In Leo Szilard’s 1929 thought experiment, a Maxwell demon manipulates a one-molecule gas to reverse the second law of thermodynamics. The demon must fail, Szilard argued, since there is hidden entropy creation in the demon’s collecting of information. This thought experiment is an inconsistent muddle of improper idealizations. It diverted an already successful literature of exorcism into degenerating speculations about about a connection between thermodynamic entropy and information. These confusions persist today in a voluminous literature. Narrative conventions in a thought (...) experiment allow thought experimenters great latitude in deciding which processes are typical and bear generalization and which can be idealized away as incidental. For its abuse of this latitude and the extent of the harm it caused, this thought experiment merits the distinction of the worst thought experiment in science. (shrink)

. The origin and history of Landauer’s principle is traced through the development of the thermodynamics of computation at IBM from 1959 to 1982. This development was characterized by multiple conceptual shifts: memory came to be seen not as information storage, but as delayed information transmission; information itself was seen not as a disembodied logical entity, but as participating in the physical world; and logical irreversibility was connected with physical, thermodynamic, irreversibility. These conceptual shifts were characterized by an ambivalence opposing (...) strong metaphysical claims to practical considerations. Three sorts of practical considerations are discussed. First, these conceptual shifts engaged materials central to IBM’s business practice. Second, arguments for metaphysical certainties were made with reference to the practical functioning of typical computers. Third, arguments for metaphysical certainties were made in the context of establishing the thermodynamics of information as a sub-discipline of physics. (shrink)

The thermodynamics of computation assumes that computational processes at the molecular level can be brought arbitrarily close to thermodynamic reversibility and that thermodynamic entropy creation is unavoidable only in data erasure or the merging of computational paths, in accord with Landauer’s principle. The no-go result shows that fluctuations preclude completion of thermodynamically reversible processes. Completion can be achieved only by irreversible processes that create thermodynamic entropy in excess of the Landauer limit.

1. Approximations of arbitrarily large but finite systems are often mistaken for infinite idealizations in statistical and thermal physics. The problem is illustrated by thermodynamically reversible processes. They are approximations of processes requiring arbitrarily long, but finite times to complete, not processes requiring an actual infinity of time.2. The present debate over whether phase transitions comprise a failure of reduction is confounded by a confusion of two senses of “level”: the molecular versus the thermodynamic level and the few component versus (...) the many component level. (shrink)

Brownian computers are supposed to illustrate how logically reversible mathematical operations can be computed by physical processes that are thermodynamically reversible or nearly so. In fact, they are thermodynamically irreversible processes that are the analog of an uncontrolled expansion of a gas into a vacuum.

Ladyman, Presnell, and Short proposed a model of the implementation of logical operations by physical processes in order to clarify the exact statement of Landauer's Principle, and then offered a new proof of the latter based on the construction of a thermodynamic cycle, arguing that if Landauer's Principle were false it would be possible to harness a machine that violated it to produce a violation of the second law of thermodynamics. In a recent paper in this journal, John Norton directly (...) challenges the consistency of that proof. In the present paper we defend the proof given by Ladyman et al. against his critique. In particular, contrary to what Norton claims, we argue that the processes used in the proof cannot be used to construct a cycle that enacts erasure in a thermodynamically reversible way, and that he does not show that the processes used in the proof violate the Second Law of Thermodynamics. (shrink)

This paper makes a novel linkage between the multiple-computations theorem in philosophy of mind and Landauer’s principle in physics. The multiple-computations theorem implies that certain physical systems implement simultaneously more than one computation. Landauer’s principle implies that the physical implementation of “logically irreversible” functions is accompanied by minimal entropy increase. We show that the multiple-computations theorem is incompatible with, or at least challenges, the universal validity of Landauer’s principle. To this end we provide accounts of both ideas in terms of (...) low-level fundamental concepts in statistical mechanics, thus providing a deeper understanding of these ideas than their standard formulations given in the high-level terms of thermodynamics and cognitive science. Since Landauer’s principle is pivotal in the attempts to derive the universal validity of the second law of thermodynamics in statistical mechanics, our result entails that the multiple-computations theorem has crucial implications with respect to the second law. Finally, our analysis contributes to the understanding of notions, such as “logical irreversibility,” “entropy increase,” “implementing a computation,” in terms of fundamental physics, and to resolving open questions in the literature of both fields, such as: what could it possibly mean that a certain physical process implements a certain computation. (shrink)

We show that the so-called Multiple-Computations Theorem in cognitive science and philosophy of mind challenges Landauer’s Principle in physics. Since the orthodox wisdom in statistical physics is that Landauer’s Principle is implied by, or is the mechanical equivalent of, the Second Law of thermodynamics, our argument shows that the Multiple-Computations Theorem challenges the universal validity of the Second Law of thermodynamics itself. We construct two examples of computations carried out by one and the same dynamical process with respect to which (...) Landauer’s principle implies contradictory predictions concerning the entropy increase. Our two examples are based on a weak version of the Multiple-Computations Theorem, which is quite uncontroversial, and therefore they amount to a clear refutation of the universal validity of Landauer’s Principle. We consider some responses to this argument that do not attempt to single out one computation over the others, and we show that they do not work. We further consider ways out of the argument by externalist approaches supporting the computational theory of the mind who propose that the interaction of a computing system with the environment is enough to select a single computation over the others. We show on physical grounds that this approach fails too. We then reverse the direction of our challenge and formulate a dilemma for supporters of the computational theory of the mind: they must reject the causal closure of physic; or else they must accept on a priori grounds that Landauer’s Principle and the Second Law of thermodynamics are not universally valid. Finally, we present our version of a type–type mind-brain identity theory called Flat Physicalism, which is based on the paradigm case of statistical mechanics, and we show that it circumvents the challenge from Landauer’s Principle and the Multiple-Computations Theorem and does not fall prey to our dilemma. (shrink)

After more than 60 years, Shannon’s research continues to raise fundamental questions, such as the one formulated by R. Luce, which is still unanswered: “Why is information theory not very applicable to psychological problems, despite apparent similarities of concepts?” On this topic, S. Pinker, one of the foremost defenders of the widespread computational theory of mind, has argued that thought is simply a type of computation, and that the gap between human cognition and computational models may be illusory. In this (...) context, in his latest book, titled Thinking Fast and Slow, D. Kahneman provides further theoretical interpretation by differentiating the two assumed systems of the cognitive functioning of the human mind. He calls them intuition (system 1) determined to be an associative (automatic, fast and perceptual) machine, and reasoning (system 2) required to be voluntary and to operate logical-deductively. In this paper, we propose a mathematical approach inspired by Ausubel’s meaningful learning theory for investigating, from the constructivist perspective, information processing in the working memory of cognizers. Specifically, a thought experiment is performed utilizing the mind of a dual-natured creature known as Maxwell’s demon: a tiny “man–machine” solely equipped with the characteristics of system 1, which prevents it from reasoning. The calculation presented here shows that the Ausubelian learning schema, when inserted into the creature’s memory, leads to a Shannon-Hartley-like model that, in turn, converges exactly to the fundamental thermodynamic principle of computation, known as the Landauer limit. This result indicates that when the system 2 is shut down, both an intelligent being, as well as a binary machine, incur the same minimum energy cost per unit of information (knowledge) processed (acquired), which mathematically shows the computational attribute of the system 1, as Kahneman theorized. This finding links information theory to human psychological features and opens the possibility to experimentally test the computational theory of mind by means of Landauer’s energy cost, which can pave a way toward the conception of a multi-bit reasoning machine. (shrink)

Even though the second law of thermodynamics holds the supreme position among the laws of nature, as stated by many distinguished scientists, notably Eddington and Einstein, its position appears to be also quite peculiar. Given the atomic nature of matter, whose behavior is well described by statistical physics, the second law could not hold unconditionally, but only statistically. It is not an absolute law. As a result of this, in the present paper we try to argue that we have not (...) yet any truly cogent argument to exclude its possible macroscopic violation. Even Landauer's information-theoretic principle seems to fall short of the initial expectations of being the fundamental ‘physical’ reason of all Maxwell's demons failure. Here we propose a modified Szilard engine which operates without any steps in the process resembling the creation or destruction of information. We argue that the information-based exorcisms must be wrong, or at the very least superfluous, and that the real physical reason why such engines cannot work lies in the ubiquity of thermal fluctuations.We see in the above peculiar features the main motivation and rationale for pursuing exploratory research to challenge the second law, which is still ongoing and probably richer than ever. A quite thorough description of some of these challenges is also given. (shrink)

In this paper I aim to demonstrate that Rudolf Carnap's analysis of the application of information theory within physics, an intellectual-historical precedent of current philosophical criticisms toward this tendency, is justified. First, Carnap and Bar-Hillel (1952) underlined the unjustified ‘semantification’ of Shannon entropy Furthermore, Carnap criticized the ‘physicalization’ of Shannon entropy, but that criticism was not accepted by the physics community of the 1950s (Köhler 2001). Finally, in the posthumously published "Two Essays on Entropy" Carnap (1977) developed a critical assessment (...) of entropy concepts that showed some deep conceptual and interpretative deficiencies in Jaynes (1957a) and Brillouin’s (1956) informational approaches to thermophysics. (shrink)

These preprints were automatically compiled into a PDF from the collection of papers deposited in PhilSci-Archive in conjunction with the PSA 2012.

We have a conundrum. The physical basis of information is clearly a highly active research area. Yet the power of information theory comes precisely from separating it from the detailed problems of building physical systems to perform information processing tasks. Developments in quantum information over the last two decades seem to have undermined this separation, leading to suggestions that information is itself a physical entity and must be part of our physical theories, with resource-cost implications. We will consider a variety (...) of ways in which physics seems to a affect computation, but will ultimately argue to the contrary: rejecting the claims that information is physical provides a better basis for understanding the fertile relationship between information theory and physics. instead, we will argue that the physical resource costs of information processing are to be understood through the need to consider physically embodied agents for whom information processing tasks are performed. Doing so sheds light on what it takes for something to be implementing a computational or information processing task of a given kind. (shrink)

Information theory presupposes the notion of an epistemic agent, such as a scientist or an idealized human. Despite that, information theory is increasingly invoked by physicists concerned with fundamental physics, physics at very high energies, or generally with the physics of situations in which even idealized epistemic agents cannot exist. In this paper, I shall try to determine the extent to which the application of information theory in those contexts is legitimate. I will illustrate my considerations using the case of (...) black hole thermodynamics and Bekenstein's celebrated argument for his formula for the entropy of black holes. This example is particularly pertinent to the theme of the present collection because it is widely accepted as `empirical data' in notoriously empirically deprived quantum gravity, even though the laws of black hole thermodynamics have so far evaded direct empirical confirmation. (shrink)

Equilibrium states are used as limit states to define thermodynamically reversible processes. When these processes are implemented in statistical physics, these limit states become unstable and can change with time, due to thermal fluctuations. For macroscopic systems, the changes are insignificant on ordinary time scales and what little there is can be suppressed by macroscopically negligible, entropy-creating dissipation. For systems of molecular sizes, the changes are large on short time scales and can only sometimes be suppressed with significant entropy-creating dissipation. (...) As a result, at molecular scales, thermodynamically reversible processes are impossible in principle, even as approximations, when we account for all sources of dissipation. (shrink)