A new book published by
Lawrence Erlbaum Associates, Inc.

the book page at http://www.erlbaum.com/

Duality of the Mind

A Bottom-up Approach toward Cognition

by Ron Sun

Duality of the Mind


A Summary

Synthesizing situated cognition, reinforcement learning, and hybrid connectionist modeling, a generic cognitive architecture focused on situated involvement and interaction with the world is developed in this book. The architecture notably incorporates the distinction of implicit and explicit processes.

The architecture is tested against a wide range of human learning performance data. The work described in the book demonstrates the cognitive validity and the power of the architecture, by ways of capturing a wide range of human learning phenomena. Computational properties of the architecture is explored with experiments that manipulate implicit and explicit processes to optimize performance in a range of domains. Finally, philosophical implications of the approach, on situated cognition, intentionality, symbol grounding, and consciousness, are elucidated.

In a nutshell, this book motivates, describes, and develops a broad framework for studying human cognition, based on a new approach that is characterized by its focus on the dichotomy of, and the interaction between, explicit cognition and implicit cognition.

This work can be of significant interest to a broad range of readers. For example, it can be of interest to lay readers having an interest in issues such as general principles of cognition, foundations of consciousness, or, more specifically, the role of skill learning in cognition, and so on. It can also be of interest to researchers and students of cognitive science, in that some general principles and frameworks, as well as their instantiations in computational models, may have significant bearings on the understanding of cognition, and to those of artificial intelligence, in that many models developed here (and the fundamental principles embodied in them) may be of use beyond modeling cognitive processes in particular domains and may be of use in understanding and developing intelligent systems in general.

See also the book page at

Read the comments from other cognitive scientists


Table of Contents



Chapter 1. The Essentials of Human Activities
1.1. Characteristics of Human Activities
1.2. Past Work on Human Activities
1.3. Pitfalls of Representation
1.4. Importance of Learning
1.5. A Broad Theory
1.6. Plan
1.7. Summary

Chapter 2. A Theoretical Model
2.1. Desiderata
2.1.1. \bf Implicit vs. Explicit Knowledge
2.1.2. The Interaction of Implicit and Explicit Knowledge
2.1.3. Bottom-Up Learning
2.1.4. Benefits of Incorporating Two Types of Knowledge
2.1.5. On-Line Learning
2.2. A Sketch of the Model
2.2.1. Representation
2.2.2. Learning
2.2.3. Combination and Separation
2.2.4. Summary of Basic Theoretical Hypotheses
2.3. Summary

Chapter 3. Current Implementations
3.1. Reinforcement Learning
3.2. Rule Extraction
3.3. Combining Value Functions and Rules
3.4. Plan Extraction
3.5. Modularity through Spatial Partitioning
3.6. Modularity through Temporal Segmentation
3.7. Biological Interpretations
3.8. Summary

Chapter 4. Accounting for Human Data Qualitatively
4.1. Dissociation
4.2. Division of Labor
4.3. Bottom-Up Learning
4.4. Differences in Representation of Resulting knowledge
4.5. Differences in Accessibility of Resulting Knowledge
4.6. Differences in Flexibility, Generalizability, and Robustness
4.7. Initiation of performance
4.8. Knowledge Interaction
4.9. Synergy
4.10. Summary

Chapter 5. Accounting for Human Data Quantitatively
5.1. Introduction
5.2. Simulating Serial Reaction Time Tasks
5.2.1. Simulating Lewicki et al (1987)
5.2.2. Simulating Curran and Keele (1993)
5.3. Simulating Process Control Tasks}
5.3.1. Simulating Stanley et al (1989)}
5.3.2. Simulating Berry and Broadbent (1988)
5.4. Simulating High-Level Cognitive Skill Learning Tasks
5.5. Simulating the Minefield Navigation Task
5.6. Discussions
5.6.1. Other Effects
5.6.2. Factors in Division of Labor
5.6.3. Why Synergy in the Model?
5.6.4. Uniqueness of the Model
5.7. Summary

Chapter 6. Symbol Grounding and Situated Cognition
6.1. Introduction
6.2. Symbols and Representation
6.2.1. Symbols
6.2.2. Representation
6.2.3. Intentionality
6.3. Everyday Activities and Symbol Grounding
6.3.1. Comportment
6.3.2. Conceptual Processes and Representation
6.3.3. A Dual Process Theory
6.4. Computational Analysis of Everyday Activities
6.4.1. Computational Processes of Comportment
6.4.2. Computational Processes of Conceptual Processing
6.4.3. Concept Formation
6.5. Representation and Intentionality
6.6. Further Discussions
6.7. Summary

Chapter 7. The Issue of Consciousness
7.1. Introduction
7.2. Explaining Consciousness
7.2.1. Different views
7.2.2. Analysis of Different Views
7.2.3. Further Evidence and Arguments
7.3. Functional Roles of Consciousness
7.3.1. Access Consciousness An Analysis Computational Account
7.3.2. Reflective Consciousness An Analysis Computational Account
7.3.3. Phenomenal Consciousness An Analysis Computational Account
7.4. Summary

Chapter 8. Sociocultural Factors in Cognition
8.1. Introduction
8.2. Aspects of Sociality
8.2.1. Sociality
8.2.2. Society Social Structures Culture Social Determinism?
8.3. Sociocultural Processes and Cognition
8.3.1. Inter-Agent Processes
8.3.2. Social Processes
8.4. Cognitive Modeling Incorporating Sociocultural Issues
8.4.1. Autonomous Generation of Symbolic Structures
8.4.2. Assimilation of Externally Provided Symbolic Structures
8.4.3. Sociocultural vs.\ Self-Generated Concepts
8.4.1. Accounting for Sociocultural Cognition
8.4.2. Representing Self and Others Developing Self Representation Developing Representation of Others
8.5. Further Work
8.5.1. Simulating Social Processes
8.5.2. Individual Beliefs and Sociocultural Beliefs
8.5.3. Inter-Agent Interaction
8.5.4. Forming Social Structures
8.5.5. Social Structures in Cognition
8.6. Summary

Chapter 9. Comparisons
9.1. Cognitive Models Involving Implicit Learning
9.1.1. Process Differences of the Two Levels
9.1.2. Connectionist Modeling of the Simulated Skill Learning Tasks
9.1.3. Connectionist Modeling of Other Tasks
9.1.4. Non-connectionist Modeling of Skill Learning Tasks
9.2. Top-down vs.\ Bottom-up Models
9.3. Cognitive Architectures
9.4. Objections and Responses
9.4.1. One Level or Two?
9.4.2. Need for Symbolic Representation
9.4.3. Verification of Explicit Knowledge
9.4.4. Hybridization
9.4.5. Free Parameters?
9.5. Models of Consciousness
9.6. Computational Issues
9.6.1. Models of Rule Learning
9.6.2. Models of Planning
9.6.3. Models of Spatial Partitioning
9.6.4. Models of Temporal Partitioning

Chapter 10. Conclusions
10.1. The Major Theme
10.2. Methodological Considerations
10.3. Cognitive Implications
10.4. A Final Summary


To order, go to

Lawrence Erlbaum Associates, Inc.
10 Industrial Avenue
Mahwah, NJ 07430, USA
Tel: 201-258-2200
Fax: 201-236-0072

Go to Prof. Ron Sun's Homepage