The issue of free will and determinism has been a source of controversy and debate in philosophy since Ancient Greek philosophers like Democritus, Epicurus and Lucretius first articulated those concepts in primitive form. But in recent decades it has become common practice for many-- though not all-- cognitive scientists, psychologists and neuroscientists to claim that Science has shown that free will does not exist, but rather is an illusion. Mainstream media outlets have often reported such claims uncritically (see for example https://www.psychologytoday... ; http://www.nytimes.com/2007... and http://www.telegraph.co.uk/... .
Generally, the scientific arguments against free will proceed from two directions: a) neuroscience and b) social psychology. Of course there are many complex and important arguments both for and against the existence of free will within philosophy. But this post is concerned with the scientific studies most frequently cited as evidence against free will, and the ways in which they have been, and can be, interpreted. Often popular accounts, or even technical ones, don't really define what is meant by free will with any rigor. In a future post I will discuss in some detail various definitions and theories, but for the purposes of this discussion, the crucial question is whether or not conscious decision-making processes are or are not involved in what we ordinarily take to be the implementation of freely chosen actions.
At the everyday level, we usually believe we have chosen freely because we have the conscious experience of, at least, seeming to do so. Scientific evidence cited by many neuroscientists purports to show that all decisions we attribute to free will are really made before we even become aware of them, thus undermining any coherent understanding of free will as conscious choice whatever.
Almost all such evidence comes either from the experiments of Benjamin Libet which relied on EMG and EEG data; or more recent Libet-style experiments using such data in conjunction with more recent technologies (e.g. fMRIs). The evidence that derives from social psychology experiments on conformity, groupthink and environmental influences generally, has focused on "situationism" -- the idea that people behave in the ways that they do because of environmental triggers and determinants rather than fixed traits or character. Social influences, such as being told to carry out actions by an authority figure (e.g. the classic Milgram experiments https://en.wikipedia.org/wi... ) elicit or determine behaviors that otherwise would not manifest. This weakens the case for freely chosen actions and behaviors. Some combine the 2 arguments (Libet's data and social psychology experiments on groupthink, conformity, etc.) in a pincer-like move meant to squeeze out any viable room for free will. The important point for now is that these arguments are said, by their proponents, to establish determinism, or determinism possibly combined with randomness (ala quantum events in the brain). Such randomness, it is argued, lacks purpose and as such does not provide grounds for the postulate of effective conscious free choice, since that depends on non-random sequences in nature that can be employed reliably by the conscious agent. That is, if I freely choose to begin typing something, the act of doing so will depend on non-random sequences of events in the brain that cause non-random behavior of motor neurons etc. At some level, freely chosen acts would seem to imply deterministic or mechanical sequences of events or else one's mental events would have unpredictable consequences rather than consequences that are fairly predictable based on the choices of agents.
This post, however, deals almost entirely with the Libet-type experiments, since they are, far and away, the most frequently cited evidence that "free will" does not exist. A subsequent post will deal with a) Daniel Wegner's experiments and his claim that free will is illusory and b) various psychological studies that argue from strong environmental determinants like the Milgram experiments mentioned above.
Libet Studies: Background:
German neuroscientists in the 60s took an interest in the relationship between brain activity and voluntary muscular activity. The experimenters were interested in the relationship between events in the brain as measured by an EEG, which monitors electrical activity in the brain (e.g. during seizures); and neuromuscular events in the limb or extremity using EMG (which is used in neurology to assess the health of muscles and the motor neurons that control them. It is the motor neurons that cause the relevant muscle contractions such as wrist flexion when one squeezes a ball or clicks a mouse.
The German team found a reliable pattern of electrical activity in the motor area of the brain which guides movement, such that the spike in activity preceded the activity of motor neurons in the muscle that contracted. The brain activity reliably clocked in at less than one second prior to the activity in the muscle, and it increased gradually over the remainder of that second. The interpretation was that the brain activity in the motor area was a necessary preparation for voluntary flexing. It was called the "readiness potential" (hereafter RP).
Libet-studies from the 80s to Present:
Neuroscientist Benjamin Libet had the idea of using these results to make inferences about free will. He asked about the relationship between the RP and conscious intentions to carry out acts like wrist flexion. To do this, he modified the experiment by telling participants to flex their wrists "whenever they felt like it." This "whenever" was construed by Libet as the "moment of a conscious decision." To be precise about the moment of the decision, he added a very fast-moving dot moving around a clock face marked in intervals like 5, 10, 15 etc. Each full rotation of the dots took 2 1/2 seconds. He asked subjects to "notice" where the dot "was" when they were first conscious of an urge to flex their wrist. So after flexing, subjects would report the location they recalled the dot being in when they first felt the urge to flex. Researchers repeated the experiment 40 times per test subject to avoid anomalies.
The RPs were recorded as starting at ~550 millisseconds prior to the muscular activity. That is more than 300 milliseconds after the beginning of the brain activity. The interpretation was that the brain begins ramping up, or preparing for the act of flexing ~550 milliseconds before flexion. Now the participants’ reported awareness of when they had the conscious urge was only 200 milliseconds before the muscle contraction--i.e. 300 millisceconds after the onset of the RP in the brain. Why did brain activity necessary to the flexing precede awareness of an urge to flex the wrist?
Some scientists hypothesized that the conscious decision must occur at some point before the RP brain activity commences. But the problem with that is that the test subjects consistently reported conscious urges to flex only after the RP. Libet's take on all this was that the brain non- consciously initiates the sequence of events that result in flexion, and only afterwards are people aware of an intention to act.
Still not ruling free will out, Libet ran a series of experiences on what is now called "executive veto" or "free won't." He now told subjects to plan to flex at a preset time (eliminating the "moment of intention") and asked them to "veto" the act of flexion at the last moment. Again, the RP developed at ~550 ms, but the brain activity flattened at somewhere between 1-200 milliseconds before the pre-planned time of action. That is, they did not flex despite the presence of the same RP that is taken to initiate the flexion in the other study. Another finding was that if the moment of exercising "veto power" was reported anytime later than 50ms prior to planned flexing activity, it did not prevent flexion. It appeared that if subjects let the RP ramp up long enough, then by 50ms prior to flexion the contraction was no longer stoppable. So, RP commences at 550 ms; remembered moment of experiencing urge to flex occurs at ~200 ms and "veto" is only effective if it occurs between 200 and 50 ms prior to flexion. There was, said Libet a 100- 150 ms window of opportunity to exercise free will (i.e. decide not to do X). This was dubbed "free won't" but it amounts to the same thing as free will on Libet's interpretation since it is a decision to do something, viz. prevent an action from occurring. Libet, then, reached the conclusion that we are not free to initiate decisions but can intervene so as to veto them. In short he thought he had figured out how free will works, not entirely eliminated it. Here is a 2 minute video from BBC summarizing the experimental set up and protocols:
Subsequent Interpretation in Neuroscience:
Neuroscientists were quick to conclude otherwise. For if the moment of being aware of a decision comes AFTER the brain has begun to prepare for the action, our sense of free will must be in error, and illusion. Free won't was de-emphasized, and the experiments were (and still are frequently) replicated using updated equipment. The detractors seized, at first, on the miniscule time-scales involved. How could we ask subjects to be accurate about things transpiring at speeds we almost never track or attend to? How do we know the perception of the dot on the clock is really accurate? But the degree of confirmation tended to take the sting out of such arguments. The results were and still are almost certainly too regular to be erroneous. However, the concept of free will has been most exhaustively analyzed by philosophers for centuries, and they were not all impressed by some of the conceptual assumptions and inferences made by the neuroscientists.
In no particular order, here are some of the criticisms and questions raised by several philosophers (most prominently Alfred Mele who wrote 2 books dedicated largely to the issue, one of them quite accessible https://global.oup.com/acad... -- see also his more technical Free Will and Moral Luck https://global.oup.com/acad... Some of the following points are not from Mele but other philosophers and neuroscientists, and some are my own. Most however derive from Mele's work.
A) Mele analyzes and teases apart concepts like intending and wanting and other motivational terms like experiencing urges to do such-and-such. We may want to do many things at the same time, for example. I want to eat, type this post, call somebody, etc., but none of these wants is an intention. Intentional decisions include such determinations as "I will type this first and then eat." We typically don't have control over what we "want" but are passive in relation to experiencing such mental states ( e.g. "I want to have a chocolate now" is not under my control; but it is possible that whether or not I eat it is in some cases a conscious choice c.f. "the marshmallow experiment https://en.wikipedia.org/wi... .) Research on emotional regulation https://en.wikipedia.org/wi... , impulse control and delayed gratification are all relevant to these issues.
B) Libet asked subjects to wait until they "felt the urge to flex" and report the time of said urge. This awareness of the urge was then construed as the moment of a conscious decision. But having an urge to flex one's wrist and having an intention to do so are not the same things. We routinely deal selectively with competing urges, just as with wants (wants and urges are closely related, and both are passive states; decisions are active states and are said to involve judgment or deliberation which surely doesn't arise in cases of waiting until you feel an urge to flex your wrist.
C) A related concern is the question of if and when the subjects actually "chose" to do anything (as opposed to responding to an urge, which is much more akin to scratching an itch than executing a decision). I think it is possible that the decision to flex your wrist is made in the first class of experiments whenever you agree to follow Libet's instructions. He tells the subjects to flex (in the free will experiments unlike the free won't experiments). So, if they ever have to make a judgment or deliberation it is when they either accept or reject his instructions. Hypothetically, a rebellious volunteer might say "Sure, I'll do what you ask" while inwardly thinking, "Like Hell I will! " If such a person decided to not flex when asked to during the instruction giving period of the experiment, would he or she end up flexing "against his own decision?" I personally doubt it, and am not aware of this being studied. Compliance with instructions is a choice. The moment of an urge to move is usually not.
D) The task of flexing is non-purposive. It seems to me that according to results obtained by neuroscientists and social psychologists (e.g. Daniel Kahnneman and Amos Tversky) even more important tasks than flexing a muscle actually are carried out non-consciously by what is sometimes called "System 1" brain activity, or "auto-pilot." I don't remember which turnstyle I used at the subway station earlier, or which door I used to enter the building I'm in at the moment. Such choices are close to being inconsequential, though they are still necessary functional choices I must make since I can't follow Yogi Berra's advice ("When you reach a fork in the road take it.") But when I click a mouse-- after 1/2 a minute or 3 minutes during an experiment, say-- makes no difference (unless I'm in quite a hurry!) This decision resembles the thought experiment (from the middle ages) of Burridan's Ass https://en.wikipedia.org/wi... Imagine a donkey posed situated exactly halfway between 2 equadistant sources of food. The meals are identical, and set out at exactly the same distance from the ass, one on the left and one on the right. There is no reason to prefer the left hand meal to the right hand meal, and so, it was argued, the animal is paralyzed by indecision with no reason to act this or that way(although neuroscience has shown that animals do not freeze in such situations, but evince non-predictable behaviors-- and that includes humans when faced with no-stakes decisions like which door to use in the lobby of a large building). Still, the point here is important, i.e. when there is no REASON to prefer or choose X over Y or vice versa, it makes little sense to say an agent is confronted with a "DECISION."
In the history of philosophy and psychology decisions occur when different potential actions entail relevantly different outcomes. Picking a door or turnstyle to walk through doesn't quite qualify, I think. These are differences that don't make a difference, if you will.
E. Mele makes a related point. He distinguishes between "Important" and "Unimportant" decisions, pointing out that we have never seen a scientific study of such important decisions as choosing to accept a new job, or to move to new place or get married. These have very different structures conceptually and temporally. They don't necessarily get made at one moment that can be clocked, but involve extended thinking and deliberating. Even if the final tipping point were (hypothetically) non-conscious, would that mean that all the conscious rational thought would have been unrelated to the non-consciously implemented outcome? There's no reason to assume so, and it seems odd to imagine so. Some advocates of free will have long held that only some kinds of decisions, and also some aspects of decision-making are under our full conscious control. Why must it be black or white? Can't conscious and unconscious processes both have a role in "important decisions?"
F. Having made these distinctions, Mele points out that "no free will" proponents extrapolate from inconsequential acts like flexing one's wrist to ALL POSSIBLE DECISIONS FORMERLY THOUGHT OF AS FREE. The extreme inductive generalization from a particular (and possibly irrelevant) case to each and every case (i.e. "There is no free will at all") is unwarranted.
G. Some philosophers have discussed the difference between cognition and meta-cognition with reference to these studies. Libet asked subjects to "remember where the timer was at the moment of feeling the desire or urge to flex" but there is an important difference between the actual experience of wanting X and the recollection that I wanted X at Time 1 or 2. Logically, the experienced desire must precede the recollection of the moment of that experienced desire. Memories, like shadows, follow events and never precede them. Might not this explain at least some of the delay between RP and moment of [recollected] choice?
H. Finally, if we are to believe, even after all of the above, that the summarized results represent "non-conscious decisions." than it seems strange indeed that all subjects have brains that have near-identical intervals of time between decision and execution. If we are taking these studies to be paradigmatic of ALL alleged decision-making , then why do most of our daily decisions have variable elipses between "time of conscious urge/intention" and voluntary actions? It may take me longer to move my arm than you when I "decide" to pick brand X instead of brand Y of some item at the store. Maybe I'll feel an urge for brand X, but nothing happens for 2 seconds (rather than picking it within 200 milliseconds). The uniformity of temporal elipsis in so many different cases at least deserves to be explained.
Summing all this up, philosophers and some psychologists and neuroscientists have interrogated the concepts and interpretations used by Libet and his many successors in the "free will vs. science" debates. Thus far, very few responses to these questions, concerns and criticisms by Mele et al. have been articulated.
Meanwhile, the whole issue has prodded philosophers and psychologists to reexamine the concept of free will. Several new theories of free will have emerged, partly shaped by these debates. Some of these theories are compatible in various differing ways with scientific determinism, others are not. I hope to discuss some of these in another post. But the next post will deal with other scientific claims to have "disproved" free will. The next post will take up the work of Daniel Wegner who hold's that free will is illusory, and claims made by social psychologists that environmental factors may determine much if not all behavior. Classic conformity studies such as the troubling Milgram experiments https://en.wikipedia.org/wi... and others will be discussed then.
Suggested links and readings:
Here are some good resources for learning more about the topics of this post:
-Ted Honderich is, perhaps, the most outspoken proponent of hard determinism. This is his website with tons of links and entries on various philosophers and scientists: http://www.informationphilo... and this is a short excerpt explaining his position taken from one of his own books: http://www.informationphilo...
-Robert Kane (1930-2017) was one of the most outspoken proponents of libertarian free will (a strong version of free will). This is his website: http://www.roberthilarykane... This book https://www.amazon.com/Cont... is his readable and learned intro to the topic:
-Alfred Mele is the "man in the middle" not pushing for radical free will as Kane does, but avoiding hard determinism. I have drawn significantly on some of his arguments and included links to some of his books. Here is his website followed by a link to a short video in which he explains his perspective. http://myweb.fsu.edu/amele/...
This is a 15 minute video interview with Alfred Mele on the question, "Does Free Will Exist?" on Big Think. If you click on his name, there are several related videos and interviews on related topics.
https://bigthink.com/videos/does-free-will-exist
________________________________________________________________
Possible Question:
Do you believe Libet-type studies have ruled out the existence of free will as conscious decision making behavior? If so, how would you meet some of Mele's objections or those of others mentioned?
No comments:
Post a Comment