Flanagan, O. (1992): Consciousness Reconsidered. Cambridge/MA & London: MIT-Press. UK: œ 11.75, Continental Europe: $ 17.50.

As Flanagan remarks at the outset, many philosophers and researchers in the cognitive and neurosciences today believe that a naturalistic solution to the mind-body problem will eventually be found. Optimistic attitudes of this sort are usually inspired by the remarkable theoretical success so far achieved under the information-processing approach. The information-processing approach rests on a number of ubiquitous background assumptions. The most central of these is that treating human beings and their brains as information-processing systems may open precisely those levels of description under which the classical philosophical questions can be reformulated and finally answered. Many philosophers today try to treat traditional puzzles in this field - like the mind-brain-relationship, intentionality or self-awareness - by analyzing them in a top-down-fashion, i.e. by reformulating them in an information-processing terminology. In doing so one of their goals, usually, is eventually to naturalize those theoretical problems, to transform them into problems tractable by empirical science.

The standard approach in cognitive neuroscience, however, completely leaves out the question of how phenomenal consciousness could emerge from the flow of information in our central nervous systems. If we are looking for a functional analysis or an algorithmic description of a cognitive phenomenon, the question of wether this phenomenon actually is a conscious phenomenon at first sight seems irrelevant. And within current philosophy of mind itself a considerable number of authors seem to share a very strong "modal intuition", i.e. an intuition about what is always possible: No matter how rich any empirical description of a given type of system and the specific mechanisms subserving the assumed fact that it is conscious may be - it is always logically possible that this system is actually completely unconscious. There could always, as David Chalmers would say, be a Zombie Twin. This problem (Owen Flanagan calls it "conscious inessentialism") may have its root in confusing what beings like ourselves can mentally simulate with what is logically possible and in the simple fact that the term "consciousness" is a very ambiguous, idiosyncratic and ill-defined term of Western folk psychology, presently lacking substantial content. That general background-situation may lead one to return to the dead-end streets of metaphysics, skepticism or to a philosophical version of what Flanagan and others have polemically called "New Mysterianism". On the other hand there always is the alternative of eliminative materialism, the proposal of a terminological revolution - which basically is a cultural revolution. Our author, however, points out that this strategy might at present be premature and that the problem of consciousness can nevertheless be tackled. He proposes a general research programm, which he dubs "constructive naturalism": A broad interdisciplinary effort in constructing a naturalistic theory of consciousness, "coordinating all our informational sources at once" (3). And indeed one of the attractive characteristics of Flanagan's book consists in its combination of a thorough cognitive optimism and taking the problem of consiousness and all its intuitive and pretheoretical roots seriously. "We can make intelligible the existence of consciousness in the natural world. Against the eliminativist, I maintain that the concept of consciousness, despite its shortcomings, is needed, at least at the beginning of inquiry, to mark what is in need of explanation. Phenomenal, qualitative consciousness is what needs to be explained. Indeed, the method I propose for attacking the problem of consciousness requires that we use, but treat as revisable, our ordinary first-person modes for taxonomizing subjectivity." (2)

Another strength of Flanagan's book consists in touching a very large part of the canonical "hot spots" and central issues of the current philosophical debate on consciousness. He achieves this while never getting lost in overly technical considerations. He also makes the relevant interdisciplinary connections of the philosophical discussion clear without drowning non-professional readers in a flood of details or citations which are just too specific for the general audience. This fact will make the book attractive reading for a lot of other consciousness-researchers, e.g. in the neuro- and cognitive sciences. Consciousness Reconsidered is mostly pleasant and easy reading, but at the same time it successfully escapes the dangers of becoming one of those shallow pieces of philosophical generalism presently produced in increasing numbers by authors who never had a serious interest in a theory of mind, but who now, finding themselves in the "decade of the brain", suddenly feel the urge to jump the train.

Flanagan uses the second chapter to investigate claims that the concept of consciousness is just too inconsistent, empty or ambiguous to allow for a refining discussion which might yield serious theoretical progress. He comes to the conclusion that there is a minimal phenomenological core in the concept, not to be given up and capable of binding the many different semantic aspects together: "Conscious experience names the class of mental states or events that involve awareness. A conscious experience is a state such that there is something it is like to be in it." (31) Flanagan then goes on to point out the relevance of brain research to the project of an interdisciplinary theory of consciousness. He does so by introducing the reader to a number of basic but important new concepts like neural vectors, activation spaces and partitions of such spaces. He proceeds by introducing Craikian mental models and Calvinian Darwin Machines, drawing his reader's attention to the interesting fact that, although many of our "higher" cognitive capacities like conscious thought and the activation of higher-order forms of mental content seem to be serial processes in the sense of classical von-Neumann-cognitivism, we are systems which employ a massively parallel architecture in achieving their epistemic goals. Flanagan is right in observing that the functional profile of our own type of central nervous system might indeed turn out to be so complex and specific that we might have to seriously reexamine Putnam's early multiple-realization-argument against proposed type-identities between mental and physical states, which were so very important in getting functionalism off the ground at that time. Not many people seem to think in this direction today. But Flanagan is perfectly right in observing that a combination of a biologically plausible teleofunctionalism with a PDP-inspired microfunctionalism satisfying the "bottom-up-constraints" of an advancing neuroinformatics might well give us at least very informative (if not strict) mappings between mental events and brain processes and in this way lead the way back to new versions of good old identity-theory. "Massive parallelism is compatible with identity theory, and no philosophical song and dance about the possibility of Martians and robots with our kinds of minds should make us abandon the prospect that there might turn out to be some, perhaps many, interesting species-specific type identities." (47) In this sense even the most interesting phenomenological differentiation of all - the subjective segregation of self and non-self - should somehow be mirrored in the underlying processes on the level of information flow and neural activation patterns. This is possibly true of qualia like pain or subjective colour-impressions and their subserving processes describable by neural vectors too. At this point Flanagan introduces two useful new concepts: "experiential sensitivity" and "informational sensitivity". The latter property can be instantiated by an entirely unconscious organism, the former helps to further define the explanandum - subjective consciousness refers to experiential sensitivity. If a naturalistic theory of mind under the information-processing approach should be successful, it would have to reduce experiental content to informational content. Are there good reasons to believe that this is presently possible in a conceptually convincing manner?

In the following chapters we find a discussion of qualia, which again blends empirical data with elements of the current philosophical debate and a turn towards the more general problem of the subjectivity of the mental, exemplified by authors like Nagel, McGinn and Jackson. Again, the material is organized and presented in a way that makes it attractive to readers interested in a representative and comprehensive tour of the field, although on a second reading Flanagan sometimes does not do full justice to the analytical depth of the problems discussed. Take the very beginning of chapter 5 as an example: Is it really uncontroversial that "There is something it is like to be a subject of experience." ? Is this a grammatically correct sentence, a clear thought at all - or just an intuitively appealing fa‡on de parler? On the other hand in what follows a simple but popular misunderstanding of the goals of scientific theories is being dissolved, namely that it does belong to criteria for the quality of a description of the internal states of a system, say, the qualitative content of sensory awareness, that this description - just by the act of cognitively understanding it - must activate the respective states in other systems, i.e. enable them to perform a phenomenal emulation. Chapter 6 offers a critical discussion of McGinn's "cognitive-closure-argument" and tries to refute principled forms of skepticism. Chapter 7 discusses the possibility of an epiphenomenalist solution to the mind-body problem, the idea that phenomenal content might be causally inefficacious with regard to the generation of external behavior: Could it be that the reasons for the survival of our species lie exclusively in our ever-increased informational sensitivity to features in our environment, and that our capacities to experience fear and lust or pleasure and pain simply came along as "free riders", as Flanagan (132) puts it? Flanagan argues that the causal role of consciousness has been overestimated, but, after citing different neuropsychological evidences and discussing Ned Block's attempt to differentiate the concept of "consciousness" itself, as quite often in his book, steers a middle course and comes to the conclusion that epiphenomenalism is probably false, but that there are a lot of open empirical questions concerning its role in the fixing of informational content: "The biggest problem the epiphenomenalist faces is explaining how, given the massive connectivity of the brain, any feature as common, well-structured and multimodal as phenomenal consciousness could supervene on certain neural processes without having interesting and important causal effects in other parts of the neural network." (150)

In the remaining four chapters a second major topic gradually emerges from Flanagan's lively interdisciplinary considerations. It is a philosophical theme which will without doubt be of great interest to many of his readers and which will certainly constitute one of the central issues in the future of consciousness research, namely The Illusion of the Mind's "I" - to quote the title of his ninth chapter. Before this there had been a step back in time, back to William James' "master metaphor" of the stream of consciousness. The discussion of William James serves as a beautiful introduction to the problems of the continuity of consciousness and the unity of the phenomenal self. What makes the stream of consciousness one stream of consciousness for the system itself, what gives us a subjective identity, a sense of coherent and authoritative me-ness, of "personal sameness"? (167) And indeed it is counterintuitive how something so smooth and homogeneous as the general background of our phenomenal awareness should emerge from myriads of single events on the neural level, catching us in an illusion of continuity and naive realism. On the other hand we need an explanation of how discrete objects can be found in the "free waters" of consciousness, how object formation, scene segmentation and situated perceptual episodes can take place through completely natural self-organizing processes in our brains. First-person experience does not help at all, especially if we are looking for a convincing naturalistic explanation for the subjective "sameness" of the most interesting representational object, the mental model of the self: Is it an object in the stream or is it the bed of the river? The answer to both questions may be found in empirical work on temporal coding, as conducted by Christoph von der Malsburg and Wolf Singer. Should it actually turn out that it is the time-constants of neuronal algorithms which we have to investigate if we are searching for a naturalistic explanation of higher order coherence, object formation, constancy of mental content etc., then Owen Flanagan will have been one of the very first philosophers of mind having seen this and commented on it. Although his treatment of the binding problem is extremely brief and he only cites Crick and Koch, it demonstrates again how up-to-date and empirically well informed his treatment of the canonical philosophical problems surrounding the notion of consciousness is.

Flanagan also seems to see that the greatest theoretical achievement of a future theory of mind might be the final departure from the myth of the transcendental ego. Taking his start from William James he writes: "I will speak of the model of the self to refer to the highest-order model of the self that contains the various components of the material, social, and spiritual "me" as proper parts or aspects." (181) and "The posit of a mind's "I" is unnecessary. First, whatever is captured is an object, a model, a thing thought of or represented. The spiritual me, the material me, and the social me, even when integrated in thought, are things seen, not the seer. The obvious questions arise. What does the seeing? Who is the seer? The answer is that the whole organism, with its functional nervous system, is the seer." (182) and finally he tells his readers: "The system is dynamic. No part of the system resembles the soul, the ego, the self, or the "I" of traditional metaphysics. There is no mind's "I" that stands behind all experience as the condition of its possibility. This is because there is nothing that stands behind anything without something else standing behind it as the condition of its possibility. This is true even for the whole organism." (188) It must be clear that this kind of thetic criticism will probably not even raise the attention of historically oriented continental philosophers, German Kantians or specialists in transcendental idealism. So it cannot serve as a booster charge for the potential theoretical revolution mentioned above. But even for someone who thinks that a full blown naturalistic theory of phenomenal content - subjectivity, perspectivalness etc. included - along the lines of constructive naturalism is indeed desirable and soon within reach, it may appear that Flanagan has never actually addressed the core of the philosophical problem: What turns a mental model of the system into a mental model of the self? It is as necessary as it is easy to introduce a new terminology of mental self-representation, but a representational relation itself is - as for instance Husserl, a great defender of the transcendental ego, pointed out - not enough to constitute seeing something as something. What transforms a complex, multimodal activation pattern in our brain modelling the organism and a large number of its abstract, non-spatial and higher-order properties from a mere model of the organism into a phenomenal self? Simply introducing the term "self-model" might be a kind of terminological cheating or of intellectual self-deception - that is, if we take the problem seriously, if we actually want to know how that mysterious (and illusory) phenomenal property of "prereflexive self-intimacy" could come about as a purely biological phenomenon. The answer is missing, but it is implicitly given at many places in Flanagan's book: It is the simple fact that we are systems that never experience their self-generated mental models as models and so get caught in a naive realism on the phenomenal level of subjective experience. If this is true of the mental model of the self too, then it becomes clear how a phenomenal self could in fact emerge from the flow of information in our brains. On the other hand it must be noted that one of the main advantages of the way in which Flanagan develops his concept of a "self-model" lies in the fact that he clearly describes its constructive aspects and the impact which the social environment has on its content and functional profile. If the exciting project dubbed a Unified Theory of Consciousness by Flanagan in his last and concluding chapter should ever become a reality, then he has written a very enjoyable interdisciplinary introduction into its current canon of problems and helped a lot in defining the theoretical problems constituting its logical landscape.

Thomas Metzinger