[Puzzles in language change] Nov 4: Blythe&Croft 2012

[Igor Yanovich]

Dear everyone, 

We meet tomorrow (Nov 4) to discuss Blythe & Croft 2012 (sent in an earlier email). 

The goal of the paper is to check which kinds of evolutionary settings (as we discussed before with relation to Reali & Griffiths's paper) lead to the emergence of an S-curve of change. The methodology is as follows. B&C define several ways in which a linguistic change may proceed - based on, e.g., whether there is a prestige differential between the modelled speakers. Then they check which trajectories linguistic changes under different settings show. Their conclusion is that the only sensible setting where an S-curve emerges is one where the innovative variant is reproduced preferentially, i.e. has a greater fitness. This is in opposition to e.g. cases when an equally fit innovation just spreads through the population by chance, without any functional force driving the change. 

If you have only limited time, but would like to skim some parts of the article, I'd suggest the order Sec 1 -> Sec 4 -> Sec 3. The terminology such as "neutral interactor selection" is not at all intuitive, and, I find, also not very clearly explained in the text. So if you are confused in Sec.3, that's likely not your fault. 

See you tomorrow! 
 -Igor