Archive for February, 2009

on the flaws of universal standards

February 5th, 2009

We are accustomed to ethical systems where rules apply universally to everyone. Such systems are everywhere: the legal system of every country, road regulations, the rules of just about every social organization. Universal rules are the only practical solution to the problem of anarchy in most cases. They are also based on the assumption that the greatest order and justice will be achieved through universality.


To see why this is a flawed assumption you need look no further than the classic teen movie about the kid who gets bullied. The standard response from the authorities is always the same one: "report it to us, we will fix the problem". So they sit down the bully for a talking to, and of course it changes nothing because the intervention is too light, it has no effect on him. Why doesn't it? Because the bully [in the movie] is a lot more self confident than the average kid, so an intervention that would have taken effect on most kids doesn't make a dent in him. The guiding principle is that the an action has a universal appropriate response associated with it.

Parents discover the same thing. The ideal model is that all their kids get the same treatment and play by the same rules. No preference for one over the other. But the reality is that kids are different, and they need different stimuli. A kid who's thick skinned won't respond to a light reprimand, whereas a sensitive kid will be overwhelmed by the same harsh beat down that would have been appropriate for his robust sibling. One size does not fit all.

Most clearcut of all are monetary incentives. If one guy gets pulled over for speeding in a beat up old car worth 1000 bucks and another guy gets caught in a 100,000 sports car, should they pay the same fine? In the interest of fairness, yes. But if you're trying to prevent speeding, then each should get a fine that will be high enough to set a sufficient incentive. If you make them both pay 100 then the guy in the sports car doesn't even notice it and let's get real, people don't buy sports cars out of a strong dedication to the speed limit.

Universal rules are in a sense the least bad solution. And people intuitively accept them, because even as they feel judged harshly they can take solace in the fact that everyone gets the same treatment for the same infraction. But that doesn't mean universality promotes that most harmony. Sometimes people even act out in protest against being treated as replaceable cogs in a machine, and all it takes to pacify them is a little individual attention.

Academic performance

Another area where universal standards rule is academic testing. The idea is that if you give every student the same test then you can determine how well each person has absorbed the same material. Of course, what you end up measuring in part is test taking aptitude. But there is a more serious problem with this. Ideally, what you would like to measure is not so much knowledge of this specific material, which may be obsolete a few years down the road. Instead, what a potential employer would be interested in is learning aptitude. And taking it one step further, perhaps even the aptitude to learn how to learn.

To clarify this point, suppose you are taking a class on discrete mathematics. It's entirely possible that you won't ever find a use for this knowledge in your professional life, so to determine your knowledge of discrete math isn't particularly interesting. What's more interesting is to have some metric of how successful your learning process was. So if you knew all the material beforehand, you would score a zero. And if you think about it, that is a far more just way to measure performance. A student with no relevant background would get credit for the work he had to put in to compensate for his deficit.