1. What Does Consciousness Do?

We say we become conscious when we awake in the morning, but we are also conscious when we are asleep and dreaming. We are conscious when we become drunk, and for those of us who resort to mind altering drugs we are conscious even then. Consciousness is also present in people who are psychotic, or in fugue states in which they have lost their sense of who they are, yet are otherwise fully conscious.

Even though consciousness is present in all these different states, our strong tendency is to associate consciousness with that one form it assumes when we awake. We believe that we are truly conscious then. This strong tendency persists in cognitive science and neuroscience where it is referred to as reflective consciousness, or that form of consciousness in which we know we are conscious and what we are conscious of. In other forms of consciousness, reflectiveness presumably is absent or greatly attenuated. If this is the case, then answering our question stated in the title becomes complicated. It is not simply "What does consciousness do?”, but what kind of consciousness are we talking about that does it.

I will argue that there is only one kind of consciousness. To talk of more than one kind of consciousness is to conflate a variety of mental functions with what is primarily and exclusively an experience, or if one prefers, a unique and irreducible phenomenon. One might liken it to that moment when suddenly and unexpectedly a fireworks display turns night into day and we are transfixed and totally conscious of only one thing -- the exploding fireworks. At that moment there is no reflectiveness only consciousness, pure and simple.

We should not speak of dream consciousness, but of consciousness accompanying dreaming; similarly when we speak of waking consciousness, we should speak of consciousness occurring in the waking state. It is not consciousness that is reflective, but the functions of the state of mind in the waking state that are capable of reflection. One important implication of this unconflation of conscious experience from aboutness, and state is that it allows for an important role for unconscious processes for which there is a growing body of confirmatory evidence, both psychological and neurophysiological (Shevrin & Dickman, 1980/2003). Reflectiveness can occur unconsciously as well as accompany consciousness; decision making can occur unconsciously as well as consciously (Lau & Passingham, 2010).

When I use the term consciousness I am referring to a unitary phenomenon that is the same in its ontology, or nature, regardless of when and under what circumstances it occurs; adjectives qualifying consciousness itself are not required and are misleading. The adjectives belong to what I will call psychological states that may function entirely unconsciously, or at other times are accompanied by consciousness. Dreams are generated by certain psychological conditions instantiated by physiological conditions beautifully captured by our deepening knowledge of the sleep-dream cycle. We know that most dreams never become conscious, or that they may be conscious at the time of dreaming, but fail to leave a stable memory trace that can be recalled at waking, or it might leave a stable memory trace that simply remains unconscious but still capable of influencing other psychological states that do emerge in consciousness. In cognitive science these delayed emergences are referred to as implicit insofar as there is no consciousness of their true source. (This latter phenomenon will figure largely in my subsequent effort for a more complete answer to our question).

Two examples, one clinical and one experimental, might introduce this important phenomenon of implicit recall without consciousness of its true source. (For some readers the choice of a clinical example, and in particular one from psychoanalysis, would seem surprising. Its particular value in this context will I hope become apparent in our further examination of the issue of implicit recall.) A patient in psychoanalysis states that the reason why she would never want to become pregnant is that she could not tolerate the bodily deformations that accompany pregnancy. She felt this thought was important to understanding her conflicted sexual feelings and, as the session was drawing to a close, promised to continue with it in the next session a day later. When the next session arrived there was no explicit continuation of this line of thought, but at one point, almost as an aside, she said that she found herself thinking of how the heroine of Pearl Buck"s novel, The Good Earth, went out into a field to give birth, in point of fact a distorted rendering of why she went out into the field, and thus very likely already influenced by the underlying true source, or primed, the preferred term in cognitive science to refer to an unconscious influence on a conscious process,. But surprisingly this did not awaken the memory of the previous day; rather it was an implicit reference to her conflict over sexuality and pregnancy; the true source remaining, as we say, "unconscious" for reasons that are likely functional rather than organic in nature.

Cognitive investigators of memory have discovered a similar phenomenon in organic amnesiacs who cannot recall or recognize words from a list originally learned; but when presented with a word stem like win _ _ _ will more likely complete it with a word from the original list like window rather than a word equally likely but not on the list (e.g. winter). This apparently successful retrieval does not, however, result in eliciting any recall of the link to the list, just as the memory of the Pearl Buck heroine did not activate the personal pregnancy talk of the previous session. For the amnesiac the true source appears to be unconscious but for organic reasons, generally thought to be a lesion in a part of the brain called the hippocampus and related brain regions associated with memory formation, although another explanation will be offered below bearing on the nature and role of consciousness.

Experiments of the kind just described have given rise to a memory typology divided between declarative and procedural memory (Clark & Squire, 1998). Thus, memories of the kind, “Yesterday I recall having lunch with Bill at the Student Union” are considered to be episodic or autobiographical memories formed consciously and retrieved consciously by reflective consciousness, while memories of the sort as the Pearl Buck allusion, or the window completion are considered to be “implicit”, or procedural, lacking in any conscious awareness of source and not experienced consciously as a memory. Moreover, the claim has been made that reflective consciousness is essential for the formation and retrieval of autobiographical memories (Clark & Squire, 1998). According to this view in trace conditioning (described below) a new stimulus initially without aversive connotations acquires a negative meaning when repeatedly linked to an electric shock. Acquiring this negative meaning requires reflective consciousness of the relationship between the two stimuli, and this awareness of a relationship across an interval is, according to this view, at the heart of what is meant by autobiographical, or episodic memory: two or more events linked in a particular context, as my lunch with Bill.

This form of conditioning is referred to as trace conditioning because the initially neutral stimulus forms a memory trace that, during the acquisition of the conditioning, is related to the subsequently appearing aversive stimulus, the electric shock. The capacity to form this relationship between the memory trace and the later occurring shock, across the trace interval, is what requires a special form of consciousness, reflective consciousness. Conversely, the apprehension of this relationship cannot occur unconsciously. The implication of this role for reflective consciousness in forming autobiographical memories led Schachter (1998) to speculate that the ability to achieve trace conditioning might be used as a guide to determining whether an animal is capable of consciousness. If the Schachter position is in fact correct then the position taken in this paper is seriously undermined. A particular form of consciousness, reflective consciousness, is required and performs a function I have assigned to psychological states.

There is considerable controversy over whether consciousness is in fact required for trace conditioning, and more generally that it is required for the formation and retrieval of episodic memories. Given that the hypothesis for requiring consciousness is stated as a universal then even one well done study showing that it is not necessary would put the hypothesis in danger. In fact there have been several such studies, from our own laboratory and other laboratories. From our own laboratory Wong et al. (2004), a replication of a previous study by Wong et al. (1997), found in a trace conditioning paradigm in which the stimuli were presented unconsciously that trace conditioning can occur in the absence of conscious awareness of the relationship between the two stimuli. Reflective consciousness was not necessary, or for that matter any consciousness. The ability to form autobiographical memories was a function of a psychological state that could be unconscious or conscious; therefore it was not reflective consciousness that was required, but a psychological state having the ability reflectively to form an autobiographical memory. The upshot of these experiments for our thesis is that it is not consciousness that is essential to forming autobiographical memories, but an underlying psychological state. Consciousness remains the same kind of experience across many different states.

Lastly, in a comprehensive and brilliant examination of the role of consciousness in memory drawing upon a number of subliminal studies similar to the ones cited above, Henke (2010) concluded that, “consciousness is, in fact, a poor criterion for distinguishing between types of memory (p. 523).” In place of consciousness she provides convincing evidence that differences in processing modes can account for differences in memory like declarative and procedural. As she further states, “Importantly, conscious awareness … does not determine which memory system is engaged … and the term episodic memory can refer to both conscious and unconscious episodic memories (p.527).” Henke"s position considerably weakens if not outright disproves the Clark and Squire (1998) hypothesis requiring reflective consciousness for the formation and retrieval of episodic memories, thus its standing as a negative instance against a very different role for consciousness proposed in this paper loses its salience.

Let us now turn to considering what consciousness does do as consciousness.

2. The Central Role of Consciousness in Distinguishing Perception From Memory

We usually have no difficulty in knowing when we are perceiving and when we are remembering. In fact we have no difficulty in doing both at the same time -- perceiving one thing and remembering another even when the object involved is the same. We can see an old friend while summoning a memory of how he looked the last time you saw him that might lead you to say under your breath, “My how he"s changed”. Distinguishing a memory from a perception, simple as it sounds, I will argue was a hard won evolutionary triumph in which the emergence of consciousness played a key role, and continues to play that role.

Richard Gregory (1980, 1996) was one of the few psychologists who developed a way of understanding what consciousness does as consciousness rather than as a synonym for a psychological state. A considerable amount of his original research has demonstrated that perception is not simply a replica of an external object, but is fashioned by considerable influence from past experience to the extent that one could speak of a perception as an hypothesis of what the object is. This compelling research then led Gregory to wonder if a perception was compounded of so much from the past, how was the contribution from the present identified? To put it simply, past experience might lead one to the hypothesis that at the corner you will encounter a traffic light that might be red or green; but it will not be able to predict that at the moment you reach the corner whether it will be red or green. For Gregory this is where consciousness enters the picture. Consciousness functions to capture the present moment as when the fireworks explodes. It does so because it is particularly sensitive to qualia, those immediate, intense experiences of color, sound, smell, and touch that are much stronger than their counterparts in memory. Gregory acknowledged that he had no explanation of how consciousness accomplished this feat. He also recognized that qualia can be quite intense in dreams, but do not have the same present moment, perceptual result. Here he might have overlooked the fact that dreams are experienced as real, as nighttime hallucinations, and thus consciousness might be doing the same job, but in the absence of any external input, the present moment experience becomes an hallucination. This might also be the case with psychotic hallucinations in which the major input might be internal rather than external.

We will return to Gregory after examining two other attempts to discover what consciousness accomplishes apart from what psychological states accomplish. The first is Freud"s attempt. In his initial effort he conceptualized consciousness as modeled after a sense organ (Freud, 1900), except that the stimuli activating the consciousness sense organ came from within the person and were in the form of unconscious processes. Freud started with the assumption that all mental processes were initially unconscious and incapable of being directly experienced or intrinsically conscious. He proposed that they were organized quantitatively as brain frequencies lacking any qualitative character. Qualitative might also be read as lacking in qualia. Only when they impinged on the conscious “sense organ” were they transformed into qualitative experiences, much as light energy impinging on the retina is transformed from various frequencies into the qualia of conscious colors. Freud further limited the nature of this qualitative transformation to experiences of pleasure and unpleasure. In short, consciousness served the purpose of tagging or flagging ongoing unconscious processes as either agreeable or disagreeable, certainly a valuable function. For Freud, the criteria for pleasure and unpleasure related primarily to satisfying basic instinctual requirements. If the unconscious psychological state would result in a satisfying outcome it was tagged as pleasure consciously, and if not then it was tagged consciously as unpleasure. In either case, unlike Gregory, the focus shifts considerably away from sensation and perception to emotion and motivation. Later on, as if recognizing this omission, Freud added to consciousness the ability to provide “indications of reality” (Freud, 1911), but he never integrated these two functions, nor did he specify how the transformations to pleasure and unpleasure and how the “indications of reality” were achieved.

We next consider the important contribution of Edelman who formulated a comprehensive theory of brain and mind (Edelman, 1989) which in several respects parallel Freud"s broadening of the problem from perceptual qualia. Edelman focused on the appearance in evolution of what he refers to as primary consciousness which he distinguished from secondary consciousness. The latter is what we have already discussed as reflective consciousness and that I argued was a conflation of the experience of consciousness and the operation of a psychological state capable of reflection. But we can put this issue aside and concentrate on what Edelman meant by primary consciousness. According to Edelman, primary consciousness emerged as a means to provide what he refers to as a “value free” input that can modify those already existing organizations of past experience. By value Edelman meant the organism"s biological needs necessary to maintaining itself. In short, the kind of input that determines the light is red right now, not green as you had hoped and desired to be, or experienced the last time you were at that corner. Unless these past value organized memories could be modified by value free current inputs the organism would succumb fairly quickly in a world in which traffic lights are not subject to our value driven needs.

Our conclusion from this overview of three different views of consciousness is straightforward. Consciousness is our one adaptive resource that tells it as it is. Perception as Gregory has shown is greatly influenced by hypotheses and expectations formed on the basis of past experience. Memory as Schachter and others have shown is often an unreliable indicator of its true source. Thought, judgment, fantasy, and dreams, as well as emotions and desires, have each their own contribution to make to adaptive acts, but are all dependent on what consciousness contributes to correct or modify past experience simply by what it registers -- for Gregory the immediate qualia, for Freud whether it is pleasurable or unpleasurable, for Edelman whether it reveals what is value free.

3. A Generalized Tagging Function of Consciousness

The recent discovery of mirror neurons may tell us something about the role of consciousness as experience operating totally internally and thus demonstrating a generalized function not limited to the purely perceptual as proposed by Gregory. This general role is hinted at both by Freud and Edelman, but deserves a more detailed treatment.

Our account of what mirror neurons do as part of a complex system will be a simplification of a fascinating and complicated process (for more technical accounts see Jeannerod, 1997, Lindner, et al. 2010, Mukamel, et al. 2010). When observing someone throw a football the mirror neurons will be activated in much the same way as when we ourselves throw a football. How do we then distinguish between whether it is in fact not ourselves? Apparently there are two sets of neurons, one reflecting the observed experience and the other reflecting one"s own experience very likely related to proprioceptive feedback from musculature involved in actually throwing a football. Here we have two events associated with two different psychological states. One a perception of an external event embodied in our own musculature; the other, the experience of an intention and the execution of that intention mirroring the content of the perception also embodied in our own musculature. We would agree that it is important to distinguish between these two events, much as it would be important to distinguish between the expected traffic light and the perceived traffic light. It is precisely at this juncture that the tagging function of consciousness exercises its vital role. But in this instance consciousness is distinguishing between two internal events -- the movement of one"s own rather than another"s limbs as registered by the mirror neurons. We suggest that this is an instance of consciousness as experience tagging the difference between two internal events, and using the same means- the presence of qualia, in this instance proprioceptive feedback at some stage of motor activation.

But there are many other internal processes that must be distinguished from each other if our inner life is not to become a mish mash of images, fantasies, memories, perceptions, thoughts, dreams, and so on- and to make matters worse, often sharing the same content. We propose to generalize the mirror neuron model from motor acts to all of the categories of psychological states -- images, fantasies, memories, dreams, thoughts, and so on. In short, for each of these mental events as with the mirror neurons for motor acts, there is a feedback loop that enters consciousness, and once it enters consciousness it is tagged as that function. But most important, the feedback loop needn't enter consciousness; and if this happens the mental function involved is not tagged for what it is (e.g., memory, image, etc.) and when the memory or image returns and does become conscious it is not experienced as a memory or image as they were at the source, but as belonging to whatever the psychological state is at the time along with its accompanying state of consciousness. Our two examples illustrates how implicit memories may occur according to our view of the role of consciousness.

In the case of the amnesiac the supposition is that certain brain lesions produced a defect in episodic memory formation. We suggest that the defect may be in the tagging function of consciousness. In the absence of any tagging the words are simply registered as words, not as a memory of words, or an imagining of words. These words can be activated later when the word stems are presented, but they cannot be recognized as belonging to a previous experience because, although conscious, they have not been tagged as such because of the proposed defect in the tagging function.

In the case of the Good Earth example we hypothesize that a very different cause is at work resulting in the source being unavailable. The personal pregnancy thoughts of the previous day were at the time conscious and could be tagged as a memory. Yet the next day these thoughts entered consciousness implicitly and not referred to its previous source as an episodic memory. If we hypothesize that repression was at work, then we might further hypothesize that repression undoes the conscious tagging function. From a clinical standpoint, it is not unusual for repressed contents to emerge implicitly as in the Good Earth example, something that Freud called displacements -- a shift away from the original source not because of organic insult, but in order to avoid unpleasure that, according to Freud, will be detected by his unique sense organ model of what consciousness does.

The tagging of neuronal functions is not in itself offered as a new idea; what might be new is its application to the qualitatively different categories of psychological states such as perception, memory, dreaming, and so on. It is of interest, however, that neuronal memory tagging has been reported that tags one set of neurons for one memory function and another set for another so that, for example, a recently formed memory can be distinguished from its later version in delayed recall (Lesburgučres et al. 2011; Sweatt, 2011). Both synaptic and epigenetic (molecular) levels of neuronal organization are involved. These findings lead us to expect that the tagging function of consciousness might involve highly complex neuronal mechanisms varying in their nature depending on what functions are being tagged.

To briefly summarize: Implicit memories may occur because of 1) a failure in the tagging function of consciousness as in amnesia, 2) a removal of the tag as in repression, 3) no tagging occurs because the psychological state content has never been in consciousness, a condition found in subliminal perception in which the effects of the untagged content can be traced in so-called priming effects. In all three cases the true source of the implicit recall remains unconscious because the tagging function of consciousness has in one way or another been compromised. (See Shevrin, 1998 for a fuller account of these three pathways to implicit memory).

4. Conclusion

What does consciousness do? Something most important: It keeps the different categories of psychological states and what they are about distinct and separate, and identifiable by source, thus maintaining our internal life organized and stable, making it possible for us to distinguish past from present, imagined from real, dreams from thoughts and fantasies. What started in evolutionary terms as the capacity of consciousness to distinguish perception from memory, Edelman"s primary consciousness, developed into a general tagging function as psychological states became more complex and varied.