Complexity Systems Simulation Seminar Series (CS^4):Language structure is a trade-off between compression and expression Event

Language is unique among the communication systems of the natural world in that it is structured: we convey complex meanings by constructing complex utterances in a rule-governed manner. While structure in language is sometimes explained as a consequence of biological evolution of the human capacity for language, I will present a mix of experimental and simulation data which shows that linguistic structure arises as a result of cultural evolution. Languages change as a consequence of being learned and used in populations, in response to pressures for compressibility (arising from language learning) and expressivity (arising from language use). Pressure for compressibility but not expressivity leads to degenerate languages which are easy to learn but not useful for communication. Pressure for expressivity but not compressibility leads to languages which are communicatively functional but unstructured and unlike human language. Only when both pressures are at play do we see structure emerge: structured languages are compressible and yet still communicatively functional. Importantly, the extent to which languages need to be compressible and expressive depends on the structure of the populations they are transmitted and used in: we should expect to see differences in language structure which are modulated by differences in the social structure of human populations, which would be unexpected if language structure were a reflection of universal features of human biology."

All CS4 talks are free and refreshments will be provided from 5pm. No registration is required. For videos of previous talks and details of future talks please visit: http://cs4southampton.wordpress.com

The CS4 seminar series 2012/13 will take place fortnightly on Wednesday afternoons at 4-5pm, building 53 room 4025 and will be pitched for a general academic audience from a wide variety of backgrounds.

In this seminar series, leading researchers will explore the challenges related to a particular area of complex science (e.g., climate science, market behaviour, molecular dynamics, neuroscience, networks science etc.) and/or a piece of work addressing these challenges through modelling, empirical work, philosophy or analysis.

Each seminar comprises of a 45 minute talk and will be followed by a discussion. There will be additional time to network pre and post each seminar.

Seminars are free and open to everyone, with no registration required. However to help us manage numbers, if you would like to attend one of our seminar series please can you inform Alison Simmance at: a.simmance@southampton.ac.uk .