(Warning: neuroscience weeds and references to gruesome animal research.)
The vast majority of neuroscientists see consciousness as a cortical phenomenon. It may be crucially dependent on sub-cortical and sub-cerebral structures, but subjective experience itself exists mainly or entirely in the neocortex. In this view, the brainstem only produces reflex responses, with anything more sophisticated coming from higher level structures.
But there’s a small but vocal minority in neuroscience who see it differently. Views in this camp vary somewhat. Some are more cautious and see the brainstem as perhaps providing a more primal version of consciousness with the cortex providing higher level aspects of it, while others see the brainstem as the primary or even sole source of consciousness.
A scientist often cited for this view is Bjorn Merker and his paper: Consciousness without a cerebral cortex: A challenge for neuroscience and medicine. (A PDF of the paper is publicly available.)
To understand what Merker is proposing, consider this diagram from the paper:
In each of the four images, the large oval on top is the cortex and overall cerebrum, while the small oval is the brainstem. The white sections in each image are where consciousness is proposed to reside, with the grey being non-conscious processes. The two top images reflect, more or less, mainstream neuroscience, with consciousness being entirely a cerebral phenomenon, although in the top right image, it is more crucially dependent on sub-cortical but cerebral structures such as the thalamus and basal ganglia.
The bottom images reflect the minority camp, with the bottom left one reflecting more cautious views involving consciousness spanning both the brainstem and cortex, and the bottom right one the more uncompromising version that only the brainstem is conscious, with the cerebral structures only supplying pre-conscious content.
Merker in the main paper seems to argue for the bottom right view, although in his response to the commentary that was published with the paper , he seems to back off a bit, retreating toward the bottom left view. (Unfortunately, the commentary and response are pay-walled, but can be found here.)
So how does Merker reach this conclusion? There’s a lot in this paper, and this post is going to necessarily be selective and highly summarized. If you’re interested in the details, I highly recommend the paper itself. It’s a fascinating read for anyone interested in neuroscience, albeit a very technical one.
Merker first cites the work of neurosurgeons Wilder Penfield and Herbert Jasper in the middle 20th century, who performed surgeries on patients with severe epileptic seizures. It was often necessary for them to remove large tracts of the patient’s neocortex. While undergoing the procedure, the patients were kept conscious with a local anesthetic so the surgeons could communicate with them and know if they were damaging their cognition.
In these procedures, Penfield and Jasper were impressed by the fact that removal of cortical sections never seemed to interrupt the consciousness of the patient. They proposed that consciousness must be maintained in lower level structures.
Merker then discusses the Sprague effect. When one side of a cat’s visual cortex is removed, despite functional eyes, the cat becomes unresponsive to half of its visual field. In human patients, similar damage results in cortical blindness (and sometimes the phenomenon of blindsight). However, when the cat’s upper brainstem is additionally damaged in a certain manner, some of the cat’s abilities to respond to visual stimuli returns.
Merker also discusses the abilities of rats that have been decorticated, that is, had their neocortex removed but with the rest of the brain left intact. These rats often retain a remarkable ability to navigate and engage in customary behavior, including reproduction, despite losing many cognitive abilities detectable to a trained observer.
Finally, Merker discusses hydranencephalic children. These are children who typically suffer a stroke in the womb that destroys much of their brain. Generally they are born with only the brainstem and a few lower level cerebral structures. Their cognitive ability seems to be roughly limited to that of newborns, although they never move beyond that stage. Despite substantially missing a neocortex, they reportedly display powerful indications of a sort of primal consciousness.
There are issues with all these lines of evidence that weaken Merker’s case. Some of them Merker admits to, but then summarily dismisses. For example, in the case of the cat, another interpretation is that the followup damage to the upper brainstem merely destroys the cat’s ability to inhibit its reflexive reactions to visual stimuli, and decorticated rats retain a lot of cerebral structures that mainstream neuroscience sees as sufficient for habitual behavior.
But I’m going to focus on a broader issue. As neuroscientist Anton Coenen asked in his commentary, “But what kind of consciousness is this?” When we use the word “consciousness”, we can mean all kinds of things, but there are at least three broad meanings that often get conflated:
- Being awake and responsive to stimuli
- Awareness with phenomenal experience
- Self reflection
When we see behavior indicating 1, we tend to assume that all three versions are present. In the case of a healthy developed human, it’s usually a safe assumption. But the further we get from healthy humans, the weaker that assumption becomes. In non-human animals, 3 may be limited to only a few primate species, and many patients in a vegetative state seem to have 1 without 2 or 3.
On inferring which level of consciousness is present based on behavior, I’m going to quote Richard Feynman on scientific observation:
The first principle is that you must not fool yourself — and you are the easiest person to fool.
Nowhere is this principle more needed than when using behavior to assess mental states in non-human animals and brain injured humans. We have to be careful about taking affect displays such as crying, facial expressions, avoidance reflexes, etc, as evidence. As intuitively powerful as they are, affect displays do not necessarily indicate conscious affective feelings. Human psychology studies show that many affect displays are unconscious. This is why body language and unguarded facial expressions are often cited as better indicators of mental states than the more conscious behavior.
The consciousness hierarchy above highlights how important it is to be clear about which type of consciousness we’re discussing. Merker, to his credit, explicitly identifies the definition he’s working with: information integration for action. And, despite my quibbles above, I do think he makes a good case that integration for action happens in the upper brainstem. But integration for action only meets the first level of consciousness.
Consider the phenomenon of mind wandering. I can be driving to work, mowing the lawn, taking a shower, or doing a host of other complicated physical tasks with little if any conscious thought going into what I’m physically doing. When driving, I can be thinking about the next blog post I’m going to write or how I’m going to handle a presentation at work. Clearly some part of my brain is doing integration for action in order for me to drive, but it doesn’t seem to be the parts we normally label as conscious, at least until something about the driving requires that I focus on it, on what needs to happen next.
In practice, most of the habitual automatic but learned behaviors described above are controlled by my basal ganglia, sub-cortical structures above the brainstem. But if a loud noise causes a startle reflex, that is handled by the brainstem. The frontal lobe cortex only seems to be involved when some degree of planning is needed, even if only planning for the next few seconds, utilizing integration for planning.
Merker seems right that what happens in the brainstem is the final integration for action, and that all action goes through it. But the brainstem itself only appears to have its reflexive reactions, reactions which can be inhibited from higher level structures. Whether those inhibitions arrive are driven by higher level integrations. These structures process an enormous amount of information that never makes it down to the brainstem.
For example, about 10% of the axons from the retina project to the superior colliculus in the upper brainstem region. Most of the remainder, including all of the axons from the color sensitive cone cells, project to the thalamus and visual cortex. This means that the redness of red and many other conscious qualities only happen in the cerebrum. That information is used by the cortex to decide which reflexive reactions in the brainstem to allow and which to inhibit. The superior colliculus does have low resolution colorless images, but we appear to have no introspective access to them.
None of this is to imply that the brainstem isn’t crucial for consciousness, particularly the first level. It arouses the cerebrum, provides the underlying impulses and valences that form the core of feelings, and generally drives the overall system toward homeostasis. Everything above it is an elaboration of those functions. But that doesn’t mean it has phenomenal awareness. What it means is what we call phenomenal consciousness is itself an elaboration of its fundamental functions.
So perhaps a better way of saying this is that Merker and those of similar disposition aren’t wrong about the brainstem’s primacy. The more cautious views aren’t even wrong that the brainstem has lower level consciousness in the sense of the first level described above. They’re only wrong to the extent they claim that primacy includes phenomenal or self reflective consciousness.
As is often the case, much of the differences between mainstream neuroscience and the more cautious views in the minority camp amount to differences in what people are willing to call “conscious.”
Unless of course I’m missing something?