The Detector Approach
Understanding the Hard Problem - How Consciousness can fit into our Natural Orders
Jan Holmgren October 1997
In the late sixties, I became fascinated by the intricacies in understanding consciousness. In 1981, I gathered my efforts in a sketchy paper, "Hypothesis on Consciousness" (HoC). On some occasions in the years to come, I sent HoC to persons who published on such issues. Generally, the response was friendly, but reluctant. Recently, on the Internet, I found a new openness towards attitudes similar to mine, and I decided to use the net to make my ideas public. The original text of HoC is left unchanged, with this introduction relating it to the current discussion, and especially to the work of David J. Chalmers.
Chalmers, in his book "The Conscious Mind: In Search of a Fundamental Theory" (CM), substantially improved speculation on consciousness. He explicitly stated something I had tried to aim at long ago: taking consciousness and science seriously, and taking consciousness to be a natural phenomenon. Chalmers started as a mathematician and information theorizer; my interests have been in biology and theories on evolution (by profession I am an architect). I claim that I have taken the more fundamental grasp, and that my approach demonstrates some of Chalmers' results to be wrong.
Accepting some darkness
In the introduction to CM, Chalmers states that we are entirely in the dark about how consciousness fits into the natural order. At the core of my view, which I would now like to call "the detector approach" (as distinct from for example "the first-person approach" and "the third-person approach"), there is a clear answer to that problem (P. Hut & R.N. Shepard suggest a similar move in their contribution "Turning 'the Hard Problem' Upside Down and Sideways" in "Explaining Consciousness - The 'Hard Problem'" (EC), ed. J. Shear). Being fundamentally detectors every one of us, in every moment, is in the middle of a huge "darkness" (or rather nothingness). The detector approach establishes a natural dualism between (1) detection and (2) what is outside detection. Since detection takes place in fall-outs of consciousness within a neural structure with capacity for memory, the dualism takes on the shape of a three-entity relation (not "body - mind", but "world - underlying structure - fall-out of consciousness").
The illusion is that it seems to us that we, in every moment, experience almost everything there is to experience. The explanation is simple: we do not at all experience what we do not experience. Also, we easily move in consciousness inside the parts of the neural structures, which are impregnated with memories: in the extensive world internalized within every one of us. We do not experience any of the darkness outside or around what we experience in a certain moment; we simply have not the faintest hint about it.
This view may seem innocent, but some persons react with alarm. In the modernist view typical of natural science, from the age of Enlightenment onwards, omnioptimism is the good manner. Romanticism, brooding on the dark and elusive sides of life, should be avoided. Perhaps, any darkness around us can be felt to be similar to death: an endless, empty, and cold nothingness. It may be understandable if some of us prefer to forget about, and even choose to deny, that there is anything in principle unknowable.
Still, the worlds we know of (each of us obviously has his or her own world, often very different, but in parts more or less similar), and their natural orders, are entirely experienced inside consciousness within single human individuals. Science, in common efforts, can help us individuals to widen our worlds, to make accessible for consciousness more comprehensive areas of knowledge and experience, and to move towards some consensus by learning to share at least some parts of each other's worldviews. The magnitude of the remaining darkness can never be assessed. The detector approach may seem rather self-evident, and would probably, if accepted by the scientific community, rather easily gain a general acceptance.
To me it is obvious that the step from animal to man is an enormously important one. Much in the behavioral repertoire of man is unique in the world. Take science, just for an example. Must we really dismiss the obvious explanation: that man is unique in having acquired access to consciousness? Biologists, who are generally aware of the utterly stereotyped behavior of all animals, often seem to be clear about this crucial step. For Chalmers "consciousness is a feature of the world that we would not predict from the physical facts". Quite on the contrary, I think it is obvious from the physical (high-level) facts that something new has entered the world. However, the philosophical problems around emergence have been blocking this possibility for an independent explanatory role for consciousness.
The crucial move here is to see consciousness as a natural phenomenon (similar to light, gravitation, etc). The process of acquiring access to a new natural phenomenon (often never before exploited by any organism) is omnipresent in the history of biological evolution, so this move gives consciousness a natural place in a biological worldview. What remains special for it is its role as the medium for conscious detection, which should be expected to cause problems around self-reference. This is exactly what we can observe.
Many have had doubts about Charles Darwin's theories. To quote G. Bernard Shaw on natural selection, "there is a hideous fatalism about it, a ghastly and damnable reduction of beauty and intelligence, of strength and purpose, of honour and aspiration". Natural selection seems indeed to do away with values such as truth and goodness, and Darwin himself commented: "With me the horrid doubt always arises, whether the convictions of man's mind, which has been developed from the mind of lower animals, are of any value or at all trustworthy." (quotes from D. Lack "Swifts in a Tower").
Conscious selection in domestic breeding and conscious manipulation of genetic material are striking examples of man's capability to act consciously outside constraints, which otherwise are absolute in the animal kingdom. As soon as we accept consciousness as a natural phenomenon, we also should accept its ontology. Truth, goodness, etc. exist, and can affect the world. As long as we are not willing to say that much, we leave the field wide open for sheer mysticism. Absurdly enough, creationism is still not defeated, and the main explanation may be just this: the natural sciences have tended to leave us with unnecessarily restricted and colorless worldviews.
In CM, Chalmers leaves a fundamental ambiguity in the concept "physical", which is fatal for some of his results. For example, talking about "the physical" or "the physical domain" in the crucial characterization of possible options on p. 161, he says: "conscious experience arises from the physical according to some laws of nature, but is not itself physical". What is the demarcation of the physical here? For consciousness, to be a natural phenomenon and obey laws of nature obviously is not enough for it to count as a physical phenomenon in Chalmers' worldview. This unclear tendency towards an essential dualism prevails throughout the book.
The detector approach makes possible a simple and clear answer to this (compatible for example with C.J.S. Clarke's contribution "The Nonlocality of Mind" in EC). In evolutionary time we have achieved neural structures that can exploit consciousness, rather like eyes can help an organism to exploit the phenomenon light. For obvious reasons, these mechanisms have become effective in handling things around us (and indeed they give us a very good grasp of the world), while they are much less efficient in understanding our inner structures and functions. However, in evolving methods for abstract information processing, we have extended our abilities to understand and deal with our inner structures in increasingly sophisticated ways. For example, we can apply different attitudes towards different parts of our inner structures; "the first-person approach" and "the third-person approach" are examples of such different aspects. "The detector approach" is the most fundamental aspect possible, and only establishes the relation that each of us experiences something and does not experience what is simultaneously outside, "in the dark". Relations like supervenience and explanatory relevance, for example, are much more sophisticated conceptual structures.
Since all we experience takes place in consciousness, each of us lives in a world of remarkable unity. Everything we deal with, concepts, perceptions, actions, etc., are "made of the same stuff" ("We are such stuff As dreams are made on, "), however difficult to capture. Reduction is generally understood to mean functional representation by use of elementary (often physical) concepts. Now, accepting the existence of qualities (see HoC), we can add a slightly different notion, namely methodical suppressing of qualities. Physical science is a conceptual building made up of complexes of concepts, rules, etc., all inside consciousness. An important character, shared with other natural sciences, is the systematic minimizing of the roles of qualities.
On this view, there exists just one single world, of which we all are well-integrated parts. The topological situation for all human individuals is that each of us experiences only parts of the world. All "domains" we can distinguish must be defined within our inner structures, in principle in sets of concepts. We then must decide how to define different domains; so, let us consider "the physical domain".
Some "things" in the world seem to have a rather curious existence; for example, a rainbow "exists" only where an observer is present. Still, the rainbow obviously belongs to the physical domain, thanks to a comprehensive theoretical structure explaining it. The experienced existence of a physical "thing" (in this case a rainbow, in direct experiences and in the form of complexes of concepts) must be thought to be located in a brain. Now, is a rainbow anything at all? For example, can it affect anything in the world? Yes, obviously children are often painting rainbows, probably experiences of rainbows inspired Newton to experiment with prisms, etc. According to Chalmers, we now may distinguish between two levels in the inner life: psychological and phenomenal. On the psychological level we are inside the physical domain (which is closed), while on the phenomenal level we are outside the physical domain. To me it is utterly unclear how this would sort out the situation I have sketched. For example, ghosts also have the "rainbowy" characteristic of existing only where an observer is present. Certainly, ghosts will have both psychological and phenomenal aspects, children paint ghosts, etc, but should this let them into the physical domain?
I think we have two options: (a) to say that everything everywhere in all times (including consciousness, and even ghosts) is inside the physical domain, (b) to let the present state of physical science decide what is inside the physical domain. Obviously the concept "physical domain" will have very different contents owing to our choice, but this is a semantic issue not affecting the implicit worldview. Both options may amount to essential monism; in (b) everything that is not at present inside the physical domain may be so potentially. Now, what can we say about ghosts? Obviously ghosts exist in some meaning, and our problem is to fit them into our general framework of concepts. Perhaps we may speculate that the borders between physical science and say chemical science, and even biological science, in the future can dissolve, so we can have an extensive united field of knowledge. Perhaps even ghosts can fit in somewhere. When Chalmers talks about "the physical", he seems to have such an extended field of knowledge in mind, including almost everything.
To me it is more natural to opt for (b) with physics in its present state, and count with numerous different sciences covering different domains or aspects of the universe. To me, there would be no problem to have a science on ghosts, and probably there already exists one somewhere. On this view, the physical domain can not be causally closed until physical science is complete, and then tells us that it is causally closed. Obviously, we are far from that situation, and I suspect that man will never, owing to topological problems with self-reference in consciousness, come really close to it. But, in fact, this is an empirical problem.
Linked with the physical is the possibility for computation. Chalmers says: "there is every reason to believe that the laws of physics, at least as currently understood, are computable, and that human behavior is a consequence of physical laws." Computation is systematic evolution of microevents in time, and most of the time it goes on "in the dark". Results are attainable when they reach some kind of terminal, where a structure with special physical properties "reacts" or "exposes something" (often visually on a screen). Since man is a result of an enormously complex biological evolution, still badly understood, the sudden jump from rather simple computational systems to AI pleaders' expectations for consciousness within these systems is just surprising.
However, the detector approach leaves no hope for such simplistic reasoning. Natural evolution is known to be effective in exploiting new phenomena. If not even broadly experienced organisms very nearly similar to man, like chimps or gorillas, have gained access to consciousness, how can one expect it to appear in simple mechanical devices? Even if chimps and gorillas should turn out to be computable, the behavior of man will never be wholly computable, since consciousness, of topological reasons, will not be well defined in space and time (see HoC). The independent explanatory relevance of consciousness in the detector approach, supplying the medium for detection, also makes most of Chalmers' thinking on zombies invalid. Chimps and gorillas may be said to be a kind of zombies, but a human zombie without consciousness is not even a logical possibility. Conscious detection is fundamental, permeates behavior, and has affected the structure in human brains in evolutionary time.
In his conclusion in EC, Chalmers takes a stand in five "choice points". I here declare my meaning on each of the five points, without further argument:
F.J. Varela (in "Neurophenomenology: A Methodological Remedy for the Hard Problem" in EC) interestingly argues for phenomenological investigation. If we adopt the detector approach, most human activities can be thought to contribute to such a project. A result of the new openness, very much a result of Chalmers' good work, may be the appeal that many kinds of human endeavors should contribute to our worldviews. To understand the hard problem, I think, is to accept the great importance of qualities in our worlds, and their enormously varied influences in common human life, in the arts, in politics, in different religions, etc. A common field of knowledge cannot be restricted or closed, and certainly cannot be restricted to abstract theories within the natural sciences.
J. Shear (in "The Hard Problem: Closing the Empirical Gap" in EC) talks about "pure consciousness" that can exist entirely independently of qualities (qualia). However, he seems to be thinking of meditative states, where both qualities and functions are extremely suppressed by volitional mechanisms in the brain. I believe that the natural sciences should be characterized by their methodical suppression of the influences of qualities; they aim at functional explanations (in my terminology in HoC: they deal with distributions). D.C. Dennett ("Facing Backwards on the Problem of Consciousness" in EC) perhaps is the most consequent proponent of the view that functional explanation gives full explanation, so qualities do not exist. Although he may be determined to stay within the domains of the natural sciences, it seems premature, from that perspective, to decide on whether qualities exist, which, again from that perspective, must be an empirical question that is far from a final answer. Also, the theories of the natural sciences are obviously far from quality-free; a preliminary mapping of quality-laden concepts in crucial points in scientific theories might be interesting. For all of us, the rich interplay of distributions and qualities will never cease to intrigue and fascinate.
Hypothesis on Consciousness
The Graphic Detector Approach
The Nursery Detector Approach