Blog Coordinator

« Why Regress Arguments Fail | Main | The Origins of Imagination »



Feed You can follow this conversation by subscribing to the comment feed for this post.

The above was modified from a grant application that I recently co-wrote with philosopher Carolyn Dicey Jennings at U.C.Merced

Hi Peter,

Nice to meet you here, and thanks for another provocative post. A friendly request for elaboration and clarification:

As you are envisioning it, what would an instance of "the automatization of self-control" be like? In what sense would it be "automatic", as opposed to not-automatic? And how would it resemble common forms of "self-control", such that it would merit that title?

I ask because I'm worried that as soon as an instance of voluntary, attentionally demanding, effortful control over one's own act of decision-making becomes "automatic", we've shifted topics entirely and are no longer talking about self-control.

Hi Peter,

I am enjoying your posts a lot! I have a question about the contrast between automatic and self-controlled action/behavior/cognition.

I lean towards a view that both effortful and automatic behaviors/actions require will-power, albeit different amounts or strengths. If this is right, then the difference in will-power between , say, effortful self-controlled cognition/behavior and automatic cognition/behavior is one of degree, not kind. (These claims are part of a larger view "The Network Theory of Will-power" here).

It is unclear whether you would accept this. You seem to contrast self-controlled and automatic actions, but I wonder if you could say a bit more about the nature of this contrast (e.g., if it is a difference in kind or degree? how might we systematically individuate these actions?).

Thanks again for these fantastic posts!

Nick, I think there is no question that there can be degrees of willpower exertion as endogenous attentional/cognitive load varies. I think the automatization of processes can relieve the need for attentional involvement. Data show that people improve on a dual task with practice. If attention's role is the binding of representations with other representations, and the binding of representations with operations over representations, and if such bindings can 'harden' as new chunked or 'bound representations' and 'chunked processes,' then attention has functioned like a good teacher, who makes herself/himself no longer necessary, once the student has learned. Here are some relevant paragraphs from my book:

In general, most processes that can be dealt with automatically seem to be placed on the unconscious side of the conscious–unconscious divide. Some processes, like monitoring blood pressure, cannot be moved to the conscious side of that divide. Other processes that are normally unconscious, such as breathing, can be moved to the conscious side of the divide when they need to be brought under voluntary control. In general, only processes that require open-ended and endogenously controlled (as opposed to closed-ended or ballistic, exogenously controlled) computations of what to do next are made conscious. If there is only one possible sensible course of action, there is no need to make that process conscious. To do so would waste time and energy in a system designed to minimize both. It would be like having a window pop up in an operating system that needlessly states “the optimal course of action is x, which will now be implemented”; it only makes sense to invoke such an interrupt when a choice between two or more genuinely possible processes or courses of action needs to be made via endogenous consideration of options in working memory.

The consciousness associated with full endogenous attentional engagement may not be needed when multiple courses of action are possible but the optimal path can be selected on the basis of information available in the input. For example, there are multiple ways of driving down the highway, but the process may unfold largely automatically as long as nothing too unexpected happens—that is, as long as we can act according to script. There may be processes that generate possible courses of action automatically, and there may be processes that select the optimal option automatically when we are in auto-pilot mode driving down the highway. For example, the basal ganglia may select and switch to the optimal motoric (e.g., Chakravarthy et al., 2010), cognitive (Cools et al., 2003, 2004; Ragozzino et al., 2002; Ravizza & Ciranni, 2002; Ravizza & Ivry, 2001; Stocco et al., 2010), or emotional (Saint-Cyr et al., 1995; the amygdala also plays a role in emotional switching: Herry et al., 2008) action program at each moment as needed. This can happen automatically, analogous to how an automatic transmission selects the optimal gear when driving.

But when something happens that is unexpected because it has not been learned or genetically preprogrammed in an action script, the “automatic transmission mode” of action and thought selection may shift into “stick shift mode,” as actions and thoughts are now selected endogenously or voluntarily on the basis of information beyond what can be handled by any action script. That is, concrete and stimulus-driven basal ganglia action program selection (Cools et al., 2006) can be handed over to and influenced by executive cortical circuitry that incorporates information beyond that present in the stimulus, including abstract relations such as rules for behaving (Cameron et al., 2010). Say a clown skips across the highway. This event lies outside information available within the driving script, so control is handed over to endogenous attentional, planning, and selection circuitry that can decide what to do in a context-sensitive and flexible manner.

As such, the only time most of our actions require a highest level executive control (with associated conscious feelings of willing or exertion of willpower) will be cases where the action involves a decision point that cannot be preprogrammed or a conflict that must be resolved, but cannot be resolved by faster, automatic processing. For a simple action such as repeatedly lifting a finger, or even a complex one such as driving a car down an open highway, there might be no associated conscious feeling of willing even though brain circuits are generating possible courses of action and choosing among them. However, should something outside the script that is followed by an automatized mid-level manager happen, like the clown leaping across the road, control is passed to the highest level voluntary and attentionally loading executive control system (and the feeling of willing or exerting willpower may arise) as one of multiple possible courses of action are considered (e.g., braking, honking, waving, swerving to the left or right), one action plan is endogenously selected, others are inhibited, and the selected one is implemented. The feeling of willing or exerting willpower generally arises in situations where there is either some conflict that must be resolved with endogenous attentional control, as in the Stroop task, or some process that must be endogenously considered, selected, or inhibited, as in the clown example just given.

More common than feelings of willing may be the sense of authorship over an action, which could arise as a form of efference copy. In short, we may feel authorship but no effortful willing when we are in automatic transmission mode of action selection and implementation, but feel both authorship and effortful willing when in stick shift mode, facing a complex situation or one that requires endogenous attentional selection and inhibition of prepotent responses.

Given that the feeling of willing (W) may not typically arise for the kinds of nondeliberative actions that Libet and his followers have studied, the entire research program may be in need of revision. Libet’s standard paradigm, in asking subjects to report the timing of W, may be asking them to report the timing of something which they do not particularly experience. Estimated times of W may then be cognitive inferences of the sort “Given that I moved at time x, I must have felt an urge to move at x minus y milliseconds,” instead of being reports of the time when an urge to move was actually experienced.

Michael, I think all of us have automatized many executive operations to a point of 'mid-level managers' who can make decisions without instantiating highest-level executive control associated with exhausting attentional load. Please read the response I just wrote above, to Nick, before reading the rest of this.

A subset of automatized executive control processes are involved in handling the control of the emotions and desires, so count as automatized self-control. Consider, cases where the neural basis of executive control has been destroyed. The famous case of Phineas Gage is a good one. Before the metal rod blew out his frontal cortex, he was responsible, considerate and a good planner. He was in fact the foreman among the men laying tracks in Cavendish Vermont on that day, not too far from where I now live. After the accident he basically became a walking 'Id,' cursing, fighting, chasing skirts and so on. Did these ancient desire, agression and emotion circuits come into existence because of the accident? No, they became disinhibited. Was their inhibition attentionally exhausting while he was the foreman? No. He did not have to exert attentional control in order not to act like a beast, I would argue, because he had automatized self-control with regard to sex, food, agression and so forth.

It is actually an empirical question the degree to which automatized inhibition of desires/emotions requires attention. The evidence so far suggests that people get better on dual tasks with practice.

A less dramatic example would be my own kids. When they were younger they loved anything inappropriate, and would laugh over, and say the most inappropriate things. As they got older they stopped doing this. They learned that there were times when such 'toilet talk' was OK, and other times when it was not. They got to a point where we did not need to remind them to do this or a whole host of other naughty things. I would say they wired up automatic frontal inhibition over prepotent responses associated with more ancient circuitry. They did not simply grab whatever they wanted, or blurt out whatever came to mind. They learned to control themselves to a point where they no longer needed to exert any attentional effort in order to do so.

Hi Peter,

Thanks for your reply, and for suggesting that I read your response to Nick, which I did.

Since the case is so controversial, the evidence scant and unreliable, and the man is long dead, we shouldn’t quibble too much over how best to interpret the case of Phineas Gage. However, I will say that, given your sketch of the case, there is an alternative way to interpret what happened to PG. The alternative interpretation is that the aggressive impulses *did* in fact come into existence because of the destruction of those regions of his brain, and precisely because his brain was damaged in that way, the newly acquired aggressive impulses could not be controlled. Crucially, this interpretation does not involve the assumption that, prior to PG’s accident, there were automatized executive control processes involved in handling the control of his emotions and desires. But like I said, the case is controversial, so I won't nitpick, at least not too much.

About your second example, involving your kids: in general, there exist at least two ways that one can inhibit an impulse:

(a) While the impulse in question is present to conscious awareness, one suppresses it with an equivalently potent yet opposite impulse

(b) Find a way to prevent the impulse from emerging into conscious awareness in the first place

Although I have not met your kids, I would say that they have adopted the second strategy. They are not suppressing impulses or desires to blurt out silly remarks. Rather, such urges are no longer present. Crucially, since those urges are no longer present, this would not be a form of inhibition, and so this would not be an instance of self-control. This is why, in my original reply, my worry was that you have changed the subject: in the cases that you’ve described there is no self-control in view, since there is no volitional governing of one’s mental operations in the attempt to resist first-order impulses or desires so as to maximize fulfillment of one’s long-term goals or higher-order desires. Whatever might be going on in such cases, they do not seem to merit the title “self-control”.

Michael, I will reply tonight. Now I have to drive back to NH from NYC. Sorry for the delay.

Thanks Peter! That was very helpful! I wish you well!

Hi, Peter! Once again you've added an enormous contribution of reflective, quality discussions here. Wow! These threads have really been something. I really hope you'll come back here next year as well! : )

I had been waiting to see if you were going to start an actual thread on Libet stuff. But since your month is almost up, maybe I should just address the topic in THIS thread since I saw where you mentioned Libet in one of the comments above.

More specifically, I hope at some point to get your thoughts on Libet's idea of FREE WON'T. I'd also be interested in what you think about some of the experiments done in relation to the whole Free Won't/Conscious "Veto" idea that Libet came up with.

Although I've seen many comments here over the years about Libet, the Readiness Potential, the Haynes-Soon Experiments, etc., I've never really seen a whole lot discussed on the whole idea of Free Won't or about the experiments done to "supposedly" try to find it.

Frankly, the whole idea of "Free Won't" the way Libet envisioned it has never made an ounce of sense to me on either a philosophical or scientific level(In other words, the so-called idea that we're supposed to be able to "Veto" ourselves or one of our decisions at the very last moment while using some kind of independent, conscious process lying totally outside of causation).

It makes no sense.

Moreover, Libet's experiements and the interpretations given to his "Readiness Potential" have taken a real hit in recent years in terms of credibility.

Yet even so, certain Willusionist researchers (ones who love to deny Free Will----Patrick Haggard) continue to test for things like FREE WON'T (again, a ridiculous concept from the beginning) while assuring us (of course) that we don't have that either!

"There Is No Free Won’t: Antecedent Brain Activity Predicts Decisions to Inhibit"

Any thoughts on all this, Peter? It seems to me that if Libet and his experiments have taken such a hit, then trying to then find things like a "Free Won't" seem like a big waste of time. Unless of course some of these researchers haven't read the memo that Libet's results have taken such a thumping in recent years.

The comments to this entry are closed.

Books about Agency

3QD Prize 2014: Marcus Arvan