Barang

Notes from the underground

Updated 1 August 2024

Draft

Trying to write something.

Draft

There is surprisingly little discussion of whether teletransportation can be verified (or falsified) empirically. I think the question is more interesting than most people realize..

The issue is whether the person who emerges from the transporter is the same as the person who entered; or just a freshly-created lookalike, the one who entered having died. Empirically speaking, what everyone agrees on is this: we cannot determine the answer “from the outside,” i.e., from a third-person vantage point. It would be futile to observe someone else enter the machine and thereafter “check” if that same person emerged on Mars: there would be no telling if the person who emerged was the same as the one who entered.

The interesting question however is whether we can determine the answer “from the inside,” from the first-person point of view. At risk to your own life, could you not try the machine out for yourself in order to see what happens?

This simple first-person empirical check does appear to work in at least one type of case. If there was such a thing as an afterlife, then, if the machine killed you, you’d find yourself waking up in some fiery place like purgatory (say), and you’d realize (let’s presume) that you were dead. Thus:

Captain Kirk, so the story goes, disintegrates at place p and reassembles at place p*. But perhaps, instead, he dies at p and a doppelgänger emerges at p*. What is the difference? One way of illustrating the difference is to suppose there is an afterlife: a heaven, or hell, increasingly supplemented by yet more Captain Kirks all cursing the day they ever stepped into the molecular disintegrator.

This is Kathleen Wilkes, lamenting the “inconclusive nature” of science fantasy. But her words are to our purpose too: if the transporter killed you, then you could discover this by trying it out, assuming there was such a thing as an afterlife. Of course, if there was no afterlife, then you’d learn nothing: you’d just lose consciousness and it’d be over. But the potential for falsification is clearly there since we cannot in general rule out the possibility of an afterlife.

If falsification of teletransportation is thus possible, what of verification? What happens if the machine worked and really did “send” you to Mars? You’d press the button and find yourself waking up on Mars. Would this assure you in the same way that teletransportation was real and that you really had been transported to Mars? Take Parfit’s story above:

I press the button. As predicted, I lose and seem at once to regain consciousness, but in a different cubicle. Examining my new body, I find no change at all. Even the cut on my upper lip, from this morning’s shave, is still there.

Would this satisfy Parfit himself, if no one else, that the machine had worked? Would it constitute an empirical verification of teletransportation?

As we know, the matter is not that simple, because of the following obvious snag. Even if the machine did work, and you did find yourself waking up on Mars, you’d be unable to tell (upon waking) whether you were the person who entered the machine, or whether that person died, and you were just a freshly-minted lookalike with false beliefs of having previously entered the machine. After all, anyone—real or copy—who emerged on Mars would “remember” having earlier entered the machine on Earth. So you’d be none the wiser as to whether you were “real” or “copy,” and thus none the wiser as to whether the machine had really worked.

For example, in Dan Dennett’s story, a woman on Mars with a broken spaceship uses a teletransporter to get back to Earth. Upon reuniting with her daughter Sarah, it “hits her”:

Am I, really, the same person who kissed this little girl good-bye three years ago? Am I this eight-year-old child’s mother or am I actually a brand-new human being, only several hours old, in spite of my memories—or apparent memories—of days and years before that? Did this child’s mother recently die on Mars, dismantled and destroyed in the chamber of a Teleclone Mark IV?

Did I die on Mars? No, certainly I did not die on Mars, since I am alive on Earth. Perhaps, though, someone died on Mars—Sarah’s mother. Then I am not Sarah’s mother. But I must be! The whole point of getting into the Teleclone was to return home to my family! But I keep forgetting; maybe I never got into that Teleclone on Mars. Maybe that was someone else—if it ever happened at all. Is that infernal machine a teleporter—a mode of transportation—or, as the brand name suggests, a sort of murdering twinmaker?

So it seems that, if the machine worked, you could not discover this even by trying it out for yourself. This may be why the question is seldom discussed of whether teletransportation is subject to empirical test. At least where verification of the phenomenon is concerned—and this is arguably the more important case—the answer may be thought to be obviously no.

⚹

This may be thought to settle the matter but, in fact, I will suggest now that the foregoing sceptical view is confused, and ultimately incorrect. If the teletransporter works, I believe that you can verify this by trying the machine out for yourself.

The sceptic claims that if you entered the machine, pressed the button, and found yourself waking up on Mars, you’d be unable to tell (upon waking) whether your memory of having entered the machine was genuine or false. You wouldn’t know if you were the person who originally entered the machine, or just a freshly-minted copy with false memories of having done so. So you’d be unable to tell if the machine had worked. (If you happened to be the copy, the machine would in fact have failed.)

The alleged threat here is that of “being the copy.” As you wake up on Mars, the thought that you might be the copy is supposed to give you pause. But for the copy, you could just go by your memory of having entered the machine and declare that the machine had worked. But—wait—you might be the copy! Then your memory would be false and the machine would in fact have failed! You’d be in danger of making a wrong judgement here if you went by your “memory.”

This reasoning is confused. Consider first that it would not be possible for you to enter the machine and then wake up on Mars in the body of your copy. This could not happen since your copy and you are two distinct individuals. So you don’t have to worry about the touted possibility of “being the copy”: there is no such possibility. Put another way, there is no danger of you—the person originally entering the machine—waking up on Mars with a false memory of having entered the machine. Nor is there any danger of you waking up on Mars and judging wrongly that the machine had worked when, in fact, it had failed. These things are impossible within the context of our story.

It will be objected that the touted possibility (“You might be the copy”) is not that of you entering the machine and then waking up on Mars in the body of your copy, which would of course be impossible, but that of the person now waking up on Mars—the one being addressed (“you”)—being the copy. This latter is certainly possible since if the machine should fail,

the copy would wake up on Mars, and this might be who the sceptic happens now to be addressing.

But this doesn’t help matters. If you were the copy in this sense, there’d still be no “danger” of making a wrong judgement. The reason is that your false memory (and attendant danger) would now be irrelevant: it wouldn’t in this case matter if you went by your “memory” and judged wrongly that the machine had worked. It wouldn’t matter because, relative to our purposes, nothing turns on whether

the copy is able to accurately divine his true situation.

Recall that we were trying to ascertain whether you, the person who originally entered the machine, would be able to tell that the machine had worked, should you find yourself waking up on Mars. The one being interrogated here is you: would you be able to tell? Now, should the machine fail, you’d die, and your copy would awaken on Mars, but notice that we have no essential interest in whether your copy would (in that event) be able to tell that the machine had failed. In a different context, we might care about this—see shortly below—but if we are trying simply to establish whether teletransportation can be verified in the first person, i.e., by you, then it doesn’t matter if your copy, should he come into existence, gets it wrong. It doesn’t matter what your copy judges; he is not being interrogated at all. What matters is that you, the person who entered the machine, get it right.

For comparison, consider a different case where both you and your copy are equally subject to interrogation, so that it matters that he gets it right too. Consider the scenario once mentioned by Russell wherein the whole world suddenly springs into being with everyone having false memories of an

unreal past. How do you know that this did not just happen?

Here, as in our case, either you are who you remember yourself to be (the real you) or else a freshly-minted individual with false memories of an unreal past (the copy). But notice here that either way, you are being interrogated by the sceptic: did the world just spring into being? The sceptic claims that “you” cannot reliably tell. In this case, it seems to me that the sceptic has a point because the threat of “being the copy” is now genuine. If you just went by your memory, you might get it wrong, because you might be the copy. This is exactly as before but, this time, getting it wrong qua being the copy is just as bad as getting it

wrong qua being real. This is clear from the context and it makes all the difference.

It’s as though God had randomly subjected either the real person or the copy to interrogation, creating the appropriate world for the purpose. It should be clear that this randomly chosen individual (“you”) could do no better than chance in divining whether or not the world had just sprung into being. God might demonstrate this by conducting repeated trials: in the long run, “you” would be able to divine the correct answer only half the time, at best.

Likewise for Dennett’s protagonist above, who wonders whether she’s real or copy, whether the machine had worked or failed. This case is close to ours, but the difference is that Dennett’s protagonist means to be interrogating herself regardless of whether she’s real or copy. (She wants to know either way.) This is again clear from the context, and scepticism now has a foothold for the same reason as before: getting it wrong qua being the copy would be just as bad as getting it wrong qua being real. So, in this case too, it seems to me that the touted scepticism is warranted: Dennett’s protagonist is unable to tell whether she’s real or copy. (God might again demonstrate this by means of repeated trials.)

Neither Russell’s nor Dennett’s case should be confused with ours. Our case involves seeing if you can tell whether the teletransporter works by trying it out for yourself. So, essentially, only you, the person who enters the machine, stand to be interrogated over whether or not the machine had worked. Your copy, should he materialize, is essentially irrelevant.

Indeed, your copy here is something of an interloper, who pilfers your memories and runs into trouble as a result. If he wakes up on Mars and believes that the machine had worked, he may—to his undoing—try to schedule a “return” trip, etc. None of this has anything to do with you: you could never get into such trouble. Put another way, should God conduct repeated trials to see whether you could tell that the machine had worked (whenever it did), God would set aside worlds where the person waking up on Mars was your copy, and focus on worlds where this person was you, the one who entered the machine. You’d then be interrogated on whether or not the machine had worked. Could you reliably tell? Well, if you always went by your memory, you’d always get it right.

So we’re back where we started, but armed with the realization that your copy is a red herring. If you entered the machine, pressed the button, and found yourself waking up on Mars, you may simply go by your apparent memory, judge that the machine had worked, and not worry about “being the copy.” Your judgement would be bound to be correct, even if there was some sense—the (irrelevant) Russell-Dennett one—in which “you” couldn’t tell whether or not you were the copy. It’d be bound to be correct just because it was you, the person who originally entered the machine, who was making the judgement. Even God, by repeated test, would have to acknowledge that you could reliably tell that the machine had worked. So should you so much as seem to find yourself waking up on Mars, you may be completely confident—indeed certain—that the machine had worked. You’d know that the machine had worked. “That’s what your copy would think!” an objector persists. But, no, this is the same mistake again: overlooking that, relative to our purposes, what your copy thinks or judges is immaterial.

My conclusion, therefore, is that if the teletransporter works, you can discover this by trying it out for yourself. You’d find yourself waking up on Mars and you’d know that you had survived and that the thing had worked. The objection, “Not so fast! You might be the copy!” is just confused. In one sense, as pointed out, you couldn’t possibly be the copy; in another, it would be irrelevant if “you” were, because it doesn’t matter if the copy gets it wrong. Notice also that, in this latter case, where the copy (“you”) wakes up on Mars and judges wrongly that the machine had worked, the real you might well find yourself waking up in purgatory (as before), realizing correctly that the machine had failed. This makes it even clearer that the misadventures of the copy are just irrelevant: the real you would always get it right. The sceptical objection seems to work only because it confuses our context with various other ones like Russell’s and Dennett’s above in which this sort of objection would work.

⚹

Does a teletransporter really work? The most convincing demonstration of this, it seems to me, would be a first-person empirical one. Just try out the machine and see. If it works, you’d find out. The natural thought that such a demonstration would be impossible is, I think, wrong.

Let me now show how this sort of empirical demonstration can be used to decide another vexing problem in this area, viz., the question of what happens in fission.

###############

------------

This may be thought to settle the matter but, in fact, I will suggest now that the foregoing sceptical view is more confused than correct. I believe that the sceptic has overlooked an important feature of our predicament.

The sceptic claims that if you entered the machine, pressed the button, and found yourself waking up on Mars, you’d be unable to tell (upon waking) whether your memory of having entered the machine was genuine or false. You wouldn’t know if you were the person who originally entered the machine, or just a freshly-minted copy with false memories of having done so. So you’d be unable to tell if the machine had worked. If you happened to be the copy, the machine would in fact have failed.

The alleged threat here is that of “being the copy.” As you wake up on Mars, the thought that you might be the copy is supposed to give you pause. But for the copy, you could just go by your memory of having entered the machine and declare that the machine had worked. But—wait—you might be the copy! Then your memory would be false and the machine would in fact have failed! You’d be in danger of making a wrong judgement here if you just went by your “memory.”

This threat of “being the copy” seems plausible at first sight, but I believe it is ultimately illusory. Note first that the sceptic cannot mean that you, the person who originally entered the machine, might be the copy. This would be impossible since you and your copy are distinct individuals. What the sceptic means rather is that the person now waking up on Mars—the one being addressed (“you”)—might be the copy. But this threat is illusory too, though in a more subtle way. If you were the copy in this sense, then it wouldn’t actually matter if you went by your memory and ended up judging wrongly that the machine had worked, because the ability of your copy to descry his true situation is nowhere under scrutiny here. This is what the sceptic seems to me to have overlooked.

Recall that we were trying to ascertain whether you, the person who originally entered the machine, would be able to tell that the machine had worked, should you find yourself waking up on Mars. The challenge here is being issued to you: can you tell? Now, should the machine fail, you’d die, and your copy would awaken on Mars, but notice that we have no essential interest in whether your copy would (in that event) be able to tell that the machine had failed. In a different context, we might care about this—see shortly below—but if we are trying simply to establish whether teletransportation can be verified in the first person, i.e., by you, then it doesn’t matter if your copy, should he come into existence, gets it wrong. No challenge is being issued to your copy at all. What matters is that you, the person who entered the machine, get it right.

To see the difference, consider a context where both you and your copy are “challenged,” so that it matters that he gets it right too. Consider the scenario once mentioned by Russell wherein the whole world suddenly springs into being with everyone having false memories of an unreal past. How do you know that this did not just happen?

Here, as in our case, either you are who you remember yourself to be (the real you) or else a freshly-minted individual with false memories of an unreal past (the copy). But notice here that either way, you are being challenged by the sceptic: did the world just spring into being? The sceptic claims that “you” cannot reliably tell which. In this case, I am inclined to side with the sceptic because the threat of being the copy is now genuine. If you just went by your memory, you might get it wrong, because you might be the copy. This is exactly as before but, this time, getting it wrong qua being the copy

is just as bad as getting it wrong qua being real.

It is as though God were to pick either the real person or the copy at random to be thus challenged, creating the appropriate world for the purpose. This randomly chosen person (“you”) could clearly do no better than chance in divining whether or not the world had just sprung into being. God might demonstrate this by conducting repeated trials: in the long run, “you” would be able to divine the correct answer only half the time, at best.

In our case, in contrast, only the person who originally entered the machine is being tested, or challenged. The copy, should he come into existence, is irrelevant. Should God conduct repeated trials to see whether the person waking up on Mars could reliably tell whether or not the machine had worked, he’d ignore worlds where the person waking up on Mars was the copy, and focus on worlds where this person was the one who originally entered the machine. This person would then be challenged to say whether or not the machine had worked, the question being whether he can reliably tell.

And it should now be clear that if this person always went by his apparent memory and averred that the machine had worked, he’d always get it right! Even God would have to admit that he could reliably tell. In other words, if you entered the machine, pressed the button, and found yourself waking up on Mars, you may simply go by your apparent memory, judge that the machine had worked, and not worry about “being the copy.” Your judgement would be bound to be correct, even if there was some sense (see below) in which you couldn’t tell whether or not you were the copy. It’d be correct just because it was you who was making the judgement. Should your copy find himself waking up on Mars and attempt the corresponding judgement, he’d get it wrong, but not you. You could never get this judgement wrong.

Put another way, there is no danger of you waking up on Mars and judging wrongly that the machine had worked when, in fact, it had failed. (There is no danger of you waking up in your copy’s body.) And so you really should have no hesitation in judging that the machine had worked, should you so much as appear to find yourself waking up on Mars. Your judgement here could never be wrong, so you may confidently judge that you had survived the trip and that the machine had worked; indeed, you may be certain of it. In this sense, you’d know that the machine had worked, should you appear to find yourself waking up on Mars.

It may be thought that this is not a real case of knowing that the machine had worked. Waking up on Mars, you wouldn’t really know whether you were real or copy, and thus whether or not the machine had worked. What you’d know is that if you were copy, it wouldn’t matter what you judged (since, as before, the copy is not being tested), so you might as well judge that you were real and that the machine had worked. But you wouldn’t really know whether this was true. It’d just be a “safe” judgement to make, in that you could never be faulted, either way, for judging as much.

This is the same error as before, only in a different form. There is doubtless a sense in which “you” wouldn’t know whether you were real or copy—just as there was a sense in Russell’s case in which “you” wouldn’t know whether or not the universe had just sprung into being—but the unknowing person (“you”) here is the randomly chosen person previously mentioned: either you or your copy, picked at random. This person wouldn’t reliably be able to say whether they were real or copy, or whether the machine had worked or failed, or whether their memories were genuine or false, etc.

But you, on the other hand, the person who entered the machine, would know that you were real, in the same way that you’d know that the machine had worked. You could never get this wrong if you went simply by your memory: if God were to test you, he’d concur. It may also help to see it this way. Suppose that “you” were actually the copy, judging wrongly that the machine had worked. Then the real you would be somewhere else—in purgatory, let us assume—finding out that the machine had failed. This should make it clear that you really don’t have to worry about “being the copy”: in the only sense that matters, you could never be the copy: the two of you are different individuals.

You don’t have to worry about being the copy because you could never end up in your copy’s body. “But that’s exactly what your copy would think!,” you retort. But this again fails to see that what your copy would think is irrelevant.

Your copy is essentially an interloper who pilfers your memories and gets into all sorts of trouble because of that. For example, if he believes that the machine had worked, he may—to his downfall—try to schedule a “return” trip, etc. None of this has anything to do with you: you could never get into such trouble, nor are you obliged to protect your copy from harm.

It may be objected that—wait—you might turn out to be the copy, in which case your “confident judgement” would be badly wrong. Surely you’d need to rule out that you’re the copy in order to be confident that the machine had worked—but how could you do that? This is the same error again.

Imagine for a moment that there was no complication of any doppelgänger waking up on Mars, so that if the machine killed you, no one would wake up on Mars. Then, if you entered the machine, pressed the button, and found yourself waking up on Mars, you’d know at once that the machine had worked. I take it that this needs no explanation. Now restore the complication of the doppelgänger: should the machine kill you, a copy of you would wake up on Mars with false memories of having entered the machine, and so on. The sceptic contends that, this time, if you entered the machine, pressed the button, and found yourself waking up on Mars, you’d be unable to tell if you were real or copy, and thus be unable to tell if the machine had worked. (If you happened to be the copy, the machine would in fact have failed.)

This contention of the sceptic seems rather doubtful to me, even as it manages to sound both natural and familiar. It embodies a kind of thinking that works well in certain philosophical contexts, which is what lends it an air of plausibility. A little thought, however, will show that it doesn’t quite work in our case.

A context where it does work is the scenario once mentioned by Russell in which the entire world suddenly springs into being with everyone having false memories of an unreal past. How do you know that this did not just happen? In this case, as in ours, either you are who you remember yourself to be—the real you—or else a freshly-minted individual with false memories of an unreal past—the copy. In this case, I concur with the sceptic that there’s no saying which one. You simply can’t tell. Derivatively, you can’t tell whether or not the universe has just sprung into being.

But this case of Russell’s differs from ours in one crucial respect. In Russell’s case, it doesn’t matter if you—the subject of the epistemic predicament—happen to be real or copy: either way, you are being “challenged” to say whether or not the universe has just sprung into being. And the sceptic is claiming that you cannot reliably tell which. More vividly, imagine God randomly picking either the real person or the copy to be challenged in this fashion, creating the appropriate world for the purpose. The claim is then that this randomly chosen person (“you”) can do no better than chance in divining whether he or she was real or copy. God might demonstrate this by conducting repeated trials: in the long run, “you” would be seen to divine the correct answer only half the time, at best.

In our case, however, things are different. Only the real you (and not the copy) is being challenged to say whether or not the machine had worked. After all, we were only trying to ascertain whether you—the person who entered the machine—would be able to tell that the machine had worked, should you find yourself waking up on Mars. This challenge is addressed to you: can you tell? Now, should the machine fail, you’d die, and your copy would awaken on Mars, but notice that there is no concern here with whether your copy would (in that event) be able to tell that the machine had failed. In a different context, we might be concerned with thist—see shortly below—but it doesn’t exist for our purposes. No challenge is being issued to your copy at all.

In divine terms, imagine God creating a world in which teletransportation works and then challenging someone who just took a trip in the machine to say whether or not

This makes a difference because you may now act as if you were in the “uncomplicated” case, where the potential existence of your copy was not an issue. Notice, in particular, that if you found yourself waking up on Mars, you’d be bound to make a correct judgement if you judged that the machine had worked: you could never get this judgement wrong. So just ignore the complication of your copy, go by the appearances, and you’ll be fine. Put another way, there is no danger of you waking up on Mars and judging wrongly that the machine had worked when, in fact, it had failed. The sceptic wants to protest here that “Hold on, you might be the copy,” implying that your judgement that the machine had worked might be wrong. But this doesn’t mean that you might wake up in your copy’s body, which is impossible. It means only that the person now waking up on Mars and judging that the machine had worked might be the copy. And the reply is that: yes, but it doesn’t matter if the copy’s judgement is wrong. This is what the sceptic seems to me to have overlooked. Our case cannot be compared with Russell’s.

Indeed, in our case, your copy is essentially an interloper who pilfers your memories and gets into all sorts of trouble because of that. For example, if he believes that the machine had worked, he may—to his downfall—try to schedule a “return” trip, etc. None of this has anything to do with you: you could never get into such trouble, nor are you obliged to protect your copy from harm.

Simply focus on your well-being, ignore the potential existence of your copy, and you will be fine. Suppose, for example, that you enter the machine and find yourself waking up on Mars. Then, exactly as in the uncomplicated way, you’d be able to tell that the machine had worked. The sceptic might protest that—hold on—you might be the copy, in which case the machine would have failed; how can you tell? But notice again that this is your copy’s problem, not yours. After all, the sceptic doesn’t mean that you—the person who entered the machine—might be the copy, which would be impossible, but that the person now waking up on Mars might be the copy. But this has nothing to do with you. You could never wake up on Mars and believe wrongly that the machine had worked. If you awaken on Mars and judge that the machine had worked, you’d be bound to be right. As before, simply focus on your well-being, ignore the potential existence of your copy, and you will be fine: the alleged complication of your copy is just a red herring.

Your concern is simply to establish that the machine had worked should you find yourself waking up on Mars. You can’t be concerned also with doing this in such a way that your copy be able to establish that the machine had failed should he find himself waking up on Mars! You think what. Got so much time ah?

If God were to recreate the whole situation repeatedly in such a way that the machine worked each time, then the person “challenged” to say whether or not the machine had worked would always be you, and never your copy. And you’d always be right if you said that the machine worked.

The sceptic claims that if you entered the machine, pressed the button, and found yourself waking up on Mars, you’d be unable to tell (upon waking) whether your memory of entering the machine was genuine or false. You’d be unable to tell if you were the person who originally entered the machine, or just a freshly-minted copy with false memories of having done so. As such, you wouldn’t know if the machine had worked. (If you happened to be the copy, the machine would in fact have failed.)

This reasoning is extremely natural. There is indeed a sense in which, if you woke up on Mars, you wouldn’t know whether the propositions mentioned above were true—e.g., you wouldn’t know whether the machine had worked—and the reasoning above captures this philosophically familiar sceptical sense well. But what the sceptic overlooks is that, in the nature of our case, there is another sense in which you would know that those propositions were true—e.g., you would know that the machine had worked. (The question then arises of which sense is pertinent for our purposes.)

To see this other sense, notice that if you entered the machine, pressed the button, and found yourself waking up on Mars, you’d be bound to make a correct judgement if you judged that the machine had worked. Your judgement would be correct even if—as the sceptic contends—you couldn’t tell whether the machine had worked. It’d be correct just because it was you who was making the judgement. Should your copy find himself waking up on Mars and attempt the corresponding judgement, he’d get it wrong, but not you. You could never get this judgement wrong.

Indeed, it should be plain to you, as you press the button, that there is no danger of you waking up on Mars and judging wrongly that the machine had worked when, in fact, it had failed. (There is no danger of you waking up in your copy’s body.) The sceptic would have you believe that if you pressed the button and found yourself waking up on Mars, you wouldn’t really know whether the machine had worked: should you judge as much, you could well be wrong. Now, in one sense, as I have conceded, this is correct, and I will say more about this sense below. But notice also the other sense in which there is absolutely no danger here of you being wrong. As just explained, should you find yourself waking up on Mars, you couldn’t be wrong in judging that the machine had worked. (Only your copy could get this judgement wrong.)

From this point of view, you should have no hesitation in judging that the machine had worked, should you so much as appear to find yourself waking up on Mars. Your judgement here could never be wrong, so you may confidently judge that you had survived the trip and that the machine had worked; indeed, you may be certain of it. In this sense, you’d know that the machine had worked, should you appear to find yourself waking up on Mars.

It may be objected that—wait—you might turn out to be the copy, in which case your “confident judgement” would be badly wrong. Surely you’d need to rule out that you’re the copy in order to be confident that the machine had worked—but how could you do that? This objection is natural but it’s just another reflection of the ambiguity currently being discussed. I mentioned above that there is a sense in which, upon waking up on Mars, you wouldn’t know whether the machine had worked, and the sceptic is just pointing this out in a different way: “You could be the copy” is just another way of saying, “You don’t know whether the machine worked.”

It doesn’t matter if “you” turn out to be the copy, judging wrongly that the machine had worked, because the ability of your copy to descry his situation is not under scrutiny here. Recall that we were trying to ascertain whether you—the person who entered the machine—would be able to tell that the machine had worked, should you find yourself waking up on Mars. Should the machine fail, you’d die, and your copy would find himself awakening on Mars, but there was never a concern with whether he would (in that event) be able to tell that the machine had failed. What your copy thinks or judges, should he come into being, is irrelevant for our purposes. What matters is only whether you are able to correctly judge that the machine had worked, should you find yourself waking up on Mars.

Your copy, in our context, is essentially an interloper who pilfers your memories and gets into all sorts of trouble because of that. But none of this has to do with you, or your ability to tell that the machine had worked, should you find yourself waking up on Mars.

I find it useful to think of this in terms of an epistemological experiment conducted by God. He creates a world where the machine works and watches you enter the machine and press the button. You thereafter wake up on Mars, whereupon God asks if you can tell whether the machine had worked. Just say, “Yes,” and you will always be right. Even God must acknowledge that you know the answer. (You always get it right.)

To see this, consider what happens if, upon waking up on Mars above, you make the “naive” judgement that the machine had worked, e.g., you just go by the appearances. (“This thing works!” you aver.) Well, the sceptic would retort as before that—hold on a second—your judgement may be incorrect: perhaps the machine actually failed—the original subject having died—and you’re just the copy with false memories of having entered the machine. The sceptic draws attention here to the possibility of your judgement being incorrect, but notice that the person at risk of making this incorrect judgement is not you, the one who entered the machine, but your copy, the one with false memories.

Note. This is obscured by the formulation, “Your judgement may be incorrect,” or “Your memories may be false,” which may suggest that only one person (you) is involved, whereas in fact there are two (you and your copy). Contrast this with a case which genuinely involves one person: imagine having a neurological syndrome that saddles you with false memories from time to time. Here, too, we might have occasion to say, “Are you sure that happened? Your judgement may be incorrect. Your memory may be false.” In this case, it really is your memory that is in question, whereas, in ours, there’s nothing wrong with your memory and it’s only your copy’s memory that is false. (Likewise for your judgement.)

But, now, why should it matter if your copy makes an incorrect judgement as to whether or not the machine had worked?—specifically, if he judges that the machine had worked when it had actually failed? The sceptic may have forgotten what we were trying to do.

Recall that we were trying to ascertain whether you—the person who entered the machine—would be able to tell that the machine had worked, should you find yourself waking up on Mars. Should the machine fail, you’d die, and your copy would find himself awakening on Mars, but there was never a concern with whether he would (in that event) be able to tell that the machine had failed. What your copy thinks or judges, should he come into being, is irrelevant for our purposes.

Note. In a different context, what your copy thinks or judges may matter: see further below.

What matters is only whether you are able to correctly judge that the machine had worked, should you find yourself waking up on Mars. And notice now that, in a way, nothing could be easier. After all, if you pressed the button and found yourself waking up on Mars, you’d be bound to make a correct judgement if you judged that the machine had worked. It would be correct just because it was you who was making the judgement. Your copy would be wrong if he attempted the parallel judgement, but not you. You could never get this judgement wrong.

So there as you stand with your finger on the button, it should be plain to you that nothing can go wrong: there is no danger of you waking up on Mars and judging wrongly that the machine had worked when, in fact, it had failed. (There is no danger of you waking up in your copy’s body.) The sceptic contends that if you pressed the button and found yourself waking up on Mars, you wouldn’t really know whether the machine had worked: should you be inclined to think so, you could well be wrong. But there is no such danger of being wrong. (The only person who could be wrong here is your copy.)

You should have no hesitation therefore in judging that the machine had worked should you find yourself waking up on Mars. You know in advance that you could never get this judgement wrong, so you may confidently judge that the machine had worked.

The sceptic may retort that you could not reasonably make this judgement unless you knew (upon thus waking) that you were the one who entered the machine. Only the person who entered the machine would get this judgement right; the copy would get it wrong. So you need to rule out that you’re the copy before you can “confidently” make this judgement. But then we’re back to square one. How do you rule out that you’re the copy?

Unfortunately, this is the same confusion. When the sceptic says that you could be the copy, he doesn’t mean you—the person who entered the machine—could be the copy. Obviously, the person who entered the machine couldn’t be the copy: ex hypothesi, they are different persons. The sceptic means rather that it could be your copy who was now waking up on Mars and attempting “confidently” (though erroneously) to judge that the machine had worked. Now, this is certainly true, but we have already seen that it is irrelevant since your copy is not you. It wouldn’t matter if your copy incorrectly judged that the machine had worked, when in fact it had failed. So you don’t have to “rule out” that you’re the copy, any more than you have to “rule out” that your memories are false, or that your judgement that the machine had worked is incorrect. Were any of these possibilities to come true (and they are all of a piece), they would be true of your copy and may safely be ignored.

An equally confused retort would be that, upon waking up on Mars, you’d just be saying that the machine had worked without really knowing that it had. All you’d really know is that your words would be true if you were real, and irrelevant if you were copy. But you wouldn’t really know if you were real or copy.

This is the same confusion in a different form. There is admittedly a sense in which “you” wouldn’t know if you were real or copy, but it is irrelevant in the same way. In this same sense, “you” wouldn’t know that the machine had worked; “you” could be wrong if you judged that it did; and so on. In each of these formulations, the pronoun “you” potentially refers to either you or your copy, as though one of you were chosen at random and put to the test. There are sceptical contexts in which this formulation would be relevant and natural.

For instance, consider the scenario once mentioned by Russell in which the entire world suddenly springs into being with everyone having false memories of an unreal past. Can you rule out that this did not just happen? Like our case, either you are who you remember yourself to be—“the real you”—or else a freshly-minted individual with false memories of an unreal past—“the copy”. And, in this case, I agree that there appears to be no saying which one: you just don’t know. Here, it doesn’t matter if you—the subject of the epistemic predicament—happen to be real or copy. Should you hazard a guess as to whether the universe just sprung into being, and you happen to get it wrong, it doesn’t matter if you get it wrong qua being real or copy: it would be just as bad either way. The real and the copy are on a par in this respect: both of them are under the same epistemic obligation to get their “guesses” right. This just has to do with the context in which Russell raises his sceptical scenario. You get no special treatment if you happen to be real, as opposed to copy.

An incorrect judgement on your copy’s part would be your copy’s problem, not yours. For example, if, on the basis of this incorrect judgement, your copy schedules a “return trip,” that would be his undoing, not yours.

What all of this means is that, if you enter the machine, press the button, and find yourself waking up on Mars, the skeptic has nothing on you if you judged that the machine had worked. Your copy might get this judgment wrong should he attempt the parallel judgement upon waking up on Mars, but not you. Indeed, it should be clear anyway that you could never get this judgment wrong. It would be correct solely in virtue of the fact that it was you who was making the judgement.

So there as you stand with your finger on the button, it should be evident to you that nothing can go wrong: there is absolutely no danger of you waking up on Mars and wrongly judging that the machine had worked when, in fact, it had failed. (There is no danger of you waking up in your copy’s body.) The sceptic would have you believe that, if you found yourself waking up on Mars, you wouldn’t really know whether the machine had worked: should you be inclined to think so, you could very well be wrong. But there is no such danger of being wrong!

It is important to see that there are two distinct people here, one with a genuine memory (you), the other with a false memory (your copy), rather than a single person whose memory may either be genuine or false, as might be inadvertently suggested by the ambiguous formulation, “Your memory might be false.”

In the case with one person, the sceptic’s retort would indeed have a point. Imagine having a neurological syndrome that saddles you with false memories from time to time. Then we might on occasion say to you, “Did that really happen? Your memory might be false.” Here, your memory would genuinely be in question, whereas, in our case, your memory is beyond reproach and it is only your copy’s memory that is false. As mentioned, our case revolves around two people, one of whom remembers correctly, the other wrongly, whereas the neurological case revolves around a single person who sometimes remembers correctly, sometimes wrongly. These cases should not be confused.

Likewise, “Your memory of entering the machine might be false,” doesn’t mean that the real you might now be waking up on Mars with a false memory of entering the machine—which again wouldn’t be possible in the context of our story. The sceptic means rather that it might be the copy (with said false memory) who was now waking up on Mars.

The point here is not to confuse our case with a superficially similar one in which your memory really might be in question. Imagine having a neurological syndrome that saddles you with false memories from time to time. Then we might on occasion say to you, “Did that really happen? Your memory might be false.” Here, your memory would genuinely be in question, whereas, in our case, your memory is beyond reproach and it is only your copy’s memory that is false. Our case revolves around two people, one of whom remembers correctly, the other wrongly, whereas the neurological case revolves around a single person who sometimes remembers correctly, sometimes wrongly.

More akin to our case is the scenario once mentioned by Russell in which the entire world suddenly springs into being with everyone having false memories of an unreal past. Can you rule out that this did not just happen? Like our case (and unlike the neurological one), this case revolves around two people: either you are who you remember yourself to be—“the real you”—or else a freshly-minted individual with false memories of an unreal past—“the copy”. And there appears to be no saying which one. Here too we can say, “Your memory might be false,” meaning again, not that the real you might have false memories, but that the person currently in Russell’s predicament might be the copy.

Similarly, when Russell says, “You can’t tell whether or not the universe has just sprung into being,” he means that any opinion you dare to venture here would really be no better than a guess. Thus, should you aver that the universe has been around for a while, then you’d be right only if you were real and wrong if you were copy. Conversely, dare you aver (for whatever reason) that the universe did just spring into being, then you’d be right if you were copy and wrong if you were real. Either way, since you can’t tell whether you are real or copy, you can’t tell whether or not the universe has just sprung into being, and no purpose would be served by guessing.

All of this is exactly as in our case.

But there is a crucial difference between Russell’s case and ours. In Russell’s case, it doesn’t matter if you—the subject of the epistemic predicament—happen to be real or copy. In particular, should you hazard a guess as to whether the universe just sprung into being, and you happen to get it wrong, it doesn’t matter if you get it wrong qua being real or copy: it would be just as bad either way. The real and the copy are on a par in this respect: both of them are under the same epistemic obligation to get their “guesses” right. This just has to do with the context in which Russell raises his sceptical scenario. There is nothing which favours you being either real or copy, so far as Russell’s predicament itself is concerned.

suppose for eg that you judged that the machine had worked. In the nature of the case your copy would judge the same, but he would be wrong.

Let me now say that, in Russell’s case, my sympathies do lie with the sceptic: I agree that you can’t tell whether you are real or copy, i.e., whether or not the universe has just sprung into being.

But there is a crucial difference between Russell’s case and ours. When Russell says, “You can’t tell whether your memories are genuine or false,” the pronoun ‘you’ may refer to either you or your copy, depending on who Russell happens to be addressing. Russell is essentially addressing one of you at random, and his words are true just in case this randomized individual is unable to tell whether his memories are genuine or false. If we use a capitalized ‘YOU’ to denote this randomized individual, then, in Russell’s case, YOU can’t tell whether YOUR memories are genuine or false—e.g., over repeated testing, YOU’D do no better than chance in getting the answer right.

But there is a crucial difference between Russell’s case and ours. Russell’s challenge—“Can you tell whether or not the universe has just sprung into being?”—is not directed to you in particular, as opposed to your copy. It doesn’t matter if the person in Russell’s predicament is you or your copy: the claim (to which I’m sympathetic) is that neither of you can tell whether or not the universe has just sprung into being.

In our case, in contrast, the challenge is understood to be specifically directed to you, the one with genuine memories, the person who enters the teletransporter in an attempt to verify if the machine really works. Our question is whether you can tell that the machine had worked should you find yourself waking up on Mars. It is irrelevant whether your copy can tell that the machine had failed, should he find himself waking up on Mars: it was no part of our brief to care about that. And so the sceptic’s claim, “You can’t tell whether the machine has worked or failed,” proves to be irrelevant. The claim is true if understood in Russell’s sense, as being directed indifferently to either you or your copy. But

To see this clearly, imagine that you find yourself waking up on Mars and judge confidently that the machine must have worked. The sceptic taps you on the shoulder and reminds you that—hold on—you might be the copy, the implication being that the machine would in that case have failed and your confident judgement would be wrong.

The sceptic means in this way to warn you against rushing to judge that the machine had worked should you find yourself waking up on Mars. “You might be the copy,” is the admonition, the point being that, in that case, your (rushed) judgement would be false.

Unfortunately for the sceptic, however, this admonition overlooks that, in the context of our investigation, it doesn’t matter whether the copy makes a correct judgement as to whether the machine had worked or failed. Our concern was only with whether you—the one who entered the machine—would be able to tell that the machine had worked should you find yourself waking up on Mars. There was no concern with whether your copy would be able to tell that the machine had failed should he find himself waking up on Mars. In a different context, this might matter, but not in ours.

For example,

Indeed, as you step into the teletransporter and press the button, you might not care about the fate of your copy at all. Suppose, for instance, that the machine fails, your copy wakes up on Mars, and mistakenly thinks that the machine had worked. This might not bode well for him—e.g., he might schedule “another” trip and die as a result—but the thought of this need not bother you at all since ex hypothesi your copy is not you. At any rate, the sceptic cannot reasonably require that you be able to tell that the machine had worked should you find yourself waking up on Mars, as well as that your copy be able to tell that the machine had failed should he find himself waking up on Mars, since it is only your ability to descry your situation that is under investigation.

So the sceptic’s cry, “You might be the copy,” is neither here nor there; an appropriate reply would be, “If I were the copy, I would be irrelevant.” But, without this cry, the sceptic has no case against you, e.g., if you confidently judged that the machine had worked upon waking up on Mars. Indeed, notice that if you entered the machine, pressed the button, and found yourself waking up on Mars, you’d be bound to make a correct judgement if you judged that the machine had worked. It would be correct solely in virtue of the fact that it was you who was making the judgement. Your copy would be wrong if he attempted the parallel judgement, but not you. You could never get this judgement wrong.

So there as you stand with your finger on the button, it should be evident to you that nothing can go wrong: there is absolutely no danger of you waking up on Mars and wrongly judging that the machine had worked when, in fact, it had failed. (There is no danger of you waking up in your copy’s body.) The sceptic would have you believe that, if you found yourself waking up on Mars, you wouldn’t really know whether the machine had worked: should you be inclined to think so, you could very well be wrong. But there is no such danger of being wrong! Only your copy could get this wrong; but, as before, he is not you.

This looks straightforward at first glance but, to begin to see where it goes wrong, consider an opposing point of view. Consider that if you entered the machine, pressed the button, and found yourself waking up on Mars, you’d be bound to make a correct judgement if you judged that the machine had worked. Your judgement would be correct even if you couldn’t tell whether your memories were genuine or false. It would be correct just because it was you who was making the judgement. Your copy would be wrong if he attempted the parallel judgement, but not you. You could never get this judgement wrong.

The sceptic might retort that you could not reasonably make this judgement unless you knew (upon waking up on Mars) that you were the one who entered the machine. After all, only the person who entered the machine would get this judgement right; the copy would get it wrong. So you need to rule out that you’re the copy before you can “confidently” make this judgement. But then we’re back to square one. How do you rule out that you’re the copy?

Unfortunately, this retort is confused. When the sceptic says that “you” could be the copy, he doesn’t mean you—the person who entered the machine. Obviously, the person who entered the machine couldn’t be the copy: ex hypothesi, they are different persons. What the sceptic means rather is that it could be the copy who is now waking up on Mars and attempting confidently (though erroneously) to judge that the machine had worked. Now, this may well be true, but it is also irrelevant, since this person would not be you. As mentioned previously, you could never get this judgement wrong; only your copy could.

To see what’s going on here, it helps to consider a similar scenario once mentioned by Russell in which the entire world suddenly springs into being with everyone having false memories of an unreal past. Can you rule out that this did not just happen? Like our case, this one revolves around two people: either you are who you remember yourself to be (the genuine article) or else a freshly-created individual with false memories of an unreal past (the copy). And there appears to be no telling which one. In this case, my sympathies lie with the sceptic: you simply cannot tell whether you are real or copy, i.e., whether or not the universe has just materialized.

This case is usefully compared with ours because it may seem on the surface to be the same sort of case. Just as you cannot tell in Russell’s case whether or not the universe has just materialized, you cannot tell upon waking up on Mars whether or not the machine had worked. In either case, you may be real or copy and you just cannot tell.

But there is a crucial difference between Russell’s case and ours. In Russell’s case, it doesn’t matter whether the subject of the epistemic predicament happens to be the real person or the copy: either way, the subject is being held to account for what he or she may claim to know. Suppose, for instance, that you consider Russell’s scenario absurd and contend that you are obviously real. The retort would be that, for all that you can tell, you are the copy, in which case your contention would be wrong. Notice in this way how the copy is potentially being held to account no less than the real person is. It doesn’t matter if you are real or copy: you are required either way to be right.

In our case, however, only the real person is being held to account. We were trying to ascertain whether you—the person who entered the machine—could tell that the machine had worked should you find yourself waking up on Mars. There is no concern with whether your copy could tell that the machine had failed should he find himself waking up on Mars. In a different context, we might care about this, but it is of no consequence where our current question of first-person empirical verification of teletransportation is concerned: it doesn’t matter if your copy gets it wrong, so long as you get it right. An example of a different context is given in Dennett’s tale above, in which it doesn’t matter if the protagonist is real or copy: she wants to be right either way.

Accordingly, if you find yourself waking up on Mars and contend that the machine had worked, the sceptic can no longer retort, as in Russell’s case, that, for all that you can tell, you might be the copy and that your contention might be false. If you were the copy, it wouldn’t matter that your contention was false because the copy is not being held to account.

The bottom line, therefore, is that you may enter the machine, press the button, and simply resolve to judge that the machine had worked should you so much as seem to find yourself waking up on Mars. You could never get this judgement wrong and would have every claim upon so waking to know that the machine had worked. Your copy, should he come into existence, would erroneously make the same judgement, but this has got nothing to do with you. Indeed, simultaneously as your copy makes this wrong judgement, you could well be making the correct judgement that the machine had failed, e.g., by finding yourself waking up in purgatory. In testing the machine, we want you to be able to tell whether the machine had worked or failed. This refers to A and B below, as opposed to A and C, which is what the sceptic has erroneously focused on.

In our case, in contrast, we care primarily about you (lower case), the one with genuine memories, the person who enters the teletransporter in an attempt to verify if the machine really works. Our question is whether you can tell whether the machine had worked should you find yourself waking up on Mars. It is irrelevant whether your copy can tell, should he find himself waking up on Mars: it was no part of our brief to care about that. And so the sceptic’s claim, “You can’t tell whether your memories are genuine or false,” needs to be handled here with care because it may well be true that YOU can’t tell whether YOUR memories are genuine or false, without it being true that you can’t tell whether your memories are genuine or false.

In my opinion, the sceptic stumbles in failing to heed this difference. It’s not true that you—the person who entered the machine—can’t tell whether your memories are genuine, should you find yourself waking up on Mars. (Likewise, whether the machine had worked, whether you were the one who entered the machine, etc.) As explained above, if you go simply by “how it seems,” you’re bound to get the answer right. By any measure, you would be able to tell. Tested repeatedly, for instance, you’d always give the correct answer and there’d be no grounds whatever for saying that you couldn’t tell, or didn’t know. This strategy of going by how it seems doesn’t work in Russell’s case because it is implicitly required there that the correct answer be given by the person selected at random between you and your copy. In contrast, it is part of the nature of our case that it doesn’t matter what answer your copy gives and that the correct answer only be given by you. The person being addressed in our case is you, specifically, as opposed to the one selected at random between you and your copy.

In this case, you do need to know whether you are real or copy, because it is required that you get the answer right either way.

In our case, by contrast, what the copy ends up judging, should he find himself waking up on Mars, is unimportant. The requirement is only that the person who entered the machine—you—have good grounds to judge that the machine had worked. According to the sceptic, you cannot have such grounds unless you know that you are the person who entered the machine. Now it is true that, in the sense in which the sceptic intends it, you don’t know whether you are the person who entered the machine. But, as just explained, it does not follow that you have to know this in order to have good grounds for judging that the machine had worked. Indeed, given that you cannot get this judgement wrong, you have all the grounds that you need for making it. In this same sense, you also do know that you are the person who entered the machine. You know this in the sense that, knowing that if you so judge as much, you could never get it wrong, you should have no hesitation in judging as much.

but this time it literally is your memory that is in question. In our case, in contrast, despite the same sceptical form of words being used—“You can’t tell if your memory is genuine or false”—your memory is never at fault.

There is a sense in which the danger in question exists but it goes beyond our original concern. Our concern was with whether the person who entered the machine, supposing he woke up on Mars, would know whether the machine had worked or failed. Another concern might be whether this person’s copy (should he come into existence) would know. When the sceptic says that if you woke up on Mars, you would not know whether the machine had worked or failed, the pronoun ‘you’ refers indifferently to either you or your copy, as though God had picked one of “you” at random and asked, ‘Did the machine work or fail?’ This randomly picked subject would do no better than chance in answering the question. This is the sense in which “you” don’t know whether the machine had worked or failed.

But it was no part of our brief to care whether your copy gets the answer right. Our concern was only with whether you, the one who entered the machine, do.

Why does the sceptic think there is such a danger? Well, he claims that if you entered the machine, pressed the button, and found yourself waking up on Mars, you’d be unable to tell (upon so waking) whether your memory of having entered the machine was genuine or false. You wouldn’t know if you were the one who originally entered the machine, or just a freshly-minted copy with false memories of having done so. As such, you’d be unable to tell if the machine had worked or failed. (If you happened to be the copy, the machine would in fact have failed.)

re is another sort of case where there is such a danger, with which ours is easily confused. Suppose you had a neurological syndrome (say) that saddled you with false memories from time to time. In some given instance of this case,

we can likewise say, “You don’t really know if your memory is genuine or fake,” but this time it really is your memory that is in question and the corresponding scepticism concerning your past really does have a bite. But our case is not of this sort.

Why could you not just resolve to judge that the machine had worked if, upon pressing the button, you found (or seemed to find) yourself waking up on Mars?

I will argue that these simple facts constitute a basis for resisting the sceptical provocation. Very roughly, knowing these facts as you press the button, you can simply resolve to judge that the machine had worked should you find yourself waking up on Mars. Not only would your judgement be bound to be correct, it would also be well-founded, and so you could claim, upon so waking and following up on the resolution, to know that the machine had worked.

The sceptic claims that, upon waking up on Mars, you would not be able to tell that the machine had worked. There is a sense in which this is true, as I will make clear below, but the fact that your judgement that the machine had worked could never be wrong suggests also that there may be another sense in which you would so be able to tell. After all, if you could never get the judgement wrong, why should you hesitate in judging that the machine had worked should you find yourself waking up on Mars? I will argue that this line of thinking is sound and is more relevant for our purposes.

To this line of thinking, the sceptic would naturally retort that, upon waking up on Mars, you wouldn’t know whether it was you or your copy who was waking up, and that this is what generates the sceptical threat. You need to know that it was you—the one who entered the machine—that was waking up, otherwise you’d have no grounds to conclude that “your” memories were genuine and that the machine had worked.

This retort would have a point if it were required quite generally that the person waking up on Mars (whether he be real or copy) be able to tell whether the machine had worked or failed. But this is not required. It is not required that your copy be able to tell that the machine had failed, but only that you be able to tell that the machine had worked.

So our case differs from a superficially-similar case once mentioned by Russell in which the entire world sprang into being a moment ago with everyone having false memories of an unreal past: how can you tell whether or not this happened? In this case, you do need to know whether you are real or copy, because it is required that you get the answer right either way.

In our case, by contrast, what the copy ends up judging, should he find himself waking up on Mars, is unimportant. The requirement is only that the person who entered the machine—you—have good grounds to judge that the machine had worked. According to the sceptic, you cannot have such grounds unless you know that you are the person who entered the machine. Now it is true that, in the sense in which the sceptic intends it, you don’t know whether you are the person who entered the machine. But, as just explained, it does not follow that you have to know this in order to have good grounds for judging that the machine had worked. Indeed, given that you cannot get this judgement wrong, you have all the grounds that you need for making it. In this same sense, you also do know that you are the person who entered the machine. You know this in the sense that, knowing that if you so judge as much, you could never get it wrong, you should have no hesitation in judging as much.

Here,

overlooks one thing, viz., that it doesn’t matter if your copy gets things wrong, e.g., that he “has no grounds” to conclude that his memories are genuine. What matters is only that you—the person who entered the machine—has such grounds.

There is a confusion here. The sceptical concern is that your experience of waking up on Mars is indistinguishable (“from the inside”) from that of your copy’s experience, so how can you tell, upon waking up on Mars, whether you are real or copy? The answer, as before, is that you cannot possibly be copy: there is no such possibility as that of entering the machine, pressing the button, and then waking up on Mars in your copy’s body.

The apparent force of the sceptical provocation is due to confusing our scenario with another one (or two). Consider a case where you have a neurological syndrome (say) that saddles you with false memories from time to time. In some given instance of this case, we can likewise say, “You don’t really know if your memory is genuine or fake,” but this time it really is your memory that is in question and the corresponding scepticism concerning your past really does have a bite. But our case is not of this sort.

In our case, if you entered the machine and found yourself waking up on Mars, there’d be no question of your memory being unreliable. As pointed out, the person with false memories is always your copy and never you. So there’d be no reason to doubt your memory. You’d know that you were the one who entered the machine; you’d remember it. The sceptical assertion, “You wouldn’t really know whether your memory was genuine or fake” is no pronouncement on your memory here, but means something different. (The pronoun ‘you’ doesn’t refer to you in particular, but indiscriminately to either you or your copy.) It’s a way of saying that if someone was placed at random in either you or your copy’s shoes, they would be unable to say who they were and whether their memory was genuine or fake.

To make all of this out, let me spell out the ingenuous view a little and then explain why I think the sceptical complications above fail to undermine it.

Suppose then that you enter the machine on Earth, press the green button, and duly find yourself waking up on Mars. Ordinarily, at this point—i.e., modulo the sceptical complications to be addressed below—you’d realize that you were alive and know right away that the machine had worked, in the same way that, if you found yourself waking up in purgatory, you’d realize that you were dead and know right away that the machine had failed.

In either case, your memory would tell you that your last act was to enter the machine and press the green button, allowing you to draw the appropriate conclusion. We’d take you to know that the machine had worked in the one case, just as we’d take you to know that it had failed in the other.

This is the ingenuous view of the matter.

In the first of these two cases, however, the sceptical complication is that if you found yourself waking up on Mars, you wouldn’t really know who you were: the person who entered the machine or a freshly-minted copy with false memories of having done so. So you wouldn’t really know if your memory was genuine or fake, and thus wouldn’t be able to tell if the machine had worked or failed. (If you happened to be the copy, the machine would in fact have failed.)

I think that this complication is illusory and not the hazard that it first seems to be. One way to see this is to notice that the person with false memories here is your copy and not you: your memory is never at any point in question. So the assertion, “You wouldn’t really know

if your memory was genuine or fake,” is potentially very misleading in this context.

To see this clearly, consider a different case with which ours is easily confused: you have a neurological syndrome (say) that saddles you with false memories from time to time. In some given instance of this case, we can likewise say, “You don’t really know if your memory is genuine or fake,” but this time it really is your memory that is in question and the corresponding scepticism concerning your past really does have a bite. But our case is not of this sort.

In our case, if you entered the machine and found yourself waking up on Mars, there’d be no question of your memory being unreliable. As pointed out, the person with false memories is always your copy and never you. So there’d be no reason to doubt your memory. You’d know that you were the one who entered the machine; you’d remember it. The sceptical assertion, “You wouldn’t really know whether your memory was genuine or fake” is no pronouncement on your memory here, but means something different. (The pronoun ‘you’ doesn’t refer to you in particular, but indiscriminately to either you or your copy.) It’s a way of saying that if someone was placed at random in either you or your copy’s shoes, they would be unable to say who they were and whether their memory was genuine or fake.

Alternatively, it’s a way of saying that you and your copy would have phenomenally indistinguishable experiences, including memory experiences. The sceptic takes this to show that if you entered the machine and found yourself waking up on Mars, you wouldn’t know whether you were the real you—the one who entered the machine—or just the copy. Your experiences would be phenomenally indistinguishable, so how could you tell who you were?

The answer again is that such a claim as that you might be your copy is ambiguous. In one sense, you obviously could not be the copy: he is a different person. How could you enter the machine and subsequently find yourself in the body of your copy? That would be metaphysically impossible. In this sense, you obviously know that you are not your copy. The sceptical claim that “you might be your copy” only makes sense if understood in a different way, viz., as saying that “these” might be the experiences of your copy.

it is still not that simple because of the existence of countervailing reasons for upholding the opposite view, viz., that if the machine works, then you can discover this by trying the thing out for yourself.

These countervailing reasons are hardly ever mentioned in the literature, and perhaps not even generally noticed, but we do need to consider them because they threaten to generate a paradox at this point, wherein distinct lines of reasoning—sceptical and countervailing—are seen to lead to contrary conclusions.

Let’s see what these countervailing reasons are before considering their paradoxical ramifications.

We may take our cue from the sceptic, who points out that if the machine worked and you found yourself waking up on Mars, you wouldn’t be able to tell (upon so waking) who you were: the person who entered the machine or a freshly-minted copy with false memories of having done so. As such, you wouldn’t be able to tell if the machine had really worked. (If you happened to be the copy, the machine would in fact have failed.)

This line of reasoning is fine as far as it goes but it does overlook one important fact, viz., that, necessarily, if you entered the machine and found yourself waking up on Mars, you would judge correctly if you judged that the machine had worked. Your judgement would be correct even if you were unable to tell who you were upon waking, or whether your memories were genuine or fake. It would be correct solely in virtue of the fact that it was you who was making the judgement. Your copy would be wrong if he attempted the parallel judgement, but not you. You could never get this judgement wrong.

Knowing this as you enter the machine and press the green button, you should have no hesitation in resolving to judge that the machine had worked were you to subsequently find yourself waking up on Mars, or even just seeming to do so. You’d know that you could never get this judgement wrong.

Suppose then that you enter the machine on Earth, press the green button, and duly find yourself waking up on Mars. Ordinarily, at this point—i.e., modulo the important complication to be discussed below—you’d know right away that the machine had worked, in the same way that, if you found yourself waking up in purgatory, you’d know right away that the machine had failed. In either case, your memory would tell you that the last thing you did was enter the machine and press the green button, allowing you to draw the appropriate conclusion. We’d take you to know that the machine had worked in the one case, just as we’d take you to know that it had failed in the other.

The complication, however (in the first of these two cases), is that if you found yourself waking up on Mars, you wouldn’t really know who you were: the person who entered the machine or a freshly-minted copy with false memories of having done so. As such, you couldn’t be sure that the machine had worked. (If you happened to be the copy, the machine would in fact have failed!)

This is of course the sceptic rearing his head again. Note, however, that this is not just a repetition of his previous result that you wouldn’t know whether the machine had worked or failed; the sceptic is also using this result to block the “ordinary” demonstration that you would know that the machine had worked.

So the “ordinary” reason for thinking that you would know that the machine had worked had you found yourself awakening on Mars appears to be lost: you’d ordinarily need the aid of memory to know that the machine had worked, but you have no way of knowing if your memories are genuine or fake.

These sceptical considerations actually play a dual role in our dialectic. They serve to show that you would not know whether the machine had worked (as we previously saw) but they also aim to undercut our ordinary reason for thinking that you would know that the machine had worked (as we just saw). These two roles should be separated because I think that the sceptical considerations succeed only in the first of these two roles.

They do not succeed in the second role because they do not touch the fact that, necessarily, if you found yourself waking up on Mars, you would be correct to judge that the machine had worked. Notice that this judgement of yours would be correct even if you did not know whether your memories were genuine or fake. It would be correct solely in virtue of the fact that it was you who was making the judgement. Your copy would be wrong if he attempted the parallel judgement, but not you. You could never get this judgement wrong.

⚹

This may be thought to settle the matter but, in fact, it is still not that simple because there exists a little-noticed countervailing reason for upholding the opposite view, viz., that if the machine works, then you can discover this by trying the machine out for yourself.

If the countervailing reason were sound, then we’d have the makings of a paradox, with distinct arguments leading to contrary conclusions. But let’s see what the countervailing reason is before considering its paradoxical ramifications.

The idea behind all this talk of first-person verification is that if the machine doesn’t kill you, then you’d find yourself waking up on Mars, still very much alive, and you’d know thereby—modulo the complication about to be discussed—that the machine had worked.

An idea of this sort would normally need no explaining. It would be the same if you wanted to verify for yourself that consuming Japanese pufferfish (say) doesn’t kill you. You could just venture into your kitchen, cook up the dish and see: if it didn’t kill you, then you’d find yourself still in your kitchen, very much alive, and you’d know thereby that the dish was safe.

The complication in the teletransportation case, however—the sceptical complication—is that if you found yourself waking up on Mars, you wouldn’t know who you were: the person who entered the machine or a freshly-minted copy with false memories of having done so. So you’d be in no position to judge that you were still alive and that the machine had worked. (If you were the copy waking up on Mars, this would mean that the real you—the one who entered the machine—had died and that the machine had failed!)

What is little noticed, however, is that this complication is actually harmless. For it does not touch the fact that, necessarily, if you found yourself waking up on Mars, you would be correct to judge that you were still alive and that the machine had worked. Indeed, this judgement of yours would be correct even if you did not know whether you were real or copy. It would be correct solely in virtue of the fact that it was you who was making the judgement. Your copy would be wrong if he attempted the parallel judgement, but not you. You could never get this judgement wrong.

Our case differs therefore from a superficially-similar case once mentioned by Russell in which the entire world sprang into being a moment ago with everyone having false memories of an unreal past: how do you know this did not happen? In this case too, you do not know whether you are “real” or “copy,” but notice that the pronoun ‘you’ in this case may potentially refer to either character. (Were you to judge that you and the entire past were real, there is a chance that you would be wrong, since you might in fact be “copy.”)

In our case, by contrast, the pronoun ‘you’ in the clause, ‘If you found yourself waking up on Mars … ,’ only ever refers to the real you, the person who entered the teletransporter, and never to your copy. This stems from the nature of the case: ours is one in which you, qua person who entered the teletransporter, are attempting to verify whether the machine works. So the referent of ‘you’ is fixed independently of the subsequent sceptical question of whether you are real or copy. (In Russell’s case, in contrast, the referent of ‘you’ is fixed precisely by the answer to that sceptical question.) And so, as before: if you found yourself waking up on Mars, or even just seemed to find yourself doing so, then if you judged that you were real and that the machine had worked, your judgement would automatically be right. It would be a “safe judgement” to make.

Knowing all this, you should have no hesitation in judging that the machine had worked should you seem to find yourself waking up on Mars. Indeed, you’d know that the machine had worked, this being the same thing.

More precisely, remember that verification in this context is a temporally extended process that begins on Earth: you enter the machine, press the button, and see “what happens next.” You know that one of two things will happen next: either you will find yourself waking up on Mars (if the machine works), or else you will find yourself waking up in purgatory, or perhaps lose consciousness altogether (if the machine fails).

So if what transpires next is an awakening on Mars, you may conclude that the machine worked, because you know that otherwise, if the machine had failed, either an awakening in purgatory or else nothing whatever (loss of consciousness) would have transpired. Q.E.D.

This countervailing argument, when properly understood, also seems undeniable to me. I take it to show that, if the machine works, you can (rather straightforwardly) verify this by trying the machine out for yourself. Just press the button and see what happens. If the next thing that happens is an awakening on Mars, then you may conclude, as just explained, that the machine worked.

But let’s make sure that we understand the argument correctly.

A natural “sceptical” objection to the argument is that if you found yourself waking up on Mars, you would not know who “you” were: the person who entered the machine or a freshly-minted copy with false memories of having done so. So you could not rightly judge that the machine must have worked on the countervailing grounds that otherwise “you” would have found yourself waking up in purgatory (or else lost consciousness altogether). After all, if you were the copy, then if the machine had failed, you would have found yourself waking up on Mars!

But this misses the point, which is that from your starting point of getting into the machine and pressing the button, you could never wake up in the body of your copy. So if you found yourself waking up on Mars, you can justifiably claim that the machine had worked: indeed, you could never get this wrong. Anyone who enters the machine and finds themselves waking up on Mars will never be wrong in averring thereby that the machine works. The sceptical worry was that, for all you know, the machine might have failed and you were just the copy. But this doesn’t matter. If you were the copy, then you’d be wrong to aver that the machine had worked, but the person who entered the machine would still get it right: he or she would find themselves waking up in purgatory and realize that the machine had failed. It is the person who entered the machine that we care about here: he or she is the one who is attempting to verify whether the machine works.

Unfortunately, this objection is overly-focused on the “sceptical” way of thinking and misses the point of the countervailing logic. For example, the objection takes for granted that you wouldn’t know who “you” were if you found yourself waking up on Mars. But a proponent of the countervailing logic would contend the opposite. Remember again that verification in this context is a temporally extended process that begins on Earth: you enter the machine, press the button, and see “what happens next.” But notice how obvious it is from this point of view that, whatever happens next, your consciousness could never “jump” into the body of your copy who was awakening on Mars—or for that matter into anyone else’s body. So, whatever “happens next” (e.g., an awakening on Mars, as it may be) will be something that happens to you, the one who entered the machine. Now, you certainly know this as you press the button; the question is whether you will continue to know this when you awaken, e.g., if the machine works and you find yourself awakening on Mars. The answer, it seems to me, is yes, because nothing in our science fantasy so far impugns the fact that, barring special circumstances such as forgetfulness or amnesia, absent in our fantasy, you retain your knowledge over time in the normal way. So if the machine works and you find yourself awakening on Mars then, as you thus awaken, you know (trivially) that this awakening is what “happens next” and (by retention of knowledge) that it is happening to the person who entered the machine. So you know that you are the one that entered the machine.

if what happens next is an awakening on Mars, you can be sure that you are not the copy; derivatively that the machine must have worked.

This is also clear if we imagine simply that the machine fails. In this case, when you press the button, you will either find yourself waking up in purgatory or else lose consciousness altogether. Your copy finds himself waking up on Mars but his experience has nothing to do with you: as before, you certainly won’t find yourself waking up in his body. For you, the subject, either the purgatory experience or else outright unconsciousness happens next; there is no question of your consciousness suddenly jumping into the body of your copy on Mars. As before then, if the Mars experience happens next, then you can’t possibly be the copy and so the machine really must have worked.

In fact, you can’t be the copy whatever happens next, because

It is important to realize that none of this denies what the sceptic maintains, viz., that if you find yourself waking up on Mars, then you won’t be able to tell whether you are the original or the copy. This has already been conceded. The point rather is that we can simultaneously prove the opposite: that you will be able to tell. So if you find yourself waking up on Mars, you both won’t and will be able to tell who you are. (It’s a paradox.)

doesn’t kill you is no more difficult than verifying that ingesting rat poision (say) doesn’t kill you.

⚹

This may be thought to settle the matter but, in fact, we now run up against a difficulty of sorts—a paradox, in effect—because there is a different way of looking at the issue which seems to deliver the opposite result.

The sceptical worry was that, if you found yourself waking up on Mars, you’d be unable to tell who “you” were: the person who originally entered the machine and pressed the button, or just a freshly-minted copy with false memories of having done so. So you’d be unable to verify that the machine had worked.

This logic is undeniable in its own way.

But here’s another way of looking at things – a rather straightforward way, as it turns out. Remember that verification in this context is a temporally extended process that begins on Earth: you enter the machine, press the button, and “see what happens next.” You know that one of two things will happen upon pressing the button: either you’ll find yourself waking up on Mars (if the machine works), or else you’ll find yourself waking up in purgatory, or perhaps lose consciousness altogether (if the machine fails).

So if you find yourself waking up on Mars, you can be sure that the machine must have worked, because otherwise (if the machine had failed) you would have found yourself waking up in purgatory or else lost consciousness altogether!

This simple countervailing logic also seems undeniable to me!

I take it to show that, if the machine works, you can (straightforwardly) verify this by trying the machine out for yourself. Just press the button and see what happens. If the next thing that happens is an awakening on Mars, then, as just explained, you may conclude that the machine has worked. This is the paradox-generating “opposite result” mentioned above.

This countervailing logic may seem almost too simple, however, so some clarification is called for.

The obvious “sceptical” objection is that if you found yourself waking up on Mars, you would not know who “you” were: the person who entered the machine or a freshly-minted copy with false memories of having done so. For all that you know, you might be the copy waking up on Mars, i.e., even as the real you undergoes his appointed purgatory experience. So the fact that you would have found yourself waking up in purgatory should the machine have failed—this being the central premise of the countervailing logic above—does not entitle you to conclude that the machine must have worked were you to find yourself waking up on Mars. Because you wouldn’t be able to tell who “you” were if you found yourself waking up on Mars.

Unfortunately, this objection is overly-focused on the “sceptical” way of thinking and misses the point of the countervailing logic.

Remember again that verification in this context is a temporally extended process that begins on Earth: you enter the machine, press the button, and “see what happens next.” But notice how obvious it is from this point of view that, whatever happens next, your consciousness could never “jump” into the body of your copy who was awakening on Mars. So if what happens next is an awakening on Mars, you can be sure that you are not the copy, and that the machine must have worked.

This is also clear if we imagine simply that the machine fails. In this case, when you press the button, you will either find yourself waking up in purgatory or else lose consciousness altogether. Your copy finds himself waking up on Mars but his experience has nothing to do with you: as before, you certainly won’t find yourself waking up in his body. For you, the subject, either the purgatory experience or else outright unconsciousness happens next; there is no question of your consciousness suddenly jumping into the body of your copy on Mars. As before then, if the Mars experience happens next, then the machine really must have worked.

It is important to realize that none of this denies what the sceptic maintains, viz., that if you find yourself waking up on Mars, then you won’t be able to tell whether you are the original or the copy. This has already been conceded. The point rather is that we can simultaneously prove the opposite: that you will be able to tell. So if you find yourself waking up on Mars, you both won’t and will be able to tell who you are. (It’s a paradox.)

This also explains why the reasoning is not available to your “copy,” supposing he or she were to come into existence. Your copy, awakening on Mars, might likewise reason, “The machine must have worked because otherwise a purgatory experience would have occurred.” But this reasoning is unsound because the bit after “otherwise” is not true of your copy.

It’s not often in philosophy that you come across a veridical paradox of this sort but the above seems to me to be one. One line of reasoning shows that teletransportation cannot be verified in the first person; another shows that it can. Neither reasoning refutes the other: each just goes about establishing its own conclusion. (The sceptical reasoning is often voiced but the countervailing logic is usually overlooked.) I can’t see anything wrong with either reasoning, taken in its own terms, so the blame must lie with some underlying assumption that they share in common.

In fact, a pair of related assumptions may be held jointly to blame and I will suggest that, when all is said and done, the countervailing logic is the one with the balance of truth on its side.

oooooooooooooooooooooooooo

oooooooooooooooooooooooooo

2. Is teletransportation empirically verifiable?

There is surprisingly little discussion of whether teletransportation can be verified (or falsified) empirically. I think the question is more interesting than most people realize.

The issue is whether the person who emerges from the transporter is the same person as the one who entered; or just a freshly-created lookalike, the one who entered having died. Empirically speaking, what everyone agrees on is this: we cannot determine the answer “from the outside,” i.e., from a third-person vantage point. It would be futile to observe someone else enter the machine and thereafter “check” if that same person emerged on Mars: there would be no way to tell if the person who emerged was the same as the one who entered.

The interesting question, however, is whether we can determine the answer “from the inside,” from the first-person point of view. At risk to your own life, could you not try the machine out for yourself in order to see what happens?

This simple first-person empirical check does appear to work in at least one type of case. If there was such a thing as an afterlife, then, if the machine killed you, you’d find yourself waking up in some fiery place like purgatory (say), and you’d realize (let’s presume) that you were dead. Thus:

Would this satisfy Parfit himself, if no one else, that the machine had worked? Would it constitute an empirical verification of teletransportation?

As we know, the matter is not that simple, because of the following obvious snag. Even if the machine did work, and you did find yourself waking up on Mars, you’d be unable to tell (upon waking) whether you were the person who originally entered the machine, or whether that person died, and you were just a freshly-minted lookalike with false beliefs of having previously entered the machine. After all, anyone—real or copy—who emerged on Mars would “remember” having earlier entered the machine on Earth. So you’d be none the wiser as to whether you were “real” or “copy,” and thus none the wiser as to whether the machine had really worked.

For example, in Dan Dennett’s story, a woman on Mars with a broken spaceship uses a teletransporter to get back to Earth. Upon reuniting with her daughter Sarah, it “hits her”:

⚹

But here’s another way of looking at things – a rather straightforward way, as it turns out. Remember that verification in this context is a temporally extended process that begins on Earth: you enter the machine, press the button, and “see what happens next.” You know that one of two things will happen upon pressing the button: either you’ll find yourself waking up on Mars (if the machine works), or else you’ll find yourself waking up in purgatory, if not lose consciousness altogether (if the machine fails).

So if you find yourself waking up on Mars, you can be sure that the machine must have worked, because otherwise (if the machine had failed) you would have found yourself waking up in purgatory or else lost consciousness altogether!

This simple countervailing logic also seems undeniable to me!

I take it to show that, if the machine works, you can (straightforwardly) verify this by trying the machine out for yourself. Press the button and see what happens. If the next thing that happens is an experience of waking up on Mars, then, as just explained, the machine must have worked. This is the paradox-generating “opposite result” mentioned above.

This countervailing logic may seem almost too simple, however, so some clarification is called for.

The obvious “sceptical” objection is that if you found yourself waking up on Mars, you would not know who “you” were: the person who entered the machine or a freshly-minted copy with false memories of having done so. For all that you know, you might be the copy waking up on Mars, i.e., even as the real you undergoes his appointed purgatory experience. So the fact that you would have found yourself waking up in purgatory should the machine have failed—this being the central premise of the countervailing logic above—does not entitle you to conclude that the machine must have worked were you to find yourself waking up on Mars. Because you wouldn’t be able to tell who “you” were if you found yourself waking up on Mars.

Unfortunately, this objection is overly-focused on the “sceptical” way of thinking and misses the point of the countervailing logic.

Remember again that verification in this context is a temporally extended process that begins on Earth: you enter the machine, press the button, and “see what happens next.” But notice how obvious it is from this point of view that, whatever happens next, your consciousness could never “jump” into the body of your copy who was awakening on Mars. So if what happens next is a Mars awakening, you can be sure that you are not the copy, and that the machine must have worked.

This is also clear if we imagine simply that the machine fails. In this case, when you press the button, you will either find yourself waking up in purgatory or else lose consciousness altogether. Your copy finds himself waking up on Mars but his experience has nothing to do with you: as before, you certainly won’t find yourself waking up in his body. For you, the subject, either the purgatory experience or else outright unconsciousness happens next; there is no question of your consciousness suddenly jumping into the body of your copy on Mars. As before then, if the Mars experience happens next, then the machine must have worked.

It is important to realize that none of this denies what the sceptic maintains, viz., that if you find yourself waking up on Mars, then you won’t be able to tell whether you are the original or the copy. This has already been conceded. The point rather is that we can also prove that you will be able to tell. So if you find yourself waking up on Mars, you both won’t and will be able to tell who you are. (It’s a paradox.)

This also explains why the reasoning is not available to your “copy,” supposing he or she were to come into existence. Your copy, awakening on Mars, might likewise reason, “The machine must have worked because otherwise a purgatory experience would have occurred.” But this reasoning is unsound because the bit after “otherwise” is not true of your copy.

It’s not often in philosophy that you come across a veridical paradox of this sort but the above seems to me to be one. One line of reasoning shows that teletransportation cannot be verified in the first person; another shows that it can. Neither reasoning refutes the other: each just goes about establishing its own conclusion. (The sceptical reasoning is often voiced but the countervailing logic is usually overlooked.) I can’t see anything wrong with either reasoning, taken in its own terms, so the blame must lie with some underlying assumption that they share in common.

In fact, a pair of related assumptions may be held jointly to blame and I will suggest that, when all is said and done, the countervailing logic is the one with the balance of truth on its side.

⚹

To see this, notice first that the contradiction arises only in the case where you find yourself waking up on Mars—the contradiction being that, upon so waking, you’d be unable to tell whether the teletransporter had worked, and yet you’d be able to tell that it did work.

There is no contradiction if you don’t find yourself waking up on Mars—i.e., if you lose consciousness altogether or find yourself waking up in a place like purgatory. In this case, you just die, while your doppelgänger awakens on Mars, unable to tell whether the teletransporter had worked.

So a superficial conclusion might be that, on pain of contradiction, you cannot possibly find yourself waking up on Mars! (This is tantamount to concluding that teletransportation is impossible.)

The conclusion is superficial however because all this talk of the “case” where you find yourself waking up on Mars, as opposed to the “case” where you don’t, presupposes that there is a genuine distinction between the two cases to begin with. That is to say, it presupposes that there is a genuine distinction between the “case” where the teletransporter works (and sends you to Mars) and the “case” where it doesn’t (and you die). But this presupposition is philosophically controversial and potentially open to question. Thus, some people see no substantial difference between the two alleged “cases,” viz., being teletransported to Mars and being killed by the teletransporter (and replaced by a doppelgänger on Mars). These are the same case, they say, only described in different words: your alleged doppelgänger is you. Parfit himself, most noticeably, holds a position of this sort and considers the difference between these two “cases” to be just verbal. (This position is easier to hold if you ignore the possibility of an afterlife and assume that consciousness ends with death.)

So it is open to us to abandon this prior supposition instead and, if we do that, then the opposite conclusion (from before) may be drawn. Our conclusion would now be that teletransportation essentially works! For we would now be saying that the hypothesis of you dying in the teletransporter and being replaced on Mars by a copy is essentially empty, much like the traditional sceptical hypothesis of the universe surreptitiously doubling in size overnight: there is no such possibility. It is you who wakes up on Mars; the suggested alternative is bogus.

Nothing so far tells us which supposition we should abandon but they can’t both be true, on pain of contradiction. What we may conclude therefore is that if the aforementioned distinction is genuine—the “prior supposition” is in order—then you won’t find yourself waking up on Mars; conversely that if you do find yourself waking up on Mars, then the aforementioned distinction must be bogus.

And notice how this latter tells us something important about the empirical verification of teletransportation, viz., that it can be verified after all in the first-person. Our bottom line is that, if you find yourself waking up on Mars, then you may infer that there is no distinction between you and your doppelgänger and so there is no longer any sceptical scenario to worry about: you may safely conclude that the machine has sent you to Mars. This is what I meant above in saying that the paradox shows that the countervailing logic has the balance of truth on its side.

⚹

What the foregoing paradox shows therefore is that the sceptical reasoning often voiced as a counterpoint to first-person attempts to empirically verify the reality of teletransportation is misguided. Superficially, the sceptical reasoning is sound, but a more careful look undermines it altogether. This is not specific to the phenomenon of teletransportation but applies as well to, e.g.,

different conclusion should be drawn, viz., finding yourself awakening on Mars would be proof enough that the teletransporter had worked, i.e., because the sceptical reasoning mentioned previously would no longer have any purchase.

and so empirical verification of teletransportation is possible, after all.

This would be hasty, of course, since other assumptions are required to generate the contradiction, and some other one of these might be to blame. The obvious candidate is the assumption that there is a genuine difference between you waking on Mars and your doppelgänger doing so. This assumption is required, of course, for the sceptical conclusion that, finding yourself waking on Mars, you cannot tell who “you” are.

the case where the teletransporter works (you subsequently awaken on Mars) and the case where the teletransporter fails to work (you die and a doppelgänger awakens on Mars).

Perhaps there is no such difference: your doppelgänger is you.

The only sensible culprit is the assumption that there is a difference between you waking up on Mars and your copy doing so. The foregoing paradox suggests that there is no such difference: your copy is you.

From this point of view, there is no question of your doppelgänger waking up on Mars instead of you. Any doppelgänger of yours that wakes up on Mars is you, and so if teletransportation works, you can indeed verify this for yourself by pressing the button and seeing if you wake up on Mars.

This may be thought to settle the matter, but it turns out, alas, that these considerations are still too simple. While they admittedly do seem to show that teletransportation cannot be verified by trying out the procedure for yourself, they run up against a counterargument that seems to show the opposite: that it can be so verified. A paradox now threatens to arise, with argument and counterargument supporting contradictory conclusions. I will argue that the paradox is real and forces us to abandon an assumption that both arguments have in common.

Here’s the counterargument. To perform the empirical test, you simply press the button and see what happens. If the machine works, you’ll find yourself waking up on Mars. If it doesn’t work, then you’ll die, meaning either that you’ll lose consciousness permanently, or you’ll find yourself waking up in an afterlife. Let’s go with the afterlife for the moment, e.g., that, if you die, you’ll awaken in a fiery place. This assumption is not really necessary but it imposes a certain symmetry upon the argument that helps to bring out its logic. As before then, just press the button and see what happens. Clearly now, if you don’t find yourself waking up in a fiery place, then you’ll know that the machine has worked. (Because if the machine did not work, you’d find yourself waking up in a fiery place.) In particular, if you find yourself waking up on Mars, you’ll know that the machine has worked. So the originally-mooted first-person empirical verification of teletransportation is possible, after all.

The sceptical worry was that, if you found yourself waking up on Mars, you’d be unable to tell who “you” were: the one who pressed the button or just a freshly-minted copy. So you’d be unable to tell if the machine had worked. This logic is undeniable in its own way. But consider also how obvious it would be to you at the point of pressing the button that you could never wake up on Mars in the body of your copy. After all, a copy of you could exist only if the machine did not work, whereupon you’d find yourself waking up in a fiery place, rather than on Mars, let alone on Mars in the body of your copy. And so I repeat: if you found yourself waking up on Mars, you’d know that the machine had worked, because otherwise you’d have found yourself waking up in a fiery place. This logic also seems undeniable to me.

-------------

Recall that the point of the “obvious snag” is that anyone waking up on Mars would be in one of two indistinguishable scenarios: they might really have arrived from Earth, or they might be a freshly-minted person with false memories of having come from Earth. They would have no way of telling which scenario was theirs. All of this sounds perfectly reasonable and I think that, given all the necessary assumptions, this line of reasoning is correct.

But consider now the following “counterargument,” which seems to be just as correct. Consider how obvious it would be to you at the point of departure, with your finger on the green button, that you could never end up in the second of these two scenarios. How could you possibly wake up on Mars in the body of a freshly-minted person? If the machine killed you, you’d be permanently knocked out, if not in purgatory (see below). A freshly-minted person would awaken on Mars, thinking he had survived the trip, but you wouldn’t have any such thoughts: you’d be dead. On the other hand, were you to survive and wake up on Mars, there’d be no freshly-minted person to speak of. Either way, you could never end up in the second scenario mooted above. And you know this as you press the green button. So if you subsequently find yourself waking up on Mars, you’d know that you were in the first scenario: you’d know that the machine had worked. In other words, if the machine works, verification “from the inside” is possible.

This counterargument is even more compelling if you assume, as before, that there is an afterlife. Then if you pressed the green button and found yourself in a fiery place, you’d know that the machine had killed you, whereas if you found yourself on Mars, you’d know that the machine had worked (because otherwise you’d have found yourself in a fiery place). The appeal to an afterlife is not really necessary, but it may help us to see the essential point: if you found yourself waking up on Mars, you’d know that the machine had worked, because the last thing that could have happened is that you died and woke up in the body of a freshly-minted person.

It is not often in philosophy that you encounter a veridical paradox but I am driven to conclude that the foregoing is one. I can’t see anything wrong with either the “obvious snag” or the “counterargument.” Both seem to be entirely logical but they lead to contradictory conclusions. Assuming that the teletransporter works, the obvious snag shows that you cannot establish this by trying it out, whereas the counterargument shows that you can. Neither is refuted by trumpeting the other because each simply goes about establishing its own conclusion.

There is one important difference between the two lines of reasoning. One is conducted at the point of waking up on Mars, whereas the other is conducted at the point of departure. Does this make a difference?

I believe that both lines of reasoning are correct and that the blame must lie with some underlying assumption that they have in common. The contradiction arises only if we assume both that teletransportation is real and that there is a genuine distinction between you waking up on Mars and your doppelgänger doing so. So at least one of these assumptions must be false. Either teletransportation is not real (the machine just kills you and your doppelgänger wakes up on Mars), or else there is no genuine distinction, after all, between you waking up on Mars and your doppelgänger doing so: any “doppelgänger” of yours who wakes up on Mars is you; seeming to wake up on Mars is waking up on Mars.

It follows that, if teletransportation is real, then you can discover this by trying it out and finding yourself waking up on Mars. So the reality of teletransportation stands to be demonstrated empirically after all.

From ostrich notes to self 4, 58

Some quotes from Michael Cerullo's "Uploading and Branching Identity" (2014).

Nagel coined the term day-person to refer to the possibility that each evening after we go to sleep we die and a new conscious being with false memories awakes the next day (Parfit 1984). There is of course no way to rule out that we are all day-people, or even that we are minute-people that are born a new each minute with false memories. [27, n. 12]‘’

-------- Older one

The interesting question, however, is whether we can determine the answer “from the inside,” from the first-person point of view. At risk to your own life, could you not try the machine out for yourself in order to see what happens? Admittedly, if the transporter killed you, then you’d learn nothing: you’d just lose consciousness and it’d be over. But if the machine worked, and you found yourself waking up on Mars, would this not settle the matter, if only for yourself? Would you not have established to your satisfaction that you really had been transported to Mars? Take Parfit’s story above:

Would this not satisfy Parfit himself, if no one else, that the machine had worked? Would this not constitute empirical verification of the reality of teletransportation?

Naturally, it’s not that simple, because of the following obvious snag. Even if the machine did work, and you did find yourself waking up on Mars, you’d be unable to tell (upon waking) whether you really did enter the machine on Earth, or whether you were just a freshly-minted lookalike with false beliefs of having previously entered the machine. After all, anyone—real or copy—who emerged from the machine would “remember” having earlier entered the machine. So you’d be none the wiser as to whether you were “real” or “copy,” and thus none the wiser as to whether the machine had really worked.

For example, in Dan Dennett’s story, a woman on Mars with a broken spaceship uses a teletransporter to get back to Earth. Upon reuniting with her daughter Sarah, it “hits her”:

So it seems that you cannot determine if the machine works even if you tried it out for yourself. The reality of teletransportation cannot be established either from the “outside” or from the “inside;” so it cannot be established empirically, after all. This may ultimately be why the question is seldom discussed of whether teletransportation is amenable to empirical verification. The answer may be thought in this way to be obviously no.

This may be thought to settle the matter, but it turns out that these considerations are still too simple: there’s one final twist to the tale. The point of the “obvious snag” is that anyone waking up on Mars would be in one of two indistinguishable scenarios: they might really have arrived from Earth, or they might be a freshly-minted person with false memories of having come from Earth. They would have no way of telling which scenario was theirs. All of this sounds perfectly reasonable and I think that, in its own way, this line of reasoning is quite correct.

But consider this counterargument, which, in its own way, seems to be just as correct. Consider how obvious it would be to you at the point of departure, with your finger on the green button, that you could never end up in the second of these two scenarios. How could you possibly wake up on Mars in the body of a freshly-minted person? If the machine killed you, you’d be permanently knocked out. A freshly-minted person would awaken on Mars, thinking he had survived the trip, but you wouldn’t have any such thoughts: you’d be dead. On the other hand, were you to survive and wake up on Mars, there’d be no freshly-minted person to speak of. Either way, you could never end up in the second scenario mooted above. And you know this as you press the green button. So if you subsequently find yourself waking up on Mars, you’d know that you were in the first scenario: you’d know that the machine had worked. In other words, if the machine works, verification “from the inside” is possible.

Note: I'm not saying that anyone who finds themselves waking up on Mars would know that they were in the first scenario mentioned. Clearly, a freshly-minted person waking up on Mars would not know this.

This counterargument is even more compelling if you imagine, as some people believe, that there is an afterlife. I assumed above that consciousness is permanently extinguished when we die, but suppose instead, as the Catholics believe, that the first station after death is purgatory, sometimes described as a fiery place where your spirit gets purified before ascent into heaven. Then if you pressed the green button and found yourself in a fiery place, you’d know that the machine had killed you, whereas if you found yourself on Mars, you’d know that the machine had worked (because otherwise you’d have found yourself in a fiery place). The appeal to an afterlife is not really necessary, but it may help us to see the essential point: if you found yourself waking up on Mars, you’d know that the machine had worked, because the last thing that could have happened is that you died and woke up in the body of a freshly-minted person.

It is not often in philosophy that you encounter a veridical paradox but I am driven to conclude that the foregoing is one. I can’t see anything wrong with either the “obvious snag” or the “counterargument.” Both seem to be entirely logical but they lead to contradictory conclusions. Assuming that the teletransporter works, the obvious snag shows that you cannot establish this by trying it out, whereas the counterargument shows that you can. Neither is refuted by trumpeting the other because each simply goes about establishing its own conclusion.

There is one important difference between the two lines of reasoning. One is conducted at the point of waking up on Mars, whereas the other is conducted at the point of departure. Does this make a difference?

I believe that both lines of reasoning are correct and that the blame must lie with some underlying assumption that they have in common. The contradiction arises only if we assume both that teletransportation is real and that there is a genuine distinction between you waking up on Mars and your doppelgänger doing so. So at least one of these assumptions must be false. Either teletransportation is not real (the machine just kills you and your doppelgänger wakes up on Mars), or else there is no genuine distinction, after all, between you waking up on Mars and your doppelgänger doing so: any “doppelgänger” of yours who wakes up on Mars is you; seeming to wake up on Mars is waking up on Mars.

It follows that, if teletransportation is real, then you can discover this by trying it out and finding yourself waking up on Mars. So the reality of teletransportation stands to be demonstrated empirically after all.

From ostrich notes to self 4, 58

Some quotes from Michael Cerullo's "Uploading and Branching Identity" (2014).

Nagel coined the term day-person to refer to the possibility that each evening after we go to sleep we die and a new conscious being with false memories awakes the next day (Parfit 1984). There is of course no way to rule out that we are all day-people, or even that we are minute-people that are born a new each minute with false memories. [27, n. 12]‘’

1. The teletransporter

I enter the Teletransporter. I have been to Mars before, but only by the old method, a space-ship journey taking several weeks. This machine will send me at the speed of light. I merely have to press the green button. Like others, I am nervous. Will it work?

– Derek Parfit, Reasons and Persons (1984)

Derek Parfit was not the first to discuss the metaphysical perplexities surrounding teletransportation, fission, fusion, and the like, but his discussions of them were often quite gripping. The passage above, for instance, continues:

I remind myself what I have been told to expect. When I press the button, I shall lose consciousness, and then wake up at what seems a moment later. In fact I shall have been unconscious for about an hour. The Scanner here on Earth will destroy my brain and body, while recording the exact states of all of my cells. It will then transmit this information by radio. Travelling at the speed of light, the message will take three minutes to reach the Replicator on Mars. This will then create, out of new matter, a brain and body exactly like mine. It will be in this body that I shall wake up.

Though I believe that this is what will happen, I still hesitate. But then I remember seeing my wife grin when, at breakfast today, I revealed my nervousness. As she reminded me, she has been often teletransported, and there is nothing wrong with her. I press the button. As predicted, I lose and seem at once to regain consciousness, but in a different cubicle. Examining my new body, I find no change at all. Even the cut on my upper lip, from this morning’s shave, is still there.

Can a person really be teletransported to Mars (or anywhere else) in this way? The question is vexed. Assuming that the relevant physical procedures can be carried out, some philosophers are prepared to regard this as a genuine form of transport: the “fastest way of travelling,” as Parfit aptly puts it. Others, however, consider it nothing so much as a way of killing oneself!

These latter believe that the people who emerge from these “transporters” are not identical to the ones who entered but are just duplicates who look and act like them, and who unsuspectingly think they are them. The truth, they aver, is that the ones who entered died the instant the machines destroyed their brains and bodies, and the ones who emerged are mere “copies.”

There are sensible arguments on both sides of this debate, but nothing decisive has been advanced, and philosophers remain split on the matter. In what follows, I will assume that the debate is broadly familiar by way of background context: see, e.g., the penultimate section of David Chalmers’s, ‘The Singularity: A Philosophical Analysis’ (2010). I don’t really plan on examining it here because I see little prospect of resolving the matter a priori, which is how the debate is usually conducted. My plan, rather, will be to sidestep this armchair debate and ask if the matter might instead be settled empirically. Assuming that we had one of these teletransporters at our disposal, could we not simply try it out and see if the device “worked”? For example, suppose that I get into the machine, press the green button, and promptly find myself waking up on Mars, much as Parfit describes above. Would this not prove, if only to my own satisfaction, that I had indeed been transported to Mars? Empirical considerations like these are somewhat underexplored in the literature and I’d like to show below what we can learn from pursuing them. But first a word about Parfit’s own take on the controversy.

⚹

As is well-known, Parfit has a somewhat nuanced take on the issue. The question is whether the person who emerges from the teletransporter is the same person as the one who entered. Parfit thinks this is the wrong question to ask because he thinks that’s not what we really care about here. According to Parfit, what really matters to us is whether the person who emerges is psychologically continuous with the one who entered. We can see this, he says, if we consider a related case where personal identity and psychological continuity threaten to come apart, this being the traditional conundrum where a man’s brain is divided and each half transplanted into a new body of its own. Which of the two resulting men would be the same person as the original man? This question of identity is known to be just as vexed. There seem to be just three possibilities, “Neither”, “Both,” or “Just the one,” but each of these answers faces well-known difficulties.

Parfit points out two things about this case. First, although the question of personal identity is vexed, the issue of psychological continuity is simple: both men are psychologically continuous with the original one. Second, even without settling the question of personal identity, we can see that this must be a case of “survival” because we know that a man can survive with just half of his brain; what more if he has two stabs at it? So it must be psychological continuity, and not personal identity, that matters for “survival.” In the same way, we need not bother with whether the person who emerges from the teletransporter is the same person as the one who entered: this question is likewise irrelevant. (Parfit thinks it’s a relatively insubstantial verbal issue that may not even have an answer.) What really concerns us is that the person who emerges should be psychologically continuous with the one who entered. Since this is so ex hypothesi, teletransportation is as good as survival.

Parfit develops this account carefully over four chapters of his book; the above is a bare summary. I don’t know that many philosophers have been persuaded by the account, however. Consider, in particular, a case just like the one we started with, except that your body on Earth is not destroyed until an hour after your data has been sent to Mars. (This is Parfit’s “branch-line” case.) So, as your replica comes to life on Mars, you remain alive here on Earth with an hour to spare before you die. You even get to talk to your replica on Mars via video link in this final hour. Parfit ventures that you should regard this case too as one of “survival,” notwithstanding that, here on Earth, you are about to die. For you have a replica on Mars and, even though he is clearly (this time) not you, this should not matter, Parfit says, because what matters (as before) is not that he is the same person as you, but that he is psychologically continuous with you, give or take an hour. This position does not seem very credible, in my opinion, despite Parfit’s vigorous defense of it, and Parfit himself admits that “this is one of the cases where my view is hardest to believe.” (289)

I mention Parfit’s account because it is a well-known point of reference. I will say more about it below by way of contrast with my own attempt to make sense of teletransportation and fission. My account differs not just from Parfit’s, but also from other accounts that I’ve seen in the literature, so I think I want to write it down. As mentioned, the idea is to skirt the armchair debate and ask if the controversy surrounding teletransportation might instead be settled empirically.

---------------

Philosophers and their colours

Modern physics does undoubtedly offer many affronts to common sense: but it does not appear that the denial that objects are coloured is one of them.

– Michael Dummett, ‘Common Sense and Physics’

Is colour merely in the mind or is it really out there in the world?

Some people say that the scientific verdict is clear: colour is in the mind—or perhaps somewhere in the brain—and not “out there” in the world. Objects in themselves are not really coloured; the colours you see are just a construction in your mind created when your eyes and brain are stimulated by light.

Indeed, this view is often taken as established fact in the realm of popular science. For instance, neuroscientist David Eagleman writes:

We think of color as a fundamental quality of the world around us. But in the outside world, color doesn’t actually exist. When electromagnetic radiation hits an object, some of it bounces off and is captured by our eyes. We can distinguish between millions of combinations of wavelengths – but it is only inside our heads that any of this becomes color. Color is an interpretation of wavelengths, one that only exists internally. Show source
The Brain: The Story of You, Pantheon Books 2015, p. 57.

The view is also found in sober academic publications:

We know from psychophysical and neurophysiological investigations that color is created somewhere in the brain, although the exact location of this process is still unknown, and we even have no idea what entities the sensations called color are … In short, colors appear only at a first naïve glance to be located in objects. Source
Backhaus & Menzel, ‘Conclusions From Color-Vision of Insects,’ Behavioral and Brain Sciences, vol. 15, issue 1, p. 28.

As quoted in Byrne & Hilbert, ‘Color Realism and Color Science,’ Behavioral and Brain Sciences, vol. 26, issue 1, p. 3. See this paper for similar quotes by other prominent scientists.

The average layman may not know it, but this is actually an early modern, 17th-century, view of colour (sound, smell, taste, etc.). People like Galileo, Descartes, Newton and Locke all espoused it in one form or another. Galileo famously said:

I think that tastes, odors, colors, and so on are no more than mere names so far as the object in which we place them is concerned, and that they reside only in the consciousness. Hence if the living creature were removed, all these qualities would be wiped away and annihilated. Source
The Assayer 1623, trans. Stillman Drake, online copy here.

Likewise Locke:

The particular bulk, number, figure, and motion of the parts of fire, or snow, are really in them, whether any one’s senses perceive them or no; and therefore they may be called real qualities, because they really exist in those bodies: but light, heat, whiteness or coldness, are no more really in them, than sickness or pain is in manna. Source
Essay Concerning Human Understanding 1690, Book II, Chapter VIII, section 17.

Here’s a less well-known quote from Robert Desgabets:

By the most ancient, the most strange, and the most deeply rooted of errors, we attribute to the external world what are called sensible and corporeal qualities, such as light, heat, taste, odor … Source
Supplément de la Philosophie de M. Descartes 1675. As quoted in the epigraph to Walter Ott’s Descartes, Malebranche and the Crisis of Perception, Oxford 2017.

This view remains common to this day and has even seeped into the popular consciousness. Students, for example, don’t bat an eyelid at the suggestion that colour is only in the mind, and not out there in the world. They regard this as nothing but enlightened common sense!

You would not know this, however, if you were to consult a bunch of contemporary philosophers. The view that colour is only in the mind, and not out there in the world is often considered by contemporary philosophers to be—well—a 17th century view! The following jibe from Byrne and Hilbert captures the tone:

Although contemporary color science would be quite unrecognizable to Galileo, this is one respect in which he is perfectly up to date. Source
‘Color Realism and Color Science,’ Behavioral and Brain Sciences, vol. 26, issue 1, p. 4.

Or consider this expression of dismay:

According to projectivists, science has revealed something very important about colours, something which used to be understood clearly, but which recent philosophers have failed to grasp. People with colour vision … represent the world as containing objects, spatially separated from us, which have various different colours. Projectivists say that when people represent things to themselves that way, they are misrepresenting the world. Material objects do not have the colour properties we represent them as having …

According to the projectivists, Galileo, Newton, Locke, and others were right about colours. Armstrong, Smart, Shoemaker, Peacocke, McDowell, Wiggins, McGinn, and many others among our contemporaries are all wrong. Source
John Bigelow, John Collins & Robert Pargetter, ‘Colouring in the World,’ Mind 1990

The contemporary philosophers cited here (Armstrong, Smart, ...) all believe that colours are genuine features of the world at large, siding largely with common sense on this score against the likes of Galileo, Eagleman and the others cited above. According to Armstrong, for example:

For myself, I think that the only plausible way that a Materialist can deal with the secondary qualities is to completely reverse the whole programme started by Galileo, a programme that has persisted for so long. What we should do is put these qualities back into the physical world again … Source
The Mind-Body Problem: An Opinionated Introduction, 1999, p. 124.

They disagree over some details but they are united in rejecting the broad Galilean idea that colours are mere “projections” of the mind onto the external world, i.e., features that the mind “reads into” the world that are not really there. The Galilean idea may draw some superficial support from science, they would concede, but, over the centuries, philosophers have learnt that the idea is not as easy to sustain as one might first think. So what happened? How did it get to be like this?

This is a long and complex story but I will try and give an idea of it in as concise a way as I can. The number of positions one can take on colour is quite surprising: it is not as simple as just colour is in your mind or out there in the world. I will sketch the main positions and some of their motivations.

Let's start with philosophers who demur at the oddness of saying things like roses are not red, or that grass is not green, or that the sky is not blue. This goes against ordinary language, they say, and is unnecessarily radical. Our colour language serves a useful function and anyone who gives it up will soon find themselves forced to invent a replacement language to serve the very same function. Suppose for instance that you were convinced by colour science that roses are not really red, grass is not really green, and so on. You would soon find yourself using new words like "schmred", "schgreen", and so on. Roses are not red, but they are schmred, you point out, meaning that induce in us a certain red sensation. A philosopher will now point out to you that the English word "red" actually functions in the way that your "schmred" functions, so you did not have to invent a new word in the first place.

The matter is not quite as settled though as some would have you believe. Among philosophers, the range of views that have been taken on colour is somewhat more diverse. Some philosophers do continue to hold something like the early modern of colour mentioned. An example would be Thomas Metzinger:

But there is another variant of the view, on which colours do not exist anywhere, not even in the mind. They are just wholesale illusions. Larry Hardin:

A slightly different view is to say that the term "colour" has two senses: a physical one and a mental one, loosely speaking. Thus objects are coloured in one sense, but colours "as we see them" belong in the mind. Thomas Reid held this view and so does Brian McLaughlin:

Is this just a verbal dispute?

------------------

Older entry

Essay on colour

1. The commonsense view

The phenomenon of colour is a fairly perplexing one. While we know a great deal about it through centuries of inspired scientific detective work, some disconcertingly basic aspects of the phenomenon continue to remain mysterious, shrouded in (as yet) unresolved issues surrounding the larger mind-body problem. In a review of a recent study of colour, writer Malcolm Harris remarks:

To read [this book] as a layman feels like being let in on a shocking secret: Neither scientists nor philosophers know for sure what color is.
—‘Does Color Even Exist?’ (2015), www.newrepublic.com

This sentiment seems to me pretty accurate.

At first sight, colour appears to be a simple property of material things in our environment. We speak of leaves being green and flowers being yellow much as we speak of them having various shapes and sizes. That is to say, like shape and size, we quite naturally take the colours we see to be features of our environment that exist independently of our perceptions of them. Asked to imagine a world devoid of perceivers, for example, I believe that many people would quite naturally imagine it in colour, i.e., just as they would see it. As Colin McGinn says:

… when we see an object as red we see it as having a simple, monadic, local property of the object’s surface. The color is perceived as intrinsic to the object, in much the way that shape and size are perceived as intrinsic.
—‘Another Look at Color’ (1996), p. 132

Alison Simmons puts it this way:

The proposal that there is an ontological difference between a kumquat’s shape and its color does not suggest itself to perceptual experience: both look to be out there in the kumquat.
—‘Perception in Early Modern Philosophy’ (2013), p. 188

That colour is an objective, intrinsic property of the things around us may reasonably be called the commonsense view of colour. This means that the view may fairly be attributed to the average layman—at least before he or she becomes “tainted” by the modern scientific account of colour perception in the brain (see below). Aristotle himself, that ancient embodiment of common sense, seemed to embrace the view, as evinced by his speculations concerning how the eye perceives colour:

If what has colour is placed in immediate contact with the eye, it cannot be seen. Colour sets in movement not the sense organ but what is transparent, e.g. the air, and that, extending continuously from the object to the organ, sets the latter in movement.
—De Anima (c. 350 BC), Book 2, Chapter 7

This account of colour perception is no longer credible but Aristotle did seem to accept that colour was an “objective feature” of the world, capable (via the air) of affecting our eyes, whereupon it is seen. Sarah Broadie attests as much:

Aristotle speaks as an unabashed realist about objects of sense. He never questions the existence of what later came to be called the “external” world: a cosmos of physical substances there independently of our knowledge and perception. And he never doubts that the objects composing that world really are as they present themselves to us in sense experience: pungent, fragrant, warm and cold, soft and hard, full of sounds and colors, just as we perceive them to be.
—‘Aristotle’s Perceptual Realism’ (1993), p. 137

The objectivity of colour continued to figure as common sense for more than a thousand years thereafter, as Peter King reminds us:

... during the Middle Ages colors, sounds, and smells were all thought to be unproblematically qualitative features of the external world, no more having a special and inexplicable ‘phenomenal’ dimension than, say, the shape of the chair.
—‘Why isn’t the Mind-Body Problem Medieval?’ (2007), p. 204

So much for a rough statement of the commonsense view of colour. One might no doubt ask further questions of it, e.g., what colour is supposed to be, on this view, over and above something “out there” in the world. When the layman affirms that blood is red, for example, what does he suppose this property of redness to be? What property does he mean to be attributing to blood when he says that blood is red? It took a long time for natural philosophers to make any headway with a question like this (see below); common sense itself has no answer beyond a shrug of the shoulders and the solace that we can recognize the property when we see it.

But never mind this—for the commonsense view of colour is (in any case) very much on the backfoot these days, as most people know. It has come under sustained fire over the past four centuries from various scientific discoveries that seem to suggest that, whatever common sense may aver, there is in fact no such thing as colour (sound, smell, taste, etc.) in the world at large. The world at large consists rather of “qualitatively bland” particles and fields vibrating in space and time – or whatever the latest science says – with no apparent room in it, or even need, for any such quality as colour.

Thus, when you look at a so-called red apple, the apple reflects selected wavelengths of light into your eyes and its “redness” is then generated in your brain as part of your visual experience of the apple. The apple itself is not red in the way you see it to be; the redness is just an illusory product of your mind.

This broad scientific picture is nowadays so familiar and entrenched that one might be surprised to discover that many contemporary philosophers still uphold the commonsense view of colour – the view that colour is “out there” in the world, just as Aristotle, Alhazen and Aquinas supposed it to be. One might well wonder why. Are contemporary philosophers mired in the middle ages, in denial over the findings of enlightenment science? What is the truth about colour anyway? Is it simply “in the mind” or “out there” in the world?

In what follows, I’d like to do three things. The first is to rehearse the broad scientific considerations that have persuaded many people since at least the 17th century that colour is just “in the mind” or is merely “a product of our brains,” as opposed to being “out there” in the world as common sense would have it. This tale is pretty well-worn so I will run through it just to set things up. In recent decades, even more sophisticated considerations about colour-processing in the brain have added fuel to this scientific fire but these additional sophistications (e.g., opponent processing) won’t matter much for my purposes and I will mention them only when they matter.

Second, I’ll explain why, in the face of these considerations, philosophers nevertheless often cling to the commonsense view of colour. Scientists and others often find this hard to understand because philosophers don’t always bother to explain where they are coming from or what they are up to; they also sometimes speak in gnomic terms – see the cartoon. So it may be worth pausing to explain this to the general reader. There is in fact a clear point behind the philosophical resistance to the alleged findings of

science – a point that scientists often seem to be unaware of – but (on the flip side) I will contend also that philosophers, for their part, do not ultimately succeed in rejuvenating the view that colours are “out there” in the world.

To anticipate this last claim a little, a certain familiar difficulty has always bedevilled philosophical attempts to “restore” colours to the external world, viz., that it makes no apparent sense to identify colours (as plain folk know and love them) with physical properties of objects like surface spectral reflectances or microstructural properties, and so on – this being the standard way of “restoring” colours to the external world. The entities being equated seem to belong to completely different categories. No matter how philosophers twist and turn, it seems to me that they are unable to shake off this essential difficulty: in one form or another, it haunts their every attempt to restore colour to the external world. I will illustrate this below with numerous examples of the twisting and turning and I hope that this throws some light on what philosophers are up to because these twists and turns are often unannounced in point of both existence and purpose. My overall conclusion will be that while philosophers are right to resist the scientists’ inclination to relegate colour to the mind, they don’t really have a good way of restoring colour back to the world either. The result is a kind of impasse or aporia in which the place of colour in the natural world is left mysterious. As the Danish physicist Niels Bohr once said in a slightly different context, some essentially new ideas are needed here.

Finally, I’ll say what I think the truth about colour is. At least, I’ll sketch the outlines of an essentially new idea. Whether colour is in the mind or out there in the world is a slightly disingenuous question, of course, since other options may be possible. Some people think of colour as a relation of some sort between mind and world. For example, the redness of an object is simply its disposition to produce certain characteristic kinds of visual sensations in normal human perceivers. On a view like this, the choice is false of whether colour is in the mind or in the world, since, it some sense, it partakes of both. The disadvantage of this approach however is that it continues to run into the familiar difficulty

mentioned above: there is no obvious sense to the claim that colours, as plain folk know them, are just relations or dispositions of a certain kind.

I will propose a completely different way of thinking about colour instead. The colour of an object, I will suggest, has something essentially to do with the way in which the object “stands out” in perception, i.e., with its “perceptual salience,” as I will sometimes say. This notion may seem unfamiliar but many everyday words in our language in fact revolve around it. A bright light, for example, is one which is easily noticed (perceptually very salient) as opposed to a dim one. Likewise, material objects can be bright, luminous or phosphorecent; dark, dull or unreflective. The paint on your walls can be glossy or matt and clothes can be understated or loud. These adjectives apply to everyday things but the point of having them in our language is not so much to characterize the things in themselves as the manner in which they strike us perceptually. There is indeed a diverse family of such words, ranging over all five senses, and functioning in quite different ways – they are not all mere variants of “perceptually conspicuous” and “perceptually inconspicuous,” it should be stressed. A gunshot or a scream may be loud or faint, but it may also be muffled. A star in the sky may be bright or dim, but it may also blink or twinkle. Glass is transparent, which says something not just about the difficulty of seeing it, but also about the ease of seeing what’s behind. Other things are opaque or diaphanous, even invisible. Perfume can be overpowering, food can be tasteless, chameleon-like creatures blend into the background. And so on – it’s a fairly wide-ranging list. I believe that colour terms like ‘red,’ ‘green’ and ‘blue’ belong in this list as well, occupying a special niche of their own. I’ll introduce this novel view fully below and explain why I like it. To see this at once, jump to sections 6-10.

As mentioned, however, I’m going to begin with the broad scientific attack on the commonsense view of colour because everything unfolds pretty naturally from there. I characterized the commonsense view above as holding that colours are objective, intrinsic properties of everyday things in our environment, but I should now mention that some philosophers think that this takes common sense too far. Take (ripe) strawberries, for example. Arguably, all that may safely be attributed to common sense, where colour is concerned, is simply that strawberries are red, i.e., that the statement, ‘Strawberries are red,’ is perfectly true. Common sense need take no further stand on whether the redness is “out there” on the strawberries, or is one of their “intrinsic properties,” or anything remotely metaphysical like that. Here’s a recent expression of a similar point:

... it does not seem so implausible that there is a pan-historically and cross-culturally universal commonsense view that holds that colors are properties of ordinary objects or their surfaces, without making any claims about the perceiver-dependent or independent nature of those properties.
—M. Chirimuuta, Outside Color (2015), p. 32

As a methodological principle, many philosophers think that we should not contradict common sense if we can avoid it – I will explain why below. So a satisfactory account of colour should ideally entail (at least) that strawberries are red (bananas are yellow, grass is green, etc.), just as minimal common sense avers. This “thin” reading of common sense will serve our purposes as well because the broad scientific attack on common sense that we are about to consider does deny even that strawberries are red, and so really does amount to an attack on common sense even thus thinly construed.

2. The early modern view

Some time in the 17th century, natural philosophers began to distinguish the “primary qualities” of material objects from their “secondary” ones.

Very roughly, they suspected that secondary qualities (like colour, sound, smell and taste) were not genuine features of material objects at all, despite initial appearances, but were mere artefacts of sensory perception, misleadingly “projected” by our minds onto the external world. In contrast, primary qualities (like shape, size, motion and number) were deemed to be genuine features of material objects, faithfully represented by our minds in sensory perception.

These must have been extraordinary claims at the time but, over the centuries, this broad scientific view has been gradually developed, come to be widely known, and even passed into popular culture – students no longer bat an eyelid at the suggestion that colour, unlike shape, is “only in the mind.” Indeed, I think it is fair to say that, very broadly speaking, the scientific jury has already made its mind up on this question although, as hinted at above, the philosophical jury is still very much out.

But let’s consider the basic science first.

Writing in 1936, the German psychologist Wolfgang Metzger had this to say about the ordinary attitude towards seeing:

For people who naively look around, their own eyes appear to be a kind of window. As soon as the curtains, the eyelids, are opened there “is” a visible world of things and of other beings out there. Nothing could arouse the suspicion that any of its recognizable properties might originate in the observer or could be codetermined by the observer’s nature – except perhaps for the effects of greater or lesser transparency of the “window panes.” (Wolfgang Metzger, The Laws of Seeing, MIT 2006 English translation, Introduction.)

In this passage, Metzger acknowledges the “transparency” or “diaphanousness” of seeing (as philosophers often call it) but regards it as a potentially misleading thing. In particular, it makes us overlook the possibility that we do not see things as they are “in themselves,” but rather contribute ourselves in some way (by the very act of seeing them) towards their appearances. Metzger was thinking inter alia about the feature of colour:

Colors are now recognized as the result of external influences on particular parts of the human body. They are called “sensations” in order to express that they belong together with heat or pressure, which one really feels on one’s own body when touching a hot iron or carrying a heavy sack on one’s back. Nevertheless it is strange that these odd sensations stubbornly seem to cling to the external surfaces of objects and cannot, not even with considerable effort, be sensed (like the pain of blinding light) as within the eyes. But in view of the incontrovertible evidence of physics, we cannot be fooled into thinking that light rays (and even light sources) are themselves colored. They only evoke the experience of certain colors “within” us depending on their wavelength, and therefore the colors themselves are a property of the human nature (i.e., the brain, not the external world). (Ibid.)

Metzger was writing in the twentieth century but this way of thinking about colour goes back to Galileo in the seventeenth, who was one of the first great scientists to articulate the view:

I think that tastes, odors, colors, and so on are no more than mere names so far as the object in which we place them is concerned, and that they reside only in the consciousness. Hence if the living creature were removed, all these qualities would be wiped away and annihilated. (The Assayer 1623, trans. Stillman Drake, online copy here.)

The broad idea is that colour is “in the mind” and not really a property of the things in our environment, as is natural to think. Alternately, it is said that colour is in the consciousness, or in the head, or in the brain, or even in the eyes, and so on. (See later. I will say “in the mind” for now.) Similar sentiments may be found among many other early modern philosophers from Descartes and Newton to Locke and Hume, and the view continues to be found in contemporary psychology texts. Here is psychologist Stephen Palmer in a work from 1999:

People universally believe that objects look colored because they are colored, just as we experience them. The sky looks blue because it is blue, grass looks green because it is green, and blood looks red because it is red. As surprising as it may seem, these beliefs are fundamentally mistaken. Neither objects nor lights are actually ‘colored’ in anything like the way we experience them. Rather, color is a psychological property of our visual experiences when we look at objects and lights, not a physical property of those objects or lights. The colors we see are based on physical properties of objects and lights that cause us to see them as colored, to be sure, but these physical properties are different in important ways from the colors we perceive. (Vision Science: Photons to Phenomenology, MIT Press 1999, p. 95.)

One reason why scientists tend to uphold a view of this sort is that they seem to have no use for any such property as “colour” (as we normally understand this) in explaining how the inanimate world works. The inanimate world, in the eyes of science, is a world of myriad moving lifeless things of different shapes, sizes and other such “primary qualities.” Various laws govern how they interact with one another but no law ever turns on the “colour” of anything. For example, the law of gravity says that the force with which two bodies attract each other depends on their masses and distance apart, but hardly on their colour. (And so on – no law of physics ever mentions the property of colour.)

Even in cases where colour might at first seem to figure in a scientific generalization, the reference to colour turns out to be otiose. Black things might heat up faster in the sun than white ones but this can be explained without mentioning their colours at all but by their relative tendencies to absorb or reflect certain kinds of radiation. Here’s another example from D. M. Armstrong:

Does the high-pitched sound shatter the glass? To superficial observation it might appear to do so. But investigation reveals that it is vibrations in the air which really break the glass. (‘Smart and the Secondary Qualities.’)

As far as science is concerned, in other words, there might as well be no such thing as colour (sound, smell, etc.) in the world “out there.” Sydney Shoemaker sums this up very well:

Since the 17th century, a central issue in metaphysics and epistemology has concerned the status of what, following John Locke, have come to be called “secondary qualities”: colors, odors, tastes, sounds, warmth and coldness, etc. The problem is posed by the fact that while these qualities are experienced as belonging to objects in our external environment – the apple is experienced as red, the rose as fragrant, the lemon as sour, etc. – the scientific world-view, as developed in the 17th century and subsequently, seems to allow no place for them among the objective properties of material bodies. A common solution has been to deny, in one way or another, that these qualities do objectively inhere in external things in the way we experience them as doing. Galileo seems to have held that colors, etc., are properties of our sensations rather than of external things. (‘Qualities and Qualia: What’s in the Mind?,’ Philosophy and Phenomenological Research, vol. 50, 1990, opening paragraph.)

Indeed, where science is concerned, colour needs to come into it only when sentient beings enter the picture. Put a man in a field and you now need to explain why he sees the field as being green. Without the man’s act of seeing, the “greenness” of the field would never get a mention. Colour makes an appearance only when mind encounters world, but not before. And so it is natural to think that colour has something to do with the way in which the mind “receives” or “represents” the world and is not a feature of the external world as such, i.e., the world apart from the mind.

As mentioned previously, this kind of picture has now infiltrated popular culture to the extent of threatening to be the new common sense. I don’t think it has quite reached that stage but it is getting pretty close. Roughly speaking, colourless things out there in the world send signals to our brains (through our eyes) and colours are then conjured up in our brains which we mistakenly project onto the things that we see. The colours are only ever “in our minds” but an unthinking person (the common man) may fail to realize this. This is the broad picture.

It is worth comparing this early modern conception of colour perception to the Aristotelian/medieval conception that preceded it. Aristotle’s conception was briefly mentioned above:

Colour sets in movement not the sense organ but what is transparent, e.g. the air, and that, extending continuously from the object to the organ, sets the latter in movement.

Descartes

There are two basic reactions one might have to this broad picture. Some consider it a profound discovery of science that fully deserves to be passed down to the common man. This is no doubt the view of Palmer, quoted above, and numerous others eminent scientists as well. Here’s a similar sentiment from neurologist Semir Zeki, quoted in Byrne & Hilbert, ‘Color realism and color science’ (2003):

The results described here ... suggest that the nervous system, rather than analyze colours, takes what information there is in the external environment, namely, the reflectance of different surfaces for different wavelengths of light, and transforms that information to construct colours, using its own algorithms to do so. In other words, it constructs something which is a property of the brain, not the world outside. (‘Color coding in the visual cortex’ 1983, p. 764.)

Others however consider this broad picture to be a piece of misguided metaphysics amounting to a pretty serious error. For those in the second camp – and here reside many contemporary philosophers – the following quote captures something of the dismay and despair:

According to projectivists, science has revealed something very important about colours, something which used to be understood clearly, but which recent philosophers have failed to grasp. People with colour vision represent the world to themselves in a specific way. They represent the world as containing objects, spatially separated from us, which have various different colours. Projectivists say that when people represent things to themselves that way, they are misrepresenting the world. Material objects do not have the colour properties we represent them as having. [Footnote: According to the projectivists, Galileo, Newton, Locke, and others were right about colours. Armstrong, Smart, Shoemaker, Peacocke, McDowell, Wiggins, McGinn, and many others among our contemporaries are all wrong.] (Bigelow, Collins & Pargetter, ‘Colouring in the World.’)

The contemporary philosophers cited here (Armstrong, Smart, Shoemaker, Peacocke, ...) all believe that colours are genuine features of the world at large, siding largely with common sense on this score against the likes of Palmer and Zeki above. They disagree over some details—see below—but they are united in rejecting the broad Galilean idea that colours are mere “projections” of the mind onto the external world, i.e., features that the mind “reads into” the world that are not really there. The Galilean idea may draw some superficial support from science, they would concede, but, over the centuries, philosophers have learnt that the idea is not as easy to sustain as one might first think. In fact, it is the thin edge of that wedge known nowadays as the “hard problem of consciousness.” And so it behooves us to consider whether colours might not be better reinstated in the external world, where they seemed in the first place to be.—This is the broad stance.

To fully appreciate this point of view, we need to unravel a few historical threads that are often glossed over in the contemporary philosophical literature on colour – possibly because they are regarded as well-known. It is convenient to start with the writings of the 17th-century English philosopher John Locke, who lived in the same era as Galileo, Descartes, Boyle and Newton, and was deeply immersed in the revolutionary scientific discoveries of his age, while being an out-and-out philosopher to boot. His pregnant remarks in his Essay Concerning Human Understanding (1690) on the “secondary qualities” (including colour) were among the most sophisticated of his time and have given philosophers plenty to chew on since then.

--- OLD ---
1. The commonsense view

The phenomenon of colour is a fairly perplexing one. Although we know a great deal about it through centuries of heroic scientific detective work, some rather basic aspects of the phenomenon continue to remain mysterious, shrouded as they are in larger, unresolved issues surrounding the perennial mind-body problem. (See below.)

At first sight, colour appears to be a simple (monadic) property of material things in our environment. We speak of leaves being green and buses being yellow much as we speak of them as being thusly shaped, sized or numbered. Indeed, like shape, size and number, we quite naturally take the colours we see to be features of our environment that exist independently of our perceptions of them. Asked to imagine a world devoid of perceivers, for example, I believe that many people would quite naturally imagine it in colour, i.e., just as they would see it. Colin McGinn puts it this way:

... when we see an object as red we see it as having a simple, monadic, local property of the object’s surface. The color is perceived as intrinsic to the object, in much the way that shape and size are perceived as intrinsic. No relation to perceivers enters into how the color appears; the color is perceived as wholly on the object, not as somehow straddling the gap between it and the perceiver. (‘Another Look at Color,’ Journal of Philosophy, vol. 93, 1996.)

Alison Simmons uses a more exotic example:

The proposal that there is an ontological difference between a kumquat’s shape and its color does not suggest itself to perceptual experience: both look to be out there in the kumquat. (‘Perception in Early Modern Philosophy,’ p. 188.)

This ordinary attitude towards colour may be mistaken – or not; see below – but it is certainly understandable. For, whenever we see, we normally just “take in” whatever we see, paying scant attention to the fact that we are engaged in an act of seeing. Our attention is normally focused on the things that we see, rather than on the fact that we see them. Indeed, we tend to regard the act of seeing itself as doing little more than “enabling” us to know various aspects of our environment (which is what usually interests us). In particular, under good viewing conditions, we do not imagine that the act of seeing in any way distorts the nature of whatever we see, in the way that stepping into a puddle to gauge its depth (say) might well alter that depth. And so when we see a coloured thing, we naturally take the colour we see to be an intrinsic feature of the thing – the act of seeing the thing, we are apt to suppose, simply “lets us know” what colour it is.

Some philosophers speak in this vein of the “transparency” or “diaphanousness” of seeing, often meaning thereby to endorse the ordinary attitude towards seeing just described. Where colour is concerned, in particular, these philosophers are apt to agree with the common man that our physical environment is coloured in pretty much the way we see it to be. Michael Watkins expresses this vividly in his book, Rediscovering Colors: A Study in Pollyanna Realism:

Pollyanna, the heroine of my story, endorses a simple theory. It claims that objects are colored just as they are shaped; typically, objects appear to have the colors they have because of the colors they have; objects were colored before there were observers to perceive those colors; and objects have their colors in the dark, when no one is watching, and sometimes (though not usually) when they appear to have some color they do not have. Colors, Pollyanna believes, are intrinsic, observer-independent, properties of objects that are causally responsible for objects appearing colored. Colors, in other words, are just as Pollyanna experiences them most of the time. Pollyanna doesn’t have much to say about colors. She doesn’t worry about the ontological status of colors, or whether objects are really colored, or whether colors are properties of objects or only “in her head.” Why worry? The answers seem obvious. Pollyanna can see that objects are colored and she can see that they are really colored too. Indeed, Pollyanna can’t help but believe that objects are colored and she suspects that’s true of everyone else (although she acknowledges, of course, that some believe that they don’t believe it). Once upon a time, almost everyone agreed with Pollyanna. Those were happy days, before the dark days of scientism, before philosophers lost their metaphysical nerve. (Rediscovering Colors: A Study in Pollyanna Realism, Springer 2002, pp. 1-2.)

As Watkins laments, however, a commonsense view of this sort has found itself increasingly on the backfoot over the centuries. It has come under attack primarily from various scientific discoveries which seem to suggest that there is no such thing as colour in the world at large. The world at large is supposed to consist rather of nothing but particles and fields vibrating in space and time – or something like that, depending on the latest science – with no apparent room in it (or even need) for anything like colour. A scientific picture of this sort is nowadays so familiar that philosophers who continue to defend the commonsense view of colour may be suspected of being in denial over the findings of science. Worse, these philosophers also squabble over the “correct way” to defend the commonsense view and this is also part of Watkins’ lament above, for his way of defending the view is considered even in philosophical circles to be particularly audacious, as we will see.

In what follows, I would like to do two things. The first is to explain why philosophers often tend to defend the commonsense view of colour in the face of the apparently strong scientific case to the contrary.

Scientists (and other interested observers) sometimes find this hard to understand because philosophers don’t always bother to explain (except in “insider’s language”) where they are coming from and what they are up to. So I think it may be worth pausing to explain in somewhat plainer English what these philosophers take themselves to be doing. I will also explain some of their internal squabbles over “how best” to defend the commonsense view.

For better or worse, I will not cover this in great detail. The philosophical literature on colour is vast and confusing and there is no real point to scouring it in detail. I will say enough to enable a scientist (say) to make substantial sense of it, and also enough to justify a certain contention of mine, viz., that while much of this literature is interesting in its own right, it also tends (disappointingly) to bypass the really “hard issue” when it comes to the phenomenon of colour. The hard issue is essentially that of understanding where colour fits into the natural world. Our current scientific worldview seems to have no natural home for colour, so what is this thing we call colour and where does it fit into the overall scheme of things? Note that the issue is not resolved simply by declaring, as a scientist might, that colour is a sensory illusion that exists “only in the mind” – for what is the mind and what is it for something to exist “in” the mind? If you look inside the brain, for example, you won’t find any colours there, so what are these colours and where are they hiding? This is one place where philosophers are one step ahead of scientists because they more clearly see the difficulties involved in trying to “relegate” colour to the mind. Scientists tend to assume that they will eventually figure out the mind and sort this residual issue out but philosophers have had their eye on this “residual issue” for a long time now and know it to be a bit of a Gordian knot.

So when philosophers talk about colour, they are often wrestling with this Gordian knot – the “hard issue” of colour; the issue of finding a comfortable home for colour in the natural world. At least, that’s one of the things they are supposed to be doing, whatever else they may have to say about colour. But, as I said, their writings on this score often tend to disappoint. Many philosophers of colour give every appearance of addressing the hard issue without in my opinion really doing so. All too often, certain (related) “easy questions” are raised and addressed while the really “hard question” is skirted. A similar thing happens with philosophical work on consciousness, as is well-known. David Chalmers famously remarked that much of this work addresses the “easy problems” of consciousness without tackling the really “hard” one. I believe that something similar is true of much philosophizing about colour. It is easiest to make this charge out in the same breath as explaining the specific moves that philosophers make when they talk about colour, so that’s what I’ll do below. (See sections 2 to 5.)

The second thing I will do is to set the philosophical literature aside and grapple directly with the hard issue myself. I will suggest a relatively novel way of “locating” colour in the natural world that is not found in the literature. It is not found there partly because, as mentioned, the literature tends to skirt the hard issue, but also (I think) because it is not an approach that comes readily to mind – it is not the first thing you would think of. Briefly, however, there is an offbeat way of approaching the phenomenon of colour that has largely been overlooked but it is of a sort that I think we need to consider if we are to properly meet the challenge of finding a place for colour in the overall scheme of things. (See sections 6 to 10.)

We should begin however with the broad scientific attack on the commonsense view of colour because everything unfolds pretty naturally from there. Some time in the 16th century, natural philosophers began to distinguish the “primary qualities” of material objects from their “secondary” ones. Very roughly, they suspected that the secondary qualities (like colour, sound, smell and taste) were not genuine features of material objects at all, despite initial appearances, but were mere artefacts of sensory perception, misleadingly “projected” by our minds onto the external world. In contrast, the primary qualities (like shape, size, motion and number) were deemed to be genuine features of material objects, faithfully represented by our minds in sensory perception. These must have been extraordinary claims at the time but, over the centuries, this broad scientific view has been gradually developed, come to be widely known, and even passed into popular culture – students no longer bat an eyelid at the suggestion that colour, unlike shape, is “only in the mind.” Indeed, I think it is fair to say that, very broadly speaking, the scientific jury has already made its mind up on this question although, as hinted at above, the philosophical jury is still very much out.

But let’s consider the basic science first and save the philosophy for later.

2. The early modern view

Writing in 1936, the German psychologist Wolfgang Metzger had this to say about the ordinary attitude towards seeing:

For people who naively look around, their own eyes appear to be a kind of window. As soon as the curtains, the eyelids, are opened there “is” a visible world of things and of other beings out there. Nothing could arouse the suspicion that any of its recognizable properties might originate in the observer or could be codetermined by the observer’s nature – except perhaps for the effects of greater or lesser transparency of the “window panes.” (Wolfgang Metzger, The Laws of Seeing, MIT 2006 English translation, Introduction.)

In this passage, Metzger acknowledges the “transparency” or “diaphanousness” of seeing (discussed briefly in the previous section) but regards it as a potentially misleading thing. In particular, it makes us overlook the possibility that we do not see things as they are “in themselves,” but rather contribute ourselves in some way (by the very act of seeing them) towards their appearances. Metzger was thinking inter alia about the feature of colour:

Colors are now recognized as the result of external influences on particular parts of the human body. They are called “sensations” in order to express that they belong together with heat or pressure, which one really feels on one’s own body when touching a hot iron or carrying a heavy sack on one’s back. Nevertheless it is strange that these odd sensations stubbornly seem to cling to the external surfaces of objects and cannot, not even with considerable effort, be sensed (like the pain of blinding light) as within the eyes. But in view of the incontrovertible evidence of physics, we cannot be fooled into thinking that light rays (and even light sources) are themselves colored. They only evoke the experience of certain colors “within” us depending on their wavelength, and therefore the colors themselves are a property of the human nature (i.e., the brain, not the external world). (Ibid.)

This way of thinking about colour goes back to Galileo, who was one of the first great scientists to articulate the view:

I think that tastes, odors, colors, and so on are no more than mere names so far as the object in which we place them is concerned, and that they reside only in the consciousness. Hence if the living creature were removed, all these qualities would be wiped away and annihilated. (The Assayer 1623, trans. Stillman Drake, online copy here.)

People universally believe that objects look colored because they are colored, just as we experience them. The sky looks blue because it is blue, grass looks green because it is green, and blood looks red because it is red. As surprising as it may seem, these beliefs are fundamentally mistaken. Neither objects nor lights are actually ‘colored’ in anything like the way we experience them. Rather, color is a psychological property of our visual experiences when we look at objects and lights, not a physical property of those objects or lights. The colors we see are based on physical properties of objects and lights that cause us to see them as colored, to be sure, but these physical properties are different in important ways from the colors we perceive. (Vision Science: Photons to Phenomenology, MIT Press 1999, p. 95.)

Does the high-pitched sound shatter the glass? To superficial observation it might appear to do so. But investigation reveals that it is vibrations in the air which really break the glass. (‘Smart and the Secondary Qualities.’)

As far as science is concerned, in other words, there might as well be no such thing as colour (sound, smell, etc.) in the world “out there.” Sydney Shoemaker sums this up very well:

Since the 17th century, a central issue in metaphysics and epistemology has concerned the status of what, following John Locke, have come to be called “secondary qualities”: colors, odors, tastes, sounds, warmth and coldness, etc. The problem is posed by the fact that while these qualities are experienced as belonging to objects in our external environment – the apple is experienced as red, the rose as fragrant, the lemon as sour, etc. – the scientific world-view, as developed in the 17th century and subsequently, seems to allow no place for them among the objective properties of material bodies. A common solution has been to deny, in one way or another, that these qualities do objectively inhere in external things in the way we experience them as doing. Galileo seems to have held that colors, etc., are properties of our sensations rather than of external things. (‘Qualities and Qualia: What’s in the Mind?,’ Philosophy and Phenomenological Research, vol. 50, 1990, opening paragraph.)

Indeed, where science is concerned, colour needs to come into it only when sentient beings enter the picture. Put a man in a field and you now need to explain why he sees the field as being green. Without the man’s act of seeing, the “greenness” of the field would never get a mention. Colour makes an appearance only when mind encounters world, but not before. And so it is natural to think that colour has something to do with the way in which the mind “receives” or “represents” the world and is not a feature of the external world as such, i.e., the world apart from the mind.

As mentioned previously, this kind of picture has now infiltrated popular culture to the extent of threatening to be the new common sense. I don’t think it has quite reached that stage but it is getting pretty close. Roughly speaking, colourless things out there in the world send signals to our brains (through our eyes) and colours are then conjured up in our brains which we mistakenly project onto the things that we see. The colours are only ever “in our minds” but an unthinking person (the common man) may fail to realize this. This is the broad picture.

There are two basic reactions one might have to this broad picture. Some consider it a profound discovery of science that fully deserves to be passed down to the common man. This is no doubt the view of Palmer, quoted above, and numerous others eminent scientists as well. Here’s a similar sentiment from neurologist Semir Zeki, quoted in Byrne & Hilbert, ‘Color realism and color science’ (2003):

The results described here ... suggest that the nervous system, rather than analyze colours, takes what information there is in the external environment, namely, the reflectance of different surfaces for different wavelengths of light, and transforms that information to construct colours, using its own algorithms to do so. In other words, it constructs something which is a property of the brain, not the world outside. (‘Color coding in the visual cortex’ 1983, p. 764.)

According to projectivists, science has revealed something very important about colours, something which used to be understood clearly, but which recent philosophers have failed to grasp. People with colour vision represent the world to themselves in a specific way. They represent the world as containing objects, spatially separated from us, which have various different colours. Projectivists say that when people represent things to themselves that way, they are misrepresenting the world. Material objects do not have the colour properties we represent them as having. [Footnote: According to the projectivists, Galileo, Newton, Locke, and others were right about colours. Armstrong, Smart, Shoemaker, Peacocke, McDowell, Wiggins, McGinn, and many others among our contemporaries are all wrong.] (Bigelow, Collins & Pargetter, ‘Colouring in the World.’)

John Locke (1632-1704) is often regarded as one of the great enlightenment thinkers. He taught himself medicine, wrote widely on economics and politics, epistemology and the philosophy of mind – and, where the specific issue of secondary qualities was concerned, accepted and championed the broad Galilean outlook described in the previous section.

Locke’s remarks on colour and the other secondary qualities are found in chapter 8 of book 2 of his Essay, a chapter innocently titled, ‘Some further considerations concerning our simple ideas.’ This chapter contains twenty-six short sections and is well-known to philosophers of colour, who have pored over Locke’s words with a fine-toothed comb. That Locke accepted the broad Galilean outlook is not in doubt and evident from passages like these:

What I have said concerning colours and smells may be understood also of tastes and sounds, and other the like sensible qualities; which, whatever reality we by mistake attribute to them, are in truth nothing in the objects themselves, but powers to produce various sensations in us, and depend on those primary qualities, viz. bulk, figure, texture, and motions of parts; as I have said. (Section 14.)

The particular bulk, number, figure, and motion of the parts of fire, or snow, are really in them, whether any one’s senses perceive them or no; and therefore they may be called real qualities, because they really exist in those bodies: but light, heat, whiteness or coldness, are no more really in them, than sickness or pain is in manna. Take away the sensation of them; let not the eyes see light, or colours, nor the ears hear sounds; let the palate not taste, nor the nose smell; and all colours, tastes, odours, and sounds, as they are such particular ideas, vanish and cease, and are reduced to their causes, i.e. bulk, figure, and motion of parts. (Section 17.)

For all the brilliance of Locke and the early moderns, however, the idea of evicting colours, sound and smells (etc.) from the material world and relocating them to the mind has proven over time to be rather difficult to sustain. Locke often spoke of colours, sounds and smells as being “in us,” as did Galileo and Descartes. They did not mean, of course, literally in us, i.e., inside our bodies, along with our hearts and spleens. These entities were supposed rather to be “sensations” or “ideas,” with attendant mental connotations. They were also said to be “in the mind” or “in the consciousness” and the like. Malebranche spoke of them as being in the soul:

Our eyes represent colors to us on the surface of bodies and light in the air and in the sun; our ears make us hear sounds as if spread out through the air and in the resounding bodies; and if we believe what the other senses report, heat will be in fire, sweetness will be in sugar, musk will have an odor, and all the sensible qualities will be in the bodies that seem to exude or diffuse them. Yet it is certain ... that all these qualities do not exist outside the soul that perceives them ... (Elucidations, 6.)

The language here is not crucial since the underlying conceptual apparatus was relatively new and being developed as they wrote. (The old scholastic or Aristotelian ways of speaking were gradually being displaced.) This well-known passage from Descartes expresses the same embryonic conception:

It is clear then that when we say we perceive colors in objects, it is really just the same as saying that we perceive in objects something as to whose nature we are ignorant but which produces in us a very clear and vivid sensation, what we call the sensation of color. (Principles of Philosophy, para. 70.)

The trouble arises when we try to get clearer on what it could mean to say that colours, sounds and smells reside “within us.” Here, the early moderns seemed to do little better than treat the mind or the soul as an “immaterial realm” capable of housing entities that seemed out of place in the material world. External physical impulses, transmitted via our sense organs into our skulls, conspired somehow to generate sensations of colour, sound and smell in this convenient immaterial realm. Or, in another version, the physical impulses were said to correlate with the occurrence of these sensations in the immaterial realm, e.g., through the grace of God. Bells and whistles notwithstanding, the essential early modern account never really got much better than this. Most of the early moderns were immaterialists about the mind in this sketchy way and seemed content to trust that future developments would clarify the nature of immaterial minds and their relation to material bodies. A notable exception was Thomas Hobbes, who thought that such talk of immaterial minds bordered on the nonsensical.

Five hundred years on, Hobbes seems to have been vindicated, and it is no longer credible to think of the mind – assuming this concept is still useful – in the early modern way. If anything, the mind is more credibly identified nowadays with the physical brain, somewhat as Hobbes suspected, although this view has difficulties of its own (see below). We haven’t quite deciphered the mind and its place in nature, nor even the specific phenomenon of sense perception, but it does increasingly seem that any broad “dualism” on which colours, sounds and smells are evicted from the material world and housed in an immaterial realm is not a likely contender. The early moderns cannot reasonably have foreseen this as they had a lot on their plate – they were not just endeavouring to develop a new way of thinking about sense perception but were also busy keeping up with the latest discoveries in science such as the discovery of the retina and the nature of light.

There is another difficulty that they did foresee, however, and which has proven to be the deeper one.

On the Galilean view, as mentioned, colourless things out there in the world send signals to our brains (through our eyes) and colours are then conjured up in our immaterial minds, which we mistakenly project onto the things that we see. But how does a physical signal transform itself into an immaterial sensation? In the case of sound, Galileo says:

Sounds are made and heard by us when the air—without any special property of “sonority” or “transonority”—is ruffled by a rapid tremor into very minute waves and moves certain cartilages of a tympanum in our ear. External means capable of thus ruffling the air are very numerous, but for the most part they may be reduced to the trembling of some body which pushes the air and disturbs it. Waves are propagated very rapidly in this way, and high tones are produced by frequent waves and low tones by sparse ones.

This is a promising start at a mechanical explanation but notice that Galileo says nothing about how frequent or sparse waves (in the physical ear) are able to produce high or low tones (in the immaterial mind). No doubt, he believed that the answer would be revealed in time but there is a clear difficulty here about what is supposedly happening at the interface between body and mind, where material impulses are transformed into immaterial sensations.

Half a century later, Isaac Newton had discovered the refraction of light and the associated colours of the rainbow – a connection that Galileo knew nothing of – but he too professed ignorance over how “phantasms” of colour are generated in our minds:

But, to determine more absolutely, what Light is, after what manner refracted, and by what modes or actions it produceth in our minds the Phantasms of Colours, is not so easie. (‘New Theory about Light and Colors’ 1672, online copy here.)

Locke himself was certainly aware of the difficulty:

We are so far from knowing what figure, size, or motion of parts produce a yellow colour, a sweet taste, or a sharp sound, that we can by no means conceive how any size, figure, or motion of any particles, can possibly produce in us the idea of any colour, taste, or sound whatsoever; there is no conceivable connexion betwixt the one and the other. (Essay, Book 4, Chapter 3, Section 13.)

Of all the difficulties faced by the early moderns in their attempts to make sense of sense-perception, this one has persisted to this day. Nowadays known as the “hard problem of consciousness,” it is not resolved simply by replacing the notion of an immaterial mind with a physically more respectable conception. If anything this makes things harder. Thus, when you see a red object (say), physical waves are transmitted to your eyes and brain, and countless neurons fire in your visual cortex – and then what? Few would nowadays say that a “red sensation” is then generated in your “immaterial mind,” but one struggles to find a better story. Ever since the 17th century, we have known that activity in the brain causes us to “see red,” but how this phenomenon of “seeing red” is supposed to fit within the brain remains a baffling question. An immaterial mind at least sounds like the right sort of thing in which an episode of “seeing red” might be generated and housed, but how do you make sense of the phenomenon of “seeing red” if you think that it occurs inside the head but reject the idea of an immaterial mind? Hobbes himself, arch materialist, did no better than Galileo, Newton or Locke in explaining what was going on when colours, sounds and smells are generated “within us.” He writes, in the case of sound:

... the clapper hath no sound in [the bell], but motion, and maketh motion in the internal parts of the bell; so the bell hath motion, and not sound, that imparteth motion to the air; and the air hath motion, but not sound; the air imparteth motion by the ear and nerve unto the brain; and the brain hath motion but not sound ...

... from the brain, [the motion] reboundeth back into the nerves outward, and thence it becometh an apparition without, which we call sound. (Hobbes 1658, II.9, quoted in Alison Simmons, ‘Perception in Early Modern Philosophy.’)

As Simmons comments, “It’s not at all clear that this is an improvement on the dualist’s response.”

The difficulty that the early moderns faced thus remains very much alive, only transmuted into a different form. Indeed, contemporary materialists have not progressed very far beyond Hobbes. And here we encounter one of the main reasons why contemporary philosophers are suspicious of the view that colours (sounds, smells, etc.) are all “in the head.” For those who propound this doctrine usually either mean something like “in the mind,” with intimations of immateriality lurking not far behind, or else they are unable to say what exactly they mean by “in the head.” They are simply gesturing at some embryonic conception of what is going on in the way of the early moderns. If we accept that there are no such things as immaterial minds and that the only things within our heads are our brains, then the claim that colours, sounds and smells are “in the head” is considerably less attractive – whatever could you possibly mean?

As mentioned, philosophers are keenly aware of this difficulty and it vividly informs their reluctance to accept the standard sketch often found in psychology texts that colours are “generated in the brain.” It is difficult to know what scientists think of this difficulty but their attitude towards it often seems to philosophers to verge on the cavalier:

We know from psychophysical and neurophysiological investigations that color is created somewhere in the brain, although the exact location of this process is still unknown, and we even have no idea what entities the sensations called color are. Because the scientiﬁc and philosophical discussion of color should not go beyond our actual knowledge ... it is not at all a matter of taste or philosophical viewpoint where colors can be thought to be located. In short, colors appear only at a first (naive) glance to be located in objects; closer inspection shows that colors are produced in the brain ... (Backhaus & Menzel 1992, p. 28)

This is what philosophers call a “promissory note,” i.e., a blank cheque that nobody quite knows how to cash. Scientists sometimes say – as is implicit in the remarks above – that it is just a matter of time before we figure out what these “colour sensations” are and how they are created in the brain, but this plea of ignorance can only last so long before it starts to wear thin – and it has been 500 years.

When contemporary philosophers write about colour, these considerations are in the back of their minds. And this is one reason why they are often prepared to reconsider what science has supposedly shown us about colour. Consider the general stance of David Armstrong, captured here by his former student Keith Campbell:

Armstrong’s thought about colour has been conditioned not only by the search for system; it has been made more difficult by a further condition that he has always imposed: The system may admit of no nonmaterial realm of concrete substances. So the classic Lockean strategy, that of providing a refuge for colour in the immaterial perceiving minds of creatures with colour vision, has never been a serious option for him ... It is not that Armstrong does not believe in minds. On the contrary, he holds that minds are substances we know to exist. The problem is that, with minds being functional systems whose categorical basis consists in nervous tissue, there is no more comfortable place for the colours in the brain than there is on the surfaces of the material objects that seem to bear them. (Keith Campbell, ‘David Armstrong and Realism about Colour.’)

Armstrong himself puts it this way:

In the manifest image of the world , to adopt Wilfrid Sellars' s expression, colour, heat and cold, sound, taste and smell, play a conspicuous part . Yet as far back as Galileo (with intellectual precedents stretching back to the Greek atomists) physicists could find no place for these qualities as intrinsic properties of physical things ...

What is a scientific realist to do? A traditional solution is to use the mind as an ontological dustbin, or sink, for the secondary qualities. Locke gave us a classical formulation. For him, the surface of a ripe Jonathan apple can properly be said to be red. But what constitutes its redness is only this. The surface has nothing but the primary qualities. In virtue of certain of these properties, however, properties of the micro-structure of the surface, it has the power to produce in the minds of normal perceivers in standard conditions (this last fills out Locke a little) ideas, or sense-impressions or sense-data, mental objects which have a certain simple quality, a quality unfamiliar to persons blind from birth ...

[But if] we have not merely accepted a scientific realism about the physical world, as Locke did, but have also made the mind part of that world, then there is no hiding place down in the mind for the sensible secondary qualities. The same reasons that made one want to exclude them from the physical world will make one want to exclude them from the mind. (Armstrong, ibid.)

If we agree that everything (including the mind) is material, then there is no longer any gain to relocating colours (sounds, smells, etc.) to the mind. At least, the early modern reason is no longer available. The question then is whether there are other reasons for relocating colours to the mind. Quite a few have been offered but the sting of attack has largely been defused and philosophers generally have no trouble warding off these other objections while counterattacking for good measure. (See below.)

Here’s how Sydney Shoemaker puts it:

It is, at any rate, one of the cliches of the history of philosophy that the problem of secondary qualities first faced in the 17th century was solved, or swept under the rug, by relegating them to the "dustbin of the mind." It is also a cliche of recent discussion of these issues that this solution becomes problematic if one abandons the dualistic view of mind taken for granted by the 17th century writers in favor of a materialist or physicalist view. A dualist can hold that there is no place for redness-as-experienced in material bodies, and nevertheless allow for its existence, by locating it in the non-material realm he takes the mind to be. A materialist obviously cannot do this, since according to him there is no non-material realm. A materialist who does not deny that there is such a thing as redness-as-experienced must allow that there is, after all, a place for this property in the material world. And if he nevertheless agrees with the 17th century writers in holding that such properties are instantiated only in the mind, i.e., as properties of sense-experiences or the like, he must explain how they can be instantiated in one part of the material world, namely the neural substrate of our perceptual experience, if they cannot be instantiated in the parts of the material world in which we seem to experience them as being instantiated, namely the surfaces of objects like ripe tomatoes ... What options does a materialist have? One is to say that the 17th century writers got the whole discussion off on the wrong foot, and that there is no cogent objection to a straightforward identification of the secondary qualities with "primary" ones, i.e., with properties that have a place in the scientific world picture ... (Shoemaker, ibid.)

Shoemaker himself does not think that colours are out there in the world, but proposes a more sophisticated view which need not concern us here.

Second, that philosophers sometimes use terms like ‘colour,’ ‘sound,’ ‘smell’ (etc.) in an ambiguous way which is apt to mislead those who are not mindful of the ambiguity. Consider the old riddle, doubtless inspired by the Galilean view of secondary qualities: if a tree falls in a forest and no one is around, does it make a sound? A straightforward Galilean answer would be that it doesn’t, since sound is only ever a “sensation” in the mind. A more nuanced answer however is that, in the sense in which sound is a “sensation” or a “mental experience,” the tree doesn’t make a sound, whereas, in the sense in which sound is a physical disturbance in the air, it certainly does. By these nuanced lights, while the term ‘sound’ perhaps has one sense in which it denotes something “in the mind,” it has another legitimate sense in which it straightforwardly denotes something in the external world. Given this distinction, it is possible for a philosopher (say) to claim that secondary qualities like colour, sound and smell are located in the external world even as he or she holds firmly to the broad Galilean view outlined above. This philosopher would simply have to be using terms

like ‘colour,’ ‘sound’ and ‘smell’ in the second of the two senses mentioned. This is a potential source of confusion to someone not in the know but, as we shall see, a position like this is in fact often attributed to Locke. It gets even more confusing when 20th-century philosophers enter the picture because many of them operate with an unspoken theoretical model on which the second of the two senses above is the “correct” way to use terms like ‘colour,’ ‘sound’ and ‘smell,’ so that there can be no question as to whether colour, sound and smell are out there in the external world – of course they are. It is not only outsiders who stand to be confused by these linguistic twists and turns; philosophers themselves often confuse each other.

Menu