Consciousness is the greatest mystery. Its most fundamental nature has yet to be uncovered and while many discoveries about the intricate workings of the brain have shed light on the various information processing and sensory-perceptual operations, there remains a great controversy about where in this immense system consciousness emerges. As enigmatic a subject as this is, however, there can nevertheless be delineated some basic aspects or properties of consciousness which can inform researchers in their pursuit of its neural correlates. The first, and perhaps most important, of these is the availability of the contents of consciousness to voluntary control. If some datum processed in the brain can be voluntarily incorporated into an intentional strategy then that datum is said to be conscious. If it cannot it must be unconscious. It seems inconceivable to postulate some conscious knowledge which nevertheless cannot be used to direct action in non-pathological individuals. For instance, if I am aware of the presence or absence of some stimulus, and if this is helpful or harmful to my goals, there would be no grounds to assert that it could not inform my subsequent behaviour. This point is especially relevant in certain experimental methodologies that demonstrate a qualitative difference in the reactions of subjects to two different stimuli. Debner and Jacoby’s (1994) word exclusion paradigm is one such experiment where subjects who were presented with stimuli under one condition responded correctly by excluding that stimuli in a subsequent task, whereas subjects who were presented with the same stimuli under a different condition could not exclude it. These findings strictly constrain any explanation of this phenomenon which does not allow that one group of subjects was aware of the stimuli and the other was not. It stretches the bounds of the imagination to conclude that both groups were aware of the stimuli, but some aspect of the presentation conditions of one subset of these precluded its ability to inform subsequent behaviour. Indeed, this experiment highlights how central the property of voluntary control is to the common conception of consciousness.

The second important aspect of consciousness is that its contents can be reported, both internally and externally. If they cannot, they would not be said to be conscious. To be conscious of something is somewhat similar to being in a state of self-report. There is a ‘stream of consciousness’ whose contents are brought ‘to the surface’ in a process very similar to that of report, albeit an internal one which may or may not be declarative. If some datum in the brain is not available for report in a non-pathological individual, it seems fair to say that the individual does not have explicit knowledge of it. While implicit knowledge is possible, as for instance latent memories or subliminal percepts, the very fact that these cannot be reported is the basis for calling them unconscious. Testing for this aspect experimentally is more difficult than the first one we examined, however, because of the criticism that is always applicable to such experiments. For instance, Marcel (1983) and Greenwald et al (1996) demonstrated that stimuli which subjects could not report on nevertheless influenced their behaviour in subsequent tasks, indicating that these unreportable stimuli were processed to some extent, and thus that they are an instance of unconscious perception. However, such experiments which attempt to arrive at a dissociation between two measures of perception, one identifying the absence of conscious perception and the other identifying the presence of unconscious perception, usually rely on either subjective or objective measures of conscious perception, both of which are subject to some criticism. If a subjective measure is used to demonstrate the lack of consciousness, by relying on a subject’s report, it can be argued that this negative report may reflect a lack of confidence in responding or other bias arising from the uncertainty associated with the perception of the degraded stimulus. If an objective measure is used whereby subjects are given a forced choice upon which they perform at chance level, it can be argued that while minimising the conditions for conscious perception this procedure also minimises the conditions for unconscious perception and thus leads to an underestimation of the influence of unconscious perception.

Finally, self-reference is another of the fundamental features of consciousness which, according to William James, in his Principles of Psychology, is its most important aspect and the key to its elucidation. This aspect consists in the creation of two separate levels of representation in the brain, one of which deals with perceptual reality, and the other with the self. These two are then yoked to one another such that all perception and action become centred around this self-representation. The contents of consciousness are always of the form “I think so-and-so” or “I feel so-and-so” and not “thoughts of so-and-so” or “feelings of so-and-so”. This “I” is the mental symbol which signifies our representation of our selves. While this seems at least as fundamental as the previous two features of consciousness described, it presents far greater challenges in the elucidation of its mechanisms, since it is a higher order representation than perception, which operates relatively computationally upon sense-data incoming from sense-organs which can be controlled, and the effects of which can be traced. Representation of a self, on the other hand, does not seem to proceed as mechanistically from a clear and identifiable source but instead seems to emerge from the brain itself. Perhaps a good methodology for studying this aspect, at least to start with, may be to use stimuli which are self-referents, such as a photo of the subject whose brain is imaged while it is presented compared with the effects elicited by presentation of a picture of a stranger. Maybe this will show activation profiles which somehow relate to the brain areas that are responsible for representation of the self since the subject will undoubtedly recognise herself and so the brain areas which mediate this ‘self’ image must be activated in this process.

It is this impenetrability and almost ineffability of consciousness which so violently engages our research pursuits. But rather than reaching immediately for key to unlocking the mysteries of our phenomenal states, we may begin where progress is made possible by technology. Cognitive psychology must start by elucidating the causal structure of the brain’s computational functions. Philosophy can retreat to the realm of the phenomenal states for the time being, until science finds a method to examine even these most ephemeral of occurrences. Until then, we must be content with what the neuroscientists and the psychologists can offer.