http://www.elea.org/Miracles/
Copyright © 1999, Allan Randall

 
 

 Miracles and Quantum Suicide

Allan F. Randall
Dept. of Philosophy, York University
Toronto, Ontario, Canada
randall@elea.org, http://www.elea.org/

I. Introduction

Are miracles possible? This essay explores the possibility of generating miracles using the well-established principles of quantum mechanics. It is not a New Age tract on crystals and auras, and those looking for Òself-helpÓ tips on miracle-working will probably be disappointed, as the results are largely (although not totally) negative. Reliably generating a major miracle (for instance, invoking a magical green elf to appear in a puff of smoke) turns out to be impossible. This may not seem surprising, but defending this result is harder than you might think, and the road we take to get there will provide us with a better understanding of quantum mechanics. On the other hand, minor "miracles", like winning the lottery, can be generated reasonably reliably. Unfortunately, the win is risky and comes at a great cost, one you are not likely willing to pay.

Finally, I will examine what may be the most desirable miracle of all: immortality. Assuming there is no God who will save us in the end, or soul that departs the body at death, then immortality would seem to be pretty obviously in the Òmajor miraclesÓ category and hence impossible. But this issue will turn out to be particularly tricky, with its own complications that set it apart.

The Interpretation of Quantum Mechanics

Quantum mechanics is our current best theory as to how the world works. If our understanding of the universe is to progress, we must take our best theories seriously. True, they may turn out in the end to be incomplete or totally false, but it would be unscientific in the extreme to assume such before there is real empirical evidence that the theory has a problem. Currently, there are no such empirical holes in quantum mechanics. I will therefore adopt a completely literal interpretation of quantum mechanics. I will not spend much time defending this view, but will simply assume it, and explore the consequences. More detailed arguments for taking quantum mechanics seriously can be found elsewhere.[1],[2],[3] It will be helpful if the reader is generally familiar with the essentials of the quantum measurement problem, although a brief (but inadequate) overview will be given for those who have not encountered it before.

The classic problem with quantum mechanics is that, in spite of the fact that it provides us with a remarkably accurate way of predicting the behaviour of the very small, if scaled up to the very large, it seems to produce paradoxical results.[4] The quantum equations do not yield a single outcome to any given situation, but multiple outcomes in ÒsuperpositionÓ. The mathematical description of this superposition is called the quantum ÒwavefunctionÓ. The classic example is SchršdingerÕs Cat experiment, wherein a cat is killed or not depending on whether a radioactive sample decaysÑgiven that the quantum equations predict a 50% chance of decay. Quantum mechanics predicts that the cat will be in a superposition of being both dead and alive. This seems bizarre, even contradictory, since we never see cats that are both alive and dead. Yet quantum mechanics, if it is truly an accurate picture of the world, must scale up to the macroscopic level, else it is an incomplete theory. After all, cats in the end are supposedly just large collections of interacting subatomic particles, so the laws of quantum mechanics ought to apply to a large cat as much as to a small collection of electrons. Some have thus argued from SchršdingerÕs Cat that quantum mechanics is therefore disproved as contradictory (Schršdinger himself took this view). The theory must  be incomplete, since our empirical observations disprove what it predicts: we do not see cats in superposition.

But this argument is flawed, since it assumes that the human being observing the cat is not a quantum object. If we are taking the theory seriously, and insisting that the cat is in a superposition, then we must assume that the human observer is also just a collection of interacting particles, obeying the laws of quantum physics. Thus, the observer watching the cat is also in a superposition, with one version of the observer seeing a live cat and the other a dead cat. Another way of putting this is to say that there are two ÒworldsÓ or Òparallel universesÓ, one containing a human looking at a dead cat and the other a similar but not identical human looking at a live cat.

This interpretation of quantum theory, popularly called the Òmany-worlds interpretationÓ, was originated by Hugh Everett as the Òrelative state formulationÓ of quantum theory.[5] But really, it is the only type of interpretation of quantum mechanics that is possible. This can be seen by looking at the different types of interpretations that have been offered. Traditionally, it has been said that the various interpretations fall roughly into three main categories:

(1) The Copenhagen interpretation and its variants. These all assume that at some point between the subatomic events and human observation, a classical nonquantum object (or process, or something) intervenes and ÒcollapsesÓ the superposition into only one of its possibilities. Thus, this collapse prevents the quantum system from developing into more than one world on the macroscopic level. When Niels Bohr originally defended this view, he claimed that it was a true interpretation of quantum theory, and that quantum theory could thus be consistently said to be complete. Although he seems to have convinced an entire generation of physicists of this delusion, it is quite straightforwardly false, since a nonquantum something-or-other is required to collapse the wavefunction. So Copenhagenism cannot possibly be correct if quantum theory is complete. Somehow, there is something extra, not explained by quantum mechanics, that collapses the wavefunction. This something extra was easy for laboratory physicists to just ignore for many years, but modern cosmologists have found this impossible, since they seek to study the wavefunction of the entire universe, of everything there is. One cannot so easily ignore the requirement that something extra outside the entire universe collapses it into a single history. This problem with Copenhagenism is now widely recognized, and there are relatively few who continue to support it, at least as a pure ÒinterpretationÓ of quantum theory.

(2) Hidden variables theories. This includes all those theories that postulate the incompleteness of quantum mechanics explicitly. They claim that there is something hidden going on, not yet discovered, that reduces the superposition of possibilities. These theories, most notably the deBoglie-Bohm pilot wave, have traditionally been recognized as not strictly interpretational, but as actually postulating the incompleteness of the theory. Of course, Copenhagenism is really no less a hidden variables theory than any other. I place it above in a separate category only because it was for many years misclassified as a pure interpretation.

(3) Many-worlds and its variants. The notion of many-worlds is thus the only pure interpretation of quantum theory currently on the table. There are, it should be noted, numerous interpretations that do not use the term "many worlds" that nonetheless fall into this category. I am not claiming that there is only one conception of many worlds that is possible, only that any consistent way of interpreting quantum theory will involve, one way or another, an ensemble of worlds equally real to our own. Whether or not we call it Òmany-worldsÓ is not important. What is important is that these alternate worlds (or states of affairs, or universes, or histories, or whatever one wants to call them) are all there in the wavefunction of the universe, and they are alternatives we have traditionally considered, for empirical reasons, as ÒcontradictoryÓ (although they are not logically or mathematically contradictory). Their existence is mandated by quantum mechanics. No re-interpretation can get rid of the alternate worlds, simply because they are there in the mathematical machinery of our most thoroughy tested scientific theory ever. One can postulate hidden "world-eliminating" forces at work, but these are not required to explain our empirical data, all of which is in accord with quantum theory. I will thus do what I think any good scientist ought to do under such circunstances: accept the theory as our current best guess, while remaining open to its being disproved in the future.

Quantum Miracles

It has often been noted that almost any result you can dream up exists in the wave function, as long as one does not look to actually violate any laws of quantum physics, such as the conservation of mass/energy. For instance, there is a slice of the wavefunction where all the gas particles in the room you are now in suddenly rush into one corner of the room, suffocating you. The quantum wavefunction tells us that there are remarkably few worlds where this happens, but nonetheless, there are a tiny fraction of worlds where it does indeed occur. But since the probability of its occurence is so small, we never actually observe it in practise.

So all kinds of crazy things that would once have been said to be almost certainly impossible (making a book levitate in the air, making a magical green elf appear in front of you in a puff of smoke, etc.) are not only possible, but actually do get realized in some worlds, according to quantum theory. However, any one version of you will still only experience one of the worlds, so these ÒmiraclesÓ are highly improbable, even though they are nonetheless there in the universal wavefunction. Worlds exist where our greatest dreams and worst nightmares come true.

So the question arises: can I not do something to change the probabilities so that the more desirable alternatives are more probable? I will call such feats Òmiracle-workingÓ, since we are trying to increase the probability of worlds close to some ideal of what we want, in a way that would violate our usual notions of what is possible, or at least probable. Many New Age adherents have speculated about such possibilities, but most quantum physicists dismiss such talk as nonsense, of course. However, I would argue that the possibility of miracle-working should be looked at more closely, as its refutation is by no means as trivial as many suppose.

The first thing to note is that it is, in fact, quite easy to act to change the probabilities so that what you observe is closer to your ideal. In fact, this might be said to be the very definition of life (or at least integral to its definition). As living feedback control systems, we act all the time to bring the quantum probabilities closer to our ideal. However, we seem to have a limited ability to do this. For convenience, I will divide such "purposive actions" into three classes, although there is of course no solid dividing line between them:

Purposive Action:

     (1) Nonmiraculous (natural): actions that make one's ideal highly probable, where that ideal does not violate our usual notions of what is probable. For example, you work hard to start a new business so that you will be likely to make lots of money.

     (2) Minor miracles (beating the odds): actions that make one's ideal highly probable, where that ideal would normally be considered relatively improbable, but not in itself miraculous, perhaps even being observed to happen on occasion. An example would be winning the lottery. This certainly happens to some people, and in itself is not a miraculous event, but it would certainly still be a miracle of sorts if we could beat the odds and be able to guarantee a win reliably.

     (3) Major miracles (supernatural): actions that make one's ideal highly probable, where that ideal would normally be considered highly unlikely, but not literally impossible, to ever be observed. For instance, causing a small magical green elf to appear in a puff of smoke.

Again, I must stress that this is not an absolute distinction. According to quantum mechanics, all these results are there in the universal wavefunction, no matter how bizarre and improbable they may otherwise seem. Note that I have not listed above with the others the following fourth kind of action:

     (4) Impossibilities: actions that make one's ideal highly probable, where that ideal is inconsistent, incoherent or violates known laws of physics. For instance, causing a square circle to appear before me, or violating the conservation of mass/energy.

This fourth kind of action is trivially impossible, so it is not really a legitimate kind of action at all, since the ideal you are shooting for is actually a confused idea in your mind, or a physical impossibility. It is thus not even in the universal wavefunction at all.

Now let us look again at all three legitimate actions one at a time.

Nonmiraculous (natural) actions

We all know these are possible. We struggle all our lives to make a world closer to our ideal, but we seem constrained by a certain distribution of the probabilities, which we have limited control over.

Minor miracles

I will use the lottery example for this, since the actual probabilities are well established, and it is easy to see that the chances of winning the big prize are extremely remote. Of course, the mere fact of winning the lottery is not itself a true miracle, since there are people who win the lottery. But if you could do something that would guarantee (or almost guarantee) that you would win the lottery, then that would seem miraculous. So we will call that a "minor" miracle.

Say you are playing a lottery with a 1 in 10 million chance of winning the jackpot. The universe where you win is in the universal wavefunctionÑthat seems pretty straighforward. In fact, it should be fairly straightforward to show that up to a point in time fairly close to the draw, all possible lottery numbers are about equally probable. Lottery numbers are usually chosen by a machine that mixes up a bunch of small balls with numbers written on themÑa process that physicists call "chaotic", meaning that a very small quantum uncertainty very quickly makes a difference at the macroscopic level (as with SchršdingerÕs Cat), so that all the different possible lottery numbers are about equally probable right up to some time fairly close to the actual draw.

The result of the draw is 10 million different worlds, one for each possible lottery number (we will pretend for the sake of convenience that the lottery draw is the only thing distinguishing the different worlds). You exist in all 10 million, in one of which you are the big winner. This means that at the moment of the draw, you split into 10 million different versions of yourself. So it is just dumb luck if you find yourself with a winning number, since there are 9,999,999 other slightly different copies of you out there in other universes with unlucky numbers.

The probability 1/10,000,000 is subjective, being the probability of your being in a world where you win the lottery. By saying it is subjective, I do not mean that it is not based on an objectively existing wavefunction, just that the probabilities involved are probabilities of possible experiences that you could have. Actually, all probabilities are subjective in a weaker sense, since they involve an arbitrary classification of equipossible results into different kinds of results. All the individual results are equally possibleÑwe cannot say that one is more probable than another, only that one kind  of result is more probable than the other kinds. In the case of the lottery, we are classifying the results as either positive or negative lottery wins. But here, the probabilities are not only subjective because of this subjectively chosen classification scheme, but also in a stronger sense, because the items being classified themselves include you, the observer, as part of the system.

For instance, take two worlds, A and B. You win the lottery in world A, but lose in world B. Now imagine an analysis of the universal wavefunction that splits up the "worlds" in such a way that the left half of your brain in world A was classified instead as part of world X and the right half as part of world Y, and vice-versa, so the right half of your brain in world B was part of world X and the left half part of world Y. This might be a technically allowable analysis of the wavefunction (one can theoretically use any arbitary "basis" language one wants to describe the wave function). However, it makes no sense to choose such a basis, since it would destroy the unity of your consciousness. Not that this unity has some power to define how the wavefunction is constituted, but rather, it is simply not possible for you to experience the world in any other way than in a way that has a unified consciousness in each "world".

As a simpler example, imagine that a world resulted from the quantum equations where you did not even exist. This world would be automatically ruled out of consideration for the strong type of subjectivity, and it would be placed in neither the positive nor the negative lottery group, since you can neither win nor lose the lottery in a world where you do not exist. Curiously, all physical quantum measurements are ultimately in this strongly subjective category, since they involve the probabilities of a single human observer experiencing certain results. This is where talk of ÒcollapseÓ of the wavefunction is actually useful. From an objective stance, we say that SchršdingerÕs cat is in a superposition, and we are likewise in a superposition of observing it. But all actual physical measurements are made by people who are part of the system being measured, and they only end up experiencing one of the outcomes. This is, in a sense, a ÒcollapseÓ of the wavefunction, but not an objective, ÒrealÓ collapse, merely a subjective perspective from which the wavefunction appears to collapse, since the observer has bifurcated into two different consciousnesses. So long as the probabilities we are computing have this strong subjectivity, then only one path of consciousness through the wavefunction can be followed, since we are by definition considering the probabilities of a single observer having certain experiences.

Actually, the more objective stance that recognizes two superimposed observers observing superimposed cats is not completely objective. It is a ÒweakerÓ subjectivity. But it still classifies outcomes according to certain criteria that place different observers in different worlds. It turns out that to split the wavefunction into different worlds requires this arbitrary analysis of the wavefunction that already recognizes that different streams of consciousnesses belong to different ÒworldsÓ. So this perspective is more objective than strong subjectivity, since it does not privilege a particular, single consciousness, but it is still not completely objective, since it does assume that what is important is consciousness. So it is a weaker subjectivity, a more strongly objective stance. The most objective stance of all would simply take the universal quantum wavefunction at face value, as it was, without splitting it or analyzing it into separate worlds, or caring about consciousness at all. But such a viewpoint would no longer be one concerning physical measurement. I would argue that it would be the perspective of pure mathematics, although this is a controversial question that need not be settled for our purposes here.

What we are trying to do in our miracle-working is ensure that there are far more results in the ÒpositiveÓ lottery group than in the ÒnegativeÓ group, where the only possibilities placed in either group are worlds where we do experience an outcome. This is a relatively objective stance that defines a world in terms of the consciousness of a particular person, and Òcollapses outÓ any worlds that do not contain said consciousness, but does not insist on only a single consciousness, allowing a superposition of different versions of the same person. This requires, of course, some kind of ultimately arbitrary definition of what distinguishes two people as separate individuals, but I think we should all be able to agree such a theory might be possible, or else we have no basis for a notion of personal identity.

I really must stress here that all this talk about different ways to analyze the wavefunction, collapsing out certain parts of it, in no way changes the nature of the overall universal wavefunction, which contains all the possibilities, there to be analyzed from whatever stance we like. We are taking these various perspectives and stances because by definition the issue we are exploring is how things seem from the perspective of a particular person, and what the person should justifiably be expected to experience in the future.

So is there any way that you can act to ensure that the vast majority of worlds are in the positive group, thus practically guaranteeing that you will win the lottery? There is at least one way to do this, although I always hesitate to tell anyone, in case they might actually try it. So please allow me to insert the following disclaimer:

The following technique for winning the lottery, or performing other minor miracles, is strongly recommended AGAINST. It is a philosophical thought experiment only, related for the purposes of stimulating intellectual discussion. It would be extremely unwise to actually try it out and I take no responsibility for anyone unbalanced enough to do so.

So having said that, here is the technique: simply kill yourself if you do not win the lottery! The best way to do this would be to have a machine automatically monitor the lottery results. You then go to bed before the draw, and have the machine quietly perform the execution while you sleep. That way, you are never in any way conscious or aware in any universe where you lose the lottery, and all such undesirable worlds are automatically eliminated from both the winning and the losing groups.

In other words, if you are never conscious or aware in these universes, then they are automatically no longer part of your physical wavefunction! They completely drop out, since the probabilities in the wavefunction are strongly subjective: they state the probability of your observing some class or kind of result.

From the objective point of view, all results are there in the universal wavefunction and no one is more probable than any other. But physical measurements are always strongly subjective and thus always involve a subjective collapse of the wavefunction. By killing yourself if you lose, before you have a chance to even be conscious in such a world, you are never at any point alive and aware in any world where your lottery number loses, and, from your own purely subjective point of view, these universes don't really matterÑyou in fact are not even in them. In summary, the only way I can think to generate miracles is to simply exit those universes where the miracle does not occur, and the only way I can think to do that is to commit suicide.

There are several really good reasons not to go out and do this! First of all, do you really have enough faith in the idea of many worlds to risk your life based on it? I certainly think the theory makes all kinds of scientific and philosophical sense, and is indeed our current best guess at the truth, but I for one would never trust it enough to go out and kill myself. But even if you could somehow be 100% certain of the theoryÕs truth, you had still better be very sure you do your probability calculations properly. If you mess up, the result will not be millions of dollars, but a failed suicide attempt that may leave you crippled for life.

Besides which, to perform such a quantum suicide would be extremely cruel to all your loved ones, and, some might argue, would in a sense be the ultimate selfish act. To commit suicide in this way assumes that all that matters is the wavefunction from your subjective point of view. What about the subjective point of view of others? For your loved ones, the chances of your survival are still 1/10,000,000, so they will not be so impressed. From their point of view, there is no minor miracle hereÑyou are almost certainly killing yourself. Most likely, you would not appreciate their doing the same thing.

Such an act would perhaps make more sense for someone who has no loved ones at all. If you really have nobody in the world who means anything to you at all, then at least no one else would be directly hurt if you killed yourself. On the other hand, you would be "opting out" of society (or at least 9,999,999/10,000,000 of future societies). You are refusing to be part of the global solutionÑrefusing to do your part in making the world a better place, and instead exiting the world for your own selfish reasons. Of course, from your subjective point of view, you are not exiting the world at all, but this assumes that only your own subjective view matters.

On the other hand, it could be argued that if someone wants to so exit the world, they have the right to do so. The moral issues here are not all that different from those surrounding regular suicide. There are some differences, however. If many-worlds were proven to the satisfaction of society as a whole, there might be a rash of suicides, as people take advantage of their ability to (subjectively) work miracles. This would be disruptive enough to society as a whole, that a societal rule against it might make sense even to those who would not oppose suicide under other circumstances.

My guess is that even given iron-clad proof of other worlds, the vast majority of people would still instinctively avoid suicide. There might, however, be a certain percentage of people who would do it. Exactly how many is anybodyÕs guess. We could expect, however, that in the vast majority of worlds, the people who survive will be those who are not prone to taking such a purely subjective view. So even if the existence of many-worlds were proven, natural selection would probably push our species away from a state of continual mass suicide (unless, of course, the suicides themselves proved to be adaptive, perhaps as a means of population controlÑbut I will leave further debate on this issue for another time).

Major (supernatural) miracles

Let us put aside moral issues for now, and stick with the purpose of the thought experiment: to explore the extent to which it is possible to make miracles happen, regardless of the wisdom or ethics of actually doing it. We have seen that there is a way to make a minor miracle occur. Now we will try to apply the same technique to create major miracles. It might seem that, if there is no absolute dividing line between minor and major miracles, and the distinction is only a matter of degree, as I claimed earlier, then we ought to be able to apply the same quantum suicide trick, killing ourselves in any universe where a green magical elf does not appear in a puff of smoke. However, there are numerous difficulties with this, which I will discuss one at a time. Some of the problems listed below might also be a concern in creating minor miracles, but we will see that such problems are surmountable for minor miracles with some clever design and engineering. For major miracles, however, there are (I believe) insurmountable difficulties.

(1) Automatic detection of the ideal.

Take the example of the elf. I want a magical green elf to appear before me in a puff of smoke. Unfortunately, it is going to be harder in this instance than in the lottery case to hook up a machine to detect the result automatically and kill me while I sleep, since it certainly seems like one would need a human being to detect the presence of a green magical elf. This may be a problem for the lottery miracle as wellÑbut we can probably solve the problem there with some ingenuity. Perhaps the lottery results can be obtained without human intervention from an Internet site, for instance. Unfortunately, while current computer technology can provide lottery results automatically, we would need an advanced technology not available today to detect a magic elf automatically. Perhaps, in the future, we could build an artificial intelligence (AI) that is able to detect magic green elves just as well as a human can. Of course, we would have to decide what the criteria for ÒmagicalÓ and ÒelfÓ are, but that can perhaps be worked out with some thought.

However, even if no adequate AI is forthcoming, this is really just another ethical issue, not a technical one, since the only reason we want to detect the ideal with a machine rather than a human is to prevent having to bring in a human accomplice who would be required to commit murder. But, putting such ethical issues aside, there is nothing preventing us from having a friend perform the detection and execution, without the need for an AI. There are practical problems, of courseÑlike finding an accomplice who is willing to risk the murder charge that would almost certainly follow.

Another possibility would be to just perform the suicide yourself, reasoning that you will only be conscious for a brief time in the worlds without the elf, or lottery win, or whatever. However, most of us would find it discomforting to say the least, knowing that doomed versions of us that know they have only seconds to live will exist for even a brief time. So it is reasonable to do everything we can to make sure no version of ourselves is ever conscious of their eminent death.

In summary, so long as reliable detection is possible, whether we can do so automatically without human involvement is not a fundamental stumbling block. What is at stake here is more our peace of mind in being able to perform the experiment without undo stress on those involved, such as our other selves or an accomplice who could get into legal trouble.

On the other hand, there are some kinds of major miracles where this problem clearly would be insurmountable, such as when the ideal we seek is not something any human (or AI) would be able to recognize, even were it to happen. For instance, we can forget about using this technique to discover some truth that we do not currently know. We cannot advance science simply by killing ourselves in all universes where scientific papers explaining all the mysteries of the unvierse do not appear fully written before us. Weeding out the garbage from the legitimate breakthroughs would likely be a far tougher job than just going ahead and figuring it all out for ourselves in the first place (i.e., actually doing the experimental science).

The only miracles even under consideration here, then, are those that we feel confident we, or an accomplice, could recognize were they to happen. The magic green elf seems to be in this category, as one could spend some time testing and interviewing the elf before making a decision. Unfortunately, as we will see in #3 below, the fact that we could detect the elf were he to appear does not mean that we can detect that he has not appeared, and this will turn out to be a far greater problem.

(2) Detection time

Detection time would have been a problem for the lottery example except that you were willing to fall asleep and wake up to your idealÑyou did not require that it happen before your eyes. So the fact that it took a certain amount of time for the lottery draw to occur did not matter to you. But if you want a green elf to appear before your very eyes in a puff of smoke, detection time becomes a problem. However, this really has nothing to do with whether it is a major or a minor miracle, so we will consider that this is a potential problem in either case. Because it takes time for your friend to recognize whether a green elf has appeared, you will be aware and conscious in the many many universes where it does not happen, for at least a brief time, before being killed. Again, most of us would not be happy with that (although perhaps that is just a matter of mental outlook).

There are two solutions. You could go with the artificially intelligent detection of the green elf and require that the elf appear almost instantlyÑfast enough that its presence or absence could be detected by the artificially intelligent machine faster than you could become aware of the result yourself. Of course, the problem here is that artificial intelligence technology is not good enough yet, and may never be. So, for the time being, let us modify our requirements slightly and have your human, naturally intelligent, friend kill you only if a magical green elf does not appear while you sleep (say within a certain amount of time of falling asleep). This way, you go to bed and wake up with a magical green elf by your bedside. Okay, so we have to dispense with the puff of smoke and instant appearance, but they unfortunately present technical problems, and I think we can agree that having a magical green elf at our bedside when we wake up is still a pretty major miracle indeed.

In summary, then, we must choose a miracle for which we can detect its presence fast enough so that the killing can occur before we can be conscious of a failure (or any result of the failure whatsoever).

(3) Accuracy of detection

Problems (1) and (2) seem surmountable, so long as you always have a human accomplice and allow the miracle to occur while you sleep. However, the accuracy of the detection process is a real problem, not so easily worked around, particularly in the negative caseÑdetecting that our ideal has not appeared. While it is true that your friend might be perfectly able to recognize a green elf if such a being were to appear, this does not at all mean that if the friend believes there is a green elf in the room, that there is one there, and that it will continue to be there, acting lawfully like a magic green elf in the future when you wake up (whatever the laws governing magical green elves are!).

For instance, your friend could be hallucinating an elf, in which case, when you wake up you will see nothing unusual at all, not even for a brief instant. Even if you could somehow guarantee against hallucination and delusion (perhaps with an artificial intelligence?), the molecules in the room could still form into a temporary shell that looks like a green elf, but is empty inside and will break apart into a random mess by the time you wake up, or very shortly thereafter. All of these possibilities certainly seem like they would be far more probable than a real actual probabilistically stable magical green elf appearing. Even the artificial intelligence would not actually be immune from delusion or hallucination (perhaps you could make one that was less prone to it than humans, but its hallucinating would likely still be far more probable than a real actual green elf appearing).

The fact is that there are just all kinds of ways in which the detection of the green elf could fail in some way that is far more probable than a real (stable) green elf appearing. This is not so much of a problem with a minor miracle, like winning the lottery, since it is relatively easy to detect a winning lottery number accurately with a probability greater than that of not winning.

(4) Reliability of Suicide Method

Problem (3) could be considered an aspect of the more general problem of choosing a suicide method, if we lump the detection technique in as part of the overall suicide method, which includes many other things that could potentially go wrong. For instance, if a gun is used, what is the probability of its misfiring? What is the probability of the bullet vaporising before even reaching a vital organ? These probabilities can be made quite low, granted, but they are still almost certainly much higher than the probability of a stable magic green elf appearing (even if we could accurately detect it). On the other hand, they probably need not be higher than the probability of winning the lotteryÑso the minor miracle is still doable.

On the other hand, it is probably easy to underestimate the probability of suicide failure, so even for a minor miracle, we had better be very careful to make the suicide method as reliable as possible, since if we screw things up, the most probable result might be that we survive the whole experience, but with severe paralysis or other debilitating injury--and no lottery winning to boot!

So problems (3) and (4) can be combined into the more general problem of maximizing the reliability of detecting the absence of the ideal we desire and performing the suicide. Let us say the probability of successfully doing this is p(suicide) and the probability of failing is p(~suicide). The probability of the ideal we seek actually occuring is p(ideal) and of its not occuring is p(~ideal). Then if the following conditions hold (where >> means Òmuch-greater-thanÓ and << means Òmuch-less-thanÓ), it is highly improbable that we can perform the miracle:

Conditions for failure of miracle generation:

     p(suicide) << p(~ideal)
     p(~suicide) >> p(ideal)

The lottery certainly seems doable, although with odds like 1/10,000,000, we still need to work very hard to make the detection and suicide methods as reliable as possible. The green elf, on the other hand, is certainly unachievable with any reliability at all. The actual result of such an experiment will almost certainly be a failed suicide attempt, perhaps resulting in extreme pain or permanent injury.

We could even decide to take the above requirement as actually defining the distinction between major and minor miracles. Although this is really an arbitrary distinction, if we are going to draw the line somewhere, this is not an unreasonable place to draw it. Perhaps even more interesting are cases that may be on the border line between non-miraculous and minor, and between minor and major. But I think it may be very difficult to determine the probabilities well enough to make such judgements (especially between minor and major). The actual situations are far too complex for us to be certain of the probabilities; we can only make educated guesses.

Immortality

So beating the odds at a lottery is possible, but only at a cost few would be willing to pay. More dramatic miracles, that actually seem to accomplish the impossible, are probably just thatÑimpossible. But what about what may be the most desirable miracle of all: immortality? One might think this is surely in the same category as the magical green elf, a major miracle not achievable in practice. But this is not so obvious. Immortality is a very special kind of miracle, with surprising properties when viewed in the light of quantum mechanics. What happens if you try the lottery trick here? That would mean that you kill yourself in all universes where you do not survive! But that is, of course, already done for you, by definition! No consideration need be given here to methods of suicide and their probabilities of failure, since mere survival is all we care about. As long as you do survive in even a tiny percentage of worlds, then immortality is automatically yours! This means that you are already immortal, so long as your continued survival is not (1) a logical contradiction or (2) a physical impossibility. Yet if a green elf or a levitating chair are logically and physically possible (as they are generally considered to be), then surely oneÕs continued survival is not literally impossible! It is this issue we will now explore.

The case of accidental death is the most straightforward. Quantum theory gives you automatic protection against this sort of thing (getting run over by a bus, for instance). Such accidents are subject to all kinds of chaotic effects, and one need not go back very far before the accident to find that it is only one quantum possibility out of many, and usually not the most probable one at that. So it makes no sense, assuming the literal truth of quantum mechanics, to go around worrying about accidental death, at least not for fear of your own life. It does, of course, make sense to worry about losing a loved one in this manner, and it does make sense to avoid accidents out of concern for those who would be affected by your death. And it also makes sense for another completely selfish reason: doing what you can to avoid accidental death will also tend to prevent accidents that injure you, but do not quite kill you. So while the fear of accidental death may no longer be rational given quantum theory, the fear of accidental near-death experiences is still quite real.

I said earlier that there will be at least a small number of worlds in which you survive, considering your state shortly before the accident. But what about mere moments before the accident? You may still be conscious as the bus looms over you. Does that person, so close to death, have any hope of survival, or must s/he simply take comfort in the fact that there are alternate worlds that split off from his/her own only moments ago in which s/he does survive. Should it even matter? What if I told you that I was going to erase all your memories of the past five hours. Would that bother you greatly? Probably to an extent, yes, but you would hardly think it of the same seriousness as death. Many people have had the experience of losing a few hours of memory due to a drinking binge or other cause, and while it may be disconcerting up to a point, it is something we shrug off fairly easily.

So let us assume for now that there is a brief, consciously discernible time before the moment of death/survival when there is no room in the quantum equations for survival, whether due to physical or logical impossibility. You look up at the looming bus and there is literally nobody in any future world who can remember such an event. Is there any reason to view this any differently from the classical (nonquantum) situation involving a single world, where you similarly get hit, but survive and lose your memory from just before the accident?

Of course, even if you decide losing a few minutes, hours or even days of memory is acceptable, there will still be a point where you will consider the memory loss to be more than an inconvenience and in some sense a true loss of self. What if you lost your memory of the past ten years? The thought of that probably terrifies you and is surely not all that different from your fear of death. Unlike the last ten minutes, the last ten years are vital to your very sense of self-identity and who you are. So there is no clear-cut answer as to how much memory loss is acceptable. Most of us would simply become more and more uncomfortable with the idea of greater and greater memory losses.

However, a loss as great as ten years is not analogous to what happens in accidental death. One need only look hours or minutes (perhaps even seconds) before the accident to clearly see that tiny differences could prevent the accident from occuring, even without having to invoke major quantum miracles. And if we are willing to invoke major miracles, it is hard to see how there would even be a consciously discernible time during which we need face death. Exactly how close we need get to the accident for death to be 100% physically assured is not clear, but it seems hard to believe that the time scale involved is very large, and perhaps it is not significant at all. Keep in mind that anything logically consistent that does not violate the laws of physics is found somewhere in the wavefunction, even if in only a tiny percentage of it. Even major miracles like the bus suddenly levitating in the air, and thus missing you, should have some finite, nonzero probability. Recall that we are no longer worrying about the probability of this major miracle in comparison to the probability of suicide failure. Here, your death is by far the most probable outcome, from any point of view just slightly more objective than your own, so no matter how unlikely the quantum miracle required to save your life, it is now the dominant factor in your subjective wave function. Most likely, even a fraction of a second before the accident, there will be some possible event that will save your life in some small percentage of worlds that is, in fact, far more likely than the bus levitating. So we are back to feeling invincible. Nothing can kill us it appears, for we will always survive in some world or other, even considering a point in time very close to when death would normally be expected to occur.

But before you go jumping in front of any buses, recall that this invincibility is no justification for recklessness. True, you will survive in some world or other, but the most probable such worlds will almost certainly not involve levitating buses (which was just given as an extreme example), but instead will most likely involve your being seriously injured, perhaps crippled for the rest of your life.

Of course, for those who seek immortality, what really matters here is not whether we can prevent accidental death, but whether we can avoid the aging process, since that is the inevitable death that seems to await us all, even if we do manage to avoid accident and disease.

Given the current state of medicine, it would appear that we are all doomed. Every last human on Earth is suffering from a fatal disease called aging. Try as we might, it will kill us in the end, as there is no known cure with present technology. This seems to put it in the class of requiring a major miracle to save us, in line with green elves and levitating chairs. It is unlikely that continued survival in spite of the aging process is literally physically or logically impossible. If there is some tiny percentage of worlds where molecules by chance conspire to levitate chairs, then there is surely also some perhaps even tinier percentage where they conspire to keep us alive at least a little bit longer in spite of advanced age.

This is not, it would seem, a kind of immortality you should desire, since the most probable such major miracle is not going to restore your health and youth. All that is required here is that your consciousness and personal sense of identity be retained. This will most likely leave you in a state of advanced ill health. According to this argument, you are doomed to live for all eternity in a state of progressively worsening health, kept barely alive and conscious by quantum miracles that are just miraculous enough to keep you alive, but no more so.

Egads! This is not the utopic vision of future immortality one may have hoped for. Rather, this is a fate that seems more on par with the traditional notion of hell. But note something very strange and odd about the situation we are now faced with, according to the argument advanced thus far. In the case of all other major miracles, they were just not remotely attainable. If it had been possible to generate a magic green elf, or make the air rush to the corner of the room, then we would have proved that it is possible to willfully violate observed laws of (classical, macroscopic) thermodynamics![6] So as physicists, we all breathed a sigh of relief when we realized the green elf was actually unattainableÑthermodynamics remained intact. Winning the lottery, as a minor miracle, does not violate thermodynamics.

Yet here we are facing the conclusion that the laws of thermodynamics must be violated repeatedly and continuously from some point of advanced age onward, indefinitely into the future (that is, either forever, or until our continued survival is a logical or physical impossibility). It has been said that oneÕs pet theory of the universe can violate almost any law without being in the same trouble it is in if it violates the laws of thermodynamics, and I tend to agree with this rule of thumbÑwe had better be very suspicious of the conclusion we are now facing. After all, we initially thought that perhaps major miracles were possible via the quantum suicide method, yet more careful thinking showed us that there were fundamental reasons why thermodynamics would remain intact. Perhaps the same is the case here (although a violation of thermodynamics is not in itself enough to reject a theory; at this point it simply raises our suspicions).

So let us look more carefully at what actually would, in practice, be likely to happen if our above scenario of infinitely progressing decrepitude came to pass. First of all, we are talking about forever here, or at least until survival is a physical or logical impossibility, which is quite likely to be until the end of the universe or relatively close to it. It seems highly unlikely that, given such a long time period, that continued quantum miracles would really be required, since at some point medical technology would surely be able to restore our health. Even today, while immortality is still a dream, it is not an outlandish or far-fetched one. All that would really be required to keep someone alive indefinitely would be:

(a) the ability to clone and grow a new body, without the development of a new personÑi.e.,  without the development of a cerebral cortex. Given the current state of technology, this does not seem like that remote a possibility. We could have our bodies cloned, and perform a thorough frontal lobotomy early on in the process. Actually, preventing a cerebral cortex from developing would not strictly be required, although it would be needed to avoid ethical problems (at the very leastÑthe procedure would likely still evoke controversy!).

(b) the ability to transplant the brain of an aging patient into the new, younger body. This, again, is beyond our current abilities, but not that far off. Brain transplants have actually been done on chimpanzees, and although the result was paralyzed chimpanzees, they did survive the operation for a time, and a more successful result probably just requires better microsurgery techniques. Again, nothing outlandish or fundamentally new is required.

(c) the ability to repair or replace brain tissue as it dies off. Although transplanting your brain into a fresh body would give you some time, your brain does not have the ability to regenerate itself naturally. However, there have been some successes with transplanting fetal tissue into damaged brains (such as those of patients with ParkinsonÕs disease), and so it is again not that outlandish to imagine a technology that would be able to repair wear-and-tear brain damage now and then.

With these three medical techniques in place, we could keep people alive perhaps indefinitely (barring accidental death, which we have already shown to be a minor problem). This is not to mention other possibilities that may be more fanciful and futuristic, such as downloading oneÕs mind into a robotic or a virtual body. The point is that some kind of immortality is probably going to be available through technology at some point in the future. Even if we disagree as to how easy it will be, this is a question of "when" not "if". So even if the required technology arrives too late to save us, it should be in place for a future generation. Let us guess, for the sake of argument, that these advances are in place 100 years from now. At some time before that, you expect to die. However, let us say, as we argued earlier, that there will be a tiny percentage of worlds where you survive via major miracles, but in a decrepit state. Since major miracles are required, these worlds are extremely physically unlikely. That means if there is any remotely or vaguely possible nonmiraculous chain of events that could lead to the right technology being discovered in time to save your life before quantum miracles are required, then the vast majority of worlds that have you in them 100 years from now will have the required technologies in place in time to save you nonmiraculously.

Of course, this is from your point of view right now (and I am presuming here you are young enough to look at least a few decades ahead). But what about when you are 98 years old and about to die, and still no immortality technology is available? You need either the right technology or a major miracle right now (or at least very very soon) if you are to survive nonmiraculously. At this point, it might take a major miracle to generate the technology in time to save you, and we are back to where we started, in violation of thermodynamics (again, not absolutely against the rules, but definitely cause for worry).

Figuring out exactly what the probabilities are is, of course, far beyond our abilities, given our current knowledge. However, it should be noted that, when calculating quantum probabilities, the future does contribute to the probability of current events, at least in a way. It does not cause them in the usual sense of Òcause and effectÓ (what Aristotle called Òefficient causeÓ). But it does cause them in the sense that the quantum probabilities at any point in the universe depend upon the whole structure of the universe, everywhere in space, considered together (which can include points in the future as well). This is because it is the whole observable universe that is Òmore or less probableÓ given the quantum equations. We cannot really consider the room in which we lie on our death-bed, on the tiny planet Earth, as wholly separate from the rest of the universe.

The most common example of this sort of thing is the case of delayed collapse of the wavefunction. A photon on its way to Earth from a distant galaxy is refracted around an obstruction, and could go either to the left or to the right on its way around the object (which might be, for example, an intervening galaxy). There are thus two worlds, one where the photon goes around the object to the left and one where it goes to the right. Which way it goes has no affect on our consciousness until we observe through our instruments which way the photon actually went. So at the moment we make the observation, the wavefunction collapses into one possible history, which is now the way it was all along from our point of view, even though a moment ago both possible histories were there in superposition. This is the same as saying that a moment ago, two universes with different photon states had identical copies of you, but now the two copies are different, since one has observed a left-path photon and the other a right-path photon. From your point of view, the universal wave-function has "collapsed".

This subjective ÒcollapseÓ applies right back to the moment the photon began to veer around the object, which may be by now millions of years ago. Our observation now helps to determine what happened millions of years ago, far away. This is because the possibilities that are in superposition are of the entire experimental setup, including Earth plus galaxy plus photon, over millions of years, and what we do now to observe the photon may affect what is more or less probable for it to do when it begins its trip. Such correlations are part and parcel of quantum theory. They are not, under the many worlds interpretation, to be thought of as faster than light or backwards in time causality, however, as they are in some interpretations, since one could never use them to send information faster than light or backward in time in any particular universe. This is because each world still obeys the laws of physics, including the prohibition against faster than light travel. What appears as backwards in time causality is actually just an expression of the relative probabilities of various kinds of possible histories of the universe. Each kind of possible history has a certain probability, so that what we choose to do now in observing the photon can affect the probability of what it does at the start of its travels, even though each possible history involves no backwards through time causality. In each history, the photon is emitted, follows a certain path, and is absorbed. However, what we choose to do to absorb the photon can affect the probabilities of what it will do when it is emitted, since the probabilities depend on the whole ensemble of kinds of universes, which depends on the entire experimental setup, even though each possible universe has only local, slower than light, effects. The result is apparent faster than light and backwards through time correlations, even though each individual universe can never involve information transmission faster than light or backwards in time.

Now assume that there are two possible histories of the universe, one in which we manage to make the required medical discoveries in time to save ourselves (and not just our descendents) from death nonmiraculously, and another in which we do not. On the face of it, it seems that there is a 50-50 chance that we will surviveÑfrom a point of view slightly more objective than our own, say that of our family doctor (although he may be unaware of what this probability is). However, the actual probabilities might be quite other than 50-50, since these probabilities depend on the entire experimental setup, and not just how many different mental states will result. Even if there are only two possible resulting kinds of mental states, one where our doctor is aware of our death and one where s/he is aware of our survival, in actual fact there may be far more world histories that result in one than in the other.

It would seem then that far more worlds will involve our death than our survival. But recall that if our survival does not occur nonmiraculously, it will occur miraculously. Given this, as an extreme case, it would appear that what is actually most likely is that we will find out that we have been in a universe all along that will at some point in the future have the medical technology in time to save us. This should be far more probable than being saved by quantum miracles and in a state of decrepitude until the technology becomes available. In other words, of all the world histories it is possible to be in, the entire world histories that involve medical technologies developing in time to save us will be far more common than those without. In other words, to be a person in a universe with the required technology is a more probable kind of person to be than to be one in a world without, since the latter kind of world must be a world that violates our commonly understood laws of thermodynamics.

The quite legitimate objection to this argument is that the probabilities of our survival in the future should not affect our chances of finding ourselves here now. If we require thermodynamic miracles in the future to save us, fine, why should that make our present universe less likely, future and all? What matters in terms of what kind of universe I find myself in now is the probability that said universe will produce something like me up to this point in time. If such a universe will in all probability end up killing me off in the vast majority of futures, then that makes the decrepit me of the future who has survived major miracles an unlikely person to be, but it unfortunately does not make the me now who will die in the majority of futures any more probable.

Unfortunately, this line of reasoning leads us back to observing that, even given that the majority of futures from our doctor's point of view involve our death, only the miraculous ones where we survive in a decrepit state are relevant from our own personal point of view. So we are back to expecting a violation of thermodynamics in our own personal future.

Yet if such a person is so unlikely a person to be, as the universe involved is so improbable, how can we now be so probable, since such a future naturally goes with usÑgiven our current conscious state, we are guaranteed such a future. There is something oddly paradoxical about this result. The problem here is that we are being too loose about which probabilities are the fundamental ones, the ones that should actually matter to us.

From a rigorous philosophical perspective, what begs for explanation is simply my current conscious state now, not my entire life-line. In fact, there may well be no absolute metaphysical fact of the matter that makes me-now a future version of a me-from-the-past. I have memories that make me believe, rightly or wrongly, that I am that person, and that is all that really matters. So we should not be initially concerned with the probabilities of this or that happening to us in the future. We must be first concerned instead with the probability of being the kind of person we are right now. In other words, the philosophically prior question to ask is not Òwhat is likely to happen to me tomorrow or ten years from now?Ó, but rather, Òam I, right now, a likely sort of person to be?Ó

If we assume this latter question as fundamental, we get a version of Carter's strong anthropic principle, in which the universe we are in has the properties it does because we are a rather typical instance of a consciousness, and our universe is rather typical of the environments that must necessarily accompany such consciousnesses (the consciousness and the environment being together one object, of course, not absolutely separable). Thus, it is extremely unlikely to ever find oneself in the situation of being a person who has survived multiple supernatural miracles to preserve their lifeÑlet us call them Òsupernatural immortalsÓ. On the other hand, one might quite likely find oneself to be a person who has either not yet reached that point (call them Ònatural mortalsÓ), or a person in a universe that allows them to continue on through nonmiraculous means such as an appropriately advanced medical science (Ònatural immortalsÓ).

So, to sum up, we have the following types of consciousness (note that the term ÒimmortalÓ refers to living either forever, or up to some point where continued life is either a physical or logical impossibility):

(1)  supernatural immortals (highly improbable): someone who, for some period of their lives, is in an increasingly decrepit state of health and kept alive by the continuous action of supernatural miracles. The supernatural miracles may stop, and the person may enter a phase of their life more like that of the natural immortal, but as they have gone through a supernatural period, their universe and consciousness are still highly improbable, and they are still classified as supernaturalÑnot naturalÑimmortals.

(2)  natural mortals (highly probable): someone who is in a universe that cannot keep them alive through natural means, but whose current state is highly probable. Destined to become supernatural immortals.

(3)  natural immortals (highly probable): someone who is in a universe that can keep them alive indefinitely (until it becomes a physical or logical impossibility), through natural means.

It would seem that we have no reason to believe that either natural mortals (#2) or immortals (#3) are particularly improbable. One type might be more common than the other, but both could be, so far as we can tell at the moment, reasonably common. And since supernatural immortals (#1) are exceedingly uncommon, one might reasonably expect not to find oneself in that situation, and we can thus effectively eliminate it from consideration. Ironically, however, one of the probable types (#2) is destined to turn into the improbable type (#1), once we drop the philosophically prior question of "who is me now?" and ask the inevitable question "who will I be?". Natural mortals always turn into supernatural immortals.

So is there a problem with this? What if it really is highly probable to be naturally mortal, and yet one's future is exceedingly improbable? Since there is no necessary metaphysical connection (that we know of) between the future you and the current you, this presents no fundamental contradiction, only an apparent one. Yet this apparent one is quite disturbing. It seems to indicate that our current consciousness is reasonably likely, and yet we are destined to become something highly unlikely.

But is this the most reasonable way to view the situation? As mentioned earlier, many people lose memory due to a drinking binge, and yet we would probably not suggest that when they are in the drunk stage, living out memories that will be forgotten, that they consider as valid futures only those that include memories of precisely this current consciousness. Indeed, such futures might be exceedingly improbable, perhaps even requiring some sort of miracle. Yet, in projecting from our current natural mortality to a future supernatural immortality, we are doing something very much like this. Why not presume, as we would likely suggest to the drunk person, that we should consider only reasonably probable consciousnesses, and draw only from those the ones that are most similar to ourselves as candidates for possible future (or past) selves?

But, one might object, this is fine for the case of drunken amnesia, since there are still many future selves that, while they lack the ability to remember being our current consciousness, nonetheless are justifiably the same person as us now. But in the case of the natural 110 year old mortal, there are no probable beings that could reasonably be described as their future selves. So, it seems, we are forced in this case to consider our future to be miraculous, albeit minimally miraculous, since what we can expect to happen to us will be the least miraculous way possible to keep us alive.  The paradoxical thing about supernatural miracles is that, just when you think you have come up with some situation that mandates them, you realize that delusion or hallucination could naturally account for the same phenomenon (as when we thought we could produce the magic elf, and then realized that we would be much more likely to just end up hallucinating the elf). In order to consider the most probable supernatural miracle, then, we must try our best to consider whether some natural chain of events would be more likely. One tends naturally to be suspicious of such a result, and yet there it is. Let us examine this possibility in more detail. Assume for now, that we are indeed a type 2 consciousness: a natural mortal. At the advanced age of 110, we reach the limit of natural (minor) miracles, and the first supernatural miracle occurs to keep us aliveÑwe are at that magical borderline between our natural mortal existence, and our new supernatural immortal existence. Let us try to make a reasonably educated guess as to the most likely such miracle. This means not going for the most extreme cases, like the cells in our body spontaneously regenerating just enough to keep us alive. Instead, let us try to come up with something as plausible as we can that is nonetheless consistent with our life experience as a natural mortal up to this point. Consider some possibilities:

Time travel: If time travel is at all possible--even in the far future, at such great cost that they can go back in time only once--then it seems reasonable to presume that in some possible future, someone will travel back in time to save us. But this would no longer be a supernatural miracle. (One might call it a natural miracle, of course, but so is everything else that has ever occurred that had anything to do with allowing us to be born or keeping us alive.) So if time travel is possible, supernatural immortality is effectively ruled out, and we are almost certainly natural immortals (presuming there will be immortality technology at some point in the future, even if distant). Of course, that is a big "if", since time travel might very well not be possible without a major miracle.

Delusion/hallucination: What about the situation where you find out, after reaching 110 years of age, that you were simply living in a universe that allowed natural immortality all along, but were delusional? You find, in fact, that you have been in an insane asylum for years, a psychotic who really believed themselves to be a natural immortal. This may seem implausible, but is it not perhaps more likely that there is a world were such a being legitimately has memories of being you-now than that the laws of thermodynamics would be broken to keep you alive? It seems that there must be. So, if it makes sense to speculate being the perpetually decrepit individual, it makes even more sense to imagine you will turn into a delusional person in another universe! You will skip universes! Bizarre and unlikely yes, and perhaps a major miracle, but surely more probable than continual regeneration.

Alien visitation: Perhaps aliens visit us and give us the secret to immortality. Unlikely, perhaps, but possibly more likely than violation of thermodynamics. However, unless there really are aliens likely to do this in our universe, this will probably have to be called a major miracle.

Premature medical breakthrough: An individual or research group discovers the secret to immortality years ahead of time. Even if some miracles are required for this, it is, again, surely more likely than spontaneous regeneration. I would guess that this scenario is the most likely considered so far. After all, since medical technology is expected to eventually cure us of aging anyway, would it not seem that if a supernatural miracle is required to keep us alive, that it would be a minimal miracle, that simply allowed the technology to exist earlier than otherwise anticipated? Somehow, the natural mortal must get this technology before one would normally expect. However, while this may be reasonably probable (even if perhaps still a miracle) for our current generation, it would certainly seem that to save someone in ancient Egypt, for instance, would require a stupendous miracle, even through this route. How is someone in ancient Egypt to get life-saving medical help without drastic supernatual miracles? On the other hand, wouldn't it still be more likely than continual spontaneous regeneration?Ñperhaps ancient scientists just get really really really lucky quite a few times in a row.

Even if all of these possibilities require major miracles, they still illustrate that jumping to the conclusion that miraculous decrepitude is the most likely scenario is premature. The number of possible ways for a major miracle to occur is truly staggering. We have only touched the tip of the iceberg of possible scenarios we might dream up. And remember, that the important scenario is whichever one is the most likely one, and there are probably vast numbers of scenarios I am not even thinking of.

We are now faced with three possible scenarios:

(1) We are already natural immortals: there is some reasonably probable way of achieving immortality, either by (a) a natural miracle similar to one of the above scenarios, (b) by some natural miracle we have not yet considered, or (c) by completely nonmiraculous eventÑwe simply are already in a world with the technology coming very soon to save us. In any of these cases, we are natural immortals. Celebrate.

(2)  There is no reasonably probable way to achieve immortality, but there are major miracles that result in it, and projection into the future is metaphysically arbitrary. We could choose to interpret this in either of two ways: (a) we are in a probable world with an improbable personal futureÑwe are natural morals destined to miraculously turn into supernatural immortals, or (b) we could assume instead that it is simply philosophically illegitimate to focus on incredibly improbable future selves, no matter how like us they are. It makes more sense to focus on those probable selves most like us. If none are like us enough to meet minimal personal identity requirements, then the whole project of extrapolation into the future is just untenable.

(3)  The only logically and physically possible worlds really are ones in which we die. We are natural mortals. Make your funeral preparations and say goodbye to your loved ones.

Scenario #1 is uncontroversially cause for celebration, but we have no real reason to maintain that it must be true. Scenario #2a is problematic, since we can call its philosophial legitimacy into question. Projecting ourselves into highly improbable future selves may be philosophically untenable. This leads to scenario #2b, which assumes it is completely fallacious to extropolate our current consciousness into the future like this. The view in #2a, that we must view our future selves as supernatural, goes something like this: "the only futures that are relevant to me-now are those for which me-now is a valid past state. Therefore, those are the ones I will use to talk about my future probabilities, even if they are all wildly and fantastically improbable."

"But wait!" scenario #2b chimes in, "there is no metaphysical principle tying you to these futures as opposed to other futures of selves that are close to being future-you but not quite, or for that matter selves that are vaguely like you or quite unlike you. If we are as strict as we possibly can be about applying Descartes' Cogito principle ("I think, therefore I am")Ñi.e. that what begs for explanation is just our current conscious stateÑall we can really do is presume that our current conscious state is a reasonably probable one out of the set of all conscious states (including ours and all other possible creatures). So it is not really legitimate to declare quantum miracles highly probable, just because they are required to maintain a future self that is literally our future. All we can really say is that certain consciousnesses are more probable than others, and one in which quantum miracles had to occur is highly improbable, and so one should not expect to find oneself in such a state. So a more realistic view of the future may be just to realize that it involves many possible consciousnesses, and that some of these will satisfy our arbitrary conception of "me" and some won't. But whether we focus on those that do or not, it is not sensible to view our future as consisting of quantum-miraculous selves. We are free to focus on future consciousnesses that are very much but not exactly future versions of "me-now", but only if we choose relatively probable ones. It is thus more reasonable to extropolate oneself into the future by imagining a self that is as consistent as possible with one's current self, but that has reasonably high probability, than by imagining a more strictly consistent future self that has extremely low probability (i.e. that violates classical, macroscopic thermodynamics).

This leaves us with the rather bizarre suggestion that our most likely future self is someone maximally similar to us, but not really quite us. It might even be someone in a future manifestly not a future of this current world--but this is a result we should view with great suspicion. Am I to conclude that my most probable future either contains a future-me, or a future-not-quite-me, or a future-kind-of-vaguely-like-me? There seems to be something wrong with this last possibility. While it may seem reasonable to extropolate into the future towards a self that is almost a future-me-now, extropolating into someone who is just a different person altogether is plain silly, and violates the whole idea of extrapolating ourselves into the future in the first place. So perhaps we must say that this is a case of true death, since we cannot extrapolate into the future exact-me, and it makes no sense to extrapolate into the not-me. But we cannot view it as true death if there is even one exact-me in the future, no matter how improbable, since we must extrapolate into this improbable self if we decide that extrapolating into all available probable ones is untenable. So are we not then stuck with extrapolating into the highly improbable supernatural self, if we are to extrapolate at all? It may well be, depending on how the actual probabilities work out, that we are left with the choice of #2a, projecting into an improbable immortality, or #2b, simply not projecting at all, which is entirely different than projecting into death. In this case, there is simply no legitimacy into trying to project into the future at all. The question of immortality has not been answered, it has been rendered meaningless.

The curious thing is that the choice between #2a and #2b is utterly arbitrary. The two scenarios describe the same objective state of affairs, only differing in how we emotionally relate to the future, not in anything real at all. There is no objective dilemma here; there is only a dilemma if we insist on being able to interpret the universal wavefunction in terms of some kind of single unified life-line. But there may be no such clear-cut interpretation. Perhaps we will die, and there will follow huge numbers of future natural immortals more or less similar to us, and only a tiny few miraculous ones that more legitimately are us. The less probable future selves that really are us are incredibly unlikely people to be. There is no objective paradox here just because we are now highly likely people to beÑonly a dilemma as to how to arbitrarily project oneself into the future, an operation we are not metaphysically obligated to perform in the first place.

Of course, some may argue against my assumption that projection into the future is metaphysically arbitrary, which I will call the "assumption of subjective personhood". Subjective personhood holds if the mind is purely computational like a machine (an idea known as the strong AI postulate). Although I personally tend to assume subjective personhood, since anything else seems a form of mysticism, there is no clear-cut argument either way, and many people reject the idea, assuming instead what I will call the "assumption of objective personhood", that there is an exact future self that is metaphysically connected to our current self in a special way that makes it absolutely a future version of us, regardless of our concept of personal idenity. In this case, #2b is impossible, and the only interpretation of #2 is #2a, that we are natural mortals destined for supernatural immortality. #1 and #3 are still as much possibilities under metaphysical personhood as under subjective personhood. Please note that the assumption of subjective personhood is not suggesting that our persons have no objective identity. It is suggesting only that there is no reason to suppose that there is an absolute standard by which to determine that a future person is the same person as some particular past person.

Scenario #3 is a scenario we have neglected up to now, in our optimism that perhaps some kind of immortality was possible. However, the sad truth might very well be that, for usü as opposed perhaps to those lucky enough to live in the high-tech future, there is not even a miraculous future in which we can be said to exist in any real sense at all. The argument against this, and in favour of #1 or #2, is that surely there is some kind of miracle that could keep you alive, even if highly improbable. This has intuitive appeal, considering how vast the space of possibilities isÑit must surely have any incredible thing we can imagine. But it is not entirely clear that this is so. A miraculous future might quite possibly be very unstable and unpredictable. Such a future might be more like a half-conscious dream-state with no clear linear direction of time or lawful behaviour, than anything like a solidly envisionable future in something we would call a "world" at all. And what sense does it make to extrapolate into such a state and call it the "future"Ñeven if we do decide that projection into the future is in principle legitimate? In the end, of course, even in this case, it may be quite arbitrary what one chooses to extrapolate into, since all that science really need provide us is a reasonable explanation for who we are right now. In any case, it seems quite possible to me, even likely, that a miraculous future self would lack many of the features required for a stable self identity at all, so #3, true death, isÑsorryÑa real possibility.

Of course, #1 is still a possibility, but I see no reason to conclude that it must be the case. I am personally highly suspicious of #2, since it requires either taking seriously a future that violates thermodynamics, or (objectively no different) rejecting the whole notion of "future" altogetherÑnot a reason to eliminate it, but certainly a reason to be suspicious of it. On the other hand, the fanciful sounding nature of the many-worlds hypothesis, which is the main assumption of this entire essay, was also, I think, reason enough to be highly suspicious of it before there was solid experimental evidence for it, in spite of the fact that there were philosophical systems that predicted it (such as those of Bradley, and in a sense Leibniz). So we have no solid rational argument against #2. I remain, however, suspicious of it, given that there are other reasonable options.

Finally, #3 seems, like #1, to be reasonable, but I again see no reason to prefer it. The biggest reason to suspect #1 is the simple fact that we do seem to be living very close to the point in history where the required medical technology might become possible. But again, this is too unclear to justify any firm conclusions.

In summary, then, I see no reason to currently make any bold claims as to the mortality or immortality of the soul based on quantum physics. We may well be immortal--in which case it will probably be medical science that saves us, not a mystical spiritual soul or quantum miracles--or we may be quite mortal and destined to die. A more complete understanding of quantum mechanics as a rational theory, which we have only just begun to explore in this paper, may help in all this. It may be that the tools to properly answer the death question, and its relation to quantum theory, are just not at hand in our current scientific arsenal. On the other hand, the possible existence of many worlds, which quantum mechanics at the very least makes a real possibility, does make the whole question of mortality much more difficult to figure out. The set of all possibilities is just so extreme, so vast--and issues of probabilities and stability so subtle--that even trying to guess at the actual probability distribution just defeats us.

Conclusion

To sum up, although you definitely don't want to use the quantum suicide technique, if quantum theory is literally correct, and there are other worlds, it is possible to force minor "miracles" to occurÑevents that are highly unlikely but nonetheless are observed to happen on occasion and do not seem to violate the laws of classical physics (or more precisely, are more probable than the failure of detecting their absence and committing suicide).

An obvious question arises: is there any way to do this without having to resort to suicide? I have tried to think of other ways, and certainly the existence of one way suggests that it is at least reasonable to ask whether there are others, but I personally cannot think of any other way to acomplish it. Unless you actually die in the undesirable worlds, it seems that you are at the mercy of relatively uncontrollable probabilities, and must stick with the traditional non-miraculous ways that all living control systems use to control their environment and make their future world vaguely and imperfectly closer to their ideal.

Another question is whether there is any general way around the problem of reliability for major miracles. Again, it is possible, but I cannot think of any. As of now, although the supernatural thought experiments are intriguing, they seem to be just that: thought experiments of the fanciful variety that could not in practise be carried out.

And finally, although quantum mechanics makes it much more difficult to decide on the question of immortality, it is still entirely unclear how to decide the issue. We may or may not be immortal, and if we are, it will very likely come to us, not through supernatural and mystical intervention, but through breakthroughs in medical technology that seem in no way miraculous, but are rather the result of hard sweat and work by determined humans through the ages.


[1]Allan F. Randall. Quantum Superposition, Necessity and the Identity of Indiscernibles. http://www.io.org/~randall/Indiscernibles/, Toronto, 1996.

[2]Michael Clive Price. Many Worlds Theory FAQ. http://kuoi.asui.uidaho.edu/~kamikaze/documents/many-worlds-faq.html, 1995.

[3]David Deutsch. The Fabric of Reality. Penguin Press, London, 1997.

[4]For an introduction to the different interpretations of quantum theory see the following two volumes, both of which are excellant: P.C.W. Davies and J.R. Brown. The Ghost In The Atom. Cambridge University Press, Cambridge, UK, 1986. Also: Nick Herbert. Quantum Reality. Anchor Press/Doubleday, Garden City, New York, 1985.

[5]Hugh Everett, III. Ò'Relative state' formulation of quantum mechanics,Ó In: The Many Worlds Interpretation of Quantum Mechanics, DeWitt & Graham (Eds.), pp. 141Ð149. Princeton U. Press, Princeton, 1957, 1973.

[6]The laws of thermodynamics are not counted amongst the laws of physics that cannot be violated in the quantum wavefunction. This is because the laws of physics do not actually dictate against the laws of thermodynamics being violated! Thermodynamic laws are probabilistic, not absolute. It is possible that this applies to other physical laws as well, and that they all amount to logical necessity plus probability theory, but such a grand unification and reduction of physics to the (synthetic) a priori has not yet been achieved, so we will assume that there are some absolute physical (nonmathematical) laws that cannot be violated.