128141 FS FS Research Seminar I (2018S)

In this seminar we familiarize ourselves with experimental ways of studying language and language change. Our focus will be on experiments in the Iterated Learning paradigm. Iterated Learning experiments work a bit like the game 'Chinese Whispers' (Gmn 'Stille Post'). They are meant to test what properties languages have (or are likely to get) because languages need to be shared and transmitted among speakers and speaker generations in order to do their job in communication. In a typical Iterated Learning experiment, the first 'generation' of participants is made to learn an artificial mini language (constructed for the purpose of testing some specific hypothesis), and then they are asked to use it. Normally, they acquire it imperfectly and this shows in the way they use it. Next, the output of that first generation of participants is used as learning input for a second generation of participants, who are also asked to use it and who usually change it a bit more. Then, their output is taught to a third generation, and the game is repeated. Basically, there is no limit to the number of times the procedure can be repeated, but one usually stops after 6 to 10 generations. Finally, one looks at the ways in which the mini language has changed and asks if one’s initial hypotheses would have predicted the observed changes.
Iterated Learning experiments have proved to be extremely useful in testing hypotheses about language transmission and language change and have thrown new light on the general question why languages are as they are.
In the seminar, participants will design experiments of their own. This involves (a) developing testable hypotheses (e.g. irregular morphology is dispreferred); (b) designing an artificial mini language by which these hypotheses can be tested (How will the language change if the hypothesis is correct?); (c) thinking of an appropriate experimental setup (What will participants be asked to do with the language? Will they use it in interaction, or will they perform tasks individually?). Ideally, experiments will not only be designed, but also (d) tried out (piloted), so we’ll also have the chance to (e) analyse results, and interpret them with regard to our initial hypotheses.
The overall goal of the seminar is therefore twofold. On the one hand, we will gain some expertise in a new method, and on the other we will deepen our understanding of why languages are as they are.