Today I interview Marcel Brass, one of the leading researchers in the brain science of free will and intentional action.
Download CPBD episode 085 with Marcel Brass. Total time is 51:18.
Marcel Brass links:
- Marcel Brass at Ghent University
Links for things we discussed:
- Broca’s area
- positron emission tomography, transcranial magnetic stimulation, single-cell recording, electroencephalography, magnetoencephalography
- experimental psychology, cognitive neuroscience, neuroscience
- neuroscience of free will
- Benjamin Libet
- Matsuhashi & Hallett, “The timing of the conscious intention to move” (2008)
- Supplementary motor area (SMA)
- Banks & Isham, “We infer rather than perceive the moment we decided to act” (2009)
- Kuhn & Brass, “Retrospective construction of the judgment of free choice” (2009)
- Brass, “Unconscious determinants of free decisions in the human brain” (2008)
- Two experiments showing a connection between fatalistic beliefs and propensity to cheat
- Saul Smilansky
LUKE: Dr. Marcel Brass is a professor of experimental psychology at Ghent University in Belgium and he also works with the Center for Cognitive Neuroscience at Ghent University. He is the author of numerous publications in psychology, neuroscience and philosophy.
Marcel, welcome to the show!
MARCEL: Yeah, thanks for having me.
LUKE: Marcel, I want to ask you about recent work in the neuroscience of free will but first I’m hoping you could explain to us more generally, how much we currently know about the brain?
It seems like we know a lot about how individual neurons communicate and we know a few things about which areas of the brain are usually associated with certain functions, but we seem to know very little about the broad logical structure of the brain and how it does what it does. Is that right?
MARCEL: Yes, I absolutely agree. We have learned a lot about brain function during the last 15 years since this new brain imaging techniques such as functional MRI developed.
But we are also faced with a number of problems that we didn’t expect when we started to investigate the brain with such brain imaging techniques.
One major problem is that this localization approach – that means attaching specific brain functions to specific regions of the brain – turned out to be much more difficult than we expected, and this has a number of reasons.
First it turned out that, I mean there only are a restricted number of brain regions but we have in principle an infinite number of cognitive operations. So what happens is a lot of cognitive functions have been attached to specific brain areas.
For example, Broca’s area is a region in the frontal lateral cortex, so in the anterior part of the brain, and language people would say that this is an area that’s involved in language production while researchers doing action research would say it’s involved in imitation. And there are maybe twenty or thirty other functions that have been related to this brain area.
So now the crucial question is: what is the real function of this region?
And I think the problem occurs because traditionally from experimental psychology, we are determined by specific disciplines. So for example, when I’m a memory researcher or a language researcher I would interpret my results in terms of language. But when I’m an action researcher I interpret my results in terms of motor control for example.
At the moment we have problems to integrate these different perspectives. But this is absolutely necessary in order to get a broader view of cognitive operations that are implemented in specific brain areas. And this is one major problem at the moment.
Another problem is that the cognitive role of specific brain areas is not only determined by the specific brain area but usually the cognitive function in a specific, broader cognitive operation is determined by networks of brain regions.
So to come back to my example, Broca’s Area, together with other language areas might be involved in language production but together with other motor areas might be involved in imitation.
So we have to look at networks of brain activity rather than on single brain areas. And where we have started to do that, this is a very complicated issue in a way and we are only starting to understand the networks that are involved in specific tasks.
I think the third problem is that we don’t really have an understanding of the information flow in the brain. So the problem with the most commonly used technique in cognitive neuroscience memory function, magnetic imaging – resonance imaging – is that it doesn’t have a very good temporal resolution.
So we can’t really see online how brain activation interacts or changes across brain regions and therefore we don’t really have a good understanding of the information flow within the brain.
In some domains we have, in the visual domain I think there is a very good understanding how that works, also from monkey research.
But in more complex domains like research on cognitive control or intentional action, we don’t really have a good understanding how informations are transferred between different brain areas. And I think that’s the third major problem we are facing.
LUKE: So Marcel, you just spoke about a lot of the big-scale results that are coming out of using fMRI to look at the activity between different brain regions. It seems like we’re also making some inroads into determining the structure of the brain by looking at the very small level, at individual neurons and how they’re talking to each other.
I think, for example, some of our most developed studies there regard how we perceive visual stimuli. Is that right? Could you talk a little bit about the progress that we’re making in that front?
MARCEL: Yeah, of course. I mean, there are different levels to look at brain function. And I’m a psychologist, therefore my expertise lies primarily in the systems level, to look at the role of larger brain regions that are a few centimeters large in cognitive function.
But of course you can also look at single-cell activity. So for example, you can show a monkey an object, and then record the activity of a single neuron during this experiment.
And this also gives you information about the neural processing that takes place in specific brain areas, but then on the single-cell level. And in a way, this is a very fruitful approach, because when you look at the single-cell level, you measure directly the neural activity- the electrical activity – of the neuron.
So you have a very good idea about the temporal resolution. And also you can study specifically which kind of information encoded in this single cell.
But on the other hand, you have a broader perspective. So you have in the brain regions millions of cells that are firing in synchrony. So the problem is always to bridge the gap between this single-cell level and the systems level; the level of brain areas.
Yeah, it’s interesting. I think to some degree these informations are complementary, but sometimes they also lead to completely different conclusions.
So the problem is really to integrate these different levels. And if you ask a single-cell person, so a person who is doing single-cell recordings in monkeys, about a specific brain area, what this brain area is doing, you might end up with a completely different answer than if you ask somebody doing functional brain imaging, who looks more at the larger-scale activity of the brain.
That is certainly a challenge in the future, to integrate these different levels of investigation.
LUKE: So, Marcel, it seems that neuroscience in particular is very driven by the tools that we have available to us. And I think most people are familiar with fMRI, but what are some of the other tools that we have available, and what tools are we trying to develop?
MARCEL: First of all, in my opinion, if you are interested in cognitive function, the most important tool, so to say, is the experiment. So Experimental Psychology; the experimental method.
Because if you don’t design good experiments, you can’t come up with good answers on the question, which brain areas, for example, are involved in specific cognitive operations. But of course, we have different dependent measures.
So the most traditional measure was reaction time. You show participants, for example, specific stimuli, and then you measure how fast they react. And there is a long history of research with reaction time, more than a hundred years. So this is a very interesting tool, still.
Even though we have now these new fancy brain imaging techniques. From a brain perspective, we have these new brain imaging techniques such as Functional Magnetic Resonance Imaging and Positron Emission Tomography, which measure not really the neural activity directly, but where the blood flows.
So we have a very indirect measure of the neural activity. That is the reason why these brain measures don’t have a very good temporal resolution. But they have a relatively good spatial resolution. Then we have also intervention techniques, for example, Transcranial Magnetic Stimulation.
In Transcranial Magnetic Stimulation, you induce a very strong magnetic pulse over a specific brain region. And this disturbs the operation of this brain region for a relatively short period in time, so you can see what happens if you – in a way – knock out the specific brain area. And this is very crucial because most of the other techniques we are using are correlational techniques.
So we find activation, for example, in a specific visual area when we show a specific stimulus. When we show a face, we find activation in Fusiform face area. But the question is, is this activation really necessary to perceive a face? And this is difficult to answer from this correlational methods, such as a functional MRI.
In order to draw this conclusion, we need these intervention methods, where you knock out a specific brain region and then can show that participants can’t perceive a face anymore, for example.
Then we have older techniques to measure brain activity. One is single-cell recording, that is already relatively old. But the problem is that you can’t do it in humans, or only in exceptional cases, in epileptic patients, where you place electrode grids for diagnostic purposes.
But in principle you, of course, can’t do single-cell recordings in humans because you have to place an electrode in your brain, so you have to open the skull.
Then there are methods that allow to measure electrical activity from outside. This is EEG – electroencephalogram. And this is also relatively an old method that has a relatively long tradition. And we are still using these methods.
And then we have methods that work similar, like MEG, where you don’t look at the electrical field, but the magnetic field, so the spatial resolution of this method is much better than with EEG. Now that’s in principle… these are in principle the most common tools we use in cognitive neuroscience at the moment.
They all have their advantages and disadvantages, that is crucial to know. There is not one method that solves all the problems.
LUKE: And what’s the difference between experimental psychology, neuroscience, cognitive science, cognitive neuroscience?
MARCEL: So experimental psychology is, in principle, a discipline that is relatively old. It is about 100 years old. And it started primarily with reaction time research. So you present specific stimuli and then you record how long it takes participants to respond to the stimuli. And in this discipline, the experimental method develops to investigate cognitive operations. And we got quite far with that.
But more recently these new methods were developed, such as EEG, fMRI, and PET, with which you record electrical activity.
So in principle, you have a different dependent measure. You use the same kind of experiments, but you measure something different than reaction time. You measure neural activity. And this signal, of course, is very rich and gives us a lot of information.
So cognitive neuroscience is about investigating cognitive operation with brain activity as the dependent measure. That’s at least how I would understand it.
And neuroscience is even a broader discipline. Neuroscience, in general, is interested in the functioning of the brain on different levels. On the molecular level, on the single-cell level, on the systems level.
So it includes cognitive neuroscience, but it’s interested in the understanding of the brain, while cognitive neuroscience is primarily interested in understanding the cognitive system, using brain activity as a measure.
LUKE: Well, Marcel, thank you very much. That’s a wonderful introduction to some recent study of the brain.
I’d love to talk to you now about free will. For centuries, philosophers have argued about whether or not we have what I call contra-causal free will, which is that kind of free will that would allow us to be unmoved movers of our own actions, at least some of the time, for some actions.
But recently, it looks like science, and neuroscience in particular, might be able to answer this question for us. My understanding is that Benjamin Libet really good things going here, is that right?
MARCEL: Yeah. I absolutely agree. First of all, I do not really think that neuroscience will solve the old philosophical problem whether free will exists or not. And I’m not even sure whether any science will solve this problem.
But neuroscience can inform us about the basis of our experience of free will. So our subjective experience of will and how it is related to our behavior. And certainly the experiment of Benjamin Libet was very influencing in getting this research going.
So first of all, it might be good to briefly describe the experiment that Benjamin Libet carried out. So in this experiment, participants have to carry out simple actions. They have to press a key or release a button. And they can do that whenever they want. So not completely, but they have a specific period of time where they can decide to press a key.
While they are doing that, they are watching a clock hand rotating. And after each trial, they tell the experimenter on this clock when exactly they decided to press a key. So their subjective experience of will as Libet called it, the “W judgment,” – the will judgment.
Now you can first relate this W judgment to the response. So you know when participants actually press a key. And then you can look at this subjective experience of will and it turned out that about two hundred milliseconds before participants press a key, they say that they experience the intention to press a key.
Now the ingenious idea of Libet was to relate this subjective experience of will to neural activity. So he used EEG; electrical activity to in a way see where the brain already starts to process information before we decide to act in a way.
Benjamin Libet measured EEG potential in a mortor potential that is called Bereitschaftspotential, or readiness potential, and this Bereitschaftspotential precedes voluntary action. So if you press a key and you average back the activity in the EEG you find this Bereitschaftspotential.
Now the crucial question was whether the Bereitschaftspotential started to rise before participants said that they intended to press a key or afterwards. And a somewhat surprising finding was the Bereitschaftspotential starts about a second before participants indicate that they are going to press a key. So this seems to indicate that our decision to act is preceded by brain activity for more than a second or so.
Benjamin Libet concluded from this that in principle, our intentional decision to act doesn’t get this thing going in a way, but rather is a consequence of unconscious brain… unconscious processes that are reflected in this brain activity.
So our subjective experience of will is not what is initiating the whole thing, but is a consequence of something that is already going on in the brain.
By relating the subjective experience of willing something to brain activity, he tried to draw conclusions about the question: how free we can come up with our decision or to what degree we are determined by brain activity in our decisions.
LUKE: Now, some people criticize that conclusion that Libet made. How did those criticisms go?
MARCEL: So, first of all, there is a lot of methodological criticism about the study.
One problem of the Libet experiment is that it relies on recall. So participants do not directly tell you when they have an intention. But they tell it afterwards. So they have to recall the position of the clock, in a way.
The second problem is that this Libet clock, or it in principle a Vunt clock, it wasn’t already introduced in the beginning of the 20th century. But this Libet clock introduced some biases. So it is very difficult to determine the exact position of the moving stimulus.
And we know that from other research on perception of moving objects, for example. So there is another bias related to that.
And finally, introducing this report of intention, might change the intention itself in a way. So because you have to report on your intention, the way you formed your intention might be change. So that’s another methodological critique.
But then they are more general criticisms of the method. So one crucial question is: Whether the readiness potential – this brainwave – is always followed by an action. And this is difficult to figure out because you back average this brainwave from the action. So it’s difficult to observe when you don’t have an action.
But in principle, it could well be that you sometimes, show a gratuitous potential but then you don’t act. But if this would be the case, then it could be possible that they always readiness potentials, but we can nevertheless decide on whether this leads to an action or not.
MARCEL: So that would also completely change the conclusions you can draw from this experiment.
Finally, there is of course a very general or more philosophical question that is crucial namely: the Libet experiment is based on the idea that willful influences on our behavior are in a way conscious; so the whole experiment is about it becoming aware of the intention to act.
But if you assume that the will can also operate on unconscious processes or that you can have pre-conscious intentions also, that would also of course change the whole idea because then discovery judgment wouldn’t tell us how do you think about our volition or our will.
LUKE: Very interesting. Now there’s a more recent study that maybe an improvement on Libet’s work by Masao Matsuhashi and Mark Hallett.
Could you explain that study, and what it’s implication might be for our understanding of intentional action?
MARCEL: I think it’s a very interesting and clever study that they did. So as I said, one criticism of the Libet’s experiment is that you have to read this clock, and afterwards report on the position of the clock when you form your intention.
So they try to overcome this methodological critique by using a completely different method to determine the intention to act. So the way that they did this was, the participants had to carry all the Libet kind of experiment. So they had to press the key, whenever they wanted to press the key.
But during their experiment, they played back tones at specific points. And when participants heard the tone, and were forming the intentions at that point, they should stop the action. But if they heard the tone at any other point in time, so when they haven’t formed any intention to act, they should simply contend.
Now, this also looks at the distribution of the tones to which participants reacted. And from this distribution, you can in principle infer indirectly the time when participants form the intention to act because when they form the intention to act they would not respond. So this trial wouldn’t be in the distribution.
So you can indirectly infer from this shape of the distribution when participants form the intentions to act, because in this situation they wouldn’t act, and this wouldn’t occur in this distribution.
So with this indirect method, that has the advantage that it is not based on recall because participants immediately report indirectly by not acting that they had the intention to act. They showed that the intention to act already started more than a second before participants actually responded. So this is a few hundred milliseconds earlier that reported by the Libet method.
Interestingly, nevertheless this intention to act started after the onset of the readiness potential. So in this sense, results of this study are very similar to the Libet study experiment.
Even though one has to say that in some subject, the intention started before the readiness potential but in most of the subjects it started after the readiness potential.
So what this study in a way shows is that the intention to act seems to start much earlier than has been demonstrated by Libet. That is a little bit the problem of the whole method to use subjective experience to determine when participants intend to act because it depends firmly on the method.
So you could imagine the comparison between this experiment and the Libet experiment in the following way.
Let’s assume you were sitting in a conference and somebody asked you whether you want a coffee or not. The likelihood that you say, “Yes I want a coffee,” is presumably much higher than if you would have decided yourself to drink coffee. So there might be an intention to drink coffee, but you wouldn’t have taken a coffee at that point yet.
But if somebody ask you, you say, “Oh yes, I want a coffee.” So it depends a little bit on the degree or the strength of the intention when you can detect it in a way and the method how you detect it.
LUKE: Now some other studies have tried to get at this issue using that tool that you described, Transcranial Magnetic Stimulation. How does that work?
MARCEL: I think you are referring to… I hope you are referring to a study by Hi Kwong Lau and what he was doing is again a very clever experiment.
So we know that our experience of intending to act is related to a specific brain area and he himself has done a study where he showed that. But it’s also known from other experimental studies on intentional action that a specific part of the brain pre-SMA or more precisely pre-SMA-SMA complex seem to be involved in this intention to act.
He was interested in the question, what happens if you disturb the functioning of this brain region? So the question was whether this would change the experience of will. And it does. So it changes the onset of the experience of will, so the W judgment.
That is in itself not extremely surprising. Of course if there is a brain region that is involved in the formation of the intention and you knock out this brain region of course, it changes the way you experience your intention.
What was surprising though, was that it also changes the time of your intention when you stimulate after participants executed the action. So he stimulated after participants pressed a key and nevertheless you found a shift of the W judgment. And this seemed to suggest that brain processes happening after you already executed the action change the way you report your intention to act.
This, of course, is not very comfortable with the view that in principle the reported time of intention is a very trustworthy indicator of your intention because I mean how can the intention be influenced by something that happens after you act?
There is now some evidence, also behavioral evidence, that this is a case that in a way your experience of will is a reconstruction to some degree at least, of information that happened after you already acted. So it’s not only influenced by what you are thinking at that point but also what happens afterwards.
And there is a nice behavioral study that has shown this point as well. They used also Libet’s kind of experiment but after participants press the key they played back a tone. So the key press had an auditory consequence. You decide to press a key and then you hear a tone.
Now they will vary the interval of the tone after the key press and what they found was that depending on the interval of the tone the W judgment changed.
So your reported intention of will changes depending on what happens after the action. And again, this seemed to indicate that this reported intention to act is not always a good indicator of your intention, but it seemed to be also influenced by information integration after you already acted – so by a kind of reconstruction of your intention.
LUKE: Now, in their Transcranial Magnetic Stimulation study, in what way did that interruption of that brain region, how did that actually change the experience of intending?
MARCEL: That’s a very interesting question and I think it’s very difficult to answer what exactly is happening, why it shifted the W judgment.
I think what we can say at the moment, and we also did an EEG study with a second experiment, I told you, this experiment by Banks & Isham. And what we think is that our experience of intending to act is presumably related to the intention, but it is also related to the processing of information after you have acted.
So it’s a kind of integration of information over a longer period of time. And this integration of information might be changed when you knock out briefly this brain area that has been related to the formation of the intention.
How exactly that works is, of course, a very difficult question. The crucial point in the Lau’s experiment is that stimulation after the action changes your experience, or at least the recorded time.
LUKE: Right. Now, Marcel, you’ve done some research on this issue as well. One of your studies was conducted with Simone Kuehn? Did I pronounce that correctly?
MARCEL: Yes. Simone Kuehn.
LUKE: How did that study work?
MARCEL: The logic of the study is a little bit similar to the studies I told you about. These studies that show that information that happened after you responded seemed to influence your experience of intention.
In this experiment we asked participants to respond to a letter by pressing a key. And they should do that as fast as possible. So participants are really in a mode where they press as fast as possible when they see a specific letter.
Now, in some trials, the letter changed color. And in this case, participants had to decide whether they want to press the key or whether they stop pressing the key and don’t press the key.
So they have a decision, they are decision trials, where they can decide between pressing and not pressing. The point is, it depends a little bit on when you introduce this change of the color of the cue. If you do that too early, participants can’t stop their response. So they will simply press the key. Because the information that they should think about comes too late, so they can’t really respond to that.
If you present this signal later, then participants first have to stop pressing the key, and then they have to decide whether they want to press the key or not. And this takes time. It takes a few hundred milliseconds to come up with this decision whether you want to press a key or not.
Now in some cases, from an outside perspective, it looks very similar. So when you, for example, press the key and then the signal appears, but you can’t stop the key press because the signal appeared too early, so you can’t really stop your key press anymore.
And then you have the situation where you want to press the key, then the signal occurs, and then you decide, no, I don’t press immediately, I first decide on it, then you decide and press the key. So in principle, in both cases, participants press the key, but in one case, they have an additional decision to do it, and in the other case they simply respond to the stimulus.
And because you can’t distinguish these two trials easily, trial tests easily, we asked participants after each trials whether they decided or simply weren’t able to stop. So what you would expect is that in cases where participants report that they decided to press the key, their response times should be much slower.
Because in these cases they first have to stop the action, and then implement the decision and then press the key. So they should be a few hundred milliseconds slower than in trials where they simply press the key, or where they get the signal but can’t stop themselves from pressing the key.
And in most of the trials, this is really the case. So in two-thirds of the trials, participants are about 500 milliseconds slower if they decide to press the key, compared to the situation where they simply press the key.
However, and this was really surprising, in one-third of the trials, participants tell us afterward that they decided to press the key, but they are nevertheless as fast as if they didn’t come up with a decision, but simply responded. So from these reaction times, from these extremely fast reaction times, one can really doubt the report of the subjects.
Because if they really have decided on it, they should have been much slower. In this sense, you could either assume that they were lying, and we try to exclude this possibility by correlating the data with live questionnaires and stuff like that, or sometimes they don’t have a very good insight into their own intentions.
Because they think afterward, “Oh, yes, I intended to press the key, ” while in fact, they didn’t really decide on it. They simply pressed the key. So this data also suggests that sometimes our experience of intending to act seems to be a reconstruction of things that happen afterward. So you interpret your behavior in this specific way retrospectively, even though there was no intention in the first place.
LUKE: What are some of the broader conclusions that you might consider drawing from that study that you did?
MARCEL: I mean, it is very difficult to draw extreme conclusions. Because, I mean, as we showed, in most of the trials, participants judged their intentions correctly. Only in some trials did they misjudge their intentions.
What I think about these results and other results is that our experience of intention to act is influenced by a number of factors. It is influenced by maybe our deliberation about the decision. It is influenced by information that happened after we responded.
So it is, to some degree and in some situations, a reconstructive process. And a third influence on our actions might be unconscious informations that are in the system and bias our decision.
So in principle I would conclude from this experiment and other experiments that our experience of will is a mixture of a number of factors.
LUKE: Marcel, what are some of the other experiments that you’ve worked on to get at the nature of intentional action and free will?
MARCEL: One experiment that I think is interesting in this context, and also relates to what I said before, is an experiment I did in collaboration with John-Dylan Haynes.
So there we tried a kind of fMRI version of the Libet trials. And the goal of this experiment was to predict decisions of participants from brain activity, and to see how far in advance we can predict the decision of the subject. So in this experiment, participants had to decide between two response alternatives. They could press a left or right key.
They could determine themselves when they pressed the key and which key they pressed, and we measured brain activity with fMRI while the participants were coming up with this decision. And then we used a specific way to analyze brain activity to predict their decisions from brain activity. So what we were looking at, we looked at specific pattern of activity that predicted what participants will later do.
And the interesting finding in this experiment was that already, six or eight seconds before you report that you will press the key, we could predict the decision to some degree, with a 60 percent accuracy.
I mean, this is far from being deterministic, that’s a very crucial point, because I mean chance is 50 percent, so we are a little bit above chance; 10 percent.
But on the other hand, this experiment shows that already eight or ten seconds before you come up with a decision, there are informations in the system that bias your decision. And obviously you have no idea about these biases.
This is an interesting study because, yeah, it shows our experience of will is to some degree a reconstruction of what happened. But it is also determined by information that are in the system, and is also pre-determined to some degree by this information.
LUKE: Well, yeah, six to eight seconds is a very long time.
Are there other experiments that you could conduct, that perhaps you would be able to predict people’s choice with a greater degree of accuracy, but say, maybe, two or three seconds before the action? Do you think that’s plausible?
MARCEL: Yeah, of course. I mean, the closer you get to the action, the easier it is to predict the decision. And so if you look at the point in time where they really act, and you look at motor cortex – the area where the motor command is generated — you can predict with an 80 or 90 percent probability.
The problem with fMRI is that the temporal resolution isn’t very good. So if you get too close to the response, you don’t really know any more whether you are really predicting or postdicting.
In this sense, it was very crucial to show that already very early on you can predict the decisions from brain activity, because it’s eight seconds in advance. You can’t argue that, because of the inaccuracy of the measure you can’t draw the conclusion.
I would assume that the closer you get to the decision, the better you would be able to predict. But this, in my opinion, hasn’t been systematically investigated. But it’s a very interesting question.
LUKE: Do you have other experiments that you know of that you think are really interesting results in trying to investigate the nature of free will through the tools of cognitive neuroscience?
MARCEL: Recently I have started a series of experiments which is motivated by very nice social, psychological finding. It provides a completely different perspective on this question of free will.
Then you’d be a social psychologist. It doesn’t really matter whether we believe in free will or not. So in other words, what happens if it turns out that free will is an illusion? That it doesn’t exist? Would it change our behavior?
They developed a very clever way to investigate that. They gave participants brief text vignettes, with statements. And one group of participants got a statement where these texts claimed that free will is an illusion; that it doesn’t exist; that scientists has demonstrated that free will doesn’t exist.
So participants had to read these text messages and then afterwards they had to do something else. But I will come to that. And they had another group where participants had to read messages that supported the idea that free will exists.
Reading these text messages seemed to influence your belief in free will. So that you can use a questionnaire to measure that and participants who have read the text vignette that claimed that free will does not exist showed lower scores on this free will questionnaire.
So, reading these messages seemed to influence your belief in free will, at least on a very short term basis.
Now, the interesting idea of these social psychologists – it was first Foss and Scoler – was that they then brought participants in a situation where they could cheat. And what they found was that participants who read the text passages, the text vignette that stated that free will does not exist, cheated more than the other group.
Other experiments have shown that questioning belief in free will or providing a deterministic world view leads to anti-social behavior. And I found that very interesting, because this is something… we have to debate about free will. But this addresses really whether this debate might have any consequences.
Now, I’m a cognitive psychologist. I’m not a social psychologist. So when I first heard about these experiments, I thought, “OK, they can show that these beliefs about free will influence very complex social behavior.” But for me the question was, “Why would they also influence basic motor cognition?”
What we did in a recent study – it’s not published yet… but it will hopefully soon. We did a similar kind of manipulation. So, one group of participants had to read texts that stated that free will is an illusion and doesn’t exist, and science has shown that it doesn’t exist.
Another group of subjects read a text that was not about free will at all. It had a similar complexity. It was about consciousness.
Then we asked participants to carry out the Libet task. They didn’t know that there was any relationship between reading the text in the beginning and then carrying out a Libet task. So we left participants to carry out the Libet task and measured EEG as well.
I was expecting that maybe the W judgment might be changed in the group that their belief in free will was questioned. But what we actually found was that the readiness potential changes. So, participants who read the text that free will is an illusion showed lower readiness potential.
The change in the readiness potential occurs about one and a half seconds before participants press a key, and more than a second before they reported that they found intention. So, these data seem to suggest that such high-level belief about free will and determinism affect brain correlates of pre-conscious motor preparation.
I think this is a very interesting finding in a way. Because it shows that this debate on whether you believe in free will seem to affect your basic motor processes.
LUKE: So, I’m not quite understanding, how does, for example, reading the passage about how science has disproved free will and all that kind of thing… How does that affect the readiness potential, and what does that mean?
MARCEL: That’s a very interesting question. I mean and we can at the moment only speculate about the precise mechanisms — how these high-level beliefs about free will influence our basic motor cognition.
I mean, the way we think about it is, that in a way, if you believe that free will doesn’t exist, this has an influence on your self-efficacy. So the control you experience over your actions.
We assume that these beliefs influence your experience of self-efficacy. And this, in turn, influences how you are involved in a specific task. So if you think that it doesn’t matter anyway what I’m doing, you might put less effort into such a task.
And this then leads to a modulation of the readiness potential.
LUKE: OK. Got you.
MARCEL: But this is at the moment very speculative. Because I mean, that this is simply the first study that shows we can find these influences of relatively high-level beliefs on basic motor cognition. And what I’m now interested in is really investigating the functional and neural mechanisms that modulate this influence.
LUKE: Those studies about the effects of belief in free will on people’s behavior are really interesting to me as well.
I’ve read that some researchers or some philosophers think that this indicates maybe we should not… Let’s say science discovers a lot of things that really place heavy, serious doubt on the existence of free will, and some people like Saul Smilansky will argue that the scientists and philosophers who are investigating free will shouldn’t tell the public this.
That they just shouldn’t let the cat out of the bag because society will crumble, and people will have less self-efficacy, and they will be less motivated to act morally and less motivated to act at all. [laughs]
It’s a very interesting debate to have. I don’t even know how you go about arguing those kinds of things, but it sounds like you’re interested in investigating those issues as well.
MARCEL: In a way, our data support that, and also the social-psychological data support this basic idea. I have to admit that I would question that you can withhold the information in a way, I think that’s not really the way to solve this problem.
But of course you have to be a little bit careful about the conclusions you draw from your data. And in my opinion there is not much evidence that free will doesn’t exist. There is also not much evidence that it does exist.
Of course these data seem to indicate that one should be a little bit careful about coming up with statements about free will.
But on the other hand, I also have to say that I think you can’t really completely question belief in free will and people. I mean, even people that would argue that free will doesn’t exist. I think in everyday life, they don’t really implement this knowledge, because you can’t, in a way.
You have to believe that your intentions are effective. If you wouldn’t believe that, yeah, I think yeah… suicide would be the only solution, in a way.
MARCEL: So in this sense, you might be able to modulate these intentions about free will or belief in free will. But I think everybody in fact behaves as if free will exists.
If you lose this experience or conviction, then you are really in trouble, and we see specific pathologies where this is the case, in depression, or other pathologies where people have the impression that they can’t influence the world.
If they can’t control the environment, they have serious psychological problems. Because our belief that we can control our environment is extremely crucial for our health.
LUKE: Well Marcel, it’s been a pleasure speaking with you. Thanks for coming on the show.
MARCEL: Thanks again for inviting me.
Previous post: News Bits