On the Interpretation of Neuroscientific Findings

By now, many readers will know that I’m a devotee of Peter Hacker’s Wittgenstein-infused critique of neuroscientific research. I recently came across a podcast of a lecture that he gave on the topic and the corresponding YouTube video above. Here’s the iTunes podcast link (it’s episode #92). So, if you’re interested to learn more about this but don’t want to commit to reading the whole tome, the lecture will get you the gist of it in just under an hour.

Here, I’ll take a stab at applying this kind of analysis to newly-published work regarding the integration of brain activities while driving and listening to different kinds of audio. But first it’s necessary to review some foundational research on “split brain” patients. Once upon a time, severe generalized epilepsies were treated with commissurotomy (corpus callosotomy), the idea being that if the epileptic discharges could at least be confined to one brain hemisphere, this would be less disabling (and life-threatening) than if the patients kept having generalized seizures. Starting in the 1960s, Profs. Roger Sperry and Michael Gazzaniga published some extremely interesting studies on these patients. Here’s a diagram of one of their experimental setups:

Image result for split brain shovel claw

The subject fixes his vision on the center of the screen, while two different images are projected such that each is visible only in the left or the right visual field. He is asked to point with his right hand to the picture that corresponds to what he sees on the screen; he chooses the chicken. When asked why he so chose, he responds that the chicken claw goes with the chicken. When asked to point with his left hand to the picture that corresponds to what he sees, he chooses the shovel. And here’s the fascinating part: When asked why he chose the shovel, he doesn’t say anything about the snow but rather that the shovel is needed to clean out the chicken shed!

How should we interpret this? The researchers’ explanation is that the left hemisphere “sees” the chicken claw, and since language functions are lateralized to the left hemisphere, this hemisphere can state what it saw. The right hemisphere, however, “sees” the snow scene but can’t speak and is disconnected from the language-endowed left hemisphere. Therefore, the left hemisphere, which is unaware of what the right hemisphere “saw”, confabulates an explanation based on what it “saw”.

Prof. Hacker argues that this is a nonsensical explanation. Neither brains nor their individual hemispheres can see, think, be aware of anything, transfer information, or talk–people do those things. Certainly, people must have properly functioning brains in order to exercise their perceptual, attentional, cognitive, and linguistic abilities, but it doesn’t make sense to attribute to some portion of a person’s anatomy what is conceptually only properly attributable to the person. The conceptual nature of the argument is critical: It is not the case that these “split brain” experiments have been contravened by new empirical research or invalidated by the discovery of a design flaw. Rather, this perfectly valid empirical research has been interpreted in a way that transgresses the bounds of sense. Questions of sense and nonsense precede questions of truth and falsity.

How should the remarkable behavior of these subjects be explained? Prof. Hacker describes it as a dissociation of normally coordinated capacities. Normally, our visuoperceptual and language capacities are fully coordinated. In post-commissurotomy subjects however, these capacities are partially dissociated and result in confabulation under this experimental condition. The point may become clearer if one considers it from a linguistic perspective: Does the subject see the snow scene or does he not see it? As ascertained by the subject’s motor response, he saw it; as ascertained by his verbal response, he did not. The very meaning of the concept “to see” has been strained by the dissociation of neurological functions resulting from commissurotomy.

As fascinating as these particular experiments are, such dissociation of functions should not seem esoteric to us as clinical neurologists; we traffic in them! Consider the relatively mundane case of supranuclear gaze palsy: We ask the patient to look down, but he can’t. Then we ask him to tilt his head down a bit and then rapidly tilt it upwards. We see that the eyes move downward. The voluntary control of vertical gaze is thus dissociated from the reflex; can he “look down” or not? Or consider a patient with vascular dementia: We give him a few words to remember and then ask for them later in the evaluation. He can’t restate the words spontaneously, but he can select them from multiple choice lists; did he “remember” them or not? Examples like this abound–localization and differential diagnosis in clinical neurology depend heavily on an understanding of such structure-function relationships.

This brings us to the present research, which sought to demonstrate a functional split in brain functions, as opposed to the structural split resulting from commissurotomy. Subjects were scanned with fMRI while performing a driving simulation under two different auditory conditions. In one, they were listening to GPS driving instructions–an “integrated” task in that the subject had to integrate the auditory instructions with her visual perceptions and motor responses. In the other, they listened to a radio show–a “split” task in that the subject would presumably not integrate what she heard on the radio with the visuoperceptual and motor aspects of driving. And indeed, the fMRI results showed much higher neural system integration during the GPS condition than during the radio show condition. The paper describes how such integration was measured, but I’m not knowledgeable enough to critique the experimental methods in such detail. I want to stipulate that the fMRI methods are valid and instead focus on what I think is a conceptual problem with one of the study’s interpretations (emphasis added):

An intriguing question is what happens to consciousness when driving while listening in the split condition. Is there a single conscious stream, with attention deployed primarily to a dominant task, typically listening, and much less to driving? Alternatively, does driving become unconscious, as on autopilot? Or, does a normally integrated conscious stream split into two separate conscious streams that coexist within the same brain, as indicated by studies of patients with an anatomically split brain? 

What does it mean to have “two conscious streams” and do post-commissurotomy patients have such? As with the case of “seeing”, “looking down”, and “remembering”, a lot turns on the definition of “conscious”. There are at least three types of consciousness: One can speak of being conscious as opposed to unconscious–sleeping, post-ictal, or comatose for whatever reason. One can be conscious of something, such as the distance between my vehicle and the one in front of me, the fact that I’m late for a meeting, or the pleasant aroma of the coffee I’m drinking. One can also be self-conscious, which itself has several different aspects. When considering structural or functional segregations of brain function, it seems that the second sense of “consciousness” is the most relevant one.

Most of us have had the experience of driving from point A to point B but being so lost in thought that we don’t recall any of the details of the journey. Or we may come to the sudden realization that we’ve overshot our highway exit. But does it make sense to say that, in addition to the stream of consciousness that we can recall–our cogitations along the way, the songs we heard on the radio, etc.–there was a second stream of consciousness that we can’t recall? That this stream of consciousness was responsible for the operation of the vehicle? No, I think that the term “conscious” means that if we weren’t aware of it along the way, and can’t recall anything about it later, then we simply weren’t conscious of it. People are able to carry out many activities without being conscious of them.

Again, let’s consider a mundane example: Think about the last time you walked from your car to your house. You may recall what you were carrying, what you were thinking about, etc. However, you probably don’t recall much, if anything, about the mechanics of the walking–the height of your step, the length of your stride, how it felt when your right heel struck the floor upon your first step into the house, how it felt in the knee and hip. Undoubtedly, there was neural activity underlying all of this, and specifically neural activity in the brain and spinal cord. I hypothesize that if it were possible to conduct an fMRI study of this behavior, one would find results similar to the “functional split” condition in the driving experiment above. That is, there would be no integration between the neural activity underlying your cogitations while walking and that underlying the walking itself. But it doesn’t follow from this that there was a second “conscious stream” that was responsible for the walking, any more than it implies the existence of a “spinal soul“.

I’ve never had the opportunity to examine a post-commissurotomy patient. My understanding is that they function quite normally in most situations, and strikingly abnormally under certain specific conditions. Undoubtedly, the study of the relationships between the behavior of such patients and their structural and functional neuroanatomy is very interesting and leads to insights about the neural conditions that must obtain in order for a person to exercise capacities such as perception and speech. But I don’t think it makes sense to say that such patients have two streams of consciousness. I assume that, like all people, they are conscious of some things and not conscious of others. Uniquely, they exhibit under certain conditions a dissociation between normally coordinated functions.

Likewise, the present research provides insights into the functional neuroanatomy underlying driving, listening, driving while concentrating on what’s being said and listening while not concentrating on driving. This is all very interesting, and I think the best interpretation vis a vis consciousness is, as the authors offer in the first two (related) hypotheses quoted above, that the subjects simply weren’t attending to the driving task while listening to the radio program–they were conscious of the latter but not the former. The study doesn’t show that there are two conscious streams because that concept just doesn’t make sense.


Sasai, S., Boly, M., Mensen, A., & Tononi, G. (2016). Functional split brain in a driving/listening paradigm Proceedings of the National Academy of Sciences, 113 (50), 14444-14449 DOI: 10.1073/pnas.1613200113

About Justin A. Sattin

I'm a vascular neurologist and residency program director. I created this blog in order to share some thoughts with my resident and other colleagues, and to foster my own learning as well.
This entry was posted in Medical Knowledge and tagged , . Bookmark the permalink.

One Response to On the Interpretation of Neuroscientific Findings

  1. Michel Accad says:

    Thanks for the comments and the link to Peter Hacker’s lecture. One can only hope for a return to more sensible Aristotelian anthropology!

Comments are closed.