« Medical Ideology | Main | Silly Consensus »

April 05, 2009


From a submission wrestling perspective, "choke to submit" seems fairly obvious.

That seems like a very speculative account -- might it be more parsimonious to just connect it to more familiar phenomena like the fact that people are more risk-averse on gains and more risk-seeking on losses? (Where "choking" is just the natural performance consequence of risk-induced fear.)

We could confirm this relationship by seeing if the same kind of "choking" happens the other way around: if people choke more, or less, when they are threatened with a 400 rather than a 4 rupee penalty for failing. And if so, then we'd need an evolutionary explanation for why people get scared when they're risking a big gain, or get desperate and reckless when facing a big loss, etc., not some kind of submission thing.

(By "if so," that is, I mean that the relationship would be confirmed if people choked less on the 400 rupee punishment -- since that would confirm that the performance losses were tracking induced risk preferences.)

I have absolutely no evidence, but the story I would have told is something like, "large sums at stake register as a large risk, and thus trigger our fight-or-flight response. Since our neural systems evolved when in most danger scenarios it was more important to take some action quickly than it was to take the best possible action, the fight-or-flight response drives us to take an ill-considered action quickly and thus makes it hard to stop and think clearly about the right answer." I'm making three or four factual claims in there that I have no particular evidence for, but it sounds right and matches my introspective experience--anyone actually know one way or the other on any of it?

Why do we experience vertigo when looking down from heights? Fear I can see, but why vertigo, the worst possible reaction? It seems like a similar but even more bizarre question.

On further reflection, I would test Robin's hypothesis by seeing if people are less likely to choke in front of others believed to be of lower rank - say, if no one is (believed to be) watching except children.

Eliezer, re your proposed test, see my added to the post. Re vertigo, I have no idea.

Paul, these tests can be framed as gains or losses; I doubt such framing would make much difference.

Jadagul, fight-or-flight seems plausible, but also seems somewhat compatible with my story.

Jayson, you lost me.

I don't see why there needs to be a reason for choking - it could just be a consequence of human physiology*, as described by the Yerkes-Dodson law. Arousal influences task performance with an upside-down U-shape, meaning that there's an optimal level of arousal for each task - people perform worse if they are too calm or too agitated. Because arousal facilitates the dominant response, the optimal level of arousal is lower for difficult, complex, novel tasks and higher for easy, simple, familiar tasks. High incentives just produce too much arousal.

* And it's not unique to humans - the same effect has been found in other species, including cockroaches

Eliezer - when I was on a ski lift a long time ago, I started thinking about its safety...the chance of it collapsing was very small, I thought, so the only thing that could go wrong was if I got scared of the height and fainted - at which point I did actually start feeling intense vertigo and anxiety and thought I might actually faint! I managed to overcome it, but it was really puzzling for the reason you mentioned: why would my body decide to enter the very state that could do me the most harm?

The only thing I can think of now is the "Doomsday device" scenario put forth as an EvPsych explanation for maladaptive behavior like suicide or homicidal rampage - the triggering of the reaction doesn't benefit the person who carries out the behavior, but it does help copies of the "gene for suicide" by giving credence to the pre-suicidal threats of others. Perhaps knowing that vertigo is the reaction your body will have - knowing your body will betray you - will make you even less likely to seek out heights, more so than if the reaction were "only" fear, which you might master.

I have a feeling that choking under large reward conditions could be explained by the Doomsday model in the same manner.

If you choke around a certain type of person, then try raising the status of other people to the level of the choke-inducer. It may be counterintuitive, but i find that this works as well as lowering the status of the choke-inducer and has the benefit of inducing a more positive outlook in myself. I raise status by recognizing the worth, dignity, and beauty of people. This works for me.

@Robin re Jayson,

For anyone who spends more time thinking about Brazilian jiu jitsu (or similar pursuits) than social signalling, your title evokes submission training rather than evolutionary tricks to avoid showing up the boss.

One further point of clarification -- in submission wrestling, "to submit" is typically a transitive verb used as a shortened form of "to force someone to submit", e.g., "I submitted him with a rear naked choke." Your title evoked this usage, not the usual intransitive usage.

Perhaps it's a simple numbers game. When facing superior numbers it might be best to show your belly, signal insecurity, etc.

Flopsweat strikes the new teacher in front of the kindergarten class, too.

"it was dangerous to do well on an important mental task in front of a large group,"

Welcome to my childhood.
There was a time when I didn't have this instinct at all, I eventually learned to dumb myself down, I'm not sure how much of that I really internalized.

Is it possible that a lot of high stakes choking is due to conditioning from childhood, and not instinct?

In natural environments, great heights are fairly rare. Staying away from them is probably a better strategy than performing well near them. Instant fear will get you away temporarily. Feeling like you almost DID fall off may keep you away better.

Notice the weasel words so I don't appear overconfident?

Added: I first overcame my fear of public speaking by thinking the audience was beneath me, and my fear of asking women out by thinking they were not as far above me as I had previously thought.

A good way to do this might be to read news websites for women like jezebel.com.

Why do we experience vertigo when looking down from heights? Fear I can see, but why vertigo, the worst possible reaction? It seems like a similar but even more bizarre question.

I've noticed that I'm affected by vertigo when I look up something--a tall telephone pole, for instance. Perhaps vertigo was a way to discourage our ancestors from climbing up rock faces.

Perhaps related to self handicapping? If you fail obviously enough that it seems due to shyness or some reaction to the situation, you don't have be judged on your capability at whatever the task is. Seeming shy seems less bad than seeming wrong or stupid. This seems true for me, at least partly.

Status of viewers making a difference seems compatible with almost any hypothesis, as it means the stakes are higher, which is what we are already looking at. More anecdotal evidence: I find talking to people below me (including children) enjoyable, but often can hardly think talking in front of even one person whose opinion matters to me.

More anecdotal evidence that doesn't fit with Robin's hypothesis, but does mine: at high school I did well in sciencey things and the head of science dept raved about what a great student I was. In his class later on I was paralyzed with nervousness, and did badly. Can't be due to possibility of unseating higher status others, as already the recognised best. Also neither science ability nor being liked by science teacher are tickets to much general status at high school. Seems more likely that I couldn't stand possibility of not being best at something my self worth required, or of not being liked by one person who did, so couldn't bear to test self. Failing spectacularly means he thinks there was something wrong, rather than that I'm a bit less great as his expectations have me.

Pete: "the triggering of the reaction [vertigo] doesn't benefit the person who carries out the behavior"

Yet arguably it does, if it keeps you away from heights in the future. The risk of death right now might be outweighed by that payoff later.

Added: I first overcame my fear of public speaking by thinking the audience was beneath me

I wonder what this says about people who never had any trouble with public speaking.

I, um... can't actually remember ever choking because other people were watching me. Um, ever.

(On my first occasion in public speaking at a Foresight Gathering in 2000, I remember thinking, just before I went on stage, "So this is when people are supposed to panic? Well, I don't see how that helps anything so I'll be sure not to do it.")

Unnamed, we can ask why performance vs. arousal is U-shaped, why we evolved to let arousal get past the peak, why arousal pushes the dominant response, or why doing well on mental tasks isn't the dominant response.

Pete, James, John, and Allan, yes vertigo might helpfully push us to avoid places we might fall.

John, my claim was about relative status; so yes you can raise yourself or lower them.

Fenn, kindergarten teachers know their performance will be evaluated by parents and colleagues.

Katja, I doubt your being the best science student made you feel like the highest status person in your high school science class.

Eliezer, I suspect you already thought of yourself as high status then relative to your audience.

Robin: I think Unnamed's post brings some evidence favoring my hypothesis (alternatively: my hypothesis is a just-so story for his evidence, presented before I'd seen his evidence). You'll note that the more purely physical and the more drilled an activity is, the more delayed the peak of the U; there's at least a theory floating around (mentioned in the Wikipedia article) that the inverted U is a combination of 2 factors: increase in performance due to arousal, and a gradual shutdown of higher cognitive processes.

Also, Malcom Gladwell wrote an interesting article on choking a while back. On reflection and rereading the article, I might be talking more about what he describes as "panic," and choking might be something completely different (and might well be what you're explaining).

For some reason, I always found public performances to be thrilling rather than frightening. I suspect that, physiologically, it's probably a very similar response to the stereotypical "stage fright", but I interpret it differently.

"Fenn, kindergarten teachers know their performance will be evaluated by parents and colleagues."

Fuzzy. So now the source of nerves doesn't lie in the act of speaking before the audience?

"but it does help copies of the "gene for suicide" by giving credence to the pre-suicidal threats of others"

going to this kind of positive feedback loop would pretty much invalidate evolution all together.

There is a well known effect in sports of the "contract year." This is the year a player's contract expires, and if he perform particularly well, he may reap considerable rewards in his new contract. Players frequently do much better in a contract year than in the following year, post-contract. This would seem to contradict the study except that (a) these are physical feats, not mental, and (b) perhaps professional athletes are more likely to be the type of people that are better able to rise to the occasion.

This post provides interesting food for thought. On the one hand, the Solomon Asch social conformity studies--now rather dated--come to mind. In those experiments, subjects more frequently repeated obviously incorrect answers to questions after a confederate gave an incorrect answer. The test subjects demonstrated they knew the correct answer, but were apparently willing to self-handicap to conform their opinion to that of the confederate(s). On the other hand, if one accepts that humans are cognitive misers, a conflict between or overload of systems 1 and 2 cognition may give rise to "choking," as in attempting to appear calm and composed while performing a demanding task in front of a group. Finite cognitive resources may simply be spread too thinly. Finally, what appears to be choking may in fact be a case of reversion to mean performance--a so-called winner's curse.

Robin, while that was admittedly what I was hinting at in a joking way, I suspect that there are people who have a kind of permanent adequate-status indicator, what I believe is sometimes called "self-confidence", so that they won't choke up in front of an arbitrary high-ranked audience. People like this have others who they think are awesome, but they wouldn't quite fear them enough to choke up. I think a lot of entrepreneurs have this, even though they consider others as genuinely above them in the universal lattice.

Again, it seems like a lot of you are making evolution capable of explaining anything (and therefore nothing). For example, in the vertigo comments, if you can explain choking as making you avoid situations where you'll fail, then there's not much evolution can't explain. Any potential observation could be rationalized as helping the gene.

I also wanted to point out that the exact opposite effect was observed in the infamous case of Joshua Bell, the violinist who started losing confidence when he played for random passers-by, and who stated how much more confident he was playing for (high-status) people who shelled out a lot of money to see him. Of course, in that case I see it as a case of the Emporer's New Clothes ... the music's greatness is due to the illusion of thinking that other people think it's great, while in a blind test no one's actually willing to listen to it.

@Eliezer et al:
"Why do we experience vertigo when looking down from heights? Fear I can see, but why vertigo, the worst possible reaction?"

Maybe the vertigo mutation dates a long way back: individuals who experienced anything less destructive than vertigo would overcome simple fear and climb up the trees to pick juicy, nutritious fruits, taking a different evolutionary path that didn't lead to Homo erectus & C.
Possibly becoming extinct later due to climate changes etc.

Please define success in life.

Vertigo makes a lot more sense if it's not something evolved for specifically but is rather an accidental response to an improbable situation. Straight-down cliffs of 100 feet or more must have been pretty rare in the ancestral environment, and the height at which people experience vertigo is significantly higher than the "you probably die if you fall here" height.

Vertigo happens because part of your brain is saying "that way is down" and another part is saying "that way can't possibly be down, the ground is too far away".

I think choking is a death spiral that starts with being nervous about making a bad impression (which in the ancestral environment might lower your status in the minds of everyone you know, potentially having catastrophic effects, such as making you undesirable to all females in your tribe [if you're male]). It becomes a genuine feedback spiral when thinking about your nervousness and the potential bad outcomes and possibly focusing your attention on the physical symptoms of distress cause you to become even more nervous, which causes you to perceive the danger to be greater, your symptoms to get stronger, etc...

The physical symptoms alone impair performance for any task having a high cognitive component. In a situation of perceived danger (to your body or status or ...), blood in the brain is shunted away from areas of higher cognition and all areas not related to basic survival and routed to muscles in preparation for fighting or fleeing. The resultant hit to cognitive ability augments the performance impairments that would already occur due to inattention as a result of focusing so much on the physical symptoms of anxiety and the potential bad outcomes.

Robin, exactly - seems irrelevant to status. I meant 'if status related, still shouldn't be problem'. But still choke, so seems to only require potential loss, not potential status change. So needs non-status-dependent explanation, unless I'm just a weird exception.

An interesting question. You might get some insight by looking at studies of stutterers/stammerers, i.e. people who literally choke on their words.

While we know stuttering is physiological, there is a psychological/situational component to the severity of the symptoms. Most stutterers can be completely fluent when they talk to themselves, their pets, small children. Many find the pressure of public speaking or dating makes their problem worse. Some find when they are acting, faking accents, or speaking to complete strangers they can be fluent.

To my knowledge there are no convincing studies of why, most theories simply suggest that stressful situations 'overload' the ability of the speech system to work as intended. But it's not this simple, because certain kinds of stress have a beneficial effect.

We didn't evolve in situations where we needed to be able to perform complex cognitive tasks under extreme pressure. It was much more adaptive to curl up into a ball and whimper in the corner (i.e. show submissiveness to the guy who would otherwise kill you) in situations where you weren't dominant.

Anybody who was ever a nerd in childhood knows that the ability to think of witty things to say under pressure is much less adaptive in the face of bullies than the ability to cower and hide.

Robin, the difference with the performance-arousal U-curve is that the answers to most of the questions aren't specific to humans, and that the poor performance is a consequence of a process that often improves performance. If there's an evolutionary story it goes back to our distant pre-primate ancestors (like Jadagul's fight-or-flight story). This doesn't mean that status is completely irrelevant - arousal will generally be higher when the stakes are higher, and a high status audience could raise the stakes just like large financial incentives do. But it seems unlikely that we're dealing with an evolutionary adaptation based on status, and especially unlikely that it's an adaptation to perform worse that's based on status.

I'm less confident about the answers to your specific questions than the rest of what I've said, but I'll give it a try. The dominant response is something like the process that follows the strongest pathways in the brain. If doing the task is at all difficult or complicated, and not well-practiced (or built-in by evolution), then successful performance will require branching off into less obvious alternatives, overriding your first instinct, and so on. That means that favoring the dominant response will hurt your performance. Arousal may have evolved to favor the dominant response because high-stress situations were important ones where an animal benefited from acting decisively (as by fighting or fleeing). The peak is different for each task, so if evolution capped the maximum arousal level it would be at a level that is past the peak for the kinds of tasks that we've been discussing, like public speaking.

Another possible explanation might be fear of loss aversion, or just plain loss of aversion probabilistically. So, suddenly I have this opportunity to make a lot of money. But, now that I have this possibility, I know that I will be really unhappy if I do not get it, as I will have already built some expectation of getting into my psyche. Would simply rather not have this opportunity and thus the fear of the pain of losing (not getting) it.

"Anybody who was ever a nerd in childhood knows that the ability to think of witty things to say under pressure is much less adaptive in the face of bullies than the ability to cower and hide."

Both of these options are indicators of the failure of the alpha-wolf or avoidance strategies. Better to be the bully, beat up the bully or, failing both, avoid the bully--than be the bully's target.

Another possibility: maybe we like to orchestrate silent coups. Maybe we want to impress, but we don't want our methods to be known, thereby eroding the perception that our authority came about effortlessly and naturally. This would also help cover out tracks so that our tactics couldn't be used against us later. So we hide our effort and shroud ourselves in an aura of infallibility by not betting our reputation on small ventures and instead staying prepared and making decisive bids for better positions. So it isn't out of lack of confidence that we maintain the status quo, it is out of a desire to force others to internally justify our ascent to higher status thus making us more secure in the new position.

I guess this only makes sense if we see the group as hostile or dumb. I think Robin's conclusions hold more water in modern society.

Also, regression to the mean doesn't seem like an adequate explanation. Mathematically it must be true, but it doesn't really tell us anything about causes.

Consider another piece of conventional sports wisdom: the tactic of "icing" an opponent by calling a timeout before an important play (i.e. a potential go-ahead free throw or field goal). This is usually done with the explicit goal of "making him think about it," the hope being that an extended time to consider the impending task (and the presumed increase in anxiety this creates) will lead to poorer performance.

The fight-flight explanation for choking doesn't seem able to handle this, as it would predict that coaches would want opponents to have to react to high-pressure situations as soon as possible to maximize choking. This example still may reinforce the Yerkes-Dodson Law angle, though, because of the important distinction that this is a simple physical and not a complex mental task. Perhaps an optimal level of arousal for a familiar mechanical task like shooting a free throw would be achieved by an immediate reaction. Calling a timeout could interfere with this, not so much by increasing anxiety and nerves, but by decreasing arousal--killing the athlete's buzz. The problem I see with this is that anyone who has ever shot a free throw knows that subjectively it feels like you'll perform better if you're calm, not excited--ergo rituals like a certain number of dribbles or a deep breath before shooting. (Not to mention the western movie truism in the post.)

This post and many of the comments seems to take as a given that evolution had to have some good reason for the way humans behave. Perhaps it is just a random "bad" trait that has survived that has not been bred out yet. Or, perhaps natural selection has been lessened to an extent, and so this bad trait is on a comeback.

That being said, it is logical to argue that choking is correlated to not trying to be the alpha male/female, which could mean one is more likely to survive. 5 studs with no fear and a desire to be the best would end up fighting to be the one alpha male, and would have a 20% chance of survival, where as people in a tribe who wanted to be competent enough to justify their existence, but not too successful to be deemed a threat, would likely have a higher survival ratio.

The comments to this entry are closed.

Less Wrong (sister site)

May 2009

Sun Mon Tue Wed Thu Fri Sat
          1 2
3 4 5 6 7 8 9
10 11 12 13 14 15 16
17 18 19 20 21 22 23
24 25 26 27 28 29 30