UPDATE 12/1/19: Having read Sean Carroll’s book, Something Deeply Hidden: Quantum Worlds and the Emergence of Spacetime (2019), and having listened to the several companion podcast episodes he released in conjunction with the book’s publication, and having read various responses by others (like this one at Forbes, by Chad Orzel: “Many Worlds, but Too Much Metaphor”), I feel I now have a better understanding of the many-worlds formulation (Carroll shuns the word ‘interpretation’) of quantum mechanics than I did when writing the below blog post. But it’s still a naive understanding, and the below questions remain for me items of curiosity.
The many-worlds interpretation of quantum mechanics includes the thesis that, given a set of possible outcomes, every one of those outcomes is certain to happen in at least one world that branches off from the current one. I’d like to ask several naive questions related to that thesis, the underlying theme being: Do experts tend to play too fast and loose with their conception of possible outcomes, and, if so, what implications might this have for (many-worlds) quantum mechanical reality more broadly?
I’ve heard people make rather silly claims about the sorts of things that may happen in branching worlds. These generally have not come from experts. For example, someone once told me that the many-worlds thesis allows for Jesus to be both the son of the Christian God and merely a mortal prophet; and, even more absurdly, that it allows for atheists and believers in God (in the thoroughly traditional sense) to be simultaneously correct in their beliefs. I take these things to be obviously impossible, and thus absent from any branch.
I’ve heard experts endorse strikingly strange—though not obviously impossible—cases, but I’ll hold off for a moment on giving examples. First, I’d like to outline three kinds of impossibility. My goal isn’t to rigorously parse the space of possibility, but rather to tidy up what I mean by “impossibility” while also demonstrating that some events are not only impossible, but may be so in differing and subtle ways.
Logical Impossibility: Something is logically impossible if it contains a clear contradiction. It is logically impossible for a glass to be simultaneously full of water and empty, or for a flipped (fair, double-sided) coin to land simultaneously heads-up and tails-up. The coin turning into a frog and hopping away is not logically impossible, at least not obviously so. Perhaps such a thing would entail some contradictions that could be unearthed with a little digging. But such examples are often used by philosophers and physicists as not only being logically possible, but maybe even physically possible (as we’ll see below).
Metaphysical Impossibility: This is a tough one. I want to say, with self-conscious circularity, that something is metaphysically impossible if its impossibility is of a notably metaphysical flavor. There are better definitions to point to (e.g., typically involving ways the world cannot have turned out), but this could lead us down a rabbit hole that might, unhelpfully, lead us back to logical impossibility. So I’ll just give a few examples. (At your own risk, enter that rabbit hole here.)
It is metaphysically impossible for the flipped coin to transform into a square circle; or for me to have been born Genghis Kahn; or for the physical texture of smoothness to exist disembodied from some smooth-surfaced object (or from some collection of particles thus arranged, in the event that composite objects are metaphysically—or, more specifically, mereologically—impossible).
That last example posits that there is some accidental property of the surface of the object (a certain arrangement of physical stuff) that is in part responsible for the secondary quality (i.e., as experienced by an observer) of smoothness. Something similar can be said of color, and if you think about such things hard enough, you’ll start to wonder about relations like smoother than and taller than: for tallness to happen, there must be something that is tall, and there must be something the first thing is taller than.
Physical Impossibility: Something is physically impossible if it defies the laws of physics. It is physically impossible for me to lift 15 trillion tons or to run faster than the speed of light. I assume it’d be physically impossible for a human to grow 2,000 feet tall, much less to grow bigger than the universe.
All of these classifications matter here, but the many-worlds interpretation only commits us to the thesis that everything that is physically possible (i.e., that is compatible with the laws of physics) happens. Of course, most impossible events do not clearly fall into just one of these categories. If something is logically impossible it is also metaphysically and physically impossible. And something could be logically possible, yet metaphysically and physically impossible. To say that it is impossible for water to be simultaneously frozen and boiling entails, I presume, a number of contradictory propositions about those sorts of particles while in those sorts of states. The same goes, unless I’m mistaken, for when proposing that some bit of water may freeze deep within the larger environment of a boiling pot on a hot stove—I take it, that is, that this isn’t just a statement about the extremely low probability of a spontaneous lowering of entropy, but rather posits an outright impossibility.
With this talk of entropy, I begin to test the boundaries of impossibility, and do so in order to make salient what it is I’m really interested in here, which is the vagueness surrounding what things are indeed impossible, and thus will not arise in a branching world.
Before pushing hard against those boundaries, here are a few more things I take to be obviously impossible. There will not be a branch…
…in which π (i.e., the ratio of a circle’s circumference to its diameter) suddenly equals 12, particularly when assuming nothing else about the world is changed—including the things the word “circle” refers to;
…in which 2 + 3 = 5 while 3 + 2 = 17 (again, assuming those numerals refer to the exact same quantities they refer to right now);
…in which my past suddenly changes so that, instead of being born in Philadelphia, I was born in Mexico City (though my brain may rearrange itself so that my childhood memories change);
…in which, at a given moment, Sonia is taller than Jeff, and Jeff is taller than Marcus, and Marcus is taller than Sonia.
There are plenty of astoundingly strange things, however, that may not only be possible, but that the many-worlds interpretation predicts, with 100% certainty, will happen somewhere. For instance, a phenomenon I’ve elsewhere described as Spooky Cards is guaranteed to happen, somewhere. An alpha-numeric deck of cards reveals, for weeks on end and by way of random draws, correct answers to any question you ask—What did I dream last night? What’s tomorrow’s winning lottery number?—in any language you like. The cards at one point spell out: “Stop testing us, we are prescient.”
As I put it in that writing, perhaps worlds of magic are those in which the highly improbable has just so happens to happen often—every time you try to walk through a wall, you happen to phase smoothly through it; every time you guess a die roll or the occurrent contents of another’s mind or what the next word of any novel will be, you happen to get it all exactly right, every time.
The probabilities of such things are unfathomably low. But they are possible given our current understanding of logic, metaphysics, and physics. It may also be possible for a coin to transmogrify into a lively frog. But, as we’ll see, it gets worse than that.
2. From Weird to Worse: Uncharacteristic Behavior and the Spontaneous Rearrangement of Brain Stuff
I have recently been learning a lot from physicist Sean Carroll, especially by way of his Great Courses lectures on time (more about which soon) and his discussions with other physicists. I’ve especially enjoyed his discussions with philosopher David Albert, two of which I’ll reference below (there’s a there’s a third one I won’t refer to here, but is worth sharing: “Science Saturday: Time’s Arrow | Sean Carroll & David Albert“).
To be clear, I am particularly interested in the things such experts say and endorse when speaking to the public—a public that doesn’t usually read the books—much less the academic articles—in order to get into the details. Popular communicators of physics seem to be taking more care in this regard lately, due to people relying on quantum mechanics to ground all sorts of ridiculous claims (e.g., about human consciousness grounding reality). One thing I like about Carroll is that he has proven himself time and again, and on several topics—quantum mechanics, time, entropy, the fundamentality of information—to be one of the most self-consciously careful I know of, thus his eagerness to chat with philosophers like Albert.
So, I will rely on Carroll’s output as a point of reference, while drawing from other experts when appropriate (e.g., to present a corroborating or differing opinion). Let’s begin with the following exchange on Sam Harris’s Making Sense podcast, Episode 124: “In Search of Reality” (4/21/18) (starting at 16:20; or listen on YouTube):
Harris: In the many-worlds version, everything that can happen is happening.
Harris: Right? So in some sense there is only the actual. Most of it’s not in this universe.
Carroll: That’s right.
Harris: But it’s still happening. So there was some [“minute”?] possibility that I might have picked this up and put it down and then picked up again and put it down and did that 75 times…
Harris: …to the consternation of everyone in the room.
Carroll: Hopefully the probability is low, but yes.
Harris: But if there’s a non-zero possibility of that, that happened somewhere. Right?
Carroll: Yes, that’s right.
Harris: And every other conceivable adumbration of that.
Harris: So I was singing “The Star-Spangled Banner” at one point while I was doing that.
Carroll: Yep. It gets worse than that, yeah. But yes.
Harris: If you heard me sing you know it doesn’t get much worse than that.
Carroll: Maybe it doesn’t get worse.
Harris: But this is supposed to be science, right? But this sounds like the strangest and least believable idea on offer. How is it that science, after centuries of being apparently rigorous and parsimonious and hard-headed, finally disgorges a picture of reality which seems to be the least believable thing anyone’s ever thought of?
Carroll, after confirming essentially that anything that can happen does happen, answers Harris’ question: “it is the simplest, purest, most parsimonious way of making sense of the data.” I’ve heard this elsewhere stated as something like, “it’s our best current interpretation of our equations.” Note that this isn’t just in support of quantum mechanics, which has proven to be mightily robust, but is in support of favoring the many-world interpretation of quantum mechanics over other interpretations (more about which in a moment).
I am not here to challenge the many-worlds thesis, however. At least not directly. What I am interested in is the “worse” Carroll refers to. To begin heading in that direction, let’s dwell a moment on the scenario Harris describes.
It is difficult for me to imagine that Harris would do something so uncharacteristic as to pick up and put down an object 75 times. What sort of change in chemistry would be required for him to act completely counter to his character and history? To suddenly strangle Carroll, flash the crowd, set himself on fire, or, most unthinkably, coherently and sincerely denounce meditation while claiming to believe in free will? Sure, it’s physically possible for Harris to enunciate words to the effect of disavowing meditation; but what sort of change in his body would it take for him to spontaneously do so while actually meaning those words? On what grounds are we committed to the (psychological) possibility of such a thing?
It seems to me that just because we can imagine someone physically doing something, it does not mean that, on the die being rolled, as it were, that thing appears on one of its faces. In short, it does not guarantee that the thing is in the sample space.
But if an outcome is ruled out of the sample space, there must be some principled grounds for doing so (and, even then, we might leave it in for practical reasons; I won’t explore this here). It cannot be ruled out on grounds of bizarreness. Rather, it must fall somewhere within the space of impossibility. Let’s put this thought on the back-burner for now, set to simmer.
Note also Harris’ use of the word “conceivable.” I take it that he did so intending its ordinary usage, but in metaphysics, it has—or aspires to have—a technical application that I’ll refer back to shortly. So back-burner this thought as well. For now, let’s continue to consider examples from the discussions of experts.
As bad as the above-imagined outcomes are, they still aren’t in the class of “worse,” assuming that by “worse” Carroll meant “difficult to believe” or “highly improbable.” To get closer to that, let’s dip into one of the aforementioned fascinating discussions with Albert, from Carroll’s Mindscape podcast, Episode 36: “David Albert on Quantum Measurement and the Problems with Many-Worlds” (3/4/19). Here, Albert speaks of his experience as a freshly minted physics PhD, working as a postdoc with Yakir Aharonov in Israel (starts at 10:45; or listen on YouTube):
I don’t know that I’ve ever met anybody in my life who had as quick and as sure an intuitive grasp, which is almost a contradiction in terms, about how quantum mechanical systems are going to behave. And it was just amazing. …
… he wasn’t so focused on what I would now call philosophical problems, but he just had this intuition about how you could make quantum mechanical systems do astonishing things. There were stories about him when he used to be at Yeshiva University in New York, where he would walk into people’s offices in the philosophy department just in order to have something to do that evening, and say, “Describe a physical effect which is obviously impossible.” So people would say, “I don’t know, a man turns into an elephant,” or something, and Aharonov would go home and figure out a quantum mechanical way that this could happen. And this was the kind of exercise that lay behind many things that he discovered.
“A man turns into an elephant” is heading us in the direction of “worse” possibilities. To get yet closer, I’ll turn to another popular communicator of science, physicist Max Tegmark. His 2014 book Our Mathematical Universe: My Quest for the Ultimate Nature of Reality has a wonderful and clear account of the two major competing interpretations of quantum mechanics (i.e., the Copenhagen and many-worlds interpretations), told in the context of Tegmark’s own grapplings with quantum mechanics as a college student.
Tegmark appeared on Harris’ podcast to discuss the book: Episode 18: “The Multiverse & You (& You & You & You…).” In the part of that exchange I’m about to cite, they are discussing what Tegmark calls the Level I Multiverse, which is simply the idea that “in an infinite space created by inflation, everything that can happen according to the laws of physics does happen. And it happens an infinite number of times” (Tegmark 123). This is not explicitly the many-worlds interpretation of quantum mechanics, whose branching worlds Tegmark refers to as Level III multiverses, but it might as well be. In fact, Tegmark goes so far as to say that, “In this sense, the Level I and Level III multiverses are unified: as long as you have an infinite space with a Level I multiverse, you can ignore all its Level III parallel universes, since they’re in practice all just identical copies” (Tegmark 224).
It’s also worth noting that Tegmark endorses the many-worlds thesis: “Since I believe that Hugh Everett’s parallel universes are real, I can’t help thinking about what they’re like” (Tegmark 196).
Harris: Things begin to get very weird given this fact that inflation—which, as you said, is the best current picture of how things got started—given that inflation predicts a universe of infinite extent, infinite space, infinite matter, and therefore you have a universe in which everything that is possible is in fact actual, everything happens. Everything happens, in fact, an infinite number of times. Which is to say that you and I have this podcast an infinite number of times and an infinite number of different ways. …
We’re still talking about the Level I Multiverse here, so we’re just talking about, you know, if you could travel far enough fast enough away, you’d arrive on some planet disconcertingly like Earth where you and I are having a virtually identical podcast but for a single change in term, or I just decide to shave off my eyebrows in the middle of this conversation.
Tegmark: [brief chuckle] Exactly. Or I switch to talking French. … [skipping ahead about two minutes] … So if you have an infinite number of other regions equally big, and you roll the dice again in all of them, then you can calculate that if you go about a googolplex meters away, you will indeed end up with just what you described: a universe that’s extremely similar to this one except that one minute ago you all of a sudden decided to start speaking Hungarian instead.
Here we have yet two more bewildering scenarios in which to imagine Harris. The first—shaving off his eyebrows—is not so bad, though it does pose some difficult questions. Apparently, there is a sudden change in both Harris and Tegmark, such that Harris not only utters the words, “I just decide to shave off my eyebrows in the middle of this conversation,” but actually does shave off his eyebrows while neither Harris nor Tegmark notice; or if they do notice, they don’t mention it. That is, if everything else about the situation is identical, then their brains and their locutions and so on are identical save for whatever change is required so that Harris shaves his eyebrows. Unless the idea here is that the podcast is identical up until he shaves off his eyebrows. Fine, let’s assume that.
Tegmark next mentions the possibility that he himself suddenly switches to speaking French in another world. Maybe Tegmark knows French. Or maybe he meant that he would suddenly start speaking French despite not knowing that language. This is the more interesting option and the one I’ll assume he means us to imagine, particularly given the second bewildering scenario Tegmark asks us to imagine for Harris, who I’ll safely assume does not know Hungarian: “one minute ago you all of a sudden decided to start speaking Hungarian.”
Here we are now, as far as I’m concerned, well into a world of “worse.”
What I consider the worse are possibilities involving the spontaneous rearrangement of the parts composing a well-functioning human brain—often noted to be the most complex bit of machinery in the known universe—into some new, equally well-functioning human brain, but one of an entirely different character. This strikes me as far, far, far more astonishing than, say, a flipped coin happening to phase through the floor while leaving undisturbed the structural integrity of all bodies involved, or of an elephant turning into a man. Wait! Unless that man has a fully formed human brain and speaks Hungarian. In that case, it’s pretty worse.
I suppose the idea here is that one’s brain parts (and those of the tongue and throat) spontaneously rearrange themselves so that Harris just so happens to now be proficient in Hungarian. Or, maybe, so that Harris begins to sort of involuntarily ejaculate sounds that a Hungarian-speaking person would understand. I’m not sure which would be harder to believe, but I assume Tegmark means the former, given that Harris “decides” to start speaking Hungarian. Presumably, though, if the former is possible, so is the latter.
Perhaps, though, they are both logically impossible—in the sense that, were we (or the brilliant Aharonov) to try to give thoroughly detailed account of them, we’d run into contradictions. But let’s suppose them not only logically possible, but physically possible.
Think of what this means! An instant from now, our world will branch into worlds in which all English-speaking brains suddenly become Hungarian-speaking brains, and vice versa. And, in some of these worlds the particles of all text written in English will coincidentally rearrange themselves into perfectly translated Hungarian, and vice versa. If that text is on a computer, then the code will appropriate rearrange itself, and the particles responsible for human memories and for audio and visual recordings in those languages will similarly switch, and historical accounts involving philology, etymology, how languages comes to be spoken such and such a way in such and such a region—all of this and more will spontaneously just so happen to change, somewhere. Imagine the changes being so thorough that nobody notices.
My money is on this being impossible, that enough digging would reveal some logical or metaphysical impossibilities, thus ruling such events out of the sample space—a space that is identical to the space of physical possibility. I believe this could be accomplished just at the stage where Harris alone is speaking Hungarian, but I won’t push this here. Instead, I’ll simply acknowledge that, if such a thing does not defy the laws of physics, it is possible, and, moreover, if the many-worlds thesis is correct, then such worlds have come to exist, and will again many times over, with a probability of 100%.
I think the stage now sufficiently set for my naive questions.
3. Some Naive Questions About Many-Worlds Quantum Mechanics
Some of these questions need setting up. Apologies for length. The broader theme, again, is whether difficulties in mapping the space of possibility has any substantive implications for many-worlds quantum mechanics. I’ll actually put a finer point on this: Does that difficulty threaten to introduce to Everettian (i.e., many-worlds) interpretation the sort of vagueness that leads some theorists to reject Copenhagen interpretations?
3.1 Does rejecting the conceivability of philosophical zombies encourage us to reject other strange possibilities involving brains, or does the acceptance of other strange possibilities involving brains as conceivable encourage us to accept philosophical zombies?
Recall my characterization of spontaneous brain rearrangement as the worst of the “worse,” so to speak. The characterization is due to my inability to conceive of, for example, an English-speaking brain suddenly becoming a Hungarian-speaking brain. I take this to be similar to the most common criticism of the famous thought experiment in which we are asked to conceive of philosophical zombies (henceforth, just “zombies”). That is, humanoids who are in every way identical to fully functioning, active, conscious human beings, but who are no more conscious than a pile of fresh laundry (maybe even less so, depending on your views about panpsychism).
The point of the zombie thought experiment is to open up the possibility for a kind of dualism (e.g., one in which there are both non-physical and mental properties of physical substance). That is, the possibility that, if you can imagine two identical beings, one of whom has mental content and the other doesn’t, then perhaps it makes sense to say that there is a separation between the brain physical and mental properties.
Some of the details will get spelled out a little more below, but if you want to go deeper into the arguments regarding zombies, dualism, and conceivability, a good place to start is the Stanford Encyclopedia Entry on Zombies. My hope here is to rely on the experiment only as a kind of gateway for questioning the conceivability of other kinds of brain-related possibilities.
The effectiveness of the zombie experiment hinges on one’s ability, or inability, to (convince oneself that one can) conceive of such creatures. Conceivability has a special place in metaphysics. Part of what grounds Descartes’ argument for mind-body separation (in his case substance, rather than property, dualism) is that those things can be conceived of as separate—unlike, say, the roundness of a ball and the ball itself. Also inconceivable is a square circle; if you disagree, try, for example, to truly and vividly imagine, much less construct, a square whose parameter-to-diagonal ratio is simultaneously 21.5 and π. And I, personally, cannot conceive of both determinism being true (in the way Carroll uses that term) and free will being true (insomuch as I can make coherent sense of that term, a difficulty, in fact, that is largely responsible for both libertarianism and compatibilism being inconceivable for me).
Many folks, including Carroll, reject that zombies are conceivable. (Note that I don’t know Tegmark’s opinion on zombies. Nor will I speculate about what Carroll would say about Tegmark’s Hungarian example.) From the above-linked podcast with Harris (starts at 11:20):
Carroll: Then, of course, there’s the famous philosophical zombie thought experiment. Can you imagine something that acts like a person but has no inner sensations? Does not know what it is like to be a person, but acts exactly like a person would act? And to me the answer is, no, that’s not even conceivable. Because if you asked a zombie, “What are you feeling right now?,” it would say it’s feeling something, because otherwise it would be acting differently, right?
Carroll: So, why is the zombie lying all the time about feeling something inside? How do you know you’re not a zombie yourself?
Harris: Yeah, yeah.
Carroll: So, I actually just don’t think that’s possible and I think that ultimately the attempts to wriggle around basing reality in stuff obeying the laws of physics don’t quite hold together.
Now, this was spoken in front of a live audience, a situation in which I wouldn’t expect someone to present the best version of their take on a complicated idea. Still, I was a little surprised to hear Carroll base his argument principally on external behavior (more about this in a moment), and was a lot surprised at his question, “How do you know you’re not a zombie yourself?” I was also surprised that the question seemed to be happily endorsed by Harris, who has an undergraduate degree in philosophy, a PhD in neuroscience, and teaches meditation (I enjoyed his book on that subject).
Anyway, it’s an easy question to answer. I know I’m not a zombie because I have internal sensations. And that’s that. To give it any more thought than this is silly. It’d be unintelligible for me to employ my conscious mind in the task of wondering whether I have a conscious mind.
Or maybe I’m missing something here. Is this question meant as a criticism of the zombie thought experiment? The most famous version of the experiment is due to philosopher David Chalmers, and it never asks you, the experimenter, to ponder such questions, nor does it imply that you should or could seriously ponder them. Indeed, Chalmers has been quick to point out—rightly so and in no way inconsistent with the experiment—that “consciousness is the thing we know about more directly than we know about anything in the world… one thing I can’t doubt is that I’m conscious”; which is just a restatement of Descartes’ Cogito (“I think, therefore I am”).
Or maybe Carroll means to show how hard it is to discern whether those around you are conscious, by pointing out how hard it would be to convince others that you are conscious: You say you have internal sensation, but that’s just what a zombie would say. And here I’ve fallen for the trap. Except there’s still a problem here. Carroll’s argument depends on behavior that he a moment ago said only makes sense if the exhibitor of the behavior is in fact conscious.
So, I actually thought this question might have been asked in error due to the energizing dynamics of the live setting—audience enthusiasm can sometimes imbue intellectual performers with an alacrity for the silly. Maybe he got carried away. The fair thing to do, then, in order to get a more focused picture of what he means by the impossibility of zombies—i.e., of what it means to try to “wriggle around” the laws of physics—is to look at Carroll’s 2016 book, The Big Picture: On the Origins of Life, Meaning, and the Universe Itself, the 41st chapter of which is called “Zombies and Stories.” The book is made up of rather concise, clearly written chapters, so it’s not difficult to pick out his key claims (though, as always, I recommend reading the book for yourself).
The first thing to note is that he does not merely rely on behavior. Now, to be clear, I am sympathetic to arguments against conceivability on the grounds of behavior, and, in fact, have at times posed that criticism myself, but not against behavior per se, rather than against the idea that the behavior would arise with creatures that share our evolutionary history, which at some point seemed to become inextricably mixed up with—and in fact dependent on—conscious experience. Thus our present situation as beings who cry at the movies and fall into complicated fits of nostalgia at the smell of moth balls (well, I do, anyway).
I’m no longer as moved by this argument as I used to be. The question isn’t how the beings come to be, but how they’re made up. On this point, Carroll raises a criticism I agree with. From his book:
Zombies would be collections of particles in exactly the same arrangement as would ordinarily make up a person, obeying the same laws of physics and therefore behaving in precisely the same way, but lacking the mental properties that account for inner experience. As far as you can tell by talking to them, all of your friends and loved ones are secretly zombies. And they can’t be sure you’re not a zombie. Perhaps they have suspicions.
The big question about zombies is a simple one: can they possibly exist? If they can, it’s a knockout argument against the idea that consciousness can be explained in completely physical terms. If you can have two identical collections of atoms, both of which take the form of a human being, but one has consciousness and the other does not, then consciousness cannot be purely physical. There must be something else going on, not necessarily a disembodied spirit, but at least a mental aspect in addition to the physical configuration.
When we talk about whether zombies are possible, we don’t necessarily mean physically possible. …
As long as zombies are conceivable or logically possible, Chalmers argues, then we know that consciousness is not purely physical, regardless of whether zombies could exist in our world. Because then we would know that consciousness can’t simply be attributed to what matter is doing: the same behavior of matter could happen with or without conscious experience. (Carroll, 356–357)
Carroll’s talk of “identical collections of atoms, both of which take the form of a human being” outlines the territory in which most of the conceivability debate happens. I share Carroll’s (and many other people’s) concerns: it is inconceivable to me that there could exist—say, right here in my living room—two identically arranged brains attached to two humanoids engaged in an exciting discussion about music, while one humanoid is conscious and the other a zombie. This is a clear strike against dualism. For me at least. But I don’t see it as a clear strike against the conceivability of zombies altogether. Namely because, once dualism is set aside, there is no reason why the brains in question must be identical.
To see this, it’s helpful to look more closely at Carroll’s account of the connection between experience (a.k.a. consciousness or inner sensation or mind) and behavior. Namely, he seemingly implies that behavior isn’t enough to rule out a humanoid’s being a zombie. Or does he imply this? It’s ambiguous. Also, he regrettably again raises the surprising question from a moment ago:
Deciding whether something is “conceivable” is harder than it might seem at first glance. We can conjure up an image in our mind of someone that looks and acts just like a human being, but who is dead inside, with no inner experiences. But can we really do so without imagining any differences in the physical behavior of them versus an ordinary person?
Imagine a zombie stubbed its toe. It would cry out in pain, because that’s what a human would do, and zombies behave just like humans. (Otherwise we would be able to recognize zombies by observing their external behavior.) When you stub your toe, certain electrochemical signals bounce around your connectome, and the exact same signals bounce around the zombie connectome. If you asked it why it cried out, it could say, “Because I stubbed my toe and it hurts.” When a human says something like that, we presume it’s telling the truth. But the zombie must be lying, because zombies have no mental states such as “experiencing pain.” Why do zombies lie all the time?
For that matter, are you sure you’re not a zombie? You think you’re not, because you have access to your own mental experiences. You can write about them in your journal or sing songs about them in a coffee shop. But a zombie version of you would do those things as well. Your zombie doppelgänger would swear in all sincerity that it had inner experiences, just as you would. You don’t think you’re a zombie, but that’s just what a zombie would say.
The problem is that the notion of “inner mental states” isn’t one that merely goes along for the ride as we interact with the world. It has an important role to play in accounting for how people behave. (Carroll 357–358)
Again, I am 100% certain that I have inner mental states. I would suggest these revisions: “You think, and in fact know with certainty, you’re not, because you have access to your mental experience. … You don’t think you’re a zombie, and in fact know with certainty that you’re not one, but that’s just what a zombie would say.” Though, again, the question of whether I’m sure I’m not a zombie seems to be designed to demonstrate just how difficult it would be to convince someone else that I’m not one—in which case, behavior alone isn’t enough. But apparently, if I take Carroll’s question at face value, neither is my own inner experience; at least, not enough to be sure.
In the end, I’m not sure what work the question is supposed to be doing, and I’m not entirely sure whether his case against the inconceivability of zombies relies strictly on the “atom-for-atom-identical” argument, or if behavior alone is meant to be enough to do that job, despite his asking whether we can really imagine human-like zombies “without imagining any differences in the physical behavior of them versus an ordinary person.”
Maybe these are meant to be two different levels of talk about the same thing (as with the free will case; see below). But I’m unsure, given his insistence, for example, that “that’s just what a zombie would say.” Or does he mean talk of behavior to entail talk of atoms qua atoms (unlike in the free will case). I don’t think so; identical behavior doesn’t entail identical—or even similar—brain states.
At any rate, whether Carroll is committed to zombie’s being behaviorally inconceivable (i.e., with no talk of atoms or brain stuff) is not essential to the argument I’m about to pose, but it does strengthen its impact, as my argument is meant to raise a challenge to any zombie-denying many-worlds proponent (who think not only that highly improbable events are nevertheless possible, but that they are certain to happen at every moment). Since Carroll unambiguously argues from behavior on Harris’ podcast, and at least somewhat seems to do so in his book, I’ll assume that behavior matters for him to a significant degree.
It is obvious that differently arranged brains can result in consciousness, if only because my brain is not now arranged precisely as it was five minutes ago. Moreover, there are conscious people with quite differently arranged brains than mine—people, for instance, who’ve had their left hemisphere removed—who are doing just fine. So, it’s conceivable that there’s no specific relation between particles that must be present within a brain for it to be conscious. There may be tell-tale markers for consciousness, but this may well amount to a kind of family of brain arrangements, rather than some specific arrangement. Indeed, if all the conscious human brains we look out do turn out to have some sort of structural signature, we can’t know for sure that there isn’t some other sort of similarly appearing arrangement that would produce consciousness.
Here comes the move to zombie-conceivability.
It strikes me as conceivable that a brain could be nearly identical to every conscious human brain in the world, yet not be conscious. I’m not necessarily referring to something like blind-sight, by the way, wherein a person’s behavior is clearly influenced by visual input though the person is not aware of any visual input (i.e., the person sincerely reports blindness); in that case, there may be some conscious part of the brain the body’s “driver” simply doesn’t have direct access to. There’s more to talk about there, but I’ll set it aside, as it’s not the sort of thing I mean here.
What I’m referring to is a humanoid with a brain that appears exactly like we’d expect a conscious human’s brain to appear, but there is no phenomenological account of the thing whatsoever. I’ll take this a step further. It’s conceivable that the zombie brain’s arrangement could be identical to that of a conscious brain given a snapshot of each, but the way the zombie brain functions—what the neurons do, etc.—does not result in consciousness (so the two brains would not appear identical over several snapshots of sufficiently fine granularity).
Here comes the naive question about many-worlds quantum mechanics.
If it is possible for a conscious human to have a massive, spontaneous rearrangement of brain stuff, such that the human suddenly becomes an entirely different person or goes from being an English-speaker to a Hungarian-speaker, why can the same not be true for a shift of brain stuff such that the person is not conscious?
More crucially, since there’s no logical problem with a person, say, spasmodically uttering Hungarian (without understanding the words), why could the person-turned-zombie not do so while in this new non-conscious state? Why could the zombie not do so indefinitely, as a sheer matter of chance?
Finally, if many-worlds quantum mechanics is true, would it not be guaranteed that such things happen constantly? Would it not be guaranteed, in fact, that, a moment from now, everyone on Earth except for Sean Carroll will become such a creature, but will—as a pure matter of chance—continue behaving just as they always do?
Or is there something simply inconceivable about such scenarios? I think there is, both here and in the other spontaneous-brain-rearrangement cases. Is it intelligible to claim one conceivable and the other not? One possible and the other not? As Carroll notes in the above quote, it’s often difficult to be sure that something is conceivable. To borrow his example: I can only think it conceivable that my glass full of water contains no oxygen if I’m unaware that “water is made of molecules with one oxygen atom and two hydrogens” (Carroll 258). I suspect something similar is going on with many of the “worse” cases. At any rate, there must be some things we can imagine—i.e., think we can conceive of—that, were we to look more closely, would be rendered inconceivable.
Do many-worlds advocates—popular communicators in particular—play far too fast and loose with the sorts of scenarios they endorse as guaranteed to happen in the multiverse?
3.2 Is there a non-empty region of space carved out for moral impossibility?
There are some interesting sub-regions in the space of physical impossibility. Might one of those contain events of the sort some humans view as morally salient? I’ll elaborate on the question by way of Carroll’s discussion of free will.
I should first point out that his views seem to be idiosyncratic among the (I admit, small) circle of theoretical physics experts to whom I’ve paid some attention, only some of whom I’ve observed weighing in on the question. For example, in one of the earlier mentioned discussions, Albert tries to talk Carroll out of—or to at least get Carroll worrying about—what Albert describes as a “watered-down conception” of free will. Albert is a skeptic about free will, as I am, and perhaps on similar grounds. At least, we both, as he puts it in the above-linked video, “don’t know how to coherently formulate what the words mean.” My skepticism is not strictly grounded, then, in a commitment to determinism.
Carroll describes himself as a compatibilist about free will, which is how many philosophers today describe themselves, though he seems to mean something different by this term than what most (if not all) of those philosophers mean by it. He characterizes the position as positing the compatibility of determinism (or the laws of physics) with talking about humans as if they have free will. See, for example, the lecture in his Great Courses lecture series on time, “Lecture 12: Memory, Causality, and Action,” where he states (at 23:10):
There are people called “compatibilists” who think that the laws of physics do govern everything and, nevertheless, we should talk in the language of free will because we don’t know what will happen next. Even though the laws of physics tell us what will happen next if we know the microstate. We don’t know the microstate and therefore the way to speak about the world as if we are making choices. …
… a philosophical libertarian is one who believes [in free will because] our human choices overcome the laws of physics. You are not determined by the laws of physics. I am not one of those because I’m a physicists. But I do believe that free will is the right way to talk. …
… Personally, even though I believe in the laws of physics, even though I believe that the future is determined at least probabilistically, if we know the microstate of the universe, I don’t mind talking about free will. I believe that it’s ok to say that we have free will because we don’t know what the microstate of is; we have the macrostate of the universe. When we use the language of free will, we’re not describing the individual microstates of the universe. We’re not talking about atoms and molecules.
The usual meaning of compatibilism, as I understand it in the current discourse, is not just that determinism is compatible with the semantics of free will, but rather that determinism is compatible with moral responsibility, a condition that may or may not be characterized as “free will.” That is, a person can be seen as deserving punishment even if their actions are determined. So, this isn’t just semantics and it’s not just a casual term catchall term for saying I believe in the laws of physics and in free will (though maybe this is how non-philosophers often mean the term these days).
I’ve discussed different sorts of compatibilism (e.g., semi-compatibilism and semantic compatibilism) and other conceptions of free will elsewhere, and won’t get bogged down in such things today (for example, see: “Causation and Free Will: Compatibilist vs. Incompatibilist Intuitions” and “Frankfurt Cases & Moral v. Legal Responsibility”) Instead, I’ll move on, but will refer to Carroll’s position as semantic compatibilism, which essentially means that, even if we reject free will, words like “responsibility” and “choice” are still intelligible.
This may or may not, however, be a fair characterization of Carroll’s position, which I frankly find impossible to understand. Particularly if he means to say that folks deserve (in the usually understanding of that time) to be punished for misdeeds—i.e., that malefactors owe some degree of personal suffering to the world proportional to the severity of their malefactions.
Harris, in their discussion that I’ve been citing, seems to exhibit similar frustration in making sense of Carroll’s position, though neither he nor Carroll mentions desert or punishment. Rather, time is mostly spent on Harris’ attempt to reconcile Carroll’s beliefs about physics with his belief in free will, or at least with his belief that it’s ok to talk as if humans author their own choices in some deep, free-will-friendly sense. Incidentally, Harris rejects free will on deterministic grounds, and largely does so from the perspective of a neuroscientist. From the podcast (starting at 43:30):
Harris: Do you feel that your experience is compatible—let’s just say that determinism is true…
Harris: …and provably so, so that we could have people with the right scanners backstage actually anticipating everything we’re going to say before we say it.
Harris: So we could see a printout of everything we said here before it could possibly have been recorded, say, or there’s some way of proving to us that we are mere puppets. Is your conscious experience compatible with that fact?
Carroll: Sure! Yeah.
Harris: Ok, so that’s not an affront. The fact that everything you said tonight could have been predicted…
Harris: …is fine. That maps onto your experience.
Carroll: It could have been predicted by the imaginary Laplace’s demon in the back room. It couldn’t have been predicted by me.
Carroll goes on to state essentially the same point he makes in the lecture, about how we can talk at one level about atoms and molecules, and at another level about humans making choices, but not at both levels at the same time. (In a longer discussion, it would be interesting to compare this to the aforementioned talk of zombies at the levels of atoms and behavior.)
So far, I find Carroll’s position on free will impossible to understand as anything but an outright denial of free will, though he persists in rejecting this characterization of his position. More to the point, I find it difficult to align the position with the many-worlds thesis, in which the best prediction is: you will do everything that you it is possible for you to do.
To be clear, I’m not worried about Carroll agreeing with the deterministic, Newtonian, Laplacian picture while also accepting the irreducible chanciness of quantum mechanics. Earlier in the conversation with Harris, in fact, Carroll takes care to pause and ask that we put aside the uncertainties due to quantum mechanics so that we may contemplate Laplace’s demon (at 27:40). He is similarly careful in the above-cited Great Courses lecture (“the future is determined at least probabilistically”); and throughout the above-linked conversations with Albert; and in his above-cited book, The Big Picture (where he refers to a “quantum version of Laplace’s demon [p 378]).*
[*Overthought Rhetorical Disclaimer: While I really am not worried, I admit to making a rhetorical move, both here and throughout this writing. I’d like to be transparent, and careful, about it. The move is to show that it’s often difficult to parse out what theoretical physicists (and other theoretically minded folks) truly believe and what they claim to believe on the grounds of best explanation from evidence or argument—i.e., what they rate and accept as the best working theory, right now. So doing, I assume my motivation is to cast doubt on the ideas themselves.
A key distinction here is between acceptance of a theory and belief in what the theory proposes. These are often difficult to disentangle in practice, but need not always be. For example, Albert believes that the Copenhagen and many-worlds interpretations fail, though he seems to hold the many-worlds view in much higher regard than he does the Copenhagen view (more on this below), and he rejects the many-worlds view for problems internal to the theory, rather than due to finding a different view more intuitively appealing. (The accounts I’ve seen him give of these internal problems strike me as convincing. From what I understand, Albert favors a dynamical collapse interpretation.)
Something I value in Carroll’s work is his sober presentation of these ideas. He has often qualified the many-worlds thesis, and even quantum mechanics itself, as constituting persuasive pictures of the reality, but as possibly wrong. Though maybe he’s just being perfunctorily careful, or is even protecting his reputation. His forthcoming book, slated for publication this fall, is titled Something Deeply Hidden: Quantum Worlds and the Emergence of Spacetime, is being promoted as follows (e.g., at Amazon):
Academics discourage students from working on the “dead end” of quantum foundations. Putting his professional reputation on the line, Carroll says that crisis can now come to an end. We just have to accept that there is more than one of us in the universe. There are many, many Sean Carrolls. Many of every one of us.
The Many Worlds Theory of quantum behavior says that every time there is a quantum event, a world splits off with everything in it the same, except in that other world the quantum event didn’t happen. Since the beginning of the universe about fourteen billion years ago, this has happened about ten to the tenth to the one-hundredth power times. Step-by-step in Carroll’s uniquely lucid way, he sets out the major objections to this utterly mind-blowing notion until his case is inescapably established.
Does the book amount to his dropping the perfunctory care he’s taken surrounding certain controversial notions in physics, or does this just mean he’s thought more about it and his beliefs have strengthened? Maybe this and other questions will be cleared up in the book, which I intend to read.
Whatever the case, I strongly second the description of Carroll’s explanations as “lucid,” particularly when it comes to ideas from physics. His lucidity is, in fact, why I’m drawing so much in this writing from his public discussions. Despite his lucidity, I have a difficult time making sense of what he believes about physics (i.e., the micro-level of the world), and how he reconciles those ideas with what he claims to believe about the world at the macro-level.
It is telling to me that such difficulties arise even with such a lucid and careful interlocutor and explainer. Also of value are his discussions with Albert, who, as a philosopher, demands a greater degree of precision and coherence than physicists are generally expected to aspire to. Carroll himself not only says this himself, but holds that intellectual ethos in high regard and as important (e.g., he notes this in the intro to the Great Courses lecture as well as in the intro to the recent Mindscape podcast that features Albert as a guest). I am, in fact, grateful to Carroll not only for aspiring to that ethos himself, but to his public promotion of the importance of philosophy in this respect.
That said, I’ll pick on someone else for just a moment. Recall above that Tegmark wrote, “I believe that Hugh Everett’s parallel universes [i.e., the many-worlds universes] are real,” something Tegmark refers to in his appearance on Harris’ podcast (e.g., when he flies in a plane, he tells himself that, in the overwhelming majority of universes, it doesn’t crash; at 53:40).
I don’t know what to make of such claims, given that he also seems to speak at times as though even just the basic notions of quantum mechanics are to be taken seriously on the grounds that they are the best interpretation of beautiful and simple equations (e.g., at 49:33); and it allows you to make predictions (e.g., at 12:58, “These physics theories, even though they sound crazy, as you yourself said here, they actually make predictions that we can actually test. That is really the crux of it. If you take a theory of quantum mechanics seriously, for example, and assume that particles can be in several places at once, then you predict that you should be able to build this thing called a transistor which you can combine in vast numbers and build this thing called a cellphone…”); predictive power is a good reason to make use of a theory, but not to believe that it’s model of the world is veridical.
In another podcast with Harris (Episode 120: What Is and What Matters, also with philosopher and writer Rebecca Goldstein), Tegmark says of modal realism (on YouTube at 42:44), “if you loosely speaking take it to mean that everything that could exist does exist, I find that an interesting idea, but it’s a little bit too wishy-washy to be really scientifically testable… the various theories of physics that give you some kind of multiverse … those are more restrictive, in a way: it’s not like everything I can think about after I have too much wine exists, but rather, if you have some particular equations in physics: Does it have this solution or have another solution too? Maybe that exists. That’s the kind of alternative reality that these theories tend to give. …”
I would love to get some examples of the sorts of things he “can think about” after having too much wine! Also, he admits to not knowing much about modal realism, but still manages to hit upon ways it can get silly, depending on whom you talk to. This includes round squares, and, well, the general idea that “There are objects of which it is true that there are no such objects,”1 the most oft-quoted line from philosopher Alexius Meinong (1853–1920), who describes two different meanings of “there are” here and, well, I won’t say any more about this here. Let’s just say there are philosophy majors sitting in a metaphysics class right now discussing weirder, and certainly more arcane, things than many-worlds quantum mechanics. And the dissenting student can’t just say, “this is too weird”—she has to engage the argument, the logic, the squiggly notation and abstractions.
Tegmark goes on to essentially say that for something to happen, the probability must be non-zero, and that consideration of the vast number of things that can happen (at 45:19) provides “very disturbing notions, but you can’t dismiss it just by saying it sounds too weird, right? The way you dismiss would be to falsify the physics theories, Alan Guth’s equations. … that’s how we’re going to sort it out.” Fair enough, but I still struggle to figure out the distinction between most robust and elegant equations and what I really believe about the world.
Perhaps on the one hand there is, for theoretical physicists, what-I-have-faith-about talk, and at other times there’s let’s-be-good-scientists talk? I often hear physicists say that good science has skepticism built into it, but then go on to make claims with seeming certainty about propositions that are not universally held even among those in their subfield; this often begins looking like arguments from intuition (a source which these folks are usually very quick to point out as irrelevant), or even from faith (I take it this suggestion would be rejected with even greater vehemence than that of intuition).
How else, though, to make sense of two people with essentially the same educations and access to information developing so many different “best interpretations” of the same data?
To be clear, I have heard it said, such as by Carroll in his Great Courses series on Time—Lecture 8: “Time in Quantum Mechanics” (at 23 minutes)—that, while physicists in general seem to take for granted the Copenhagen interpretation[s], “what you’ll find is that in the subset of physicists who think hard about quantum mechanics, nobody believes in the Copenhagen interpretation.”
Tegmark, on the podcast with Harris and Goldstein, had this to say about the Copenhagen versus Everettian interpretations, though he did not indicate whether the physicists questioned were sub-specialists (at 1:27:20): “There is violent disagreement within the physics community, great controversies. In recent polls, well, I’ve been doing a bunch of polls at physics conferences for the past 20 years and the Copenhagen interpretation has been going steadily down. Right now, it’s between 30 and 50% who work a lot in quantum information tend to vote for the many-worlds interpretations, and about 50% tend to vote for the none-of-the-above interpretation because it’s all so confusing.”
He goes on to note that there are also some minority views floating around, and to rate all this as good and healthy because it means not everyone is looking in the same place for the same thing, so “we’re much more likely to find the actual truth.”
To find the actual truth. Again, fair enough, but it doesn’t help me understand what people actually believe—e.g., are they voting in the poll for the best bet for making better predictions, or for what they think is actually true? And, again, I’m often told, “built into all of this is science stuff is the virtue of open-mindedness.” Fine, but what to make of the many, many instances of “belief” talk? Of Carroll’s forthcoming book’s blurb stating his aim to settle the many-worlds question (broadly, if not in the details)? How much is sincere, perfunctory, showmanship, marketing (for book sales and grant funding), or just a matter of a kind of epistemological contextualism where one actually believes one thing in one context (at home in the dark and under the covers) and another thing in another context (at a physics conference) and might hedge in yet another (at a live podcast taping)?
(I have only mentioned three foundations theorists here [Carroll, Tegmark, Albert]; there are obviously many others who’ve written books and articles and appeared on podcasts and so on. Things could be made yet more confusing by introducing some of those folks, but let’s not.)
Finally, this returns me to the point of this disclaimer.
My persistent worrying about such (potential) incoherence borders on ad hominem, a border I’m trying not to cross, for example, by calling out irrelevant instances of hypocrisy. Nevertheless, I do think the extent to which experts (and this goes for philosophers, too!) are consistent in their dealings with an idea matters, and sometimes this extends also to their non-verbal behavior, or their behavior in daily life more broadly, which may to some degree align with or belie their purported beliefs.
This doesn’t make those beliefs wrong—an astronomer can intelligibly speak of the sun rising in the east, and she can exhale cigarette smoke while sincerely uttering the words, “It’s bad to smoke, don’t do it”—but it seems that a mismatch in purported beliefs and behavior does often raise questions; e.g., about how convincing the evidence really is or, in a more familiar case, about the true beliefs of the person who claims to care deeply about factory farming while consistently otherwise behaving no differently than someone who obviously doesn’t care about factory farming.]
Carroll acknowledges in The Big Picture the difficulty of making sense of punishment under compatibilism, though he leaves the question open and does not address desert, which is really what’s at issue here; we may, for instance, decide that punishment is necessary for maintaining social order, while understanding that no one actually deserves to be punished. I do not know if this aligns with what Carroll means—that is, that we’re better off talking as if people deserve to be punished. What he does explicitly state are claims like, “There’s a sense in which you do have free will. There’s also a sense in which you don’t” (p 378), which refers to the macrostate versus microstate levels, respectively; and that the concept of free will clearly “plays a useful role when we choose to describe human beings as people” (p 379).
So, while the discussion in the book is more nuanced than I’ve observed from him elsewhere, it doesn’t clear things up for me, including with regards to the naive questions about many-worlds quantum mechanics I’ll now pose.
What does it mean for free will if you are guaranteed to do everything you can do? If many-worlds quantum mechanics predicts with 100% certainty that, at every turn, I will do the most morally reprehensible thing possible, how is there any sense I can be said to have acted freely or, if we decide I’m acting un-freely, how can it be said I’m acting as a morally responsible agent? That is, as someone who deserves to be punished? Particularly given that the reason why I perform every possible action comes down, strictly, to something about the laws of physics—representable by some beautiful, simple equation—that guarantees that the events populating any given sample space all literally occur.
There is a yet more difficult naive question to ask here, which I’ll phrase several ways.
How do we determine what someone can do? How do we work out what events populate that sample space?
Is there, or could there conceivably be, some non-empty region in the space of impossibility dedicated to moral impossibility?
More precisely, are there morally salient actions that are physically impossible for some individuals but not others? I don’t mean due to differences in, say, physical strength, but in what we call psychological constitution.
For example, I mentioned that I cannot imagine Harris suddenly endeavoring to strangle Carroll during their podcast. I cannot even imagine Harris deliberating over such a thing, which is to say I can’t imagine it being a viable option for him to even contemplate. Perhaps there is some probability that this could happen—some spontaneous change in Harris’ chemistry overtakes his mind and body, suddenly turning him into a ecstatic murderer. I need to clarify further that I’m not talking about the possibility of some mad scientist hitting Harris with a mind-control beam that forces Harris to strangle Carroll. I’m talking about a very different sort of possibility: one in which Harris literally desires to strangle Carroll and then does so.
We might say that, metaphysically, Harris is now a different person, but this too is a different question, and one we may ask about any person from moment to moment—though one that really may carry some weight in some cases, as when someone suffers a brain tumor or entered prison at 18 and is now 88. These questions get at both the metaphysical difficulties and stakes of notions surrounding free will and identity (a starting place: Are those people who count as “me” all those people who roughly look like me and were conceived of by all the women and men who look roughly like my mother and father, who themselves are all the children of parents who look roughly like their parents, who themselves are… you get the idea?). It would be a distraction to get further into those questions here, but they are worth acknowledging.
The question now is simpler (though still hard): Given who Harris is precisely at one moment, what would it take for him to spontaneously take on every conceivable action in the next, including ones of moral salience that are atrociously out of character?
Maybe such a spontaneous transformation is impossible. Maybe if we dig deep and wide enough, we’ll encounter some flavor of impossibility; a logical contradiction, maybe.
If so, then maybe some compatibilist notion of free will can be preserved under many-worlds quantum mechanics. That is, maybe we would view any “truly moral” person as being immune to many-worlds branching—to only have morally permissible acts in their repertoire of possible branches. There may be many such branches, but they’re all within the space of moral permissibility. This would, of course, have nothing to do with the moral salience, as perceived by humans, of such acts, but must be strictly due to some principled sort of impossibility for an event that just so happens to have, for some humans, moral salience. I would remain skeptical about free will under such a view, and for reasons similar to those that ground my current skepticism (e.g., one does not get to choose one’s repertoire of possible branches—i.e., one’s sample space).
Otherwise, it seems to me that we simply have to say that everyone does every possible good and bad thing that, to put it crudely, we imagine to be physically possible for them to do. And that includes you. And by “you” I do mean “you,” at least at the moment of branching; and this is how I’ve heard everyone characterize the thing. Really, it’s the only way to characterize it. That is, it’s nonsensical to say, “I will do the right thing, but the version of me that branches off will do the wrong thing.” You can equally say, “I will do the wrong thing, but the version of me that branches off will do the right thing.” The good- and bad- doers are both equally you, at least until you’ve performed the act, in which case there are indeed two actors, though they are descended, as it were, from the exact same person. (I said I wouldn’t discuss personal identity, but here it creeps in.)
To nudge us towards the intuition—yes, intuition; do what you will with it—that maybe there’s something horribly wrong with this picture, the next time you’re in public, ask yourself what sort of rearrangement of the current state of the world (no fair changing the past) would be required so that, in the next moment, something in you spontaneously shifts and you begin attempting to do unspeakably ghastly things to all the children in the vicinity while spasmodically growling out passages from Nazi pamphlets translated into perfect English from the original German. Those movements from your body may be physically possible, but is this scenario logically-metaphysically-physically possible? If it is possible, many-worlds quantum mechanics predicts that it will happen with 100% certainty, and it will be you who does it, where “you” is exactly the person you are at the moment of contemplating this shift.
3.3 Is it conceivable that there ever may be only one possible outcome?
The many-worlds thesis predicts that everything that can happen will happen. But what can happen is bound by the space of possibility. Is it conceivable that there could ever be just one event in that space, so that only one thing can happen? Could such a thing be so expansive as to apply to an entire world? Or might such a world be guaranteed to develop eventually? In which case such a world guaranteed to exist? That is, a world that never branches? What would be the implications of this? For free will? For the progression of branched worlds (e.g., might all worlds quickly concretize into a non-branching worlds)? For quantum mechanics in general and within those “concretized” worlds?
3.4 Does vagueness in making sense of the space of (im)possibility threaten many-worlds interpretations with anything like the problems that threaten Copenhagen interpretations by the vagueness surrounding wave function collapse?
My (limited!) understanding is that the great appeal of Hugh Everett‘s contribution—i.e., the many-worlds thesis—to quantum mechanics is a precise solution to the measurement problem, a problem that is shrouded in vagueness when treated by the so-called Copenhagen interpretation, which was dominant at the time Everett was working (i.e., in the 1950s and as a graduate student who would soon leave the field due to scornful treatment). The vagueness Everett avoided was that involved in determining what counts as an observation or measurement or appropriate interaction with a particle such that gives rise to wave function collapse.
Such determinations are established according to vague or imprecise criteria (see below for an example), thereby, as I’ve heard Copenhagen rejectors put it, sweeping under the rug certain inconvenient implications or results of the math of quantum mechanics that Everett simply took at face value. Another way to put this is that it draws an arbitrary line between the macroscopic classical world (i.e., the observer’s world) and the microscopic quantum mechanical world (e.g., that of the electron in a superposition)—a distinction that doesn’t matter for most working for physicists, but does matters for those dealing, at the very least, in foundations. To quote Carroll:
…quantum mechanics is not just a theory of the very small. It’s a theory of the whole world, including the very large. Quantum mechanics should apply to me and you as well as it applies to individual electrons. If that is true, then the Copenhagen interpretation can’t be right. The Copenhagen interpretation relies on the fact that you and I are not quantum mechanical, that we are classical things. But no one really believes that the real world is made of small particles that obey quantum mechanics, and that then when you take all those small particles and then put them together to make a human being they suddenly stop obeying quantum mechanics. (Great Courses Mysteries of Modern Physics: Time, Lecture 8: “Time in Quantum Mechanics” [24:55])
The solution offered by the many-worlds interpretation is that there’s no need to say that, for example, the particle that was spinning in two directions at once while unobserved concretizes into one specific outcome once observed. Rather, when you observer the particle, one version of you literally sees it spinning one way in one universe, and another version of you sees it spinning the other way in a different universe. So, this keeps both events in play, rather than saying that only one of them gets to be the event that “actually happened in the end.” The wave function splits rather than collapses, so there’s no need to define what counts as an observation (or measurement, etc.; these terms are synonymous). What this preservers, in contrast to the Copenhagen model, is that the observed particle has a wave function but so do you, the observer, as you are also treated as a quantum mechanical system (i.e., a body made up of small particles).
This also means there is one wave function for the entire universe, which, I take it, somehow settles worries like, “What about conservation of energy? Where’s all the matter coming from for these new universes?” I couldn’t tell you how that works, but it apparently is all going to come down to a single, deterministic equation. Notice the word “deterministic” here, which calls again to mind the discussion of free will, or, more precisely, moral responsibility and desert:
Another benefit is that the evolution of the [universe’s] wave function is completely deterministic. … Schrödinger’s equation … tells you how the wave function evolves. In the many-worlds interpretation, the evolution of the universe is just as deterministic and reversible as it was for Laplace and Newton. If you tell me the wave function of the universe now, and you tell me Schrödinger’s equation, I can evolve that wave function forward in time, I can evolve it backward in time, and no information is lost. You may ask why it seems that information is lost when wave functions collapse [Dan Note: does he mean “split” or “evolve” here?], and that’s because I don’t know where I am in the wave function. There are different copies of me that are in the many universes that are produced, that’s why information seems to be lost. (Carroll, Lecture 8: “Time in Quantum Mechanics” [28:55]
To paint a yet more vivid picture of the sorts of problems at issue, I’ll turn again to a public discussion between Carroll and Albert. This won’t do justice to scope of their public discussions (which, in turn, don’t do justice to the discussion happening at the academic level), but it’s a start.
In the earlier-cited Mindscape podcast discussion, Albert tells the following story (which I’ve heard him tell elsewhere), and is one of my favorite examples not just of the difficulties surrounding solving the measurement problem by way of wave function collapse, but of the silliness that was taken seriously in those discussions not so long ago, as well as the vagueness and misguidedness of connecting observation with consciousness (something that has given leeway to a lot of popular misconceptions about quantum mechanics as a kind of catchall for whatever mystical ideas to which one would like to cling). This is not to make fun of what are very smart people, so much as to take it in as a kind of humbling epistemological lesson. Albert refers here to physicist Eugene Wigner (starts at 27:45).
Albert: There’s a famous essay of Wigner’s that listeners might be interested to look at called, “On the Mind-Body Problem”…
[NOTE from Dan: It’s actually called “Remarks on the Mind-Body Question,” and originally appeared I. J. Good (ed.), The Scientist Speculates: An Anthology of Partly-Baked Ideas. Heineman (1961). It also appears in the 2004 collection Philosophy of Mind: A Guide and Anthology (1st Edition). Or get it as a PDF here.]
Albert: … where he’s thrilled about this idea for the following reason. It’s been thought for a long time that the picture of the world we get from physics is hostile to certain ideas we have about ourselves as agents and as thinkers. There’s no room in this picture for mind, there’s no room in this picture for agency or freedom, so on and so forth. Wigner had this idea that it was precisely conscious agency that caused departures from the standard quantum mechanical equations of motion, that caused the collapse of the wave function. Indeed, he thought there was in this a new definition of the difference between animate and inanimate physical objects. What you mean by an inanimate physical object is precisely one that evolves according to the standard quantum mechanical equations of motion, and animate ones are ones that don’t, that can give rise to collapses of wave functions. And Wigner thought this was great because physics had for so long seemed hostile to the idea of a nonphysical mind somewhere in the world. It turns out, so he thought, not only is it not hostile, it requires such a thing in order to do its ordinary physical job.
Obviously this distinction between animate and inanimate is just as vague as anything else. I remember—I swear on anything you want me to swear on—that as a graduate student, I witnessed Wigner at a conference, in response to a question, speculating to the effect that he thought dogs could likely collapse wave functions but mice probably not. And you just sit there and you say, “This is not good.”
Albert: “This is not the way to do physics.” It remained in that kind of terrible state for about 40 years.
Carroll here interjects, with emphatic agreement from Albert, that it’s not the “ridiculousness” of the claim, but rather the “vagueness” of the claim that is a problem. Also, again, it’s understood that “to collapse the wave function” could be replace with “perform a measurement” (another wording I’ve heard Albert use, in fact, when telling this story), or with “observe.” Also, to be fair, current physicists thinking about these things seem to be careful about explicitly removing consciousness from the Copenhagen story (a potential manifestation of that here being whether mice are significantly less conscious than dogs), and being clear that any sort of (macro?) interaction does the trick (e.g., a water droplet or a falling leaf colliding with the particle would constitute an observation). But this still doesn’t clean up the measurement problem, at least not to the satisfaction of proponents of many-worlds interpretations, such as Carroll and Tegmark. (Albert, again, does not think the many-worlds interpretation works.)
Now for some questions.
Does the vagueness surrounding the space of (im)possibility pose to many-worlds interpretations anything like the sort of vagueness described above in the Copenhagen picture?
If so, what sorts of problems? Are they just considered probabilistic, if they’re considered at all?
3.5 For the Future: Probability
Finally, there are special concerns here about what probability means in the many-worlds context, in which everything that can happen does happen. I recently noted this elsewhere, so I’ll just quote myself:
See also the following conversation between physicist Sean Carroll and philosopher David Albert on the difficulties of making sense of probability under a many-worlds interpretation of quantum mechanics, one difficulty therein being that it seems to threaten the surprisingness of seeing the same result over many trials. Something Albert goes so far as to refer to as astonishing at one point, and which he seems to think signals deeper theoretical challenges to the many-worlds interpretation. For example, getting the same result over and over again—Albert’s specific example is finding a million x-spin up particles z-spin up, when there’s a 1/2 probability for the z-spin being up or down—would seem to count as evidence against quantum mechanics, but the many-worlds interpretation predicts with 100% certainty that there will be such a world in which this happens. They agree that such a thing would “amount to a gigantic surprise,” and, moreover, they seem to agree that surprise is justified. How to justify that surprise seems to be their point of disagreement.
Carroll seems to think the surprise is accounted for by reflecting back on the rarity of one’s situation, while Albert seems to find this unintelligible, along the lines of asking a person about to be perfectly cloned if they expect—or would be surprised—to end up on the left or on the right post-cloning. There is more nuance here than I’m doing just to, so I recommend watching the video to see for yourself (at YouTube; starts at 40:42): Science Saturday: Problems in Quantum Mechanics | Sean Carroll & David Albert (recorded July 2018). I’m particularly fascinated by Albert’s comment that attempting to make sense of probability in many-worlds scenarios is pushing us to think more deeply about what probability is. I’ve begun collecting papers to get a better handle on what he means. (From my post “Why 92 Heads in a Row Is Not Not Surprising.”)
I share Albert’s concerns and have many other questions about probability in this context, but will just give a sample here. I’ve heard Carroll say, and Tegmark imply, that anything with non-zero probability will happen somewhere. But what about things with zero probability that actually are possible, as with continuous uniform distributions? For example, do infinitely many worlds branch out if you ask a random number generator to produce an irrational number between 0 and 1? I imagine, rather, that it is only the number of worlds corresponding to the computer’s deterministically finite granularity. But what about a human asked to name a number between 0 and 1. Not all possibilities are open, but which are? Or are we to take it there’s some world in which I’d start rattling off numbers—0.109586584022940511….—and continue the project until I die? There must be some sample space of possible outcomes, confirmed by the laws of physics in the production of literal worlds in which each outcome occurs. And if we understand the laws of physics, we should, in principle, be able to figure this out.
In a related question, when I draw a dot on a piece of paper, does this generate infinitely many worlds corresponding to the theoretically infinitely many points the dot could have covered, or do spatial gaps in the paper and ink somehow prohibit this? What about as I move the paper around in space?
Similarly, are there infinitely many versions of me at heights within some range? If so, what are the boundaries and how do we define those boundaries non-arbitrarily? Again, there must be an answer confirmed by the branches. Or is Tegmark correct when he writes that “neither quantum suicide nor quantum immortality actually works, because they depend crucially on something that I don’t think exists in nature: an infinitely divisible mathematical continuum” (Tegmark, 220)? I believe Tegmark is correct, but, even so, the question remains of which of my possible heights are in the sample space and which aren’t?
And with clearly discrete states (e.g., brain states), how are intermediate states parsed, sample-space wise? What sort of reference class is there for a brain halfway rearranged from English- to Hungarian-speaking? Wouldn’t most such transformations result in vegetative brains? How could such a transformation be modeled? It doesn’t happen one brain cell at a time in some organized, dominos-cascading fashion, but how does it happen? I think the question worth asking because it raises the possibility of independence of states such that may lower the probability of its occurring; that is, an intermediate state may result in a kind of brain-death that would simply stop any further useful transformation.
Remember, these transformations have no teleological backdrop—they aren’t about moral salience or organizing a brain so that it now speaks a new language. Halfway between an English-speaking brain and Hungarian-speaking brain isn’t one that remembers half of English and now knows half of Hungarian. The transformations are, rather, just random reconfigurations of particles that happen to yield, for example, the ability to speak Hungarian. But the results would mostly, I take it, be more like what you get when you shake up a can of red and blue paint: you increase entropy in the can, rather than decrease it.
At any rate, there is, in principle, a sample space that could be viewed by an outside deserver. That space just is the branches that actually occur. My principal perplexity—which connects questions about probability and possibility—is in the modeling of that sample space from within a given branch.
I’ll hold onto these and other questions until I’ve done some background reading, starting with some more in-depth articles I’ve collected by Albert and others. Also a good starting place is the relevant section in the Stanford Encyclopedia of Philosophy entry, “Many-Worlds Interpretation of Quantum Mechanics”: Probability in the MWI.
In the meanwhile, perhaps some friendly and generous expert with some moments to spare will indulge my questions (and won’t hold back on the math, if math is required—this is a math-friendly zone).
My hunch is that the answer to most of these questions will be, “The inability of humans to figure out which worlds are possible has no bearing on the fact that worlds branch.” If so, can the Copenhagen proponent similarly say that “The inability of humans to figure out what counts as an observation has no bearing on the fact that the wave function collapses”? Where does this comparison go wrong?
Lastly, I suppose I should note explicitly that I neither believe nor accept the many-worlds interpretation. Given what I’ve seen of the public discussion surrounding it (e.g., Albert’s rejection, and he’s not alone), and my inability to make sense of talk like, “everything that can possibly happen does happen, somewhere” (what does “possible” mean and how to relate that to probability?), the many-worlds thesis is psychologically impossible for me to believe or accept or even to be agnostic about. Maybe learning the equations would convince me. Or maybe not. Not all those who know the equations are converts. Maybe I’ll be moved by another interpretation about which I’ve recently heard glowing endorsements: relational quantum mechanics, whose principle originator, Carlo Rovelli, has also appeared on Carroll’s podcast.
But my aim here hasn’t been to challenge the many-worlds interpretation. Rather, it has been to pose some naive questions about quantum mechanics as someone who has not rigorously studied the topic. Nor do I intend to, though I am interested in what physicists claim to be true about the world, particularly for what it means about things like free will/moral responsibility/desert, probability, and other areas that I do aspire to study rigorously. Just as I’m interested in the informal and informal models for representing the world being developed within other disciplines, sub-disciplines, sub-sub-disciplines and so on, and, just as importantly, the degree to which those developing and relying on those models believe the models to be veridical. Also of critical interest are the tools being used to arrive at those models and beliefs, as well as how those things are translated for discussions in non-specialist—including popular and interdisciplinary—contexts.
UDPATE: Since writing the above post, I’ve encountered a clearer expression of Carroll’s reasons for believing in free will, in this 11/5/18 episode of his Mindscape podcast: Alex Rosenberg on Naturalism, History, and Theory of Mind (Episode 21) (also available at YouTube). The discussion begins at 15:10. Carroll makes a clarifying comparison:
My view of free will is that it is exactly as real tables and chairs and baseball and chess. It is something that is not found in the fundamental laws of physics—that if you describe the world at the most reductionistic level it wouldn’t be there. But we have other levels that we describe it on. I can say that my chair is made of atoms, but that doesn’t meant that the chair stops existing.
He says more than this, but even just this snippet helps illuminate his view of free will as something of a genuinely existing, emergent phenomenon that supervenes on lower-level goings-on. I still disagree with him, but can make more sense of this than I could of his other articulations of his view, which struck me as relying on fundamentally “as if it were true” reasoning. My disagreement with Carroll, by the way, is very much in line with Rosenberg’s comments in this discussion.
Enjoy or find this post useful? Please consider pitching in a dollar or three to help me do a better job of populating this website with worthwhile words and music. Let me know what you'd like to see more of while you're at it. Transaction handled by PayPal.
Or click the banner to shop at Amazon (at no extra cost: it just gives me some of what would have gone to Amazon).
- Meinong, A. (1904), appears translated as “The Theory of Objects” in Realism and the Background of Phenomenology (1960), R. Chisholm (ed).