Evolution TreeThere are many, many different ways of implementing consciousness, each with its own advantages and weaknesses, and it may well be that lots of them have been tried out during the course of evolution. So say Rodrick Wallace and Roger G. Wallace, in a paper (pdf) full of daunting mathematics and airy speculations.

The Wallaces take Bernard J Baars’ Global Workspace theory as their starting point. The Global Workspace theory sees consciousness as providing a general access function: a fleeting memory capacity that connects brain modules that otherwise function separately. It’s as though consciousness were a noticeboard where the different functions can post problems, warnings, discoveries and data which other functions may be able to use or respond to. Consciousness is there to provide an overall integrative function.

The Wallaces have developed a detailed analysis of the requirements of such a function, and they have come to the conclusion that many different architectures and organisations are capable of giving rise to something with the essential properties. Indeed, they see no barrier to a slow form of consciousness in all sorts of creatures: a kind of ‘paraconsciousness’, as they call it. The details of this notion are tantalisingly vague, but it seems the idea is that even trees might have processes which allow them a very slow and very dim kind of awareness. The Wallaces also see analogues of the Global Workspace in systems made up of multiple organisms: societies or ant colonies. The idea that such higher-level systems have at least some of the properties of consciousness is reminiscent of Ned Block’s ‘Chinese Nation’ thought experiment, in which the whole population of China is somehow dragooned into hand-simulating the nervous activity of a brain. The Chinese Nation idea, however, was supposed to demonstrate the weakness of functionalism by pointing out how absurd it is to think that this kind of gigantic hand-simulation of mental functions would give rise to anything like a mind. I suppose this perhaps shows how different our perceptions of intuitive plausibility can be.

The idea of higher level workspaces has a role, however, in overcoming some of the objections to the Wallace’s ideas. The most obvious objection, of course, is that they take the Global Workspace for granted: they say it is rapidly becoming the most favoured model among researchers, but even if that is true, it’s a long way short of being established. It remains quite possible that the theory is entirely wrong; or perhaps more likely, that an integrating function does exist, but does not constitute consciousness.

At the risk of being speciesist, it certainly looks as if the Wallaces are thinking of a form of consciousness somewhat short of the grand human version we are normally concerned with: they readily attribute it to a range of animals all the way down to cephalopods (who would surely achieve consciousness through a very different organisational structure from ours, given their tendency to rely on large neural ganglia distributed around the body rather than just a single central brain).

In fact the Wallaces are primarily interested in the palaeontology of consciousness, the road to which is, they say, wide and open (A rather optimistic conclusion, I think, since about the best evidence we can hope for is the shape of a fossil skull. The Wallaces’ own conclusions reinforce the common sense assumption that you can’t tell much about how a strange brain may have worked merely from its shape) . What strange forms of intellect may have flourished during the great Cambrian explosion, when many bizarre animals with strange body plans flourished!

However, they do have a view about the special qualities of purely human consciousness, and this is where the higher level workspaces come in. It is likely that all the different forms of consciousness tried out by evolution had different strengths and weaknesses: most would have suffered from some areas of ‘inattentional blindness’, but the creatures that had the smallest blind spots in the least important places would have survived best. Perhaps the unique trick of human consciousness, the one which fitted it for a distinctly human way of surviving, was to have a special facility for tuning in to higher-level workspaces: for fitting into complex societies and networks of communication and decision making.

This idea is more or less an aside in the Wallaces’ paper, but I can see some appeal in it, and I suspect it might in fact provide a more fruitful avenue to pursue than the attempt to reconstruct the minds of Acanthostega or Tulerpeton.


  1. 1. Eric Thomson says:

    Interesting ideas!

    At the risk of being speciesist, it certainly looks as if the Wallaces are thinking of a form of consciousness somewhat short of the grand human version we are normally concerned with: they readily attribute it to a range of animals all the way down to cephalopods.

    Cephalopods have pretty complicated nervous systems. Like the leech, they have ganglia distributed all over the place, but also a more voluminous ‘head brain’ that guides decision making. I worked on the leech nervous system in grad school, and by the end I wasn’t sure if they were conscious or not.

    They represent sensory information, make decisions based on those sensory inputs, and behave accordingly (e.g., when they are feeding, leeches are very hard to perturb as it is such a special event). Are we far along in consciousness studies to say with confidence that the leech is not conscious? Clearly not.

    OTOH, I am not sure the leech has ‘modules’. For that matter, I’m not sure humans have them, but my bet is that key aspects of the global workspace idea can be salvaged even if modular organization is a theorists’ fiction.

    The ideas about the time scale being flexible, of having creatures conscious even though the relevant time scales are very long, is very interesting. If we played someone’s brain in ‘slow motion’, would they still be conscious?

  2. 2. Peter says:

    Of course, there’s consciousness and consciousness – if you stunned a leech somehow you would have every right to speak of it’s being no longer conscious. It would be harder to say whether it was in any sense self-aware or capable of reflecting on its destiny or indeed, of experiencing pain – something you might want to know about if you proposed to stun them on a regular basis!

    I remember seeing a film about an octopus which was kept in a lab where they also had, if I remember rightly, a glass tank of crabs. The crabs began to disappear: the scientists had not realised that the octopus was able to clamber out of its own tank and into the one with the crabs. It did this every night, but the really cunning part was that it went back to itsd own tank again afterwards. Some would say that was evidence that the octopus had a ‘theory of mind’ – it could predict what the scientists would do if they found it in the crab tank.

  3. 3. Eric Thomson says:

    I agree: even if leeches feel pleasures and pains, they almost certainly don’t think about the past or future (they are awful at learning and have no operant conditioning to speak of). But if they have qualia, if they have any consciousness at all, that would make them a great model system for consciousness.

Leave a Reply