In his classic work “The Nature of Thermodynamics” (1943), P. W. Bridgman described thermodynamics as a branch of physics that carries a distinct human element. His operational approach to the philosophy of science emphasises grounding physical theories in human activities, such as controlled experiments conducted by scientists. Classical thermodynamics highlights this aspect by focusing on machines that generate work and chemical reactions that yield products, as well as various phenomena resulting from interventions, such as mixing.
When examining the connection between thermodynamics and the economy, this operational perspective is particularly evident in its application within engineering. In another influential work, “The Entropy Law and the Economic Process” (1971), Georgescu-Roegen pointed out that economics has largely ignored thermodynamics, which paved the way for the emergence of ecological economics. He argued that both energy and entropy are often overlooked in economic discussions, particularly emphasising entropy as the forgotten concept. While energy is a critical input in production and is frequently considered in discussions of economic growth, thermodynamics does not form the foundation of production functions in economics. This trend persists to this day: energy remains a significant topic in economic growth discourse, while entropy is seldom mentioned.
This pattern traces back to Victorian scientific and social thought, particularly regarding thermodynamics—the science of steam engines and industrialisation (see my related post). During this time, thermodynamics was primarily focused on concepts such as work, efficiency, and waste. In what I refer to as the first stage of developing a pragmatist thermodynamics, the idea of efficiency gained considerable influence not only in science but also in everyday economic life, merging with the social perspective of productivism. In this framework, energy is central, while entropy is viewed merely as waste in the form of unusable energy, often represented as “heat.”
In a recent paper, Charles Lineweaver argues that these perspectives overlook the importance of entropy in open, non-equilibrium evolving systems, which he refers to as far-from-equilibrium dissipative systems. These systems encompass all types of environments in which artefacts like machines exist, including both the technosphere and the biosphere. I would compare the distinction between classical thermodynamics and the thermodynamics of open non-equilibrium systems to the difference between functional and evolutionary biology. In classical thermodynamics, entropy is simply released into the environment, treating the system and its surroundings as a closed entity. In this view, entropy is defined for the combined system and environment, allowing the Second Law of Thermodynamics to apply. In contrast, when examining open systems embedded in their environment, we also consider how that environment is shaped by the system’s evolution, and how this feedback influences system evolution, typically through a set of non-linear mechanisms. Lineweaver emphasises that this distinction resembles the difference between velocity and acceleration. In open evolving systems, the acceleration of entropy increase, or specifically, of the entropy production by the system, is of key importance. This situation reflects a significant epistemic Necker cube effect: In classical thermodynamics, which is pragmatist in nature, the concept of production focuses on the energy expended to generate work. However, in non-equilibrium thermodynamics, the approach takes a counterintuitive turn: systems evolve to produce entropy, with work being merely an intermediate step toward this ultimate goal. As Lineweaver metaphorically puts it, we shift from the idea of “We Eat Food” to “Food Has Produced Us to Eat It.”
Like other researchers exploring the Maximum Entropy approach in relation to human technology and the technosphere, such as Peter Haff, I have systematically elaborated on the consequences of this perspective for economics in my work, *Foundations of Economic Evolution* (2013). From this viewpoint, Pragmatist Classical Thermodynamics places us in a state of epistemic alienation, as we often fail to recognise the true workings of the Second Law of Thermodynamics in our efforts to maximise productivity and minimise waste. Enhancing efficiency directly leads to an increase in entropy production within the context of evolving and competing systems, such as life and technology. This idea was first clearly articulated by Alfred Lotka, who integrated thermodynamics with evolutionary theory. However, there is a precursor to this thought: the scientist and philosopher Charles S. Peirce. As I discuss extensively in my work, I am inspired by Short’s systematic overview of Peirce’s related writings. In this sense, Pragmatist thermodynamics of open evolving systems can also be interpreted as a scientific expression of philosophical pragmatism.
Pragmatist thermodynamics 2.0 acknowledges that human economic actions and engineering tend to maximise entropy production as an unintended consequence of activities aimed at traditional productivity. This perspective inverts the analytical approach and redefines human production as centred on mitigating the adverse environmental impacts of production. At a fundamental level, these can be understood as an increase in entropy, as argued by Georgescu-Roegen in his seminal work. We can conceptualise this as moving from efficiency of production to sustainability of production as criterion for designing the technosphere.
One important consequence of this discussion is the ability to develop a thermodynamic definition and criterion for sustainability. Bernhard Wessling has proposed a thermodynamic approach to sustainability based on two key premises. First, we need to view production as embedded within a broader array of systems. Sustainability, therefore, pertains to this entire system rather than just focusing on the specific aspects of production technology. For instance, while industrial agriculture aims to minimise harmful environmental impacts, it often overlooks the overall biospheric context. In contrast, organic agriculture considers this broader context and cares for the sustainability of the entire array of embedded biospheric systems. Second, we can use a straightforward criterion for quantifying sustainability: between two processes, the one that produces less entropy is considered more sustainable. We can refine this idea using Lineweaver’s approach: a sustainable production process is one that minimises the rate of increase of entropy production in the interaction between the system and its environment, or slows down the trend towards maximum entropy production. In economic terms, this translates to a deceleration of economic growth while still maintaining a positive growth rate. Metaphorically speaking, the car is a sustainable artefact if it approaches minimum acceleration to attain a certain velocity. Obviously, what matters here is the nature of the competition between cars, as Lotka had already argued for life in general: Natural selection tends towards maximising energy throughput by favouring rapid deployment of a certain performance, such as speed. Humans have the capacity to rein in the forces of technological competition and selection, in analogy to the “slow food movement.” A sustainable technosphere is shaped by “slow technology.”
The technosphere, our agency, our planet 