| |
Barry Dainton's book, Stream of Consciousness: Unity and Continuity in Conscious Experience, can be purchased from Amazon.Com |
|
|---|
The Introspectibility Thesis
Cody S. Gilmore
Department of Philosophy
Princeton University
1879 Hall
Princeton, NJ 08544
U.S.A.
gilmore@princeton.edu
Copyright (c) Cody S. Gilmore 2003
PSYCHE, 9(05), February 2003
http://psyche.cs.monash.edu.au/v9/psyche-9-05-gilmore.html
KEYWORDS: unity of consciousness, co-consciousness, introspection, introspectibility, headache argument, shrub argument, split-brain argument
COMMENTARY ON: Dainton, B. (2000). Stream of Consciousness: Unity and Continuity in Conscious Experience. London: Routledge.
ABSTRACT: According to what Barry Dainton calls the 'Strong Introspectibility thesis', it is a necessary truth that mental states S and S* are co-conscious (experienced together) if and only if they are 'jointly introspectible', i.e., if and only if it is possible for there to be some single state of introspective awareness that represents both S and S*. Dainton offers two arguments for the conclusion that joint introspectibility is unnecessary for co-consciousness. In these comments I attempt to show, first, that Dainton's arguments fail, and, second, that joint introspectibility is actually insufficient for co-consciousness. (As to whether it is also unnecessary, I take no stance.)
Barry Dainton's fascinating and insightful book on the unity of consciousness explores a huge swath of heretofore poorly charted territory. In these comments I shall confine my attention to just one small corner of this territory, the so-called 'Strong Introspectibility thesis':
Co-consciousness is constituted by introspectibility: experiences are co-conscious [i.e., experienced together] because they are introspected or introspectible. A group of token experiences are co-conscious if and only if they are either the actual or potential objects of a single introspective awareness (2000: 35, Dainton's emphasis).
The I-thesis (as I shall call it) is a reductive account of co-consciousness; i.e., it purports to give informative, individually necessary and jointly sufficient conditions for a pair of mental states to be co-conscious. It says: joint introspectibility is necessary and sufficient for co-consciousness.
Since Dainton regards the I-thesis as one of the main rivals to his own (non-reductive) account of co-consciousness, and since his defense of his own account consists entirely of a series of objections to its main rivals, much rests on Dainton's case against the I-thesis. Accordingly, he aims for redundancy. As I read him, he attempts to refute the I-thesis twice over: he offers two arguments each of which purports to establish that a pair of mental states can be co-conscious despite failing to be jointly introspectible, hence that joint introspectibility is unnecessary for co-consciousness.
In these comments I shall attempt to show, first, that both of Dainton's arguments against the I-thesis are unsuccessful and, second, that joint introspectibility is actually insufficient for co-consciousness. (As to whether it is also unnecessary, I take no stance.)
Before we begin our examination of the I-thesis itself, we ought to contrast this thesis with two of its close relatives. First is what we might call the Introspection thesis:
Necessarily, for any (token) mental states S and S*, S and S* are co-conscious if and only if they are jointly introspected, i.e., if and only if there is some single state of introspective awareness that represents both S and S*. (See, e.g., Parfit 1984: 250-51)
The Introspection thesis says, in other words, that 'joint introspectedness' is both necessary and sufficient for co-consciousness. Dainton has very little patience for this claim. He takes it to be obvious that we often have pairs of co-conscious perceptual experiences (e.g.) that fail to be jointly introspected, hence that joint introspectedness is unnecessary for co-consciousness.<3>
But won't it always be true in such cases that the relevant co-conscious experiences at least could have been jointly introspected, even if they were not actually so? It is this suggestion that gives rise to the Introspectibility thesis, which takes it to be a necessary truth that experiences are co-conscious just in case they are at least jointly introspectible.
There are, however, a number of different things one could mean by saying that a pair of experiences 'could have been' jointly introspected. To see this, consider an attempt by Christopher Hill to refute the I-thesis:
There are lower animals who appear to have unified sensory fields but who do not display the sort of conceptual sophistication that a being must have in order to be capable of having introspective beliefs. Consider raccoons. (1991: 232).
Call this the 'raccoon argument'. Hill is probably right on the following counts: (i) raccoons have co-conscious experiences, (ii) raccoons lack the 'cognitive capacity' to be introspectively aware of their experiences; i.e., raccoons do not actually have the right sort of physical or causal-functional make-up to engage in this sort of introspection. Therefore, it seems to me that Hill's raccoon argument succeeds in refuting the following version of the I-thesis:
For any mental states S and S*, any possible world w and any time t in w, S and S* are co-conscious at t in w if and only if: (i) S and S* occur at t in w, and (ii) the owner of S and S* in w has a physical or causal-functional make-up at t in w that gives it the cognitive capacity to jointly introspect S and S* at t in w.
Call this the 'ICC-thesis'. It states that joint introspectibility understood in terms of actual cognitive capacities is both necessary and sufficient for co-consciousness. As the raccoon argument demonstrates, the ICC-thesis is implausible: this sort of joint introspectibility seems to be unnecessary for co-consciousness. A raccoon could have co-conscious experiences despite lacking the actual cognitive capacity to introspect them jointly.<2>
Nevertheless, for any co-conscious raccoon experiences e and e*, there is probably some sense in which e and e* 'could have been' jointly introspected. Suppose, for example, that e and e* are co-conscious experiences that belong to raccoon R. Then presumably there is a possible world in which, prior to the occurrence of e and e*, R undergoes some brain-enhancing procedure and thereby acquires the cognitive capacity for the relevant sort of introspection. When e and e* occur, R can and does jointly introspect them. For this reason it seems to me that the raccoon argument does not refute the following version of the I-thesis:
For any mental states S and S*, any possible world w, and any time t in w, S and S* are co-conscious at t in w if and only if (i) S and S* occur at t in w, and (ii) S and S* are jointly introspectible; i.e., there is some possible world in which S and S* are jointly introspected.
Call this the 'IP-thesis'. It says that two mental states are co-conscious just in case it is possible (somehow or other) for these states to be jointly introspected. In what follows, I shall construe Dainton's I-thesis as the IP-thesis.
We turn now to the first of Dainton's two arguments against the I-thesis. The argument has three main premises: (i) some pairs of co-conscious experiences have at least one member that is not introspected, (ii) introspection affects the phenomenal character of its objects,<3> and (ii) each (token) mental state has its phenomenal character essentially. From this appealing starting point, Dainton reasons as follows:
The idea that we could have introspected experiences we did not in fact introspect only makes sense if . . . introspection . . . does not alter the character of the relevant experiences. Since the identity of an experience depends in part on its phenomenal character, if introspection affected the character of an experience we clearly could not introspect experiences that we did not in fact introspect. It is also clear that . . . introspection does influence the character of experience. . . . Suppose that five minutes ago I was not . . . introspecting my slight headache. According to the . . . I-thesis, I could have been attentively aware of this sensation if I had chosen to be. But if I had so chosen my headache would very probably have intensified, and so the pain I would have been reflectively aware of would not have been numerically the same pain as I actually had. . . . Since our experience does exhibit this dependency, it will often be the case that experiences we did not introspect could not have been introspected" (2000: 37-38, my emphasis).
Now, to complete the argument, consider Dainton's headache experience e and suppose that it is co-conscious with some other experience e*: e could not have been introspected, and a fortiori e and e* could not have been jointly introspected. So e and e* are co-conscious despite failing to be jointly introspectible, hence the I-thesis is false.
Call this the 'headache argument'. Is it successful? No, not even if we grant its premises. The flaw lies in Dainton's inference from (1) to (2):
(1) If it had been the case that
(A) Dainton was engaged in the relevant sort of headache-directed introspection at time t,
then it would not have been the case that
(B) the object of this introspection was e.
(Rather, the object of his introspection would have been some phenomenally distinct, hence numerically distinct, experience, e'. If we let '[]-->' be the counterfactual conditional sign and 'not-' be the negation sign, then we can express (1) symbolically as follows: A[]--> not-B.)
(2) Therefore, it could not have been the case both that:
(A) Dainton was engaged in the relevant sort of headache-directed introspection at t, and
(B) the object of this introspection was e.
(Letting 'POSS' be the possibility operator, we can express (2) symbolically thus: not-[POSS(A&B)].)
It is uncontroversial that the counterfactual conditional p[]-->q does not entail the denial of compossibility not-[POSS(p&q)]. Equivalently, it is uncontroversial that the counterfactual conditional is weaker than the strict conditional.<4> This can be seen very easily by considering a different pair of propositions with the relevant logical forms:
(1*) If Gore had won the 2000 election, Laura Bush would not have become First Lady in 2001. Symbolically: A*[]-->not-B*.
(2*) It is not possible that both Gore won the 2000 election and Laura Bush became First Lady in 2001. Symbolically: not-[POSS(A*&B*)].
(1*) clearly does not entail (2*) since, despite the probable truth of the former, the latter is plainly false. After all, it could have been the case, to take just one specific course of events, that Gore won the election, became a widower, and shortly thereafter married a recently divorced Laura Bush, who then ascended to First Lady-hood as Gore's wife upon his inauguration. There are possible worlds in which this occurs. But it seems likely that the closest possible world in which Gore won the election is a world in which Laura Bush did not become First Lady, hence that (2*) is true. (I adopt possible worlds semantics here only for the sake of vividness; my claim that the relevant counterfactuals do not entail the corresponding denials of compossibility is independent of this semantics.)
Similarly, we can formulate a number of more or less specific hypotheses that entail (1) while entailing the negation of (2). Here is one.
H1 The principle that introspection affects the phenomenal character of its objects is a counterfactual supporting but contingent truth, much like a law of nature. Since this principle is counterfactual supporting, it is true that if Dainton had been engaged in headache-directed introspection at t, his headache would have had a different phenomenal character. And since the principle is contingent, there are possible worlds in which introspection does not affect the character of its objects. In some of these worlds, Dainton can and does engage in headache-directed introspection at t without thereby intensifying his headache sensations, hence without preventing his actual headache, e, from occurring. In some such world, e occurs and is introspected.
Since Dainton has given us no reason to believe that it is a necessary truth that introspection affects the phenomenal character of its objects, he has given us no reason to reject H1; and consequently he has given us no reason to accept (2) or to reject the I-thesis.
Even if he had established the necessity of the relevant principle, however, he still would have failed to make his case. This can be seen by considering H2.
H2 It is a necessary truth that introspection affects the phenomenal character of its objects. In the actual world, w, Dainton had a highly stressful morning, and his psychological tension built throughout the afternoon. By 6:00 p.m. (time t), Dainton had a headache, e. On a scale of 1 to 10, 10 being the most intense, e was a 'level 3' headache; and, at 6:00 p.m., Dainton was not introspectively aware of e.
However, in nearby possible world w*, on the very same day, things turned out somewhat differently. Dainton's day was less stressful and his headache sensations were less severe throughout the afternoon. At 5:58 p.m., Dainton had a 'level 2' headache, of which he was not introspectively aware. At 5:59 p.m., Dainton happened to turn his introspective attention toward his headache sensations, with the result that they intensified somewhat. At 6:00 p.m., Dainton had a 'level 3' headache, e, of which he was introspectively aware.
The main idea behind H2 is straightforward. As things actually happened, e was a non-introspected, 'level 3' headache sensation preceded by a long, monotonous series of the same. If we confine our attention to those possible worlds in which these preceding conditions are held constant, we may thereby rule out the possibility of e's being introspected. How so? Any possible world in which these preceding conditions obtain and in which Dainton begins to introspect just in time for e will be a world in which the monotony of the series is disrupted, the headache sensations intensify, and e (which is essentially a 'level 3' headache) is thereby kept from occurring. (Or so it might be argued by one who accepted Dainton's premises.)
But if we vary the preceding conditions in the manner indicated by H2, we can make it turn out both that e occurs and that introspective awareness is directed its way: e could be the result of Dainton's introspecting, and thereby intensifying, a series of 'level 2' headache sensations.
Each of the two foregoing hypotheses entails (1) while entailing the negation of (2). Since Dainton has done nothing to cast doubt on either H1 or H2, I conclude that the headache argument gives us no reason to think that joint introspectibility is unnecessary for co-consciousness or that the I-thesis is false.
The following passage presents a second argument against the I-thesis:
Suppose you do introspect some part of your current experience. . . . This introspected experience remains co-conscious with the remainder of your experience. . . . Given this, what is responsible for the unity of the introspected experience with the non-introspected experience? One thing seems certain: it cannot be any form of introspection. When you focus your attention onto the shrub [e.g.,] your . . . auditory experiences [etc.] all remain co-conscious with your visual experiences. These experiential relationships cannot be explained in terms of introspectibility, for they are not even potential objects of introspection. If you were to try to . . . introspect these relationships you would have to stop introspecting your experience of the shrub. As this example makes plain, the co-consciousness of experiences which are not being . . . introspected with experiences which are is not something which can be . . . introspected. . . . This provides another reason for rejecting the . . . I-thesis: there is at least one form of co-consciousness that is in principle non-introspectible (2000: 36-37, Dainton's emphasis).
Let us suppose for a moment that the relevant "experiential relationships" are not even "potential objects of introspection." Does it follow that the I-thesis is false? Not without some additional premises. The I-thesis claims only this: that when two experiences are co-conscious, this is in virtue of the fact that they are jointly introspectible, that they could be among the objects of a single state of introspective awareness. It does not claim that when two experiences are co-conscious, this is in virtue of the fact that their co-consciousness is introspectible; indeed, the I-thesis is perfectly consistent with Dainton's claim that certain instances of co-consciousness cannot themselves be introspected.
To convert the passage quoted above into a logically valid argument against the I-thesis, we will have to fill in some tacit assumptions. I suggest that the following reconstruction strikes the best balance between fidelity to Dainton's words and plausibility of premises:
(1) cav is an instance of co-consciousness that holds between ea, which is a non-introspected auditory experience, and ev, which is an introspected visual experience of a shrub.
(2) Necessarily: for any experiences e and e* and any instance c of co-consciousness, if c holds between e and e*, then: (i) c is an experience in is own right, and (ii) c is an experience that is co-conscious both with e and with e*.<5>
(3) Therefore, by (1) and (2), cav is an experience that is co-conscious both with ea and with ev.
(4) For any instance c of co-consciousness, if c holds between two experiences only one of which is introspected, then, necessarily, c is non-introspected.
(5) Therefore, by (1) and (4), necessarily, cav is non-introspected; and a fortiori it is necessary that cav is not jointly introspected together with ea (or ev). In other words, cav and ea are not jointly introspectible.
(6) Therefore, by (1), (3), and (5), cav and ea are co-conscious experiences despite failing to be jointly introspectible; and consequently the I-thesis is false.
For the sake of argument, I will grant (1) and (2). In particular, I will concede the following questionable moves: the decision to posit such entities as instances of co-consciousness, the decision to treat these entities as experiences (or mental states, or phenomenal items) in their own right, and the assumption that these entities are always co-conscious with the experiences they relate.<6>
The only remaining premise is (4), which states that if c is an instance of co-consciousness that holds between an introspected and a non-introspected experience, then c itself is "in principle non-introspectible". What does Dainton have to say in its defense? "The co-consciousness of experiences which are not being introspected [such as your auditory experience] with those which are [such as your visual experience of a shrub] is not something which can be introspected." This is made "plain", Dainton thinks, by an example: "if you were to try to introspect these relationships, you would have to stop introspecting your experience of the shrub."
But why couldn't you simultaneously introspect both (i) the relevant instance of co-consciousness and (ii) your experience of the shrub? Is Dainton assuming that it's impossible to be introspectively aware of more that one experience at a time? If so, I deny his assumption. But in any case he has failed to provide us with a good reason for accepting (4).
Here is an alternative attempt at defending (4). Consider the following principle, which I find plausible:
(A) It is impossible to introspect a given instance of co-consciousness without simultaneously introspecting the two experiences related by the given instance; i.e., necessarily, for any experiences e and e*, any instance c of co-consciousness, and any mental state S, if c holds between e and e*, then, if S introspectively represents c, S also introspectively represents e and e*.
If (A) is true, and if you were to try to introspect the instance of co-consciousness that holds between your introspected visual experience and your non-introspected auditory experience, then, far from having to stop introspecting your visual experience (as Dainton claims), you would actually have to start introspecting your previously non-introspected auditory experience. Therefore, if (A) is true, it is impossible to introspect an instance of co-consciousness that holds between an introspected experience and a non-introspected experience.
The problem with this line of reasoning is that (A), though plausible, does not entail (4); rather, it entails
(4*) Necessarily, for any instance c of co-consciousness, if c holds between two experiences only one of which is introspected, then c is non-introspected.<7>
Contrast this with (4):
(4) For any instance c of co-consciousness, if c holds between two experiences only one of which is introspected, then, necessarily, c is non-introspected.<8>
Let H be the property being an instance of co-consciousness that holds between two experiences only one of which is introspected; and let N be the property being introspected. Now it is plausible that for any possible world w, anything that has H in w fails to have N in w; i.e., (4*) is plausible. But it is not plausible that anything that has H in the actual world has H in every possible world; on the contrary it seems that there are some entities that have H in the actual world despite existing in other possible worlds where they do not have H. Consider, for example, our instance cav: in the actual world it holds between an introspected and a non-introspected experience (hence, by (4*), we must conclude that cav is not introspected in the actual world); but it is plausible to suppose that in other possible worlds cav holds between two introspected experiences. Moreover, it seems plausible to suppose that in some of these other worlds, cav itself is introspected. Therefore (4) is apparently false.
Even with all of our concessions in place, then, the shrub argument does nothing to suggest that joint introspectibility is not necessary for co-consciousness; hence this argument remains powerless to harm the I-thesis.
I would now like to suggest a very different line of attack against the I-thesis. The foregoing arguments both purport to show that some pairs of mental states succeed in being co-conscious despite failing to be jointly introspectible, hence that joint introspectibility is not necessary for co-consciousness. But if my criticisms are correct, each of these arguments fails. A better way to challenge the I-thesis, I think, is to find a possible counterexample to the sufficiency claim, i.e., to find a possible case involving a pair of mental states that fail to be co-conscious despite succeeding in being jointly introspectible.
Here is one such case. Suppose that I have had a futuristic device installed in my brain that allows me to connect or disconnect my cerebral hemispheres at will (as described by Parfit (1984: 246)). When the hemispheres are connected, each mental state supported by any part of my brain is co-conscious with every other such state. But when the hemispheres are disconnected, each of them supports its own set of mental states, i.e., (i) each mental state supported by the right hemisphere is co-conscious with every other such state, (ii) each mental state supported by the left hemisphere is co-conscious with every other such state, and (iii) no mental state supported by the left is co-conscious with any mental state supported by the right. We might also add that the following sort of situation cannot occur: experiences e and e* are jointly introspected despite being supported by distinct, disconnected hemispheres.
Now consider a possible world w at which the following is true. During T, a brief period of disconnectedness, my right hemisphere supports (inter alia) a visual experience, ev, and my left hemisphere supports an auditory experience, ea. Then, since ev and ea are supported by distinct, disconnected hemispheres, they fail to be co-conscious, and they also fail to be jointly introspected.
But it is plausible to suppose that ea and ev* could have been jointly introspected, since their respective hemispheres could have been connected throughout T, the period when ev and ea occur. That is to say, it seems likely that there is a possible world w* at which my hemispheres are connected throughout T and at which there is a single higher-order state S that introspectively represents both ev and ea. (And if it is not possible that all of this be true of me, then surely it is possible that it all be true of some sentient being. Let us waive this complication.)
There is no good reason to deny that these very experiences occur at some possible world w* where my hemispheres are connected throughout T. This is true even if each (token) mental state is so 'modally fragile' that it has its spatiotemporal location, its physical basis, its phenomenal character, and its owner essentially, i.e., in every possible world in which it occurs. After all, in light of the set-up of the case, it is clear that all these features can remain constant from possible world w (where my hemispheres are disconnected throughout T), to a possible world w* where my hemispheres are connected throughout T.<9> ,<10> Nor is there any good reason to deny that at some such world where my hemispheres are connected throughout T, the relevant experiences are jointly introspected.
Thus it seems that we have our possible counterexample to the sufficiency clause of the thesis that joint introspectibility is necessary and sufficient for co-consciousness: ev and ea are jointly introspectible in w despite failing to be co-conscious there.
It should be pointed out that this 'split-brain argument' does not count against the ICC-thesis, which says that mental states S and S* are co-conscious at time t in possible world w if and only if they occur at t in w and their owner has, at t in w, the cognitive capacity to introspect S and S* jointly. In the world w described in our case above, my hemispheres are disconnected throughout T. Hence it is plausible to say that throughout T in w I lack the cognitive capacity to introspect ev and ea jointly.
Our results can be summarized informally as follows. My actually having the cognitive capacity to jointly introspect my experiences e and e* may be a sufficient condition for these experiences to be co-conscious; but, as the raccoon argument apparently shows, this is not a necessary condition. And my being such that it is possible for e and e* to be jointly introspected may be a necessary condition for these experiences to be co-conscious (given the failure of the headache and shrub arguments); but, as the split-brain argument apparently shows, this is not a sufficient condition.<11>
<1>. See, e.g., Dainton (2000: 34) and Hill (1991: 231-2).
<2>. Dainton has suggested in correspondence that while raccoons probably lack the concepts to form introspective thought-like representations (HOTs) of their experiences, they may have the capacity to form (non-conceptual) perception-like representations (HOPs) of these experiences. (On the distinction between HOTs and HOPs, see Block et. al. (1997: sect. X).) Perhaps this is so. But surely there at least could be creatures that have co-conscious experiences despite lacking the cognitive capacity to form any sort of higher-order representations. Indeed I would guess that there actually are such creatures: frogs might be a good example.
<3>. Actually Dainton distinguishes two kinds of introspection - active and passive - and restricts his claim that introspection affects the character of its objects to the case of active introspection. I ignore passive introspection in what follows for two reasons: (i) taking it into account would complicate the discussion, and (ii) we should be suspicious of it. Insofar as one is willing to concede the character-affecting powers of some form of introspection, one should be unwilling to depend upon the inertness of any form of introspection. In particular, it seems reasonable to suppose that if some forms of introspection are more active and character-affecting than others, this is because there is a whole continuum of such forms, ranging from the most active and character-affecting to the least active and least character-affecting. And if any form of introspection is character-affecting, then it seems likely that all such forms are at least somewhat character-affecting.
<4>. See, e.g., the entry on 'Counterfactuals' in Audi (1995: 163).
<5>. The following two claims, both of which Dainton seems to accept, might provide some motivation for (2): (a) necessarily, for any experiences e and e*, e and e* are co-conscious iff e and e* have a fusion that is an experience in its own right; (b) necessarily, for any experiences e and e*, if e is a part of e*, then e and e* are co-conscious. See Dainton (2000: ch. 9) and Bayne (2001).
<6>. Do these assumptions lead to an infinite regress of experiences, and if so is the regress vicious? On these questions see Bayne (2001).
<7>. Letting 'NEC' be the necessity operator and 'A' be the universal quantifier, we can express (4*) symbolically as follows: NEC[Ax(Hx --> not-Nx)].
<8>. Symbolically: Ax[Hx --> NEC(not-Nx)].
<9>. I assume here that the relevant experiences can fail to be co-conscious in w without having different owners there.
<10>. In correspondence, Dainton has noted that certain forms of interexperiential holism would prevent ea and ev from occurring both in worlds where they are co-conscious and in worlds where they are not. In particular, if we say that each experience has all of its phenomenal properties essentially, and if we count such relational, haecceitistic properties as being co-conscious with the particular experience ev as phenomenal, then of course ea either has this property essentially or lacks it essentially. In my view, the strongest form of holism that is not obviously absurd is what Dainton calls 'Weak C-holism' (2000: 223), which I state in my own words as follows: each experience has its (intrinsic) phenomenal character essentially, and each experience has its context type (but not its context token) essentially. (The context token of an experience e can be defined as the set of token experiences co-conscious with e; we can say that two context tokens C and C* are of the same context type iff the members of the sets C and C* can be put into a one-to-one correspondence in such a way that each member e of C is paired with a member of C* that has the same intrinsic phenomenal character as e.) Although I find Weak C-holism implausible, it may be worthwhile to show that it does not stand in the way of the split-brain argument. To see this, just suppose that the set of experiences supported by my left hemisphere throughout T in w is of the same context type as the set of experiences supported by my right hemisphere throughout T in w, and that each of these sets is of the same context type as the set of experiences supported by my entire (connected) brain throughout T in some world w*.
More specifically. Let ea be a particular auditory experience, and let ev be a particular visual experience. Let ea* be a particular auditory experience that has the same intrinsic phenomenal character as ea but that is numerically distinct from ea. Let ev* be a particular visual experience that is numerically distinct from ev but that has the same intrinsic phenomenal character as ev. Let “I(e)” mean: e is introspected. Assume any two co-conscious experiences e and e* have a unique fusion that is an experience, and let “e+e*” refer to the fusion of e and e*. Then world w can be represented as follows:
My left stream: I(ea*+ev) My right stream: I(ea+ev*)
And here is world w*:
My single unified stream: I(ea+ev)
Here’s the basic idea: In world w, where my hemispheres are split, ea is co-conscious with and jointly introspected with ev*, an intrinsic duplicate of ev; and ev is co-conscious with and jointly introspected with ea*, an intrinsic duplicate of ea. So my right and left streams are qualitatively exactly alike in w. Moreover, each of these two streams is qualitatively exactly similar to the single stream that I have in w*, where ea and ev are co-conscious and jointly introspected.
Assuming the possibility of this set-up, it seems to follow that one can accept both the split-brain argument and Weak C-holism.
<11>. I would like to thank Barry Dainton for his helpful comments. A version of this paper was written with support from a Princeton University Graduate Fellowship, which I gratefully acknowledge.
Audi, R., ed. (1995). The Cambridge Dictionary of Philosophy. Cambridge: Cambridge University Press.
Bayne, T. (2001). Review of Barry Dainton, Stream of Consciousness. Journal of Consciousness Studies, 8, 79-92.
Block, N., Flanagan, O., and Guezeldere, G., eds. (1997). The Nature of Consciousness: Philosophical and Scientific Debates. Cambridge, MA: MIT Press.
Dainton, B. (2000). Stream of Consciousness: Unity and Continuity in Conscious Experience. London: Routledge.
Hill, C. (1991). Sensations: A Defense of Type Materialism. Cambridge University Press: Cambridge.
Parfit, D. (1984). Reasons and Persons. Oxford: Oxford University Press.
