WSU: Free Will and Neuroscience with Alfred Mele

what i'm going to do is talk about three different kinds of scientific experiment using three different kinds of technology that are claimed by some people to show that free will is an illusion that free will doesn't exist and um i wanted to start really by talking just a little bit about my motivation uh for writing on this topic because you know there are all kinds of things a person can write on and this is something now i've been writing on for a few years and i'm going to talk about three problems that all three of these experiments have in common in generating the conclusion that there's no free will but now motivation so actually i've i worked on theory of free will for a very long time and i also track science of action because i'm very interested in how actions are produced all kinds of actions just simple actions complicated actions and so i knew about the neuroscientific studies that way back in the 80s actually that were claimed to show that there's no free will and i thought oh you know they don't show it and i don't think people are going to take it seriously but eventually this stuff made big news and it was online a lot in newspapers and so on and just to give you a feel for it i'm going to read you uh from two relatively recent uh online articles about free will this one the first one is from science now daily news it was uh 2008. your mind might be made up before you know it researchers have found patterns of brain activity that predict people's decisions up to 10 seconds before they're aware they've made a choice the result was hard for some to stomach because it suggested that the unconscious brain calls the shots making free will and illusory afterthought the experiment the author's talking about there is is one that i'll talk about a bit later it's done with fmri i'll explain that later and here's another one this also is from 2008 this was december it's from science news and the article was called the decider here it is free will is not the defining feature of humanness modern neuroscience implies but is rather an illusion that endures only because biochemical complexity conceals the mechanisms of decision making and so stuff like this that motivated me to you know start writing a book on it but a serious book a book for scientists philosophers and uh grad students and eventually i wrote the book free that was mentioned and that's a book i tried to reach everybody with and i hope it will do it and the dialogue that was mentioned is a book for undergrads explaining the scientific critique of free will in those books i go into scientific assaults on free will from a bunch of different angles including social psychology and cognitive psychology and neuroscience but i can't do all that now so i'll just talk about the three different kinds of neuroscientific argument for no free will oh okay so when i was writing this book um i got an email out of the blue from somebody i didn't know and here i'm going to read it to you and i ended up putting this in my preface after getting her permission to do so i recently purchased a dvd by dr steven wolinski he explains from the point of view of neuroscience that there is no such thing as free will as we can only perceive an action after it has already occurred can you please help me with this i can understand that i don't know what thought will occur next but that that has already happened is beyond comprehension thank you as i am in a lot of despair now it's kind of amusing and sad at the same time and that was the sort of effect i was worried about people would read that they don't have any free will and they'd get depressed now until 2008 we didn't have any hard evidence on the effect that the news that there's no free will has on people and then we started getting some hard evidence so uh two social psychologists uh kathleen voss and jonathan schooler uh did a study in which uh the subject pool was divided into three groups there was a group that read passages saying that nobody has free will there was a group that read neutral passages and there was a group that read passages saying that people do have free will and then the next task was to take a little math quiz and they were told the cover story was uh the program is glitchy so that if they don't press the space bar right after the question shows up then the answer is just going to show up on the screen in which case they can cheat of course because the answer is right there there's also a version of this in which they get a dollar for every correct answer and it turns out that the people who read that the no free will passages cheat significantly more often than the other two groups and the other two groups behave essentially the same which by the way is evidence that free will is a kind of default assumption because telling people they have free will doesn't change their behavior at all just keeps them the same as the control group and then my friend roy baumeister who's also at florida state university did a study in which he had just two groups and one group read neutral passages that's the control group and the other group read no free will statements like you're not in control everything's in the cards stuff like that and the next task was to serve snacks to people who were about to walk into the room and they were told two things about these people they have to eat everything you put on their plate and they've all indicated that they really hate spicy food and the people who read the no free will passages doled out way more of the spicy salsa option than the other group did so they're lowering confidence and free will increases what looks like aggressive behavior and you might wonder what's going on there well what might be going on is they're thinking hmm no free will i guess i'm not responsible for anything i do i've got this urge why not go for it you know punish the person with the spicy salsa so we do have evidence that the news that there's no free will increases misbehavior now if the news is true then we have to deal with it we have to figure out what to do but if it isn't true then we can sort of shove it off to the side or at least that's what i've been trying to do for a few years now before this multi-million dollar grant i had a four and a half million dollar grant on free will in science and i i uh the money was given out in subgrants and i subgranted money to people who disbelieve in free will people who believe it and so on but we did a lot of nice neuroscientific experiments and social psych experiments and we know more than uh now than we used to okay so just a little bit of terminology so decisions and intentions because all of these studies are about well deciding to do something at a certain time or coming to have an intention at a certain time so the way to think of deciding for our purposes is as a momentary action of forming an intention to do something so i might decide now to snap my fingers now to give an example and i might decide now while i'm talking to you to find a way to take a break between this session and then my next one tonight at eight o'clock and in both cases the deciding is a little mental action needn't be conscious i'm not saying it has to be conscious of forming an intention to do something intentions you can think of however you normally do because it turns out most people think of intentions in basically the same way and then among decisions and intentions there are proximal ones there are decisions now to do things now or intentions now to do things now and there are what i call distal ones decisions now to do things later intentions now to do things later and in all of these studies it's proximal or very nearly proximal decisions and intentions that are being studied okay now pretty soon this uh my daughter is an artist she's actually a wonderful artist and i have no artistic talent at all so don't count on this to look exactly like the eeg chart i'm talking about but in my mind you know it's close but i can't really tell because i'm sort of art blind um so i'll talk about the experiment shortly this first one but i just want to give you some empirical claims some claims from the literature so now the first one i'm quoting from benjamin libit benjamin libit is the neuroscientist who got the no free will ball rolling in neuroscience and actually his position position's a little more subtle than he's usually given credit for and i'll talk a bit about that but anyway here's the quotation the brain decides to initiate or at least prepare to initiate certain actions before there is any reportable subjective awareness that such a decision has taken place so the claim is you know the brain's making these decisions unconsciously and we pick up on the decisions after they've been made the next one because brain activity in the supplementary motor area consistently preceded the conscious decision it has been argued that the brain had already unconsciously made a decision to move even before the subject became aware of it and i'll talk about that now we get a theoretical claim that links these alleged findings to free will if the act now process is initiated unconsciously then conscious free will is not doing it and then we get a claim that was libid again another claim from libit where he generalizes from uh his apparent findings in the experiment i'm about to describe to all cases of deciding here it is our overall findings do suggest some fundamental characteristics of the simpler acts that may be applicable to all consciously intended acts and even to responsibility and free will so now let me um i think i'm done with all that let me uh describe to you the original experiment that got this going so your task as a subject is to flex your wrist whenever you want and you're watching a very fast clock it makes a complete revolution in roughly two and a half seconds so it ticks around um you're hooked up for eeg readings so we're measuring electrical conductivity on the scalp and uh also muscle motion and when subjects are regularly reminded to be spontaneous that is not think at all in advance about when to flex um and you always do at least 40 trials because you need a whole bunch of data so you can do some averaging of the statistic of the results the eeg ramp up starts about 550 milliseconds before the wrist muscle moves but the average time of first reported awareness of the decision or urge or intention different terms are used is about 200 milliseconds before the muscle moves this they call w time so the idea is after you flex your wrist usually you wait for the clock to stop moving and then the way you do it now anyway is you move a cursor to the spot on the clock where you think it was when you were first aware of your urge decision intention or whatever so the average time there is 200 milliseconds the eeg ramp up starts here at 550 milliseconds a little more than half a second before the muscle burst and so back to those quotations now what libit is saying is well look what's going on is that the brain is deciding here and the person doesn't become aware of the decision for another third of a second roughly and then we get the generalization to all decisions so all of your decisions you make unconsciously and then later on you know after you make them a bit later third of a second you become aware of them okay now there's another aspect to libit study that i want to talk about before i get on to the next study so libbet thought that there was a tiny window of opportunity for something like free will he thought that free will could never initiate actions but what it could do is veto an intention to do a thing or veto a decision to do a thing it could kill it cancel it um so now i can add a little bit to my chart so what libit thought then is that i should probably stand here because okay what libit thought is that once you became conscious of your intention to flex now or your decision to flex now you had about a 100 millisecond window of opportunity to veto it and he thought he had two different kinds of evidence for that one kind is that after the experiment quite a few subjects actually said you know sometimes i had the urge to flex right then but i just didn't act on it i let it go and i waited for the next urge to come along so there's that kind of anecdotal evidence for veto power they felt the urge but they didn't act on it but another thing libit did then was to do a veto study he was trying to find evidence for veto power and in the study what the subjects were supposed to do their instructions were these prepare to flex when the spot hits a certain point on the clock like the nine o'clock point say so prepare to flex when the spot hits there on the clock but don't do it and when he did that experiment he got an eeg ramp up starting about one second before the designated time like if it was the nine o'clock point one second before that uh it started ramping up there and then remember the but don't do it part and it started petering out 250 to 150 milliseconds before the designated time and then what libit said is uh well look this is evidence of veto power because here these subjects intended to flex when the spot on the clock at the nine o'clock point uh and then they canceled it now there's a kind of flaw in that experiment that i could go on and on and try to explain in a technical way but it's really easy to see if we just do our own experiment so a free one so we're gonna do it now and i'll ask you to cooperate usually i can see the audience pretty well but the lights are a little bright so i can't see all of you so what i want you to do i'm gonna count one two three and i want you to prepare to snap your fingers like this when i get to three but don't do it okay so you prepare to do it but don't do it and i sort of like you to pay attention to what's going on in your head as i'm counting to three can you do that okay so let's try it so you remember now so you prepare to do it but you don't do it okay one two three okay good now question how many of you intended to snap your fingers when i got to three answer nobody and how do you know that well you know it's sort of introspectively you can also infer it maybe from your behavior what did you intend to do well you intended to prepare to flex when i got to three i mean snap when i got to three but not to do it okay so what libit is measuring here isn't the power to veto an intention you had to snap your fingers when i said three or an intention his subjects had uh to raise a wrist when it got to the nine o'clock point what he's measuring is something else and then what you might wonder is what is he measuring you know what's going on back then well it could be preparation or it could be thinking about doing this thing at a certain time and i'll come back to that and actually you can do other kinds of studies where people are just thinking about doing this thing you know they're not going to do it where they're preparing it or where they're watching somebody else do it you get similar charts the 2008 study that was mentioned in that first quotation from science news that i read you this was done with fmri so what you're reading are changes in blood flow in the brain it's a slower kind of process than the eeg because it takes blood oxygen about three to four seconds or so to get to where it was most being used in this task so while doing the fmri subjects had the task of pressing either this button with this finger or that button with that finger and it was up to them what to do and uh the scientists in the paper claimed that they could predict with high accuracy the high accuracy part is a quotation uh seven to ten seconds in a in advance of a button press which button the person would press next and that's the i'm going to read that passage to you again so this is the first one i read your mind might be made up before you know it researchers have found patterns of brain activity that predict people's decisions up to 10 seconds before they're aware they've made a choice so the thought is look the brain has already made that choice you know seven to ten seconds in advance and you don't become aware of it for quite a long time the result was hard to stomach because it suggested that the unconscious brain calls the shots now here's the connection to free will making free will and illusory afterthought okay so we'll see a problem with that experiment in a minute here's a study done by itzhak freed and co-authors really interesting study so sometimes when epilepsy patients can't be well treated with drugs they opt for surgery and part of the skull is removed and then if they consent they can be involved in science experiments where an electrode grid is placed directly on the brain or electrodes are put inserted directly into the brain and then they can do various tasks with my big questions in free will grant we funded one study like this it wasn't the freed study it was a very interesting study and by the way there's on pbs a series called closer to truth and there's an eight episode series on closer to truth and in one of the episodes uh you can see our eeg study i mean our depth electrode study it's really cool um okay so now the task here is just to click a key on a keyboard whenever you want that's that's the task and they were taking readings directly from the brain and 800 milliseconds in advance of a key press they could predict within a window of about 300 milliseconds so it's not pinpointing it but they could predict with 80 accuracy when the person would press the key and what they were also measuring was when the person would say the person became aware of the intention and that was about a hundred milliseconds before muscle motion began and 200 milliseconds before the key made contact with the switch under it okay that's 80 accuracy that's that's pretty good all right so i said i would describe uh three different experiments you know using three different sets of technology and then i would try to explain why we can't really justify the conclusion that there's no free will based on these findings and i'm going to get to that part now um always at the end when i give a talk like this somebody will ask me what does free will mean anyway and a nice thing about this talk is we don't have to get so much into the semantics of it because we can see that there are problems just getting to the claim that decisions are made unconsciously in these experiments and that's the key claim that's how they get to the conclusion that there's no free will well look you're not making your decisions consciously and you need to make them consciously in order to be acting freely okay so now there are three different problems i'm going to talk about and each one is a problem for each of the three experiments the first one is so back to libit what happens at minus 550 milliseconds in libit study that's where the eeg ramp up starts and you so you should be wondering well is that where the decision is made that's libit's claim and it's it's an important claim to get his argument to go well one question you'd want to ask in that connection then is how long does it take a decision to do something now or an intention to do something now to generate muscle motion does it take 550 milliseconds might it take less time for example now we do have already and we've had for years indirect evidence about this so you can do reaction time studies in which a subject knows what he or she is supposed to do when the subject gets a go signal so it might be when you hear a beep then what we want you to do as quickly as possible is flex your wrist now to get data that would be even more useful in answering our question about libit what you'd want is a reaction time experiment where subjects are watching a clock in fact a limit clock a fast clock like this because watching the clock is going to divide attention and it's going to slow down reaction time and now there is a study like this it was done by patrick haggard and elena magno it was published in 1999 as i recall and the mean time between the sounding of the ghost signal and the beginning of muscle motions called the muscle burst was 231 milliseconds so the time between uh intention and muscle motion is going to be less than 231 milliseconds because it takes a little time to detect the beep and then if the way it works is you detect the beep that generates the intention that then generates the muscle motion well the intention is closer to the muscle motion than 231 milliseconds that's one study here's another one this is a recent one it's on my handout it was done by trevina in miller 2010 it's a study i had suggested back in my 2009 book uh effective intentions and the idea is this so have a study where people have two options they can press this button or that button and set it up so that you tell them don't make your decision until you hear the beep and the beep instead of being a specific ghost signal like press that button or press that button is to be treated as a decide signal so we say when you hear the beep decide which button to press and then go ahead and press it right away and what happened in that study is that there was no difference in pre-tone eeg so the eeg before the decide signal was the same no matter whether the person then went left or then went right which is evidence that no decision had been made prior to the decide signal and the mean time between the side signal sounding and muscle motion beginning was 150 milliseconds so if the decision is made there it's really close to the beginning of muscle motion it's in here somewhere around 200 milliseconds rather than way back here but then if the decisions are being made around there um they're being made during w time you know the time at which the person is conscious of making them okay all right so now back to the soon at all study the study done with fmri where they said that with high accuracy they could predict seven to ten seconds in advance which button the person would press next well the high accuracy was sixty percent now that's significant that's ten points above chance but it really is only sixty percent and just to illustrate what that amounts to notice this if i had a fair coin and you had two buttons you could press and you're going to press one or the other next time i could predict with 50 percent accuracy just by flipping my coin right which one you're going to press next and they're doing 10 points better than that um so what are they detecting mine they are detecting something they're 10 points better than chance they're not just guessing what are they detecting my hunch is they're detecting a certain kind of bias so in this study what you're supposed to do is press this button next or that button next you do it a whole bunch of times you're not supposed to fall into any patterns so you have to keep track in a rough and ready kind of way and i think what's going on is that as you go through the task you will build up in yourself for the next press a certain bias or inclination that you're not even aware of to go left next time or right and maybe it's one that just increases the probability that you'll go that way but not by much and that's why we get the 60 accuracy rate but notice that if it's that kind of thing that's going on it's it's no problem at all for free will because it's not as though your brain has definitely decided back then all it's done is given you a little nudge in a certain direction a nudge that you could go against apparently okay so and then the freed study with depth electrodes well there the response the claim is going to be similar their accuracy rating is way higher it's 80 percent which is impressive but the 80 figure doesn't indicate that a decision has already been made at let's say minus 800 milliseconds and it's utterly consistent with the hypothesis that what's going on there is you're getting closer to making a decision to press now but you're not all the way there yet you're getting closer in a way that doesn't ensure that you'll go all the way and make your decision at that point in time uh and i i think that that's quite likely now i've made all these uh points before and sometimes people will respond well even if you're right let's go back to libit that's sort of the simplest one to see because we have our picture even if you're right that the decision isn't made here and even if it's true that the decision is made around here around the time you are you say you're conscious of making it they say well the point of no return has been hit back here and the point of no return is a point such that once you get to it you can't stop the process in any natural way okay so they could claim that and so they could say well you know this consciousness of a decision isn't really doing any work because everything's already been settled back then even if a decision hasn't been made back then um all right now see this is where it'd be good for you to have the handout because there's a quotation on it from libits method it's about limit libit's method and doing the study i'm going to read it to you i'm going to read it slowly because it's crucial in the absence of the muscle's electrical signal when being activated there was no trigger to initiate the computer's recording of any preceding brain activity now here's the significance of that if you want to know whether the point of no return is hit back here what you'd want to be able to do is inspect brain activity oh starting from there and let's say halfway to zero or so and see whether you can ever find any cases where you get a signal that looks like this right this part of the signal um but you get no muscle motion at zero and what he's saying in that passage that i quoted to you is that he's using the muscle motion itself as the signal to the computer to make a record of the preceding second or so of brain activity so if you're using the muscle motion to signal the computer to make a record of that brain activity then you don't have any record of any brain activity that happens when a little bit later there's no muscle motion right so he hasn't been able to look for the point of no return and people have worked on ways of you know trying to solve that problem but uh here's just one more thing about that this will make it uh really clear i hope so when you do the eeg stuff you do something called back averaging and um what happens is you need then something to signal the computer to let's say save that preceding a second of brain activity so you can match it with others and come out with a result and his signal in the main study was the muscle motion now in that veto study that i talked about so there was a signal and it wasn't muscle motion because there was no muscle motion right that was like the snapping the finger thing where you don't actually do it the signal was the clock hitting the nine o'clock point okay so you always need a signal when you're doing eeg um okay so we don't have good evidence that the point of no return is hit here in fact we don't have any evidence because of the way the experiment was conducted and libid himself wasn't claiming that the point of no return was hit there because he thought we had veto power so he wasn't looking for a point of no return and we don't have good evidence that the point of no return is hit seven to ten seconds in advance in the fmri study or eight tenths of a second in advance in the depth electrode study the first problem was then i'm just to repeat we don't have good evidence that decisions are made early at half a second before the muscle burst 7 to 10 seconds 800 milliseconds we don't have good evidence for that we don't have good evidence for the claim that the muscle burst i mean that the point of no return is hit that early and now we're going to go on to the third problem now here's the third one this is sort of a counter factual or hypothetical issue i put it this way so even if it were true that all proximal decisions are made unconsciously in these experiments would we be entitled to generalize to all proximal decisions or all decisions proximal or otherwise and remember the proximal decisions or decisions you make now to do a thing now so what you could do then i'm just to simplify it is suppose for the sake of argument that these experiments really had shown that in them all the decisions to do things right now were made unconsciously just suppose it and remember the generalization claim i read to you at the beginning because libit wants to get from his results to all cases of deciding now here i think it helps for me to describe my own experience as a subject in a libit style experiment so years ago it was 2005. i gave a lecture on free will in neuroscience at the national institutes of health to a very smart group of neuroscientists who work on muscle motion and action production and the deal was i give my lecture we do the questions and answer then i'm a subject in their experiment and then after i do the experiment they take me out to dinner okay so that was a plan sounded like a good plan and i wanted to see what it felt like to be a subject in that experiment and my plan was to be a very naive subject so i went in there without any you know preconceived plan i thought what i would do is i'd wait for urges to flex now to pop up in consciousness and then as soon as they did i thought i'd flex and i'd be watching the clock in the meantime so that after i flexed i could move the cursor and say where the spot was on it when i first became aware of my urge and what i noticed is that nothing was happening that is nothing was popping up in my head no urges to do it no intentions to do it no decisions just some questions and my questions were how do people do this how do they do the task and uh how am i going to do it what's my plan going to be because frankly by that time i was getting a bit hungry and i knew i had to do it before i could go out to dinner and i also didn't want to look like an uncooperative jerk so my plan was to say now to myself silently and treat that as a decision marker you could say treat it as a ghost signal and then flex as soon as i could in response to the silent now saying now why was it so hard for me to have these urges or whatever it's because what i was doing was arbitrarily picking a moment at which to begin flexing it's like when you go to the supermarket and you have your shopping list and maybe if you're like i like peanuts a lot so you have a 16 ounce jar of planters peanuts on your shopping list and you get to the peanut display and you just pick one you don't think about which one is better than which because you know they're all the same right um so that's what this is like it's arbitrary picking so bearden's ass i want to work bearden's ass in here because last time with the handout when the audience had the handout i forgot bearden's ass and i made a mental note never to forget it again and now i knew it was coming up and i'm not going to forget it um okay so bearden's ass it's a it's a famous old parable there's a donkey that is hyper rational so it will never do a thing unless it has a better reason to do it than anything else okay that's bearden's ass and one day it's walking around and it's hungry and it finds itself equidistant between two equivalent bales of hay they're just as attractive they're just as large and so on and they're each just as far from the donkey as the other one so it's hyper rational and so what it does is it starves to death because it doesn't have a better reason to go this way than that way but we're not built like that fortunately because you know be hard to survive what we do is we make random choices and there's no need for any conscious processing to make those random choices you just pick okay now remember that in libit study and all these studies subjects are told to be spontaneous so they're not supposed to think in advance about when to flex or which button to click or anything of the sort so how much like a lot of ordinary real decisions that we make are these decisions that are really just arbitrary choices and so just think about some tough decision you you had to make for different age groups i give different examples but i can't see too well because the lights are bright so let's do divorce let's do that one it looks like some of us are old enough for that and so um you uh you think about it day after day geez you know should i do it uh what about the kids oh now they're maybe old enough where it could be okay and you know there's all this conscious processing and then uh eventually in the end you make a decision or you might have a nice job but you get a really nice job offer and then you think geez i could leave all my friends here and take this other offer or and you might even i mean if you're like me and when you get to a case like that you make a list of pros and cons and you cross off things that are equivalent and um so this is nothing like uh arbitrary picking but if that's right then you're not in a position to generalize from alleged findings about arbitrary decisions to all decisions also where conscious processing is preceding the decision uh the processing itself will raise the probability that the decision is conscious so the more conscious thought goes into a decision the more likely that the decision will be conscious whereas in these experiments uh consciousness is shoved off to the side um i made a little note to say something about the free will literature and i think i should so there's there's a view called uh restrictivism in the philosophical free will literature and according to this view you exercise free will only in cases in which you are torn between competing options often the options might be what you think you ought morally to do what your moral duty is and what you're tempted to do but sometimes the options could be long-term self-interest and sort of short-term temptation and the restrictivists these are theorists now they're philosophers say that's the only place where free will can be at work now if that's right the arbitrary picking studies are not about free will at all because they're not about being torn between competing options they're about something else random picking but even if the restrictivists are wrong we should see that there is a huge difference between randomly deciding and then deciding after payne's taking conscious reasoning and given that there is a huge difference we're in no position to generalize from what we find about random picking to all these other cases of decision and if that's right then all those decisions uh the free will with respect to them isn't challenged isn't threatened

Channel: World Science U

Views: 936

Rating: 4.5384617 out of 5

Keywords: Alfred Mele, Philosophy, Free Will Project, Do we have free will?, free will, Aristotle, philosophy of mind, determinism, free will is an illusion, incompatibilism, Determinism vs Free Will, Crash Course, Does Free Will Exist?, World Science U, WSU, Brian Greene, World, Science, festival, New York City, 2020

Id: IQnq9goFJF8

Length: 39min 37sec (2377 seconds)

Published: Fri Sep 11 2020