MoN13: Thirteenth Mathematics of Networks meeting, Wednesday 10th September, Imperial

Dario Papavassiliou (Warwick) The evolutionary network of the Germanic languages

The concept of linguistic evolution far predates the now more familiar paradigm of biological evolution. By the time of Darwin the idea of languages undergoing birth, change, splitting into separate dialects and eventually distinct languages, and death, had achieved widespread enough credence that he was able to cite it as an illustrative example of the evolutionary process. As in biological evolution, binary trees provide a useful and informative approximate representation of the historical relationships between existing languages, ignoring the complications of lateral transfer of linguistic features (analogous to hybridisation). Here we investigate several statistical techniques (minimal spanning trees; maximum parsimony trees; and simulated history using Markov chain Monte Carlo) for inference of the phylogeny of languages, using as our data set the early Germanic languages (Anglian, Kentish and West Saxon dialects of Old English; Old Frisian; Old Saxon; Old High German; Old Norse; and Gothic) as documented by H. F. Nielsen. It is common in the literature, both mathematical and linguistic, to base such analyses on lexical data, using Swadesh lists. We instead use morphological and phonological data and compare our results with those obtained by lexical comparison. The phylogeny is tested for robustness by filtering data by different grammatical categories. Results from both methods are consistent with one another and robust against different choices of distance metric, provided that enough traits are present in the filtered data set, and show a good agreement with the accepted history of Germanic languages. By comparison to known history of migrations of Germanic peoples, we find that phonological traits are more sensitive than morphological traits to later contact, i.e. following separation and divergence of languages. This may provide a means of distinguishing the influence of inheritance and lateral transfer on the development of a language.

Return to previous page

Contact: Keith Briggs (mailto:keith.briggs_at_bt_dot_com) or Richard G. Clegg (