Melbourne Linguistics in the pub, Tuesday 1st September 2015
When: 6:00 - 8:00 pm, Tuesday 1st September
Where: Function room (towards the rear, on the left), Naughton's Parkville hotel (NOTE CHANGE OF VENUE DUE TO RENOVATIONS), 43 Royal Parade, Parkville Melbourne Victoria (03) 9347 2255
Roger Blench (Kay Williamson Educational Foundation, Cambridge) will be leading a discussion on this topic:
'New mathematical methods' in linguistics constitute the greatest intellectual fraud in the discipline since Chomsky (click title to see slides from seminar on this topic)
Historical linguistics, and increasingly other areas of linguistics, have been shaken up by the application of 'new mathematical methods' based on Bayesian phylogenies and neighbour net approaches. These papers are usually published in 'hard science' journals rather than linguistics outlets and claim to put historical and phylogenetic linguistics on a scientific basis. However, despite being 'not lexicostatistics' they remain founded on lexical cognacy. The presentation argues that the application of these methods to historical linguistics does not in fact pass any usual test for 'science' and that therefore these claims are overblown and essentially do not advance knowledge. The focus of the argument is on two issues, the problems with cognacy, in particular establishing the difference between borrowing and inheritance, and the assertion that we can date the early diversification of language families.
The presentation does not in any way claim that these publications are intentional chicanery; the authors are evidently serious, merely that they have been misled, as have the editors of 'Science'. Equally it does not claim there are defects in Bayesian statistics; this is all about their inappropriate application.
Check RNLD MLIP page for background reading (to be uploaded shortly): http://www.rnld.org/mlip
Food and drinks available at the venue