|
Interdisciplinary cognitive science analyses human communication in terms of how people deal with information--when they talk and listen, read and write, and interpret other media. It uses concepts derived from computation, and also contributes to the design of information technology. It combines insights from artificial intelligence, linguistics, logic, philosophy and psychology. An outline of the relations between these non-school subjects is provided.
The first part of the course examines in depth two examples which have been extensively in the psychology laboratory as failures of human ratioanality. We choose these examples for what they show us of the different perspectives of logic, psychology, linguistics and anthropology in understanding the mind. We argue that the observations do more to show how these disciplines have failed to communicate than to show how their subjects are irrational. So these examples illustrate why cognitive science is necessary.
Contrary to what you might think, the language that you use is systematic: it can be described in terms of rules and general principles. We investigate its dual nature, considering both its physical, external forms--the words we speak, hear, read or write--and the less observable--the thoughts or actions those words inspire.
We will examine the nature of meaning in language and investigate how we can build characterisations of the meanings of simple sentences in systematic ways. We look at how words are grouped into larger units, and how sentences connect together to form a conversation or text. We will examine the challenges that face us in trying to develop a scientific model of the relationship between language and meaning. Ambiguity is perhaps the aspect of human language that is most puzzling and intriguing.
We examine the extent to which the models described above are psychologically accurate. For example, when do people detect ambiguity in what they hear? Does it pattern with the ambiguity detected in the model of language developed in previous parts of the course?
Explicit models of language allow the development of a range of technology for the computer-based manipulation of human language. We will consider examples of such systems, examining a range of methodologies that people use, and a range of applications, such as the retrieval of information from databases, and automatic translation from English into French.
Speakers leave things they want to convey unsaid, and hearers infer more from an utterance than what is actually said. This contrasts powerfully with the explicit nature of communication with machines. We will examine how people use unspoken assumptions when communicating. We will show that, perhaps surprisingly, some aspects of the way unspoken assumptions are used in human conversation can be modelled in a formal, systematic way. On the other hand, some aspects of this process present serious challenges for formal modelling, and to the development of computer programs for interpreting language. In many cases, the data you will analyse will be drawn from your subjective experience in a number of tasks.
Communication is not just about passing on information, but also about learning new interpretations and concepts. This section of the course examines the relationship between human interpretation and reasoning and logical frameworks for analysing them. It looks at cognitive changes that happen when students learn to formalise their own natural languages.
Human beings continually create systems for communicating in novel media such as technical languages, diagrams and interactive computer programs. We'll examine how Cognitive science provides methods for analysing these systems, and observing and predicting their effects in practice.
The course ends by presenting and assessing some arguments that computational accounts of cognition are fundamentally incoherent. We think these critiques are valuable ways of sharpening your understanding of cognitive science, but that they are themselves flawed critiques--now you should be in a position to make your own judgements.
|
Informatics Forum, 10 Crichton Street, Edinburgh, EH8 9AB, Scotland, UK
Tel: +44 131 651 5661, Fax: +44 131 651 1426, E-mail: school-office@inf.ed.ac.uk Please contact our webadmin with any comments or corrections. Logging and Cookies Unless explicitly stated otherwise, all material is copyright © The University of Edinburgh |
