Blog Coordinator

« Moral Psychology (Vol. 4): Free Will and Moral Responsibility | Main | CFP: The Moral Domain »



Feed You can follow this conversation by subscribing to the comment feed for this post.

In my view, there are various problems of free will. Perhaps the most prominent one is this: we seem to have free will, and yet it seems incompatible with both determinism and with indeterminism (and thus seems impossible). What's worse, there are arguments that purport to demonstrate these incompatibilities. If the P2P model "explains, and promises a new resolution to, the problem of free will", I imagine it should tell us how these arguments go wrong. Perhaps it can tell us which premises are false. Do you think it does this? If so, which premises of, say, the Mind and Consequence Arguments, does the P2P model recommend we reject?

Hi Andrew: Thanks for your comment!

Frankly, I think the argument that free will is impossible is a non sequitur. Look, you don't think it is impossible for electrons to orbit the nucleus of an atom, do you? Of course not. They orbit the nucleus of atoms because it is a *primitive* law of nature that they do it in accordance with certain equations. But now once you admit that there are *some* primitives in nature -- and of there are -- there is nothing, in principle, that rules out genuine libertarian free will (i.e. thoughts and perceptions as uncaused causes that enable us to will our actions on the basis of *nothing* beyond them) as a coherent possibility.

The idea of consciousness being an uncaused cause may not seem to "make sense" -- but notice, the fact that electrons just do orbit nuclei (without any further explanation) "makes no sense" either. But this just a bad way of putting it. They *both* make sense once we realize that certain features of reality must be primitives. Primitives are, by definition, things that lack any further explanation. They are, well, primitive. Asking for an explanation for them is wrongheaded in just the same way that it is wrongheaded to ask for an explanation of an axiom in, say, logic or mathematics.

Anyway, as to the Mind and Consequence arguments, it should be pretty clear which premises I mean to deny. If the kind of libertarian free will my model makes room for exists, then our actions as the result of consciousness itself *being* a primitive uncaused cause not determined by any laws of nature, including indeterministic laws.

One final point from my 2013 paper is worth reiterating here. In the paper, I admit there are two possibilities if the P2P model is true:

1. Our actions are caused by our consciousness being this kind of uncaused cause in a higher reference frame (viz. the frame of reference of an observer outside the simulation).

2. Our actions in that higher reference frame are *determined* by laws of nature in that reference frame.

I admit in the paper that we may not be able to know for sure which of these is actually the case--since, after all, we are trapped within this reference frame (if we are in a simulation, we cannot directly observe what is going on at the higher level outside of it).

In which case, in the end, the P2P Model leaves room for libertarian free will but also leaves open the possibility that the P2P simulation is deterministic "all the way down."

Even so -- even though I do not purport to "prove" that we have libertarian free will -- I think it is a real (possible) victory to show how libertarianism is consistent with our evidence and *could* be a part of a unified model that explains fundamental features of our world not explained by other physical or metaphysical theories.

Finally, I am currently working on some arguments against possibility (2), and in favor of the notion that consciousness *must* be an uncaused cause, not just the result of deterministic processes in a higher reference frame. But that must wait for another day. :)

Hi Marcus,

Just to pick up on Andrew's thread...

There probably are some arguments for the impossibility of free will that are non sequiturs, but the one Andrew suggested isn't. It goes (roughly) like this: Either determinism is true or indeterminism is true, but determinism and indeterminism are both incompatible with free will. So, free will is impossible. From your response, it sounds like you deny the claim that indeterminism is incompatible with free will. But it's not clear to me how your model gives us reason to think that this claim is false (or that any particular argument for this claim is unsound).

What your model gives us is a way to be libertarians while admitting that our best neuroscientific theories may well give us good reason to believe that the brain operates deterministically. (It wouldn't *really* operate deterministically, but given your P2P model, we'd expect it to look as though it does.)

It's a further task, however, to argue that any action caused by an uncaused agent cause would be free. (This would involve, among other things, responding to the Mind argument.)

Hi Neal: Thanks for your comment! Fair enough - I agree that I should address those issues in future work. But, for now, a couple of quick comments.

First, in my view, there are already many independent reasons to doubt the mind argument(see e.g. Shabo 2013, Graham 2010, Franklin 2011, and others).

Second, in my own view, the mind argument is clearly question-begging. In making the mind argument, van Inwagen assumes that if the mental states (e.g. desires) that cause behavior are uncaused, they *just occur* -- and hence, are not under the agent's control. But this is precisely what I, as a libertarian, want to deny. I want to say that our choices do not "just occur", but that, in a completely primitive way, we *choose* them. Now, of course, you can doubt this kind of libertarian primitivism on other grounds (say, an Occam's Razor argument for physicalism, where physicalism leaves no room for libertarian free will). And that's fine. I'll deny *that* argument on other grounds (namely, that the P2P model explains things about QM that physicalism doesn't). Anyway, my point is just that I think the mind argument is question-begging. The argument assumes as a premise, without argument, that primitive libertarian free choice does not exist.

Concerning Shabo's new "assimilation argument", where he argues that libertarians must *explain* how free will can have the power to settle which one many possible futures occur, there are two issues.

First, I think it too is a non sequitur. In his view, the libertarian must *explain* how free choice can settle which of two possible futures occur. This, again, is what I think libertarians should deny. If free will is primitive (and, I want to say, there are good reasons for taking it to be), no explanation is called for.

Second, Shabo thinks that a libertarian must explain how some indeterminacies involve free libertarian choice but others do not. But the P2P hypothesis has the resources for doing just this! Allow me to explain.

In a P2P video game, there are two distinct levels of indeterminacy. The first, background indeterminacy is comprised by the P2P network as a whole. So, for example, a rock in a P2P simulation does not have a determinate location in space time because "its location" is just a superposition of the parallel representations on the different machines. Call this indeterminacy the P2P wave function. In a P2P a second indeterminacy on top of this one occurs at level of the characters you control (their behavior cannot be fully predicted within the sim because of the choices made outside of the sim. Call this the Choice wave function. On the P2P model, the first kind of indeterminacy (indeterminacy in nature) is only slightly affected by free choice (since our choices do affect the entire network). However, the second class of indeterminacies -- each person's "Choice wave function" -- is predominantly the result of free choice ex nihilo.

In this way, the P2P model provides a clear, unified picture of why some indeterminacies involve libertarian free will and others do not.

In any case, thanks again for your comment. Good stuff to think about!

Marcus, I don't *want* your version of free will. It comes out of nowhere. Its decisions don't relate to the (real, empirical, this-here-possible-world) me I know and love. My character, my beliefs, my values: these aren't pulling its strings. It's ex nihilo - and I wish it to go back there.

Paul: To quote the great Mick Jagger, you can't always get what you want, but if you try sometime, you just might get what you need! ;)

On a more serious note, the P2P Model does not deny the "real, empirical, this-here-possible-world" you that you know and love. It just simply holds that there is more to you than the forces and particles of contemporary physics. The model is dualistic, holding that there are two *aspects* of you: a physical brain/body interacting with a non-physical conscious will. It also doesn't hold that free will comes out of nowhere. It holds that it comes (ex nihilo) from your conscious *will* interacting with the physical information/character/dispositions encoded in your brain (so, it *doesn't* deny these features of you).

Is all this really so bad? (I don't think so!) In fact, isn't this how we experience ourselves? You *have* beliefs, desires, and character traits -- but it is up to *you* (your conscious will) how you act!

Marcus, I think it should be up to me to define what counts as part of me. Wielding Torek's Razor, I shave off your dualistic extra and reject any claim to its being me. (It wasn't making that claim anyway. All claims of interest emanate from empirical this-here-possible-world mouths.) If such an alien being exists and disturbs my trajectory through life, I regard that as a violation.

Paul: with respect, it is not up to us to define what is true any more than it is up to Young Earth Creationists to define, by fiat, that we did not evolve from primates. By a similar token, critics of General Relativity once held that the theory "must be false" because, "by definition", space and time must be absolute. We all know how well that definition fared (not well!). Indeed, philosophical definitions have a rather bad habit in history of being refuted by empirical science. (Note: I do not mean to compare myself to these two fellows [Darwin & Einstein]. They are just a couple of particularly famous examples!).

For better or worse, what is true is *true*, whether or not like we like it, and regardless of whether it fits with how we like to define things. The P2P theory will ultimately stand or fall on its merits.

1. I really liked the part about quantum physics as a result of the intersubjective nature of reality
2. I do not understand (mea culpa) why you say that consciousness is the "hardware"
3. When clocks were the last fashionable invention, people used clock-metaphors to explain things they could not understand. They also used the first computing machines (like the one invented by Pascal) as a model to understand the way a mind works, and now everyone compares the brain to a computer. Is not it the case that we are conditioned by what we know? (Cf this post: )
In this sense it is hard for me to think that the world is *really* (as you seems to imply in your comments) a P2P simulation rather than thinking that the existence of P2P simulations made us think at the comparison.

Hi Elisa: Thanks for your comment! Here are a few thoughts in reply.

On (1): thanks! :)

On (2): I'm understanding software and hardware here in roughly the way computer scientists do. On the P2P Model, physical reality is just a string of quantitative information (1's and 0's) -- in other words, software. Consciousness, in contrast, is the *medium* that reads the software (which is what computer hardware does). This is only very rough, but I'll have to leave a full discussion for elsewhere (I hope to write a paper on this soon!).

On (3): The P2P model is not merely a metaphor. Metaphors are *metaphors* precisely because they map onto the things they are metaphorical descriptions of only very imperfectly (viz. the brain only functions like a clock in a few broad respects; in most respects, it is not like a clock [the brain does not literally have gears, etc.]). If the P2P Model is correct, reality is *literally* functionally identical to a P2P simulation in almost every respect. The reason I say this is that the P2P Model holds that a P2P structure is *literally* necessary to explain a number of observed features of the quantum world.

That being said, the P2P Model involves at least one -- and possibly two -- features that are distinct from the kinds of P2P simulations we have created. The first difference is that on the P2P Model, consciousness is a fundamental feature of the structure (which, obviously, is not true of P2P simulations we have created; in P2P simulations we have created, the measurement device is just another structure, viz. the computer console/processor).

Another possible difference that I am thinking about concerns the analog/digital divide. In its present form -- the form worked out in my 2013 paper -- the P2P Hypothesis holds that physical reality is fundamentally digital/quantized. Future observations may reveal, to the contrary, that our reality is fundamentally analog. This wouldn't rule out the P2P Hypothesis per se, but it would rule out a digital version of the model (but again, we will have to wait and see. The jury is still out on whether our reality is digital or analog).

Anyway, thanks again for your comment. I hope this reply helps! :)

Marcus, it's not about whether the metaphysics you've laid out is *true*. Suppose that all the things you posit do indeed exist. There remain a set of questions of interpretation, especially with regard to the identity of persons. And your interpretation is wrong - or at least, not mandatory.

Even if, in addition to my brain activities in this universe that I call my thoughts and desires and decisions and plans etc., there is a dualistic entity in the multiverse that has influenced the course of my life in this universe, it remains a further claim to posit that the dualistic entity is me. I reject that claim.

Paul: Interpretations (in science, at least) ultimately give way to facts.

Example #1: as Steven Hawking point out in his recent book "The Grand Design" (2010), Ptolemy's geocentric model of heavenly orbits was *never* truly "disproven" by Copernicus' heliocentric model. One can always interpret heavenly phenomena -- even today -- in a way that conforms to Ptolemy's model (e.g., by stipulating that planetary epicyles fit observation). But nobody adopts this interpretation. Why? Because it is hopelessly ad hoc. Copernicus' geocentric model gives a simpler, more unified account of celestial phenomena -- and it is precisely because of this that everyone accepts the Copernican interpretation over the Ptolemaic one.

Example #2: by a similar token, modern biology has never fully "disproven" the animistic theory that there is a special life force that gives life to all and only living things. There is always a way of interpreting observed phenomena such that, indeed, there is such a special force (one could, for instance, just stipulate that anything that does not reproduce organically lacks a "life force"). But, of course, nobody adopts this interpretation. Why? Because, again, it is redundant and contrived. We can explain everything in question with a far *simpler* interpretation: namely, biochemistry *without* a "life force." Which is why everyone accepts that simpler interpretation.

With this in mind, suppose the P2P Model is right, and there really is (A) an entity that is coincident with *all* of your conscious experiences and choices (the visual experiences you are having right now, as you are reading this), and (B) this entity is *necessary* for explaining quantum phenomena, etc. Here are two possible interpretations of the nature of this entity:

(1) You are distinct from it -- even though it experiences everything you do, wills everything you do, etc. [Your favored interpretation]

(2) You are identical to it precisely because it experiences everything you do, wills everything you do, etc. [My interpretation]

Which interpretation provides the simpler, less redundant, more unified interpretation of the phenomena we observe? Which is more like Copernican theory and less like Ptolemaic theory? Which is more like biochemistry and less like animism?

The answer to all of these questions is simple: interpretation (2).

Interpretation (1) -- the interpretation you seem to favor -- is contrived, redundant, etc., in all the ways that Ptolemaic astronomy and animism are contrived. Such attempts to "define away" new phenomena, again, haven't fared well in history. And for good reason. Contrived interpretations are contrived. Non-contrived interpretations simply respond, in the most minimal and non-contrived way, to observed facts.


Well sure, if you posit (A) an entity that is coincident with *all* your experiences and choices, then that's you. But (A) is a new thesis, relative to your summary above. (It is there, though, in your full paper, I see now. Shoulda read it sooner.)

Worse, (A) is not necessary or even helpful in explaining any quantum phenomena, qualia, or anything else. The whole physical universe could be "read" (a la a DVD drive reading information) by a single "reader" switching in an uncaused way between multiple tracks on the storage medium. Or there could be one "reader" for every quantum-mechanical collapse event, and so many many readers per human being. Since information propagates across the physical world, there is no mystery about how we experience a unified coherent timeline despite multiple non-physical readers being involved. Or on the single-reader-of-the-universe option, since information propagates very weakly from one brain to another compared to intra-cranial propagation, there is again no mystery as to how most of us forget that we are all One. (Hat tip to some philosopher/theologians in a broadly Hindu tradition.)

Of course, neither or these options, nor the myriad others that could be conceived in between, would preserve a further fact of personal identity. But no further fact is in evidence, so it doesn't need explaining (nor, in my view, is it even to be wished for).

One also wonders why, in your preferred option, these nonphysical readers just happen to latch onto the physical universe just at the moments and locales when people are born (give or take a few months). Or if not, what was their subjective experience before then?

Hello Marcus,

I am not sure I can understand the 100% of what you say in this post. Still, here’s a couple of comments from somebody who is hostile to metaphysics, and even more to metaphysics done along the lines of the Matrix.
First point: you seem to consider the fact of showing a possible implementation of a form of libertarian free will as a point of strength of your theory. But, to put it a bit brutally: why should I care for libertarian free will if I came to be convinced that I live in a simulation – that is to say, that reality as I am used to think of it is essentially illusory? I mean, the good in libertarian free will (for those who think there is some good in it) is primarily that it vindicates our intuitions about self-determination. But your position vindicates those intuitions only at the price of screwing intuitions about reality that are even more basic. Don’t you think that the cure is, as it were, worse than the disease?
Second point: you may want to say that your theory is based on scientific evidence and that as such it deserves attention irrespective of vindicating or failing to vindicate our intuitions. But I think that the substantive metaphysical conclusions you reach can only be reached by coupling scientific evidence with substantive (and controversial) philosophical assumptions. To get started, you must be committed to the idea of a ‘fundamental’ layer of reality; this is already a substantive philosophical assumption (I think you also make substantive and controversial assumptions about the nature of consciousness, but I won’t enter into that, for I am not sure). Once you have this on the table, it seems to me you must either admit that such fundamental reality is something we cannot know (something akin to the Kantian noumenon), or that we can in some way know it. You seem committed to the idea that we can know it and that we can theorize about it by employing a certain vocabulary (of quantum physics? or computer science?). I see no reason to accept this last step – no reason to think that there is one single vocabulary that connects with a privileged layer of reality. It is a trivial fact about mankind that there are different vocabularies we employ, different language games that we play in the interaction with the world. There are certainly many interesting and difficult issues about how the different areas of our linguo-conceptual practices interact and influence each other, how we revise them and abandon some parts of them. But I see no reason to think that only one vocabulary connects with ‘fundamental’ Reality.

Marcus, do you have a list of concrete predictions, which, if they do not turn out as you predict, could falsify your matrix-like theory? Note that it is not enough to say finding x is consistent with my theory or that oddity y from quantum theory can be understood in light of my theory.

Paul: There is no way to definitively rule out those alternative possibilities, any more than we can rule out alternative explanations of any event (see e.g. the general problem of underdetermination of theory by evidence in philosophy of science). There are always, strictly speaking, an *infinite* number of theories consistent with our evidence at any given time. So, indeed, it is possible that "we are all One", and that there is only one reading mechanism comprising all of our (apparently distinct) consciousnesses.

I do not -- and cannot -- purport to disprove that possibility. What I can do is provide a general model (the P2P Hypothesis) that explains fundamental features of the quantum world that are not yet otherwise explained, as well the problem of free will and the mind-body problem, and leave it to be determined whether that is the *best* model. That is all I purport to have done. There may indeed be other possible models. All I am claiming is that this one is implied by several compelling lines of evidence, has a number of theoretical virtues (explanatory power, etc.), and is therefore a viable model worth further investigation.

This is also my answer to your final question: about why consciousness becomes hooked to our bodies when we are born. That is indeed a mystery with many possible different explanations. But notice: *every* theory of our world has to posit mysteries, including physicalism (why is there a physical world at all?). Mysteries are disturbing, but if I am right, the P2P model, while admittedly creating some new mysteries, *explains* some mysteries not yet explained (quantum phenomena). In any case, I do hope to tackle the P2P model's mysteries in future work.

Stefano: Thanks for your great comment!

On your first point: I am simply not interested in intuitions about reality. I am interested in *reality*. Allow me to explain by once again using a few famous cases (again, I do not purport to compare myself to these fellows. They are just famous examples of what I'm interested in).

Kepler's elliptical model of planetary orbit was accused of screwing with the neo-Aristotelian intuition that heavenly objects must move in circles because (1) the heavens, being heavens, must be perfect, and (2) circles are perfect. Well, so much the worse for Aristotelian intuitions. They are false.

Copernicus' heliocentric model was accused of screwing with the intuition that we must be at the center of the universe. Well, so much the worse for the intuition. It's false.

Darwin's theory of the origin of species was accused of screwing with the intuition that we are beings of a different order than non-human animals. Again, so much the worse for the intuition.

The theories of relativity were accused of screwing with the intuition that space and time "obviously" must be absolute.

Finally, Einstein accused quantum mechanics of screwing with the intuitions that (A) God does not play dice, and (B) action cannot operate at a distance (viz. entanglement). Again, so much the worse for Einstein's intuitions.

Intuitions about reality, as far as I am concerned, have little value. They have been disproven by scientific theory and observation again and again throughout history. I am interested in theory and observation, not intuition.

On your second point: I entirely agree that the model is reached only by combining scientific observation/theory with controversial philosophical positions. But a few things to note here.

First, I am myself prepared to argue for many of these controversial positions individually. (Example: I am presently working on a paper arguing that *any* reality must be dualistic, comprised of Kantian noumena and phenomena).

Second, and perhaps more to the point, all I am doing at the present time is saying, "Look, here is a model that *explains* things about reality not yet explained. The explanation may involve controversial philosophical positions, but for all that, the model does some very important things that no other model -- no model in physics, and no model in metaphysics -- does. Regardless of whether its individual philosophical assumptions are controversial, this is an important result. Quantum mechanics has long seemed -- physically and metaphyscially -- a complete mystery. Now, at least, there is at least one metaphysical model -- one admittedly based on controversial assumptions -- that promises to provide a unified explanation of why the world has quantum mechanical features." In other words, I am more than happy to admit that the model is predicated on lots of philosophically controversial stuff. At the same time, I want to say that *if* that controversial stuff explains features about our reality not yet explained in other terms, this is a new reason to take that controversial stuff seriously.

Hope this helps. Thanks again for the great comment!

Hi Peter: Thanks for your comment!

Yes, if you read pp. 39-42 of my (2013) paper, you will see that the digital version of the P2P Hypothesis I propose makes three specific predictions unique to it as a theory (and so which, if all three were confirmed, would provide unique confirmation of the P2P Model). The first prediction is related to the holographic principle. The second is related to quantum chromodynamics (i.e. observations of lattices/edges to space-time). The final one is broad prediction related to quantum mechanics (i.e. violation of the normal wave-function) at at a neural level.

As I mention in the post above, the prediction about quantum mechanics at a neural level seems to have some tentative support. Observations about the first two predictions are not yet complete. They could confirm or falsify the theory. We will have to wait and see.

I am not yet sure whether an analog version of the P2P Hypothesis is empirically testable. I need to do further research on the analog/digital distinction to see whether an analog version would have specific predictions, and what those predictions might be.

Hope this helps - thanks again for your comment!

The comments to this entry are closed.

Books about Agency

3QD Prize 2014: Marcus Arvan