Learning Inconsistent Systems and Language Change – R. Alex Schumacher, Northwestern University

Natural languages are highly structured, but they are never perfectly regular. In language contact situations, learners face the challenge of learning inconsistent input. Inconsistent input is a major factor in language change, like in the change of pidgin to creole. In a series of experiments, we investigated how inconsistent systems are learned, namely whether they are made more regular or learned as inconsistent, paying particular attention to the role of bias and individual differences. We taught participants an artificial language with an unexpected type of number marking (Singulative marking) mixed with a familiar system (Plural marking), using a novel paradigm, unscaled adaptive tracking. Post-training evaluation revealed that learners made systems that had more of the Plural system more regular, and systems that had more of the Singulative system were highly variable, expect when the input was already completely regular. This finding shows that individual learners bring potent biases to the task of learning in consistent systems, which determines whether they will make a system more regular or not. Consequently, languages may be changed based on what types of system are mixed, and what expectations individuals bring to the task.

Bio: Alex is a researcher for the Wordovators project. He received a B.A in Linguistics from the University of Iowa, and is currently a Ph.D in Linguistics at Northwestern University

– See more at: http://www.oerc.ox.ac.uk/languagechange
~Seminar, open to all, coffee and cakes provided,