SOPhiA 2019

Salzburgiense Concilium Omnibus Philosophis Analyticis

SOPhiA ToolsDE-pageEN-page

Programme - Talk

Memory and Landauer's Principle
(Philosophy of Science, English)

Why do we know more about the past than the future? One natural explanation of this ''knowledge asymmetry'' is the fact that we have records of the past but not the future. Being obviously physical in nature, we can expect the asymmetry of records to be grounded in a yet more fundamental time-asymmetry (as the adage goes, ''no asymmetry in, no asymmetry out''). The Second Law of Thermodynamics states that the entropy of the universe increases following any process. Since Boltzmann, various attempts have been made ground the knowledge asymmetry in this ''thermodynamic asymmetry''. Today, the belief that this has met success is a received view in physics and a popular view in philosophy. In this talk I examine a salient account in this vein that appeals to Landauer's Principle.



Landauer's Principle consists of two claims. The first is that any realistic computer (of which humans are a biological example), if it is not to simply be a growing catalogue of information, must implement logical erasure. The second claim is that that logical erasure necessarily invokes entropic increase in the computer's surroundings. Hence, Landauer's Principle is used to explain the knowledge asymmetry as follows. Our experience of a knowledge asymmetry is a product of our brain's computations, and since these computations necessarily align with the thermodynamic asymmetry, the latter therefore grounds the knowledge asymmetry.



This account, however, faces two obstacles. First, since it only characterises computational systems, it does not explain why many external records (fossils, photos, etc.), which do not constitute computational systems, tell us so much more about the past than the future. Second, and more seriously, this account only manages to align the thermodynamic asymmetry with an asymmetry inherent to computation. It does not, however, establish that the latter must align with a knowledge asymmetry. As it stands, takes it for granted that the information written on a computer is more informative of the world's lower-entropy states (which lie in the past) than its higher entropy states (which lie in the future), and therefore begs the question it purports to answer. Whatever the merits of Landauer's Principle, it currently does not explain the knowledge asymmetry.

Chair:
Time: 17:00-17:30, 18 September 2019 (Wednesday)
Location: SR 1.005

Athamos Stradis 
(King's College London, United Kingdom)

I am in the fourth year of the PhD programme in Philosophy and King's College London. My background is in astrophysics, which I studied at UCL, during which time I became interested in philosophy through the study of quantum mechanics, statistical mechanics, and cosmology. My current research is in philosophy of physics, philosophy of science, and metaphysics.

Testability and Meaning deco