Eric Raimy Colloquium

November 19
3:30 - 5:00 pm
University of Wisconsin at Madison

For the most part, the current theory for linear order in linguistic structures is 'left to right' graphic presentation for both syntax and phonology. I will trace a rough history of linear order in syntax (Chomsky 1965 to Berwick and Chomsky 2011) and phonology (SPE to Raimy 2000) to demonstrate the lack of development of this topic. Considering how different data structures support ease of computation for certain types of phenomena that may or may not occur in natural language, I will propose a general theory of precedence for linguistic structures. This theory of precedence ends up matching general observations about how natural language works but adds a new perspective about what are 'economical' processes. Demonstrations of simple apps that serialize structures and learn reduplication patterns will be presented. These formalizations create new questions for computation and representation in language. Having a better understanding of the data structures involved in language makes it easier to identify them through cognitive neuroscience and this particular proposal aligns closely with Arbib's (2005) solution to Lashley's (1951) 'The problem of serial order in behavior' and Hasslemo's (2012) work on epistemic memory in the hippocampus.