihood of getting things wrong. The inherent benefit of the approach is to deal with things more comprehensively, and following from this, to reduce blunders associated with discounting the limits of our established thinking and practice, rather than to eliminate the possibility of errors altogether. As always, avoiding errors is in large part a matter of the diligence with which any inquiry is conducted—and in important respects, results from being systematic in that inquiry, rather than necessarily systemic.

With systemic inquiry, given that it often roves widely across disciplinary domains in the attempt to be comprehensive, there is perhaps even an increased likelihood of errors where either the inquirer is dealing with domains with which she or he isn’t sufficiently familiar, or where the scope of inquiry is such that those domains are not or cannot be dealt with in sufficient depth. An important response to this from within the systems approach itself is to make inquiry participatory, by encouraging and allowing for as broad a range of perspectives as possible. So for instance, in the present case this could be achieved through the discussion that’s allowed by conducting the inquiry via a blog platform. The possibility that this might eventuate in a more robust way remains open.

In the meantime though, the inquiry proceeds, and in the next post I’ll return to the task proper of considering what will be entailed in exploring the broad theme of energy efficiency from within a systemic worldview.


Note 1 While there are a great many introductions to this field, Ray Ison’s Systems Practice: How to Act in a Climate-Change World [1] stands out for me for the way in which it situates this in a comprehensive account of why the field is practically important.


[1] Ison, Ray. (2010). Systems Practice: How to Act in a Climate-Change World. London: Springer.

[2] Boltzmann, Ludwig. (1964). Lectures on gas theory (S. G. Brush, Trans.). Berkeley: University of California Press.

[3] Corning, Peter, & Kline, Stephen J. (1998a). Thermodynamics, information and life revisited, part I: to be or entropy. Systems Research and Behavioral Science, 15, 273-295.

[4] Corning, Peter, & Kline, Stephen J. (1998b). Thermodynamics, information and life revisited, part II: ‘thermoeconomics’ and ‘control information’. Systems Research and Behavioral Science, 15, 453-482.

[5] Kline, Stephen J. (1999). The low-down on entropy and interpretive thermodynamics. La Cañada: DCW Industries.

[6] von Bertalanffy, Ludwig. (1968). General System Theory: Foundations, Development, Application. New York: George Brazillier.

One thought on “Slaying systems disorders

  1. Pingback: The engineering view of systemic efficiency: available energy | Beyond this Brief Anomaly

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s