The Ravenous Brain: How the New Science of Consciousness Explains Our Insatiable Search for Meaning

by Bor, Daniel

  So certainly not all brain areas equally contribute to consciousness. The above example also partially answers another question from the list—about consciousness simply relating to number of neurons. If neuron count equated to levels of awareness, then the cerebellum, with perhaps 80 percent of the entire brain’s neurons, would be the most conscious part. That clearly isn’t the case.

  In fact, the most critical region for consciousness is also one of the smallest and—in some ways—the least interesting, for the purposes of this book. This brain area is known as the reticular formation. It is part of the brain stem, the most primitive of brain regions and another component of the reptilian brain. The reticular formation controls the sleep-wake cycle through a complex set of subregions that each play a part in a chemical and neuronal cascade of activity. These actions allow us both to wake up and to enter different modes of sleep. For instance, when we dream, it is the reticular formation that pushes signals down to the spinal cord, actively paralyzing the rest of the body so that we don’t start actually crashing into walls when we’re dreaming we’re running around. Damage to the reticular formation is usually a pretty absolute business. Either the patient will die or he will be totally robbed of consciousness in a deep coma, unable to ever wake up.

  But while we definitely need our reticular formation in order to be aware, it doesn’t follow that our consciousness actually occurs in this primitive brain area. My PC will simply not turn on without the power supply unit, but this box is the dumbest part of my computer, having nothing to do with the processing that gives me a functioning operating system.

  A somewhat more relevant brain region is one of the main output regions for the reticular formation, the thalamus. It sits just above the brain stem, right in the center of the brain. This particular region is very special because it acts as a hub, the Grand Central Station of the brain. Its neurons receive inputs from, and send outputs to, almost every other brain region. Historically, it was viewed mainly as a sensory conduit, because the thalamus is usually the first port of call for the senses before the information is shunted to the cortex. For instance, for vision, our eyes report what they see via a thick information highway known as the optic tract, which flows into a part of the thalamus known as the lateral geniculate nucleus. This then shunts the information on to the primary visual cortex. More recently, though, the role of the thalamus has been considerably upgraded from passive relay. It is now known to be a sophisticated information-filtering and -organizing device and is thought to play a central role in consciousness.

  Patients with extensive damage to the thalamus tend to enter a so-called vegetative state. This is distinguished from coma in that vegetative patients show signs of wakefulness—they may open their eyes more in the daytime, for instance. It’s as if their reticular formation is still partially carrying out its duty of waking up the brain, but there isn’t enough neural coherence for them actually to be aware of anything when they are awake. As the thalamus is a key player in generating consciousness, it’s natural that its functions have been discussed both in terms of information processing and attention: Along with helping with the flow of information between most regions of the brain, the thalamus helps to point our attention in various directions. This explains how those in a vegetative state with damage to their thalamus can appear to be awake, but lack any directed consciousness. Imagine a life where you were alert in some sense, but your attention was never directed toward any object, any goal, any thought—would we even call such a state conscious?

  Returning to the e-mail analogy, the thalamus in our large company would take on the role of the IT department. All e-mail coming in from outside (sensory input) first filters through the servers in this large department before being sent on its way to the other departments for which it was intended. And most internal e-mails also pass through the buzzing IT servers in this center of the building before finding their correct recipients. But this IT department is a particularly proactive one. It tries to clamp down on spam before it even turns up in anyone’s inbox. And if the situation calls for it, the IT managers won’t hesitate to delay any e-mails relating to a particular subject—say pop-music listening, while adding “PRIORITY” to the subject headers of any e-mails that relate closely to the main company project of the moment.

  If the IT department servers crash (the thalamus is damaged, causing a vegetative state), there may still be power to the building from the generators in the basement—the lights will still go on in the evening and off in the daytime, and people can still use their computers, and chat in person with their office neighbors to pass on any interesting piece of local gossip. But communication won’t flow effectively through the building, the large group of managers who make the decisions will largely be ignorant of what is happening, and little real work will get done.

  BLINDSIGHT PATIENTS LEADING YOU UP A BLIND ALLEY

  It would be incredible, though, if the thalamus were the sole region in the brain that created consciousness. It is a relatively primitive region, present in any animal that has a backbone. Although it is clear that the thalamus is important for consciousness, and is directly involved in producing our experiences, it is likely that other regions are also critical.

  When the experimental study of the neuroscience of consciousness was in its infancy thirty years ago, there was really only one topic in town: blindsight. When I was an undergraduate in the mid-1990s, and we examined the question of how the brain supports consciousness, we studied little else. Blindsight still figured prominently when it was my turn to teach undergraduates some years later. Therefore, it would be remiss of me not to mention this subject, even if I don’t think it casts that much light on the field.

  Blindsight is a paradoxical, puzzling neurological condition. If a patient sustains damage to her primary visual cortex, at the very back of the brain, then she will be blind, at least in some part of her visual field. It turns out, though, that in this context “blind” can be an ambiguous term.

  The traditional story is that these patients will have no conscious experience in their damaged field of vision. If you move an object left or right in this patient’s “blind” region and ask her to tell you which direction it went, she will laugh at you, saying she sees nothing and so obviously can’t tell you—for all she knows you didn’t even move any object. If the experimenter then asks the blindsight patient to humor him—just to take a random guess, either left or right, even if she doesn’t really see a thing—lo and behold, even though she has no visual experience of this movement, she will guess correctly, much to her surprise. So the information is still there somewhere in the brain, but the experience of it, the actual seeing part, is missing. Similar stories occur for “blindtouch,” where information about touch sensations is available, even when the feel of them isn’t, and so on. Blindsight and similar conditions, the surface story argues, demonstrates a sharp fracture between sensory data and the experience of that data—a potential problem for anyone who links consciousness with information. This traditional story also implies that our experiences are generated from the primary sensory regions of our brain, since losing these regions abolishes our experiences of those specific stimuli.

  This, as I say, is a traditional view. The trouble is that it is far too simplistic. First off, most of these patients usually are only partially blinded. Some of the time they do actually have a visual experience and some of the time they don’t, even if it’s clear that their entire primary visual cortex is missing. Some patients either seem to recover slowly, or get better with practice, generating further doubt as to exactly what level of visual awareness they are reporting. And the sensory information itself isn’t perfectly intact—far from it. The patients guess better than chance, but for some patients and some aspects of vision, not by much. In line with this, some patients have been shown, when carefully tested, not to have a total absence of awareness, but merely a degraded form of consciousness, in a way that closely matches their reduced ability t
o detect visual objects.

  So we have a rather more murky reality than the traditional story typically reveals, where there is an indeterminate, shifting level of sensory experience, and a far reduced amount of information available relating to that sense.

  Thus there is no strong dissociation between information and experience here. Instead, both are reduced, but in idiosyncratic, fickle ways. It’s as if there were faulty wiring in a house, causing only one bulb out of a dozen to turn on in a room—but this single illumination only flickers on and off, creating a dim, indeterminate, shadow-laden view. In this light-impoverished place, sometimes the best that can be achieved is a hunch based on a slightly desperate internal analysis of the meager perceptual input. This may be right a little more than wrong, but you so wish the lights worked properly, and you could clearly see everything around you.

  One final point about blindsight is that although it seems at the very least to follow from such cases that the primary visual cortex is the seat for visual awareness, since this kind of experience is still heavily disrupted in these patients, even this conclusion is overreaching. An equally plausible interpretation is that the primary visual cortex is merely an early conduit for the kind of information that only later in the brain will become conscious. In analogous fashion, if a person went blind because of damage to his eyes, it would be shockingly premature suddenly to claim that all aspects of visual awareness occurred in his eyes.

  The take-home message, therefore, is that although blindsight is a strange condition, it in itself doesn’t explain too much about consciousness. But if you view blindsight instead as a patient group that sometimes can be conscious of, and sometimes be blind to, the same kind of visual stimulus, that in itself makes them potentially an invaluable set of people by which to study consciousness. The trick is to control performance, by carefully tweaking the stimuli, so that the patients are usually correct in guessing its direction of motion regardless of whether they are aware of it or not. Then, with everything controlled for, you can place the patient in the scanner with these stimuli and discover which additional brain areas will light up when the patients are conscious of the moving dot, compared with when they can’t see it (but can still correctly guess its direction). Larry Weiskrantz, who coined the term “blindsight” in the mid-1970s, carried out exactly this experiment in 1997, along with colleagues, including Arash Sahraie, on one blindsight patient in the fMRI scanner. The main additional activation for the aware condition was the lateral prefrontal cortex, which has already cropped up a few times in this book; its function is closely tied to working memory and other complex cognitive processes. But for the moment, it’s worth emphasizing that the patient’s main additional brain activity when he was aware of these moving visual objects was in generalist nonsensory regions, those that adapt their function to the task at hand.

  VISUAL HIGHWAYS TOWARD CONSCIOUSNESS

  Blindsight, although not necessarily informative on its own for how the brain represents consciousness, was useful in opening the floodgates for other researchers from neighboring fields to join the party. In the mid-1990s, monkey vision researchers also became interested in the question of whether the primary visual cortex was really the seat of visual awareness. Blazing a trail in this field was Nikos Logothetis and his team, who carried out a series of groundbreaking studies using electrodes implanted into various places in the monkey brain to record the activity from individual neurons. The monkeys were performing a now classic task in the consciousness literature known as the binocular rivalry test. If you present one image to one eye—say, a face—and a completely different image to the other eye—for instance, a house—then a person does not actually see some chimeric amalgam of a face and house. Instead, the subject will experience only a house, followed at some later time by just a face, and then a little later the visual image will flip back to a house, and so on (see Figure 6, top).

  If you give the binocular rivalry test to humans, you merely need to spend a few minutes explaining to the participant that she should press a button to indicate when her experience switches from a face to a house or if the reverse switch occurs. Monkeys can do the same task, except you need months of training to get them to carry it out.

  Logothetis discovered that the primary visual cortex wasn’t nearly as vital for consciousness as the blindsight evidence initially hinted at. Although primary visual cortex neurons in these monkeys showed an accurate representation of what was actually presented to the eyes, neuronal firing was poorly related to what the monkey “perceived”: Only a fifth of the neurons seemed to match the contents of the monkey’s reported “experiences.” So the primary visual cortex, rather than the seat of visual awareness, is instead mainly a filtered copy of what comes into the eyes. We may lose much of our visual consciousness if we lose this brain area, but that’s merely because the primary visual cortex is the first and main cortical stopover for visual information, regardless of whether or not it is conscious.

  To reinforce this point: Have you ever wondered why the world doesn’t go black every time you blink your eyes? Whenever you blink, activity in the primary visual cortex massively reduces to reflect the darkness. However, more advanced visual regions, whose main purpose is not to reflect the visual world, but to explain and predict it, can generate a sense of visual continuity by allowing our most recent perception of our surroundings to persist during the momentary gap of the eye blink.

  Instead of its role in awareness, the primary visual cortex performs basic processing to pick out the crude raw features of the visual world. It carries a map of that world in an organized way, relating closely to what the eye sees.23

  It’s worth pausing for a moment for a quick primer on the human visual brain. Let me return to the manager responsible for yellow in the large company, since he’s very much in the middle of all the action and has a particularly good perspective on the visual system:

  Once the company receives information from the security cameras at the front of the building, above the main entrance, large, thick cables take the data from these two cameras (eyes), via the IT servers in the middle of the building (thalamus), to the back of the building, where the grunts sift through it (primary visual cortex). These are not the most advanced employees in the world, which is why they’re paid minimum wage, and they never give orders—they just carry out basic tasks and accept simple instructions. Each grunt is responsible for a single pixel coming from the cameras. The grunts even sit in a vertical scaffold, so that the top right grunt works on the bottom left pixel, the one below him works on the camera pixel above that and so on. If they all held up a color based on what they were looking at, you could almost see a picture from the overall scaffold.

  These grunts send a flurry of e-mails to slightly better paid workers who are still toward the back of the building, but in offices in front of them. I’m one of these people. I don’t care about single pixels—that’s totally beneath me. I manage about a hundred pixels. So a hundred grunts, each responsible for a single pixel, constantly send me e-mails on whatever colors the cameras have seen in that location. And if yellow comes up from enough of the grunts, I get really excited and send out my e-mails. Offices near me get info about movement in just the same way, and they also don’t care about single pixels anymore, but a small region of space.

  We send out our e-mails to offices further to the front of the building. As I said before, if I send e-mails out about yellow, and neighbors send e-mails about black, then there’s some guy we e-mail who gets excited about this potential striped spy nearby. So he actually doesn’t care where in space something happens—it could be absolutely anywhere in the picture—he just cares about what it is.

  So the primary visual cortex right at the back has neurons that each code for a tiny region of space, and collectively you could recreate what the eye sees by a closely corresponding map of primary visual neurons. These primary visual neurons then send out their information to slightly more advanced, later visual regi
ons, a little more to the front, which will be combining this cruder data together to start extracting the important features. A general rule is that the further forward you go in the visual system, moving away from the primary visual cortex, a given neuron will represent more refined information about an ever larger region of space, as increasing levels of data grouping and meaning extraction occur. Here is the factory floor of visual chunk recognition.

  From the primary visual cortex right at the back, part of the information might go forward to V4, which processes color, with each neuron coding for a slightly larger region of space than in the primary visual cortex, and then a few more steps will lead to one of the endpoints at the inferotemporal cortex, much further forward in the brain, already in a different lobe, the temporal lobes, where some forms of object recognition occur. A neuron may activate now if a chair, say, turns up in any region of the left half of space.

  How does a neuron in the inferotemporal cortex represent the notion of a wasp? Aside from the fact that thousands of inferotemporal neurons will collectively be involved in this memory, these advanced neurons probably store their information via a hierarchy of connections with earlier visual regions. For the inferotemporal cortex to represent a wasp, perhaps it needs to have neurons connected with those for yellow and black in V4, those for a furry texture in V3, and so on.