What do you think?
Rate this book
Neurophilosophy at Work
In this collection of essays, Paul Churchland explores the unfolding impact of the several empirical sciences of the mind, especially cognitive neurobiology and computational neuroscience on a variety of traditional issues central to the discipline of philosophy. Representing Churchland's most recent research, they continue his research program, launched over thirty years ago, and which has evolved into the field of neurophilosophy.
262 pages, Paperback
First published January 5, 2007
6 people are currently reading
124 people want to read
About the author
Paul M. Churchland
18 books64 followersPaul Churchland is a philosopher noted for his studies in neurophilosophy and the philosophy of mind. He is currently a Professor at the University of California, San Diego, where he holds the Valtz Chair of Philosophy. Churchland holds a joint appointment with the Cognitive Science Faculty and the Institute for Neural Computation. He earned his Ph.D. from the University of Pittsburgh in 1969 under the direction of Wilfrid Sellars. Churchland is the husband of philosopher Patricia Churchland, and the father of two children.
Churchland began his professional career as an instructor at the University of Pittsburgh in 1969; he also lectured at the University of Toronto from 1967-69. In 1969, Churchland took a position at the University of Manitoba, where he would teach for fifteen years: as an assistant professor (69 - 74) and associate professor (74 - 79), and then as a full professor from 1979 - 1984. Professor Churchland joined the Institute for Advanced Study at Princeton University in 1982, staying as a member until 1983. He joined the faculty at the University of California, San Diego in 1983, serving as Department Chair from 1986 - 1990.
Churchland has supervised a number of PhD students, including P.D. Magnus (now at the University at Albany) and Philip Brey (now at the University of Twente).
Along with his wife, Churchland is a major proponent of eliminative materialism, which claims that everyday mental concepts such as beliefs, feelings and desires are theoretical constructs without coherent definition; hence we should not expect such concepts to be a necessary part of a scientific understanding of the brain. Just as a modern understanding of science has no need for concepts such as luck or witchcraft to explain the world, Churchland argues that a future neuroscience is likely to have no need for "beliefs" or "feelings" to explain the mind. Instead, the use of objective phenomena such as neurons and their interaction should suffice. He points out that the history of science has seen many previous concepts discarded, such as phlogiston, caloric, the luminiferous ether, and vital forces.
Churchland began his professional career as an instructor at the University of Pittsburgh in 1969; he also lectured at the University of Toronto from 1967-69. In 1969, Churchland took a position at the University of Manitoba, where he would teach for fifteen years: as an assistant professor (69 - 74) and associate professor (74 - 79), and then as a full professor from 1979 - 1984. Professor Churchland joined the Institute for Advanced Study at Princeton University in 1982, staying as a member until 1983. He joined the faculty at the University of California, San Diego in 1983, serving as Department Chair from 1986 - 1990.
Churchland has supervised a number of PhD students, including P.D. Magnus (now at the University at Albany) and Philip Brey (now at the University of Twente).
Along with his wife, Churchland is a major proponent of eliminative materialism, which claims that everyday mental concepts such as beliefs, feelings and desires are theoretical constructs without coherent definition; hence we should not expect such concepts to be a necessary part of a scientific understanding of the brain. Just as a modern understanding of science has no need for concepts such as luck or witchcraft to explain the world, Churchland argues that a future neuroscience is likely to have no need for "beliefs" or "feelings" to explain the mind. Instead, the use of objective phenomena such as neurons and their interaction should suffice. He points out that the history of science has seen many previous concepts discarded, such as phlogiston, caloric, the luminiferous ether, and vital forces.
Ratings & Reviews
Friends & Following
Create a free account to discover what your friends think of this book!
Community Reviews
Displaying 1 - 3 of 3 reviews
October 4, 2009
Neurophilosophy at Work is a collection of journal articles and lectures, each of which is useful for its highly specialized purpose, but which has excessive academic jargon and logical subtlety for the general reader. An example would be a lecture criticizing a Kansas school board for removing evolution and big bang theory from their required curriculum. The arguments are too erudite and convoluted to convince a beleaguered board of education, which would need a short, punchy argument, notwithstanding the noble purpose. The last Churchland book I read, The Engine of Reason the Seat of the Soul, was far more accessible, written as a popularizing textbook.
Churchland likes to set up an opposition, a battle between paradigms of cognitive activity, between his model of neural networks, with massive and recurrent parallel distributed processing (PDP) and what he regards as the prevalent, conventional model of propositional attitudes, with reasoning based on language.
Among the differences he points out between the models are the following: Humans and animals share many cognitive abilities, but they do not share language, so that the PDP model has a greater explanatory power. Cognitive abilities are exhibited in many physical activities such as sports and manual labor, which are too rapid to be explained by reasoned language. One would have to ask how many propositional attitudes can be formulated in a second, to account for such activity. Churchland goes back to Gilbert Ryle’s distinction between “knowing how and knowing that” as anticipating the rival models, and he ultimately regards propositional attitudes or reasoned language as a “conceptual scaffolding” developed late in the cognitive game, largely for social purposes.
I don’t see nearly the contrast between the two models that Churchland does. They could be regarded, it seems to me, as something like vertical and horizontal cross-sections of the same phenomenon. I will elaborate on Churchland’s model to explain.
Within PDP processing, perceptual features can be defined using vector processing, where a vector is located in n-dimensional space (called “activational space”). For example, a taste (of lemon) might be defined in the 3-dimensional space of sweet, sour and salty, or a color (of magenta) might be defined in the 3-dimensional space of red-green, yellow-blue and black-white. With enough dimensions, such vectors can be the ending points of perceptual activities, and both the starting and ending points of further cognitive activities.
Churchland acknowledges that an activational space is like a concept, and a vector within that space is an instance of that concept. The movement from one vector to another, through PDP, results in something like a judgment or inference. For example, visual input which goes through PDP may yield a vector that indicates a recognized face. This vector may go through further PDP to yield another vector which indicates an expression of surprise. Such processing amounts to a judgment. You can call it a neural network, or you can call it a propositional attitude, but it still amounts to recognizing that your friend is surprised.
Churchland also offers a neural network account of semantic reference. I am highly skeptical, but his argument is far too complicated to summarize or to critique here. There is a lot left for us to study, and a lot left for Churchland (and like-minded philosophers) to explore and develop.
Churchland likes to set up an opposition, a battle between paradigms of cognitive activity, between his model of neural networks, with massive and recurrent parallel distributed processing (PDP) and what he regards as the prevalent, conventional model of propositional attitudes, with reasoning based on language.
Among the differences he points out between the models are the following: Humans and animals share many cognitive abilities, but they do not share language, so that the PDP model has a greater explanatory power. Cognitive abilities are exhibited in many physical activities such as sports and manual labor, which are too rapid to be explained by reasoned language. One would have to ask how many propositional attitudes can be formulated in a second, to account for such activity. Churchland goes back to Gilbert Ryle’s distinction between “knowing how and knowing that” as anticipating the rival models, and he ultimately regards propositional attitudes or reasoned language as a “conceptual scaffolding” developed late in the cognitive game, largely for social purposes.
I don’t see nearly the contrast between the two models that Churchland does. They could be regarded, it seems to me, as something like vertical and horizontal cross-sections of the same phenomenon. I will elaborate on Churchland’s model to explain.
Within PDP processing, perceptual features can be defined using vector processing, where a vector is located in n-dimensional space (called “activational space”). For example, a taste (of lemon) might be defined in the 3-dimensional space of sweet, sour and salty, or a color (of magenta) might be defined in the 3-dimensional space of red-green, yellow-blue and black-white. With enough dimensions, such vectors can be the ending points of perceptual activities, and both the starting and ending points of further cognitive activities.
Churchland acknowledges that an activational space is like a concept, and a vector within that space is an instance of that concept. The movement from one vector to another, through PDP, results in something like a judgment or inference. For example, visual input which goes through PDP may yield a vector that indicates a recognized face. This vector may go through further PDP to yield another vector which indicates an expression of surprise. Such processing amounts to a judgment. You can call it a neural network, or you can call it a propositional attitude, but it still amounts to recognizing that your friend is surprised.
Churchland also offers a neural network account of semantic reference. I am highly skeptical, but his argument is far too complicated to summarize or to critique here. There is a lot left for us to study, and a lot left for Churchland (and like-minded philosophers) to explore and develop.
February 15, 2017
Different essays, some of which were extremely complex for me to understand. Some others were easier and usual involved more philosophy than algorithms. This book is also almost 10 years old, so if you need the state of the art of Neurophilosophy, it may be better to look for some other book.
Differenti saggi, alcuni dei quali troppo complessi per me. Altri invece erano di piú facile comprensione e di solito avevano a che fare piú con la filosofia che con gli algoritmi. Inoltre questo libro ha quasi 10 anni quindi magari se vi serve una situazione dello stato dell'arte, magari é meglio cercare qualcosa di piú aggiornato.
Differenti saggi, alcuni dei quali troppo complessi per me. Altri invece erano di piú facile comprensione e di solito avevano a che fare piú con la filosofia che con gli algoritmi. Inoltre questo libro ha quasi 10 anni quindi magari se vi serve una situazione dello stato dell'arte, magari é meglio cercare qualcosa di piú aggiornato.
Want to read
July 18, 2013From my notebook: neuro-phil.: book / nuerology intro >> how brain processes color
Displaying 1 - 3 of 3 reviews