The Three Doors of Serendip:
| |
Door image from http://www.woodstone.com/photoalbum7.html |
So ... you've tried out Three Doors of Serendip (Mark II), and persuaded yourself that indeed you're better off switching as opposed to staying? If you've done some serious experimenting, you'll know now that you win twice as often (or collect money twice as fast) by using the strategy of switching as you do using the strategy of staying. And that if you randomly switch and stay on succesive trials you win about half the time, which is better than always staying but less good than always switching.
What you've achieved is what we might call "experimental" understanding, which is to say a kind of understanding which results from having an hypothesis which predicts observations yet to be made and then making those observations to see whether your prediction holds or not. Actually, its a kind of understand which results from doing these things "consciously". If you started with Three Doors of Serendip (Mark 1), you'll know that the process of coming to an understanding based on observations can in fact be done "unconsciously". So, what's the difference between "intuitve" understanding and "experimental" understanding? If they're different, what's the relation between them? "Experimental" understanding is sometimes referred to as "scientific" understanding, implying that it is somehow better. Is it? Always, or only in some cases? If ever, why?
To explore those questions, let's think a bit about what you did to achieve "experimental" understanding. First, you had to have at least some intuition about how best to play the game, at least one "strategy" in mind. Then, you had to have some alternative strategy in mind, so that you could design a new set of observations to compare the results of the two. Then, you had to have made enough observations to be certain that effectively compared the two. Then you had to decide whether there could in principle be some OTHER strategy that might be better than either of the ones you tried. Only after being sure that all possible strategies had been tested could you be sure that you had adequate "experimental" understanding.
Two points follow from this:
You needed to have in mind some candidate strategy before you could design a set of observations to test it. Such a candidate strategy might have been in your mind because some one else suggested it, but more often than not, its there first as a intuition. Actually, of course, you had to make explicit at least two strategies, either or both of which may have been "intuitive". Above and beyond this, though, the process of achieving "experimental" understanding depended on some other "intutive" understandings. Among these was a "sense" of what an experiment is and how to design one, of how many observations you needed to be able to draw a conclusion, and of having actually tested ALL possible strategies.
Hold on, you might say at this point. Those aren't "intuitive" understandings, they're "rational" ones. Yeah, well, to explore that one, you'll need to go through the Serendip's third door, where we'll explore the meaning of "rational" understandings, just as we are here that of "experimental" understandings, and did earlier that of "intuitive" understandings. For the moment, its enough to recognize that the interpretation of "experimental" observations is fundamentally dependent on some OTHER kinds of understandings, which aren't usually made explicit, and certainly not themselves tested in the observations of the experiments themselves. How sure are you, for example, that there are really one three possible significantly different ways to play the Three Doors game? Maybe, in addition to always staying, always switching, and randomly staying/switching, there is some fourth possible strategy (or an infinite number of them) that might in fact prove experimentally to be better?
"Intuitive" understandings may arise from "unconscious" experimentation. They may also arise from a host of other factors of which one is not aware. The important point is that, however they arise, they are "understandings", ie beliefs about what is "right" whose origins and rationale are opaque and hence whose validity is not challengeable. The most important things about the processes involved in achieving experimental understanding is that they start form the premise that something one believes to be "right" might in fact be "wrong". By making an "intuitive" understanding conscious, one not only makes it explicit and communicable to others, one also achieves the capability to consider it to be "wrong", by virtue of imagining alternative understandings. This initial stage of achieving "experimental" understanding is central to its advantages over "intuitive" understanding: by making the intuitive "conscious" one can achieve both skepticism and the creation of alternative possibilities.
Making "intuitive" understanding explicit, and hence both communicatable and falsifiable has another great advantage. It makes it possible for efforts to achieve greater understanding to become social activities, rather than the solitary activities of individuals. One can share candidate understandings, and share efforts to evaluate them by making significant new observations. This is, in fact, the basis of the success of "science" as a method of advancing understanding. Multiple people can entertain multiple possible "understandings" and can base their own choices among them on observations made by other people. And since the "experimental" understandings have survived explict and conscious testing by observations, one knows at least what set of observations they represent and hence under what circumstances they can be trusted. And one knows that they too could become "wrong" given some future set of observations.
Link to reaction.html
Make clear that experimentation can be done unconsciously
get James, "not enough regularity" in here?
be explicit about when expt understandings better (not when want truth/reliability), yes when can be reasonably sure of possible alternatives
note that just as intuitive influences experimental so experimental influences intuitive (will react differently to choice games in future).
"Hands on" understanding intuitive, unconscious | "Experimental" understanding conscious, observational | "Rational" understanding conscious, analytical, logical |