Jun 05 2012

Brains Wide Shut

Published by

Essay: Brains wide shut?

30 April 2005
From New Scientist Print Edition.

Patricia Churchland is professor and chair of philosophy at the University of California, San Diego. She also works in consciousness studies, and has published books and has articles including:Neurophilosophy: Towards a unified science of the mind-brain (1986); Brain-Wise: Studies in neurophilosophy (2002); Ten Unsolved Problems in Neuroscience (with David Eagleman, forthcoming)

THE books-on-consciousness mills are running full tilt. Just about anyone who is conscious seems motivated to write on the subject, and most authors profess themselves emboldened to call their own contribution a theory of consciousness. Because the output is accelerating, it may be innocently assumed that something new has been discovered. Alas, the truth is quite the opposite: very little has been discovered. All this furious activity is reminiscent of the flood of speculative theories of life in the early decades of the 20th century. And for much the same reason: science is moving forward on the problem, but has not yet nailed down the answers – and no one really knows what the answers will look like.

Such a wide-open empirical playing field motivates authors to wrestle feverishly with each other, hurl mud, vote one another off the island, and draw endless boxes with connecting arrows. Naturally, it is vastly easier to hunker down in the hot tub to introspect one’s inner milieu than to do the painstaking work of neuroscience. As a result, legions of hopeful Darwins-of-consciousness flock to conferences, vying for attention, and philosophers, horrified by the spectacle of the empirical sciences treading on their sacred territory, try to scare off interlopers with extreme threats of conceptual necessities and logically adequate criteria.

Resolving disagreements about the possible nature of things comes from empirical discoveries about the actual nature of those things. Prior to such resolution, sound and fury emanates from the clash of intuitions and ideologies. And much of it, alas, does signify next to nothing.

A normal part of the maturation of any science, be it physics, astronomy, chemistry, geology or neuroscience, is that “self-evident verities” (otherwise known as “intuitions”) expire in the oxygen of factually based theories. Who now bangs his pots concerning the “intrinsic nature” of élan vital (life force)? Or considers it a conceptual necessity that space is Euclidean? And what now of the fixity of continents?

One important lesson that has escaped most philosophers and many neuroscientists is that what seems obvious can change as the science changes. With scientific development comes conceptual development, and this alters how we think about and see the world. This applies to our inner world, as well as to our outer world.

Into this fray rides – once again – the tireless figure of Daniel Dennett, the philosophy professor at Tufts University in Massachusetts whose Consciousness Explained brought him such celebrity. Fourteen years on, his new book, Sweet Dreams: Philosophical obstacles to a science of consciousness (MIT Press), is a collection of essays devoted mainly to identifying and pummelling those diehard intuitions that he believes, rightly, still obstruct the progress of cognitive neuroscience.

Among these intuitions is the idea that there could be a zombie like me in all respects – all, save that it lacks qualia. A lack of qualia means it doesn’t have the “experience” of redness when it sees a London bus, but like me would say: “Look! There is a red London bus.” Incredibly (I’m not making this up), zombie-me would have exactly the same conversations about conscious experience that I do. For example, we both say: “When I dream, I am aware of actions, such as flying, but not aware of how bizarre those actions are.” The difference is that zombie-me has neither experiences nor qualia to talk about.

Could there be such a zombie? “Perhaps not,” says the purveyor of zombies. “It is a thought-experiment-zombie.” Fine. But so what? “Well, the mere imagining of such a thing entails that consciousness cannot be a property of the brain…” Good grief. As a colleague once muttered in despair, “..his argument is not even wrong.”

Dennett is right about most of the philosophically pampered intuitions, especially those bravely predicting that “science can never, ever explain consciousness”. These intuitions and the arguments they spawn have been repeatedly exposed as confusions, fallacies, circularities, failures of imagination, arguments from ignorance and just plain bunk. By Dennett, to be sure, but by a host of other philosophers as well, including Owen Flanagan and Paul Churchland.

Puzzled by the resistance to his criticism, Dennett has gone further, trying to explain why certain dubious arguments are remarkably resilient in the face of their evident demolition – all to marginal avail, at least for the big-money players David Chalmers, Colin McGinn, Ned Block and Frank Jackson.

Why has Dennett’s remedial exercise not had greater effect? Part of the answer is old hat: it is the next generation, with less invested in the status quo, on whom the impact of new ideas is usually greatest. But the larger part of the answer, I suspect, is owed to the state of play in neuroscience.

Neuroscience is a very young science, still in search of its own exoskeleton – the fundamental principles that explain how nervous systems work. Although an enormous amount is known about the structure and function of individual neurons, how macro effects emerge from populations of neurons remains poorly understood. For example, we do not yet really understand to what degree sensory or motor systems are hierarchically organised. Deep puzzles endure over the way memories are stored and retrieved, what attention is and how it shifts, and how decisions are made and behaviours organised. And basic issues still need resolving about how neurons code information, and how information is integrated across neuronal populations.
With so many questions so open, neuroscience can’t offer an integrated, comprehensive theory of conscious phenomena to oust those old-time intuitions asserting the “metaphysical mysteriousness” of consciousness. Could it do so, some intuitions about, say, the “unity of consciousness” would undoubtedly slink off to join caloric fluid and crystal spheres in the graveyard of scientifically dead ideas.

Dennett himself, though a savvy gamester, may have misplayed his hand in an odd way. In Consciousness Explained, the title was not meant to be a joke. Yet those of us who eagerly read the book expecting to understand the neural mechanisms of mental phenomena felt somewhat let down.

“Dennett himself, though a savvy gamester, may have misplayed his hand in an odd way”

Dennett did not explain consciousness in neural terms at all. True enough, he did offer some semi-fertile new metaphors to substitute for sterile old metaphors, and that was progress of a non-trivial sort. And as in Sweet Dreams, he did joyfully beat up some silly ideas. But to describe his story as “explanation” was a bit strong.

What is Dennett’s account? First, let’s look at two cases where there is a contrast between being aware and not being aware of an event. In the first case, I have been hiking all day, and I have a blister on my heel. I am aware of the pain on my heel and aware of my fatigue. A mother bear and cubs emerge from the bush. Now I am aware neither of my pain nor my fatigue but only of the threat. In the second case, I am in the deep stage of sleep (not dreaming) and you begin to whistle Dixie. I am not aware of hearing the tune.

How does Dennett explain the difference between being aware of my blister before the bear appears and unaware after; and what is the difference between being aware of Dixie when I am awake but unaware while I am in deep sleep?

Dennett argues that at any given moment, many sensory signals, from inside and outside the body, enter the brain. Depending on conditions, some have little effect while others have a big effect. The salience of a signal (the threatening bear) and general state of arousal (being awake) are two relevant conditions among many. When I am aware of the blister, it is because the pain signal hijacks attention, behaviour, memory storage and retrieval, emotions and so on. Enter the mother bear, and the pain loses out to the fear signal, which then dominates the mechanisms for attention, decision making, memory and planning.

Dennett’s view is that consciousness of an event is what happens in the brain when a signal takes control of the aforementioned functions: that is, consciousness of an event is a matter of large areas of the brain being influenced by one of a set of competing sensory signals. None of this can happen, he insists, without language. He embraces the surprising conclusion that the non-verbal (animals, infants and the profoundly aphasic) are not genuinely conscious. He has convinced himself that acquiring a language rewires the brain to make it function as a serial digital computer, and that this rewiring is necessary for consciousness. Repeatedly hammered for pampering his own intuition, Dennett doggedly stands by his language requirement for consciousness.

Dennett concedes that he has no understanding of the neural mechanisms of the competition and control dynamics, and cheerfully waves off the brain as merely the wetware implementing the von Neumann-style software. He is strictly a software man, and proud of it.

This is not reassuring. As Dennett must know by now, the hardware-software metaphor applies to brains about as well as it applies to kidneys – poorly – and then only if you blind yourself to glaring disanalogies. Of course it might be convenient if we could understand the brain without neuroscience, but it ain’t so.

Quite simply, there is no substitute for understanding the brain at all its levels of organisation. To understand conscious phenomena, we need to understand, in neurobiological terms, such things as the difference between being aware of pain and not being aware of it, why we experience sounds differently from body position or smells, and how autobiographical memories are retrieved.

Apart from lacking “neuro-cred”, Dennett’s account turns tail when confronted with a range of clinical data from human patients. For example, neuropsychologists David Milner and Melvyn Goodale have carefully studied a patient who, following oxygen deprivation, lost the ability to see shapes. Nevertheless, when placed in front of a box with an oriented slot, and asked to “post” a card in the slot, she is consistently successful in posting the card, regardless of the slot’s orientation. The location of her lesion (the ventral stream in the visual pathway) is consistent with a wealth of physiological data identifying the specialised functions in that region.

Why is this a problem for Dennett? Because the shape signal, though not conscious, appears to “take control” to permit her to post successfully. In Dennett’s view, consciousness is just winning and taking control. Follow-up psychophysical research on normal subjects indicates that some motor behaviour is routinely guided by non-conscious visual signals that outcompete the conscious visual signals. The Milner and Goodale research is some of the best work on consciousness going.

One rather important upshot of Dennett’s claim to have explained consciousness is that the scientifically naive may have been misled into believing that his philosophical story is what an adequate scientific theory of consciousness would look like. Recognition of the inadequacies in Dennett’s positive account seemed to invigorate the naysayers’ prediction that neuroscientific explanations for conscious phenomena are forever doomed, regardless of what science discovers. The diehards remain faithful to their favourite intuitions, confident that qualia are “ontologically basic” or “transcendentally real”—pick your favourite empty slogan.

So, to avoid equivocation, let’s agree that something will be considered a theory of consciousness if, like the theory of how proteins are made, it explains the main properties in sufficient detail to satisfy four conditions: we understand how macro events emerge from the properties and organisation of the micro events; novel phenomena can be predicted; the system can be manipulated; and it is clear at what level of brain organisation the phenomenon resides.

“The unglamorous truth is that science will come to understand consciousness in much the same way as it came to understand the nature of life”

These criteria imply that a genuine explanation of the properties of conscious phenomena must characterise neurobiological mechanisms. A theory satisfying the four desiderata will not be solely a psychological account, linking various cognitive functions, it will not be just an array of boxes, labelled as cognitive, with arrows connecting the boxes. “Boxology” at the psychological level is a crucially important component of this theory, but it does not explain the neurobiological bases for the functions in the boxes. Likewise, an explanatory theory will not consist solely of detailed anatomical maps of what connects to what within the brain, though such maps are also essential to the solution.

And finding correlations for certain conscious events, perhaps via functional MRI or single-cell recordings, does not as such constitute a theory, because such correlations do not, ipso facto, reveal mechanisms. Correlating events using different measures, such as fMRI or behavioural techniques, can be extremely useful, but a roster of correlations doesn’t constitute a theory.

So, what can we expect from neuroscience? Most likely, that a theory of consciousness will co-evolve with an understanding of the fundamentals of brain function. And the details? Needless to say, I can’t answer that question. The fact is that neuroscience circa 2105 will be profoundly different from neuroscience circa 2005. Even if a kind alien left the answer on my pillow, I could not understand it without also understanding the larger (still missing) theory that embeds it. Likewise, Galileo would have remained stumped about the nature of life had the same alien left him a note describing the structure of DNA, without also teaching him the entire background of modern structural chemistry.

An empirical theory of conscious phenomena will not, or course, simply waft up out of the neural data. It will be the product of brains that create hypotheses, and that creation will draw upon psychology, neuroscience, genetics, computational theory and ethology. Reasonably enough, in Sweet Dreams Dennett warns that if you study the growing body of neurobiological data through the lenses of cock-eyed intuitions, then you impede your brain’s theory-making machinery.

As we have already seen, his worry is not idle, as some neuroscientists have unwittingly swallowed colourful blather about zombies, “intrinsic properties” and the transcendental mysteriousness of qualia. This book contains many nice remedies against a lot of sneaky rigmarole.

But the unglamorous truth is that science will come to understand the components of consciousness in pretty much the way it has come to understand the components of the nature of life. Not with a single blindingly beautiful insight, but by understanding the mass of detail at many levels of organisation: molecules, cells, networks, subsystems – and the whole system. All of which means a lot of hard empirical slog ahead before we can hand out any Nobels.

From issue 2497 of New Scientist magazine, 30 April 2005, page 46

No responses yet

Trackback URI | Comments RSS

Leave a Reply

Prove You Possess Consciousness * Time limit is exhausted. Please reload CAPTCHA.