Consciousness (IX) — Obey Libet? No Way, José.

(Note: I still consider this entry to be in rough draft form. I haven’t exactly been in the clearest state of mind. Though the rest of the essay gets better, I think, rereading the first paragraph made that especially clear to me. I don’t post these on Patreon until they’re in a refined and finished form. However, so long as I’m capable of writing but not capable of editing for clarity, I’d rather write and have sub–par work already done and ready to be refined at some point later than sit and wait until I can pull off essays that feel to me like perfect tens.)

One of the best cases for showing just how deeply bad philosophy can corrupt perfectly reasonable scientific experiment, straining data through the filter of philosophical lenses of interpretation to create something that is now not just raw data, but philosophy generated only partially in reaction to data—and also partially through the lens of conceptual filters which are themselves justified not by data, but by rationalistic considerations about what constraints a satisfactory account of the phenomena in question would need to conform to in principle—and then pretending not to have done so; pretending that the end result of this process is just plain science, refusing to defend the philosophical premises involved in it on the philosophical terms they require, and implying that these concepts therefore have the full weight of authority of a finding of Science per se, is the “science ‘on’ free will.” I write this phrase in scare–quotes because this data, in and of itself, isn’t “on” free will at all—to interpret it as bearing relevance to the question of free will is a philosophical claim about the science, and not some simple and straightforward description of what the data of that scientific investigation itself is plainly doing. I think this will become clear as I proceed through the analysis and spell out the exact details of what I mean.

There is a great deal of science that is claimed to have relevance for the question of free will, from the “readiness potentials” found in experiments from neuroscience to the “situationist effects” found in experiments from social psychology. As I spell out the way these have been argued to undermine belief in the possibility of free will, it will become clear that the crucial steps of interpretation involve philosophical assumptions—and that with different philosophical assumptions we could very well interpret that very same data to significantly different results. The crucial questions, therefore, hinge on which set of philosophical assumptions can be most well–justified—and our understanding of the data will follow from these, rather than the reverse: these philosophical assumptions are not proven or disproven by the data[1], but stand independent from it and come first; these assumptions determine how we will read the data. It only ever appears to be otherwise—it only ever appears that empirical data conclusively proves answers to the philosophical questions that determine how we interpret that data—because people interpret the empirical data through philosophical assumptions which they do not explicitly identify and recognize and then—voilà; quelle surprise!—something comes out of the other end of the process that conforms with those very assumptions. The greatest value and import of philosophy is precisely to draw these implicit assumptions out into our explicit awareness so that we can acknowledge them as the assumptions that they are—recognize that they are not, in fact, the only option—and then evaluate them in actually appropriate terms against the relevant, real alternatives.

 _______ ~.::[༒]::.~ _______

Notice: When I talk about “free will” in this post, I am talking about the robust sort of free will that entails that right up until the moment of my choice, nothing in the previous physical states of the Universe determines what my choice is going to be; and at the moment I make my choice, I alone determine what that decision will beThe idea that this is the sort of “free will” that most of us feel that we have is an empirical question, and it is separate from the question of whether we really have it. The technical term for the range of views which say that this is kind of “free will” that we feel ourselves to have (or want, or should want) is “incompatibilism.” “Compatibilists,” by contrast, argue that the only sense of the term “free will” that we do want, or should want, or perhaps that even means anything at all is the sense in which I make the decision because I want to, and not because someone else has a gun to my head—even if both my decision and my desire were absolutely set in metaphysical stone from the moment of the Big Bang. While compatibilism is a more or less uniform position, incompatibilists are divided between those who believe that we do have the metaphysical kind of free will they believe we intuitively feel ourselves to have (and these are called “libertarians”), and those who believe that we do not (and these are called “hard determinists.”)

I adopt the incompatibilist definition for two reasons: first, because given that it is the most “metaphysical” sense of the concept of “free will,” using it will even more forcefully demonstrate my point that metaphysical assumptions influence how we will interpret empirical findings of science to begin with far more than scientific findings will determine what metaphysical convictions we will come to hold, both in practice (because in practice, we are reading scientific findings through metaphysical assumptions even if we do not realize and acknowledge this at all) and in principle (because it is impossible in principle not to read scientific findings through metaphysical assumptions, and also impossible in principle to actually arrive at metaphysical convictions through scientific data alone without turning that data into actual concepts by filtering it through philosophical assumptions). This point will stand even if you think the idea that this kind of free will actually could in fact be possible is nonsense

Second, I adopt the incompatibilist definition because I think it is empirically correct. This 2010 study by Sarkissian (et al), Is Belief in Free Will a Cultural Universal? “extends previous research by presenting a cross–cultural study examining intuitions about free will and moral responsibility in subjects from the United States, Hong Kong, India and Colombia. The results revealed a striking degree of cross–cultural convergence. In all four cultural groups, the majority of participants said that (a) our universe is indeterministic and (b) moral responsibility is not compatible with determinism”—these findings, Sarkissian (et al) argue, imply “fundamental truth[s] about the way people think about human freedom.”

In the book, Free Will and Consciousness: A Determinist Account of the Illusion of Free Will, “hard determinist” Gregg D. Caruso writes: “ … I maintain that our phenomenology strongly supports an incompatibilist, libertarian, essentially agent–causal conception of free will. … compatibilists cannot simply neglect or dismiss the nature of agentive experience. … our phenomenology is rather definitive. From a first–person point of view, we feel as though we are self–determining agents who are capable of acting counter–causally. … we all experience, as Galen Strawson puts it, a sense of “radical, absolute, buckstopping up–to–me–ness in choice and actions” (2004, 380). …  In addition to experiencing a robust sense of self, we also perceive ourselves to be uncaused causes. When I perform a voluntary act, like reaching out to pick up my coffe mug, I feel as though it is I, myself, that causes the motion. We feel as though we are self–moving beings that are causally undetermined by antecedent events.”

So why does Caruso turn against these libertarian–incompatibilist intuitions? To quote from Jonathan M. S. Pearce’s review, “Caruso characterizes agent–causalism as a theory committed to a dualist picture of the self. And it is this alleged feature of agent–causal strategies that is his main target; he argues that agent–causation involves a violation of physical causal closure (pp. 29-42).” Obviously, it stands to reason that neither Pearce nor Caruso seriously consider the possibility that something like the dualist picture of the self could in fact be true—but we have already seen that this, itself, is a metaphysical question which depends on the truth or falsity of philosophical conceptual claims about the relationship between objectively measurable physical states and subjective states of conscious experience that cannot be established by “data” as such which by definition deals with only one half of this equation. Again summarizing the quotation from William James in the last entry, “In strict science, we can only write down the bare fact of concomitance [between conscious experiences and physical brain states]; and all talk about either production or transmission, as the mode of taking place, is pure superadded hypothesis, and metaphysical hypothesis at that, for we can frame no more notion of the details on the one alternative than on the other. Ask for any indication of the exact process either of transmission or of production, and Science confesses her imagination to be bankrupt.”

But notice that Caruso sees no need to defend the conception he holds of the nature of causal closure, which blithely rules out the possibility of conscious causal efficacy by definition—that this premise is both necessarily true, and necessarily true in a way that rules out the possibility of the dualist picture of the self he claims agent–causal theories depend on is simply taken for granted. And Pearce doesn’t question it either—he simply wonders if Caruso isn’t “sparring with a straw opponent,” because the notion that anyone could actually think with any degree of justification whatsoever that a dualist picture of the self just might be accurate is too absurd to even consider—so Pearce can only wonder if Caruso isn’t making the job too easy by imagining that that’s what his opponent thinks. (I’ll be exploring these arguments from causal closure in more detail later on. Suffice to say for now that I don’t think it’s anywhere near that easy. I stop just short of saying that the principle of causal closure, at least as ordinarily understood, can be downright refuted deductively.)

In any case, the point stands that these are philosophical considerations, and not anything proven by any kind of direct experiment—and notice that Caruso begins addressing them on page 15, well before the first mention of the usual introduction to scientific studies on the topic—the experiments conducted by Benjamin Libet in the 1980’s—begin to appear somewhere past page 100. The account is one which the advocate of agent causation cannot put forward “wholly without embarrassment” because it would “require giving up … one of the core principles of atomistic physicalism”: namely, that all the properties of any given thing are nothing more than the sum of its parts—and the properties which atomistic physicalism supposes the basic “parts” of reality to be composed of are all presumed by definition to be utterly mindless: lacking in intentionality, lacking experientiality, acting blindly as a passive result of inert ‘causes’ rather than ever for positively adopted ‘reasons,’ and so forth. That questioning this premise is deemed to be “embarrassing” for the advocate of agent causation, while no “embarrassment” is supposed to come for the physicalist from the complete and absolute absence of any working account of how anyone might even begin to try to derive those properties from ingredients supposed to be wholly lacking in them (see essays I–V in this series), is less a compelling argument than it is a striking illustration of intellectual fashion’s double standards. The non–physicalist is supposed to feel like a white daughter of the 1950’s confessing to her parents that she thinks she wants to date a black man: “Why, you just ought to feel ashamed of yourself!”—embarrassed for even having had the thought.

A final note: It is not my purpose here to positively “prove” that metaphysical free will does, in fact, exist—that would be too tall an order for an essay even several times this length; not least because even if consciousness is an irreducible phenomena in its own right, as I argue, this still leaves open the hypothetical possibility that consciousness could be “determined” according to its own unique kinds of rules. The question I’m concerned with here is more simply: if you do think your internal experience presents itself as possessing metaphysically free capacities for self–determining choice (and even if you don’t, that simply doesn’t change the fact that a large percentage of people very obviously do), does “science” give you overriding reasons to conclude that this sensation is nothing more than an illusion? If you really don’t share that sensation, or find any interest in the idea that the “scientific” facts may in fact still leave abundant room for the empirical possibility of it, then you can still derive value from this essay by reading them in light of the following question: “If we really should throw out the ordinary concept of free will, should that be for philosophical and conceptual reasons, or because of specific facts which rule out that possibility that have been demonstrated empirically true by ‘science’ without any involvement from or filtering through philosophical interpretations at all?” Either way, my argument is that—contrary to claims popular in some corners—“science” actually demonstrates strikingly little of direct relevance to these questions, either by “empirically” proving the inefficacy or epiphenomenal nature of conscious intention specifically or (as has been argued across the several entries preceding this one) by “empirically” proving the truth of the claim that subjective conscious intentionality and experience is either an epiphenomena of or just “identical to” the brain qua physical object in general: as William James wrote, this is as much a metaphysical hypothesis superimposed upon the mere empirical finding of “concomitance” as any other.

 _______ ~.::[༒]::.~ _______

If asked to describe our own internal experiences of our own “free will,” I think most of us would agree that our phenomenal experiences themselves indicate a difference between on the one hand our choices, and on the other hand our urges. Even those of us who feel as if we truly do have the power to ‘really’ choose between alternative decisions at the very moment of choice very clearly do not feel that we choose the moment when an urge submits itself to our conscious awareness and forces us to make a decision about it. I may believe that I have the power—at any given moment—regardless of the preceding physical states of myself and the world around me—to choose whether to continue writing this post, or to stop and take a break—without believing in any way that I control the moment at which the urge to get up and do something else will become present within my experience. If exploring the phenomenology of free will (e.g., how it ‘appears’ from the ‘inside’) is our goal, then clarifying the distinction between decisions and urges is one of the most rudimentary opening clarifications we ought to make right in the very beginning. And if you want to say that the way people experience the sense of possessing free will is inaccurate, then you have to get what it is that they feel that they experience right first. 

But this is a distinction that can only be made “from the inside.” Scientific investigation can identify physiological correlates with sensations of urge or the feeling of making a decision, but it can’t have any idea at all what sensation it is identifying physiological correlates withunless it relies on trying to correlate that physiology with a subject’s reports about their own internal subjective states. Without the intrinsic involvement of a subject doing their best to report what their own conscious experiences ‘feel like,’ scientific investigation of brain physiology can’t know what part of an individual’s concept of consciousness it is identifying aspects of. It can identify, for example, that activating a certain part of the visual cortex leads in turn to the activation of the amygdala and then leads to sweating, release of adrenaline, and so forth—but unless it relies on the subjective reports of an individual describing their own internal states of consciousness, it can’t know that “when you turned that thing on, it made me start flashing back to childhood traumas, and visualizing that made me start to panic.”

Without continually referring back to the subjective reports by subjects of their own internal states of consciousness, it would have been a mystery why activating that same location in a second subject instead lead to a release of endorphins—until they said, “when you turned that thing on, it made me start flashing back to how wonderful my childhood was, and visualizing that made me feel comforted”—and from here, we could infer that the region of the brain we were activating was probably linked to childhood experience. Now you may be thinking we could have found that the region of neurons in question which were being activated in the experiment are in some way connected to neurons that grow in either a healthy or deformed way during childhood, without relying on any subjective reports—but we just as easily could go on to find that in an equal number of subjects with childhood trauma, stimulating that part of the brain doesn’t lead to activation of the amygdala at all.And once again, we simply would have no way to form a clue about what was happening unless the subject said, “My childhood was so traumatic that I learned how to just deaden myself and stop feeling emotions at all.”

Again, you might propose that we could have identified historical differences between those subjects with traumatic childhoods in which the amygdala was activated and those with traumatic childhoods in which it was not which correlated with their “learning to stop feeling emotions,” but for any such account you might give, there is always some report that could be given that would potentially undermine that whole account (we’ve just given two examples), and you would simply have no reason to think that this had been the right place to look if you had not had the subjective report itself. Why wouldn’t you have hypothesized that the subjects with traumatic childhoods whose amygdalas did not activate were simply born with less interconnected amygdalas—a physical fact that would have nothing to do with their “response” to the trauma they went through? If you assume that the proper, correct answer would have to be in terms of the purely physical properties of the brain and that the “response” could not be a directly conscious choice of action on the part of the subject, but could only be their way of describing how what their brain, qua physical process, was doing ‘felt like’ after the fact, then you are making the philosophical assumption that epiphenomenalism is true. But there are overwhelmingly good philosophical reasons to think that epiphenomenalism is false—if it were true, then we wouldn’t even be capable of making ‘reports’ about what our conscious states of experience ‘felt like,’ because the brains that produce both our thoughts and capacity for verbal speech would have none of the capacity for causal contact with those epiphenomenal experiences that would be required in principle for it to “know” anything whatsoever about them at all. Likewise, if your answer is that it is merely an epistemological limitation that we have to rely on the reports of subjects because we can’t possibly have all the relevant physical facts, then like it or not, this is an inadvertent confession that the idea that the physical facts (construed, again, in purely mechanistic terms) would be sufficient to explain the phenomena in question is not something you “empirically” know.

Perhaps some day we will gain the ability to reconstruct what someone is visualizing simply by taking a measurement of their brain state and mapping that activity out to convert it directly into an image—but if we ever reach that day, it will only be because we spent a long stretch of time learning what brain states correlate with which aspects of subjectively experienced qualitative imagery by relying on subjects’ subjective reports in order to establish the brute facts about these correlations. This returns us, of course, to the general points established in the core philosophical part of this series in essays (IV) and (V) about qualitative experience and intentionality: physical states, qua physical states, simply are never “about” anything at all. And knowledge of physical causation cannot, in principle, give us direct knowledge about the qualitative state of subjective experience. Simply looking at the brain as a physical object doesn’t give us any reason to think any sort of stream of experience is taking place “inside of” it at all, and if we did not have the example of our own first–hand case to inform us that there appears to be some sort of correlation between subjective states of experience and physical states of brains, we would have no reason to make the inference into assuming any experiences were even happening to start with.

Not only can purely physical information not tell us what someone is experiencing, unless we already have a brute set of correlations between physical states and subjective experiences to go on which was necessarily established either by (1) first–hand knowledge of our own consciousness or (2) second–hand reports from someone else about their states of consciousness to begin with (the “knowledge argument” about “qualia”), but purely physical information cannot tell us what someone is thinking “about,” either (the “knowledge argument” applied to intentionality—see: What is it Like to be  Human (Instead of a Bat)? by Lawrence BonJour, or Bill Vallicella’s summaries of it here in Intentionality Not a ‘Hard Problem’ for Physicalists? and here in BonJour on Intentionality and Materialism). Consciousness is fundamentally and thoroughly composed of both phenomenal experience ‘on the edges’ and intentionalistic thought (which is also phenomenal) ‘at its core’—but physicalism cannot account for so much as the existence of either; and knowledge about the physical details of physical states cannot give us knowledge about the details of either[2], either—not unless we first rely on a subject’s verbal reports about their own internal state, and then simply accept whatever correlations we might happen to find between physical states and second–hand reports of intentional and experiential states as brute facts. But it simply is not clear what it is that finding these correlations establishes—not until we start the incredibly complicated intellectual work of trying very carefully to interpret them—which is fundamentally a philosophical project.

 _______ ~.::[༒]::.~ _______

 The story of “scientific” refutation of the possibility of free will begins in the 1980’s, with the scientific studies conducted by Benjamin Libet. Though now more than three decades old, these experiments still carry a large bulk of the “scientific” analysis of the implausibility of free will. In (New Atheist and practicing neuroscientist) Sam Harris’ 2012 book Free Will, he writes: “The physiologist Benjamin Libet famously used EEG to show that activity in the brain’s motor cortex can be detected some 300 milliseconds before a person feels that he has decided to move. Another lab extended this work using functional magnetic resonance imaging (fMRI): Subjects were asked to press one of two buttons while watching a “clock” composed of a random sequence of letters appearing on the screen. They reported which letter was visible at the moment they decided to press one button or the other. . . . One fact now seems indisputable: Some moments before you are aware of what you will do next—a time in which you subjectively appear to have complete freedom to behave however you please—your brain has already determined what you will do. You then become conscious of this “decision” and believe that you are in the process of making it.”

Daniel Wegner is one of the most prominent social psychologists known for his continuation of experiments aiming to prove this general sort of idea. In his discussion of Libet’s experiments in the 2002 The Illusion of Conscious Will, he explains the picture of mind that he still believes the Libet experiments prove: “Does the compass steer the ship? … [not] in any physical sense. The needle is just gliding around in the compass housing, doing no actual steering at all. It is thus tempting to relegate the little magnetic pointer to the class of epiphenomena — things that don’t really matter in determining where the ship will go. Conscious will is the mind’s compass.” In other words, preceding unconscious brain events are the cause of both our future behaviors, and our later, illusory experience of making the “choice” of those behaviors over others—it isn’t even that our experiences of choice are determined; they’re also completely superfluous to the chain of events that even lead to the actual action which we associate with the experience of choice.

How well–justified is this claim? A variety of criticisms could and have been leveled at these experiments, but I want to focus on just one of them—one I consider most obviously decisive and fatal. I argued earlier that one of the most rudimentary distinctions we should make if we want to think seriously about the way ‘free will’ seems is between urges and decisions. The set–up of the Libet experiments looses this distinction completely.

To repeat the explanation in my words, the Libet–type experiments first have a subject sit down in front of a clock, while hooked up to an EEG (or fMRI). Then, they explicitly instruct that subject to perform some simple motor activity at random—absolutely nothing is at stake in the decision; there is no goal to achieve, there are no values or variables to weigh or choose between, and no number of button presses or wrist–flicks is too high or too low. There is no way to “win,” there is no way to “fail,” and there are no alternative outcomes in the experiment for the subject to pick between. With absolutely no goals or constraints, subjects in these experiments are told to sit back and perform a perfectly purposeless motion at random for which they have absolutely no reason in principle to choose one moment over another.

Stop right there.

What do you expect might happen if you were to do that? Imagine being told to spend several minutes flipping your hand back and forth from palm–up to palm–down at random. What do you think that might feel like? For just a moment, allow yourself to pause and imagine it—or even perform the experiment (presumably minus electrodes)—before continuing.

The shocking, startling discovery these experiments find is that when test subjects tell the conductors where the clock was when they “decided” to move, there was a type of activity they designated the “readiness potential” in the brain already detectably building in the milliseconds leading up to the moment they became aware of having made the “decision” to move.

Quite simply, without even touching any of the many other angles of critique that these arguments face at all: Why should anyone interpret this as the subconscious generation of the decision itself to begin with? The answer is a plain and straightforward: they shouldn’t. What does it feel like when you set the general intention to perform a purposeless movement at random? It feels like setting the intention to sit back and allow the urge to move to randomly appear, and then waiting for it—and then moving when it occurs. Does that not in fact feel exactly like waiting for a physical “urge” to appear before acting on it? So why should the fact that a kind of brain activity precedes the decision in this very peculiar kind of case lead us in any way to even suspect on this basis that decisions in general are determined by preceding unconscious brain activity?

If anything, even those of us who think that the nature of first–hand experience offers prima facie justification for the belief that we might have the metaphysical kind of free will, sheer introspection alone should have led us to expect something like this: when we ride a bike, or type a sentence on a keyboard, or learn to play the guitar, it feels one way to initially learn how to perform the action, and it feels another way to perform once having learned. When I am learning to play the guitar, it feels as though I have to consciously deliberate each distinct individual action of placing my middle finger on the third fret of the bottom E string, my index finger on the second fret of the A string, and my ring finger on the third fret of the top E string; and then to move my middle finger to the second fret of the G string, my index finger to the second fret of the top E string, and my ring finger to the third fret of the B string; and then to move my middle finger to the second fret of the A string and my ring finger to the second fret of the D string. (Tedious!) With a little more practice, this begins to feel like “setting the intention to strum a G chord,” and then “setting the intention to strum a D chord,” and then “setting the intention to strum an Em chord” and allowing my hands to automatically fill in the rest—and with a little more practice, it simply feels like “setting the intention to play Freebird.” The more, in other words, that I consciously practice these motions, the less it feels as though I need to consciously deliberate each individual step—the more that the execution of the action becomes “automatic.” If neuroscience were to find that conscious deliberation plays no role the motion from G chord to D chord to Em when Gary Rossington plays the chords to Freebird, would that undercut anyone’s ordinary concept of free will in any way at all? I think not. And we shouldn’t look at cases where people are explicitly asked to sit passively and respond to random urges any differently—setting the general intention “to play Freebird” in advance and then allowing the details to physically carry through is no different in essence from setting the general intention “to flip my arm over at random, whenever I feel the urge” and then allowing those details to continue to physically carry through. I’ve already done the important intention setting at the point at which I decided either to begin playing Freebird, or chosen to walk inside Libet’s lab and passively follow his instructions. Anything that follows next is just quite simply categorically different—even at the most basic, subjective phenomenal level—from the kind of decision I made, at the beginning of either of these processes, to begin them.

There are, again, several other lines of critique that could be taken against drawing the determinist implications from Libet experiments—not least amongst them is that Libet himself went on to argue that we do in fact have the capacity to either allow the readiness potential to go through or to “veto” it even within the context of his own peculiar kinds of experiments (which in and of itself sounds like a confirmation of the fact that the “readiness potential” does not measure the decision itself but only a physical urge which is later decided upon). But I neither want nor need to explore that issue in the detail required here: anything further than this is simply an attempt to starve a dead horse to death. No matter what might hold true about the details of Libet–type experiments, there is just quite simply no reason nor justification for generalizing what happens when people are explicitly told to sit back and allow themselves to passively act on urges at random to what we should expect to hold true in any other sort of conscious state where decision plays a more active role at all—even if the situation is as, or more, dire in those conditions than Libet thought. The fact that a point this basic was so deeply missed by a study still hailed today as one of the most powerful pieces of “evidence” for the “scientific” impossibility of free will should give you some solid impressions about the quality of reasoning we’re dealing with when “science” is claimed to prove answers to philosophical questions about the nature of mind.

In conjunction with his claim about Benjamin Libet, you may have noticed that Sam Harris immediately followed up with a statement about “another lab [that] extended this work using functional magnetic resonance imaging (fMRI)….” The lab he refers to is Chun Siong Soon’s[3], and the summary of the 2008 study published in Nature Neuroscience can be seen here. While the activity measured in this study was still, as before, purposeless, with no goals or constrains, it did change one substantial thing. According to the way Soon (et al.) summarized their own research—in a summary paper titled “Unconscious Determinants of Free Decisions in the Brain”—“There has been a long controversy as to whether subjectively ‘free’ decisions are determined by brain activity ahead of time. We found that the outcome of a decision can be encoded in brain activity of prefrontal and parietal cortex up to 10 s before it enters awareness.” The actual point this new study was supposed to add to the already–existent debate was that it was supposed to establish the capacity of these scientific measurements to predict not just the general timing of a single choice, but now in fact which of two—count them, two—equally meaningless choices the subject would choose between. And the conclusions we are supposed to draw from this are, again, wide–reaching—returning to the summary from Harris: “One fact now seems indisputable: Some moments before you are aware of what you will do next—a time in which you subjectively appear to have complete freedom to behave however you please—your brain has already determined what you will do. You (only) then become conscious of this “decision” and believe (falsely) that you (“you”) are in the process of making it.”

What do the particular new facts drawn by this study add to the picture? There is one thing that neither Harris’ reference to this study, nor Soon (et al.)’s own summary of it in Nature Neuroscience, will clearly tell you—quoting Alfred Mele: “ … the predictions are accurate only 60 percent of the time. Using a coin, I can predict with 50–percent accuracy which button a participant will press next. And if the person agrees not to press a button for a minute (or an hour), I can make my predictions a minute (or an hour) in advance. I come out 10 points worse in accuracy, but I win big in terms of time. So what is indicated by the neural activity that Soon and colleagues measured? My money is on a slight unconscious bias toward a particular button—a bias that may give the participant about a 60–percent chance of pressing that button next.”

Notably, this 60–percent figure is a drop from a predictive value of 80–90% in cases where the moment chosen to commit a single predefined action—such as Libet’s wrist–rotating—is what is being predicted in the study. Even with the increased understanding of neurophysiology developed over the past handful of decades, and even with refined neuroimaging techniques, the predictive power of the “readiness potential” in this study still immediately drops by 20%—down to fairly little over chance—with even a slight shift of the design of the experiment towards something that comes even marginally closer to resembling the kinds of decisions in which we actually deliberate—and feel as if we deliberate freely—over a choice.

But yet again, even if the predictive value of the “readiness potential” in these expanded cases were 100%, why should even that have concerned me? When I go into Soon’s laboratory, I am walking in deliberately setting the conscious intention in advance to sit back and think about nothing other than letting myself push either one or the other button at random—and absolutely nothing weighs on the decision; I am by definition putting myself in the peculiar conscious state of waiting to act on a random urge. Even with this meaningless “choice” between two absolutely meaningless options added to the scenario, it doesn’t even feel like the kind of deliberation that subjectively presents itself as containing the power to do otherwise. I am setting the conscious intention to sit back and act randomly on one or the other urge—which, no less, it even feels as though I am willing the generation of in the first place—in other words, the act of choice that actually seems to present itself as feeling as if it possesses the metaphysical kind of freedom seems to be my decision to activate a program that says something to my body like: “you, allow meaningless urges to generate at random”, and to my brain something like: “and you, be prepared to act on them after they appear.”

The skeptic might scoff at this description of how I think things ‘seem’ ‘from the inside,’ and ridicule my ‘assumption’ that how things ‘feel’ could possibly be any indication of how reality actually is at all. But if so, this would just exactly illustrate the very central fallacy I’ve accused him of: the use of experiments like those conducted by Libet to argue that the phenomonology of decision–making is illusory has been “scientifically” proven simply does not treat that phenomenology seriously in the first place—and given how poorly, unfairly, and inaccurately it does so, we in turn are left with no reason to take the claim seriously that it proves that that phenomenology is delusional.

 _______ ~.::[༒]::.~ _______

A future post will address the conceptual issues involved in the coherency of the basic notion of metaphysical free will, as well as the question of what the practical implications of reaching either conclusion might be, and add some more details on the “science” of free will beyond Libet by discussing The Illusion of Conscious Will from Daniel Wegner. But the general point involved in both the conceptual and the “scientific” rejection of free will is quite the same: its plausibility stands or falls with the fall or stand of the premise of “atomistic physicalism.” I have argued throughout this series that not only do we have not half as much epistemic justification for believing “atomistic physicalism” to be true as is often implied—“science” most certainly has not empirically proven it; a set of philosophical assumptions through which certain findings of science are interpreted, or through which it is assumed that if a principle performs a useful methodological role in scientific investigation, then we can be guaranteed that it represents a universal metaphysical principle that holds inviolable in all places and times are in fact what are doing all the work—but in fact that there are systematic reasons for thinking any such account must necessarily fail in principle, in just the same way that attempts to draw three–dimensional figures on flat two–dimensional canvases will no matter how creative “empirical” attempts to accomplish it might ever even conceivably become.

It is not so much that neuroscientific experiments like these convince people on the sheer basis of the data they’ve collected alone that metaphysical free will is an empirical impossibility. It is much rather that they hold the philosophical view from the outset that it can’t be possible, because the mind just can’t be like that—and with this premise held in place, it is simply a matter of filling in the details about how choice is determined—but this underlying view is in turn far from proven by “science” either—recall the words of William James: “If we are talking of science positively understood, function can mean nothing more than bare concomitant variation. When the brain–activities change in one way, consciousness changes in another; when the currents pour through the occipital lobes, consciousness sees things; when through the lower frontal region, consciousness says things to itself; when they stop, she goes to sleep, etc. In strict science, we can only write down the bare fact of concomitance; and all talk about either production or transmission, as the mode of taking place, is pure superadded hypothesis, and metaphysical hypothesis at that, for we can frame no more notion of the details on the one alternative than on the other.”

The idea that the possibility of metaphysical free will is ruled out by “science” ultimately rests on interpretation of that scientific data which is drawn from the assumption being held from the outset that the conception of the nature of consciousness required to ground it is ruled out by “science”—but this, too, quite simply rests on extrapolations from what “science,” properly understood, actually informs us is so which are performed with the crucial aid of philosophical assumptions which we have every right to subject to philosophical criticism. Given that it cannot actually be demonstrated—the only kind of evidence that does, or could, exist for it directly is in principle subjective, and a skeptical hypothesis where for example the actions of consciousness may simply be determined by a unique and inscrutable set of non–physical laws can be made—metaphysical free will is not the place I would choose to stake the debate against physicalism.

However, in a cumulative case, this series has argued that physicalism cannot account for any aspect of what we actually are, because it cannot account for any aspect of the conscious experiences that we both exist within and infer the very existence of a physical universe through, in principle—consciousness is essentially composed of qualitative experiences through–and–through, with intentionalistic states of conceptual thought “about” that world of experiences in its center—and an extremely important fact about these states that is revealed about them from the direct data available immediately within them is that they come in diachronically unified, fluid and unbroken streams—our experiences, quite simply, flow, and our experiences themselves reveal this to us directly as an immediate piece of data about them—and it is in this self–evident experiential unity across time that our personal identities are found.

Together, these three aspects comprise everything that we are: temporally unified streams of subjective, qualitative experience engaging in representational, conceptual thought “about” the qualitative world we feel, taste, and qualitatively experience around us—and if physicalism supposes by assumption that atomistic forces lacking which have no properties other than those responsible for predisposing them towards various inert patterns of blind motion throughout space, then it defines the world in a way that renders it incapable in principle of accounting either for these qualitative nature of these experiences, the intentionalistic nature of these thoughts, or the temporal unity of the fluid stream composed of both.

Yet, the image which physicalism presents to us of the ultimate nature of the world—with the properties which physicalism attributes to the physical world and restricts the physical world to possessing are—is in fact an idea which we formulated as an intentionalistic, representational concept purely to explain certain aspects of the nature of these very qualitative experiences. Physicalism, in postulating that blind and inert physical processes are the sole bedrock ingredient making up reality, cuts itself off in principle from any capacity whatsoever to explain the existence of the very phenomena which ever caused any of us to have the intentionalistic thought to posit the conceptual idea that blind and inert physical processes lying somewhere inscrutably behind our qualitative experiences even exist in the first place. If our subjective, first–hand conscious experiences also indicate that consciousness is capable of making active choices between truly metaphysically open and real alternatives, and the notion that the world is built out of nothing other than blind and inert physical processes invalidates this possibility, then this is simply additional circumstantial evidence that physicalism ultimately eliminates everything that makes us what we are and grounds what most of us care about. In any case, the crux will follow from where you come down on these philosophical preceding questions; and once you have those settled, “science” actually adds surprisingly little—in fact, practically nothing—into the picture, in striking contrast to the extravagance of many popular claims.

 _______ ~.::[༒]::.~ _______

[1] Alfred Mele thinks that a particular, very strong kind of study the likes of which has not in fact been conducted yet could potentially rule out the possibility of metaphysical free will, by proving that decisions are in fact 100% predictable well in advance before the individual is consciously aware of his decision, and he explains in comments that this is why he asked that his publisher change the title of his 2014 book Free from “Why Science Can’t Disprove Free Will” to “Why Science Hasn’t Disproven Free Will.” There is a certain kind of truth to this, and I think Mele chose the better rhetorical route, as well. But notice: even if actions are 100% predictable from preceding causes, this still simply does not rule out the possibility that it is freely activated conscious intentions which in the end finalize the decision to act—for it could still simply be that peoples’ free decisions are easy to guess when you understand the matrix of conditions within which they are choosing between the options facing them. Put it this way: if we can predict that 100% of people who are starving and given the option to eat or starve to death are going to choose the latter, or that 100% who are given a choice between fillet mignon and rotting roadkill possum will choose the former, our ability to predict their decision still does not prove that the decision they made was not made with a metaphysical capacity to have chosen otherwise. If it turns out that we could measure the urges forming in a person’s brain before these percolate up into conscious awareness and from these predict what urges people are going to act on, this may be different in degree but not in kind from situations like those just mentioned. Impressions to the contrary result from the fact that determinism entails that events should be predictable in principle; the fall of a row of dominos is determined, and therefore it is possible in principle for me to predict in advance exactly when and how the seventh domino is going to fall after I push the first one with a given velocity and angle of physical force. But this does not mean that determinism equals predictability in principle [PIP]. PIP is true if determinism is true, but PIP is not true if and only if determinism is true—in other words, even though PIP would follow from determinism is being true, the truth of PIP would not entail that determinism is true, because determinism is not the only way to get predictability.

This is analogous to saying: If (claim A) someone broke into my home and stole my keys last night, then (claim B) this morning my keys would not be where I thought I left them on the table yesterday, and (claim B”) my keys are not where I thought I left them on the table yesterday, therefore (claim A”) someone broke into my home and stole my keys last night. The step from claim B” to claim A” is fallacious, even though the step from claim A to claim B is valid, because claim A is not the only way that claim B could be true. Claim B could also be true if, for example, I did not in fact put my keys where I thought I put them on the table yesterday, and in fact I put them somewhere else and have misremembered.

To return to the case, it could simply be that people so rarely choose to act contrary to their impulses that prediction of behavior is almost always possible by measuring the impulses gathering in formation subconsciously anyway. This might make the possibility of the existence of metaphysical free will ‘trivial’ for practical intents and purposes, yes—but it still simply would not settle the metaphysical question itself, any more than an empirical limitation to our ability to try to predict future behavior (say, because there are certain levels of activity in the brain which we just can’t accurately scientifically measure) would settle the metaphysical question in favor of the existence of free will (although it might similarly make the possibility of the truth of determinism ‘trivial’ for practical intents and purposes, as well). Empirical data can only settle the question of how predictable behavior is—but either way that that question is answered, the answer simply doesn’t settle the question of why it is or isn’t predictable, and whether any part of the true full explanation of that answer does (or could) involve metaphysical free will.

I think Mele chose the right rhetorical path, because saying that science “can’t” disprove free will might sound superficially to an uninitiated audience like a retreat in reaction to the fact that science seems to have disproven free will insofar as empirical evidence could possibly make the potential existence of free will look implausible, and Mele’s choice of title shifts emphasis onto the fact that the scientific evidence hasn’t even come close to properly doing this yet. But even in the most severe cases, the metaphysical question of free will still simply stands logically independent of anything empirical evidence is capable of revealing about the matter one way or another—even where in the most severe cases certain empirical answers could make certain metaphysical answers either “look implausible” or just seem irrelevant for the particular practical intents and purposes we happen to be most concerned about. The question may get less and less interesting as some reasons for considering it interesting would progressively vanish the closer empirical evidence came to confirming these kinds of claims, but even when we go all the way to a hypothetical extreme which evidence has come nowhere close to confirming yet, and the question becomes the least interesting it can get, this still doesn’t alter the fundamental logical independence of philosophical interpretive lenses from the empirical data so filtered. The philosophical interpretive lenses still determine how that empirical data is filtered, and still cannot be strictly determined by the empirical data itself.

[2] I don’t intend to endorse knowledge arguments as stand–alone arguments here. I explained in a recent entry why I strategically avoided framing my arguments against the physical reducibility of qualitative subjective consciousness in terms of the “zombie argument.” I think knowledge arguments face similar strategic issues—so even though I think they go through, for the most part I think they go through because I accept these other supporting arguments. In other words, I think the points argued by these arguments are sound—it’s just that, dialectically, they’re not “where it’s at.” I add this footnote because I don’t want anyone to get the impression that they’re in and of themselves my reason for thinking these points hold. However, because I endorse all the arguments I have made up to here, I think the truth of these arguments follows.

[3] In the Harris excerpt I read, a mention of the Soon studies followed the break after this paragraph. He may have been referring to the studies of Haggard and Eimer in this part which preceded the break, but in any case, Soon’s is one of the most recent modern “replications” of this kind of finding.

_______ ~.::[༒]::.~ _______

Further Reading:

Free will debates: Simple experiments are not so simple (


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

%d bloggers like this: