In this section, Prof. Ellen Hildreth shares how she and her colleagues help students take an interdisciplinary approach to thinking about computational modeling.
A key aim of the Brains, Minds, and Machines summer course that we try to capture in the graduate seminar is to help students learn how to think about the integration of computational modeling with empirical studies in neuroscience and cognitive science. In the Fall 2015 course, students experienced this interdisciplinary thinking in several ways.
Lectures combined work from all three fields, presented by instructors and guest speakers who bring different areas of expertise to intelligence research. Multiple instructors were always present and some classes were team-taught by faculty from different disciplines, fostering valuable intellectual exchanges. The course, for example, featured a lively discussion between Professors Josh Tenenbaum and Elizabeth Spelke, and friendly debate between Professor Laura Schulz and Postdoctoral Associate Tomer Ullman, both emphasizing critical thinking about the implications of empirical observations and models for the development of intelligence from infancy.
"Multiple instructors were always present and some classes were team-taught by faculty from different disciplines, fostering valuable intellectual exchanges."
For the recitation classes, students worked in pairs to prepare a presentation of the weekly readings and lead a class discussion in advance of the lecture on each topic. Partners from different backgrounds learned from one another.
Weekly readings drew from multiple disciplines, and writing response questions and extended problems emphasized connections between these perspectives. Weekly writing assignments, which were due the night before recitation, also asked students to submit their own question for the class discussion. The varied nature of these questions wonderfully reflected the diversity of student backgrounds. These questions were shared with the student recitation leaders and with faculty presenting the lecture that followed. This motivated students to formulate thoughtful questions and encouraged broad class discussions.
Finally, students were encouraged to pursue final projects that connect computational and empirical work on a topic related to the course content. Projects could include an implementation component involving simulation and analysis of a model, or an empirical component involving the design of a behavioral experiment with some preliminary results. Alternatively, students could read a set of papers that addressed computational and empirical aspects of a selected problem, discuss connections between the work, and provide a critique and suggestions for further studies.
