11.11.2009 - On "Unconscious Determinants of Free Decisions in the Human Brain" by Haynes et al

Gaya Ethaniel: Hello everyone :) Sorry I'm late.

Gilles Kuhn: Unconscious determinants of free decisions in the human brain. Chun Siong S. Soon, Marcel Brass, Hans-Jochen J. Heinze, John-Dylan D. Haynes Nature neuroscience, Vol. 11, No. 5. (13 May 2008), pp. 543-545.

TR Amat: http://images2.wikia.nocookie.net/philosophical-seminar/images/f/f6/Haynes_et_al_on_decision_in_the_brain.pdf

Gilles Kuhn: precisely thanks tr

Gilles Kuhn: so lets begin

TR Amat: We aim to help. :)

Teleo Aeon: Hi

Gilles Kuhn: first as you have see reading the article the experience described rely entirely on a imagining machine called a fMRI

Gilles Kuhn: functional magnetic resonance

Gilles Kuhn: last article we read was depending on other instrument the eeg (electroencephalogram)

Gilles Kuhn: i explained that eeg have very good temporal resolve and terrible spatial one

Gilles Kuhn: the fMRI has reasonable temporal resolve and very good spatial one

Gilles Kuhn: temp resolve of fmri is in the 5 sec range

Gilles Kuhn: spatial resolve in the millimeter range

Gilles Kuhn: but what are we measuring exactly in fmri ?

TR Amat: If you carry out your thinking with your brain it would seem difficult to avoid this type of scanning. :)

Gilles Kuhn: to understand it some tec explanation about fmri are necessary

Gilles Kuhn: well actually you need to be in a cumbersome scanner and not move at all not handy for mind reading in non cooperative subject ;-)

Gaya Ethaniel: :)

Gilles Kuhn: so basically a mri will first polarise all atomic spin in your brain using for that huge magnetic field (from 1 to 7 tesla earth natural field on surface is between 3 to 6 MICRO tesla ....)

Gilles Kuhn: so dont take your pacemaker or the little coin you have mistakenly ingested it will finish on the scanner really quick and violently

Gaya Ethaniel: Nice ^^;;;

Scott23 Hawker: I hope titanium dental implants are OK.

Gilles Kuhn: then it will cut the magnetic field and let depolarise all the spin which is measured thanks to the magnetic current induced

Agatha Macbeth: whoa

Gilles Kuhn: i don't know scott if titanium is magnetic sensible then no$

Gilles Kuhn: thats for mri now let explain fmri

TR Amat: Need more than a tin foil hat. :)

Gilles Kuhn: in fact the idea is to measure difference between blood that is still with oxygen and blood without

Gilles Kuhn: as they have inverse magnetic proprieties

Sartre Placebo: hi everyone

Gaya Ethaniel: Hello Satre :)

Gilles Kuhn: so as nerve cell active take more oxygen than inactive the blood flow in micro vessel near active neurons will change magnetic property

Alfred Kelberry: hi

Gilles Kuhn: basically thus fmri measure difference in the blood near the neurons and by that we can deduce after huge statistical treatment of the data which zone are more or less active in the brain with grat spatial resolution to 1 millimeter

Gaya Ethaniel: Hello Alfred :)

Alfred Kelberry: :)

Gilles Kuhn: to now all is clear no question?

Frederick Hansome: so what will happen if the earth's polarity should suddenly change?

Teleo Aeon: million dollar Q

Gilles Kuhn: ok sure because i assure you first time i had a lecture on that between med doctor physicist and engineers there was always something unclear for some group of population ;-)

Gilles Kuhn: nothing

TR Amat: So, you are OK as long as your blood contain haemoglobin, and isn't that dodgy silicone mix. :)

Gilles Kuhn: for a mri i mpean

TR Amat: fluoro-carbon*

Gilles Kuhn: earth magnetic field is 30 to 60 micro tesla we speak here of magnetic field of a hundred thousand factor more

Alfred Kelberry waves to aurora

Gilles Kuhn: ok so everyone understand more or less the fmri doing and thus is quality and limitation?

Gaya Ethaniel nods ... basics ok :)

Gilles Kuhn: i knew my seminarist were a bunch of quick genius ;-)

Gaya Ethaniel: Very kind of you :)

Gilles Kuhn: ok so let pass to the experience in itself

TR Amat: Looks at what is going on inside your brain without need for slicing. :)

Gilles Kuhn: you will see there is some similarities with libett protocol

Gilles Kuhn: basically you put a subject in the fmri (that dont hurt if he had none metallic part ...)

Gilles Kuhn: then you give him two button one left one right

Gilles Kuhn: and ask to push one of the two

Gilles Kuhn: nb the subject see letter desfiling in front of his eyes

Gilles Kuhn: the subject is instructed to remember the letter he see when he consciously decide (i e when he feel he take the decision to push whenever button he want)

Gaya Ethaniel: desfiling?

TR Amat: Displayed?

Gilles Kuhn: passing in quick succession

TR Amat: streaming?

Gaya Ethaniel: ah ok :)

Gilles Kuhn: yes streaming thanks tr

Gaya Ethaniel: TR offers a bridge between Belgium and Korea heheheh

Gilles Kuhn: so after he push the button he is show another panel that permit him to indicate which letter he remember to havesaw when he "decided"

Gilles Kuhn: naturally during al the process the fmri is working

Gilles Kuhn: and all is duly chronometred

Gilles Kuhn: so and thats the point we know three important and objectivable things

Gilles Kuhn: 1 when the subject believe he make the decision

Gilles Kuhn: 2 when he push the button

Gilles Kuhn: 3 and all his brain working during and preceding the process

Gilles Kuhn: result are let say interesting

Gilles Kuhn: you remember in the case of libett it was all about time and RP that preceded the conscious decision by 500 ms

Gaya Ethaniel: mhm

Gilles Kuhn: now thanks to statistical treatment of the data the experimentator first localised several brain zone active to 10 second before the conscious decision and related to the process

Gilles Kuhn: but more interestingly he could looking at the brain activation predict which decision the subject will take to 7 second before the conscious subject report he take the decision

Scott23 Hawker: How accurately?

Gilles Kuhn: so i repeat he can predict ;;;; yes good point and not mentioned in the article i searched for the onfo and found with a 0.6 to 0.7 accuracy which is statistically relevant

Sunfire Langer: bbiab

Gilles Kuhn: sorry sunfire?

Gaya Ethaniel: be back in a bit I think ...

Gilles Kuhn: ah

Gilles Kuhn: well if i proposed this article is that he add a very interesting component to libett studies

Gilles Kuhn: apart of the neurological zone he identify precisely

Arisia Vita: sorry to jump in Gilles, but what does this mean for free will?

Stewart Macpherson: heh. Indeed.

Frederick Hansome: good question

Gilles Kuhn: well if you can predict what choice a person will made before he is conscious to have made the say choice it tell something troubling for free will dont you think?

Stewart Macpherson: Not really.

Stewart Macpherson: Maybe if you're Descartes.

Strider Villota: I think at minimum it challenges what I thought was a possible relationship between free will and consciousness...worth thinking about

Arisia Vita: a predictable choice is no choice at all, but a response?

Gilles Kuhn: so if i can predict which choice you will have before you think you made them for you it is not let say disturbing?

Strider Villota: somewhat but you could just be predictable

Stewart Macpherson: No, so long as it's based appropriately on my belief and desires.

Gilles Kuhn: in the case arisia farewell free will at least in motor choices

Arisia Vita: not disturbing to me at all

Teleo Aeon: not in my opinion... unless you want to take peoples conditioned choices, and the biased impulse towards these as a new kind of concrete fact.

TR Amat: This is sort of like setting a random number generator running in your brain to decide when to press the button?

Gilles Kuhn: ah but it is based on brain activity starting 10 second before you think you make the choice

Stewart Macpherson: And it's not as though you're going to tell me, "Now you will do A." The time gap is much too small.

Strider Villota: in this case though apparently the decision was made significantly before you realize it. it is not easy to dismiss

Gilles Kuhn: not the button one of the two button the prediction is which button will be chose

Scott23 Hawker: Suppose that a statistician could look at a persons pattern for say 10,000 run, discover that they are not random, and they predict future choices with 60% success without any connection to the person's brain. All this would prove is that the choice is not purely random.

Scott23 Hawker: IT is difficult for humans to pick randomly.

Teleo Aeon: using your rational faculty is not dependent on irrational impulses anyhow.. rationality has the feature of impartiality.

TR Amat: They weren't allowed to roll a dice then. :)

Gilles Kuhn: well but here we dont speak statistic but single example and about that the choosing pattern was 50/50 left and right for all subject

Frederick Hansome: I think that fMRI is useful in detecting physical pathology in the brain, but pretty useless in helping us make adaptive choices

toBe Destiny: _/\_

Strider Villota: So maybe we could say that the brain pattern and the decision are correlated but not causal...is that right Scott?

Gilles Kuhn: sorry teleo but there is n rationality involved at this stage of choosing its only a motor movement choice without any other implication

TR Amat: I wonder what would have happened if they were told to only make their choice of which button to press just before they pressed it?

Scott23 Hawker: So is L R L R L R L R L R... but that is not random.

Teleo Aeon: but one must have some predisposed structuring about what is being decided upon Giles.

Gilles Kuhn: adaptive choice what do you mean

Scott23 Hawker: Strider. Yes.

TR Amat: I don't claim to have a reliable random number generator in my brain. :)

Gilles Kuhn: no the point is to ask the subject to choose between right and left button as it suit is fantasy

Frederick Hansome: adaptive choices are ones that benefit the individual

TR Amat: I think the choice is likely related to one of the bio clocks/rhythms running in the brain.

Gilles Kuhn: frederick we dont speak at thislevel of choice right now

Frederick Hansome: ok

Frederick Hansome: sorry

Gilles Kuhn: and tr the zone involved were parietal and polar frontal cortex not know to be involved in biorythms or time sense at all

Simulat Almendros: This seems to indicate that free random choice is a very slow process - which seems in accord with experience too - but we do many things that we couldn't do if all of our actions had to start happening in our brains 7 seconds before they happened with our bodies it seems

TR Amat: Most decision making, to allow a real-time response to danger, is likely in the sub-second range?

Gilles Kuhn: good point simulat

Gilles Kuhn: yes but sub second are reflexes

TR Amat: I claim a lot of my conscious decision making is pretty slow. :)

Arisia Vita: are reflexes choices?

Gilles Kuhn: libett result and other point that "choice " need something like 1 second

Gilles Kuhn: by definition no arisia

Teleo Aeon: Giles.. the subject is being instructed with language.. if something is making a random choice based on an instruction to make a choice, I can't see what one can say about any of it.. other than the subject is carrying out a motor reflex based choice of some kind.. how it may be, or may not be biased is unknown.

Arisia Vita: are thoughts reflexes?

TR Amat: You use your unconscious to handle things fast, your conscious is for novel stuff.

toBe Destiny: thoughts are not reflexes

Simulat Almendros: nods to TR

Strider Villota: One thing that bothered me is the 60%. This is significant if I tried to pick the action of another, but seems low for my own brain to pick my action.

toBe Destiny: but have a component to the production of a thought

toBe Destiny: that involves mechanisms

Frederick Hansome: conscious choices are relatively slow, but subconscious decisions are extremely rapid, in the milli-secaond range

Teleo Aeon: the body does not think. we presume it is the mind that is making a thinking choice and not the body making a tactile ofr phycial choice.

Gilles Kuhn: teleo the instruction s given way beforehand and the procedure is repeated a big number of time in sessions that are of one hour more or less (more could be qualified as torture due to the necessary immobility and the huge noise of the magnetic generator

Gilles Kuhn: mmh teleo ineresting point but the big problem indeed is in the mind body problem relationship applied to the free will problems

Gaya Ethaniel: heh :)

TR Amat: If you want a good example of complex unconscious processing, look at car driving. :)

Gilles Kuhn: yes i mentioned that one last session

Frederick Hansome: the body thinks with its brain, a process called the mind

Gaya Ethaniel: Or playing musical instruments ... :)

Gilles Kuhn: but its very conscious when you learn how to do it which is a perilous procedure for other if not you ;-)

Teleo Aeon: there are lots of undeterminable influences going on in the situation Giles.. thats for sure.. but apart from that, we always think that a mind must be the initiator of an action... we don't know this.. it is merely presumed.

TR Amat: The main job of the conscious is to train the unconscious. :)

Gaya Ethaniel: hm ... ponders ...

Gilles Kuhn: yes frederick but the problem then is consciousness is that a epiphenomena or not ?

Arisia Vita: but what drives the initiator?

toBe Destiny: I wish that somehow the knowledge that each of us had on these subjects might be sorted out and made available

TR Amat: Most conscious activity is in response to external stimuli.

Arisia Vita: response = reaction?

Stewart Macpherson: There are behavioural cues that let us predict actions long before the decision to act enters consciousnesses.

Frederick Hansome: please rephrase your question, Gilles

Gilles Kuhn: yes and teleo this presumption is at least put to test by this kind of experimental result or so claims a good number of neuroscientist

Teleo Aeon: there is little reason to press a left or right button.. there are two options.. the mind would have little reason to choose so, maybe the body takes a role there, then

toBe Destiny: there are so many cultural elements that are so close to pure mythology

toBe Destiny: and yet we have not way to dispute based on evidence

toBe Destiny: or even peer review

Gaya Ethaniel: I don't quite understand what the paper mean by 'prediction'.

TR Amat: Books and wikis are ways we share our knowledge...

Gilles Kuhn: cues are not automatic procedure stew in our case a bloody computer determine what will happen

toBe Destiny: yes

Teleo Aeon: neuroscientists reduce their dataset rquirements so low that one can often accuse them of finding a kind of result they are intending to find.

toBe Destiny: i agree

Gilles Kuhn: yes gaya the paper is not clear and very badly written probably to avoid to mention weaknesses like the 0,6 - 0,7 accuracy of prediction

Gaya Ethaniel: ah ... ok

Scott23 Hawker: The subjects may be trying to make the choices seem random, so they wan a 50 50 split. They might then try to keep track of their past choices and if they thought that had picked too many lefts, then tend to choose right for the next 10 seconds or so. They would still be making a free choice, but may have done so before picking the exact button.

TR Amat: Paper was obviously designed to upset naive philosophers. :)

Gilles Kuhn: actually i tried to be advocatus diaboli in defending the paper tonight but i have lot of critics myself

Teleo Aeon: to me the experiment indicates that the body can take a role in choice.. which I always thought anyhow.. but that doesn't preclude rational choice to me.

Stewart Macpherson: Well, if the issue is that there are events whose occurring makes it very highly probable that a subject will perform a particular action before it enters the subject's consciousness to do so, we already have those. The point isn't that it's a computer. The point is supposed to be that the "decision is already made", as it were. Of course that there are events that allow one to predict a decision doesn't suffice for the decision's actually having been made.

Gilles Kuhn: right scott but the problem is the prediction using cortex measurement before the subject decide

TR Amat: More I learn the more I think the brain does a lot of work in almost any conceivable 'mental working'.

Stewart Macpherson: The probability is very high that Gilles will tell me I'm wrong. Indeed, it is close to 1.

Sartre Placebo: /whispers night

Gilles Kuhn: sorry but if you can predict a choice then the decision was already made at least in simple situation like that

Stewart Macpherson: See? I predicted it.

Stewart Macpherson: But I won't hold it against you, because you clearly thereby had no freewill.

Gilles Kuhn: sophism stewart you dindt predicted exactly what i will do

Teleo Aeon: Choice is already accepted to operate on at least 3 levels anyhow.. the physical.. the genetic pre-dispossition/environmetal and the rational/thinking..

TR Amat: Can you predict what people will do in SL based on the appearance of their avatar? :)

Gilles Kuhn: prediction to be relevant in our context must be full prediction

Stewart Macpherson: No, the general form of argument is just invalid. And it's not clear why how fine-grainted the event description is has anything to do with it.

Gilles Kuhn: so saying he will be furious if i insult him is not very predictive as info prediction are more strong due to the improbability of what they predict

Simulat Almendros: The result here is consistent with the idea that it choice is independent of consciousness - that is the brain decides what to do and we then become conscious of the choice

Stewart Macpherson: His being furious is not a wilful action.

TR Amat: What sort of decision making process do you think worth trying to investigate with fMRI?

Arisia Vita: it is a reaction?

Scott23 Hawker: Gilles, did the paper say if the "prediction" were made before the fact? Or were these postdictions made by correlating data after the experiment?

Violette McMinnar: who or what programmes the brain then?

Gilles Kuhn: sorry but the general form of the argument can be not useful in free will debate but we have not here an argument but a experimental result

Gilles Kuhn: another good point scott but as the procedure is automated the difference is not important at my opinion

Stewart Macpherson: Simulat: I agree. Though 10s earlier suggests that the events that allow one to predict the actoin aren't constitutive of the actual choice.

Scott23 Hawker: IS the computer trained to each individual subject?

Stewart Macpherson: They are precursors that make it highly probable that a particular event will occur, a particular choice will be made.

Scott23 Hawker: It would be impressive if a machine could tell what a new individual wold do 10 sec before he did it.

Gilles Kuhn: because stewart you confound decision choice conscious and not if the conscious choice is only a reflect of a decision 10 sec before that is unconscious then the conscious choice is epiphenomenal

TR Amat: Obviously the fMRI magnetic field re-programs the brain. :)

Gilles Kuhn: no scott that would render the thing totally artefactual

Stewart Macpherson: Gilles: I haven't a clue what you were trying to say.

Gilles Kuhn: lol tr !

Teleo Aeon: pressing a left or right button doesn't really seem able to engage much thinking in my opinion.. there is little or no meaning or outcome in deciding to press one or the other.

Gilles Kuhn: yes i figured that out stewat

TR Amat: I understood there were issues of careful level adjustment in the attempts to use fMRI for lie detection, for example?

Gaya Ethaniel: heh :)

Gaya Ethaniel: We should close shop soon, people getting jittery now.

Gilles Kuhn: well for uncooperative subject i am quite sure that using fmri is pointless

TR Amat: Obviously the magnetic fields here are too strong. :)

Gaya Ethaniel: :)

Gaya Ethaniel: Well so much thinking involved, I'm not surprised.

Frederick Hansome: biofeed back is good for lid detection, not fMRI

Simulat Almendros: consciousness would only be epiphenomenal if it had no effect in the brain - but being aware of a choice could easily be able to update the ongoing brainstate so that it would evolve differently than if there was no consciousness

TR Amat: I still think it is worth asking what sort of decision making process do you think worth trying to investigate with fMRI?

Gilles Kuhn: but i have no idea there was people who thought to use fmri as lie detector well i think they will have the same problem that with eeg it will not work in a manner acceptable by any court

Gilles Kuhn: thats in part the vetoing argument of libett simulat and i quite agree

TR Amat: There are moves to use fMRI for lie detection in US courts, if I recall correctly.

Teleo Aeon: well fmri has succeded in telling me little imho. :) in regards to a freewill debate at least.. I already thought that the body can initiate action in the absence of meaningfull information.. I see that every day.

Gilles Kuhn: actually i think we can distinguish elementary choice like pushing one button (when the button is not linked to a a bomb) and complex choices like those involved in moral reasoning

TR Amat: Lots of pretty (false) colours on screens, and, lots of argument about what they might mean. :)

Teleo Aeon: some presidents even have a problem with it being linked to a Bomb Giles :D

Gilles Kuhn: yes but some of these in recent history have apparently no brain in measurable sense...

TR Amat: Don't Press The Button. :)

Teleo Aeon: heh

Teleo Aeon: we should scan them in future

Teleo Aeon: to make sure Brain matter has actually been instantiated by nature

Gilles Kuhn: well we begin to digress and the hour is over so i propose we terminate the official part we will continue in more general the subject next week

Gaya Ethaniel: Thanks everyone :)

Simulat Almendros: Thanks Gilles - very interesting

Scott23 Hawker: If the prediction program is trained with a run of x pushes, it should be compared with a computer program designed to analyze patterns not with a 50 50 random series. The pattern program should do better that 50 50 but would not use the FMRI data at all.

TR Amat: No robot presidents then? Built in the US of A? :)

Strider Villota: Thanks Gilles...very interesting

Frederick Hansome: thanks, good night all