The relationships between language variation and social structure within an urban community are far too complex to be adequately handled by traditional methods. The Tyneside Linguistic Survey offers a radical approach to sociolinguistic modelling, using multi-variate techniques. Linguistic and social clusters are discovered within a sample of Tyneside speakers by generating classifications based on a wide range of linguistic and social variables. In this study Dr. Jones-Sargent lucidly presents the background and philosophy of the TLS, the implementation of the computer model and an analysis of the results.