The problem of foreknowledge

Comments (48)

1. Interesting. It seems to be quite close to my intuitions on the subject, at least on the days of the week where I’m close to being a classical theist…

2. One thing, though — why tie foreknowledge so closely to computability? I guess you’re assuming a nominalism [1] regarding possible universes, so that they have to have been ‘simulated’ to be meaningful options for instantiation. I imagine that one could view the set of possible universes as existing more Platonically, which would allow God to select even non-computable ones for instantiation.

   On the other hand, that sort of Platonism is a very strong assumption (and probably makes little difference to how the argument cashes out in the end), so it’s probably best not to make use of it.

   [1] Apologies if this physicist is misusing the term…

   1. No, that’s a reasonable use of “nominalism”. Yes, I am not assuming that God can pluck universes out of the Platonic ether, but suppose he could. Two things follow. One is that the design of the universe would be independent of God, which is disagreeable for traditional theists. The other is that God still has to evaluate these designs, so the cognitive problem remains.

      1. Well, with regard to ‘design’ I’m inclined to think that the bit that matters isn’t so much the notion that a universe was constructed but that a given universe is *intended* and sustained in existence. So I’m not sure that the ‘traditional’ theist loses that much by conceding that point, or that at least the version of traditional theism that I believe on Thursdays doesn’t. 😉

         With regards to the evaluation problem, I’d suspect that a P-God [1] would have access to non-computable mathematical brute facts like Chaitin’s Omega (as they’d just be lying around in the Platonic Realm) which (IIRC) encode information regarding the results of non-computable problems. That might help. (Though there’s probably a class of these objects that P-God can’t have access to due to the oddities of meta-mathematics. They might not be *important* ones for the task at hand, though.)

         Anyway, I’m not going to carp too much, because this is certainly interesting, and it’s likely that if it can be gotten to work for this specific set of assumptions, then something similar can be made to work for slightly different ones, possibly even allowing for a completely non-computable universe [2].

         [1] As opposed to a p-God, which is the mature form of a p-angel…

         [2] Not that I think that we live in such a thing; while it may not be entirely computable, it’s computable enough that I can simulate bits of it…

3. God, as has been argued on a philosophical level, is called “Infinite.” Being aware of the future, past, and present argues that he (or He) [or I/it] exists in all times in the same manner at every moment. As Hawking said, “God absolutely plays with dice; He simply knows the outcome.” Rolling the dice is done for the sake of the situations dependent upon it, rather than for h/His [i/Its] sake.

   The issue of foreknowledge can only apply in a universe where [a] g/God must obey the laws pertinent to that universe, which makes h/Him [i/It] a part of it. Extending said existence outside the universe (and its laws), this ceases to be an issue.

   It should be noted that when man has actively created gods, some of them (such as Lovecraftian beings) are derived from outside of a universe, and the very premise that these entities must adhere to the perceptions of those inhabitants is fallacious: They are outside of known or knowable ken, and are thus unknowable. We would likely be as incomprehensible to those beings as they are incomprehensible to us.

   It should be said that if [a] g/God were inscrutable in this manner, it would be more of the deist’s concept than the typical christian (read: Protestant and Catholic) deity.

4. As I tried to make clear I have, as a one time student of meta-mathematics, serious doubts about a computable universe. We know as an established fact that a Turing machine cannot even compute all of the true statements of mathematics so how can the complete Universe be computable?

5. Why the obsession with a single created universe? Like the intelligent design people with their out-of-awareness assumption of a single designer (see here for more, you seem to assume that only one universe can be created by one creator.

   Given the sort of God you’ve postulated–able to run humongous simulations and assess the outcomes wrt some utility function–there’s a fourth alternative that doesn’t require putting that God outside of time. It is for that God to create a sample of all possible universes in which the instances have the characteristics that are more likely to produce the kind of world he/she/it desires. In other words, the creator’s task is to skew the distribution, not choose a single lottery ticket.

   Put another slightly different way, you’re treating God as an optimizing entity (choose the best) when it might be better treated as a satisficing entity (choose a bunch of likely-to-be-satisfactory instances and instantiate them all).

   1. Nice response, but I am trying to attend to ordinary theism. If I were God, then I’d just create lots of universes until I got the result I wanted, but then maybe God’s on a budget.

6. Interesting, but doesn’t it beg the important question?

   Is there a God?

   This isn’t trite. What’s the point of all that writing in the absence any evidence that such a God even exists?

   And even if he exists, what makes you so sure you understand his/her properties well enough to speculate on motives and behaviors? Do you know something about God’s plans? How did you find out?

   1. Larry, I am trying to explore the conceptual ramifications of views I do not hold. I thought that was clear.

7. Best comedy post ever, starting with paragraph “To answer this, Phil Dowe and I…”.

   1. >8-P

8. Interesting, but before you dive into the mechanics of a God creating and/or maintaining a universe, shouldn’t you be more precise about what the universe is? Your concept of a universe has several important assumptions or priors which may or may not correct. You should at least state those assumptions. Such as:

   1). That what is objectively observable is all that exists, and that everything that can be objectively observed is actually currently being observed. In addition to ThonyC’s arguments, these are additional reasons why the universe may not be computable, although what is computable to a God and his unknown resources may be different than what is computable to us. Some of us believe that consciousness does not exist because it is not objectively observable, and some of us don’t. Consciousness or subjectivity may not describable in objective physical or mathematical terms, or computable as we know computation.

   2). You’ve already covered the presentism vs block time models. To a being viewing the universe though time as a dimension, predictions are (relatively) easy and 100% correct. But… is it a matter of perspective or ontology – i.e. how can it be perceived both ways? Even if you conceive of a God who sees through block time, you must still allow for our perspective and observations (chaos, uncertainty principle, etc).

   3). That causation is objectively real, and that it occurs from the bottom-up. Physics only gives us a license for describing consistently correlated events in time and space, and actually makes no claims for any ultimate causes, bottom-up, top-down, extra-universe, or otherwise. Causation means something beyond that, often subjective (i.e. you cause something to happen by virtue of your will).

   1. These are all good questions, and they deserve to be discussed, but they all are in the context of ordinary metaphysics. For example many people have discussed worldlines and alternate universes, not to mention Lewisian modal realism. As to our limitations, well, they are assumed here, yes. I take no position on whether chance is Laplacean/Humean or irreducibly real and something God is subjected to as well (here, read “God” in an Einsteinian sense as “the way the world is”).

      As to causation, well that’s a morass, but I really do not see that it matters in this context. So long as whatever causation is, it is not messed about with by God, that’s all that matters. A good friend who is a philosopher of physics denies that causation is even meaningful in physics (because the physical is a manifold block, and all physics does is summarise the actual relations between classes of things), but then he’s a Platonist too, so I don’t know how much I can rely on him… 🙂

      On free will, I am a total determinist. Whatever it is that we do when deciding, and however the universe is, our choosing is an outcome of nature, and there is no slop besides the ordinary physical slop of a mildly indeterminate universe. In other words, I’m a secular Calvinist.

      1. Well yes, it’s all within the context of ordinary metaphysics, but the specific universe model that we choose may significantly affect the dynamics of the whole discussion. For example, if one were to adopt an extreme positivistic model, where only what is actually observed is real (at the macroscopic level), then the universe would be ontologically much smaller than it appears to be, compressed by its own laws. The universe would in fact be a facade – a vast exaggeration, and God’s job would be many orders of magnitude easier. And if you are a determinist on consciousness and free will, you should state that in your assumptions. Some of us may entertain the idea that consciousness is correlated with the physical, but not physical itself.

         On causation – the lack of it is actually a consequence of the scientific method. We observe events and come up with laws and formulas that describe them. But the laws are not the cause. They do not cause each individually observable event. Our minds come up with causes that may be useful for a time, but change as more data becomes available. In a way, this is actually quite consistent with a simulation, where the actual cause of each event would come from outside the observable universe, on some computing substrate that God is running, deferring actual causation to God’s level, where the fundamental physics would be much different. But I guess I was thinking along the lines of God not using a simulation, but rather some kind of universe-building method or machine, where many universes are built and one is selected. In this case, the universes would be self-sufficient, and causation matters. BTW, even if this God were able to see in block time, it would still be quite a task to search for the planet he liked among at least (100 billion) ^ 2 stars in the universe. The much more difficult task of searching though time is eliminated, though.

      2. Just some follow-up thoughts on causation:

         What we actually see all around us then, is a highly correlated physical dance of events – both microscopic and macroscopic, with no truly observable causal dependencies between events. Events are only more or less correlated in space and time. Some groups of events are more intricately correlated than others, and those are the ones our minds tend to see causal relationships in, and interpret as causal, even though there may be some stochastic behavior. Others are not so correlated (radioactive decay, quantum fluctuations, etc), and those appear utterly random, isolated, and uncaused to us.

         However… even if there is no true causality, one must still ask what makes some of these event groups so highly correlated. Where do these relationships come from? We know these correlations don’t come from our scientific laws, which are just our interpretations. If there is a source for these correlations, it is not in our observable universe. Some options are 1) those relationships exist in a hidden layer or variables, perhaps something like Bohm’s implicate layer, or 2) we are in a simulation, and those relationships exist in the real physical universe above us – if a richer superset of our reality with true causation (whatever that is) can even be understood by our simulated experience as “physical”.

9. Quantum states are indeterminate in time as well as space. On the quantum scale “before” and “after” are indistinct states and causality inherently doesn’t exist. Thus being in a state of “eternity” wouldn’t grant a being any more certainty than existing in a temporal state.

   1. Interesting point. Could you expand upon it?

      1. Space and time are not distinct phenomenon. Space is time, and vice versa. So anything a quantum particle can do in space, like exist as a distributed wave function, it can do in time. Now, on average, particles move forward in time. However, on small time-scales a particle’s position in time cannot be fixed with certainty. A particle won’t exist in a single, granular instant of time. Instead it will exist at multiple points in time simultaneously (as much as that word has any meaning in this context) with various probabilities of existing at each point.

         This means that if you interact wit a particle at any point in time that it is super-positioned over, it reacts to that interaction over all of those times. This means an effect can occur before a cause. It also allows a particle effect to be it’s own cause. This means the relationship between cause and effect breaks down at small distances over small periods of time.

         If at it’s fundamental level the Universe lacks causality, then the Universe lacks true causality at all levels. What we experience as causality is an illusion. We typically don’t notice quarks behaving badly. We don’t demarcate the passage of picoseconds during our normal lives. We don’t notice that causality only continues by sheer luck. It’s not against the laws of quantum mechanics or statistics for you to spontaneously travel into the distant past. It’s just very unlikely to happen over such large stretches of time.

         What this means for your theoretical deity is that the Universe is inherently unpredictable. Even if you set up the initial conditions; those conditions can alter themselves outside the bounds of causality. Thus, you can’t predict the outcome of any interaction because you can’t predict interactions that can alter their own initial conditions at random. Even from a timeless state with a Universal view, the Universe would simply be too fuzzy to see correctly. From the “exterior” of the Universe, such a being would perceive the Universe as we perceive individual quantum particles: fuzzy and indistinct; knowable only in part; dominated by uncertainty.

   2. Events are still ordered and related in some fashion, though, even if it’s fuzzy. Otherwise it’s hard to see how causality could emerge on a coarse-grained level, let alone how it could be possible to run numerical simulations of candidate quantum gravity theories.

      It’s not going to provide any more of a problem to a being in eternity than position-momentum fuzziness would, anyway.

      1. You assume causality where none exists. God plays dice. Get over it.

      2. That hardly addresses the point I’m making. And while you might be correct for pure QM, whether that holds up for quantum gravity (where some structure that grows into causality as we understand as the theory is coarse-grained is, IIRC, present in most theories [1]) really rather remains to be seen.

         [1] Say, the kind of structure you have in Loop Quantum Gravity or Causal Set theory.

      3. Actually, I’m not sure if we mean different things by ‘causality’ here or not. (I mean it in a sense that ‘space-time quantum A’ can be said to be related to ‘space-time quantum B’.)

         I’m not convinced that this means anything more than that the God in question understands the universe as a probability distribution of states from which certain macroscopic possibilities are selected to be instantiated, though; ‘particles’ and such like aren’t really part of the world’s furniture. What would be real would be the universes’s path integral (in a Lagrangian formulation, given that we’re ‘outside’ the spacetime) and a given set of detector clicks that are compatible with its distribution. We only see ‘fuzziness’ because we don’t have access to the world’s ontology at our level.

      4. Isn’t this a little like saying that sulphur atoms must smell like rotten eggs, otherwise it’s hard to see how rotten-egg-smell could emerge on a coarse-grained level, let alone how it could be possible to run numerical simulations of said rotten egg smell?

      5. Well, in this case it’s more that going by the models available, it looks a lot like there have to be structural constraints of some kind on whatever it is that space-time is built up from, because otherwise when you coarse-grain you don’t get the correct space-time. These constraints may be more loose and statistical than what seems to be the case on the coarse level, but they’re still constraints (such as the partial ordering requirement in causal sets, etc). It’s similar to effective field theories — the underlying theory needs to have certain properties (say, particular symmetries) in order for the higher level effective theory to have the properties it does, even if there could be several plausible underlying theories.

         It’s closer to saying that water molecules must have particular properties so that when there’s a large enough number of them they behave en masse like water as we understand it rather than the example of sulphur atoms.

      6. Or, more briefly — moving between scales like this (from Planck scale to LHC scale to semiclassical limit to classical limit in order to check that the results you get are empirically sound) is pretty much standard procedure in physics, as opposed to trying to link sense perception and chemistry, where there’s no mathematically well defined way of linking the two.

      7. It’s bad netiquette to split your posts like that.

         My point was that small structures don’t experience causality. Large structures are built from small structures. So if small structures don’t experience causality, the large structures built from them don’t experience causality. This means the Universe doesn’t experience causality. What we perceive as causality is just luck. There is a strong probability that at macroscopic levels causality-like phenomenon will emerge. These causality-like phenomenon are not the same thing as causality. At large scales there is no causality, merely the illusion of it brought on by the presence of causality-like phenomenon.

         Any theory of quantum gravity would need to describe all observed phenomenon. Because the lack of causality has been observed, quantum gravity would still predict an unpredictable world. That statement is not as impossible as it appears on first blush. While the Universe is unpredictable, Chaos theory shows that it is predictably unpredictable. We can’t predict any particular state of the Universe; we can predict what states are impossible and the probability of the remaining states. Ta da! Quantum mechanics!

         Chaos theory shows that in chaotic systems small changes lead to divergent results. If you alter the starting conditions of a simulation only slightly, that difference compounds over iterations. Thus very similar starting conditions can lead to very different processes and ending conditions.

         The lack of causality allows a system to modify its own start conditions randomly. In the context of Chaos Theory, this means the Universe is unpredictable. The condition of the Universe at any given moment is not dependent on the starting conditions of the Universe. Thus, when viewed externally (not that the word really means anything in this context), the Universe wouldn’t appear as any one particular set of conditions spread throughout time. It would instead look like a series of conditions at each moment. That is to say, the Universe would appear to be in a state of quantum indeterminacy.

         Like any quantum state, you could observe on particular feature of it, one specific condition at one specific moment. You could not, however, observe all conditions at once, or even all conditions at one particular point in time. This means the god couldn’t view all conditions at once, which would be required of an entity existing outside of time. Thus, no god could ever predict the Universe. Thus the god is not omniscient in the context of the Universe. As a consequence, such a god would also not be omnipotent.

         Suppose John’s deity creates two Universes side by side. These Universes are exactly identical down to the smallest detail. Once set running, these Universes would immediately begin altering their starting conditions at random. Thus, at any point in the progression, even at the starting moment, the state of the Universe would be indeterminate. The moment a Universe is created, it is instantly and unpredictably different than how the creator created it. Thus, while such a deity might be able to create a predictable Universe, such a Universe would not be anything like the one we reside in.

      8. Sorry about the post splitting — I’m a little distracted at the moment.

         You seem to be assuming that John’s God is observing the universe in the same way that subsystems inside the universe are observing it, and therefore collapsing the state vector. I’m not sure that follows — John’s God isn’t a system within the universe, and therefore doesn’t count as an ‘observer’ in that sense, because it’s not interacting with it in a way that could collapse the wavefunction. An uncollapsed state vector is the sort of thing it could know, in the sense that it’s a mathematical truth about a universe that can be modelled.

         Secondly, QM proper doesn’t say anything about space-time — it posits an external clock, which is partly why quantising gravity is problematic. Quantum field theory incorporates special relativity, and it seems to have a fair bit to say about causality (though what it does say is a little unclear; certainly the *interntion* is that it respects Einstein causality) .

         Finally, this is interpretation dependent (you seem to assuming some version of a Copenhagen-style account). Assume a starting state vector/wavefunction as your initial conditions. Evolve it according to the Schrodinger equation. This is completely deterministic. The only random, non-computable component is that arising from wavefunction collapse. But, as supporters of the many-worlds interpretation tell us, there’s no need to posit a collapse, in which case QM is completely deterministic, and you obtain all possible outcomes as the system is evolved and the subsystems interact to form relative-state branches. But John’s God seems only to be interested in one of those outcomes. So only one set of the branches is in fact instansiated out of all the ones available. This appears random. But only because it’s a subset of the deterministically generated whole.

         It’s even better if one goes with a transactional account — then the whole thing may well be completely determined by the past *and* future boundary conditions.

      9. It seems though, that all of this assumes that the deity in question is not in control of QM somehow, and that QM is absolutely fundamental rather than being simulated, or an observed behaviour of something deeper – like newtonian physics is to relativity. But I guess you can always make your deity do anything.

      10. Sorry about the post splitting — I’m a little distracted at the moment.

          Don’t worry about it. It’s just that it makes it harder to follow the thread.

          You seem to be assuming that John’s God is observing the universe in the same way that subsystems inside the universe are observing it, and therefore collapsing the state vector.

          How would one observe something without interacting with it? If you can’t interact in someway with a thing, then, from your perspective, it’s no different from that thing not existing.

          An uncollapsed state vector is the sort of thing it could know, in the sense that it’s a mathematical truth about a universe that can be modelled.

          Yes, it can be modeled. We model them all the time. But they wouldn’t be deterministic. They wouldn’t obey causality. And thus any given state of the Universe is effectively independent of the starting conditions. That things can be modeled in no way means that they aren’t acausal as long as the modeling is purely statistical. Thus we live in a Universe that can be modeled but does not experience true causality. Like I said, god plays dice. There are no unknown local variables. There is no causality. It doesn’t feel right. Oh, well. Get over it.

          You seem to be assuming absolute knowledge on the part of John’s god. John stated, however, that it is not an all-knowing entity. Instead it has simply run multiple simulations of the Universe and selected the set of start conditions that would lead to the existence of all the deity’s desired states being achieved. My point is that such a system is impossible because causality doesn’t exist at the quantum scales. The start conditions themselves are subject to change randomly.

          Secondly, QM proper doesn’t say anything about space-time — it posits an external clock, which is partly why quantising gravity is problematic. Quantum field theory incorporates special relativity, and it seems to have a fair bit to say about causality (though what it does say is a little unclear; certainly the *interntion* is that it respects Einstein causality) .

          The space-time continuum has been proven true. Thus quantum particles don’t exist in space and time; they exist in space-time. Anything such a particle can do in space it can therefore do in time. One of the consequences of this is that quantum particles are not merely super-positioned over space, but also over time. There are a variety of experiments that have demonstrated this phenomenon. Noble prizes have been awarded over proving this fact. Thus it is beyond question at this point if quantum-scale particles acting over short timespans exhibit causality. They simply don’t.

          Finally, this is interpretation dependent (you seem to assuming some version of a Copenhagen-style account). Assume a starting state vector/wavefunction as your initial conditions. Evolve it according to the Schrodinger equation. This is completely deterministic. The only random, non-computable component is that arising from wavefunction collapse.

          My whole point has been that you can never determine the starting conditions. The starting conditions are subject to random influence from future events. Those changes made by the future, according to Chaos Theory, will wildly alter those future events that influenced the past in the first place. Thus, the starting state can only be known as a series of possibilities of varying likelyhood. It cannot be known determinalistically.

          But, as supporters of the many-worlds interpretation tell us, there’s no need to posit a collapse, in which case QM is completely deterministic, and you obtain all possible outcomes as the system is evolved and the subsystems interact to form relative-state branches.

          Many-worlds requires multiple Universes. Copenhagen does not. As there is of yet no way to experimentally differentiate between the two. thus I apply Occham’s Razor to the problem. Many-worlds requires an extra circumstance, that other Universes exist. Copenhagen only requires the one, observed Universe. As Copenhagen requires fewer assumptions, it is preferable. Give me some reason to believe otherwise (other than the dersire for causality) and I’ll happily accept Many-worlds.

          It seems though, that all of this assumes that the deity in question is not in control of QM somehow, and that QM is absolutely fundamental rather than being simulated, or an observed behaviour of something deeper – like newtonian physics is to relativity. But I guess you can always make your deity do anything.

          John’s supposed god is not all knowing nor all powerful. Thus there is no reason for it to know all possible quantum states. That is without addressing the notion that knowing all possible quantum states is even possible.

          QM exists in our Universe. So in order for our Universe to have a creaator like the one John supposes such a creator would need to create the conditions in the obervable Universe. Such a deity might very well be able to create a Universe without quantum mechanics. It simply wouldn’t be our Universe.

          I don’t know if you’ve heard, but Newtonian physics is wrong. It’s not as wrong as those things that came before it. It just isn’t right.

      11. “How would one observe something without interacting with it? If you can’t interact in someway with a thing, then, from your perspective, it’s no different from that thing not existing.”

          If I’m simulating something, which is what this God is doing, then I’m not interacting with system in the same way that subsystems of the simulated systems are with each other. I’m defining the terms of the simulation itself, and how the subsystems interact with each other. (Copenhagen-style collapses are a property of one subsystem being defined as a Classical observer while the subsystem being observed is allowed to remain quantum. However, when John von Neumann writes this down formally, he’s neither the Classical observer nor the quantum subsystem.) I’m afraid that you’re making a category mistake here.

          “My whole point has been that you can never determine the starting conditions. The starting conditions are subject to random influence from future events. Those changes made by the future, according to Chaos Theory, will wildly alter those future events that influenced the past in the first place. Thus, the starting state can only be known as a series of possibilities of varying likelyhood. It cannot be known determinalistically.”

          The starting state is a wavefunction, a superposition of states. This can be defined by the person doing the simulating, and evolves deterministically. This is what boundary conditions are in QM.

          “Many-worlds requires multiple Universes. Copenhagen does not. As there is of yet no way to experimentally differentiate between the two. thus I apply Occham’s Razor to the problem. Many-worlds requires an extra circumstance, that other Universes exist. Copenhagen only requires the one, observed Universe. As Copenhagen requires fewer assumptions, it is preferable. Give me some reason to believe otherwise (other than the dersire for causality) and I’ll happily accept Many-worlds.”

          Well, that’s debatable. All Many Worlds requires is the unitary evolution of the state vector, not unitary evolution + collapse.

          This isn’t about Copenhagen vs. Many Worlds in the conventional sense, though. What I’ve done is describe what someone with John’s God’s aims would do to enumerate the universes it can get from simulations using quantum mechanics. Because simulating all possible universes that arise from a given starting wavefunction [1] and then picking one to make real is mathematically very easy (in principle) and *identical* to what you’d get with the Copenhagen interpretation, because only one of the evolved universes is real. Apart from the picking of the universe, there is nothing going on here other than mathematical representation and modelling using the well-defined formalism of QM. John’s God has a wavefunction and a Hamiltonian he uses to evolve it. That’s all.

          [1] In fact, if you’re studying the quantum mechanical evolution of a wavefunction, whether analytically or through numerical simulation, this is what you do — since collapse can’t be modelled except by fiat. This is all John’s God needs to do.

10. I’ve got no problem with the postulation of imaginary entities in thought experiments. In that respect, the notion of an omniscient or omnipotent god is not a lot different than that of a test particle in physics and may indeed have its uses. I do want to note, though, that there’s a problem with even a merely notional God. You can attempt to get around the foreknowledge problem by assuming that God, being eternal, is transcendent with respect to temporal events and sees what will happen in the future as we can see what is going on to the left. The trouble with this notion is the following: a God who really is outside of time simply cannot be said to act because the category of action strictly requires something like a before and after. Of course one is free to assume that God exists in something like a supertime in which something can be done–the trouble with theology is that one can assume anything at all to meet each objection as it occurs–but in the absence of this shift, shouldn’t you admit that the required deity is merely a naked light bulb in the cosmic attic, i.e. a non-doer and therefore a non-person and therefore quite irrelevant to human concerns?

11. So not the God of the Old Testament then, omipotent and omniscient, but rather just this guy who’s acting in a manner currently indistinguishable from magic.

    On the other hand, it could be the god of the Old Testement,

    “Verily, God laboured for six Days on his simulations. And on the seventh Day he sayeth, “Sod it. That’s close enough. I’m taking the day off.”

    1. Actually the God of the Tanakh varies in what he can achieve. He’s not omniscient – we know that from Genesis 3. He makes mistakes and changes his mind (repententh). He’s not even the only deity. He’s very far from the theologians’ God.

       1. A god of the sub-Planck gaps maybe?

          1. Or maybe he’s as thick as two short Plancks.

12. I haven’t studied metaphysics for a while so please bear with me, but wouldn’t these ‘visualisations’ or ‘simulations’ of possible worlds that God perceives have to model all of the events and causal chains in each universe? Wouldn’t such simulations be indistinguishable from a ‘real’ universe?

    If so, how do we know if we are in a simulation or in the ‘real’ universe? And as there are going to be far more imperfect possible worlds than the single best of all possible worlds (even if God satisfices) then it is more probable that we are not in the best of all possible worlds, and a form of dystheism is surely supported, for God would have created this universe without it being perfectly good.

    Or, would God’s ‘simulation’ of a universe be sufficiently distinct from the universe that obtains to exclude this possibility?

    1. If something “exists” in the “mind of God” (or in his simulator) I think they are not “existent” in the same sense for a theist as the realised world. There are positions that are broadly Platonist, but typically theists prefer to think of the world as separately existing from God’s mental formulation. Hence my phrase “pour in the ontological cement”.

       1. OK, so God selects one of the simulations and causes it to become the actual world. In a sense, God creates a copy of a universe that was comprised of a sort of divine [i]res cogitans[/i] out of a kind of [i]res extensa[/i] (or “pours ontological cement” onto the original) and in doing so it becomes separate.

          I suppose my argument is that even if we feel the mind of God to be separate, the simulations have still in some sense obtained – they are concrete and not just abstract possible worlds.

          So even if we distinguish between different senses of existence, I’m still not sure that as observes we would be able to say which one we are in.

          1. While I don’t know what ontological status a divine res cogitans has, in ordinary terms, a simulation and the thing being simulated are distinct. Consider a simulation of the solar system. The “planets” do not literally orbit their primary. What happens is that formal representations behave in a particular manner in a computational model. This is very different from actual plans with mass deforming spacetime. So I don’t think that a simulation is concrete; that is, the simulation is concrete (because in our world it has to be run on a physical computer), but the entities it simulates are abstract.

    2. But (and rest assured, this will be the final, no doubt misguided, shot that I will offer) when we create a computational model of a mind, one that factors in all of its functional properties, wouldn’t that simulation also be conscious?

       And thus, even if the simulation is considered abstract, the elements of the model that simulate minds would be conscious and would perceive it as if it was “real”?

       Or, would the theist state that such simulated minds would lack a soul and/or qualia – in which case they are p-zombies..?

       1. I don’t think so. Apart from my complete lack of belief in p-angels (we are all p-zombies), any perception is just itself simulation. If it isn’t real planet, then it isn’t a real brain capable of perception. The representation is not the thing being represented.

       2. The problem is that this property of your p-angels that we call subjectivity (or awareness or consciousness), and that most of us think we have, is that it is not a physically expressible or objectively observable property of any system in the universe. It is not a force, or object, or dimension, or anything quantifiable or qualifiable by physical interaction or causation. It is simply not in the language or vocabulary of physics. Physical processes – even neurological ones, all have external, objective properties verifiable by many observers. But no physical process that we can understand has hidden subjective properties. That would add something new to physics. Processes in physics, no matter now complex, have external, objective properties, and that’s all. Since most of us observe that we are subjectively aware of many things (excruciating pain can be more real than any scientific experiment), we can infer that current physics is incomplete. To say that we are unconscious p-zeds is absurd. I’ve been conscious and completely unconscious (IV valium will do the trick), and there is a huge difference. I observe that I am subjectively aware. To deny that observation is unscientific.

          Even if you find specific neural circuits that are consistently correlated with consciousness, that would still not explain how those neural processes cause subjectivity, because the property of subjectivity is simply not in the causal language of objective physics. There’s no way to get there from here. Oh yes indeed, there is a definite observed correlation between physical processes and the subjective – just have a neurosurgeon stimulate parts of your brain, or take a shot of morphine, or even just watch a movie, and you’ll see. In fact, every observable material interaction you can have appears to causally affect your subjective experience. But that doesn’t really explain anything. All you’ve done is show a correlation between some objective process (macroscopic, neural, or molecular) and your subjective experience. You haven’t explained how, nor can you with current physics. And I’m not even going to get in the more subtle problems of time and space objectivity, or the matter of conscious vs physical perspective.

          As far as the simulation is concerned, anything can be simulated *except* your consciousness. It is probably theoretically possible for objective reality to be simulated to many levels, but you must be a real conscious being to observe it.

          1. I didn’t say we weren’t conscious, I said we were p-zombies. A p-zombie thinks it is conscious, and nothing either empirically or in terms of self-reportage distinguishes a p-zombie from a p-angel, so why think that we are conscious at all in the special qualitative sense p-angels require? It’s enough that we do and experience everything as if we had those properties, not that they exist. We only think there are subjective properties, but in fact there aren’t.

             There are subjective viewpoints, and of course we all have our own personal histories and systems, but there’s nothing ontologically mysterious about that.

             Since I don’t accept this mysterious form of consciousness, what we are is conscious in a physical manner only. That can be simulated in a physical system; but physical differences will always make a difference, so the only way to fully simulate me is to make an atom-for-atom copy of me. All else is abstraction, to an arbitrary degree of precision.

             Or can you define consciousness in a manner that makes it accessible to investigation?

13. “I t’s enough that we do and experience everything as if we had those properties, not that they exist. We only think there are subjective properties, but in fact there aren’t. ”

    Well, that’s we differ. There is no “as if” for me. You are either subjectively aware, or you are not. I observe that I am subjectively aware of many things. I should not be able to do that if current physics is all there is (objective properties only). There is no room for that observation in physics. That’s all I need to say that physicalism must be incomplete.

    You may say that you cannot observe my subjectivity and that this doesn’t fit within your physical worldview, but I can respond by suggesting that you may be overly-reductionist in attempting explain everything in terms of the objective world.

14. I’m late to the party, but to choose a less computational metaphor, you are suggesting a situation where a creator acts something like an artist who performs some large number of randomized trials, then displays the favorite as a photo in a gallery?

    1. In fact, I recall seeing a painting that was executed by arranging shuffled pieces of colored paper, which was then carefully reproduced in paint (if only I could remember who it was by).

       1. Sounds about right. God’s paint is instantiation.

Comments are closed.