Hacker News Reader

> Well this argument is pretty disturbing… besides the tiny and very neat island of abstract and platonic constructs every concepts about the real world are fuzzy. If someone followed your idea completely he could not use such concepts as “green” or “blue”, or even more relevantly, such an argument can also be applied to the concept of “species”.

What's so disturbing about that? Species are also on a continuum. I think there's a popsci piece about certain lizards in California that inhabit a horseshoe shaped habitat. If your abstraction is wrong, change your abstraction.

> I also find this assertion… strange to say the least.

Historically people have had horrible things happen to them due to what race they were classified in. Now let's use those classifications to create a diverse campus, thus ensuring the continuation of the racial classification system?

If you want a bunch of people who think differently, and you're a university, there should be a more direct way to do this than using some classifier that's both biased and has negative historical connotations. Here's a couple of ways: interview the students to assess their thinking. You're a university right? The kind of place where thinking people go to practice thinking. Number two is a lottery. You want a representative sample? There you go.

> ...some of those biological differences could lead to behavioural differences...

If I want basketball players, should I recruit in the Netherlands, where people are really tall, or should I just recruit people who are good at basketball?

The biological differences are tiny and insignificant for human populations if compared to clusters based on environmental differences. An urban born white student has more in common in behavior with an urban born black student than it has with a rural born white student. If we go past borders or land masses, you can have just a few miles difference and have two individuals with complete different culture, values and language.

Beyond that, if you know nothing of an individual and wanted to determine their culture, values and health, the highest predictor would be neither where they live or race. It would be social economic status. Nothing beats clusters of high social economic status in determining behavior and outcomes. It is for example why crime research generally ignore race now days in favor of social economic status, with race seen as less scientific. Universities does the opposite which is a bit ironic.

>The biological differences are tiny and insignificant for human populations if compared to clusters based on environmental differences.

You meant “the biologically induced behaviour differences” didn’t you? Because otherwise in general it’s trivially false. It’s the reason why actually blacks in the USA suffered during decades from being prescribed drugs whose safety were nearly exclusively tested on white males.

In medicine there is many examples where genetics and environment create a difference for people. A person can have a multiple higher risk of diabetes just because they were born by parents with thrifty metabolism, caused by them or their parents eating a non-western diet.

This doesn't mean there are significant biologically different from any other human. As a species, we, humans, have a natural wide range of adeptness to the environment. Some of that can carry with people for centuries or even longer.

If we compared to what is common for all humans those differences are tiny and rather insignificant. Compared to social economic status, race differences are much smaller and tiny. If a scientist did a study comparing medical outcomes of a drug, selecting top 1% wealthy whites and compare to bottom 1% poor black, the results would not say much about how effective that drugs is based on race.

> I am pretty sure that you could yourself instantly tell apart a black albinos from a white just using the craniofacial differences.

You can. This is called anthropometry. Many measurements are highly correlated with race. Widely used in all sorts of situations. For example, in legal medicine it helps to identify bodies.

Not every real world concept is fuzzy. Some things are measurable and thus objective within a defined margin of error. For instance: annual income, height, weight, age, chemicals in the blood etc. I'm not saying we should use X or Y for admissions, simply that using non-objective criteria invariably leads to problems like what is described in the article.

I do not agree with your perspective on “objective” measures. I will take age as an example. Like every other points of your list it is a fuzzy concept. However it looks rigorous and well defined because:

1 the scale of numbers we usually use is sufficiently large to smooth away the fuzziness at the origin.

2 Age is usually artificially quantised in years. This seems transparent because a year is in itself a very precise measure of time but a lot of information is lost in this quantisation.

3. In its administrative form it seems well-defined because it actually is a platonic construct. The fact that it is measured since someone is birthed and not for example since its inception is totally arbitrary in regards to a lot of uses it’s made for. If someone was born after seven months instead of nine months, he will arbitrale counted as older.