Abstract: I will discuss some alleged evidence (from Libet, for example) that conscious intentions never play a role in producing corresponding actions and some alleged evidence (from studies of implementation intentions) that they sometimes do play this role. I then take up the question whether conscious reasoning ever plays a role in the production of intentions and actions. I set the stage for my discussion by rehearsing a familiar scientific argument for the claim that free will is an illusion.
Mele: Intentional action: Controversies, data, and core hypotheses
Mele/Cushman: Intentional Action, Folk Judgments, and Stories: Sorting Things Out
Mele: Effective Intentions (Oxford University Press, 2009) chapters 3, 4, and 7 http://bit.ly/MeleFreeWill
Mele: Free will: Action theory meets neuroscience
Mele: Real Self-Deception http://cogprints.org/295/1/MELE.html
Consciousness can modulate the intention and modify the outcome of the go signal (muscle burst or not). So maybe it is just a longer chain where there is a "go signal" that might "tell" consciousness to proceed with the action or not.ReplyDelete
I'll be back in my office on Monday. It will be much easier to reply to everyone then. In the meantime, thanks for your comments.Delete
I'm not sure what you have in mind. Is the go signal you have in mind internally generated, and is it a conscious or unconscious signal?Delete
(I asked the question live, but I repeat my comment here)ReplyDelete
I do not agree that the signal/push button experiment follow the following processing: Signal->Intention->Action, for the following reason.
The intention is already present before the subject is presented with the go signal. I propose two evidences to support my critic:
1) During a delayed memory task, (during which a monkey knows he will have to do an action once he gets the go), monkey subject exhibits preparatory activity that shuts down with movement onset.
2) The presence of false start when sprinters wait for the go signal. If intention was born after the go signal, why would they start before the go signal.
My physiological explication is that conscious unplanned arm wrist movements are initiated in the frontal area (cognitive movements) while the signal/go experiment involves the visuomotor loop involving the dorsal stream, parietal areas and motor areas, the later being priory facilitated by frontal areas/attention. (Paul Cisek was referring to this loop)
The adaptive justification for the existence of the two loops is the following, the cognitive loop enables any movements, but is slower, while the dorsal stream (...) loop enables faster movements, but is restricted to simple movement, (direct reach, on obstacle reach).
To illustrate my point, sculpting requires the cognitive loop while harpoon fishing requires the dorsal stream (...) loop.
You say "the intention is already present before the subject is present with the go signal." Question: what intention? Here's one relevant intention: When(ever) I detect the go signal, click the mouse button. Call it intention 1. Here are two theories about what happens. Theory A: Intention 1 and detection of the go signal together generate an intention to click now that in turn generates a clicking action. Theory B: Intention 1 and detection of the goal signal together generate a clicking action without generating an intention to click now. I mentioned these two theories in the Q&A period and said that I know of no evidence that favors one over the other. Are you thinking that you've provided evidence in favor of theory B?Delete
One my ask, using your firecrackers metaphor, why not just having a simpler causal chain? (1) light directly the gunpowder, (2) explosion! The conscious intention (the wick) might be present in some cases, but absent in other ones. In these cases, (1) is necessary AND sufficient to bring about (2). If that is true, it seems strange that in those cases in which an intention becomes conscious (1) is not sufficient anymore. Couldn't the conscious intention be just an emergent property - not necessary nor sufficient to bring about (2)?ReplyDelete
Yes, a conscious intention to flex now might be neither necessary nor sufficient for flexing. But my point was about a certain line of reasoning. I'm pasting in the quotation from Roediger: "Clearly conscious intention cannot cause an action if a neural event that precedes and correlates with the action comes before conscious intention." If we apply the same reasoning in the case of the firecracker: we get the following unacceptable result: the burning of a wick cannot cause an explosion because a wick-lighting event that precedes and correlates with the explosion comes before the burning. Both the lighting of the wick and the burning of the wick are causes of the explosion.Delete
The comments about "experts" like professional baseball player whose behavior is triggered and done rapidly and it seems without needing any decisions brings the interesting questions of how experience, expertise and intense learning modulates the need for consciousness in our actions. Is it still needed or our behavior has become a kind of reflex?ReplyDelete
Well of course if we assume it plays a role in our decisions which doesn't seem that clear for now!Delete
There a nice passage from William James where he says that consciousness drops out of processes when it's no longer needed. That's what's going on with gear-shifting and the like after you become an expert driver. If things go wrong, consciousness seems to kick in.Delete
I am not familiar with William James' writings (although I certainly intend to pick up some of his texts after having heard so much of him at this Summer Institute), but wouldn't what you are referring to be more akin to his concept of attentional spotlight (correct me if this concept is not rooted in James' thoughts). I think we all agree that both the professional driver and the first time driver are conscious. Same goes for the pro athlete and the amateur. It seems more appropriate to characterize differences between action production in the expert and the beginner as requiring different levels of attention, not consciousness.Delete
I really liked the reporting function of consciousness that Alfred Mele brought in his talk.ReplyDelete
COMMENTS COPIED AND PASTED FROM THE CORRESPONDING POST ON FACEBOOK :Delete
TURING CONSCIOUSNESS :
"Report what to whom? If the data are already in the brain, who else needs to be briefed. It's language that reports to other people, but the problem is explaining why that's felt, rather than just done."
GUILLAUME LOIGNON :
"Dennett mentions that possibility somewhere: if two modules do not usually communicate, language could "hack" them into doing so."
Thanks. I don't understand TC's comment, unless TC wasn't at the talk. Anyway, art of the participants' task in the Libet study is to report to the experimenter where the spot was on the clock when they first became aware of their urge, intention, or whatever.Delete
TC is me! (And though I was monitoring dozens of commentaries the while, I was indeed at the talk.)Delete
The theory that (language-like), feeling is in order to facilitate or integrate internal information "reported" from here to here inside the brain seems to me to be missing something (essential). Information is data. Data can be transferred between robots or within robots. It is not clear why either the external communication or the internal communication should be felt...
An audience member asked for clarification on why we can't measure the activity prior to the muscle trigger in the original Livet experiments. I don't know much about EEG, so I don't understand why we couldn't just look for activity prior to the muscle/reported intention that might correlate with forming the intention. Can someone expand?ReplyDelete
In the main Libet study, the muscle burst triggers the computer to make a record of the preceding second or more of brain activity. So Libet does measure activity that happens prior to the muscle burst. Maybe you're thinking about what happens when there is no muscle burst. When the muscle burst is used as the trigger, then in these situations there is no record of brain activity.Delete
Even if we grant the existence of effective intentions, is this enough to save our folk psychological conception of free will? Even on Mele's interpretation of Libet's findings, actions are INITIATED unconsciously. Mele attempts to save a causal role for conscious intentions but he still acknowledges that they are part of a causal chain that begins unconsciously in what he calls the PPG. I'm not sure this addresses in any serious way concerns over source incompatibilism. And to the extent that ordinary folk are committed to the idea that their conscious selves are the source of their actions (that they consciously initiate their actions) Libet's findings would still be troubling.ReplyDelete
That said, I still have a question about Mele's interpretation. Mele attacks Libet's claim that an unconscious decision is already made at the point of the RP. Perhaps the RP is best interpreted as an urge and not a decision. But what about the findings of Haggard and Eimer that appear to show that by the time of the LRP (which still precedes W) a decision of WHICH HAND to use appears to have already been made? If in a free choice paradigm the lateralized readiness potential correlates with the decision of which hand to flex (and not just an urge to flex), why is it not fair to say that an unconscious DECISION has already been made prior to W (even if conscious intentions are part of the causal chain further down the fuse)?
About the LRP: Why should we think that it's earliest bit is correlated with a proximal intention rather than something that might be among the causes of a proximal intention?ReplyDelete
I'm not sure what you mean by my conscious self being the source of my actions. If this requires that my conscious intentions not have any nonconscious events among their causes, then I reject the claim that my conscious self needs to be the source of my actions in order for me to act freely and morally responsibly. (Surely, neural events of which I'm not conscious are among the causes of all of my intentions.) To explain why I'm not a fan of extreme sourcehood as a requirement for free will or moral responsibility, I'd need to explain how I understand free will and moral responsibility. Fortunately, I've done that elsewhere -- for example, in my book *Free Will and Luck*.
Your latter point is a fair one. I know that you have addressed the issue of sourcehood elsewhere. All I was trying to say is that even if you grant the existence of effective intentions, you cannot disarm Libet's threat without getting into a larger discussion of what IS required for FW and MR (especially as the folk conceive of them). I know that you do that elsewhere but those arguments have to be combined with your defense of effective intentions before free will can be saved.ReplyDelete
As for the LRP, it does seem to me that this is more than just an urge. As Haggard describes it, "The LRP has a particular psychological significance in situations where the subject must choose between a left- and a right-handed action: once the LRP has begun, the selection of which action to make must be complete. That is, by LRP onset the intention has progressed from abstract stage (‘Do something or other!’) to drive a specific movement (‘Do precisely this!’)."
I have interrogations about the implications of the [goal intention vs implementation intentions] or [distal intention vs proximal intention] distinctions (I understand they are analogous) in health behavior change studies.ReplyDelete
Ajzen's intention concept was found to be the strongest predictor of a behavior, and was therefore described as a proximal predictor, as opposed to the distal predictors that are attitude, subjective norm and perceived behavior control (note: the distal-proximal distinction here is not analogous to the one found in my 1st sentence).
According to the "intention-behavior gap" currently observed in the litterature (people fail to bring about their intention, and predictive models can only explain up to 50% of the variance of the behavior), Ajzen's intention is a goal intention, a distal one.
We also know that many people do not form implementation intentions (specific plans) despite their "goal" intention to perform the behavior.
What are the implications for the Theory of Planned Behavior ? "Goal intentions" are not as strong as we tought?
Above all, what is the link between goal intentions and implementation intentions ? Is it as simple as "the strongest the goal intention, the most likely an implementation intention is to be formed" ?
What causes the transfer from one to the other? In my opinion, self-determined motivation, self-efficacy and self-regulation are involved in all this, but I don't know which one is the most important.
Dr Mele, I would like to read your impressions on this. Thank you.
Dr. Mele mentioned at one point that one way of determining the right causal explanation of the Libbet results would be to try to do an experiment where you have the characteristic EEG ramp-up of Libbet experiments but where the subject could choose not to press the button. Something (potentially external) would trigger the EEG ramp-up in a reliable way, but the subject would then have the option not to press the button. But he never mentioned if this or anything like it has been tried. Is there? It seems to me that if suppression is possible it would give strong support to his wick analogy.ReplyDelete
TO DO OR NOT TO DO?Delete
(1) As the clock goes round, every instant that the subject is not deciding to press the button is an instant that he is deciding to "not-press" the button.
(2) But ERPs need an event on which to trigger: What is the event if the subject is not doing anything? What is the premotor potential preparing to "do"?
(3) Of course, the Libet paradigm could be used to test response inhibition instead of response generation. There could be a signal (perhaps a cyclical one), in response to which the subject is asked to press the button, but he is also asked to occasionally,
at times of his own free choosing, to inhibit the response, and not-press. The not-press could be timed the same way as the presses (using the average latency of the press after the signal), but what it that likely to show, except much the same picture as for the press?
(4) Libet, in trying to explain away the prima facie implications of his findings for free will, invoked something similar to response inhibition: He said that although our pressing may be unconsciously pre-tirggered, we can always consciously over-ride or veto it, and that might be pre-willed. (That's rather grasping at straws, complicating the experimental paradigm beyond the reach of interpretability -- and the most likely truth of the matter is that the veto is unconsciously pre-triggered too, just as the pressing is.)
Xavier Dery @XavierDeryReplyDelete
RT @CynthiaRoxane: RT @sebrioux: Il ne peut y avoir de révolution que là où il y a conscience - Jean Jaurès #Citation #GGI #AssNat #TuringC
9:51 AM - 7 Jul 12 via Twicca Twitter app
Xavier Dery @XavierDeryReplyDelete
Mele: I sure hope not all of Libet's subjects did the experiment your way! Faking urges out of free will to mimic spontaneous ones? #TuringC
10:12 AM - 7 Jul 12 via Twicca Twitter app
can you please inform us from neuroscience latest changes where starts the neuronal process 1st : in conscious or subconscious mind?
FOr example, when I see a flower or read a simple word, where does the neuronal process starts?