THE COMMON DENOMINATOR OF CONSCIOUS STATES: IMPLICATIONS FOR THE BIOLOGY OF CONSCIOUSNESS Bjorn Merker, The Institute for Biomusicology, Mid Sweden University, Ostersund, Sweden Abstract: In order to distinguish the conscious state itself from its aspects and contents we need an answer to the question "if there is some- thing it is like to be conscious, what is it?" A succinct answer to this question is provided in the form of a common denominator of all conscious states. This characterization of the conscious state has implications for the systematic study of consciousness through its bearing on a number of concrete issues connected with the nature of consciousness and its relation to the biology of brains and their evolution. These are discussed with a view to delineating the characteristics of consciousness, suggesting the primary functional role of consciousness in the total economy of brain functions, and exploring the tractability of the problem of consciousness from the standpoint of ordinary science. Key words: Consciousness, subject, attention, memory, executive, decision, selection for action, behavioral priority, emotion, free will, naive realism. INTRODUCTION There is currently an upsurge of interest in the problem of consciousness from within both philosophy and natural science (Hameroff et al., 1996). One has to go back to the 1950s and the excitement generated by the discovery of the ascending reticular activating system (ARAS; Moruzzi & Magoun, 1949) to find a comparable level of serious interest in the issue, at the time reflected in conference proceedings with titles such as "Brain Mechanisms and Consciousness" (Adrian et al., 1954) and "Brain and Conscious Experience" (Eccles, 1966). Though there are significant points of continuity between the earlier and present concerns the hope pinned on the ARAS as the key to consciousness was largely premature and current interest in the problem constitutes in many ways a fresh start. In this situation it is worth noting that one reason the earlier ARAS- centered approach did not reach its goal was conceptual: the ARAS was the key to a dimension or variable or parameter of consciousness, namely alertness, "arousal" or vigilance, and not to the state of being conscious as such. Today, with data supplied by techniques such as functional brain imaging fuelling theory, there is similarly a danger of confounding the problem of consciousness as a state with one of its aspects, this time, namely, with the contents of consciousness. To avoid such errors it is necessary to have some means of conceptually characterizing the conscious state itself. That is, assuming that "there is something it is like to be conscious" (see Nagel, 1974) we need an answer to the question "if there is something it is like to be conscious, what is it?" A succinct formulation of a general answer to that question would provide a means for delineating the conscious state as such, an essential step in the systematic study of the problem of consciousness. IF THERE IS SOMETHING IT IS LIKE TO BE CONSCIOUS, WHAT IS IT? In trying to answer this question we shall adopt an indirect approach. We start by supplying a set of disparate brief sketches of the "something it is like" to be conscious that do make reference to content. We then abstract from them a single under- lying invariant of the state of being conscious that is independent of any particular content. We note that none of the sketches that follow provide adequate characterizations of consciousness by themselves. They are only meant to provide a heuristic basis for extracting a common denominator untouched by the inadequacies adhering to each sketch. We therefore ask for the reader's patience until we have arrived at that denominator. We turn first to the ordinary waking consciousness of human beings.[1] To be conscious in this sense means to find oneself in a three-dimensional panoramic world containing a potentially great multiplicity of seen, felt, heard, etc. objects among which the object we call our body occupies a privileged position around which that world extends. No special "feel" is needed for this instance to qualify as an instance of consciousness: it suffices that one find oneself surrounded by the world as we know it, or, in other words, to see and hear as we ordinarily do.[2] Nor is any particular level of complexity required of the world supplying the content of that conscious state: we can subtract one modality and channel capacity after another without for that reason altering the fact of consciousness itself. Normally, however, the experien- ced world is not only rich beyond easy verbal description but it re- mains stable despite the multiply nested coordinate transforma- tions undergone by our receptor organs as we make that world the arena of our operations with the help of combined eye, head and body movements. As we shall see, this may provide one possible clue to the "why" of consciousness, but for now, suffice it to note that finding oneself in the world - any world - is a sufficient condition of consciousness. If finding oneself in a world is a sufficient condition of conscious- ness, it is, however, not a necessary one. There are conscious states such as dreams and fantasy states that need not fit the description of finding oneself in the world, at least not in the ordi- nary waking consciousness sense of a spatially extended, stable 3-dimensional multimodal arena for potiential action. Some dream sequences do unfold within a highly structured world-like frame- work (though its stability is seldom long-lasting), while in others we are faced with fragmentary, mutable and insubstantial appearances that do not add up to a "world" as ordinarily con- strued. However fragmentary and fleeting those dream images may be, to the extent that they are experienced they are instances of consciousness, and the corresponding dream sequence a conscious state. The same applies, of course, to our third example: the appari- tions, voices, smells and other impressions engendered by hallucinatory or delirious conditions of whatever origin. The cascades of hideous demon faces that may inundate a victim of acute scopolamine intoxication represent a radical disorganization of the contents of conscious experience, but the fact of conscious- ness remains, inducing distress to the point of panic in the victim. Parenthetically we note that this example highlights the error of treating consciousness as a final flower of delicate functional sophistication on the part of highly evolved cognitive capacities. Rather, the fact that cognitive, perceptual and motoric brain function can be compromized by massive interference with basic transmitter systems to the point of inducing delirium without loss of consciousness argues, on the contrary, that consciousness is quite a robust and basic feature of the functional organization of the brain. Indeed, consciousness, by definition, is always the last function to disappear before...unconsciousness supervenes (see further Nikolinakos, 1994). The same circumstances counsel caution regarding explanations that make the state of conscious- ness dependent on delicately timed neuronal activity. We trust that the above three instantiations of the conscious state, from the ordinary waking consciousness of the adult human, over dream-states and fantasy, to hallucinations and delirium go some way toward dissociating the state of consciousness itself from its multiple potential contents. We are now ready to ask if there is any characteristic or attribute or condition that spans all these instances and other conceivable ones as the essential con- dition of consciousness as such. Is there, in other words, "some- thing it is like" to be conscious irrespective of drastically different potential contents of consciousness? We suggest that there is such a characteristic, namely "to be in the presence of" anything whatsoever. What it is like to be conscious is thus to be in the presence of something (in the sense of anything whatsoever), and the common denominator of all instances of consciousness the condition of "being in the presence of".[3] As we shall see, this formulation has consequences and implications for a number of issues in the biology of consciousness which have not generally been treated in the light of this characterization of the conscious state. THE RELATION AT THE ROOT OF THE CONSCIOUS STATE The above characterization of the common denominator of all conscious states as a condition of "being in the presence of" means that the conscious state has internal structure in the sense that in it some "x" is in the presence of some "y", with the relation "being in the presence of" designating the state of consciousness. As we have seen the "y" is the content or object of consciousness, and accordingly the "x" is that for which that content is present, namely the subject of consciousness. Note that the latter does not figure in this relation as inherently conscious: the state of consciousness is identified not with the subject but with the relation of "being in the presence of" obtaining between the subject and the object of consciousness. The problem of consciousness is thus neither being transferred from the state to one of its subdomains, nor is a much maligned "homunculus" being introduced under the pseudonym "subject".[4] Rather, a distinction has been drawn within the conscious state between two subdomains of that state: one called the subject and the other the content or objects of consciousness. The relation between subject and content that constitutes the conscious state is, morover, inherently asymmetric: the subject finds itself in the presence of objects and not the other way round. This is what Hume failed to take into account when he examined the contents of his mind for the presence of a self but found only perceptions and bundles of perceptions (i.e. contents of consciousness. See Hume, 1966). Contents of consciousness, objects of consciousness, are all one can ever be in the presence of in examining one's mind. The subject of consciousness can never be such a content or object since its occupation of the position of "being in the presence of" is what defines the relation that constitutes the conscious state.[5] If the subject saw itself, so to speak, who or what was in the presence of that content of consciousness? Or, as Schopenhauer put it metaphorically: "the eye sees all except itself". The fundamental distinction between subject and object in the constitution of the conscious state has, of course, a long history in philosophy. It was first recognized some time in the first few hundred years of our era in Indian philosophy, where the subject goes under the name of saksin ("witness") in Vedanta philosophy (Hiriyanna, 1973) and the "seeing part", as distinct from the "seen part" within the totality of consciousness, in the Alayavijnana doctrine of Mahayana Buddhism (Malalasekera, 1964). It was left to Kant to introduce an analogous distinction into the mainstream of Western philosophy (Kant, 1966),[6] where Schopenhauer (1958) treated the issue in plain language. As the most fundamental characterization of the structure of the conscious state, the distinction between subject and object has a number of consequences for our understanding of the nature of consciousness which have not always been clearly recognized and exploited in attempts at arriving at a coherent biological account of consciousness. We turn now to a number of these implications, since they may supply both heuristic guidance for and constraints on solutions to the problem of consciousness.[7] THE CONTINUITY OF CONSCIOUSNESS ACROSS TIMES AND CONTEXTS Within the relation characterizing the conscious state, the con- tents or objects of consciousness are variable, from moment to moment and from context to context. This variability defines the vast phenomenal richness and diversity of instances of conscious- ness. The subject, however, as that which finds itself in the presence of any content of consciousness whatsoever, is not characterized by such variability - at least not of a kind capable of entering consciousness (since that would require the subject of consciousness to be in the presence of itself). On this basis, the simple assumption of an unvarying subject of consciousness helps explain the continuity of an individual's con- sciousness over disparate contents, whether represented by different points in time - even when separated by periods of uncon- sciousness - or drastically different phenomenal content. However different such instances of consciousness might be, they all have one thing in common, namely the presence in them of the subject of consciousness, "unseen" but irreducibly present in each instance of consciousness. This shared aspect provides a natural basis for the continuity of consciousness across disparate instants. If experience enters memory under conservation of contextual relations (Davies, 1986), then the mnemonic record of experience would contain the "imprint" of the subject of consciousness as a tacit structural characteristic.[8] This could supply a natural basis for a sense of continuity of individual identity across varying experiences, contexts, and life circumstances, since a fundamen- tal similarity would adhere to both present experience and the memory of past experience in the form of the "perspective" of the subject of consciousness informing both. The subject would, as it were, match or fit with its own imprint on every instance of memo- ry, and would thereby be the common denominator of the entire longitudinal record of the individual's experience vouchsafed by memory. The resultant sense of continuity of identity over time and contexts would of course be a content of consciousness not to be confused with the subject of consciousness, though depen- dent on the latter for its formation. This sense of continuity of identity might in its turn provide a basis for the acquisition of more elaborate senses of identity and self-image. The "identity"-aspect of these senses and images, which have the status of contents of consciousness and nothing else, would be borrowed from, and derivative of, the more basic but at the same time more subtle (because "unseen") identity-function of the subject of consciousness. The subject of consciousness is the only feature or factor that is part of each instance of conscious- ness, and is part of every such instance throughout a lifetime without exception, and that is what lends it its role in the formation of identity at various levels of elaboration. THE SINGLENESS OF CONSCIOUSNESS If one reformulates the above exceptionless presence of the subject of consciousness in every instance of consciousness throughout a lifetime in terms of that subject's "access" to the contents of those instances of consciousness, it becomes clear that that access is nothing other than the state of consciousness itself, plain and simple. Our characterization of consciousness, in its independence of content, makes no restriction whatsoever on the possible contents that may enter the relation that defines the conscious state. They may be phenomenal, emotional, linguistic, propositional, decision-related or of any other kind. As long as the subject is in the presence of that content, whatever it is, the state of consciousness obtains. In this fundamental sense, then, an individual is equipped with only one consciousness. We hasten to add that this does not mean that the conscious state is homo- geneous or "all-or-none" in the sense of displaying no qualitative differences, gradations or internal structure beyond the distinction between subject and content, but only that distinctions such as those between "focal" and "marginal" consciousness (for which see below) do not affect the more basic attribution of the con- scious state's difference from unconsciousness to the relation of "being in the presence of" obtaining between the subject and the content of consciousness. If consciousness is indeed a relation between a subject of con- sciousness and the content of consciousness, it follows that a postulation of the existence of multiple "consciousnesses" (such as A(ccess)-consciousness and P(henomenal)-consciousness, for which see Block, 1995) involves the corrollary of multiple subjects of consciousness. This can be seen with the help of one of the examples given in Block (1995). At a certain point during an inten- se conversation one notices a persistent noise which has been present all along without one's being consciously aware of it. Block interprets this phenomenon to the effect that prior to noti- cing the noise one is P-conscious of the noise without being A-conscious of it. Since P-consciousness is supposed to be a state of consciousness, it follows from our analysis that a subject of P-consciousness must have been in the presence of the noise all along. But notice that at the same time the individual in question is supposed to be engrossed in A-consciousness of the content of the conversation, and since this A-consciousness ex- cludes the noise the subject of this A-consciousness cannot be the subject of P-consciousness (which is in the presence of the noise). Rather, it must be a separate subject of A-consciousness that is in the presence of the content of the conversation but not of the noise. Before accepting this scenario of multiple subjects of conscious- ness it is worth noting that there are alternative accounts of the psychological facts unencumbered by such consequences. For one thing it is perfectly conceivable that prior to noticing the noise one is in fact truly oblivious of it (i.e. not aware or conscious of it in any sense), despite its activation of inner ear auditory mechanisms. Dichotic listening and related experimental paradigms have demon- strated the power of attentional mechanisms in this regard (Broadbent, 1958; Treismann, 1960). In this case, consciousness is first completely taken up with the content of the conversation and then, through a redirection of attention (for whatever of a num- ber of conceivable reasons) it comes to encompass the noise. Alternatively, one is marginally aware of the noise prior to the point of becoming focally aware of it, a situation discussed further in the section on "consciousness, attention and memory" below. If so, the noise initially is at the margin of consciousness while the con- versation is in focal consciousness, upon which the noise moves to focal consciousness while the content of the conversation moves to its margin, a simple redeployment of attentional focus, in other words. These plausible alternative readings must be ruled out before there is any reason to give credence to the interpretation given by Block (1995), namely that there is such a thing as being "aware of the noise" (P-consciousness) without being "consciously aware" (A-consciousness) of it. Short of that, the P- and A-conscious- ness terminology remains a classification based on different con- tents of consciousness rather than different consciousnesses. This is not to say that a classification based on content lacks utility. On the contrary, it is essential to distinguish, say, dreams from hallucinations and these in turn from ordinary waking experience. For humans with their language-based cognitive capacities there would seem to be good reason for an additional distinction - somewhat related to Block's and following William James' (1890) distinction between a perceptual and a concepual order in human experience - between perceptual and conceptual kinds of conscious contents. Since the subject of consciousness' access to every content of consciousness is a fundamental characteristic of consciousness, it would be advisable not to use the term "access-consciousness" for any particular content-based variety of consciousness. In the more fundamental sense of the subject's access to the contents of consciousness, every instance of consciousness of whatever content is an instance of access-consciousness (for which see also Hardcastle, 1996). Finally we note that the singleness of consciousness outlined in this section taken together with the continuity of consciousness outlined in the previous section constitute the "unity of consciousness" discussed by Kant, 1966 (see also Searle, 1990). THE INERADICABLE ILLUSION OF FREEDOM OF WILL There is a third implication flowing from our delineation of the role of the subject in the constitution of consciousness. The subject is not only invariably in the presence of every actual content of con- sciousness over time, it also has potential access to every alternate content of consciousness in a given single instance of consciousness. We are referring, of course, to the alternate objects of consciousness among which attentional mechanisms and biasses select, such as those available to experimenter manipulation in dichotic listening and related experimental para- digms (Broadbent, 1958; Treismann, 1960; Posner & Snyder, 1975; Rees et al., 1997). The subject of consciousness thus occupies a "position" from which it has potential access to any and every conceivable object or content of consciousness whatsoever. It occupies, as it were, a central "vantage point" from which any potential object of con- sciousness might be accessed (see Kant 1966, p. 92; Schopen- hauer, 1958, vol. 1, p. 277). What is not thus accessible cannot be a potential object of consciousness because the subject cannot be in its presence. An intriguing possibility now suggests itself: assume that this central locus of global access is more than a passive member of the relation defining consciousness and in fact provides an arena on which potential objects of consciousness compete for the status of actual object of consciousness. Let the means of competition be any deterministic scheme of mutual inter- action (such as mutual inhibition on a global scale) capable of on- line selection of a temporarily prevailing winner among all competi- tors. In such a situation the subject of consciousness would always find itself in the presence of a content of consciousness which it had in a non-trivial sense itself been instrumental in selecting. Since intentions, goals etc. are among the potential contents of consciousness participating in the competitive process, the out- come of which is the actual content of consciousness, this means that any prevailing intention the subject finds itself in the presence of is one which it itself has, as it were, "selected". To the extent that such a situation is subject to reflective thought we suggest that it is bound to be interpreted as "freedom of will", since as far as the subject goes it always "has" (is in the presence of) the intentions it itself is the mechanism for selecting, albeit that the mechanism of selection is perfectly deterministic. Since these circumstances do not hold for any other locus of selection of the content of consciousness than the subject of consciousness, the ineradicable illusion of freedom of will provides strong circumstan- tial evidence for our conjecture that the domain of the subject of consciousness also provides an arena within which the ultimate competitive selection of the content of consciousness takes place. CONSCIOUSNESS, ATTENTION AND MEMORY In the previous two sections we have made reference to the selectivity of consciousness, that is, the fact that at any one time we tend to be aware of only a subset of the objects of conscious- ness potentially available at that time (Broadbent, 1958; Treisman, 1960; Posner & Snyder, 1975). This means that in addition to the partitioning of consciousness along its principal axis relating sub- ject to content of consciousness there is an orthogonal dimension of differentiation regulating the "access" of the subject of con- sciousness to its contents. It can be conceived of as a "tunable and mobile window" between subject and content of conscious- ness. It exhibits a profile from "focal" awareness at its center to "marginal" or "fringe" (the term is William James') access at its periphery, the specifics of which are shaped by factors such as vigilance level, task demands, and "informational load". It is mobile in the sense that its focal center can occupy alternate positions in the orthogonal dimension. Its specific position and gradient profile in a given case constitutes the attentional bias under which con- sciousness operates in a given situation.[9] It was suggested above that the subject of consciousness gains access to memory via its own contextual imprint on experience at the time it is being recorded in memory. If this is a necessary con- dition for the laying down of memory it implies that there is no memory without consciousness of the circumstances being stored in memory, a conjecture generally supported by the lack of good evidence for so called "subliminal learning", that is, learning without, or below the threshold of, consciousness (see Druckman & Bjork, 1992 and Shanks & St. John, 1994 as well as Greenwald et al., 1996). If one takes this conclusion to reflect a principle - "no learning without consciousness" - the tuning of the window of con- sciousness from focal to marginal awareness suggests that margi- nal statistical effects in such experiments might represent "margi- nal memory of marginal consciousness" rather than of "memory without consciousness". The kind of memory that is directly accessible "on command", on the other hand, would according to this interpretation reflect the contents of dominant "focal aware- ness". Anticipating a theme considered in the next section, a simple rationale for the principle of "no learning or memory without con- sciousness" suggests itself. The apparent limitation represented by the selectivity of consciousness is behaviorally reflected in our general inability to engage in disparate activities concurrently (within limits set by "informational load", for which see Broadbent, 1958; see also Rees et al., 1997). Both these limitations (for which see Treisman, 1960; Mandler, 1975; Norman & Shallice 1986) are in fact highly functional and have a common root: they serve the unity and integrity of behavior by directing it at one target or activity at a time (Allport, 1987). A hungry rabbit detecting an approaching fox while eating carrots does not eat a little and run a little, it runs a lot.[10] Given a sequential deployment of attention and behavior to one target and activity at a time, the nature of that target is not far to seek: it should be the best available estimate of the most pressing behavioral priority of the moment. The same estimate of the signi- ficance of the current situation (weighted by motivational state, past experience etc., see next section) employed for setting that behavioral priority would seem to be the ideal criterion for determi- ning what information to commit to memory as well. Information that is important enough to assume control over behavior should, if memory capacity is a resource that needs to be economized, have priority for committment to memory as well. Thus the content of consciousness simultaneously superintends behavior and enters memory. Hence, no memory without consciousness. MOBILITY AND THE FUNCTION OF CONSCIOUSNESS We have already pointed to the presence of the world as spatially extended, stable, three-dimensional, multimodal arena for potential action as a significant content of normal waking consciousness. The stability of that world is utterly unproblematic as far as the state of consciousness goes: it just lies there in all its richness awaiting our scrutiny or manipulation as the stable arena of our lives. Yet that stability is a highly derivative outcome of complex sensory and sensori-motor transformations of the first-order infor- mation supplied by our senses, and in our obliviousness, at the level of consciousness, to the fact of these transformation lies a key, we suggest, to the fundamental function or "why" of consciousness. There is a deep connection between the need for sensory guidance and the function of consciousness. The need for sensory guidance arises only with the need to orient and move about in an environment, that is, with animal mobility (since sessile forms can do little about their environment even if they should happen to per- ceive it). Yet the very same mobility that supplies selection pressures for the evolution of ever more elaborate sensory systems in mobile animals introduces a significant obstacle to the utility of the information supplied by those systems as a guide to orienta- ion and action. This is so because virtually every movement for which sensory guidance is needed introduces a displacement of the sensory organ supplying the information upon which that guidance is based, and thus confounds information drawn from the world with self-produced sensory information derived from move- ment of the sense-organs. The complexity of such confounding increases with the sophistication of the sensory and motor appara- tus evolved, which in our case involves factors such as dual eyes swiveling at fixed locations in the surface of a head in turn swivelling on a displaceable pivot (neck) attached to a trunk displaceable in all directions in space, as a highly simplified description of sources of self-motion in human vision. Natures way of dealing with the contamination of exteroceptive information by self-motion is to evolve a variety of compensatory mechanisms affecting sensory (e.g. saccadic suppression, see, e.g., Bridgeman, 1983) or motor (e.g. vestibulo-ocular reflex) pro- cesses in such a way as to cancel or subtract the effects of self- motion from the sensory information utilized in the guidance of action (see further Bell et al., 1997). The ultimate evolutionary anlage for this type of function would be the first central represen- tation of the kind of mechanism that allows an earthworm to discount cutaneous stimulation produced by self-motion as a stimulus for defensive withdrawal. In complexly mobile animals the compensatory mechanisms include the equivalent of coordinate transformations of spatially mapped sensory information between, say, retinocentric, head-centered and world-centered frames of reference. The fact that we are experientially oblivious of such com- pensatory transformations means that our consciousness is the beneficiary of the fait accomplis of their conjoint smooth functioning and thus lies at the distal, resultant end of their operation, which, judging by the result surrounding us on all sides, is an image of a stable world derived from multi-unstable sensory organs. There is a further category of processes related to mobility whose exclusion from consciousness provide a clue to the place of consciousness in the streams of processes mediating between sensory surfaces and behavior, namely motoric ones. Generally speaking we are not aware of the intricate complexity of motor control required to activate and coordinate the multitude of musculoskeletal adjustments involved in translating our intentions into behavior. Once an intention to act has arisen and a target for action has been selected, the execution of, say, a reaching move- ment takes place with smooth automaticity and without conscious- ness of the orchestration of the many muscle contractions and feedback adjustments that movement in fact involves. Consciousness is "located", in other words, at the end of or after the process through which the image of a stable world is extracted from moving exteroceptors, but before or prior to the processes through which intentions are unpacked into the details of musculo- skeletal control. In fact, consciousness is likely to be the sole function supplied with such informational characteristics, because these are ideal prerequisites for central, on-line, decision making regarding global behavioral priorities, and there is not room for more than one such function in the brain if behavior is to achieve its unitary, integrated character (Allport, 1987). Stripped of the complexities of transformation needed to synthesize a stable world and unencumbered by the details of motor execution, conscious- ness provides a summary display of an individual's current situation well suited, as we shall see, for on-line determination of the para- mount behavioral priority of the moment. Thus, for example, this summary display cannot dispense with the high degree of perceptual detail available to consciousness since that level of fine perceptual detail on occasion turns out to be crucial for determining behavioral priorities, as when one notices that one of the bars of a cage housing a starved and ferocious carnivore happens to be traversed by a hairline fracture. The above interpretation receives support from the fact of the inclusion in consciousness of a major category of information bearing directly on global decision-making processes, namely the vast array of emotional/motivational biassing variables experienced as feelings, moods and sentiments. These biassing variables span the range from physical pain over phenomena such as sexual arousal, anger, and fear to the attractiveness of the smell of fruity esters, the nagging feeling that there is something wrong with a certain argument, and a sense of sublimity. They include all the ways in which the meaning, significance as well as hedonic quality of objects or circumstances enter experience, whether on an innate or acquired basis. Though these biassing variables are truly diverse in their origins and mechanisms they have this in common that they bear on behavioral priorities and do enter consciousness. Particular- ly revealing in this regard are functions which straddle the fence between unconscious and conscious processes, such as those involved in the control of respiration. Under normal circumstances the adjustments of respiratory rate, tidal volume etc. needed to keep blood gases within normal bounds is automatic, effortless and unconscious. Should, however, the partial pressures of blood gases go out of bounds that fact intrudes most forecefully on consciousness in the form of an acute sense of panic. Why? Such a situation generally means that routine respiratory control no longer suffices but must be supplemented by urgent behavioral interventions (such as removing an obstruction or getting out of a carbon-dioxide filled pit), measures which ought to take precedence over all other concerns. It is at that point, when crucial action on the environment is of the essence, that blood gas titres "enter consciousness" in the form of an over-powering feeling. Physiological functions which never are subject to "remedial action", on the other hand, such as accomodation of the eye's lens to different target distances, are not represented in conscious- ness even when driven, as in this case, by exteroceptive informa- tion. Another example will help define further the functional role of consciousness as a modality for determining behavioral priorities. Reflex withdrawal of a limb occasioned by noxious stimulation precedes awareness of the pain, or at least is too fast to allow conscious guidance to have much influence over the act of with- drawal itself. Yet the pain forcefully enters consciousness never- theless. Preparedness for more elaborate remedial action than reflex withdrawal may of course help explain this fact, but an additional reason is supplied by the utility of experienced pain for determining future behavioral priorities. We have already sugges- ted that there is no memory without consciousness. Thus, in order for an experience involving noxious stimulation to affect future decision-priorities involved in the conscious guidance of be- havior (through any one of several conceivable ways for memory to affect the content of consciousness) the pain must enter con- sciousness and thereby memory, to become part of the mnemonic informational base affecting the setting of behavioral priorities in the future. Rare cases of congenital absence of the sense of pain bear witness to the importance of this mechanism. The above can now be summarized as follows: the conjecture that consciousness serves as a modality for selecting the most pressing behavioral priority of the moment in the light of current circumstances and past experience is supported by both the types of information that do enter consciousness and those that are excluded from it. It includes the right kinds of exteroceptive, motivational and mnemonic information to be the arena for such a central decision process, and it excludes what is not needed for such a function. Beyond a certain rudimentary level of sensori- motor complexity in mobile animals it appears to have been advantageous to segregate a variety of processes (perceptual, motivational, memory-based etc.) involved in a primary way in this central selection of behavioral priority from a variety of sensory, sensorimotor and motor processes involved in setting the stage for or following up on the decision process itself. The domain devoted to that central decision-process is, we suggest, the domain of con- sciousness. Consciousness is, in other words, the mode of operation of the mechanism through which a central decision-process responsible for on-line selection of the paramount behavioral priority of the moment accesses the information concerning current contingen- cies, motivational state and past experience on which that priority is based. This is what we suggest as the basic key to the nature, function and "why" of consciousness. We note that the decision process itself need be nothing other than the type of capacity- limited competitive selection mentioned in the previous section. This would mean that conceptions of the function of conscious- ness that link it primarily to capacity limitations of, say, selective attention or working memory (see Mandler, 1975; Posner & Rothbart, 1991; Baddeley, 1993) and those that identify it with decision- and action-related control functions (for which see Norman and Shallice, 1986; Allport, 1987; Hilgard, 1992) concern the same central function. Finally, the above formulation of the function of consciousness fits with the observation with which we started, namely the pivotal position of the subject of consciousness within the relation that constitutes of the conscious state. More generally we note that the above formulation explicitly identifies consciousness with a biological function for which there is every reason to expect a "neural implementation" in the brains of complexly mobile animals. To clarify further the bearing of that formulation on the search for such an implementation we need to consider an issue implicit in all that has gone before, an issue which must, however, be made explicit in order for an analysis of consciousness to fully constrain the search for its neural implementation. We are referring, of course, to the issue of naive realism. THE INERADICABLE ILLUSION OF THE OBJECTIVE WORLD OF REALITY, OR WHY NAIVE REALISM IS INCOMPATIBLE WITH A SCIENTIFIC APPROACH TO THE PROBLEM OF CONSCIOUSNESS To paint the issue of naive realism in the starkest possible colors we now suggest that the functional activity of conscious- ness, construed above as a specialized domain of central function related to behavioral decision-making within the total economy of brain-functions, is what everyone knows as their own presence in the world of objective reality. The distinctions between external world and body, and between body and "inner life" are, in other words, internal distinctions within the domain of consciousness, constituting the mode of organization of that domain. That domain in its entirety is in turn a subdomain of the total activity of the brain. To put it even more pointedly: the world that lies in front of and surrounds and in fact contains that which we take to be our eyes is in fact contained, in the form of neural activity, in brain tissue located behind the physical eyes (inherently inaccessible to consciousness) which actually supply that tissue with photic information. The person who, confronted with this claim, exclaims "but how can that" (pointing out at the world) "fit inside this?" (pointing to his or her head), is simply giving evidence of the persuasiveness with which the domain of consciousness performs its function as we have construed it. Such a person is probably under the natural impression that the head to which he or she is pointing is some- thing other than a cerebral image in three dimensions and multiple modalities which forms part of an image called the body that is the central object of an exquisitely structured multimodal three- dimensional cerebral image-space called the external world contrived with maximal realism for the benefit of a central decision- process associated, for logistic convenience, with that central image-body around which that image-world is deployed. (We here use image in the sense of representation, broadly construed, and note that it was Schopenhauer who first pointed out that not only the world, but this very body too, is an image in the brain. See Schopenhauer, 1958, vol. II, pp. 273 & 271). In the contrast between this version of the ontological status of the external world, including the body, and the naive realist version which interprets them as our brains were evolved to have us interpret them, namely as mind-independent physical realities to which we, somehow, have perceptual access, lies the crux of naive realism in its bearing on the problem of consciousness. The naive realist stance is the stance for which the brain's organization, and more specifically that domain of it we have analyzed under the label consciousness, was designed. It is the fundamental presupposition upon which the decision-making function of consciousness must be based. The objects and events with which consciousness presents us must be taken to be real, or the entire motivational underpinnings of behavior collapse. And they are taken to be real for the simple reason that the objects of consciousness are the only ones we ever have the privilege of being in the presence of. We lack experiential access to any other kinds of objects than the ones to which conscious- ness gives us access, and so they constitute our reality by default, as it were. Nor is there any reason to mourn this fact, since the honing action of natural selection has ensured that they are, in as far as they are contents of normal waking conscious- ness and making allowance for marginal exceptions such as per- ceptual illusions, veridical reflections of circumstances in the physical universe of which, as we have seen, they are distantly derived representations exactly because of the necessity to reconstitute those circumstances by indirect means on the basis of arrays of biological receptors etc. Exactly because natural selection has insured that the central representational spaces devoted to consciousness display veridical reflections of select aspects of the physical world the information they supply is a reliable guide in all practical matters and a good starting point for the entire gamut of natural science concerned with investigating the constitution of the physical world. But when it comes to uncovering the constitution of the mecha- nism by which consciousness supplies us with that veridical reflection it would of course be fatal to forget about the irremediab- ly indirect, representational, means by which it does so. No progress can be made in this regard on the basis of persistence in the stance of naive realism, since for it the three-dimensional stable world that surrounds us is not an elaborate display within the locus of consciousness but a mind-independent physical reality to which we, somehow, have sensory access. The stance of naive realism only has to be stripped of its onto- logical claim, namely that the world we confront directly in experience is (rather than is a representation of) a mind- independent physical universe, to appear in its true light, namely as a first-hand descriptive report from inside consciousness. The immediate perceived presence of the world on which it insists is a good description of the directness of access to the contents of consciousness afforded the subject of consciousness by the fundamental structure of consciousness as a compartment of cerebral function, according to our analysis. The stripping away of all evidence of the indirect and derivative nature of the central representation of the world from that representation is, in a sense, the very point of consciousness as we have described it. The sensori-motor transformations that go into the synthesis of the perceived, stable world had to be excluded from consciousness in order to make the illusion of mind-independence adhering to that world compelling. The naive realist is thus our primary exhibit in support of the structure of consciousness we have outlined above. He is simply unaware that the world he inhabits is a pinnacle of neural evolution on earth - a non-Cartesian theatre of reality so seamlessly constructed that even a professional thinker here and there has difficulty conceiving of it as anything other than the immediately present physical universe itself. However, beyond assisting us in the capacity of witness to this state of affairs as an endemic con- dition of consciousness, the naive realist can give little further help in exploring the how and why of consciousness. To the extent that he can be persuaded to take up the problem at all he tends to identify consciousness with certain restricted and problematic subdomains of the contents of consciousness such as "thinking", self-consciousness, an "inner life", "qualia" and the like. Generally they have this in common that they exclude the world that surrounds us from the domain of the contents of consciousness, because this world that we see and touch, the naive realist knows, is the physical universe itself rather than a representation of it. One requirement for progress on the problem of consciousness and its "neural implementation" is therefore abandonment of the naive realist perspective as far as its ontological claim is con- cerned. The issue has momentous consequences: what is in- cluded in and what excluded from the domain of contents of con- sciousness bears directly on the search for "neural implementa- tions" of the resultant characterization of the nature of the conscious state. The problem of consciousness is, in short, the only scientific problem for which the ontological stance of naive realism cannot serve as at least a reasonable starting point of inquiry, because to start from that stance in this case means to misconstrue in a fundamental sense the very subject matter to be investigated. On the topic of the ontological status of its own con- tent, and that of the experienced world in particular, conscious- ness is designed for massive and systematic deception. CONCLUSION Thus we suggest that progress on the problem of conscious- ness requires abrogation of an assumption which has been part of ordinary science so far, though not because it is in any way essential to that enterprise but simply because it has not caused any damage in its generally tacit presence in our investigations of the world of nature - until, that is, the question of the nature of con- sciousness is broached. That assumption is the naive realist assumption - roundly supported by the workings of conscious- ness - concerning the ontological status of the experienced world. Abandoning that assumption we can proceed to ask questions regarding possible modes of neural implementation of the nature and characteristics of consciousness. Of those characteristics the most fundamental one is the partitioning of consciousness into the sub-domains of subject of consciousness and contents of con- sciousness, yielding the asymmetric relation of "being in the presence of" between the former and the latter as the universal defining characteristic of consciousness. Proceeding on this basis we trust that elucidation of the nature of consciousness will turn out to be a difficult, but not insoluble, biological problem. ACKNOWLEDGMENTS: I am indebted to the late Eugene Sachs for the conceptual back- ground to this paper as well as for the "hungry rabbit" and "respiratory" examples. The support of the Province of Jaemtland Research Council is gratefully acknowledged. FOOTNOTES: 1. All our examples will be drawn from human experience, not because we deny consciousness to animals, but because ours is the only type for which we have direct evidence, albeit of a peculiarly limited nature because of the fundamental fact that we have no direct knowledge of other minds or consciousnesses. 2. This point is crucial and has been well emphasized by Dretske, 1995. 3. To readers who wonder whether this is the "in the presence of" which figures in Heidegger's writings, I can only say that my formu- lation is expressed in the terminology of ordinary language, and that the obscurity of Heidegger's language places obstacles in the way of determining what, if any, correspondence there might be between the two usages. My characterization was, in any case, made without knowledge of Heidegger's use of the phrase and should not be identified with it. 4. Note that no attributes have as yet been predicated of this subject: nothing has been said that in any way requires it to duplicate the structure or capacities of the being endowed with consciousness. It is no more than a term in a relation proposed to capture a universal characteristic of all conscious states, and might for all we know so far be represented by a single geometric point. 5. Specifically the subject of consciousness must not be confused with any kind of self-image whatsoever. The apprehension of any such notion or image, whether based on our body, appearance, personality, roles, or what not, is an instance of consciousness in which the subject of consciousness finds itself in the presence of that particular content of consciousness. See further the section on the "continuity of consciousness" below. 6. The abstruse terminology in which Kant clothed and explained his distinction (he called the subject of consciousness the "transcendental synthetic unity of apperception") may have con- tributed to its not making an impact on Western philosophy com- mensurate with its importance. Apparently it was of no avail that Schopenhauer explained the concept in plain language under the name of "pure subject of knowing". 7. We note in passing that if consciousness has internal structure it cannot be a psychological or any other kind of primitive. As composed of parts or sub-domains obeying a determinate relation (that of "being in the presence of") consciousness is a composite and hence is at least one step or level removed from the realm of irreducible primitives, and perhaps a multitude of levels thus removed. This weakens extreme versions of pessimism regarding the tractability of the problem of consciousness such as Chalmers' (1995) suggestion that consciousness be treated as a fundamental entity of nature on a par with the fundamental forces of physics. 8. Since so called "recall" or "retrieval" of memory proceeds from the vantage point of the present - the subject's vantage point - the same subject's contextual "imprint" on memory would provide a first and fundamental contextual cue in "retrieval". The subject of consciousness would, accordingly, be the ultimate functional pivot of the memory system. 9. Parenthetically we note that the mechanism of the tunable and mobile window can readily be combined with or derived from the mechanisms of competitive selection discussed in the previous section. A proposal for such a mechanism can be found in Merker, 1971. See also Taylor, 1996. 10. The "categorical" nature of behavior argues against any scheme of brain function (such as that of Dennett & Kinsbourne, 1992) that attempts to do without a central unitary decision- process. REFERENCES: Adrian, E.D. et al., Eds. (1954), Brain Mechanisms and Consciousness (Oxford: Blackwell Scientific Publications). Allport, D.A. (1987), 'Selection for action: Some behavioral and neurophysiological considerations of attention and action', in: Perspectives on perception and action, H. Heuer & A.F. Sanders, eds., (Hillsdale, N.J.: Erlbaum). Baddeley, A.D. (1993), 'Working Memory and Conscious Awareness', in: A.F. Collins & M.A. Conway, eds., Theories of memory (Hove, U.K.: Erlbaum) Bell, C., Bodznick, D., Montgomery, J. & Bastian, J. (1997), 'The Generation and Subtraction of Sensory Expectations within Cerebellum-like Structures', Brain, Behavior and Evolution 50 (suppl. 1): 17-31. Block, N. (1995), 'On a Confusion About a Function of Consciousness', Behavioral and Brain Sciences 18: 227-88. Bridgeman, B. (1983), 'Mechanisms of Space Constancy', in: Spatially Oriented Behavior, A. Hein & M. Jannerod, eds., (New York & Berlin: Springer Verlag) Broadbent, D.E. (1958), Perception and Communication (London: Pergamon). Broadbent, D.E. (1971), Decision and Stress (New York: Academic Press). Chalmers, D.J. (1995), 'Facing Up to the Problem of Consciousness', Journal of Consciousness Studies 2: 200-219. Davies, G. (1986), 'Context Effects in Episodic Memory: A Review', Cahiers de Psychologie Cognitive 6: 157-174. Dennet, D. & Kinsbourne, M. (1992), 'Time and the observer: The where and when of consciousness in the brain', Behavioral and Brain Sciences 15: 183-247. Dretske, F. (1995), Naturalizing the Mind (Cambridge, MA: MIT Press). Druckman, D. & Bjork, R.A., eds., (1992), In the Mind's Eye: Enhancing Human Performance (National Research Council). Eccles, J., Ed., (1966), Brain and Conscious Experience (New York: Springer Verlag). Greenwald, A.G., Drain, S.C. & Abrams, R.L. (1996), 'Three Cognitive Markers of Unconscious Semantic Activation', Science, 273: 1699-1702. Hameroff, S., Kaszniak, A. & Scott, A., eds., (1996) Toward a Science of Consciousness: The First Tucson Discussions and Debates (Cambridge, MA: MIT Press/Bradford Books). Hardcastle, V.G. (1996), 'Precis of Locating Consciousness', Behavioral and Brain Sciences (forthcoming). Hilgard, E.R. (1992), 'Divided Consciousness and Dissociation', Consciousness and Cognition, 1: 16-32. Hiriyanna, M. (1973), Outlines of Indian Philosophy (Bombay: Allen & Unwin (India)), pp. 342ff. Hume, D. (1966), An Enquiry Concerning Human Understanding and A Treatise of Human Nature, selections edited by Eugene Freeman, 2nd ed., (La Salle, Illinois: Open Court), pp. 257-258. James, W. (1890), Principles of Psychology, New York, Holt. Reprint: Dover, 1950. Kant, Immanuel (1966), Critique of Pure Reason, F. Max Mueller transl., (New York: Doubleday), pp. 105ff. Malalasekere, G.P., ed., (1964), The Encyclopedia of Buddhism (Ceylon: Government of Ceylon Press), vol. 1, p. 386, note 5. Mandler, G. (1975), 'Consciousness: respectable, useful, and probably necessary', in: R. Solso, ed., Information processing and cognition: The Loyola Symposium (Hillsdale, N.J.: Lawrence Erlbaum Associates). Merker, B.H. (1971), 'The Nucleus Reticularis Thalami: A Central Mechanism of Attention?', unpublished tutorial report, Dept. Psychol., Queens College, CUNY, Jan. 16, 1971, available on request from the author. Moruzzi, G. & Magoun, H.W. (1949), 'Brain stem reticular formation and activation of the EEG', EEG. & Clin. Neurophysiol. 1: 455-473. Nagel, T. (1974), 'What is it like to be a bat?', Philosophical Review, 83: 435-50. Nikolinakos, D. (1994), 'General anesthesia, consciousness, and the sceptical challenge', Journal of Philosophy 2: 88-104. Norman, D.A. & Shallice, T. (1986), 'Attention and Action: Willed and Automatic Control of Behavior', in: R. Davidson et al., eds., Consciousness and self-regulation, Vol. 4 (New York: Plenum). Posner, M.I. & Rothbart, M.K. (1991), 'Attentional mechanisms and conscious experience', in A.D. Milner & M.D. Rugg, eds., The neuropsychology of consciousness (New York: Academic Press). Posner, M.I. & Snyder, C.R.R. (1975), 'Attention and cognitive control', In R. Solso, ed., Information processing and Cognition: The Loyola Symposium (Hillsdale, N.J.: Lawrence Erlbaum Associates). Rees, G., Frith, C. D. & Lavie, N. (1997), 'Modulating Irrelevant Motion Perception by Varying Attentional Load in an Unrelated Task', Science, 278: 1616-1619. Schopenhauer, A. (1958), The World as Will and Representation, E.F.J. Payne transl., two vols., (New York: Dover). Searle, J.R. (1990), 'Consciousness, explanatory inversion, and cognitive science', Behavioral and Brain Sciences 13: 585-706. Shanks, D.R. & St. John, M.F. (1994), 'Characteristics of Dissociable Human Learning Systems', Behavioral and Brain Sciences 17: 367-447 Treisman, A.M. (1960), 'Contextual Cues in Selective Listening', Quarterly Journal of Experimental Psychology, 12: 242-248. Taylor, John G. & Alavi, F.N. (1993), 'Mathematical Analysis of a Competetive Network for Attention', in: Mathematical Approaches to Neural Networks, J.G. Taylor, ed., (Elsevier Science Publishers).