Peculiar qualia
In a recent editorial (Perception vol. 25, page 377) I suggested that sensations---qualia--- might serve to flag the present. This notion is based on the premise that human perception is heavily top--down ---so to a great extent depends on stored knowledge, derived from the individual's past experience, and ancient knowledge, inherited through the mechanisms of natural selection from ancestral successes and failures.
To be more specific: if we think of perceptions as predictive hypotheses of what is out there in the external world, there must be a continual problem for action, for what is happening at the present moment is not specified by top--down knowledge. This is as true of hypotheses of science, as for hypotheses of perception. This lack of top--down knowledge of moments in time must endanger survival when appropriate actions are required for dealing with what is happening now for living into the future. The more perception is top--down, the greater will this problem be; yet stored knowledge is essential for intelligent predictive perception and action, so it is not surprising if top--down knowledge becomes ever more important through evolution.
Individuals of `lower' species are less dependent on knowledge from the past---relying more on real-time sensory inputs. For a pure stimulus--response system there would be no problem here (even though reflexes are derived from ancient behavioural successes and failures), for stimuli arrive in real time and so signal the present. In ourselves, the reflex eye-blink to a sudden loud noise must have developed from ancestral eye-disasters, associated with sudden loud noises, which still initiate our reflex. In the same way as primitive creatures, here we do not respond to a hypothesis of what might be the cause: our eyelids respond directly to the stimulus. This is not acting on a hypothesis, so is not in this sense a perception. Such reflex responses are fast and often appropriate, but not flexible or intelligent. They are essentially geared to present stimuli.
The suggestion is that any heavily top--down system---including a robot---must have trouble with identifying the present. It has to use present inputs or stimuli to distinguish the present from the remembered past and from the anticipated future. All this seems pretty well self-evident. The question now is whether consciousness has a role to play.
To me, at least, remembering a very recent event is very different from especially visual perception. When looking at something, then closing the eyes and remembering or imagining it, the vivid qualia of vision disappear. The memory lacks the sensations of colour and so on of real-time perception. Hence memories and imaginations are not confused with perceptions. (Here I shall ignore visions of schizophrenia and dreams, considering only normal observers in typical situations.) The notion is that consciousness may come in here---by stimulus-triggered qualia flagging the present.
This notion, that qualia flag the present, does not begin to suggest how consciousness arises from brain function. But it does imply that consciouness has causal effects. This, of course, is an extremely contentious issue.
It would be a mistake to say that, if consciousness is an `emergent property' of physical brain functions and has causal effects on them, it must be separate from brain functions. Thus, for example, water has very different properties from the hydrogen and oxygen atoms from which it is created: water puts fires out, while its constituent atoms do the opposite. So emergent properties can be self-causal. One could therefore say that, if consciousness is an emergent property of brain functions, qualia could causally affect the brain that creates them---when triggered by stimuli from the present---without invoking mind--brain dualism.
The move to saying that qualia affect behaviour has many implications. In the first place it allows an evolutionary reason for consciousness, if indeed qualia aid survival beyond nonconscious processes. Second, qualia should affect evolution. This suggests, surely, that we might find evidence of qualia perhaps in ancient paths of evolutionary development.
Biologists explain behaviour without reference to consciousness. Some psychologists explain human behaviour in terms that include likes and dislikes, with the underlying assumption that hedonism has a large part to play. These psychologists at least implicitly assume that qualia are causal. This fits common sense.
A complication here is that immediate hedonism is very commonly tempered by long-term planning and expectations. Thus a poor student will suffer now to gain degrees of freedom for later rewards. (One might say that morality is deferred hedonism---delayed even to anticipated rewards after death). In spite of this complication, we do seem to see behavioural evidence of qualia in others; for example, when sitting motionless for hours at a concert. If they don't have music qualia, what on earth are they doing? Doesn't it follow that, if human behaviour is qualia-affected, at least human evolution must be affected by kinds of qualia such as likes and dislikes? If music qualia incite dancing, or violence, or war, or quiet thinking, then qualia must affect survival and reproduction---and so affect human evolution. If this is so for nonhuman animals, can we see qualia preferences in ancient paths of evolution? Can we design experiments to distinguish between effects of qualia and stimuli in animals? Can we show that a dog likes being tickled---with tickle qualia?
It would be useful to measure intensities of sensations. It was just this that Gustav Theodor Fechner (1801--1887) tried to do 150 years ago. Starting from the just-noticeable-differences (JNDs) of Weber, Fechner added (or integrated) them as units of sensation. Thus he tried to pin qualia with operational measurements.
The question is: Do sensations above threshold increase as though qualia are composed of JNDs? It was essentially this that William James rejected by saying: ``Our feeling of pink is surely not a portion of our feeling of scarlet; nor does the light of an electric arc seem to contain that of a tallow candle within itself.'' But is this a strong argument? It is very common for changes of quality to occur with changes of quantity. The trouble here, surely, is different. What we need to know is whether intensities of qualia increase lawfully with the number of JNDs above threshold. This takes us back to the dawn of experimental psychology, especially to Weber, Fechner, and Titchener. Perhaps we should look at this past to see where to go for our future progress in this most puzzling and least understood aspect of perception. Isn't it paradoxical that sensations---qualia---are what it is all about for artists, and just about everyone else---yet we, the experts, are in the dark.
Richard Gregory
Return to contents of current supplement
© 1996 Pion Ltd