Language processing: An introduction to the experimental investigation of language, above the word level
Instructors: Ted Gibson (egibson@mit.edu), Evelina Fedorenko (evelina9@mit.edu)
Location: 46-3015
Currently scheduled: Mondays 2-5. We already know of some conflicts with students who would like to take this class. So we may re-schedule the class to 4-7 or 6-9 on Mondays, or possibly some other time slot, depending on the schedules of the students who want to take the class.
This course has two goals: (1) to teach students about experimental design and basic results in language processing; and (2) to tutor students through the design and execution of an experiment of their choosing, in a language research area above the word level (e.g., syntax, semantics, pragmatics, discourse, working memory).
This is a project-oriented class. Because of the time needed to work on each project, registration is limited in this class to 10 people. If more than 10 people sign up, we will form groups so that the total number of projects does not exceed 10. We will meet with students individually (or in groups) early in the semester in order to decide on a topic area for an experiment to be run during the term. During the semester, students will design, run and analyze at least one psycholinguistic experiment. A paper presenting the study will be due at the end of the semester.
The course will meet once a week for three hours at a time. The lectures will survey some critical results from the field of sentence processing. Throughout the course we will emphasize quantitative methods for investigating language. We will also discuss how to design experimental materials to evaluate hypotheses (including basic statistics, using the language R) from all areas of language, controlling for factors not relevant to the hypothesis in question. Some later lectures will be devoted to discussing students’ experiments.
Students taking the course may come with their own hypotheses to evaluate. Alternatively, we have suggestions for projects in different areas of language. One requirement of any proposed experiment is feasibility. Consequently, most experiments will need to evaluate a question using English participants, because of the availability of this group locally. Proposed experiments on other languages are possible, but only if the experimenter can demonstrate feasibility of getting access to the relevant participant group during the term.