How did you decide to read this post? You might have seen the headline in an RSS reader or noticed it on the ScienceBlogs home page. Maybe someone emailed or tweeted the link to you. But you still had to make the decision to actually read it. How do you know when you made that decision?
In 1965 H.H. Kornhuber and L. Deeke found that brain activity precedes a conscious choice (voluntarily pressing a button) by 500 to 1,000 milliseconds. But in 1983 a team led by B. Libet found that when people were asked when they consciously decided to press a button, they said their decision came about 200 milliseconds before pressing it -- after their brain had started to process the task.
So did you "decide" to read this post after your brain had already committed to clicking on the link? It's possible, but it's also possible that there's simply a lag between when you were aware of having made a decision and when you actually decided.
Or it may be that you're only certain of having made a decision after you see evidence that that's what you chose to do: Hearing the click of the mouse and seeing the page load might be what actually makes you aware of deciding to read it. William Banks and Eve Isham have come up with a clever way of discerning whether evidence we see and hear affects our awareness of having made a decision.
They asked eight Pomona College undergraduates to watch a representation of a clock on a computer screen. While they watched, their hand was on a button that was hidden from their own view. A cursor moved around the clock's dial once every 2.6 seconds. The students were told to press the button whenever they wished, and then report exactly where the cursor was at the moment they made the decision to press the button. This was repeated 160 times for each student.
The trick was that as they pressed the button, the computer made a short beep. Unknown to the students, there was a slight delay between when the button was pressed and when the beep sounded. This delay varied randomly between between 5 and 60 milliseconds. Did the timing of the beep affect when the students believed they had decided to press the button? Here are the results:
This graph charts the delay of the "beep" from the actual button-pushing against the time viewers said they decided to push the button (measured in milliseconds before the button was pushed). The shorter the delay, the earlier the decision was made. In other words, the timing of the beep, which was randomly selected, had a significant effect on when the students said they made the decision to press the button.
Here's another way of looking at the same results:
This graph compares the time from when students said they decided to push the button to the timing of the beep. As you can see, the time from the reported decision and the beep was nearly constant: students said they decided to push the button around 130 milliseconds before they heard the beep, regardless of the actual timing of the beep.
So a randomly-timed beep has a larger effect on when we think we decided to press a button than the actual time we pressed the button.
In a second experiment, the researchers showed volunteers a video feed of their hand pressing the button superimposed on the clock. Like the beep in the first experiment, the video was randomly delayed by 5 to 60 milliseconds. The results were the same.
Does this mean we have no free will?
Not necessarily. It's still possible that the students did freely choose when to press the button, but that their awareness of that decision is affected by external factors. The awareness of the decision is never actually later than the time the button was pushed, so the cognition of making a decision is never completely divorced from the action itself.
Perhaps we're not ever conscious of having made a decision. Rather, we infer making a decision from other evidence: the sensation of pressing a button, the sound the computer makes to register our choice. That choice may be our own, but the process by which we make it remains a mystery.
Banks WP, & Isham EA (2009). We infer rather than perceive the moment we decided to act. Psychological science : a journal of the American Psychological Society / APS, 20 (1), 17-21 PMID: 19152537
- Log in to post comments
On the small scale of the awareness of a single rapid action this could be construed as alarming. But if you take an individuals actions over time I think that conscious decisions plays a more obvious role, even in those quick actions. Think of 'muscle memory' as a counter to this argument. By repeating to the same action again and again we firmly set the pathways that control behavior so that it feels like skillful behavior that is not under direct conscious control. I think pinning free will to the immediate responses to stimuli is would only be meaningful if we were all chronically impulsive; if however, through deliberated practice we can condition those immediate responses, then I think free will escapes the quirk of our biology illustrated these experiments.
It is also possible that what we perceive as a choice is really a choice not to "veto" the process that had already been started. That is to say, the process of choosing something could have a large unconscious component, causing the earlier brain activity, and what we think of as choosing is only the final "seal of approval" on an action.
I'm going to re-read this whole thing a couple of times over as soon as I have time, because I've been thinking hard about this ever since I first heard of the 1965 experiment. As soon as I started this article, this caught my eye:
Right out of the gate we can see the difficulties inherent in the topic, and the tendency of our thinking to assume its conclusions. When you say it is possible there is a lag between when you are aware of having made a decision and when you 'actually decide', it is the thing that you are identifying as "you" which changes between the two alternatives, not the thing you are identifying as "the decision". The 'you' in the first case is your consciousness, your awareness, your sentience; you WILL. The 'you' in the second case isn't; it is your brain, which you do not control, and while it is YOUR BRAIN, as much as your hand is YOUR hand, meaning it is not someone else's, that doesn't make it YOU. YOU is a thing we know is caused by the activity of the brain, but that doesn't mean any activity of the brain is you.
So essentially this "out" isn't really at all available to us. If I don't become aware of the decision until after it is made, it is unsupportable to claim that I made it. It has been common since Freud to hand-wave the matter by saying "you made the decision sub-consciously", but I think the very point of the research is that there is no such thing. There are things our brains do that we are not aware of (and so "we" are not the ones doing them) and there are things our brains do that we are aware of (and this we call "thinking" and have always, apparently mistakenly, presumed causes and results in decisions.) The latter is "you"; the former is "your brain".
The fact is your brain makes choices. Once you become aware of them (usually but not always inventing a 'reason' for preferring whatever results you are now stuck with) you call it a decision, but that's just the activity of your brain explaining its choice to itself.
And the sun rising yesterday doesn't necessarily mean it will rise tomorrow.
Free choice and free will (effectively identical) have always been a convenient myth. The main reason for maintaining these ideas is that we have not yet come up with a cogent and consistent way of explaining morality without making the assumption that, since we decide what it is we do, we are responsible for it. To recognize that free will is a myth destroys this method, supposedly leaving us in a nihilist universe where the only morality is personal preference and "Do what thou wilt" is the only law. If I am not the cause of the murderous action (meaning if it were not my decision which results in the action, rather than the action occurring beyond my control which only in retrospect provokes my 'deciding' to perform it) then how are we to justify punishing murderers?
I've got two ways to approach this: the first is the more simple. How are we to justify it? We need to do it, whether it accurately reflects reality or not, in order to maintain an orderly society. It is not vindictiveness which causes us to punish criminals (I like to pretend) but simple self-protection. So we are entitled to maintain our current thinking on crime and responsibility despite the fact that it is almost entirely based on a false claim of a link between decisions and responsibility for actions.
The second way is to forge a new link. Having recognized that it cannot be our agency (free will) which causes us to be responsible for our actions, it seems obvious that there must be something else. I presume (because it seems to work out in all cases I've examined) that it is proximity and awareness, as indicative of the potential to prevent the crime, rather than the decision to commit the crime, which causes responsibility. In most but not all cases (mental health being an obvious exception which we already, only semi-explicably, support) that means that whoever is closest to and most aware of a person committing the crime is responsible for preventing it, and that would be the person himself. Inability to control our actions does not provide freedom from responsiblity, because it is awareness, not control, of our actions that provides self-determination, the sentient consciousness that we have, for thousands of years now, erroneously referred to as "free will".
I've become convinced over the last few years that this is the correct approach. It seems to map even more precisely onto what we actually already do, both as individuals and as a society, than the 'free will' justifications we have always used in the past and continue to use today.
@tmaxPA
I think there may be another way to save free will, at least in some sense.
Assume, as you argue, that we don't consciously decide to push a button or pull a trigger. At least, not at the moment it happens. Perhaps it's our conscious reflection on past actions, as well as consideration of appropriate future actions, that determines our propensity to push to button or pull the trigger.
In these experiments, the subjects were told in advance what was expected of them. They had to consciously consider the requested actions before beginning. Arguably, it's during that conscious consideration before the act that free will comes in.
The same could apply to committing a crime. The immediate decision to pull a trigger may occur without the criminal's conscious input. However, the criminal's prior conscious reflection on both prior acts and future possible acts could establish his propensity to pull the trigger at the instant of the crime. If he had properly and consciously established acceptable 'mental guidelines' for whether it is OK to pull a trigger or not, he might not have committed the crime.
Does that make sense?
Why does this result surprise people? Given everything we know about the brain and cognition, it is completely expected. Awareness is the result of a neurological process in the brain. That process has to start before delivering its end result. These experiments simply give some notion of the time interval for that.
From what I've read before, visual and auditory stimuli take different amounts of time to process, so the brain actually generates an internal simulation of the world which, among other things, tries to synchronize them. This experiment looks like it's really studying how that process works more than anything else.
But I've also wanted to ask: is there a meaningful way to define free will? I've never been able to come up with one.
Free will, as it is usually conceived - is doomed to failure anyway. If 'free will' were to mean free of strict causal determinants, that would exactly be what we don't want - the opposite of 'determined' is 'random'. It would be random, and couldn't be connected to any prior thoughts, experiences or evaluations.
So the usual concept of 'free will' is of no use anyway - so it doesn't matter what Libet and so forth show. Aside from the fact that we do know that it is our brain that thinks, evaluates and learns, and on the basis of this makes decisions, contra-causal free will simply cannot give us what we want from it.
In my opinion the option to save 'free will' is the so-called compatibilist stance, which states that free will is compatible with determinism. Daniel Dennett's account of this in "Elbow Room - The Varieties of Freedom worth Wanting" is quite nice. He construes freedom as 'responsiveness to reason and reasons' - which we can get from a deterministic, scientific worldview.
There's often an assumption that not-conscious=not free, which is fair enough. But note that there is not reason to consider conscious choice to be particularly free either, unless you define "free" to mean "conscious". That is, there's no reason to suppose that the conscious activity of the mind is any less a matter of a machine grinding along deterministically than the unconscious activity is.
I don't think there's any need to save free will. It's not a concept that's good for anything. It's not a concept that's needed. There is nothing anyone has ever observed that requires "free will" to be explained. If there is or isn't free will, it has no consequences for anything except people's feelings. Some people get upset by the thought that we don't have it, but that's pretty much the only consequence.
I think these types of accounts are missing the point when they say that "you're" not the one making the decision if a part of your brain makes a decision before another part is aware of it. Why identify "you" with the part that's aware of the decision and not with the part that's making the actual decision? So I agree with tmaxPA
Also there is a bit of a pseudodebate since the idea of a supernatural free will has been untenable for many decades, so it's not like this should be the thing that blows it out of the water.
It is plain to see that living agents have different properties to those of inanimate objects. One of those special properties is an internal experiential point of view. Not only do we cause changes ourselves but we are also conscious of the experience of actually doing it.
When a stimulus arrives in the brain, the effect is a cascade of chemical reactions and nerve impulses. These can all be said to be âcausedâ, but the thoughts and feelings which accompany these are also âexperiencedâ.
It is only within this internal domain that free will can be said to exist.
The history of the universe is the story, but we are some of the actors within that story. Our practical reasoning and deliberation become a functioning part of the deterministic process. From the external view, everything is determined and can be laid out in a long chain of cause and effect - the story is finished. From the internal view we are still in the process of creating the story and so we have free will. It's all down to a difference in perspective.
"The second way is to forge a new link. Having recognized that it cannot be our agency (free will) which causes us to be responsible for our actions, it seems obvious that there must be something else."
Without free will, there can be no valid moral imperatives. There can be only ethics, which are collections of rules that we like. You can't spin ethics into morality.
You don't like it when people shoot you, so you are likely to support rules that make it less likely for people to shoot you, as long as the rules do not unduely abridge your freedoms. That thinking involves ethics; no necessary bearing on morality.
However, to eliminate the possibility of free will, you need to have models that reliably predict behavior. I agree that this study is damning regarding the possibility of the existence of free will, but it does not provide such a relible behavioral model. Free will may still exist. Hence, valid moral imperatives may still exist.
Saying that "From the external view, everything is determined and can be laid out in a long chain of cause and effect - the story is finished" seems to be a unrealistic, doting view of science. For example, many basic models are statistical--i.e., they can spot correlation/trends, but not cause.
@ qbsmd
From what I've read before, visual and auditory stimuli take different amounts of time to process, so the brain actually generates an internal simulation of the world which, among other things, tries to synchronize them. This experiment looks like it's really studying how that process works more than anything else.
I agree. The findings here seem as though they could be linked to the literature showing how thought processes can become automated by the cerebellum over time.
But I've also wanted to ask: is there a meaningful way to define free will? I've never been able to come up with one.
Check out Baumeister's (2008) article, where he describes it simply as 'freedom of action'. As a scientific question in psychology, it can be defined as the empirical question, to what extent are our actions free?, with the basic premise that some actions are freer than others.
I particularly like an implicit premise in Baumeister's article, that actions occurring on a micro-time scale (e.g., the decision to act is made in seconds) are likely to be less free than actions occurring on a macro-time scale (e.g., the decision to act is made over days). Qualitatively (Baumeister's example), there should be a difference between choosing to walk (or not) to the store after the rain stops vs. the choice of each footstep taken to the store. There's also another aspect of the experiment not mentioned here that suggests this qualitative difference--as part of the experimental task, participants were also given a choice not to respond.
This particular experiment does not seem to provide a straightforward understanding of free will involving decisions that involve no action.
Reference
Baumeister,R.F. (2008). Free will in scientific psychology. Perspectives on psychological science, 3, 14-19.
Readers interested in a different perspective on this topic
might locate the article by D.M. MacKay, "Freedom and a Mechanistic Universe," in the British Journal for the Phil-
osophy of Science in I believe 1987. But note that anyone who performs the "thought experiment" of walking around
trying to believe that ALL of his or her behavior is "unfree" in any meaningful sense, is likely to stumble upon what Koestler termed "Schizophrenic Ethics" --- the
need to treat at least one's own self as "free," even if
you decide that everyone else's behavior is determined by their genetics, uprbringing, cultural standards, and so on.
But note that anyone who performs the "thought experiment" of walking around trying to believe that ALL of his or her behavior is "unfree" in any meaningful sense, is likely to stumble upon what Koestler termed "Schizophrenic Ethics" ---the need to treat at least one's own self as "free
I think Seligman's work on learned helplessness and learned optimism are the best concepts addressing how social factors can influence one's understanding of self as free (i.e., self-actualization to borrow Maslow's term), assuming such ethics are learned.
even if you decide that everyone else's behavior is determined by their genetics, uprbringing, cultural standards, and so on.
Paradoxically, these ideas are a product of scientific thought. Specifically, science (i.e., experiments) are in the business of identifying causal relationships. And as Baumeister suggests, there are several reasons why 'psychologists' (and probably scientists in general) don't like the idea of free will; the most practical being it is not possible to explain a 'free' action causally. If so, there is no way to explain free will/choice through science.
However, Baumeister suggests a way around this paradox by pointing out that events (and indeed, scientific data) is probabilistic rather than deterministic; implying free will is as well.
This essentially makes science (and specific concepts like free will), the identification of factors that increase or decrease the probability of particular phenomena.
I don't think there's any need to save free will. It's not a concept that's good for anything. It's not a concept that's needed. ... If there is or isn't free will, it has no consequences for anything except people's feelings.
No, it does have consequences aside from people's feelings. As noted in comment #12, it's hard to see how people can be morally responsible for their actions if they're not free to choose whether to do one thing or the other (for instance, to kill or not kill a person). If all your bodily movements are just as involuntarily as your heart beating or an epileptic seizure, how can anyone validly praise or blame you for your actions? Now, you could get around this if you have an explanation why morality actually still makes sense without freedom of choice -- but it'd need to be a pretty counterintuitive explanation. Or you could get around it by saying morality is a mere social construction with no objective validity -- but many reasonable people simply find that position unconvincing and even appalling. So the question of whether we have free will most certainly does have consequences.
That said, I don't see what this experiment has to do with free will. As explained in the blog post, all the experiment seems to show is that events happening after a decision is made can retrospectively skew your memory of how you made the decision. Many people who believe in free will wouldn't be surprised by that fact.
Alternative interpretation: you know absolutely nothing about free will. If you thought about it for 10s, maybe you would see that there is no threat to free will here at all. How does the timing of awareness of a decision affect its being free?
Sorry to be snarky, but I've worked on this stuff for years and so have many others (including people you've heard of, like Dennett). It's not fun to see our hard work trashed like this.
I may be completely misreading all of this, but I don't see how the experiment shows that the decision is made by the brain prior to it being made consciously.
To me, the effect shown is that it's possible to affect memory -- even extremely recent memories. The subjects are being tricked into a false memory -- they are made to think that they made their decision at a different time.
I see nothing which demonstrates that they actually made the decision to press the button later, only that they have been fooled into believing it so.
I recommend interview with Rodolfo Llinas called Enter I of the Vortex (link in my name) for extremely interesting take on this question.
To quote: "I understand that free will does not exist; I understand that it is the only rational way to relate to each other, that is to assume that it does, although we deeply know that it doesnât."
People here seem to insist that there is no room for morality if conscious self does not have executive power. But we can easily observe that the entity called mind, formed by conscious self and the unconscious, must be making decisions. So, we can hold this mind-entity responsible, regardless of whether conscious self was responsible for the decisions or not.
Additionally, people here seem to insist that either all people have no free will, or nobody has free will. But this is not true at all. I believe most people do not have free will, and they miss it because they believe they have it.
Let's say I don't want to watch TV, but I have learned that every time I come home tired, I will turn it on, disregarding my "free will" on this issue. So I can decide to get rid of TV. To achieve free will you need to realize that you don't have it, and you need to control your environment so that you will achieve what you want to achieve regardless of your lack of free will. Thus, you gain free will by realizing that you are an animal who does not have free will.
There are also many interesting teachings about free will hidden in various parts of different religions. One of the esoteric teachings in many religions is how to achieve real free will, that allows you to control situations without changing your environment.
@ Marc:
The nature of the experiment written in a linear fashion is something like this:
1. Watch an object move around in circles on what is essentially a digital representation of a clock face (e.g., numbered 1-12)
2. Keep track (presumably via visual memory) of what number the object is at when you make the decision to press the button
3. Press the button
4a. auditory (beep) feedback after button press (delayed from 5-60 ms; experiment 1)
4b. Visual feedback (of your button pressing, delayed from 5-60 ms; experiment 2)
5. The feedback delays at 4a and 4b influence where one believes they made the decision to press the button press at #2.
So based on this, it does seem like the feedback at 4a and 4b are influencing visual memory. My personal critique of this study (but I also may not be understanding it completely), is that it is not really about free will.
It seems way too familiar to me as phenomena discovered by former professors of mine known as inhibition of return and representational momentum (which is quite closely related to your comment about the experimental task creating a false memory of some kind).
If there was no "free will" as described (last) @16, then you can argue that, just as the person who robbed somebody did not make a conscious choice, so the police who catch him and the jury that punishes him have no choice either, and whether he feels guilt or remorse is also no choice. In that situation, everything is determined, so we're all on the same playing field.
We know our emotions are chemical interactions within our body, caused by stimulus. Does that make them less real or unimportant? For me, I act as if I have a choice, not determined by a thousand factors acting together, and my life is fine. It is an interesting topic of study, however.
Whether someone finds the idea of relative morality appalling or are unconvinced of it is of no relevance to the facts. I can see that argument that you are making (saying this is important), just pointing out that people believe a lot of things contradicted by evidence (6k-old Earth, Ayn Rand is a great writer, etc).
I really do not see how this affects will at all. It is like a bug who coincidentally moves when I want it to move. Am I the cause of it? No. Is there a choice? Yes. As for free will's existence, I do not see what the heck the difference is between free and unfree CHOICE. Even though my environment affects who I am, It is irrelevant to choice because I am that person who chooses because of who I am. I WANTED to do this and that and that is really what it all comes down to. Could I have done differently? If I wanted to, then yes, but I don't. If I could is a question that bugs alot of people, but we try to keep a "could" towards people that already "could" but "would" and think that by doing what they wanted, they are "chained" to their choice, who are themselves, which sounds weird. Even though that me is affected by the world, I make a choice and also affect the world, for which I am a part of. So whether I could is already there, but whether I would is the question.
Free choice:
http://www.famous-poems.biz/Robert_Frost/The-Road-Not-Taken-by-Robert-F…
Unfree choice: being born
Weve done a similar experiment at our University. I agree with the comment given:
"Why does this result surprise people? Given everything we know about the brain and cognition, it is completely expected. Awareness is the result of a neurological process in the brain. That process has to start before delivering its end result. These experiments simply give some notion of the time interval for that."
The fact that there is a delay between brain activity and motor reaction is simply suggestive, I believe, of faster subconscious processes occurring before the conscious mind initiates a motor response. Phenomenon such as instinct or more complex dual processes, say holding a conversation while playing the piano, operate on similar lines.
Having said all this, however, the experiment is not nearly robust enough to make anything but a small dent in such a subtle and complex problem such as the existence of free will. (that not withstanding, its a good experiment to get the ball rolling and open up the topic to debate as evidenced by this forum thread).
The fact that there is a delay between brain activity and motor reaction is simply suggestive, I believe, of faster subconscious processes occurring before the conscious mind initiates a motor response. Phenomenon such as instinct or more complex dual processes, say holding a conversation while playing the piano, operate on similar lines.
Sounds very much like a Stroop effect analog.
Free will (as traditionally understood in philosophy) does not exist in any possible universe. See Radcliffe-Richards' Human Nature After Darwin.
Dear Mr. Munger and the Cognitive Daily Community,
We are grateful of the review of our research article (Banks and Isham,2009) and glad to see that it has generated interests in the topic of free will.
The purpose of this post is to clarify the motivation of our study and itsfindings. As mentioned in the review, Libet et al. conducted a study which showed that the moment of decision to act came approximately 800 ms after
the onset of the brain activity known as the readiness potential. Because of the delay in judgment, the findings have been interpreted to mean that we are not conscious of our action until after it has already started. Such
interpretation makes one critical assumption: the judged time of decision is predicted by and bound to the onset of the readiness potential.
Our study examined the validity of this assumption. If judged time of decision to act were predicted by the readiness potential, then it should not be influenced by any other events that come after the response has taken place.
However, this is not what we found. As your review accurately described, the delayed feedback tone, a post-action event, effectively shifted the reported
time of decision. The later the false feedback occurred, the later the reported time of decision. We do not report any internal event; rather we infer the decision time
from the apparent time of response.
What do our findings say about free will? The "conscious decision" is a retrospective report. We don't have conscious access to the processes that lead to the response and
need to guess on the basis of what we observe ourselves doing. The idea of conscious free will in an action like this is inappropriate. Consider Zen archery.
Just when do you "decide" to release the arrow? Isn't this the wrong question altogether?
WP Banks and EA Isham