Marshall Nirenberg & the genetic code Sir Frederick Gowland Hopkins: biochemistry and vitamins

What Does Implicit Cognition Tell Us About Consciousness?

A report on the first conference of the Association for the Scientific Study of Consciousness
Held at the Claremont Colleges, Claremont, California, 13-16 June, 1997.
Conference Theme: What Does Implicit Cognition Tell Us About Consciousness?

Report by Nigel J.T. Thomas (California State University, Los Angeles)
To email me, click here
Published in the Journal of Consciousness Studies (4 #5/6) 1997 pp. 393-396.
©Imprint Academic

To contact the author, click here

There was a brief inaugural session of the Association for the Scientific Study of Consciousness during the Psychonomic Society Conference in Los Angeles in November 1995, but the first full conference of the Association was held this June in the very pleasant surroundings of the Claremont Colleges. Being at this conference was very different from being at Tucson II the previous year. This was a less ballyhooed, more intimate event, maybe less exciting, and less intellectually eclectic, but also perhaps more conducive to serious scientific exchange. Certainly the roster of speakers was replete with luminaries of the consciousness studies movement, and highly respected names from psychology, neuroscience, and philosophy: Christof Koch, Bernard Baars, Ned Block, Philip Merikle, Daniel Schacter, Larry Jacoby, Walter Freeman, Valerie Hardcastle, both Churchlands, Melvyn Goodale, Owen Flanagan . . . to unfairly pick out just a few.

The Association's membership rules, and its very name, are designed to focus attention toward the less controversial and more experimentally amenable aspects of consciousness research, and away from some of the more unconventional ideas and intractable issues encountered elsewhere: even David Chalmers scarcely mentioned the 'hard problem', and, despite Stewart Hameroff's presence, the relevance of quantum theory received little attention. The conference sessions were dominated by neuroscientists and experimental, cognitive psychologists, and although there was a good leavening of philosophers, they were mostly wearing their 'scientific under-laborer' hats, rather than their more critical or metaphysical ones. I think there is an important place for both sorts of issue, and both sorts of meeting, within the consciousness studies field, and this conference appeared to succeed very well in its intended terms.

The conference was organized around a particular question – 'What does implicit cognition tell us about consciousness?' – and, although not all the presentations spoke to this issue, and a few seemed to treat it essentially as an afterthought, this organizing theme served even more to focus attention on some of the more scientifically accessible aspects of consciousness. 'Implicit cognition' is when some sort of knowledge or information, whether from perception or memory, clearly influences a person's behaviour, even though they have no conscious awareness whatsoever of that information, or that they know it. One of the best known and most striking examples is that of 'blindsight', where people with brain damage in part of V1 report being quite blind in the corresponding area of their visual field, but can nevertheless accurately point to visual targets in this area when asked to 'guess'. There were several presentations dealing with blindsight, ranging from philosophical conceptual analyses the phenomenon (from William Dibrell and from Oliver Kauffmann) to a report on ingenious experiments to confirm its occurrence in monkeys (from Alan Cowey & Petra Stoerig).

However, as we learned, there is much more to implicit cognition than blindsight. Not only do other sorts of brain lesions give rise to other effects of this type (e.g. as reported by David Milner), but numerous experiments demonstrate its effects in normal subjects too. Indeed, it would seem that much of our ordinary functioning depends on implicit cognition; very probably the lion's share of our behaviour is guided much more by 'implicit' than by conscious knowledge. And a good thing too! Often, it seems, when consciousness is deceived, say by some sort of perceptual illusion, implicit cognition ensures that we actually respond to what is really there. Bruce Bridgeman reported that, even though subjects can be induced to misreport the positions of objects (through induced motion, static position illusions, etc.), if asked to jab or point at the object they will do so quite accurately. He interprets this as implying the existence of two distinct spatial representational systems, one to handle conscious experience and another, quite separate, that subserves visually guided action.

But if there are two such systems, they do not just differ in the accuracy of the information they represent. In other experiments, subjects seem implicitly to learn successful strategies for some task, whilst having no idea of what they have learned, or of why it would be relevant. For example, Frank Durgin & David Lewis reported an experiment in which subjects moved a mouse cursor around a computer screen in order to reveal a hidden target. They were told, and clearly continued to believe throughout the experiment, that the target was in a predetermined random position that they had to find. In fact, the position of the target was not predetermined; rather its appearance was contingent on the lengths of the sweeps across the screen that the subjects made with the cursor. It was set so as to be more likely to appear after a long sweep than a short one. The subjects 'learned' to make ever longer sweeps as they undertook repeated trials of this task, but they had no conscious awareness of adopting this strategy, nor of why it would be appropriate.

Despite all the efforts made to confine the discussions to the experimentally tractable aspects of consciousness, zombies, which have become a sort of emblem of the intractability of the 'hard problem', nevertheless appeared in the titles of two papers. Having a special interest in zombies, I made sure to listen to them; but these turned out not to be the scary sort of zombie at all. These were not the non-conscious beings functionally identical to conscious humans which (if they really could exist) seem to frustrate all possibility of a full physicalist understanding of consciousness. Instead they turned out to be quite benign subsystems hypothesized to exist within conscious humans, carrying out the functions of implicit cognition whilst consciousness busies itself elsewhere. Ullin Place, a true pioneer of consciousness studies (his seminal 'Is Consciousness a Brain Process?' appeared in 1956, in the depths of the Behaviorist era) spoke on 'Consciousness and the Zombie Within', setting out a rationale for the sort of model that seemed to inform much of the experimental work reported. He argued for a non-conscious subsystem, based in the mid-brain, whose function is to protect consciousness from overload by screening and handling most routine inputs, only passing them on to consciousness if they are in some way 'problematic'.

Axel Cleeremans & Luis Jimènez, on the other hand, in 'From Zombies to Star Trek's Commander Data', argued against such a view. They rejected both 'zombie within' theories and the standard alternative, whereby (as for the TV android Data, who can introspect his own computational states) all cognition is regarded as 'really' conscious, even if the relevant experiences may sometimes be impossible to report for some reason. The choice between these extreme alternatives, they argued, is a false dichotomy arising from an explicit or implicit adherence to a symbolic picture of cognitive computation. Connectionist thinking, they claimed, allows us to escape this dilemma, and makes it clear why knowledge and learning is ordinarily implicit. This takes most of the mystery out of implicit cognition, and shifts the burden of puzzlement to where it surely belongs: to the question of how and why cognitive systems, natural or artificial, can ever produce something so remarkable as consciousness.

In fact, listening to all these accounts of the wonderful capacities of implicit cognition one might wonder why we ever are conscious. If focusing on implicit cognition is not simply a way of avoiding direct confrontation with consciousness itself, it would seem to be a way of trying to uncover its function. But is it right to think of consciousness as having a function? Surely it does not do something for a person, as their heart pumps their blood or their occipital cortex analyses visual input; rather, the consciousness is that person. The person remains if their heart is replaced by a mechanical pump, or even if their visual cortex is ablated, but if consciousness is annihilated the person goes too. I do not mean to imply that the study of implicit cognition is irrelevant to understanding consciousness, but I do fear that its relevance may be less direct than some might think. It is not normally particularly useful for a person to know what chemicals their liver is currently processing, and consequently evolution has not provided us with any easy way of finding this out. Why should we expect things to be different regarding the information that our brains are processing? Why should we expect to directly experience the computations going on in our brains any more than the processes in our livers or spleens?

The answer, I would suggest, is that we too easily assume that the mind just is the brain, and that mental processes just are the processes being carried out by the neural computer. From such a perspective, the fact that many of these processes are not consciously experienced is at first puzzling, and, when it is taken on board, when we realize that being an informational process in the brain is not enough to make something conscious, it leads to a bootless search for the mysterious X factor that raises some select few of these processes to the conscious level. Perhaps there is no such factor; perhaps none of the brain's information processing is conscious as such; if cognition is defined to mean computational information processing, then all cognition may well be implicit. We should consider the possibility that the relationship between neural information processing and conscious experiencing may be rather more indirect and complex than a simple identity. Our brains' information processing certainly has a function, it does something for us, but once we abandon identity theory, we may no longer have to say the same about consciousness itself.

In this light, I would like to mention two papers that strayed somewhat from the dominant theme and approach, but which I found of particular interest. Amy Ione's poster presentation 'Implicit Cognition and Consciousness in Scientific Speculation and Development,' was unusual in its attempt to set current discussions of consciousness in historical context. Considerations of the views of Baars, Chalmers, and Stapp were interwoven with an account of the pre-classical roots of our conceptions of truth, and set off against examples of imaginative insight -- reconfigurations of conscious experience -- from the history of science. I am not sure that the end result was entirely successful, but the attempt is exemplary. Like any 'frontier science', consciousness studies currently needs more to get a clear grasp of the problem situation, to ask the right questions, than to grab at quick answers. But this problem situation, especially our deeply contested understanding of the term 'consciousness', and perhaps even some aspects of our actual self experience (if authors like Jaynes, Snell, or Dennett are at all to be believed), is very much historically conditioned, and is unlikely to be clarified without the sort of historical perspective that Ione is attempting to introduce into the field.

I also found Richard Stevens' 'A Perceptual-Lingual Model of Consciousness' very interesting. Stevens is a very different sort of psychologist from the cognitivists who dominated the conference. He was not primarily concerned with implicit cognition either, but rather with its converse, with what consciousness itself is like. The only viable method for investigating this, he argued, is 'phenomenological reflection', i.e. a careful, extended, and disciplined attention to one's own conscious contents. Use of this method led him to conclude (apparently to his own surprise) that consciousness is entirely perceptual in its character: the contents of consciousness are all either percepts, or mental imagery in various sensory modes. Amongst these, of course, verbal imagery, inner speech, plays an especially prominent role. However, he insisted, meanings, intentions, and relations of thought, even with respect to this verbal imagery, remain entirely implicit. All that appears in consciousness itself are percepts and images.

I must confess to some mixed feelings about Stevens' claims. On the one hand his research seems to be uncomfortably like the discredited introspective psychology of the early 20th century; particularly the much maligned tradition of Titchener. (In his defence, it should be noted that he is well aware of this history, and intent on avoiding the pitfalls of the past if possible.) On the other hand, it seemed to me that his conclusions were thoroughly plausible, and, in any case, that we can hardly expect much progress in the understanding of consciousness unless we pay this sort of attention to characterizing the phenomenon itself. What precisely is consciousness studies the study of? To focus exclusively on things like implicit cognition, however artful the experiments and fascinating the results, would be to avoid the real question. If Stevens is even partly right, imagery ought to be a vital concern for consciousness studies, as it was for cognitive psychology 20 years ago. It proved contentious then, and I would venture to say that the cognitivist movement never satisfactorily resolved the issues. Standard cognitive theories avoid confronting the fact that imagery is a quintessentially conscious phenomenon. Students of consciousness cannot afford to do likewise.

If this conference had confined itself rigidly to the cognitive psychology and neuroscience of implicit cognition, then some excellent science would have been presented, but consciousness might not have been much illuminated. It seems to me, however, that by setting its bounds just a little more widely it achieved a considerable success. The more speculatively inclined were forced to confront some hard science (and some rigorous philosophy), and everyone had the opportunity to enlarge their perspectives without having their credulity strained too far. The Association and the organizers are very much to be congratulated.

To contact the author, click here
Mail me

Return to Home Page:
Imagination, Mental Imagery, Consciousness, Cognition:
Science, Philosophy & History.

Imagination, mental image, consciousness, cognition