Maria Avramidou: 'The problem of chaos in philosophy of science'
Abstract: Science typically operates under the assumption that with more research, regular and predictable patterns will be discovered in areas where regularity is not yet evident. When scientists encounter a complicated pattern, they often search for underlying periodicities that may be obscured by random noise.
Chaotic systems are characteristically hard to predict despite being governed by deterministic dynamics. They appear random due to their non-periodicity and exhibit sensitive dependence on initial conditions; a small change in the initial state of a deterministic nonlinear system can result in large differences at a later time. This suggests the critical importance of precise measurements for making accurate predictions. In practice, the reliability of predictions is compromised by three principal factors: i) the inherent inaccuracies in measurements; ii) the limitations in computational capacity; iii) the necessity of approximating continuous dynamics using discrete time series. These factors result in uncertainty about the state of the system, leading to a form of underdetermination of theory by data. Following Li and Yorke's (1975) definition of chaotic systems, I discuss the problem of building and testing faithful models of chaotic systems.
The PoP-grunch (Philosophy of Physics Graduate Lunch) is a weekly informal seminar in which graduate students in Philosophy of Physics present their work in progress.
Philosophy of Physics Graduate Lunch Seminar Convenor: Eleanor March
