Through the looking glass.

This book is a strong dissonance in a loud chorus of neuroscientists, psychologists, and philosophers who promise that, after yet another few years, one more large-scale project, one more combined effort of EEG, MEG, fMRI, and PET experts, one more imaging technology, the mystery of mind and consciousness, of feelings and thoughts, will finally be solved (or dissolved).Then the dualism of mind and matter will be overcome and replaced by a long-awaited unified science of mind/brain/gene, which will explain all the heights and subtleties of the human condition on the basis of neuronal activity – or at the very least, the chemical and physical processes in the brain. In contrast to this view (very optimistic or frightening, depending on one’s perspective), Bennett and Hacker (hereafter referred to as B&H) suggest that modern neuroscience – perhaps not the whole of it, but at least large parts – is now at a dead end, not because it did not develop technologies to answer more and more intriguing questions, but because it cannot formulate the correct questions. Not because it is unable to reach its goals, but because it has no idea which goals should be reached. Not because of the unsolved experimental problems, but because of the conceptual ones. Those who take up the task of reading the 450 pages of this book can accept or reject one or more of the ideas and conclusions of the authors – but they will be unable to go back to “science as usual,” passively accepting the dogmas so strongly shaken by B&H. This means that the book is mandatory reading material. The book has several important merits. First, the tandem of prominent authors – a neurophysiologist and a philosopher – are fully devoted to the search for truth and clarity, not to impress the public with a new bestseller. Second, it is written in excellent English, with a brilliant sense of humor. Third, it is extremely comprehensive and in this respect, reminds one of James’s Psychology. The authors consecutively analyze all forms of mental activity: sensation, pain, perception, imagination, memory, thinking, knowledge and beliefs, emotions, volition, and different kinds and subdivisions of consciousness. In addition to this, B&H briefly review the history of philosophical ideas as manifested in the classics of neuroscience from Willis and Bell to Adrian, Eccles, and Penfield. They discuss methodological issues in the study of the mind-body problem and formulate a clear, though not uncontroversial, view on what is possible and impossible for neuroscience, as well as what is possible and impossible for philosophy. They close with an analysis of the several most influential approaches in the modern philosophy of mind, as represented in the recent books of the Churchlands, Dennett, and Searle. Thus, the authors do not leave out any of the important problems of interest for cognitive and behavioral neuroscience. Their main conclusions can be summarized as follows: (1) Most modern neuroscience is based on a confused and unnatural mixture of Cartesianism, traditional British empiricism (Locke, Berkley), and materialism. Rejecting the mind/body dualism of Descartes, modern neuroscientists simply adopted (and slightly adapted) everything else in his doctrine, simply re-ascribing all the presumable functions of the mind – to the brain (or even, following Chomsky, to the hybrid of mind-brain). From Locke and Berkley, neuroscience took on subjectivism, as well as the notion that in perception, imagery, and memory the mind-brain creates “images” or “copies” of external objects – good copies in perception, weaker in imagery, yet weaker in memory. (2) As a result of this philosophical confusion, a big part of neuroscience is busy with continuous attempts to solve nonexisting problems, i.e., problems that emerge solely because of the erroneous use of words, such as introducing metaphors that are then conceived of literally. Examples of such metaphors are “maps” (as if there were open books in the brain for orientation in space) or “representations” (the verb “to represent,” like “to give,” requires a dative: to whom?). The paradigmatic case of such a pseudo-problem is the so-called binding problem, which only exists for one who believes that what we see are not real objects in the world, but small pictures of these objects in the brain. Of course, such pictures cannot exist, because if they exist, the homunculus who observes them would have to have his own mechanisms of visual perception in order to recognize the images the visual cortex shows him, and so on, ad infinitum. But as soon as we understand that “images in the brain” is merely a metaphor, that in order to see an apple we need neither a realistic painting nor an abstract drawing of it in the occipital cortex, the fact that some neurons selectively respond to some features rather than to other features, does not present a problem. B&H point out that the first version of the “binding” problem was invented by Descartes himself. He argued that the structure in which the brain contacts the immortal soul must lie exactly on the midline because, if it is lateralized, the soul would receive