A New Theory of Free Will

Just a quick note that I recently published an article in The Philosophical Forum , “A New Theory of Free Will“, that may be of interest to readers (a free PDF of the penultimate draft is available here).  Here’s the abstract: 

This paper shows that several live philosophical and scientific hypotheses – including the holographic principle and multiverse theory in quantum physics, and eternalism and mind-body dualism in philosophy – jointly imply an audacious new theory of free will. This new theory, “Libertarian Compatibilism”, holds that the physical world is an eternally existing array of two-dimensional information – a vast number of possible pasts, presents, and futures – and the mind a nonphysical entity or set of properties that “read” that physical information off to subjective conscious awareness (in much the same way that a song written on an ordinary compact-disc is only played when read by an outside medium, i.e. a CD-player). According to this theory, every possible physical “timeline” in the multiverse may be fully physically deterministic or physically-causally closed but each person’s consciousness still entirely free to choose, ex nihilo, outside of the physical order, which physically-closed timeline is experienced by conscious observers. Although Libertarian Compatibilism is admittedly fantastic, I show that it not only follows from several live scientific and philosophical hypotheses, I also show that it (1) is a far more explanatorily powerful model of quantum mechanics than more traditional interpretations (e.g. the Copenhagen, Everett, and Bohmian interpretations), (2) makes determinate, testable empirical predictions in quantum theory, and finally, (3) predicts and explains the very existence of a number of philosophical debates and positions in the philosophy of mind, time, personal identity, and free will. First, I show that whereas traditional interpretations of quantum mechanics are all philosophically problematic and roughly as ontologically “extravagant” as Libertarian Compatibilism – in that they all posit “unseen” processes – Libertarian Compatibilism is nearly identical in structure to the only working simulation that human beings have ever constructed capable of reproducing (and so explaining) every general feature of quantum mechanics we perceive: namely, massive-multiplayer-online-roleplaying videogames (or MMORPGs). Although I am not the first to suggest that our world is akin to a computer simulation, I show that existing MMORPGs (online simulations we have already created) actually reproduce every general feature of quantum mechanics within their simulated-world reference-frames. Second, I show that existing MMORPGs also replicate (and so explain) many philosophical problems we face in the philosophy of mind, time, personal identity, and free will – all while conforming to the Libertarian Compatibilist model of reality. I conclude, as such, that as fantastic and metaphysically extravagant as Libertarian Compatibilism may initially seem, it may well be true. It explains a number of features of our reality that no other physical or metaphysical theory does.

Cross-posted at Experimental Philosophy



  1. Marcus Arvan

    Hi Quentin: Thank you for your very good question. It differs from the many-minds interpretation in a number of important respects.

    First respect: in the many-minds interpretation, there is no definite macro-realm of physical reality or information (the infinity of minds is *all* there is to quantum decoherence). In contrast, on Libertarian Compatibilism (my view), there is a sense in which the macro-physical world is definite: it is an eternally array of 2-dimensional information being *measured* by minds in a higher frame-of-reference.

    I think there are a number of reasons to think my picture is more likely to be true. Among other things, my view — unlike many-minds — is structurally identical to simulations we’ve already created that reproduce quantum phenomena and “save the appearances” of our world (i.e. peer-to-peer networked MMORPGs). In contrast, we have no analogous “working model” of many-minds — and my theory is far more ontologically parsimonious to boot.

    Second respect: the many-minds interpretation holds that an infinity of minds split into *every* possible quantum “path” (eigenstate) — in essence populating a multiverse with an infinity of minds across an infinity of possible pasts, presents, and futures. As I argue and imply in the paper, this interpretation violates Occam’s Razor in important ways that my theory does not (my view does not populate an entire multiverse full of an infinity of minds — but only *our* minds, which trace out a *single* path through the multiverse, collapsing *possible* “paths” down to a single, intersubjectively experienced reality.

    Third: in many-minds, each “physical” observer has a postulated infinity of continuous minds *within* a single world (much as in 4-dimensional metaphysics different “wormstages” can imply there are two, three, etc. *person-stages* overlapping in a single body at a time). On many minds, at every quantum instant, this infinity of minds “branches out” into infinite subsets. In contrast, on my view, each person has only *one* mind which, through our choices, traverses only *one* path through the physical multiverse forward in time. Again, I think my view is far preferable (for, I think, fairly clear reasons).

    There are probably a number of other relevant differences I’m not thinking of right now. But those are some. Hope they clear things up a bit.

  2. Thank you for your response, I’ve got a better understanding of what is implied by your theory now. However I have an other question.
    When I measure, say, the spin of an electron: would you say that I choose which value it will take? I don’t have the feeling that such external events depend on my will, plus then you have the problem of other observers (this would be an interpretation close to that of Wigner which he finally abandoned). The other option is that the electron chooses its spin value when measured.

    Any electron measurement can lead to branching in the multiverse, and the number of branching (if countable) is actually far more important than those which could be subjected to human choices. You could assume that random collapses occur in addition to human choices, but I am not sure that you are willing to accept such hybrid theories (you’d inherit their drawbacks).

    In brief, your interpretation seems to entail panpsychism. Am I right?

  3. Marcus Arvan

    Hi question: thanks for another very good question. 

    I would like to have addressed this issue much more clearly in the paper, but space constraints counted against it (I feel lucky enough to have a journal even consider a 79-page paper!). Anyway, I thank you for asking, so that I can address it now.
    My theory, as I understand it, comprises a kind of *hybrid* panpsychism — but one that (contrary to your comment) avoids as far as possible the so-called “drawbacks” of either type of theory.  Allow me to explain.
    You will recall that, on my theory (Libertarian Compatibilism), our world is almost perfectly structurally identical to a peer-to-peer networked massive-multiplayer-roleplaying-game (MMORPG). Notice, next, that existing MMOPRGs have *two* distinct elements that give rise to “quantum uncertainties” and measurement problems with the simulated-world reference-frame. 
    First, the peer-to-peer network connection *itself* gives rise to random fluctuations in “measurements” observers take of objects in the simulation.  This is because each machine on a peer-to-peer network independently reads the 2-dimensional information that comprises “physical reality.” Because the machines are independent but networked together, and there is no dedicated server, there are necessarily temporal/informational gaps between how different machines read the information, giving rise to observations in the simulation of random “quantum” effects.
    But this is not the only source of “quantum uncertainty” in peer-to-peer networks. The choices of *users* in a higher reference-frame — in the case of actual MMOPRGs, ordinary videogamers — give rise to a *secondary* set of measurement uncertainties that piggyback on the first. Each gamer’s choices introduces a *non*-random set of “measurement/quantum” fluctuations within the simulation that interact with the first-order, random (peer-to-peer generated) fluctuations — but only (I think) in a manner that supervenes on the body of the “avatar” in the simulation (which is why I say in the paper that the theory predicts novel quantum/measurement effects in human brains).
    Finally, I should probably say why I think my account doesn’t have — or at least greatly minimizes — the supposed drawbacks of both panpsychism and hybrid theories. The basic story I want to tell is this: unlike traditional forms of panpsychism, etc. — which seem hopelessly mysterious and non-empirical to a lot of people — the theory I am offering provides (A) a (mostly) *non*-mysterious, empirically (and model-theoretically) respectable and compelling *explanation* not only of ( quantum mechanics (i.e. we now see that quantum effects naturally emerge from any peer-to-peer simulation), but also of (C) the “appearances” of our world more generally, including the very existence of philosophical problems ranging from the mind-body problem itself to problems in the philosophy of time, personality identity, and of course, free will.  In other words, I take myself to be giving some real empirical and theoretical respectability — not to mention, a new kind of intelligibility! — to a kind of “hybrid panpsychism.”
  4. Marcus Arvan

    Note: that should have been an “A” in my previous comment, not a smilely face. Don’t know why the comment system automatically transformed it. Must be a glitch in the matrix.

  5. Marcus Arvan

    I also just noticed your name auto-corrected to “question”! Oops.  Anyway, it just occurred to me that, in case my post wasn’t crystal clear on this, the “hybrid” aspect of the theory as follows:

    (1) Ordinary physical objects and properties will be governed by a single, random “quantum-wave function” that emerges from a peer-to-peer network connection (a first-order quantum uncertainty generated by the network), but
    (2) Non-random fluctuations should be observed “on top” of the ordinary fluctuations within the bodies/brains of animated beings with free will (i.e. us) — a second order, non-random uncertainty generated by choices in a higher frame of reference.
    (Note: these are in fact properties of actual MMOPRGs).
    One final note: I take it that another very interesting implication of the theory is that it entails a way to empirically test whether or not non-human animals in fact have conscious minds.  As anyone who has played an MMORPG game (such as the SIMS or Halo) would know, there are “animals” in those simulated realities as well. In their case, the animals do *not* have any “minds” above and beyond the programming coded into the machine, and their behavior is non-random and predictable as a result (setting aside the first-order uncertainty generated by the P2P network itself).  Accordingly, my theory suggests that if we (A) discover only a normal wave-function collapse supervenes on normal animal brains, but ( a non-random higher-order randomness supervenes on human bodies/brains, we would have strong reasons to think that animals do not have conscious minds (for again, on the theory, conscious minds exist in a higher-reference frame “on the outside” of the simulation, which we would have definitive evidence against in the case of animals…if that’s how our examination of their brains bore things out).  Of course, on the flip-side, evidence in the other direction might strongly suggest that we *aren’t* fundamentally different than animals. In either case, I take it that the fact that the theory makes these predictions is pretty cool.
  6. Actually I am quite dubious (to say the least) that mmorpg somehow “reproduce” quantum mechanics. My feeling is that they merely reproduce a classicaly branching universe with epistemic indeterminacy, which is totally different from quantum indeterminacy. If the position of a leaf is uncertain in the game, you can still be sure that it has a definite position although you don’t know it. QM cannot be interpreted interpreted that way, hence the measurement problem (e.g. would you obtain an interference pattern involving the two possible position of a leaf if you decided to measure its speed instead? Or could you reproduce an EPR experiment in a mmorpg which violate Bell’s inequality?)

    Regarding your hybrid view, how do you justify that two different sources of randomness take place in exactly the same framework, and how do you differenciate them? Then how do you explain that specific material organisations (living things) implement one kind of randomness and not the other?

  7. Marcus Arvan

    Hi again, Quentin: I appreciate your skepticism. However, I think it is misplaced. Allow me to explain why.

    When you say your feeling is that MMORPGs “merely reproduce a classically branching university with epistemic indeterminacy, which is totally different from quantum indeterminacy”, you are referring to what’s known as a dedicated server environment. In MMORPGs with a designated server, there is, indeed, a machine that represents a single “objective” position of objects and properties in the simulated-world environment. Whatever measurement problems users on the network have are, in this case, merely epistemic. But this is not how peer-to-peer (P2P) networks work. In P2P networks, no machine is “authoritative.” Rather, the simulated world itself is nothing more than many independent machines interacting. There is no “objective” placement of (say) a leaf within the game. Instead, all there is are the many different machines’ measurements, and — due to the nature of the peer-to-peer environment (information lag, etc.) — different machines (and the same machine, at different instants) necessarily take different measurements which differ in random ways, according to something very much resembling the quantum wave function we observe in our world.
    This brings me to my second point, which relates to your skepticism about whether MMORPGs reproduce would reproduce (say) an EPR experiment violating Bell’s inequality. One thing that I had to gloss over in the paper — again, due to its length — is the difference between reproducing every single general (or fundamental) feature of quantum mechanics and reproducing every single detail of quantum mechanics.  Allow me to explain the difference, and why it is important. I take the fundamental features of quantum mechanics — the features of our world that have seemed so fundamentally baffling — to be (very roughly, I realize I simplify a lot here) (A) fundamental indeterminacy/randomness in micro-objects and properties, ( measurement problems, and (C) “action-at-a-distance” (quantum tunnelling, entanglement) that cannot be accounted for in terms of any kind of classical kinematics.  My claim — and I think the evidence for it is solid — is that MMORPGs reproduce all of these things in their simulated-world reference frames.
    The theory does not claim that existing MMORPGs reproduce all of the specifics of quantum mechanics — EPR experiments, the exact specification of the quantum wave function — and for very good reason.  Every MMOPRG not only has its own in-game “physics”; it also has its own “netcoding”; not to mention the medium/manner of information transfer in a higher reference-frame (viz. information transfer via fiberoptic cables over the internet). Due to these differences, the specific “wave function” in, say, SIMS or Halo may be quite different from another.  And indeed, this is the case.  I take it, though, that this is not a “problem” for the theory.  It would be very strange for a theory to say that any world whatsoever must have exactly the same quantum-effects, given again, that different simulations can have very different informational laws and net-coding.  
    The important point is that MMORPGs reproduce the most fundamental, baffling features of quantum reality, and that all of the specifics can in principle be accounted for in terms of information transfer, net-coding, etc. — which, if I haven’t proven, I take to be a very promising hypothesis.
  8. Marcus Arvan

    Sorry — forgot to address your question about animals.  I’m not sure what you mean when you ask me how I would “justify” taking the two sources of randomness to occur.  The paper doesn’t “justify” this claim so much as it makes it as an empirical prediction. 

    Anyway, I think what you might have meant was: “if we observed these two fundamentally different types of randomness in humans and animals, what would justify ascribing consciousness to humans but not to animals?” If this was the question, my answer would be (roughly) as follows: if we observed those two types of randomness in the world, we would have evidence that our world is just like MMORPGs in that respect — and, since we know the “animals” in MMORPGs (however cute and furry and cuddly they look — and they often do in MMORPGs!) are “just programs”, we would thereby have reasons to think the animals in our reality are analogous (whereas we would have evidence that our minds are something special emanating in from a higher reference-frame).
    I think you also might have meant to be asking: “How could your theory make sense of such a fundamental difference between animals and humans — say, in naturalistic terms (viz. why would humans evolve to have a special property of existing in a higher-frame of reference but not animals)?”  If this was the question, my answer would be roughly this: I am not a God-fearing man, but if that’s what we found I might seriously consider changing my ways! Of course I jest a bit here, I’m sort of serious: I would begin to suspect more and more — and I think I would have strongly suggestive evidence — that we are truly in just a big videogame that someone (God? evil machines?) created for their (and our?) amusement.
  9. My point was that external events can cause branching (whether a leaf falls or not is a different world). As far as I can tell, you propose a collapse theory for them (introduction of noise). However then your libertarian theory becomes superfluous, since collapse theories already solve the measurement problem. Besides, you seem to believe that the libertarian theory would be at odd with current empirical data (because the will is non random). So your theory is both superfluous and speculative, which I think is a serious problem.

    About whether mmorpg reproduce QM or not (discussion below):
    EPR is central to QM and Bell’s theorem is a fundamental meta-theoretical theorem. No measurement problem without them.
    If a leaf can be observed in different positions by different observers, this is *not* QM. If a leaf is in a superposition of positions when *not* observed, then this is QM. Per my understanding, mmorpg reproduce the first alternative?

  10. Marcus Arvan

    Hi Quentin: Thanks for pressing these issues. Here are my thoughts in reply:

    (1) It’s actually true, in a P2P MMORPG, that a leaf *does* exist in a superposition when not observed. Insofar as the simulated reality *just is* the system-link of machines, and the many machines represent different values for the leaf, the network as a *whole* represents the leaf in a superposition (it’s just the case that any *particular* machine, upon taking a measurement, will “collapse” its position to a single measured value).
    (2) The libertarian theory I’m offering isn’t superfluous because it both (a) predicts *two* distinct forms of collapse, and (b) provides (what I take to be) a better/deeper explanation of each than existing explanations of quantum mechanics.
    (3) Sorry, but I wasn’t thinking carefully about EPR the first time you mentioned it. However, I think it’s fairly easy to demonstrate that the phenomena do occur in P2P MMORPGs.  First, there are clear cases of non-locally entangled states within MMORPGs (directly analogous to quantum entanglement in our world). I’ve experienced them before when playing Halo. If one steps on a particular patch of ground, another patch of ground elsewhere may instantaneously shift to a different state (with no information transfer observable to individuals in the simulated-world reference frame). This is comparable to the EPR issue of entangled particles. Next, the relevant point about the EPR case (in our world) is that if one observes two entangled particles — A and B — if one measures A as having z+ spin, this will instantaneously cause uncertainty in B’s spin (50% spin z+, 50% of spin z-). But this too is borne out in P2P MMORPGs. Although the two states are entangled, the entire *system* (as I mentioned in the leaf case above) has “noise” in it, such that (even though A and B are entangled) the properties of B are in superposition throughout the system, in which the spin of B cannot be predicted with certainty in the MMORPG — until, that is, a measurement is taken, at which point the measurement of B will cause the same uncertainty with respect to A. 

Comments are closed.