Abductive Reasoning And Learning

This book contains leading survey papers on the various aspects of Abduction, both logical and numerical approaches. Abduction is central to all areas of applied reasoning, including artificial intelligence, philosophy of science, machine learning, data mining and decision theory, as well as logic itself.

From the very beginning of their investigation of human reasoning, philosophers have identified two other forms of reasoning, besides deduction, which we now call abduction and induction. Deduction is now fairly well understood, but abduction and induction have eluded a similar level of understanding. The papers collected here address the relationship between abduction and induction and their possible integration. The approach is sometimes philosophical, sometimes that of pure logic, and some papers adopt the more task-oriented approach of AI. The book will command the attention of philosophers, logicians, AI researchers and computer scientists in general.

A study of the role of abductive inference in everyday argumentation and legal evidence Examines three areas in which abductive reasoning is especially important: medicine, science, and law. The reader is introduced to abduction and shown how it has evolved historically into the framework of conventional wisdom in logic. Discussions draw upon recent techniques used in artificial intelligence, particularly in the areas of multi-agent systems and plan recognition, to develop a dialogue model of explanation. Cases of causal explanations in law are analyzed using abductive reasoning, and all the components are finally brought together to build a new account of abductive reasoning. By clarifying the notion of abduction as a common and significant type of reasoning in everyday argumentation, Abductive Reasoning will be useful to scholars and students in many fields, including argumentation, computing and artificial intelligence, psychology and cognitive science, law, philosophy, linguistics, and speech communication and rhetoric.

“Artificial intelligence has always inspired outlandish visions—that AI is going to destroy us, save us, or at the very least radically transform us. Erik Larson exposes the vast gap between the actual science underlying AI and the dramatic claims being made for it. This is a timely, important, and even essential book.” —John Horgan, author of The End of Science Many futurists insist that AI will soon achieve human levels of intelligence. From there, it will quickly eclipse the most gifted human mind. The Myth of Artificial Intelligence argues that such claims are just that: myths. We are not on the path to developing truly intelligent machines. We don’t even know where that path might be. Erik Larson charts a journey through the landscape of AI, from Alan Turing’s early work to today’s dominant models of machine learning. Since the beginning, AI researchers and enthusiasts have equated the reasoning approaches of AI with those of human intelligence. But this is a profound mistake. Even cutting-edge AI looks nothing like human intelligence. Modern AI is based on inductive reasoning: computers make statistical correlations to determine which answer is likely to be right, allowing software to, say, detect a particular face in an image. But human reasoning is entirely different. Humans do not correlate data sets; we make conjectures sensitive to context—the best guess, given our observations and what we already know about the world. We haven’t a clue how to program this kind of reasoning, known as abduction. Yet it is the heart of common sense. Larson argues that all this AI hype is bad science and bad for science. A culture of invention thrives on exploring unknowns, not overselling existing methods. Inductive AI will continue to improve at narrow tasks, but if we are to make real progress, we must abandon futuristic talk and learn to better appreciate the only true intelligence we know—our own.

This book is about abduction, 'the logic of Sherlock Holmes', and about how some kinds of abductive reasoning can be programmed in a computer. The work brings together Artificial Intelligence and philosophy of science and is rich with implications for other areas such as, psychology, medical informatics, and linguistics. It also has subtle implications for evidence evaluation in areas such as accident investigation, confirmation of scientific theories, law, diagnosis, and financial auditing. The book is about certainty and the logico-computational foundations of knowledge; it is about inference in perception, reasoning strategies, and building expert systems.

This book gathers together novel essays on the state-of-the-art research into the logic and practice of abduction. In many ways, abduction has become established and essential to several fields, such as logic, cognitive science, artificial intelligence, philosophy of science, and methodology. In recent years this interest in abduction’s many aspects and functions has accelerated. There are evidently several different interpretations and uses for abduction. Many fundamental questions on abduction remain open. How is abduction manifested in human cognition and intelligence? What kinds or types of abduction can be discerned? What is the role for abduction in inquiry and mathematical discovery? The chapters aim at providing answer to these and other current questions. Their contributors have been at the forefront of discussions on abduction, and offer here their updated approaches to the issues that they consider central to abduction’s contemporary relevance. The book is an essential reading for any scholar or professional keeping up with disciplines impacted by the study of abductive reasoning, and its novel development and applications in various fields.

A novel defense of abduction, one of the main forms of nondeductive reasoning. With this book, Igor Douven offers the first comprehensive defense of abduction, a form of nondeductive reasoning. Abductive reasoning, which is guided by explanatory considerations, has been under normative pressure since the advent of Bayesian approaches to rationality. Douven argues that, although it deviates from Bayesian tenets, abduction is nonetheless rational. Drawing on scientific results, in particular those from reasoning research, and using computer simulations, Douven addresses the main critiques of abduction. He shows that versions of abduction can perform better than the currently popular Bayesian approaches—and can even do the sort of heavy lifting that philosophers have hoped it would do. Douven examines abduction in detail, comparing it to other modes of inference, explaining its historical roots, discussing various definitions of abduction given in the philosophical literature, and addressing the problem of underdetermination. He looks at reasoning research that investigates how judgments of explanation quality affect people’s beliefs and especially their changes of belief. He considers the two main objections to abduction, the dynamic Dutch book argument, and the inaccuracy-minimization argument, and then gives abduction a positive grounding, using agent-based models to show the superiority of abduction in some contexts. Finally, he puts abduction to work in a well-known underdetermination argument, the argument for skepticism regarding the external world.

This is the first of two volumes comprising the papers submitted for publication by the invited participants to the Tenth International Congress of Logic, Methodology and Philosophy of Science, held in Florence, August 1995. The Congress was held under the auspices of the International Union of History and Philosophy of Science, Division of Logic, Methodology and Philosophy of Science. The invited lectures published in the two volumes demonstrate much of what goes on in the fields of the Congress and give the state of the art of current research. The two volumes cover the traditional subdisciplines of mathematical logic and philosophical logic, as well as their interfaces with computer science, linguistics and philosophy. Philosophy of science is broadly represented, too, including general issues of natural sciences, social sciences and humanities. The papers in Volume One are concerned with logic, mathematical logic, the philosophy of logic and mathematics, and computer science.