In yesterday’s post I argued that a split-brain subject is not unitary qua thinker but is actually composed of two thinkers. The book also defends two further, related duality claims: that R and L are distinct subjects of experience and that they are distinct intentional agents. To many people, however, such claims seem incredible. There’s a level on which they still seem incredible to me. You can watch videos of split-brain subjects on Youtube, and even at their most “disunified,” they don’t seem psychologically dual. Naturally, they don’t seem 100% “normal”, either—but then, why would they? They’ve had major brain surgery!
There are various objections to the duality claims defended in the first part of the book. The most interesting and difficult of these consists of a large body of findings demonstrating that interhemispheric mental state interaction is not entirely indirect—not entirely mediated by action and perception. And this isn’t so surprising. Obviously all interhemispheric interaction, including indirect interaction, is carried out by non-cortical structures that remain intact after split-brain surgery. And these structures can’t be thought of as mere input- and output-channels, that merely carry incoming sensory information to and pass motor commands from the hemispheres. So, the objection from sub-cortical structures constitutes a major challenge to the simple line of argument developed in yesterday’s post, and I devote a full chapter of the book (together with a somewhat technical Appendix) to this challenge.
But I actually don’t think that this objection is, in practice, the one that carries the most persuasive weight against the duality claims. In practice what carries the most weight is just the overriding impression that a split-brain subject is one of us, psychologically speaking. Split-brain subjects just seem like one psychological being apiece. They do so to me, too. A great deal of the book is devoted to working out just why this is and what evidential import it has.
Note that to some extent, split-brain subjects are inevitably going to seem unitary qua psychological beings because they’re unitary qua animals. Each of them has one face and one pair of eyes and one mouth through which to speak. We don’t normally individuate psychological beings in terms of their psychologies; we individuate them as physical objects (usually but not always animals) and then attribute psychologies to such objects. We have presumably evolved to do so. Regardless of the facts of split-brain psychology, a split-brain subject, singular—one body, one face, one pair of eyes meeting our own—will never pre-theoretically strike us as being like two human beings psychologically speaking. The duality argument is simply that this impression is misleading.
I don’t mean that philosophers are inclined to say, “A split-brain subject, S, must be a unitary thinker, because S has one pair of eyes.” My suspicion is rather that a lot of what creates the impression of psychological unity is in fact sub-psychological. Of course, it’s an empirical question exactly which factors do evoke the unity intuition, i.e. the intuition that a split-brain subject is one of us. What philosophers usually pinpoint, qua factor, is that split-brain subjects generally behave in a unified fashion. But very little attempt has been made to specify what “unified” means here.
Sometimes it’s said that the two hemisphere systems don’t seem to be in perpetual conflict with each other. Hopefully this doesn’t mean that the only basis for distinguishing two thinking beings should be mutual physical aggression. If it means anything less than this—if it means just that R and L don’t seem to have conflicting intentions—then difficult questions arise about the evidence: given that R and L share a body, what would their having conflicting intentions look like? That is, how would we know whether their intentions were or weren’t in conflict? Moreover, it’s not totally clear why this should be a condition of their psychological individuality anyway. Multiple agents can have consonant and coordinated intentions. This is certainly true of conjoined twins, that is, psychologically distinct agents who are partially “co-embodied.”
Sometimes it’s said that split-brain subjects behave in too unified or too normal a fashion outside of experimental conditions, specifically, for the duality claims to be true. The thought, I think, is that a split-brain subject’s having two minds (being two thinkers) should at least be manifest somehow outside the lab: that is, being dual qua thinker shouldn’t be largely epiphenomenal. Although I agree with this, I question the assumption that split-brain subjects’ psychological duality isn’t manifest outside the lab.
For the most part, we don’t actually know much about how split-brain subjects behave in their day-to-day lives. I found only a single study (Ferguson, Rayport, and Corrie, 1985) devoted to studying this issue. Six split-brain subjects were asked whether they would allow an observer to follow them around for a few days and just see how they got on, interview their families, and so on. All six subjects agreed. The authors of the paper concluded that all six of the subjects exhibited a number of behavioral oddities that were apparently the ongoing consequence of their surgeries. Now, this was a single observational study, with relatively little work devoted to explaining the observations, though the authors tentatively note that persistent interhemispheric conflict might explain some of what they saw. But at a minimum, this one attempt that was made to examine split-brain subjects outside of experimental conditions didn’t paint a picture of perfect normalcy or unity. It’s too ambiguous to support the duality claims, per se. But it doesn’t undermine it.
That said, there is one respect in which I think we do know that split-brain subjects act in an (in some sense) unified fashion and which does have evidential import vis-à-vis the duality claims. It’s this: that a split-brain subject seems to think of herself as a unitary psychological being. Or—to put the point consistently with the duality claims—neither R nor L seems to recognize the existence of a second psychological being co-embodied with it—a being other than the psychological being each assumes that it is. I defend and explain the significance of this claim in tomorrow’s post. If true, it indicates an interesting psychological respect in which R and L are not like two of us.
References:
Ferguson, S., Rayport, M., and Corrie, W. 1985. Neuropsychiatric observations on behavioural consequences of corpus callosum section for seizure control. In A. Reeves (Ed.), Epilepsy and the corpus callosum (pp. 501–14). Plenum Press.
Hi Lizzie, this question is partly based on your discussion of the Hogan twins in the last post’s comments, but I figured I’d put it here because it’s also somewhat about the objection from sub-cortical structures.
In referring to the Hogan twins, you suggested they might not have ‘fully discrete minds’, suggesting some sort of overlap, though your elaboration was restricted to claims about direct interactions, I think. Something suggested by ‘not fully discrete minds’ is that token mental states (experiences, thoughts, or whatever) might be literally shared – belonging both to Krista and to Tatiana. At least formally, that looks like a different claim than just that mental states of one interact directly with mental states of the other.
And the importance of sub-cortical structures seems like it might push in the same direction: that some sort of mental activity, based wholly or partly or primarily in the sub-cortical structures, could literally belong to (=”be part of the mind of”?) both L and R. Is that implied by, or allowed by, your analysis?
Hi Luke,
You ask whether the possibility of two thinkers with non-discrete minds suggests that thinkers can share some token mental states. Certainly I don’t think that the former implies the latter. If we take the case of the Hogan twins again–suppose Krista’s eyes are closed, Tatiana sees a banana, on the basis of which Krista, too, comes to enjoy a visual experience as of a banana (perhaps we want to say even that she sees the banana)–we can equally well describe this as a case in which the girls have the same type of visual experience, rather than saying that they share a token visual experience. Indeed it seems to me that the former is clearly the better description. A funny example of this: the girls’ parents apparently learned that Tatiana dislikes ketchup when they saw her trying to scrape it off her tongue, though only Krista was eating it. One could argue that there is some sort of purely gustatory experience a token of which both girls shared–but why, given that this experience seems to have been incorporated into different psychologies?
That said, I do think the possibility is an interesting one, and I’m not sure I’m against it. I don’t know if you’re familiar with the partial unity model of split-brain consciousness (defended most explicitly by Lockwood, 1989), but it’s one according to which there are RH experiences that are unified only with other RH experiences, LH experiences that are unified only with other LH experiences, and then some (probably subcortically sustained) experiences that are unified with all of the above. The model raises issues similar to those raised by the possibility of thinkers that share token mental states. It has not been particularly popular amongst philosophers (Lockwood himself later wrote that he wasn’t sure the idea really made sense!), but I have issued (in other work) a qualified defense of it, that is, a defense against the specific objections to it that have been popular amongst philosophers. I’ll give the reference for the original Lockwood paper and for my own paper on it; the latter includes references to other philosophical works on the partial unity model.
Lockwood, M. 1989. Mind, brain and the quantum. Blackwell Publishers.
Schechter, E. 2014. Partial unity of consciousness: A preliminary defense. In D. Bennett
and C. Hill (Eds), The unity of consciousness and sensory integration (pp. 347–73). MIT Press.
Thanks! So you take minds being ‘not fully discrete’ to mean that they ‘contain mental states which directly interact’, but for a single mental state to be shared would require it to directly interact, in kind-defining ways, with enough of the other states in each of two minds, to count as belonging to both minds. Is that right?
I can see that it doesn’t look like this is going on in the Hogan twins, but it does seem like it might be happening in the split-brain phenomenon – with emotional experiences, maybe, or others that are more handled by the subcortical brain. I do know the partial-unity model, which is similar but is compatible, as I understand it, with there being only a single subject.
I was more thinking: suppose we accept the duality claims, so we think there are two minds. Why *not* think that some of their states are shared, given how much of their neural basis is shared? (I ask mostly because I think a lot of people, implicitly or explicitly, see the very idea of shared mental states as theoretically incoherent, or at least as problematic enough to want to avoid positing them unless forced.)
Do you discuss Pinto et al (2017) _Split brain: divided perception but undivided consciousness_ in the book?
Hi David,
Thanks for your question. Yes, I do discuss the Pinto paper you cite in Chapter 2, which concerns the structure of consciousness in split-brain subjects. (Pinto and I also recently participated in a debate, with Joseph LeDoux, about minds and consciousness after split-brain surgery.) I have to step out for a few hours but can speak about the paper and my thoughts about it in general terms, upon my return–or if you’d like to ask a specific question about it I can of course answer that instead!
Hi again David–
Okay so, Pinto et al. describe a number of experiments of two broad types. In one type of experiment, a stimulus is presented in one VF or the other, and the subject is asked to respond to the stimulus in some way (identifying it, say) either unimanually or verbally. Call this type of experiment a “response experiment.” For this type of experiment we can compare performance on uncrossed vs. on crossed response trials. Uncrossed trials = stimulus presented in RVF, subject asked to respond with the right hand or verbally OR stimulus presented in the LVF, subject asked to respond with the left hand; crossed response trials = stimulus presented in the LVF, subject asked to respond with the right hand or verbally, or stimulus presented in the RVF, subject asked to respond with the left hand. Pinto et al. found however that their subjects were equally accurate on crossed as on uncrossed trials.
In another type of experiment, two stimuli are presented, either both in LVF, both in RVF, or one in each VF (bilateral visual field i.e. BVF), and the subject is asked to relate or compare the stimuli in some way (saying whether they’re the same or different, say) either unimanually or verbally. Call this type of experiment a “comparison experiment.” For this type of experiment we can compare performance on unilateral comparison trials vs. on cross comparison trials. (Unilateral comparison trials: both stimuli in LVF OR both stimuli in RVF; crossed comparison trials = stimuli presented in BVF i.e. one in LVF one in RVF.) Pinto et al. found that while the subjects could perform unilateral comparisons, they were at chance on cross comparison trials.
The question is how to make sense of this pair of results. The original split-brain research said that the subjects were incapable of cross-comparisons AND of crossed responses. Or at least, when you read introductory sources on the split-brain phenomenon (including my first blog post here the other day), they state or imply that split-brain subjects have a deficit or a frank incapacity to make crossed responses–e.g. that stimuli presented on the left side of space can’t be responded to using the right hand.
As Pinto and colleagues point out in another 2017 article (“The split-brain revisited”, in Trends in Cog Sciences), however, the actual research papers (as opposed to e.g. textbook presentations of the split-brain phenomenon) present a much messier picture than this and certainly do NOT support the claim that split-brain subjects are incapable of making crossed responses. In fact they clearly show that split-brain subjects ARE capable of such responses, though often they’re better at uncrossed ones. Since Pinto et al. find in fact NO difference, with respect to accuracy, between crossed and uncrossed responses, they argue that this shows that although perception and (in a sense) cognition may be interhemispherically divided after split-brain surgery, *agency* isn’t–stimuli presented in either visual field can be responded to using any response modality. So there is ultimately one *agent* of RH and LH perceptual and cognitive information.
There’s a lot to say about the paper. It raises the question of individual differences, of psychic changes over time (the apparently frank incapacity for crossed responses is characteristic of the VERY early recovery period), of what counts as “normal” performance in split-brain subjects (e.g. what if accuracy rates are the same between crossed and uncrossed responses for a particular experiment type, but accuracy rates are sub-normal even for uncrossed responses), etc. One possibility that I don’t think Pinto et al. clearly rule out is that both hemisphere systems have the capacity to control all the relevant kinds of behavior (right hand behavior, left hand behavior, and verbal behavior of certain limited kinds–again R certainly can’t speak in full sentences, but there is some reason to think that it can speak in single words or very short phrases, which seem to have been the kinds of responses required in the verbal response trials in this paper). They also don’t rule out the involvement of cross-cueing–e.g. one hemisphere system shaping the hand into a pointing configuration while the other directs it to the appropriate target.
That said–Pinto et al.’s interpretation of their results is that they demonstrate some kind of direct interhemispheric motoric interaction, and there are a number of split-brain papers that I think do clearly demonstrate its existence. (There is interhemispheric interaction of other kinds, too, but there seems to be more of it in the motor realm.) From a philosophical perspective, the interesting question is what kinds of motoric integration suffice for integrated (unified or unitary) agency. (This, in turn, bears on the question of conscious unity, to the extent that the duality claims stand or fall together. One might think that a single agent could have a disunified consciousness–but one might not, and presumably it depends in part upon the proper theory of consciousness.)
What i argue in the book is that the truth of the 2-agents claim requires that R and L have the capacity for *intentional autonomy*–to make decisions independently of the other. Of course, their ability to actually *act* independently of the other is highly constrained, since they share a body. That said I don’t think even that is totally impossible for them; I think there is reason to think that R and L can at least initiate actions independently of and contrary to the will of the other, though it’s presumably difficult (for obvious reasons) for either of them to satisfactorily complete an action without at least the acquiescence of the other. But as for intentional autonomy–autonomy in decision-making–I do think that they have that. First, it seems as though they *can* directly interact more than they *have* to, if that makes sense. Second, it seems as though the motoric interaction they’re capable of is at a fairly low level of the motor control hierarchy–it doesn’t seem to be interaction at the level of intentions, for instance.
I’m not sure how much of this hit what your precise interest was in the Pinto paper, but hopefully it speaks to some of it!
Thank you for your comments, Elizabeth. Yes I was aware of the messiness.