Design predicts results but do blobs predict the will?
Jan Kalbitzer
Thursday, 29 May 2008 19:01 UTC
In the April 2008 issue of Nature Neuroscience, Haynes and colleagues report their results from a study based on a modification of the famous Libet experiment on free will. Test persons were placed in an fMRI scanner and given the free choice whether to press a left or a right button and announce when they became consciously aware of their decision. Analyzing the results, Haynes & co found that a distinct blob-signature appeared in Brodmann area ten and parietal cortex ten seconds prior to conscious awareness and predicted with high accuracy which button the test person would press. These findings excited the scientific community as well as common media. However, even more striking is the continuous ignorance towards the degree to which the often very simplistic design of fMRI studies influences their outcome.
Philosophers like Kierkegaard and Heidegger, concerned with human freedom, believed that free will decisions have a strong impact on a person and that the process of realizing our freedom and accepting agency for the actions in which we manifests ourselves is rather long and emotionally pretty painful (despair, anxiety). Non of the designs of our fMRI experiments so far can possibly capture the complex interaction between different brain regions while someone is considering and discussing whether he e.g. should leave a save work place or continuously give up his integrity.
I do not believe that we will never find out, I am just saying that departing from such a concept of free will, the experiment of Haynes’ group seems completely ridiculous. In other words, they observed something far from even having the potential to be a free will decision because free will needs something it is truly concerned with as well as time and interaction.
Updated 29 May 2008 20:53 UTC
-
Replies
-
I’ve not read the original paper I have to admit, but wasn’t it milliseconds, not seconds. Anyway, moot point to your argument.
I agree with you and to be honest this who debate about the nature of free will is somewhat biased, I think.
Even if the fMRI or EEG picks up a signal before conscious awareness of a decision is made, I don’t think this means we’re slaves to a set of electrical impulses. I still had to make the choice, left or right, the fact that I detect that choice a fraction of a second after my cortex fires means nothing. It’s still my brain in my body. It might also be that the coordination of firing requires the AP to fire from motor cortex before conscious detection in order to coodinate the two. Otherwise I consciously detect an unconscious nervous twitch in my hand, then “decide” to press a button.
That out of sequence motion would be a handicap, but could also underlie syndroms such as Alien Hand…
-
Ian, the astounding result was that it was seconds, not milliseconds. The data are the data, and all the proper controls were done such that there was no obvious reason or artifact for the results to come out the way that they did.
Jan, the media hype has been staggering and perhaps flying a little too fast and loose. In reality, the paper presents a temporal order of events. After that, one can speculate away as to what that order represents; after all, the concept of free will is for the realm of philosophy, not for biology anyway.
I agree that randomly choosing to press a button does not equate with complex decision-making, but that does not take away from the results, which clearly show that brain activity greatly precedes our ability to report on a decision. Perhaps the individuals were mentally debating as to when the exact moment of the decision occurred, delaying their reporting times…
-
these experiments bring out (a) the delay of consciousness and (b) the compensation for such in the use of probabilities to pre-empt events and so allow for action prior to the actual (or more so allow for synchronisation of event and action)
When we work off instincts/habits there is no delay. Introduce novelty and delay is introduced where consciousness aids in learning, in habituating to the novelty. COMBINE instinct/habit with novelty and you will get a range of behaviours some unconscious, some conscious.
Chris
IDM intro -
First of all, we are not talking about an absolute prediction but of a likelihood. It is important to keep in mind that all we have are probabilities. If we see A, there is a certain probability that B appears.
I think the more important question is: do we have influence on our conscious thought stream? And if yes, does this thought stream affect our future decisions? Given that I went for the left button, became consciously aware ten seconds after I already made the decision, and then think this was rather stupid. And then I reflect why I went for left, maybe my right hand was tired from working on the computer, etc. Then I decide next time I’ll will inhibit my impulsive for left… Does this whole consciousness stream just follow another decision that has been made? Is consciousness always delayed and if yes, to what, e.g. in cases where I think about life without making a decision?
My guess is more that many of these findings are random findings and they are biased by what is most publishable. How publishable would it be, if they had not reproduced anything? Surely in a less prestigious paper, since negative findings are difficult to distinguish from bad study design. But honestly, the same is probably true for positive findings in the same percentage of studies. I guess you’re right Noah, Freedom of will is not necessarily the best subject to study for neuroscientists. I would think that economists are probably better candidates. Problem is that neuroscience got so hip that there seems to be much less critical review of results and methodology than in other fields, especially by the public…
-
Chris, if I get you right, I pretty much agree to the model you sketch.
-
other sources:
Glimcher, P.W. (2003) “Decisions, uncertainty, and the brain : the science of neuroeconomics” MITP
Ogman, H., & Breitmeyer, B., (eds)(2006)”The First Half Second : The microgenesis and temporal dynamics of unconscious and conscious visual processes” MITP
Hirstein, W., (2006)”Brain Fiction : Self-Deception and the Riddle of Confabulation” MITP
Oaksford, M., & Chater, N., (2007)”Baysian Rationality : The probabilistic approach to human reasoning” Oxford UP
the latter produced a paper covering this area a few years back:
Oaksford, M., and Chater, N., (2001) “The probabilistic approach to human reasoning” IN Trends in Cognitive Sciences Vol 5. No8 August 2001: 349-357
and then there is:
Parsons L. M., & Osherson. D., (2001) “New Evidence for Distinct Right and Left Brain Systems for Deductive versus Probabilistic Reasoning” Cerebral Cortex, Vol. 11, No. 10, 954-965, October 2001
All of this maps to dynamics of anti-symmetry as compared to symmetry. The anti is LOCAL context, VERY fast, PARTS focused and so use of Baysian statistics (subjective probabilities) – XOR/NOT focus and inductive/abductive.
My IDM material brings out PARALLEL processing in the neural hierarchy and so emotional assessments take place unconsciously to then be suppressed by consciousness due to (a) inappropriateness of the emotional expression/assessment or (b) the high precision consciousness ‘defuses’ the less precise emotion assessment.
The path UP the hierarchy allows for the top to regulate the bottom, impose ‘rules’ for local context expression (symmetry) as well as break rules (instincts – and so break symmetry)
Frontal lobes/PFC cover high precision, high differentiation, that is representative of consciousness. Damage these areas and we fall back onto our ‘primate’ behaviours.
-
Jan: ”Problem is that neuroscience got so hip that there seems to be much less critical review of results and methodology than in other fields, especially by the public…”
...but Jan, isn’t that what makes journal club so much fun! :)
Noah: ”all the proper controls were done such that there was no obvious reason or artifact for the results to come out the way that they did”
...I certainly wasn’t suggesting any impropriety, and I apologise if I came off that way.
”the concept of free will is for the realm of philosophy, not for biology anyway.”
...perhaps a new branch for us to develop…neurophilosophy…(neuroethics exists huh?) Free will, the nature of consciousness, the soul/body issue… some of these ideas can be linked back to studies of addiction (et al. including extinction, sensitisation etc.). The very nature of self and self-determination being afected by the levels of c-Fos in my VTA… :/
-
I am just visiting a good friend who has some skills in computational models. I find it tempting to work on a model with different genotypes and socialization factors as multipliers to predict something interesting (e.g. mating behavior in humans). Reaching a high predictability there would appear more convincing to me in regard to free will because it would affect the general self-concept and not just the choice of a button.
I am not really sure if and how the question of free will is possible to discuss within the discipline of philosophy.
-
discuss within philosophy…
I believe philosophy can sketch the circumstances of what we believe is free will but philosophy can not investigate the subject. As I understand it, the main characteristic of philosophy is that it doesn’t have any subject at all, it reflects in a laid-back manner on issues that other disciplines are concerned with…
-