Rationality and Positive Psychology
Robin recently had us consider the costs of rationality, but I have been thinking about the benefits. I typically think of rationality as having instrumental value, but after reading Mihály Csíkszentmihályi’s work on flow, I began pondering its status as an intrinsically fulfilling activity. Cognitive and evolutionary psychology are major components in the study of rationality, but I haven’t seen connections drawn between rationality and positive psychology before. Csíkszentmihályi (said cheek-sent-me-high-ee) defines “flow” as a state of intense focus where you lose track of yourself and become completely involved in what you are doing. After some consideration, I am intrigued by the similarities between the practice of rationality and a flow-like state of mind.
Csíkszentmihályi identifies the following components of flow:
1. Clear goals and expectations about the task at hand.
2. Direct and immediate feedback.
3. A high degree of concentration.
4. Loss of self-consciousness.
5. Altered perception of time.
6. Challenge proportionate to your skill level.
7. Feeling of control over the situation.
8. Activity is intrinsically rewarding.
(summarized from Flow (psychology))
The first component appears directly tied to the concepts of conservation of expectation and making predictions in advance. By clearly defining your goals, how you will respond to new evidence, and what your current predictions are, you will know how to react in advance. Clear goals and expectations allow you to better recognize successes and failures. A well-defined scoring function is important to guide intelligence as an optimization process, but also allows you to be fulfilled when you do encounter success.
The second component depends on the first, but it also entails finding reliable feedback to evaluate those expectations against. Careful rationalists uses precise quantitative measures and math where possible. The precision of math can give detailed and immediate feedback, so long as it does not introduce additional unjustified complexity. As rationalists, we should actively be seeking out feedback on our beliefs and predictions. Feedback assists calibration, and if correctly used, prevents us from being trapped in bad beliefs and with poor methods.
Components three, four, and five are all closely related. These three are more of a mixed bag than the first two. On one hand, this aspect of flow represents the directive to shut up and multiply: forget yourself and your own feelings and focus only on the issue at hand. However, until your rationality instincts are sufficiently honed, becoming absorbed in a task means that you could be forgetting about unconscious biases. Like any skill though, with practice, checking biases can become automatic.
The ability to automatically check biases leads into the sixth part. The challenge of a task should be proportionate to your skill level. A disproportionately difficult task leads to frustration, and an insufficiently difficult one leads to relaxation or boredom. Relaxation is not a bad thing, but is distinct from flow and antithetical to rationality. If curiosity should lead to its own destruction, we shouldn’t allow ourselves to always be in a state of relaxation. As beginning rationalists, the challenge can simply be becoming aware of our biases and the principles of rationality. As this becomes easier and our skill increases, we can focus on more and more difficult issues to apply ourselves to. I am interested in the possibility of developing simple standard challenges that allow rationalists to build awareness without being overwhelming. Reading the origins thread, it appears religion played this role for many of us, but I think the issue is too fraught with emotion to be a reliable standard challenge.
The seventh component is a little more difficult for rationalists. Being open and willing to relinquish beliefs and acknowledge mistakes seems contrary to being in control. Nevertheless, dispelling biases and reflecting on our values means that we can be in better control of our minds and behaviors.
Finally, for many of us, curiosity is an intrinsic desire. In my case, I only need to give it more of a chance to express itself.
While I began by considering whether rationality is instrinsically fulfulling, this has really been a discussion of the practice of rationality. And even then, to be careful I should say the practice of rationality is still only instrumentally valuable; just less so than commonly thought. Eliezer thinks that preferences should be neutral to rituals of cognition, which I am inclined to agree with. That rationality tends to produce a state of flow in me is a highly contingent fact. Is it possible to lessen the costs of rationality and increase its benefits?
Does anyone else have any thoughts or experiences on this subject? Is anyone aware of a more rigorous or academic study on the relation between rationality and positive psychology?
I am, by trade, a computer programmer, and flow is a common part of my day. Programming is an inherently rational activity as a computer is as strictly rational a beast as you can imagine, and does only and exactly what you tell it to. I thought it might be interesting to relate my experiences of flow to these 8 criteria:
Clear goals and expectations about the task at hand.
While this is not actually as characteristic of programming in general as one might expect, I certainly find that the periods of flow are when I have a concrete concept of what the next part of a program to implement is, and how it should tie in to what already exists.
Direct and immediate feedback.
Of all the points above, this one is the one I’m least sure of. Certainly the dictum of “Compile early and compile often” is a good one, and a computer delivers brutally honest feedback, but still its quite possible for one to be in a flow state for three or four hours before one reaches a point at which the code might arguably be capable of compiling. Certainly its not feedback from the computer that sustains the flow during these intervals. On the other hand, there is a certain amount of internal simulation of what one has just written than one does, where you run through the code in your head and imagine what a computer would do with it, and there is a definite satisfaction to getting to the end with a result consistent with what you are trying for.
A high degree of concentration.
This is definitely the case. I believe the book “The Psychology of Computer Programming” by Gerald Weinberg ( http://en.wikipedia.org/wiki/Gerald_Weinberg ) makes a good case for programming to be as difficult a mental task as brain surgery or rocket science.
Loss of self-consciousness.
Altered perception of time.
These two, I think, go hand in hand. When there is nothing but the code in your head, then there is no other perception, either of self or of time.
Challenge proportionate to your skill level.
I’m not completely convinced of this. I have gone into flow when writing yet another instance of some piece of code I’ve written a thousand times, where the challenge is practically non-existent. This might be because one gets into the habit of flow as part of programming and one gets an almost Pavlovian response set up.
I am reminded of a research study I read about where they were studying the conditions under which human beings switch from sugar burning (which is the default and provides short-term energy) to fat burning (which is how the body sustains extended efforts). They put typical folks on a treadmill and saw how much exercise it took to have them switch metabolic modes. When they tried a professional athlete, they were surprised to discover that the switchover occurred the moment they pointed him at the treadmill. Merely anticipating exertion was enough to trigger the switchover.
But, on the third hand, I do know that I use moments where I am doing rote work to reexamine how I’ve always done things, to see if there is some reformulation of the assumptions inherent in what I am doing both in the small “Is this really the most efficient way to write this loop?” and in the large “Is iterating twice over this array really the best way of calculating this result?” This bears fruit often enough that it continues to be worth doing.
Feeling of control over the situation.
When one is programming, one is in total control, for good or for ill. When it works, you are the one that made it work, and when it fails, there is no one to blame but yourself.
Activity is intrinsically rewarding.
Again, I’m not sure how rewarding the act of coding actually is. There is a huge reward in having coded something and seeing it work and express your thoughts in the way you imagined. If it wasn’t for that bit though, I’m not sure how rewarding the work would be. Its true that part of it (the problem solving) has the fun nature of puzzles, but I don’t go into flow while problem solving, but when implementing the solutions I come up with.
EDIT: I had to remove the numbers in front of the items since they kept getting auto-changed to the same number “1”.
A quick edit will greatly improve this document: In each subsection, cite the heading you’re talking about. “The first component...” → “The first component, ‘Clear goals and expectations about the task at hand’...”
I used to consider rationality and truthseeking to be a terminal (intrinsic) value, but now I consider it secondary to happiness… or Fun Theory as Eliezer might call it.
IOW, “a rationalist should win.” And winning definitely includes Fun.
Rationality is a critical component of positive psychology, because it’s what you use to get rid of irrationally negative predictions, and thus restore your brain to its naturally overconfident positive state. In other words, in positive psychology, you want to pick and choose what biases you’re going to slice apart and which ones you’ll let stand.
More precisely, you want to ensure you’re irrationally positive about rationally-derived predictions. It’s one thing to know the risk of skydiving and be irrationally positive about doing it anyway; it’s another thing altogether to irrationally expect that you can do it without a parachute!
Thus, you want to be rational about your real-world predictions, but not necessarily rational about how much you’ll enjoy (or fail to enjoy) life, whether it has any real meaning, etc., etc. Be rational about the external world, and the effects of your actions on it. And even be rational about the operation of your brain, as a brain. But if you want to have Fun, I suggest remaining irrationally positive about how good things are in general, whether life has meaning, etc.
In these areas, it is rational to be a little irrational, if your intention is to WIN, rather than to feel good about your self-image as a person who’s willing to Sacrifice All for his/her rationality. That too, is irrational.
Rationality, or the experience of understanding the world correctly probably “feels good” to certain people because that’s just the way their minds work. I, for example, really enjoy understanding things. Others find it tedious and unpleasurable, and prefer activities that stimulate them emotionally. This might explain why many people are religious: taking part in religious ritual probably stimulates them in the way that they like to be stimulated, having correct beliefs about the world is so-so.
As I understand it, positive emotions and logic/reasoning are strongly lateralized, both are centered largely in the left frontal lobe. If the same region is being stimulated, perhaps the effect stimulates positive emotions. For me, philosophical reverie seems to be accompanied by a mild buzz.
Number 2 seems to be the primary roadblock to this comparison. You say that rationalists should have appropriate feedback systems set up, but it would be very difficult to approach the kind of feedback necessary (seconds) to achieve flow when making decisions. Until we have some sort of brain-computer interface, I doubt that feedback at a fast enough pace would be possible to feel “flow” when making rational decisions.
I always found the arts of rationality to be fun to practice, but I think they would be a lot less fun if they didn’t lead anywhere. This is probably the reason why I enjoyed Raymond Smullyan’s puzzles more than textbook exercises, because solving Smullyan’s puzzles gave some fictional benefit that you could read about, whereas a lot of textbook exercises seem disconnected from reality.
I’m not sure how to make sense of your distinction between intrinsic and instrumental benefits. Could you explain the distinction?
I think I am using them in their standard senses. Something has intrinsic value) if you directly desire or value it, and it has instrumental value or extrinsic value if it is a means to achieving something with intrinsic value. Instrumental value is often only contingently valuable as a result.
Folks with old-fashioned metaphysics use ‘intrinsic value’ to mean ‘value regardless of what anyone thinks about it’ - e.g. “Gieseking is valuable if nobody ever hears it”.
That’s why I say terminal.
They could stylise positive psychology like this: plusy, plusψ, +y or +ψ