This book offers a concise and accessible introduction to the emerging field of artificial cognitive systems. Cognition, both natural and artificial, is about anticipating the need for action and developing the capacity to predict the outcome of those actions. Drawing on artificial intelligence, developmental psychology, and cognitive neuroscience, the field of artificial cognitive systems has as its ultimate goal the creation of computer-based systems that can interact with humans and serve society in a variety of ways. This primer brings together recent work in cognitive science and cognitive robotics to offer readers a solid grounding on key issues.
The book first develops a working definition of cognitive systems—broad enough to encompass multiple views of the subject and deep enough to help in the formulation of theories and models. It surveys the cognitivist, emergent, and hybrid paradigms of cognitive science and discusses cognitive architectures derived from them. It then turns to the key issues, with chapters devoted to autonomy, embodiment, learning and development, memory and prospection, knowledge and representation, and social cognition. Ideas are introduced in an intuitive, natural order, with an emphasis on the relationships among ideas and building to an overview of the field. The main text is straightforward and succinct; sidenotes drill deeper on specific topics and provide contextual links to further reading.
Complex adaptive systems (cas), including ecosystems, governments, biological cells, and markets, are characterized by intricate hierarchical arrangements of boundaries and signals. In ecosystems, for example, niches act as semi-permeable boundaries, and smells and visual patterns serve as signals; governments have departmental hierarchies with memoranda acting as signals; and so it is with other cas. Despite a wealth of data and descriptions concerning different cas, there remain many unanswered questions about "steering" these systems. In Signals and Boundaries, John Holland argues that understanding the origin of the intricate signal/border hierarchies of these systems is the key to answering such questions. He develops an overarching framework for comparing and steering cas through the mechanisms that generate their signal/boundary hierarchies.
Holland lays out a path for developing the framework that emphasizes agents, niches, theory, and mathematical models. He discusses, among other topics, theory construction; signal-processing agents; networks as representations of signal/boundary interaction; adaptation; recombination and reproduction; the use of tagged urn models (adapted from elementary probability theory) to represent boundary hierarchies; finitely generated systems as a way to tie the models examined into a single framework; the framework itself, illustrated by a simple finitely generated version of the development of a multi-celled organism; and Markov processes.
"Simulation," writes Gary Flake in his preface, "becomes a form of experimentation in a universe of theories. The primary purpose of this book is to celebrate this fact."In this book, Gary William Flake develops in depth the simple idea that recurrent rules can produce rich and complicated behaviors. Distinguishing "agents" (e.g., molecules, cells, animals, and species) from their interactions (e.g., chemical reactions, immune system responses, sexual reproduction, and evolution), Flake argues that it is the computational properties of interactions that account for much of what we think of as "beautiful" and "interesting." From this basic thesis, Flake explores what he considers to be today's four most interesting computational topics: fractals, chaos, complex systems, and adaptation.
Each of the book's parts can be read independently, enabling even the casual reader to understand and work with the basic equations and programs. Yet the parts are bound together by the theme of the computer as a laboratory and a metaphor for understanding the universe. The inspired reader will experiment further with the ideas presented to create fractal landscapes, chaotic systems, artificial life forms, genetic algorithms, and artificial neural networks.
In development for thirty years, Soar is a general cognitive architecture that integrates knowledge-intensive reasoning, reactive execution, hierarchical reasoning, planning, and learning from experience, with the goal of creating a general computational system that has the same cognitive abilities as humans. In contrast, most AI systems are designed to solve only one type of problem, such as playing chess, searching the Internet, or scheduling aircraft departures. Soar is both a software system for agent development and a theory of what computational structures are necessary to support human-level agents. Over the years, both software system and theory have evolved. This book offers the definitive presentation of Soar from theoretical and practical perspectives, providing comprehensive descriptions of fundamental aspects and new components.
The current version of Soar features major extensions, adding reinforcement learning, semantic memory, episodic memory, mental imagery, and an appraisal-based model of emotion. This book describes details of Soar's component memories and processes and offers demonstrations of individual components, components working in combination, and real-world applications. Beyond these functional considerations, the book also proposes requirements for general cognitive architectures and explicitly evaluates how well Soar meets those requirements.
This book guides students through an exploration of the idea that thinking might be understood as a form of computation. Students make the connection between thinking and computing by learning to write computer programs for a variety of tasks that require thought, including solving puzzles, understanding natural language, recognizing objects in visual scenes, planning courses of action, and playing strategic games. The material is presented with minimal technicalities and is accessible to undergraduate students with no specialized knowledge or technical background beyond high school mathematics. Students use Prolog (without having to learn algorithms: “Prolog without tears!”), learning to express what they need as a Prolog program and letting Prolog search for answers.
After an introduction to the basic concepts, Thinking as Computation offers three chapters on Prolog, covering back-chaining, programs and queries, and how to write the sorts of Prolog programs used in the book. The book follows this with case studies of tasks that appear to require thought, then looks beyond Prolog to consider learning, explaining, and propositional reasoning. Most of the chapters conclude with short bibliographic notes and exercises. The book is based on a popular course at the University of Toronto and can be used in a variety of classroom contexts, by students ranging from first-year liberal arts undergraduates to more technically advanced computer science students.
The capacity to think about our own thinking may lie at the heart of what it means to be both human and intelligent. Philosophers and cognitive scientists have investigated these matters for many years. Researchers in artificial intelligence have gone further, attempting to implement actual machines that mimic, simulate, and perhaps even replicate this capacity, called metareasoning. In this volume, leading authorities offer a variety of perspectives—drawn from philosophy, cognitive psychology, and computer science—on reasoning about the reasoning process.
The book offers a simple model of reasoning about reason as a framework for its discussions. Following this framework, the contributors consider metalevel control of computational activities, introspective monitoring, distributed metareasoning, and, putting all these aspects of metareasoning together, models of the self. Taken together, the chapters offer an integrated narrative on metareasoning themes from both artificial intelligence and cognitive science perspectives.
Despite their apparently divergent accounts of higher cognition, cognitive theories based on neural computation and those employing symbolic computation can in fact strengthen one another. To substantiate this controversial claim, this landmark work develops in depth a cognitive architecture based in neural computation but supporting formally explicit higher-level symbolic descriptions, including new grammar formalisms.
Detailed studies in both phonology and syntax provide arguments that these grammatical theories and their neural network realizations enable deeper explanations of early acquisition, processing difficulty, cross-linguistic typology, and the possibility of genomically encoding universal principles of grammar. Foundational questions concerning the explanatory status of symbols for central problems such as the unbounded productivity of higher cognition are also given proper treatment.
The work is made accessible to scholars in different fields of cognitive science through tutorial chapters and numerous expository boxes providing background material from several disciplines. Examples common to different chapters facilitate the transition from more basic to more sophisticated treatments. Details of method, formalism, and foundation are presented in later chapters, offering a wealth of new results to specialists in psycholinguistics, language acquisition, theoretical linguistics, computational linguistics, computational neuroscience, connectionist modeling, and philosophy of mind.
The last decade has seen computational implementations of large hand-crafted natural language grammars in formal frameworks such as Tree-Adjoining Grammar (TAG), Combinatory Categorical Grammar (CCG), Head-driven Phrase Structure Grammar (HPSG), and Lexical Functional Grammar (LFG). Grammars in these frameworks typically associate linguistically motivated rich descriptions (Supertags) with words. With the availability of parse-annotated corpora, grammars in the TAG and CCG frameworks have also been automatically extracted while maintaining the linguistic relevance of the extracted Supertags. In these frameworks, Supertags are designed so that complex linguistic constraints are localized to operate within the domain of those descriptions. While this localization increases local ambiguity, the process of disambiguation (Supertagging) provides a unique way of combining linguistic and statistical information.
This volume investigates the theme of employing statistical approaches with linguistically motivated representations and its impact on Natural Language Processing tasks. In particular, the contributors describe research in which words are associated with Supertags that are the primitives of different grammar formalisms including Lexicalized Tree-Adjoining Grammar (LTAG).
Contributors: Jens Bäcker, Srinivas Bangalore, Akshar Bharati, Pierre Boullier, Tomas By, John Chen, Stephen Clark, Berthold Crysmann, James R. Curran, Kilian Foth, Robert Frank, Karin Harbusch, Mary Harper, Saša Hasan, Aravind Joshi,Vincenzo Lombardo, Takuya Matsuzaki, Alessandro Mazzei, Wolfgang Menzel, Yusuke Miyao, Richard Moot, Alexis Nasr, Günter Neumann, Martha Palmer, Owen Rambow, Rajeev Sangal, Anoop Sarkar, Giorgio Satta, Libin Shen, Patrick Sturt, Jun’ichi Tsujii, K. Vijay-Shanker, Wen Wang, Fei Xia
The conceptual history of cognitive science remains for the most part unwritten. In this groundbreaking book, Jean-Pierre Dupuy—one of the principal architects of cognitive science in France—provides an important chapter: the legacy of cybernetics. Contrary to popular belief, Dupuy argues, cybernetics represented not the anthropomorphization of the machine but the mechanization of the human. The founding fathers of cybernetics—some of the greatest minds of the twentieth century, including John von Neumann, Norbert Wiener, Warren McCulloch, and Walter Pitts—intended to construct a materialist and mechanistic science of mental behavior that would make it possible at last to resolve the ancient philosophical problem of mind and matter. The importance of cybernetics to cognitive science, Dupuy argues, lies not in its daring conception of the human mind in terms of the functioning of a machine but in the way the strengths and weaknesses of the cybernetics approach can illuminate controversies that rage today—between cognitivists and connectionists, eliminative materialists and Wittgensteinians, functionalists and anti-reductionists.
Dupuy brings to life the intellectual excitement that attended the birth of cognitive science sixty years ago. He separates the promise of cybernetic ideas from the disappointment that followed as cybernetics was rejected and consigned to intellectual oblivion. The mechanization of the mind has reemerged today as an all-encompassing paradigm in the convergence of nanotechnology, biotechnology, information technology, and cognitive science. The tensions, contradictions, paradoxes, and confusions Dupuy discerns in cybernetics offer a cautionary tale for future developments in cognitive science.