How is that the same logic I follow? I do not posit different dynamics!
Perhaps I do. Consider an isolated system before measurement, with no introduction of random phases by interaction with the environment. In this case there is still decoherence-based collapse even though there has been no collective dephasing (or whatever it’s called). Is this your understanding as well?
If a quantum computer is possible, then quantum consciousness is possible. I think you are still not understanding my position if you think I said something different in previous posts (I am open to having made a massive blunder here, but I don’t see it yet at least). I am not saying at all that there is a “special property of consciousness” that has any causal effect on dynamics. I am referring to anthropic-selection. For example if their was a consciousness that only existed whenever one plays a certain DVD, then from the point of view of that consciousness, it would be a physical law that a DVD was always playing. But this does not mean that the existence of that consciousness has any affect whatsoever on whether or not the DVD is played!
If it matters in any way that consciousness is special regarding superposition, then you must, at least implicitly. Otherwise, if the unitary dynamics applies unchanged to consciousness, then it must enter into superposition.
I don’t think so. What’s the cause of this decoherence, if not decohering interactions with the environment? To me, decoherence happens if you let a system interact with an environment, then trace out the environment, and find (with high likelihood) that the off-diagonal terms of the density matrix have been highly suppressed.
I can’t see how to interpret your previous posts in a different way than emphatically stating that consciousness/the observer does not enter into a superposition. (I’d elaborate, but my girlfriend is pressuring me to come help doing the dishes…)
Question:
Is quantum randomness the flux that makes our universe interesting?
Without it, would our universe have ended up completely symmetrical in all directions? As in possibly detailed/non-uniform due to interactions of different types of particles, but those interactions would create the same patterns in all directions?
HMHW, I think you are still somehow missing the point of how anthropic selection works. I think I’ll just quote what I wrote earlier:
I just don’t follow your objection to this concept. If consciousness existed in superposition (for example if the universal wave function were conscious) then it would experience the world differently from if it were not. The underlying laws of physics are completely unaffected.
Regarding decoherence, my understanding is that decoherence traditionally explains the dampening of interference effects as phase relationships are randomized by interaction with the environment. But there is more than just interference effects that need explaining in wave function collapse. Take the double slit experiment. Decoherence explains the loss of interference effects. But does it also explain why we observe a “blip” as though the photon is a localized particle, rather than a macroscopic wave? My understanding is that these are two separate questions, that I always put under the name of “decoherence”, because their explanation is both due to entanglement. But perhaps I should be using some different terminology. In order to explain the collapse of the photon’s wave function, you have to take into account the entanglement of the photon with the measurement apparatus. Then you have a game of dominoes of entanglement, ending ultimately with the human observer. My understanding is that the only way to tip over the dominoes so that they fall and lead back to a localized photon is with the anthropic selection from one classical subset of the observer’s wave function, which is entangled with the environment and therefore the photon, such that conservation laws enforce that whichever part of the observer’s wave function is doing the anthropic selection (ie which universe a given observer find himself in) is only consistent with one localized part of the photon’s wave function. One of course can use different language here (I am using MWI language).
Hmm, do you perhaps subscribe to some form of mind/body dualism? If not, I’m not sure how to make sense of your reply, I’m afraid…
Take your DVD example: certainly, on any physicalist conception, there would have to be some physical law that enables states in which the DVD is playing to be conscious, while states in which it’s not playing aren’t. This can be expressed in a supervenience relation: consciousness supervenes on states in which the DVD plays, but not on states in which it doesn’t.
Now, on your conception, the supervenience relation changes whenever consciousness would otherwise enter a superposition—which, again invoking a physicalist (or I think any monist) mind-body relationship, entails a different physical law acting in that case.
Take the states |“0”>|0> and |“1”>|1>: from what I gather, you would hold that consciousness supervenes on these states. However, for the state |psi> = |“0”>|0> + |“1”>|1>, you seem to be saying that it is a special property of our consciousness that it doesn’t supervene on it (as exemplified by your remark to the effect that in principle, there might be kinds of consciousness different from ours that do supervene on such a state, are hence conscious of being in a superposition). Now, the first thing is that this introduces a curious basis dependence into the supervenience relation: if you make a simple change of basis, the state |psi> is no longer superposed, and thus, presumably, consciousness supervenes on it. But again, problems with preferred bases are ubiquitous in many worlds QM. So let’s just stick with the basis used above as the preferred one and delay further questions to some other time.
The real problem is that the unitary evolution from one conscious state into a superposition of conscious states cannot change the supervenience relation. Any property of |“0”>|0> and |“1”>|1>, under the usual quantum dynamics, must also be a property of |psi>. But you assert that it’s not. Hence, you must assume a change in the usual dynamics.
Perhaps this is better illuminated by a Wigner’s friend-style argument. Let’s say you are in a closed room, with an electron in a superposed state. I am outside of this room, which is sufficiently shielded against the environment so that it can preserve quantum coherence (by whatever magical means that might be possible). At some point, you make an observation of the electron. From my point of view, you—consciousness and all—are now in a state of superposition, entangled with the electron. From your point of view, I take it, your consciousness is not in superposition. This runs into a conceptual problem with the ontic nature of the wavefunction: both of us ascribe different wavefunctions to the same reality, which is troublesome if the wavefunction ultimately is what’s real. But that’s not the real problem.
That problem comes about because I can, at least in principle, perform an interference experiment in order to confirm that the room and everything that’s physically in it—including, of course, your consciousness, if it is physical!—is, in fact, in a superposed state. But this leaves only the conclusion that your consciousness—since according to you, it is not in superposition—is not physical, or that the dynamics have to be changed. The former, of course, being the conclusion Wigner himself drew (explicitly referring to your position that consciousness is never in superposition).
The reason the MWI does not usually run into the problem is that it makes no assertion that a consciousness of |psi> is impossible, but rather uses the argument I gave (with which you disagree) in order to claim that this is not consciousness of a superposed state, but of a definite one—either of having observed |0> or |1>.
When you say this, I think you pretty decisively depart from MWI ‘orthodoxy’, such as it may be. In fact, this is perhaps the core claim of the MWI: that the universal wavefunction is conscious (in the sense that (the wave function of) my house is conscious, since it contains me as a conscious observer), and that this consciousness is exactly equivalent to many consciousnesses experiencing distinct classical worlds.
Yes! In fact, that problem (or rather, the analogous one of why one always observes alpha particles leaving straight tracks in cloud chambers) was probably the first contribution to decoherence theory (by Mott in 1929); there actually was a recent paper discussing this, which is the only way I know of it (you can find it here if you’re interested).
That’s of course von Neumann’s famous ‘chain’ of measurements, which played a pivotal role in his understanding of the collapse; basically, because you can put the collapse (or ‘cut’ between the quantum and the classical) at any point along the chain without generating any observable differences, it’s permitted to talk ‘as if’ an observation reduces a superposition to a specific state (the projection postulate), even though that may not be what actually happens (the wave function might only ‘really’ be collapsed once the chain terminates at the conscious observer, for instance, or at any other point; however, it’s crucial that the cut is, in fact, moveable).
This, I think, also illustrates where your account differs from that of the typical Everettian: to you, ultimately, the ‘real’ split only happens when consciousness is reached in the chain (effectively fixing the cut), while typical many worlds accounts are wholly agnostic regarding such a splitting-point: any system that becomes entangled with a system in superposition can be described as being in two worlds; these worlds become wholly separated once decoherence washes out the interference between them. There’s no need for consciousness, and I think your introduction of it both lessens the appeal and opens up the avenue towards new criticism regarding the approach.
Also, and on a different note, I’m not entirely sure how you’re using the phrase ‘anthropic selection’. Usually, anthropic arguments are used to rule out possible physical circumstances, that are however inconsistent with the existence of observers, and which hence are impossible to observe. But in a many-worlds split, both alternatives are consistent with the existence of observers, even the existence of me as an observer, such that I don’t see how such an argument could apply. You seem to be using it to denote the mechanism that chooses which component of the observer’s state is the ‘real’ observer; i.e. that which determines which me, if faced with the possibility of observing spin up or down in an experiment, gets selected to be the ‘real me’, the one which I have the subjective experience of becoming.
This, I’d say, is not a concept consistent with many-worlds interpretations, in which both me’s are the real me, even though I (speaking as one of those me’s) have only the experience of becoming one of them, say, the one that observed spin up. The reason this is consistent is that the other me has equal rights to the same claim. In fact, if you do mean ‘anthropic selection’ in this way, I’d say that makes your approach far more like a modal interpretation than an Everettian one.
About your first question: yes, definitely. It is thought that cosmic inflation, the exponential increase in size of the very early universe, ‘amplified’ quantum fluctuations to macroscopic scales, which then resulted in anisotropies that became the seeds of all the large-scale structure we observe in our universe. About the details of your second question, though, no idea—consult your local cosmologist.
Before replying to the rest of your post, HMHW, I have to stop you right here:
No.
No! Consider, for example, if the DVD player were rather an AI program consciousness.exe. The existence of, or the execution of, consciousness.exe, does not imply any supervenience of the output of the program! Nonetheless, from the point of view of the consciousness, it can derive physical laws for its own universe: consciousness.exe must always be running; the electricity must always be on; the computer must always be on; etc. It is not supervening, it is selecting anthropically a subset out of the larger universe. The physical laws of the larger universe are in no way affected by conciousness.exe! (I saw that you asked later in your post about my use of the phrase “anthropic selection” – I am using it in the above sense – it is exactly the same principle in spirit as how the “anthropic principle” might select universes with certain physical laws out of an anthropic landscape or modal reality, which is why I use the phrase).
[EDIT] I will reply to the rest later when I have the time, or after you have (perhaps?) come around about the supervenience thing (this time it is my gf who needs me to cook dinner).
No, but they dictate that whenever consciousness.exe runs, there is consciousness, computation being, after all, a physical process. So any step in the computation will be a physical state upon which consciousness supervenes. What you propose with your ‘anthropic selection’ still amounts to changing the supervenience relation: an instance of consciousness.exe running (a superposed state being an instance thereof, if the unitary dynamics holds, since then, if two states each are an instance of ‘consciousness.exe running’, so is their superposition), that is neverheless not accompanied by conscious experience.
But still, in the many worlds split, both branches can support consciousness; in the selection out of some landscape, those worlds are excluded which do not support consciousness.
(Just wanted to get this reply off quickly; but I don’t expect you to tackle it right away, so take your time responding to my other points first, if you wish. Also, tsk, girlfriends! Just don’t recognize when one is doing important stuff on the internet… :p)
It looks like iamnotbatman is pointing out that from the point of view of consciousness.exe, it will appear that there is a physical law that says “Electricity is always running through the computer,” even though there is no actual physical law that says that.
iamnotbatman, am I right that you’re pointing this out? (Not that it’s the whole of what you’re saying, but that it’s the main point of your consciousness.exe illustration.)
No, in fact I would completely agree; the laws of physics known by consciousness are in a sense conditioned on the fact that they permit the existence of consciousness. This is essentially my understanding of anthropic selection, as well.
The problem is that as far as I understand iamnotbatman’s argument, he’s effectively saying that consciousness.exe running does not determine the existence of consciousness. In particular, he (and please correct me if I’m wrong) seems to be claiming that there’s at least two different ways to be conscious of a superposed state, one being the ‘unified’ awareness of the superposition, the other being the ‘split’ awareness of each part of the superposition separately. But since the physics is the same in both situations, it would either underdetermine conscious experience—and thus, the supervenience of the mental on the physical would not be given—, or we would need an additional dynamic principle singling out consciousness of the ‘split’ kind whenever a superposition arises.
I on the other hand think that regular many worlds accounts would hold that there’s only one kind of conscious experience associated with a superposed state, and that simply is that of the ‘split’ awareness.
The reason why we may not want to call the superposed state “conscious” would be because it could naturally be factored into two self-aware states that did not communicate. Now, it may be that they can communicate (quantum computation would prove such a thing possible) but that the communication is not complex enough for the whole of the two states to warrant that it be called a “single consciousness” with a straight face. This is related to questions of whether the two halves of our brains are separately conscious, or is it only together through the corpus callosum that a single consciousness is possible, or are all three possible: the left hemisphere is conscious, the right hemisphere is conscious, and also, separately, the combination of the two are conscious. There is, for example, also the possibility that the superposed states are indeed fully conscious, but that they are vastly outnumbered by the non-superposed states. All of these things can live together happily, and then you have to just do that “anthropic selection” boltzmann-brain-style calculus of figuring out the branch densities and which you are most likely to find yourself in as a conscious entity. My point is simply that owing to the empirical fact that we are conscious and do not observe macroscopic superpositions, it is clear that our type of consciousness does not avail itself of superposition, and therefore each separate state (or universe) which we may find ourselves in anthropically selects, through entanglements, portions of other wave functions that are consistent, and that this is the train of dominoes that ends in the appearance of collapse.
The anthropic landscape is the same (although I’m not sure I gather the relevance of your point in any case): there are universes that do support consciousness and those that don’t. There are many universes that are exactly the same but differ in tiny ways, etc. In MWI there are also universes that don’t support consciousness (ie universes in which your brain quantum jumps outside your skull). There are the same universes in the anthropic landscape.
I don’t see how it would. Or to be more specific, if the branches decohere, then there is obviously no communication possible; but if they maintain coherence, then of course there is communication, via simple interference. Whether (large-scale) quantum computation is possible or not has no bearing on that, as far as I can tell at least. (An interesting example of communication, by the way, is the Elitzur-Vaidman bomb tester, in which a photon receives information about an interferometer path even though it can be shown never to traverse that path at all.)
I think this is explicitly giving up physicalism: there can’t be two consciousnesses supervening on the same physical state anymore than there can be two images supervening on the same set of pixels (and yes, I know about that entertaining Monroe/Einstein optical illusion, but note that here the image supervenes not only on the pixels, but on the pixels, the distance of the viewer, and certain properties of the human visual apparatus ;)). For physicalism to hold, the consciousness-properties (c-properties) must be wholly determined by the physical properties (p-properties).
Now, you might argue that there might be a second set of properties that might, in some sense, also be called ‘consciousness’-properties (let’s call them c[sub]2[/sub]-properties)—I don’t know that that’s possible but let’s for the sake of your argument assume it is. But then any physical evolution must obey the separation of c- and c[sub]2[/sub]-properties. Let’s say the c-properties concretely supervene on the total quantum state, and the c[sub]2[/sub]-properties upon the superposed components (in some preferred basis) separately. Then, consciousness of the whole state can never become consciousness of a part of a superposition, and vice-versa.
I suppose you’ll try again to appeal to anthropic selection, but before you do, consider the following: 1) if anthropic selection were able to accomplish this, it would be an extraphysical rule, and 2) this is different from how anthropic selection is usually construed: it selects from a set of universes all those which can support conscious observers, but it cannot help you select one particular universe from these. In other words, it might help you explain why the fine structure constant is about 1/137, because otherwise atoms would not be stable (or stellar fusion would not happen or whatever else), but it can’t help you explain why humans live on planet Earth, or why I had pizza for lunch today, as it can’t select this world against one in which I had, say, lasagna; both worlds are equally well compatible with the existence of conscious observers (contrary to the one in which I accidentally had tofu lasagna, and in my resulting wrath destroyed the universe as we know it).
Yes. And between them, anthropic selection can’t distinguish. But that’s exactly what you require it to do.
Yes, exactly. I have no problem with two consciousnesses supervening on, say, two parts of some superposed state; but I think problems arise if you allow that two consciousnesses supervene on both parts of the state, as well as one supervening on the state as a whole. Then it seems to me that the physical facts underdetermine the mental ones. And even if one allows this, I think (as per the Albert/Loewer argument above) that then the evolution of those consciousnesses (as regards for instance their transtemporal identification) cannot be determined by physical laws.
Concretely, if minds c[sub]1[/sub] and c[sub]2[/sub] supervene on some physical state s[sub]1[/sub], while minds c[sub]3[/sub] and c[sub]4[/sub] supervene on the state s[sub]2[/sub], and s[sub]1[/sub] evolves into s[sub]2[/sub], how is it determined whether, say, c[sub]1[/sub] evolves into c[sub]3[/sub] or into c[sub]4[/sub]? Certainly, the physical evolution is compatible with both possibilities, and thus can’t be the determining factor.
HMHW, you keep asserting supervenience and requirements of extra-physical rules, whereas I don’t yet see how you are substantiating these claims. As in the concsiousness.exe example, I still don’t understand where you are coming from in thinking that the existence of consciousness.exe could in any way superve or require extra-physical rules. The analogy between consciousness.exe and the anthropic selection I am describing is, as I see it, a good one.
Why?
I don’t see the relevance of your examples of what anthropic selection cannot do (although anthropic selection can certainly explain the probability of finding yourself on planet earth, or the probability that you had pizza for lunch today, by considering the branch density associated with all of your possible consciousnesses). Of course anthropic selection cannot predict whether you had pizza or lasagna for lunch today… because it does predict that both universes are more or less equally likely!
Ah, OK, I see there’s a terminological problem here. By supervenience of the mental on the physical, I mean that the physical facts, once fixed, fix all the mental facts, a necessary prerequisite of physicalism. A picture supervenes on the pixels it is constructed from: the facts about the pixels determine all the facts about the picture. Note that the converse doesn’t hold. In your model, this supervenience is not given, and hence, physicalism fails: the same physical facts are claimed to lead to different mental facts (as in, consciousness of the whole state versus consciousness of each part of a superposition). The same set of pixels is supposed to give rise to two different pictures.
The consciousness.exe example is indeed a good one. As I said to Frylock, it illustrates the basic problem I see with your approach: the running of consciousness.exe does not fully determine consciousness, and hence, supervenience of the mental on the physical is not given, and physicalism does not hold.
Because it would be a rule determining which consciousness c[sub]1[/sub] evolves into, something which is not fixed by the physical evolution.
Well, you repeatedly say things like:
Etc.
But this is a selection between alternatives that are equivalent with respect to what the anthropic principle cares about. The second quote makes this especially clear: if anthropic selection were to explain the appearance of wave function collapse, it would have to be able to differentiate between a state in which the observer sees spin up, and a state in which she sees spin down; but since both are states in which the observer exists, it can’t. Anthropic selection only says that you should find yourself in any of these states; it does not select one world from all possible worlds, but at best a class of worlds (those in which the observer exists). But this does not suffice to explain the collapse: we do not observe the wavefunction to collapse to some state, but to this particular one.
Forgot this part. This is again the problem of probability: there’s just no reason why anthropic selection would assign probabilities according to branch densities, or even a reason that branch densities should conform to Born’s rule. I think in order to solve the problem of probability, you’d have to do three things:
[ol]
[li]Solve the problem in the philosophy of probability of how to understand probabilities in a deterministic theory, in particular when there’s no real ‘ignorance’ probability, since in each branch, for instance ‘spin up’ is a certainty, since that’s what defines the branch; furthermore, it’s not the case that ‘spin up’ is found to the exclusion of ‘spin down’—both are observed; in different branches, maybe, but if the two things happened in different rooms, I doubt anybody would be comfortable with making any probability assignment. You could move to the conditional probability, ‘spin up is observed in room 1’, but translated to a many worlds picture, this would mean that ‘spin up is observed in branch 1’, but the branch itself is defined by the observation of ‘spin up’ within it, and the probability of spin up being observed conditioned on spin up being observed is 100%.[/li][li]If you can do the above, you then have to argue that the Born rule probabilities apply to branches. You seem to think that’s trivial, following from Gleason’s theorem as in ordinary quantum mechanics, but that’s not the case: Gleason’s theorem provides a measure on events, that is, the linear subspaces of Hilbert space; but what you need is a measure on worlds. To find a connection between one and the other is difficult because many-worlds quantum mechanics does not obey the standard eigenvalue-eigenstate link, according to which a property only has a definite value if the system is in an eigenstate of that property; that is, for instance, x-spin only is either determinately up or down if the system is in an x-spin eigenstate. But in the MWI, x-spin has a definite value in each branch even if that is not the case, and thus, the event structure and the world structure are not the same, and it’s not obvious why Gleason’s theorem should give probabilities for the latter.[/li][li]If you’ve managed to clear that hurdle as well, you still have to argue why the measure provided by Gleason’s theorem is the right one to use. In ordinary quantum mechanics, this is simple: there simply is no other measure on the space of linear subspaces of Hilbert space. But this is not the case on the space of worlds: a simple alternative is the combinatorial counting measure, which assigns equal probability to every world, and thus, to every sequence of measurement outcomes; but this doesn’t reproduce the observed quantum statistics.[/li][/ol]
So to me it seems clear that things are a lot more complicated than simply saying that the anthropic principle somehow produces the correct statistics by appealing to branch densities.
I don’t understand. You have to expand this. When you say “the running of consciousness.exe does not fully determine consciousness” … why do you think I was implying that it “fully determines consciousness” (and what do you mean by that)?
Why does there have to be a rule telling us what consciousness c[sub]1[/sub] evolves into? What does that have to do with anything?
Why?
Why should it? Neither does any interpretation of quantum mechanics that is not a hidden variable theory.
It explains collapse perfectly well. There are an infinite number of parts of the wave function, each of which is tantamount to a separate universe, and each of which includes an observation of collapse to some state. The point is that each observes collapse. That is all that needs to be explained (regarding collapse – in order to additionally reproduce QM you must assign a probability to each possible outcome). The question of why we collapse to this particular one is the entire point of the anthropic reasoning I have put forward: each observer thinks they are special, but obviously they are not: they are a tiny part of the universal wave function.
