Modelling High-Level Cognitive Processes

Richard P. Cooper
with contributions from Peter G. Yule, John Fox and David W. Glasspool

Lawrence Erlbaum Associates
(published June 2002)

From the preface:

This book is a practical guide to building computational models of high-level cognitive processes and systems. High-level processes are those central cognitive processes involved in thinking, reasoning, planning, and so on. These processes appear to share representational and processing requirements, and it is for this reason that they are considered together in this text.

The book is divided into three parts. Part 1 considers foundational and background issues. Part 2 provides a series of case studies spanning a range of cognitive domains. Part 3 reflects upon issues raised by the case studies. Teachers of cognitive modelling may use material from Part 1 to structure lectures and practical sessions, with chapters in Part 2 forming the basis of in-depth student projects.

All models discussed in this book are developed within the COGENT environment. COGENT provides a graphical interface in which models may be sketched as ``box and arrow'' diagrams. Such diagrams are common in psychological theorising, and COGENT builds on concepts such as functional modularity associated with them. Boxes within box and arrow diagrams may be fleshed-out with rules or configured through properties to obtain appropriate behaviours. COGENT also provides support for specifying experimental tasks consisting of blocks of trials analogous to those carried out within standard experimental psychology. Models may then be evaluated by running them on such tasks and comparing their behaviour with that of human participants.

What the reviewers have said:

"A much-needed book, filling a long-standing gap in the range of textbooks available. Will be welcomed by anyone who teaches cognitive modelling, at a graduate or advanced undergraduate level."
Richard M. Young, University of Hertfordshire, England.

"[...] anyone who does work through the modelling exercises all the way to the end will be very well equipped to go on and do some highly useful work on their own. A palpable hit!"
Aiden Feeney, University of Durham, England.

"... well written ... [Chapter 4] takes one nicely through from specific algorithm - heuristic - general problem-solver."
Tom Ormerod, University of Lancaster, England.

"an impressive feat ... Providing concrete models so that the reader can explore the major concepts underlying Cognitive Science. [...] a wonderful text for an advanced course [for] students interested in the Cognitive Sciences."
Lon Shapiro, Illinois Wesleyan University, Bloomington, IL.

"It's nicely done, and I like the practical exercises."
Gerhard Strube, University of Freiburg, Germany.

"very clear and well-organized."
Peter Weimer-Hastings, University of Edinburgh, Scotland.

Need more information?

Try the LEA website: www.erlbaum.com

Errata

Page 32: Typo Line -9: "it it" should read "it is"

Page 33: Typo Line +1: "styles of arrows" should read "styles of arrow"

Page 34: Typo Line +17: "component" should read "components"

Page 49: Figure 2.13 does not match its caption. The figure shows the rule editor (as stated), but the rule being edited is not one that sends all input directly to Output Words (as suggested), but one that adds all input to STS. The rule shown actually corresponds to Rule 1 in Box 2.3 and Figure 2.15.

Page 89: Misleading Typo Exercise 3.3 should state "Add the contents of Box 3.2 to Match Productions ...", not to Match Memory.

Pages 100, 102: Misleading Typo Production rules in Boxes 3.8 and 3.9 are encoded as rule/2 terms. The previous text and boxes refer to the same production rules as production/2 terms. The should either all be production/2 or all rule/2.

Page 277: Typo Example 7.3 (a): "parsed" should read "past"

Page 338: Misleading Typo Rule 2 in Box 8.8 should not be "refracted", it should be "unrefracted"

Additional Notes

COGENT has developed significantly since the book was published, meaning that some of the figures do not correspond to the current COGENT version. Users wishing to follow the book precisely should adopt COGENT version 2.1. Those who are happy with minor differences please use the most recent version. Note that there is one significant change that affects some models in Chapter 7 -- buffers now include an addition property: Grounded. When set (the default), this ensure that all elements in buffers are grounded terms (i.e., they contain no variables). Some models in Chapter 7 require ungrounded terms (e.g., those involving number/person agreement). These models will require that users deselect the grounded property in appropriate buffers.