Tuesday, May 17, 2011

The Definition

“The definition of stupidity is doing the same thing over and over and expecting a different result.” I’ve heard a lot of people say that, and it annoys me to no end. The same definition has also frequently been ascribed to the word “insanity.” But it’s really the definition of neither. Stupidity is defined as low intelligence, which is approximated with the Intelligence Quotient, or IQ; someone with an IQ significantly lower than 100, or average, might be crudely called “stupid” (although other terms are considered more polite). Insanity is an outdated term for mental illness; the field of psychology discarded it years ago, and now it’s considered more proper to name the specific type of mental illness. Neither condition has ever been measured by asking the subject whether they expect a different result when repeating the same action.

But this is pedantic; most people fond of using the above “definition” will acknowledge that the behavior it describes can be thought of as stupid or insane, whether it defines these conditions or not. Perhaps, but consider the opposite behavior: doing the same thing over and over and expecting the same result. If expecting a different result is stupid and insane, then is expecting the same result smart and sane?

Think of this example, taken from an old fable: a farmer was taking a break from plowing his field, and happened to be looking at a tree stump nearby. Suddenly, a rabbit bounded out of the brush headlong into the stump, breaking its neck and dying instantly. The farmer enjoyed rabbit stew that day, marveling at how little he had to work for it, and resolved to spend his days staring at that stump, getting easy meals by watching animals kill themselves running into it. As a result, his fields never got plowed, he never planted or harvested his crops, and he soon starved to death - all because he expected the same result (free food) from the same action (watching the stump).

Now, granted, this is a very exaggerated story; nobody is that crazy. But all cautionary tales are exaggerated. They frame a common behavior in rediculous circumstances to highlight the foolishness of the behavior. People accept Red Riding Hood as a warning against trusting strangers, even though an unscrupulous stranger is more likely to cheat you than eat you. And the hapless farmer here isn’t that different from an athlete that always wears the same underwear he wore when he made his first win, or my friend who thinks adoption is a bad idea because he knew an adopted kid who was horrifically maladjusted.

In truth, both behaviors are extreme aversions of the more moderate, sensible option: waiting for sufficient information before drawing a conclusion. As every scientist knows, an experiment must be repeated many times before a reliable conclusion can be drawn. According to the stats class I failed in college, the magic number is thirty: you need at least thirty data points to get an accurate statistical analysis. So, the common phrase shouldn’t be “if you’ve seen one, you’ve seen ‘em all” but rather, “if you’ve seen thirty, you’ve seen ‘em all.”

Of course, this cautious approach can be taken too far as well. Consider a woman who keeps an abusive boyfriend around after being beaten several times (but not quite thirty), because he could change. While it may be possible for a brutal person to change, most sensible people would agree that the woman would be much better off ditching him.

So, what is correct? How many examples must one see before it is safe to conclude all future examples will be similar? It’s hard to answer this question because to do so, one must already know the answer. Each situation I’ve described in this entry is an example of a person either reaching a conclusion without enough evidence, or not reaching an obvious conclusion despite plenty of evidence. How many such examples do we need before we can reach a conclusion regarding how and when it is appropriate to reach a conclusion? Until we get a definitive answer, it’s probably best to take each situation as unique, and middle through our decisions as well as we can. Most likely, the correct answer depends on several specific factors. For instance, what are the stakes involved? In the case of the abused girlfriend, the cost of staying with an abuser far outweighs the cost of rejecting a reformed one, so it’s best to quickly conclude that things aren’t going to change. Also, is there a logical reason to connect the action to the result? In the case of the superstitious athlete, there is no logical reason to believe his performance depends on what underwear he has on, and he should probably reserve judgment until he’s had overwhelming proof. There are probably many factors which should help deptermine what conclusions you should reach and when and how.

That said, I feel the need to point out that my definition of stupidity is “repeating clever-sounding fake definitions without stopping to check a dictionary.” But that’s just my personal opinion.

2 comments:

  1. If one is making an assertion based on personal experience or anecdotal data, then he/she must take that into account. If the farmer in the example had seen a rabbit die 9 out of 10 times, then he could assert that this is normal. However, he must be willing to change his thinking as evidence mounts. It's not wrong to assert something based on a small sample size, but keep in mind that it is small and more data means better results.

    ReplyDelete
  2. *(@&#$(*&@(*^ it ate my comment. will try to reconstruct it later. grrr.

    ReplyDelete