As the study and application of computing expand to encompass more than technology, so does the need to attend to the human component of information systems. Humans are designers, developers, programmers, data administrators, managers, documenters, and users of information systems. Learning about the human mind is important for work in computing.

It is this human component that is addressed by cognitive science. Cognitive science studies the mind. It seeks to understand how humans, or machines and non-human animals, perceive, think, reason, remember, understand language, and learn. Cognitive science is frequently seen as a convergence of research from diverse disciplines on a set of core problems and explanations. There is more agreement on the set of problems that are included in cognitive science than on the set of disciplines to be considered cognitive sciences. Notwithstanding, an introduction to cognitive science based on any of the books reviewed here will introduce students to interdisciplinary research in the various supporting areas.

Five introductory cognitive science books have appeared within the past few years. These five books share the common goal of introducing interested readers to the mysteries of cognition. We review these five from the point of view of an educator in computer science.

As would be expected with a topic as potentially complex as the mind, presentation of the subject matter varies. Among these books, two approaches have been used. One addresses the individual problems that researchers attack, such as perception, reasoning, problem solving, or language acquisition. The other presents the contributions of research by the contributing discipline, such as philosophy, linguistics, or artificial intelligence. It is our experience that each has its place in learning about the mind. For a first course, the problem approach is more exciting; for the continuing student, the disciplinary approach adds depth and focus.

Gardner

Gardner’s book traces the historical roots and development of cognitive science through the encapsulation of contributions and contributors. First published in 1985, the book was reissued in 1987 with an epilogue to update the work. In Gardner’s words,

I decided to make a comprehensive investigation of cognitive science in which I could include the long view--the philosophical origins, the histories of each of the respective fields, the current work that appears most central, and my own assessment of the prospects for this ambitious field (p. xiii).

Gardner envisions cognitive science as an integration of the cognitive sciences. In keeping with this view, he presents the contributions of the individual disciplines before he discusses the research problems. Part 1 introduces the cognitive revolution, and places the emerging field in context with earlier studies of the mind. Part 2 looks at  Gardner’s  choices for contributing disciplines: philosophy, psychology, artificial intelligence, linguistics, anthropology, and neuroscience. Each chapter not only describes research contributions but also places those contributions in perspective within the history of the study of mind. Part 3 addresses the current research problems of perception, mental imagery, categorization, reasoning, and computation. The epilogue comments primarily on parallel distributed processing, an approach to understanding the mind that was just beginning its renaissance in the scientific world.

Gardner succeeds better than most in presenting a detailed description of the origins and development of the field. The content and balance of the book are excellent. Although it does not include suggested further readings, each chapter contains a wealth of references that can be followed up if desired.

Johnson-Laird

Johnson-Laird wrote for “anyone who is interested in the scientific explanation of mental phenomena” (p. 9). Although the subject is intellectually demanding, his aim is to accommodate the reader who is able to “think, not to come equipped with a large body of technical knowledge” (p. 9).

Like Gardner, Johnson-Laird sees multiple sciences as studying the mind, but expresses doubt that a unified science of the mind has been developed. His approach to exploring cognitive science is by problem, with six parts covering computation; vision; learning, memory, and action; cogitation; communication; and the conscious and unconscious mind.

Part 1 briefly introduces methods for studying the mind, the role of symbols in mental processes, and the primacy of computation as the appropriate conceptual apparatus for theories of the mind. Part 2 considers low-level vision research, depth perception as an area of high-level vision research, and the study of scenes, shapes, and images as the precursor for developing robot vision systems. Part 3 considers learning; memory; planning and productions; parallel distributed processing; and action and control of movements. Part 4 concerns thinking in its many variations. Johnson-Laird presents chapters on deduction, induction and probability, and creativity. Part 5 considers the communication modalities of speech and hearing and examines the natural language constructs of grammar and meaning. Part 6 considers the roles of such faculties as self-reflection, free will, intentions, needs, and intuitions in human cognition.

The book is well organized and presents its material simply and understandably. Illustrations are well chosen and the suggestions for further reading are informatively annotated to assist the reader in going beyond the material in the book. The reference list features a reference to the page or pages in the book on which the citation appears. This list is in addition to name and subject indices.

Osherson et al.

Osherson was the overall editor for this three-volume series, Invitation to cognitive science. (We review this set as one entity.) As is apparent from the volume titles, this set addresses problems of research in four areas: language, vision, action, and thinking. Although arranged by problem, the end result is a series of tutorials on important areas of psychology, computer science, linguistics, philosophy, engineering, and biology.

Volume 1, Language, has nine chapters. Three chapters introduce the basic linguistic building blocks of syntax, semantics, and phonology. The next three chapters discuss the higher-level processes of speech perception, lexical processing, and sentence processing. The final three chapters examine the relationship of the organization of the language faculty (a matter of the mind) and the brain (a neurological matter); development of language by children; and philosophical concerns about the nature of rules of language, characteristics of evidence for linguistic principles, and the relationship between linguistic analysis and thought expressed in language.

Kosslyn edited the visual cognition section and Hollerback edited the action section of volume 2, Visual cognition and action. Visual cognition includes low-level vision, high-level vision, mental imagery, developmental vision, and a philosophical discussion of seeing, believing, and knowing. The section on action concentrates on motor movements of the arm, but also includes chapters on oculomotor control, planning sequential motor activity, and the apparent paradox of a scientifically described physical system and the free will displayed by freely choosing agents.

Volume 3, Thinking, addresses nine topics, primarily from the field of cognitive psychology. Basic cognitive processes are represented by chapters on memory and categorization. It is the higher-level processes that receive the majority of attention. Judgement, choice, problem-solving, and rationality are treated separately. The chapter on cognitive development addresses acquisition of conceptual structures that support higher-level cognitive processes. Volume 3 includes an epilogue that rightly pertains to all the volumes. Three chapters address artificial neural networks, the evolution of cognition, and a computer model of the mind.

In keeping with the series title, this set is suitable for newcomers to cognitive science. Unfortunately, although thoughtful questions are included by most contributors, suggested answers are not provided. Because no one teacher is likely to be expert in all of the contributing disciplines, suggested answers would be welcome. Suggestions for further readings are suitably annotated to assist readers in pursuing topics of interest.

Posner

Of the five books, this volume presents the most advanced treatment of topics. Addressed to a “new generation of researchers,” the book is divided into sections entitled “Foundations,” “Domains,” and “Assessment.”

“Foundations” consists of seven chapters that could serve as a one-semester course in the basic underlying paradigms of current cognitive science research. An introduction to the paradigm of computation establishes the perspective for the remaining chapters. Both symbolic and connectionist versions of the architecture of the mind are presented separately at length. Grammatical theory, devoted primarily to a syntactic treatment of language, is complemented by a chapter on model-theoretic semantics, which treats the property of “aboutness.” The chapter on experimental methods in cognitive science is welcome not only for its rarity, but for its coverage of psychological methods. These methods are presented simply. It would have been helpful if the author had explained similarly useful methods from the neurosciences as clearly. The final chapter in this section addresses the relation between biological and cognitive (meaning psychological) levels of investigation.

“Domains” is the repository of the problem chapters. In addition to the expected chapters on language acquisition, memory, action, categorization, problem-solving, and vision, we find chapters on several areas not usually included in mainstream cognitive science: reading, mental models, and visual attention. Inclusion of these areas points to the plasticity of the field--it does not yet have boundaries authenticated by community consensus.

“Assessment” contains two chapters, one on cultural cognition and one on philosophical issues such as intentionality, the mind-body problem, and qualia.

This collection of papers is probably most accessible to advanced undergraduate and graduate students. Prior knowledge of experimental methods will be useful, but lack of such knowledge will not preclude success for the determined reader. Because of its length and organization, the book probably requires more time (perhaps two semesters) for course coverage.

Stillings et al.

The experience of seven cooperating college instructors resulted in this book, which is designed for an undergraduate introductory course in cognitive science. The contributors represent four disciplines, judging from their departmental affiliations: psychology, linguistics, philosophy, and computer science.

The book is organized primarily by discipline: cognitive psychology, artificial intelligence, linguistics, neuroscience, and philosophy. After an introductory chapter on the nature of cognitive science, two chapters written from the perspective of cognitive psychology address such topics as cognitive architecture, representation, mental images, skill acquisition, automaticity, memory, reasoning, problem solving, and cognitive development. Two chapters on artificial intelligence address the nature of artificial intelligence, knowledge representation, search, and learning. Neuroscience, philosophy, and vision each have one chapter. Linguistics, however, rates four chapters. Language representation, language acquisition, semantics, and natural language processing are treated separately.

This book meets its objective of accessibility to undergraduates. Illustrations are used sparingly; more would be better. Annotated suggestions for further reading conclude each chapter.

Comparison

Although all the books present a decent introduction to cognitive science, differences in treatment and focus distinguish them and affect their usability as texts. Because of their problem orientation and level of presentation, we found both the Osherson set and Johnson-Laird successful in introductory courses. Their problem approach is more interesting and less tedious for students who are new to social science research literature. In a sense, we need to “hook” students with the interesting and exciting problems before we submerge them in the essential (but not thrilling) details of history and methodologies. Despite the length of the Osherson set, it can comfortably be covered in one semester, with some pruning. Both selections include annotated suggested readings, and Osherson provides questions (but no answers).

Gardner and Stillings are the two discipline-oriented books. Our experience with Gardner as the basic text for an introductory course was less than successful. We think presenting contributions by discipline is too dry and appeals less to the student than the problem approach. Comparatively, Gardner’s historical treatment enriches with detail and context, whereas Stillings et al. remains introductory in level. Gardner would be useful as an optional text to accompany one of the problem approach texts. As the required or only text, Gardner would be better in a second course aimed at students with basic knowledge in place.

Without doubt, Posner’s book presents the most advanced treatment of cognitive science. We think it would require two semesters to thoroughly cover the material. The “Foundations” chapters are particularly full of new, bedrock concepts, and thus require more time for proper assimilation. If used in a first course, even for graduate students, we would strongly consider reversing the order of the “Foundations” and “Domains” sections.