What is Cognitive Science? :: Notes: CogSci Week 1 01 - What is Cognitive Science? 02 - What is Cognitive Modelling? 03 - What is a Cognitive Technology? 04 - What is a Language? Week 2 02 - How do we make a case for innateness? 03 - Original Sim 04 - Marr's Levels of Analysis 05 - The Anatomy of a Learning Algorithm 06 - Perceptron 07 - Perceptron: Learning Algorithm 08 - Multilevel Perceptron 09 - Backpropagation More Course Webpage Built with from Grav and Hugo What is Cognitive Science? What is Cognitive Science? Cognitive science is the study of mental representations and processes. A mental representation is a description of information that exists within the mind. A mental process is a procedure for translating information into representations, representations into other representations or representations into action/behavior. Historically, we are well known for making box and arrow diagrams that describe information flow. Let’s build an example. We’ll start with a narrative. Imagine an adventurous child has escaped their high-chair and begins exploring the house, armed with their trusty screwdriver (not sonic). Of course, there is an exhausted parent nearby in the kitchen struggling to supervise the kid while working for home. The parent gets a quick glimpse at the escaped child slowly working his way towards the electric socket. He shouts, ``Who wants ice cream?'' to get the child’s attention. The child turn, smiles and runs back into the kitchen. There’s a lot of information being exchanged here. Let’s dissect it starting with the child: We have physical information in the form of light, being translated into a visual percept, the socket. This percept is then combined with the child’s knowledge of the world, an information gap is identified, what’s the socket do? This information gap is translated into a plan, screwdriver in socket. The plan is put into action. Now for the father, We have physical information in the form of light, being translated into a visual percept, child walking toward socket. That percept is then combined with the parent’s stored knowledge to infer their child’s intention, putting the screwdriver into the socket. Given the child’s intention, the parent construct plans and decides on a plan, distract the kid with ice cream. The fathers plan results in his idea to be translated into language. The language then has to be translated in a motor plan. And back to the child, Sound wave hits their ear drum translated into languages. Language informs a choice, socket or ice cream. The child’s decision becomes a plan, which then becomes an action. What’s cool here is that this simple story illustrates the breadth of cognitive science. Cognitive science is highly interdisciplinary. The translation of light to visual percepts involve neuroscience and psychology. The combination of visual percept and stored knowledge to infer the child’s intention involves anthropology and psychology. The deliberation between plans involves behavioral economics and mathematics. The use of language to enact change in the world involves linguistics and anthropology. The whole endeavor of analysing and characterizing this translation process requires philosophy and computer science. In general, we live rich mental lives, which allow us to create and represent our world, and also the worlds that we inhabit from time to time when we read fiction or watch a movie. The interdisciplinary nature of cognitive science is vital to understanding how we build mental models of the world. The aim of this endeavor is to characterize the nature of human knowledge, and how that knowledge is used processed and acquired. This is the MIT Brain and Cognitive Science party line. In this course, we are going to introduce you to the landscape of cognitive science, focusing on the kinds of questions we ask, the data we collect to answer these questions, the theories we build to based on these data and the computational models we use to implement these theories. A fundamental belief in cognitive science is that computational modelling can be used to evaluate theories, to generate new hypotheses and to guide the collection of new data.
