You enter the lab of a mad scientist and see two operating tables. On one, a second individual—“Pat” for future reference—lies in an anesthetic slumber. The other table is for you. The good(?) doctor explains the procedure for which you’ve volunteered(?) as follows:
You will be put under and, while your consciousness is offline, the mad scientist will scan the brains of you and Pat in meticulous detail. Everything will be recorded: not just the gross pattern of neural connections but the number, types, and locations of receptors at every synapse, the amount and type of neurotransmitter in every vesicle at every presynaptic terminal—every bit of physical info that encodes the “weights” of the connections between your neurons. In doing so, every detail of memory and personality will be captured.
After these hyper-detailed maps are made, some very sophisticated robots will set about re-sculpting your brain in the likeness of Pat’s and Pat’s brain in the likeness of yours. Now, there will be no transplantation of tissue here—just a complex restructuring of what’s already there. Any new cells needed will be cultured from your own. Further, let’s stipulate that while you and Pat are not related and have never met each other before, you fortuitously share all the alleles involved in brain development and function, and their states of activation will also be replicated in the brain of the other.
Now, the kicker: The scientist informs you that only one of the two bodies can be revived after the procedure, and you must choose at the outset which one.
Take a minute to note not just your decision but as much of the reasoning behind it as you can make explicit to yourself.
My inclination here is to opt for reviving the body—the original body—of Pat. If you fall into this camp as well, then I’d like you to think about what specific considerations influenced your decision. What were you trying to preserve? Your memories? Your personality? I suspect that’s not quite all. If you’re like me, you probably have an expectation that this identity of brain structure between pre-op you and post-op Pat sustains a thread of phenomenal (i.e., conscious experiential) continuity and that this thread is crucial to one’s identity and sense of self. I expect to “wake up” from this procedure in the body of Pat just as I had woken up in my pre-op body from all prior states of unconsciousness.
It is, I think, this presumed continuity of conscious experience that I am primarily seeking to preserve when I choose to have the body of Pat revived. Choosing otherwise might keep my body going, but with someone else’s brain structure in it, it would be as though my self—my person—had died and been replaced.
These intuitions may not be your own. If, for example, you believe the seat of conscious experience is an immaterial soul that has a special spatiotemporal relationship with a particular body, then you may well come to very different conclusions, and the rest of this essay may be of little more than anthropological interest to you. If, however, you found yourself at least a little reluctant to entrust your original body to the mind of a stranger, then strap in; the mad scientist isn’t through with us yet.
The Second “Hard Problem” of Consciousness
A number of philosophers of mind believe the existence of conscious experience to be a mystery of the profoundest order. The bulk of this belief rests on an assumption that consciousness is explanatorily otiose. All speech, behavior, and cognition, they contend, could be wholly explained by physical facts that make no gesture whatsoever toward first-person experience. Put another way, it is compatible with everything we know about the physical world that everyone is a “philosophical zombie”—a sophisticated meat robot that senses, perceives, speaks, behaves, believes, and desires but has no interior experience accompanying any of these activities. At the same time, nothing can be more certain to us than the existence of our own interior experience. How do we reconcile these apparent truths? How can natural science explain the existence of consciousness if it makes no possible difference to the natural world?
This is the alleged “Hard Problem” of consciousness. Personally, I am unpersuaded of it, unpersuaded of the very coherence of the idea of having ontological knowledge of something that, ex hypothesi, can have no physical effects. Nevertheless, this “epiphenomenalist” view still exerts a strong pull on the intuitions of many philosophers, and the question “Why consciousness?” remains a central preoccupation within the Philosophy of Mind.
As a child and teenager, I found myself regularly preoccupied with a related but rather more parochial question: Why this consciousness? Why did it seem as if the universe were choosing to experience itself through this particular brain out of all the billions of other particular brains seemingly adequate for the task? There ought to be a veritable symphony of first-person perspectives on the world, so why this apparent lonely solo? Why did it seem as though there were a phenomenal spotlight shining only here?
These are, of course, silly questions. The seeming is key to their answers: Of course things would seem this way to any particular conscious self, regardless of whether it was alone in the world or not. If the facts of my siloed experience are entailed by any model of the world positing multiple independent conscious selves, then those facts cannot evidentially favor a sole self model against it. This Second Hard Problem of consciousness reveals itself on analysis to be, much like the original Hard Problem, a mere pseudo-problem.
Nevertheless, I found it surprisingly difficult to keep my intuitions consistently aligned with this realization. Particularly sticky for me was this metaphor of the lone spotlight. In times when I had given my thinking over to this image, I would often wonder about the possible conditions under which the spotlight might be shifted from me and the factors that might influence where it goes. How would the next subject be chosen? If my corpus callosum were surgically severed, as in split-brain patients, such that my skull came to contain two distinct conscious selves, which one would get the spotlight? Which one would be phenomenally continuous with my pre-op conscious experience? Could they both be, and if so, what would that be like? If the mad scientist we met earlier were to copy my brain structure into two other bodies and revive them both, in which one would I wake?
For all their intuitive gravity, these too are bad questions. Their construction betrays a deep confusion about the significance of phenomenal continuity. In what follows, I want to suggest that a similar confusion plagues our default thinking about practical rationality (the normative system concerned with reasons for action rather than belief). In unpacking and poking at this confusion, my goal is not only to tease out some surprising new facts about the proper ends of practical rationality but to make certain facts already at least partially accepted vivid in novel ways. In doing so, I hope to loosen the grip of a certain stubborn, often clandestine egoism on our moral motivation.
Practical Rationality Reconsidered
Many moral theories are “rationalistic” in that they assert that for an individual to be subject to some moral obligation, that individual must have a reason to do as the obligation demands. This doesn’t mean that the individual must want to fulfill the obligation, but only that it would be rational, in some sense specified by the moral theory in question, for him to do so. A moral rationalist might say, for example, that John ought to give money to the poor even if he doesn’t want to because doing so would fulfill a value he holds dear, and this value provides him with a more compelling reason to act than any contrary desires he may have at the moment.
Rationalistic moral theories have appeal because they tie moral, universal oughts to the individualized oughts of practical rationality, for which naturalistic analyses are considered to be much easier. But these accounts have vulnerabilities as well. The central enduring challenge to this broad genus of moral theories is that posed by what David Hume called the “sensible knave.” This individual is a particularly clever rational egoist who knows how to benefit himself at the expense of others without incurring punishment, damage to his reputation, or feelings of guilt. If such an individual really could get away with furthering his own exclusive interests by behaving in putatively immoral ways, would it not be rational for him to do so? If so, then either such behavior must be moral after all or moral rationalism must be false.
Most responses to the challenge posed by the sensible knave invoke arguments to the effect that the knave’s actions do, in some roundabout way, harm him or make future harm to him more probable, such that it really is in his best long-term interest to behave morally, even when it seems he can get away with not doing so. I don’t aim to weigh in on these responses here; rather, I want to have a go at a distinct and much more fundamental line of criticism. Articulating this effectively is going to require a careful analysis of just whom the knave’s apparent rationality is supposed to benefit.
To begin with, we must note that an action taken by a present self never actually affects the present self but only some future self, which may differ from the present self in numerous ways, depending on how much time and experience has intervened between the onset of the action and the registered consequence. It’s largely in virtue of memory—specifically episodic memory—that we tend to take these discriminable conscious selves to be part of the same singular entity. Episodic memory allows a present self to call up experiences of past selves—to, in a way, simulate past selves, or at least relevant fragments of them. This ability is what sustains the “thread of phenomenal continuity” mentioned earlier (even, perhaps, across bodies, as in our initial brain swap thought experiment).
This is a very neat trick, and crucial to survival, but the accomplishment is easily overstated, ontologically speaking. Our memory is not like that of a digital computer. It is not stored in discrete locations but distributed across the brain in patterns of neural connections and distributions of connection strengths. When you have an experience, a particular pattern of neural activity is tokened. Successfully recalling this experience requires re-tokening this pattern of activity—or at least a pattern sufficiently similar to it. But each experience alters our connectivity patterns and synaptic weight distributions in countless little ways. This is simply how we learn, how we incorporate new information into our cognitive ecosystem. And these small alterations, augmenting over time, can interfere with our ability to re-token earlier experiences.
To see this, try to recall 10 experiences from the previous day. Not 10 facts about what you were doing or thinking or saying—rather, try to re-experience in as much detail as you can 10 discrete moments from that day. 10 little phenomenal vignettes. Assuming eight hours of sleep, you logged around 57,600 seconds of waking experience that day. How many of those seconds remain accessible to your present self?
How many seconds from the day before yesterday?
From the same day last week?
From the same date last month?
The phenomenal thread may be a lot more frayed and piecemeal than we tend to assume. This is likely why hardly any of us can remember experiences from before the age of 3; our brains change so dramatically during childhood that the succeeding structures are simply incapable of hosting activation patterns even remotely similar to those tokened by our earliest experiences. Personality traits too are instantiated in mutable neural structures and subject, therefore, to the same gradual change. Imagine the ship of Theseus, but where each part is replaced by a non-identical part such that the ship eventually looks nothing like its former self.
So, our reasons for treating distinguishable present and future selves as one and the same self appear on closer scrutiny to be rather ad hoc. In general, anytime a claim is made that two or more discriminable objects are “essentially the same,” there is always a tacit “for the purpose of interest” clause hiding somewhere in the background. When any defender of a knavish or otherwise egoistic conception of practical rationality makes this sort of claim, he is obligated to make said purpose explicit. If the purpose turns out to merely be the defense of an egoistic conception of practical rationality, then he is arguing in a circle. If it is some other purpose, then he must demonstrate its importance and relevance. It will not do to have the proper beneficiaries of something as central to human life as practical rationality determined by nothing more than expedient convention.
This is all getting rather abstract. Let’s see if we can’t clarify our intuitions with another thought experiment:
Suppose that at some point in the near future, in order to keep pace with technological change, it becomes necessary to take drugs that return our brains to a state of child-like hyper-plasticity. As a consequence of this constant dramatic neurological change, however, we are completely unable to recall any experiences from more than five years into our pasts. Would this fact change anything about the rationality of acting in the interests of any possible future selves five or more years out? Would our present selves in such a world be rationally justified in taking up unhealthy habits, providing the consequences wouldn’t begin to manifest for at least five years?
When I imaginatively place myself into this situation, I find I’m no more willing to discount such phenomenally isolated future selves than to discount future selves still capable of recalling (parts of) my present self. It seems to me that a young child ought not to eat paint chips, even if doing so will only impact an adult in which no phenomenal trace of that child may remain. Similarly, it seems an anterograde amnesiac would be behaving no less irrationally than a neurotypical individual in willingly acting against the interests of her future selves. A present self’s obligations to future selves, then, don’t seem to depend on the capacity of the latter to remember it.
It’s interesting to compare our intuitions here with those concerning the mad scientist brain swap case. In that example, it was presumably rational to prefer the continuance of the present conscious self (even if in another body) to its replacement in the same body with a wholly different conscious self. But that replacement entailed a very sudden change in the neural substrate of conscious experience. What if we could effect the same replacement without any detectable disruption of moment-to-moment phenomenal continuity?
Consider, then, the following variation: Instead of reconfiguring your and Pat’s brains all at once, the scientist instead fills your skulls with nanomachines that will gradually, over the course of five years, reshape your brain into that of pre-op Pat and vice-versa. At the end of this process, of necessity, all memory of the intervening five years will have been lost. Again, you are asked to choose at the outset which body to have revived.
It is much less clear to me in this case what is the most instrumentally rational choice. If I have my body revived, I will eventually lose all who I am at the moment I make the decision, but I will preserve a thread of phenomenal continuity between my current self and the self that will replace me. If I have the other body revived, my present self will be extinguished and eventually reconstituted, but it will be phenomenally continuous with some other self that is wholly different and separate from my present self. The spotlight metaphor is hardly illuminating here.
What is brought out by this thought experiment and the considerations preceding it, I think, is a deep ambiguity in our default, fundamentally egoistic notion of practical rationality: Is its proper function supposed to be (1) preserving the mental content (memories, dispositions, etc.) of the self employing it or (2) preserving the phenomenal thread that joins one experienced moment to the next?
On close examination, I think I the answer must be neither.
Let’s look at option (1) first. Suppose you were offered two routes to scientifically plausible immortality: You can either be given some treatment that halts the aging process, or you can have your present self (in the form of your brain structure and global activation pattern) plucked out of time and preserved indefinitely in some suitably safe environment. The first method means you will continue to have experiences and be changed by those experiences. If you live long enough, eventually all trace of your present self may be lost, overwritten. The second method means that your present self (and all the memories it contains) will be spared any degradation or replacement but will never experience anything new. You will live on in a sort of phenomenal stasis, a single conscious moment of infinite duration.
I don’t know about you, but I find the former kind of immortality a lot more appealing than the latter. But maybe these questions of preservation are too abstract to really engage our intuitions properly. Much of the day-to-day workings of practical rationality are taken up with the seeking of positive experiences and the avoidance of negative ones. Permit me, then, one final thought experiment:
Suppose our mad scientist has made an age-matched clone of you. He’s installed in your head a brain scanner and an emitter that will send signals to nanomachines in the clone’s head that will continually reconfigure the clone’s brain to match yours. Now, the wrinkle is that the machines are programmed to operate with a one-day delay, so that your present self will only be replicated in the clone at this time tomorrow. Now, the mad scientist, being a particularly evil sort of mad scientist, informs you that in 24 hours either you or the clone is to be brutally tortured to death and you must choose who goes on the wheel. The scientist pledges that, in the event you choose to have your own body tortured, the clone will be freed from its nanomachines and thus spared the day-lagged experience of your torment.
Now, the clone’s conscious self at the moment the torture is to be implemented will be identical to your conscious self at the moment you make the decision, but I suspect most would judge that the most instrumentally rational choice would be to save your one-day-out future self at the clone’s expense. Surely the knave, at least, would do this.
By process of elimination, then, practical rationality must ultimately serve the phenomenal thread, yes? I don’t think that’s quite right either. If preservation of the thread were the core end of practical rationality, then it would have to be reckoned irrational to do anything that severs or substantially risks severing that thread, including going under general or twilight anesthesia or simply entering deep sleep. Deep sleep, after all, is where memory consolidation takes place, a process involving considerable pruning of episodic experience.
But phenomenal continuity still feels really important. I don’t think we’re wholly mistaken on this point, but I think we misunderstand the nature of that import.
I propose the thread’s principal import is not fundamentally normative but rather epistemic. This moment-to-moment phenomenal connection makes vivid for us, in a way we find very hard to ignore, the likelihood that someone will be subjectively present when the proverbial chickens loosed by our actions come home to roost. That the future we help create, good or bad, will be experientially registered. That those experiences, good or bad, will be every bit as real, intense, and inescapable in their goodness or badness as the experiences with which we’re already intimately acquainted.
Now, those of us who are not solipsists readily accept that other selves not connected to us by threads of phenomenal continuity have first-person experiences just as real and rich as any to which we’ve ever been subject. But this fact doesn’t tend to press itself upon us with the same immediacy as does the reality of our own consciousness—unless perhaps we’re engaged in detailed empathetic imagining. This discrepancy results in decision processes that by default favor possible future selves phenomenally connected to us (or to nearer future selves connected to us) over selves, future or present, that are phenomenally isolated from us—even, as we’ve seen, if those connected future selves are just as different from the present self as .
Granted, we can’t be absolutely certain of the consciousness of others in the way we’re certain of our own, but most of us accept the existence of multiple phenomenal spotlights with little explicit doubt. And with good reason. I don’t, therefore, think this discrepancy in certainty can fully account for the degree of default favoritism we show toward phenomenally connected future selves. If I’m right, then that default favoritism is, pace the knave, fundamentally irrational. The egoistic conception of practical rationality is grounded ultimately in the same error as the lone spotlight metaphor.
See, there are also plenty of uncertainties attending our actions toward those connected future selves. How long will my body be alive to permit such selves to absorb the good consequences I’ve tried to line up for them? Will those future selves really benefit in the ways I imagined them to benefit? Might they feel guilt over some of the actions I’ve taken to secure those presumed benefits for them? These are nontrivial considerations, and the uncertainties they engender may easily dwarf the uncertainties concerning our fellows’ capacity for consciousness. The immediacy of our own moment-to-moment experience, however, can obscure this fact. The unique source of the uncertainty concerning the consciousness of unconnected others leads us to effectively overestimate the magnitude of that uncertainty. If we have no good reason to doubt that phenomenally disconnected present or future selves may be genuinely harmed (e.g., subject to some aversive experience) by some action we are contemplating, then we have no good reason to exclude this harm from our decision calculus.
To be clear: None of this implies that we ought never to prioritize our future selves or those of our kith and kin. We are, after all, often uniquely positioned to understand the plights and needs of those closest to us and may consequently be better able to help them than to help a complete stranger. What I am arguing against, rather, is the longstanding presumption that such prioritization is inherently rational.
Now a knave could, of course, simply be an out-and-out solipsist, but then his knavery can only be as rational as his arguments for solipsism are cogent (and such arguments would, again, have to be based on more than just his first-person phenomenal facts, since these facts are perfectly consistent with a multiple spotlight world). He is not entitled to a default presumption of “sensibility.”
Another knavish response might be to accept all these uncertainties about benefiting future selves and conclude that the rational course is to simply always act in the interests of immediate future selves, for whom the uncertainty of benefit will be lowest. Anyone following such a strategy consistently, though, is apt to come to ruin pretty quickly, and I submit that it’s a minimal constraint on any viable notion of practical rationality that it not extinguish itself by driving its most dutiful adherents promptly to their deaths.
Out of all the above emerges a picture of practical rationality that has as its end not the promotion of valued experience for my future selves or your future selves, but the promotion of valued experience tout court. If this seems strange, realize that it’s no stranger than our picture of epistemic rationality, which has as its end not the realization of my truth or your truth, but the realization of truth tout court. Pace traditional economists and moral antirationalists, practical rationality is not a little daemon that lives in our heads and operates for its own exclusive benefit. Rather, like epistemic rationality, it is a public repository of principles, norms, heuristics, and other tested tools that we borrow and wield, with greater or lesser facility, in attempt to realize certain valued states of the world. Not valued just by me or just by you or just by the knave. Valued tout court.
Correcting the bias…by leaning into it.
This knowledge will not be easy to consistently put into praxis. The immediacy of our own conscious experience works constantly to bias us toward thinking and behaving like solipsists, regardless of our explicit beliefs. This is why the phenomenal spotlight metaphor remains so sticky. I think, though, that we can redeploy that metaphor in a way that corrects much of the practical irrationality it tends to ferment in us.
To do this, we’ll need to talk about death. Now, there are two main ways of thinking about death using the lone spotlight metaphor. We can think of the spotlight being extinguished and the whole universe thus going dark, leaving only zombies, or we can think of the spotlight becoming untethered and shunting over to someone else such that that person becomes the universe’s new looking glass, the new vessel in which it experiences itself. The extinguishing spotlight interpretation is, of course, rather nakedly solipsistic. Since we are interested in overcoming our instrumental biases toward solipsism, let’s take a closer look at the shifting spotlight interpretation.
You could think of it, I suppose, as a kind of ontologically innocent reincarnation, one that involves no spatiotemporal transfer of immaterial souls or personality traits or memories but only of a subjective interiority, a what-it’s-like-to-be-ness. Accepting this possibility immediately prompts those questions that so confounded me as a child: Why and how does the spotlight move? What factors determine whom it settles upon next?
Confused as these questions may ultimately be, if we’re going to persist in the moving spotlight metaphor, we’re going to have to settle on some provisional answers to them. We could suppose, following certain Hindu, Buddhist, or Jain schools of thought, that spotlight transfer follows some sort of karmic principle. Or we might expect consciousness to merely jump to the physically nearest available brain. Or to the most structurally similar. Or to the one who knew you best in life. Lacking compelling evidence for any of these possibilities, however, I think the morally safest assumption—and the one that makes for the tightest analogy to the real world—is that the spotlight will be reassigned randomly.
Now, for me—an American at the intersection of a majority of the most fortunate demographic categories in my country—this is a pretty scary prospect. A random spotlight shift is almost certainly going to mean a downgrade in quality of life and opportunity for positive experience. Most of the 7.3 billion other people in the world, after all, are worse off in those respects.
To better ground me in the moral reality of this situation, I have for the past year or so engaged myself in a rather strange and macabre hobby: About every other night, before falling asleep, I try to imagine, in as vivid detail as I can manage, what it’s like to suffer through various horrific ordeals. Ordeals like:
Being consumed in flames. All my skin screaming at the top of its nerves. A Pendereckian chorus of unremitting agony.
Drowning. Cold and desperate, knowing that the gulp for which my lungs are spasming will bring the end of all I’ve ever known and felt. Too panicked to search for final peace or think of loved ones.
Starving to death on a bustling street under a merciless sun and a rain of unconcealed contempt from well-to-do passersby.
Withering away under various excruciating and humiliating diseases.
You get the idea.
All such things have happened to people—to persons, to subjects—and all will almost certainly continue happening into the foreseeable future. Now, if I give my thinking over to the moving spotlight metaphor, I have to conclude that it’s inevitable that at some point the spotlight will shine on some horrific incident like those I force myself to imagine. Perhaps it will not happen in this life, to this self, but after some number of shifts, atrocity is bound to find its way into the pilot’s seat.
This terrifies me. As viscerally as though this fate awaited me personally (the fate awaits “personally,” whether the particular person I am will be present or not). This view of death as a sudden shift of the spotlight is a powerful and strange one, a perspective freighted with neither the existential dread of annihilation nor the comfort of escape from worldly tribulations. Its promise is not oblivion but something rather more like imprisonment. Subjectivity continues after me. Interiority continues. Someone—some I, some me—will have to be present for any unvalued suffering allowed by the future I leave behind. The only thing I can do about this is try to minimize that allowance, to work toward a world in which such horrors are less likely—less numerous—than they otherwise would be.
I stress here: The moving spotlight is just a bit of conceptual scaffolding for building a wider, sturdier bridge between our motivational machinery and the demands of practical rationality, properly de-relativized. The reason these imaginative exercises so terrify and galvanize me is not because I actually think there’s a single phenomenal spotlight that shifts around to different vessels, blessing or cursing them with inner worlds and the capacity to have things go well or poorly for them; it’s because there are no first-person-accessible facts on the basis of which anyone could distinguish such a world from the one we really do inhabit: A world of countless non-shifting spotlights all shining simultaneously.
What, then, will it be like, phenomenologically speaking, for you to be dead? In either a shifting spotlight world or a multiple spotlight world, it’ll be precisely what it’s like for someone else to be alive. For better or for worse. The notion, often merely implicit, that death will be experienced as some kind of void or nothingness still presumes a trace of subjectivity where none, by definition, can persist. There is no subject to be party or host to the local nothingness death brings—the absence of such is precisely what death is. All the subjects, all the selves, all the spotlights are elsewhere, contending externally and internally with whatever world we’ve left them.
We may press the spotlight mode of thinking further. For, as we’ve mentioned already, there are events other than death that sever the phenomenal thread, if only temporarily: deep sleep, general anesthesia, seizure, coma. The spotlight could shift each night when we go offline and no subject kindled by that light would have any way of knowing. They would wake with memories of a prior life for which no one was actually present and be none the wiser as to this fact. We can go further still. One of the storied mindblowers of introductory philosophy courses is the observation that your present experience is wholly compatible with you having just been magicked into existence, false memories and all, a moment ago. The spotlight might be constantly ping-ponging between different subjects; not a one of them, you included, would be the least able to tell.
If these various moving spotlight worlds are phenomenally indistinguishable from a multiple spotlight world by anyone capable of having stakes in the distinction, and if imaginatively putting yourself in these moving spotlight worlds troubles and stirs you as it does me, then you ought to be at least as troubled and stirred by the prospect that you’re in a multiple spotlight world. After all, such a world is one in which every atrocity will be illuminated, in which someone must always be subject to the violence, the indignity, and the grief we permit or actively inflict upon each other. The intense empathy you (hopefully) feel when imagining yourself in a moving spotlight world is every bit as warranted, every bit as justified, every bit as rational in the world in which we actually find ourselves.
Every bit as needed too.