Home | Students | Course Content | Homework | Projects |Announcements
Syllabus Jul-Dec 2011
What is mind and how does it construct a reality for us? There is considerable debate on how to approach this fundamental question, reminding one of the story of the six blind men inspecting an elephant. Depending on what background you bring into the question, you see the answer very differently. Disciplines such as psychology, neuroscience, philosophy, artificial intelligence, linguistics etc. have had widely divergent positions. Cognitive science is the name given to a modern attempt to synthesize these views into a (more) coherent whole. If you accept that the mind is located in the brain, and that the brain is a device that identifies a goal, takes some sensory inputs and produces some motor actions, then one view of the mind is that it performs certain computations on the input to decide and execute a certain output. In order to do its work, the brain also creates some internal representations for structures in the world and inside us; the structure of these representations are some of the key aspects that one attempts to study in cognitive science. In this course, we shall be looking at a broad range of psychological data and performing some experiments (on ourselves as well) to try to study the operations of the mind. We shall be looking at some of the neural mechanisms that may be underlying our performance. Finally, we shall be seeking to simulate some of these capabilities on computational systems to understand the internal mechanics of these systems. We shall also consider questions of how an infant learns, and particularly, issues of modeling meaning in language. Finally, we shall consider some philosophical issues about what it means to represent something.
This is a science elective course, so there are no special pre-requisites. however, some background in AI, linguistics, psychology, or neuroscience would be helpful. This course will operate via intensive participation from you the students. in the first five weeks, we shall have overview some of the views from differing disciplines; different groups in the class will discuss the following topics. the material for this part of the course is based on the introductory chapters from the MIT encyclopedia of the cognitive sciences (MITECS) (topic links take you to the intro chapters).On each topic, there will be an overview lecture, and then the student presentations and discussions. Interaction in the class discussions carry partial credit. This will begin in the second week. Subsequently, student groups will choose specific topics which they will explore in greater depth. These may relate to
- Cognitive psychology (experiments revealing computational processes underlying cognition)
- Neuroscience (understanding at the micro-level; wetware)
- Linguistics (a prime window into cognition is through language)
- Computational intelligence (simulation and testing of cognitive models)
- perception : experiments on your friends to see how they "sense" things
- categorization and concepts : prototype theory, objects and events
- space, time and language : spatial and temporal categories
- language : lexical structure, compositionality, and semantics
- developmental learning : you may conduct experiments on babies, or simulate their learning
- evolution of social convention: multi-agent games, speech acts, diachronic processes
An excellent text, with lots of top notch essays covering the many of the topics you may wish to explore in the latter parts of the course. Great for browsing as well...
A pathbreaking collection of essays starting with infants in the first year (they learn abstract concepts like number of animacy); and how these eventually map into structures in language
Concepts may be characterized as regions in some multi-dimensional space. Do these have to be convex?
A fascinating study of cognitive processes in infancy. I consider this work as one of my most influential books from the last ten years.
Starting with plato and wittgenstein, go on to prototype theory and probabilistic models. Does a concept have to be conscious?