A pragmatic approach to free will

PinocchioEarlier this year I engaged in a discussion on doxastic voluntarism where a commenter stated that

“I use freewill to mean we can choose to change the physical sequence of events in our brains. … If we don’t have genuine freewill, then we can’t choose”,

to which I responded with

“Regardless of where one stands on free will, we agree that we engage in something called ‘choosing’. This phenomenon is universal whether we think it is performed by a ghost in the machine or it is just another cog in the chain of prior causes.

This thread of the discussion carried on a little longer without a mutual understanding and eventually ended with me saying that I would try to explain myself in a new post.

So here we are. I currently suspect that we do not have libertarian free will; that is, I doubt that there is an uncaused part of us which controls the act of choosing. This is not a certainty, but I am compelled by the evidence (and the lack of alternative evidence) that this is probably a correct description of reality. So, now that you have received this revelation, you may climb back in bed and curl up in a ball and wait for your death because you are just a cog in a chain of causes. You are no different than the computing device you are currently using. You are a powerless bag of molecules, a meat puppet dangling by the strings of chance. Upon believing that your choices are byproducts of everything else, you could, paradoxically, immediately succumb to a self-defeating fatalism or you could keep reading and take another path. What will you do? Is that even a meaningful question?

This post does not seek to argue whether or not we actually have libertarian free will. The point of this post is to consider the implications for our sense of freedom if we do not possess uncaused agency.

Wait. How do you explain our experience of choice?

Pinocchio_to_lifeGood question. Even though I have no intention here of making the case for an absence of libertarian free will, it is worth considering whether that situation is even possible. I would like to start by reflecting on some observations which are representative of things that we’ve all experienced at one time or another.

The other day the book I was reading included a comment that “…animals don’t seem to want to party, despite what we see in children’s cartoons like Madagascar.” About 30 minutes after reading that – I’m slightly embarrassed to admit – I found myself with the Katy Perry song “Firework” in my head. Upon recognizing this I was surprised, so I stewed on it a bit. This is not a song that I encounter frequently in my listening habits. When I stopped to think about this, a faint scene began to play in my mind. It was an animation of zoo animals performing circus acts. You see, about a week earlier, I spent a couple hours watching Madagascar 3 with my sons. Near the end of the movie, the main characters engage in an elaborate circus performance set to the music of – you guessed it – “Firework”. Unbeknownst to me, the reference to the Madagascar movie in the book I was reading had set in motion a network of activity, drawing on recent experience, that led to the production of a particular song in my head.

When I was a kid my brother would play the “made you flinch” game. It may be a stretch to call it a game, but the rules are basically this: at any time, you can go up to your sibling and act like you’re going to hit them and then stop short. If they react in a defensive way then you have license to actually hit them. Twice. By definition, a flinch is involuntary. After enough bruises you learn to remain vigilant and can suspend your reaction, but eventually you will be caught off-guard again. Control of the flinch is subject to awareness.

As a final example, we’re all well aware that repetition can train us to do things effortlessly and thoughtlessly even though these things required considerable conscious attention during the initial training. This includes actions like reading, riding a bike, driving a car, using a mouse, etc… Even simple math eventually becomes automatic. These well-trained processes seem to lie on the borderlands between the intentional and the unintentional, lying just below the level of consciousness and waffling in and out of our awareness. We sometimes catch ourselves unaware that we had done something, or are doing something.

As these examples show, it is possible for behavior and mental activity to arise outside of our immediate awareness and control. They do not run through the “free will” filter. If we acknowledge that this is possible then it seems reasonable to acknowledge the further possibility that choice itself, our apparent exercise of free will, restraint and deliberation, can also arise through causative factors outside of our awareness. Under this paradigm, we might say that choice is what happens when our brain deals with competing interests. Even choosing to get up and get a drink is in competition with a desire to conserve energy and stay where you are. We have a remarkable feedback system that can recall past experiences and forecast future experiences. These work themselves in to the choice equation and sometimes we can spend considerable time and energy in deliberation as the network keeps pulling up data on both sides of the tug-of-war and reconfiguring itself in response.

The insistence that we make choices independent of causative influence begs the question. It assumes that our identity is fully contained within a singular, unified, independent perspective; in short, a ghost in the machine. Yet, if we ask someone who has flinched whether they chose to flinch then they’re most likely going to say that it wasn’t a choice while at the same time agreeing that they acted. Likewise, we will not deny that it was us who performed automated tasks, even if we weren’t fully aware of what we were doing. So in some cases our action can come from some sort of involuntary aspect of our self. That is, we do not always disassociate our self identity from the actions which were not clearly “under our control”. If we accept that this is a part of who we are and that the line between voluntary and involuntary does not demarcate our identity, then I see no reason why the abolition of libertarian free will should be seen to annihilate the self and render us incapable of choice. Instead, our conception of the “self who chooses” must be revised so that it is consistent with the fact that we already include our involuntary self in our identity. We dispose of the idea that we are a singular, unified and independent soul and find that our identity is multifaceted, distributed and interdependent. Incidentally, a rare group of split-brain patients have offered us a fascinating window into how this works, as do patients who have experienced certain brain injuries (see blindsight, visual agnosia and hemispatial neglect). It appears that this distributed view of the self is the more accurate perspective.

You should believe that you can make choices

pinocchio_donkeyAs demonstrated by the original quote at the top of this post, it is common to see claims that the rejection of libertarian free will is also the rejection of choice. I will address that claim further in the next section, but first I want to briefly review why you should believe that you – this new, complex, multifaceted you – can make choices. When we believe in free will:

  1. We are less likely to harm each other and more likely to help each other (Baumeister 2009).
  2. We are more likely to act fairly and show gratitude (Vohs 2008, MacKenzie 2014; though Zwaan failed to reproduce Vohs).
  3. We show less prejudice against predetermined group membership (Zhao 2014, and accordingly, we show more prejudice against chosen group membership – Brewer 2013).
  4. We are more likely to detect errors and invest in thought (Rigoni 2014, Rigoni 2013, Lynn 2013).
  5. We are less likely to succumb to impulses and more likely to exercise self-control (Rigoni 2012, Alquist 2013, Job 2010).

Given these results, the evidence seems to suggest that we prefer the versions of ourselves who believe in free will. The pragmatist follows by suggesting that the rational thing to do is to believe that we actually possess this freedom.

But I can’t just pretend for the benefits

I completely understand the objection and agree that in the short term we can’t choose our beliefs – but I’m also pretty sure that you don’t have to pretend. Even when you think you can give a reason for your choice we can always just ask why again, and keep asking why until you get to the point of saying “I don’t know”. Eventually you will get there, which means that as far as we can tell from pure introspection, there appears to be something unexplainable going on. This is where we find our “free will”.

It is possible that there actually is no prior cause at the bottom of this search but, as we have seen, it is also possible that the prior causes are simply elusive or inaccessible. If you disagree, please explain to me how this kind of experience would differ from the experience under libertarian free will. I don’t see a difference and, introspectively, we have nothing but our experience to go on. So, if our internal experience regularly lacks a fully formed understanding of causation and if we recognize that we can choose between options, why does it matter whether or not our choice is actually uncaused? Pragmatism takes over when explanations run dry and suggests that instead of looking at causes, we should look at effects. We feel a sense of control and operate with the experience of control and this results in outcomes which accord with our choice. Is this not sufficient?

pinocchio_homeFrom a purely experiential perspective, I make choices. If there is no libertarian free will then I may end up in bed, shut off from the outside world because all prior causes led to that condition. However, it is equally true that all prior causes may lead me to fight off the melancholy and seize the day. We don’t know which is the future path of the causal chain, yet we detect an ability to direct it. The internal experience is the same; our sense of freedom is present no matter what. This is all that matters when it comes to the choices we make. You needn’t sacrifice your freedom on the alter of fatalism. You have a choice.

If you have read this, and you find yourself agreeing with my conclusions, then it is possible that your experiences have now changed you so that you are more inclined to invoke your sense of free will. Ironically, you have just been externally caused to have a greater sense of freedom. Run with it.


A few comments on Plantinga’s “Where the Conflict Really Lies”

I don’t plan on commenting on every book I read but I was compelled to address what appeared to me to be some glaring omissions and one audacious claim in the argumentation found in Alvin Plantinga’s “Where the Conflict Really Lies: Science, Religion, & Naturalism”. There were three particular assertions that caught my attention: (1) that evolutionary theory says nothing about whether it is unguided, (2) a logical proof that determinism is false, and (3) that unguided evolution cannot yield reliable beliefs (aka “the evolutionary argument against naturalism”, or EAAN).

There is no reason to believe that evolution is unguided?


Ignoring the theological implications and biblical creation accounts, Plantinga says that “The scientific theory of evolution as such is not incompatible with Christian belief; what is incompatible with it is the idea that evolution, natural selection, is unguided.” I agree that it is not necessary to assert that evolution is unguided. There is no way that we can show that some supernatural agent is not overseeing the genetic changes which drive evolution. Fair enough. However, Plantinga goes on to say that “But that [the idea that evolution is unguided] isn’t part of evolutionary theory as such; it’s instead a metaphysical or theological addition.” What struck me is that Plantinga seemingly makes this claim without engaging with the foundational reasons why evolution is generally defined to be unguided. Instead, he chooses to review the arguments which show how complexity could arise by an unguided process. Yes, those arguments don’t prove that the process is unguided but that is beside the point. The task at hand is to find the best explanation for our observations. Is the best explanation that evolution is guided, or is the best explanation that evolution is unguided? When I survey the data, I see compelling reasons for inferring an unguided process. For example:

  1. The vast majority of species that have ever existed are now extinct. Natural selection occurs by killing off creatures with the less favorable property. Competition and death are fundamental components of the evolutionary process.
  2. It is far more likely that a mutation is neutral or deleterious than beneficial.
  3. What were once beneficial adaptations can become deleterious in the face of a changing environment.
  4. Artificial selection (for example, in dogs) has produced in hundreds of generations a degree of variation that is only comparatively realized in nature over thousands of generations.

Among others, these are all characteristics of evolution which, to me, infer an unguided process. The first observation demonstrates how wasteful, vicious and “immoral” the process is. If you want to argue that we have no reason to believe that God wouldn’t create through such mechanisms then that’s fine, but at least admit that it is not how we expect an all-loving, all-powerful, super-intelligent being to act and is among the least attractive of the possible methods (e.g., special creation). The second observation highlights how the process seems to be driven by a small fraction of changes in a probabilistic paradigm, which is almost by definition the opposite of a guided process. The third observation demonstrates that the result of selection does not always lead to a long-term benefit. Again, this seems to contradict an intelligence behind the outcomes. Lastly, the final observation reveals how inefficiently slow the changes are accumulated in nature, whereas a known intelligent agent (humans) has succeeded in utilizing the exact same underlying mechanisms to realize dramatic changes in a short period of time.

On the flip side, one could argue that the amazing outcomes of evolution – the eye, flight, the brain – are all pointers toward a guiding intelligence. I understand this view; it is truly amazing what has been wrought. I feel the draw of the design explanation when I consider the remarkable intricacies of life, but I also recognize that this pull arises because the design hypothesis is easier to relate to our experience (i.e., our intuitions are biased toward that model). This is not the place to rely on intuitions, however, so we must turn to the evidence. In a twist of irony, Plantinga has already included arguments which explain how these wonders may result from an unguided process. So the counter to the argument for guided evolution has been presented and acknowledged. As I see it, this means that the reasons for thinking that evolution is unguided weren’t addressed at all and the reasons for thinking that evolution is guided were found to also fit the unguided paradigm. I cannot agree that this conflict is merely superficial.

Determinism is logically impossible?

Though it is a minor side-note in the book, it immediately caught my attention. What an audacious claim – a logical proof that determinism is false! This warranted a closer look. The argument is as follows:

  1. A natural law is of the form “If the universe (U) is causally closed, then P.”
  2. Also take the conjunction of all natural laws to be “If U is causally closed, then P.”
  3. If determinism is true then the conjunction of all natural laws (If U is causally closed, then P) and a specific past state of the universe (PAST) necessarily entails the future (F).
  4. Using N to mean Necessarily, the above statement is equivalent to: N [if (if U is causally closed then P) and PAST, then F].
  5. Becomes:  N [if (either U is not causally closed or P) and PAST, then F]
  6. Becomes:  N [if [(PAST and P) or (PAST and U is not causally closed)], then F].
  7. This takes the form N if (p or q) then r, which means that both p and q entail r, hence
  8. N [if (Past and P) then F] and N [if (PAST and U is not causally closed) then F].
  9. The right hand side of #8 is obviously false because there is clearly a possible world that (i) shares its past with the actual world, (ii) is not causally closed (because perhaps God acted) and (iii) does not share its future with the actual world. Therefore, determinism is false.

I will admit that it took me several reads to follow this argument. In the end, however, I think I see the slight of hand (whether or not this was intentional, I do not know). It was my attempt to translate this into software code that clearly revealed the problem for me. Here’s the code:

function CreateFuture(Universe, Past) {
  if(Universe.CausallyClosed) {
    P = Universe.NaturalLaws;
  return DoPhysics(P, Past);

My fellow software engineers will immediately recognize the bug in this function: if the universe is not causally closed then P is undefined and an attempt to use it to generate the future yields unpredictable results. This is the key to the problem with the proof. In step 5, Plantinga expands the proof to cover both branches of the conditional and infers that both branches are still bound within the definition of determinism. This then, of course, leads to the obvious result where the future created with defined natural laws may be different than the future created with undefined natural laws. Plantinga groups these outcomes together under the definition of determinism and declares that the internal inconsistency shows that determinism is false. This is completely invalid, however, because determinism is only defined to be the branch where the universe is causally closed. All other branches (or possible worlds) are something other than determinism. I am honestly a bit baffled that Plantinga chose to include this in the book.

Despite my rejection of this proof, I should note that I am not a strict determinist. I would consider myself something of a quasi-determinist. Quantum indeterminacy has shown us that we can’t (yet) predict all possible states, but the quantum effects adhere to a predictable distribution such that the macro-world, and even the molecular world, behaves according to the physical laws to the extent that we have accurately described them. In the absence of supernatural intervention the natural world is, for all practical purposes, deterministic.

Naturalism cannot yield reliable beliefs?

The central thesis of the evolutionary argument against naturalism (EAAN) is that, without a guiding force, there is no reason to believe that the evolutionary process would result in a belief forming system that yields true beliefs. As I read through the EAAN, I was eagerly anticipating Plantinga’s response to the following objection: evolutionary theory claims that well before any creature was conscious there were sensory systems that triggered responses which selected the population. Selection is dependent on beneficial interactions with the external world. If those interactions do not consistently and properly map to the outside world then they are less likely to be beneficial. Consciousness and belief formation are extensions of this rudimentary system. As such, the evolutionary processes which led to sensory response systems, and therefore also to consciousness and belief formation, are likely to produce reliable reflections of the outside world.

Finally, in chapter 10, part 5, section C he gets there. He introduces the objection with “Isn’t it just obvious that true beliefs will facilitate adaptive action?” and eventually follows-up with “Yes, certainly. This is indeed true. But it is also irrelevant.” Then comes the explanation: “We ordinarily think true belief leads to successful action because we also think that the beliefs cause actions, and do so by virtue of their content… But now suppose materialism were true: then, as we’ve seen, my belief will be a neural structure that has both NP [neuro-physiological] properties and also a propositional content. It is by virtue of the NP properties, however, not the content, that the belief causes what it does cause.” After providing several examples of how the content of beliefs result in action, he finishes with “Going back to materialism…If the belief had had the same NP properties but different content, it would have had the same effect on behavior.”

Plantinga immediately recognizes that materialism would deny that it is possible for two beliefs to have the same NP properties but different content. Then things get messy. He digresses into a brief discussion of how this isn’t the place to address how counterfactuals and counterpossibles should be used in argumentation. Then he closes the response with “..it doesn’t matter to the adaptiveness of the behavior (or of the neurology that causes that behavior) whether the content determined by that neurology is true.” Wait a second – isn’t that where we were before this whole objection was raised?

Is it just me, or did he completely misrepresent the naturalistic ontology of belief and then dismiss the objection to that misrepresentation without offering an explanation? It seems as if he has superimposed dualism onto naturalism and then argued against this bastard child. What really confuses me is that in the pages leading up to this he clearly defined the materialistic view as one in which belief content can be reduced to NP properties. Somehow, when it came time to address the big objection, this reduction no longer applied and content was now something completely separate from the physical. How did this happen? I re-read those pages several times and I just don’t get it. Am I in over my head? Did I miss something? I can’t help but feel like I did; but, then again, I’m far from being the only one who has seen problems with this argument. This turn of events left me bewildered and I can’t give any regard to the EAAN until this is resolved.

Closing Thoughts

This was my first encounter with any of Plantinga’s books, though I was familiar with his work and was well aware of his reputation as one of Christianity’s greatest thinkers. The writing generally lived up to the standard; the text was lucid and, in most cases, the arguments were easy to follow. It was a worthwhile read. In the end, however, I was severely disappointed that his key claims – the “apparent conflict” between theism and evolution and the “deep conflict” between naturalism and science – ultimately omit or dismiss the most relevant objections to those claims. Perhaps even more alarming was that he chose to publish a clearly flawed proof that determinism is false. Collectively these observations have done nothing but tarnish his reputation in my eyes. I had hoped for something more.