PSYCHE 11 (5), June 2005
Review of Michael Tye's “Consciousness and Persons”
Bernard W. Kobes
Department of Philosophy
Arizona State University
Box 874102
Tempe, Arizona 85287-4102
kobes@asu.edu
|
|
|
REVIEW OF: Tye, Michael. 2003. Consciousness and Persons: Unity and Identity. Cambridge, MA: MIT Press. 203 pp. ISBN: 0-262-20147-X
(c) Bernard Kobes 2005
Download review as a pdf
Consciousness has been defined as that annoying period between naps, and this grumpy definition may not be wholly facetious, if Michael Tye's latest book is right. Tye's main goal here is to develop a theory of the phenomenal unity of experience at a time, and its diachronic analog, the moment-to-moment continuity of one's experiential stream from the time one wakes up to the time consciousness lapses.
This attractively concise book is itself like a rapid stream. Divided into numbered sections of usually only one or two paragraphs each, it offers crisp arguments on a wide range of topics; these include the metaphysics of material things, the individuation of experience, the unity of the body image and bodily sensations, the unity of occurrent thoughts and moods, the specious present, the nature of the disunity in split-brain subjects, the ontology of persons, and vagueness in personal identity. Despite its brisk pace the flow is well-controlled; seldom is the exposition cloudy or turbulent.
In what follows, after surveying Tye's main positions, I highlight some strange consequences of his “one experience” thesis. I also sketch a way in which much of the spirit of Tye's explanatory project could survive even in the absence of this peculiar doctrine. I conclude with a special problem for Tye's treatment of the diachronic case.
The phenomenal unity of consciousness, the intuitive “togetherness” of experienced qualities at a time, is according to Tye a real phenomenon. However, philosophical accounts of this unity typically assume that the explanatory goal is to characterize some sort of phenomenal or experiential glue that binds smaller experiences into larger, over-arching experiences. This is a mistake. In fact we can have no phenomenal awareness of our own experience at all; as Tye and fellow “representationalists” such as Fred Dretske stress, we can be aware only of qualities that our experience represents to be instantiated in our bodies and in the world.
One's experience of the world at a time is not composed of other, smaller experiences, any more than a painting is composed of other, smaller paintings. Each of us normally has only one phenomenal experience at a time, and this is a maximal state of PANIC-poised, abstract, non-conceptual, intentional content. The unity of one's experience is wholly at the level of qualities represented in experience. These qualities include not only the colors, shapes, sounds, textures, smells, etc. of external things, but also body image, bodily sensations, auditory images accompanying thought, and moods, all of which represent states of, or events in, one's body, or desires with respect to such states and events. The state is maximal because no experience can have a proper part that is an experience.
One's experience at time t represents all of these qualities as “together”. A closure principle obtains for experiences under conjunction: given any experience e at time t, if e represents content q1 and also represents content q2, then e represents the conjunctive content (q1 & q2). (Cf. p. 37; here and in the diachronic case below, I have rephrased the closure principle for clarity.) Phenomenally unified qualities are therefore closed under conjunction.
There is similarly a uniquely correct way to individuate experiences over time. One's experience of the world over time is not composed of other, smaller experiences over time, any more than a movie is composed of other, smaller movies. The experienced present is a small temporal window, a specious present, a duration of some milliseconds that, for the experiencer, is now. In virtue of the specious present one experiences an extended event, such as a sequence of sounds or words, “all in one as a whole” (87). Thus one experiences continuity, change, and succession of qualities in the world.
Direct phenomenal unity through time is a relation between a quality q2 experienced in one specious present, and a quality q1 experienced in the immediately prior specious present, if and only if q2 is experienced as succeeding or continuing on from the prior q1. Indirect phenomenal unity through time is likewise a relation between an experienced quality q2 and an experienced quality q1, if and only if a chain can be constructed between them, where each link is between items that are directly unified. The relation of indirect phenomenal unity is therefore the ancestral of direct phenomenal unity (100).
An experience normally begins when one wakes and normally ends when one falls into a dreamless sleep; there is only one experience in any such interval, and it is not composed of other experiences. Thus the problem of what binds experiences together across time dissolves: there are no experiences (plural) that are so bound.
A principle of closure under conjunction holds likewise for diachronic unity. Given any temporally extended experience e, if e represents phenomenal content q1 as instantiated at t1 and also represents phenomenal content q2 as instantiated at t2, then e represents the conjunctive phenomenal content (q1-at-t1 & q2-at-t2) (cf. p. 101.) Experienced qualities that are (directly or indirectly) unified across time are therefore closed under conjunction.
A split-brain subject, in whom the corpus callosum has been severed, is still a single person and normally has phenomenally unified consciousness. However, under experimental conditions her consciousness can be briefly split, like a stream going around two sides of a large rock. Tye argues that in such a case transitivity of phenomenal consciousness may fail: the sensation of a pin prick in the neck region may be phenomenally unified with each of two visual images presented to the left and right hemispheres under experimental conditions, even though those experienced images are not phenomenally unified with each other.
Tye claims on behalf of his account: “Nothing that we ordinarily say about experience needs to be given up. No large bullets need to be swallowed.” (107). This is quite false, I am convinced. In fact a great deal of what we would ordinarily say about experience must be given up, on Tye's “one experience” account of phenomenal unity. If I hear music while watching the orchestra, then although the experienced sounds could have existed without the experienced sights, my actual auditory experience could not have existed without my actual visual experience, for they turn out to be strictly identical (33). Again, suppose the experienced sights actually go on longer than the experienced sounds; still, my actual visual experience does not go on longer than my actual auditory experience, for they are strictly identical (34). Suppose the music I experience lasts for two hours; strictly, on Tye's view, my experience of the music began many hours earlier, when I woke up, and lasted until many hours later, when I fell sleep.
If a late-entering concert-goer steps on my toe while taking his seat, then my token experience of pain is strictly identical to my token experience of pleasure at the music. In fact, my token experience of pain is identical to my token experience of pleasure even if there are many (sleepless) intervening hours. “I had many strange experiences today”, said by one who took no naps during the day, turns out to be strictly false. (The paraphrase “I had an experience of many strange things today”-cf. p. 97-is decidedly strained if, as we may suppose, there is no natural similarity or unity among the strange things experienced.) In the opposite direction, “I have had the experience of being woken up by a crying baby six times in a single night” also turns out to be strictly false, since on Tye's account no experience could span both sides of a dreamless sleep.
I say that, like conflicts, flirtations, lessons, celebrations, representations, and events in general (and unlike paintings and movies-artifacts with conventional boundaries), experiences as ordinarily individuated permit hierarchical nesting. A conflict between two nations that lasts ten years may have a week-long conflict between two armies as a proper part, which may in turn have smaller conflicts as proper parts, and so on. Similarly, my experience backpacking for several days may include my experience of hiking Oak Creek Canyon as a proper part, which may include my experience of encountering a bear, and so on. But on Tye's way of individuating experiences all experiences are maximal, so no nesting of experiences is possible; no experience can ever be a proper part of an experience. This amounts to a puritanical regimentation of our ordinary free and easy ways of counting experiences.
The strangeness of Tye's way of individuating experiences would be acceptable if it were a necessary part of the best account of phenomenal unity. But in fact Tye's own view suggests the elements of an account of phenomenal unity that does not depend on the “one experience” doctrine. Consistently with treating experiences as hierarchically nested, let e be a maximal experience at time t if and only if e is not a proper part of any experience, and the qualities that e represents at t are experienced as together. Togetherness may be explicated by its entailment of closure under conjunction of experienced qualities: if e represents quality q1 and also represents quality q2, then e represents the quality (q1 & q2). One might say, with Tye, that experienced qualities q1 and q2 are phenomenally unified at t if and only if (and because) there is at least one maximal experience e at t in which q1 and q2 are both represented. One might then add, against Tye, that non-maximal experiences e1 of q1 and e2 of q2 are both parts of experience e.
This is not to treat a maximal experience e as any sort of phenomenal glue joining e1 and e2. Maximal experiences have themselves as non-proper parts; no vicious regress looms. The explanatory work-such as it is on this deflationary account-is being done by the experienced togetherness of qualities q1 and q2 in a maximal experience, just as on Tye's account. The account is compatible with Tye's representationalism: phenomenal unity is still fundamentally a relation between experienced qualities, rather than a relation between experiences. My point is not at all to defend this account, but rather to note that much of the spirit of Tye's (first-order, deflationary, representationalist) project need not be hostage to the excessively regimented “one experience” doctrine.
Turning briefly to the diachronic case, I suggest that the following principle is plausible: experience e represents phenomenal content q only if q is available for cognition in virtue of e. This principle seems required by the poisedness of phenomenal content, the “P” of Tye's PANIC account of such content (cf. p. 175). Yet Tye's account of diachronic unity seems incompatible with this principle. Suppose a person S suffers sudden massive memory loss, but does not lose consciousness. Since S's conscious experience is continuous across the amnestic event, Tye would count S as having only one relevant temporally extended experience. Qualities experienced before and after the amnestic event are therefore indirectly phenomenally unified, and the relevant experienced content must be closed under conjunction. If the experience e represents phenomenal content q1-at-t1 prior to the amnestic event, and content q2-at-t2 after the event, then Tye's account entails that e represents the content (q1-at-t1 & q2-at-t2). But plainly the content (q1-at-t1 & q2-at-t2) is not at any time available for cognition in virtue of e, since given the amnesia, by the time q2 is experienced S has no memory at all of q1. So Tye's account of diachronic phenomenal unity is incompatible with a plausible principle of availability of phenomenal content, a poisedness principle apparently required by his own PANIC theory. [[A similar argument is given independently in Tim Bayne's paper, "Divided Brains & Unified Phenomenology: An Essay on Michael Tye's 'Consciousness and Persons'", forthcoming in Philosophical Psychology. ]]
Tye's account of phenomenal unity across time makes it a diachronic analog of the synchronic notion of phenomenal togetherness, which Tye explains as the relation of being represented in the same experience, where the individuation of experiences is highly regimented. In this apparently symmetric relation, memory plays no essential role. Since indirect unity is the ancestral of direct unity, it must also be transitive. But if indirect phenomenal unity is both symmetric and transitive, this creates a problem for Tye's treatment of split brain cases. There Tye argues, cogently I think, that under experimental conditions phenomenal consciousness in a single subject is briefly phenomenally disunified. A split-brain subject under experimental conditions may visually experience the word pen in one hemisphere, and knife in another, without experiencing them together. The problem is that if indirect phenomenal unity is both symmetric and transitive, then the experienced word pen and the experienced word knife will be diachronically unified, for there will be a chain of (appropriately related) specious presents stretching into the past, then back again into the future along the other branch, linking pen and knife. The two experienced words turn out to be synchronically disunified but diachronically unified-surely an unwelcome result. A natural response to this difficulty would be to make the relation of diachronic unity asymmetric. But this response is ad hoc in the absence of any role for memory in the account of unity across time, and it sits ill with Tye's fundamental idea of phenomenal togetherness as the relation of being represented in the same experience, where the individuation of experiences is highly regimented.
To sum up: the one-experience account of phenomenal unity is more revisionary with respect to common sense than Tye is prepared to acknowledge, and it is doubtful that the revision is justified by any explanatory gain. This is especially so in the diachronic case, where Tye's theory suffers from ignoring the role of memory in constituting phenomenal unity across time. For these and other reasons I am not able to accept the account. Even so, I admire the ingenuity of this rigorously first-order and severely deflationary account of phenomenal unity. [[Many thanks to Brian Fiala for helpful discussions on these topics]].
