Alexander, J McKenzie (2013) Learning to signal in a dynamic world. British journal for the philosophy of science . ISSN 0007-0882 (In Press)
Abstract
Sender-receiver games, first introduced by David Lewis in Convention, have received increased attention in recent years as a formal model for the emergence of communication. Skyrms (2010) showed that simple models of reinforcement learning often succeed in forming efficient, albeit not necessarily minimal, signalling systems for a large family of games. Later, Alexander et al. (2011) showed that reinforcement learning, combined with forgetting, frequently produced both efficient and minimal signalling systems. In this paper I define a dynamic sender-receiver game in which the state-action pairs are not held constant over time, and show that neither of these two models of learning learn to signal in this environment. However, a model of reinforcement learning with discounting of the past does learn to signal; it also gives rise to the phenomenon of linguistic drift.
| Item Type: | Article |
|---|---|
| Official URL: | http://bjps.oxfordjournals.org/ |
| Additional Information: | © The Author |
| Library of Congress subject classification: | B Philosophy. Psychology. Religion > B Philosophy (General) |
| Sets: | Departments > Philosophy, Logic and Scientific Method |
| Rights: | http://www.lse.ac.uk/library/rights/LSERO.htm |
| URL: | http://eprints.lse.ac.uk/45282/ |
Actions (login required)
 |
Record administration - authorised staff only |
