In this commentary I note that Dennett's Multiple Drafts theory is not incompatible with the "potential duality of mind" in normal subjects or split-brain patients. I then argue that if we acknowledge the slippery slope of selfhood that Multiple Drafts encourages, we can see how Puccetti's criticisms of Dennett's three "arguments" miss the mark.
2. Dennett is pretty clear on this point. Consider his aligning himself with various "Society of Mind" views, which conceive of mind "as a coalition or bundle of semi-independent agencies" (p. 260; see also Dennett's ch. 8, and his summary pp. 253-4). To conceive of mind in this way is to conceive of mental activity as existing along a rather slippery slope. One consequence of this is that it is no longer very clear just how to count or quantify "minds," as such talk becomes quite metaphorical. Depending on one's rhetorical interests, one may perfectly adequately say that normal human subjects have one mind, extremely many minds, or any arbitrary number in between (see pars. 8, 10 below). The same is true for "centers of narrative gravity," though perhaps here we tend to be more parsimonious. Thus Dennett explicitly sanctions the possibility of multiple "centers" on p. 425. Similarly, in the passage that Puccetti cites in par. 15, Dennett explicitly admits the possibility of multiple "selves" in split-brain patients. So, no incompatibility with nor problem for MD there.
3. That Puccetti misunderstands Dennett's views is evident from his summary of them: "If the mind is only an abstraction derived from these multiple drafts, then it is not based in anything readily countable... such as cerebral hemispheres" (par. 3). Whatever exactly Puccetti means by "based in," this just isn't Dennett's view. As Dennett makes clear throughout his book, he takes higher mental activity (and abstract "minds" or selves) precisely to be "based in" those very countable hemispheres. It's just that that needn't entail their own countability.
4. What these considerations suggest is that MD shouldn't be directly involved in deciding whether split-brain patients should be construed as having one or two selves, since it's compatible, in effect, with either answer. Instead, it should be appealed to rather indirectly. That is, once we loosen our hold (as MD encourages us to do) on the "all-or-none" conception of minds or selves, we can ask under what conditions we're disposed, if mainly pragmatically, to count minds and selves in the various ways, without demanding or expecting that there be facts of the matter about the count. Puccetti does not appear, after all, to be wedded specifically to "Cartesian Materialism," the view that Dennett is opposing; rather, he's attached to the "all-or-none" conception of minds and selves that tends to go along with Cartesian Materialism. But this conception he doesn't argue for. So let's see what happens, as we examine Puccetti's treatment of Dennett's three "arguments," once we give up the all-or-none conception.
5. FIRST ARGUMENT. In par. 6, Puccetti comments on the passage from Dennett in par. 5: "Dennett seems to be confusing the time it takes to demonstrate the cognitive independence of the disconnected right hemisphere under testing conditions with the post-surgical span of consciousness of that hemisphere." But that's not what's going on.
6. To see this, note Dennett's stress in the passage cited on the fact that, in those rare cases where split-mind "symptoms" are observed, they only are so in extremely contrived circumstances. Dennett doesn't mean this to imply that independent mental activity stops occurring in the two hemispheres the moment the subject leaves the lab; after all, MD sanctions lots of such mental activity occurring (i.e., agents operating) all over the brain. It's rather that, except in lab circumstances, both hemispheres have access to just about all the same information -- which deprives them, as Dennett says in the passage, of the sort of independent "autobiograph[ies] of which fully fledged selves are made." (And that in non-lab circumstances the two hemispheres have access to just about all the same information is indeed the "standard interpretation," to which even Puccetti subscribes [1973, p. 343]).
7. So Dennett uses autobiographical, not mental, independence as a criterion (or guide) for individuating selves, a fact most clearly illustrated in his admitting the possibility of "Fractional Personality Disorder" on pp. 422-3. But since autobiographical independence is a matter of degree, so too is selfhood: the more clearly we can distinguish autobiographies, the more we tend to grant separate selfhood -- or the more there are separate selves in question. (This needn't preclude there being other, perhaps less metaphysical, criteria for selfhood too, such as moral ones; cf. par. 12 below.)
8. SECOND ARGUMENT. Here Puccetti mistakenly sees Dennett contradicting himself vis-a-vis the First Argument. This is Dennett from the First Argument, as above: the brief moments in the lab during which the hemispheres are informationally distinct isn't enough to warrant distinct selfhood, which, roughly, requires more extensive and prolonged independence. (Dennett would probably grant a second self, for example, were the patient to spend the rest of his life hooked up in the lab.) But now this view is simply illustrated by the Second "Argument." Here Dennett is appealing to fairly common intuitions that the distinction between Juggle and Boggle is rather silly, not only because it's unverifiable but also because it's entirely unmotivated. If we were to accept it, after all, why not also believe there's a third character, a Dr. Doogle, who gets in there too? Or a fourth or fifth character, etc.? What this is ultimately meant to show, again, is the slippery slope characterizing autobiographical independence, and thus selfhood.
9. Is the dispute, then, the more trivial one about just how informationally distinct the hemispheres are? Puccetti calls them "profoundly nonidentical," but surely this is an exaggeration, especially given his own acknowledgment that their differences stand against a background of overwhelming similarity (e.g., in Puccetti, 1973). And it can't be that to distinguish their selves he relies not on the hemispheres' informational differences but on their different abilities, since then he'd have to admit, very counterintuitively, that in acquiring new abilities we acquire (or become) different selves. (And just which abilities matter? Bicycle riding? Algebra? Or is it that in learning to speak we acquire a different self? But why should this ability have special priority? And what about learning a second language? etc.)
10. We can take this even further by noting that if the differences between the hemispheres are sufficient for distinguishing selves then we seem obligated to go on and divide each hemisphere into many selves as well. After all, different parts of the (say) left hemisphere access different information, and have different abilities, so they should count as different selves. Now Puccetti could resist this move by arguing that these different areas should be grouped together into a single self. But then we'd ask, by what criteria would we group them? The prime candidates would no doubt include their coherence, cooperation, potential access to the same general pool of information -- all the properties in virtue of which we're inclined to group the two hemispheres into a single self! In short, Puccetti can't have it both ways: if the interhemispheric differences are sufficient to distinguish selves, then the intrahemispheric differences will be too; if the intra aren't, then neither are the inter.
11. THIRD ARGUMENT. It's not clear what Puccetti takes to be an argument in the passage he cites, nor is it clear to what end he criticizes Dennett's description of a right-hemisphere self. Puccetti denies, for example, that the disconnected right hemisphere could "want to tell the world what it is like to be [it]" (par. 10), since it doesn't have a capacity for speech per se. But that's a fairly trivial point. After all, the right hemisphere is capable of performing what we can see as some rather sophisticated speech acts (cf. cross-cuing above), which is all Dennett needs to make his point. It's also worth noting that Puccetti (1973) observes that the right hemisphere has "language comprehension and at least a rudimentary verbal conceptual scheme" (p. 342) -- so why is he denying Dennett's sketch here? In any case, even if Dennett's stressing the anxiety is mistaken, his overall description of a "right self" is certainly sufficiently accurate. Other than the anxiety, there's nothing there that Puccetti should need to criticize.
12. CONCLUSION. A single mistake, I think, runs through Puccetti's target article, namely, a failure to distinguish between having a mind (or mental activity) and having a self. It's possible, in effect, to have the former without the latter, and that point accounts for much of Dennett's interpretation of the split-brain data. Moreover, Puccetti does not confront the possibility that having minds and selves could be a matter of degree, which makes the counting problem that much more elusive. Or elusive metaphysically, to be more precise: we can use all sorts of other considerations, such as moral ones, to guide our daily practice. This last point, fleshed out, would address Puccetti's closing analysis of the left hemispherectomy and neocortical death cases.
Dennett, D. (1991). Consciousness Explained. Boston: Little, Brown.
Puccetti, R. (1973). Brain Bisection and Personal Identity. British Journal for the Philosophy of Science, 24, 339-355.
Puccetti, R. (1993) Dennett on the Split Brain. PSYCOLOQUY 4(52) split-brain.1.puccetti.