Tuesday, July 15, 2008

Layers



Bloggers make a lot more sense when they blog a lot. I don't make sense, but suspect it would be worthwhile for me to start blogging more often. I do post fairly frequently within a social network called Ned. For many reasons it's hard to take stuff from one context and place it in another. Blogs seem to have a fairly inclusive context and social networks more exclusive. So it's a matter of finding the right pattern of correspondence. I suppose I won't find what's the right sort of stuff to put here unless I start putting stuff here again.

The many blog posts I read, and the conversations generated by them contribute much to my day. Among the bloggers I love reading is Phil Jones and his several blogs. Ethan Zuckerman has generated a great conversation about homophily; i.e. birds of a feather flock together, on the Internet. It's quite a stretch to say that Phil thinks like I do. Phil Jones is a computer scientist, I am so not. Then again, a recent post of his Why Geeks are doomed in a Suit world raises some points in one context that I'd been thinking recently about in another context.

Phil begins his post pointing to a critique by Joel Spolsky of Java-oriented Computer science education. Spolsky thinks that some culling of students from CS studies is necessary, not only for the future of Computer Science, but also for the happiness of would be computer scientists. Some people will never be happy in the field. I knew that for myself never even having to venture into a demanding CS class.

If you know about computer programming, I urge you to read Phil Jones. I don't know anything about it, but this caught my attention:
The Geek's job is to make the abstraction levels fit together. Of course the only way to achieve this is with an understanding of both the higher and the lower levels : simultaneously. And the easy shifting from one to another.
Via 3Quarks Daily a fantastic blog where I get a steady fix for my xenophilia jones, there was a link to an artcle by Paul Ehrlich in Seed. Cultural Evolution has the snippet at 3Quarks along with the smart comments there, and Cultural Evolution: Does human culture evolve via natural selection, as our genes do? is the link to the whole article.

I have enormous respect for Paul Ehrlich as a scientist and public intellectual. But this article seems so wrong and sloppy in so many ways. That alone probably wouldn't have bothered me much except for Ehrlich thinking that the article describes good science towards answering his question: Does human culture evolve via natural selection, as our genes do? I emphatically don't think what he describes is good science. That puts me in a peculiar position, on one hand is Ehrlich, an esteemed scientist and on the other is me, some blogger in a basement. Who will you believe?

Oh boy, where to start. Well, Phil's observation about making abstraction levels fit together very succinctly gets to the nub of my disagreement with Ehrlich's study. Ehrlich is interested in the question of whether human culture evolves via natural selection. Ehrlich begins:
If we wish to understand and predict culture change generally, viewing culture as an evolving trait in analogy with genetic evolution is a very useful place to start.
This seems very reasonable to me. But as Ehrlich proceeds, it seems obvious that he's not very careful about the in analogy part. In analogy seems a very particular level of abstraction. It's one thing to say, "Her hair is as soft as a mink's coat." and quite another to begin collecting her hair to make a mink coat.

Ehrlich's article proceeds to tell the story of some of the ways that cultural evolution has gone off the rails into absurdity by following too closely on Darwin's heals. I suppose that's necessary to assure us that he had no intention of falling into the same pits. He asserts:
There are clear patterns in cultural evolution, which are just as prevalent as those in genetic evolution.
Then he proposes that what's necessary is a simple hypothesis which can be tested. Rather implicit in what comes next is that Ehrlich has forgotten entirely the in analogy part; that is, his hypothesis presumes that natural selection works as the engine driving cultural evolution. My head is beginning to spin because I'm not sure what culture means to Ehrlich, for example on what scale, i.e. French culture as distinct from US culture. I'm also not sure at all what exactly culture is analogous to in genetic evolution. Is it analogous to the phenotype of an organism, to a species, to an ecological system, or what? Apparently I'm dense, because there is no attempt on Ehrlich's part to make this part of the analogy clear. Is it obvious? If it is, then surely the analogy must be that culture is a kind of species. Without thinking too much about it, that analogy will break down for many purposes. In any case at this point in the article a warning is flashing in my head: Abstraction level confusion!

Here is the hypothesis that Ehrlich came up with along with Simon Levin:
We believed that the evolution of technological norms would generally be more rapid than that of ethical norms.
It concerns me that "technological norms" along with "ethical norms" seem hard to rigorously state. What exactly are they? Can you look up a list of those somewhere? But the main problem is there is no attempt to connect what "technological norms" and "ethical norms" might be in cultural evolution in analogy to genetic evolution. What are they like? Perhaps organs of an individual of a particular species?

Ehrlich then describes his false starts in looking for data that might be used to test their hypothesis. He writes:
Unfortunately, ethnographies are written more like novels than scientific studies, and it is impossible to be certain of all the crops grown and the counts of deities.
My first thought was that was just a gratuitous zinger directed towards anthropologists--They're not real scientist! Unfortunate that Ehrlich seems not to wonder why he encounters stories, and finds it justified merely to judge them unfortunate. Where there are stories it's a sure signal of relevance. The judge inquires: "Quilty or Not Quilty?" It's no good to reply: "Your honor, I have long hair."

Ehrlich was determined to ignore all complications, and locates some data about the canoes of Oceania. With that data Ehrlich has a colleague analysis it into functional traits and symbolic traits. I'm wondering what are canoes to culture in analogy to genetic evolution? What do symbolic decoration on canoes tell us about "ethical norms" and how would we know? I have several more questions, but it appears that Ehrlich is content just to have these two invented categories of canoe traits to test his hypothesis.

The conclusion after studying changes in canoes over time is the decorations change much more frequently than the functional traits; ergo
The original Ehrlich-Levin hypothesis was rejected.
Even at this point I would have been happy with the piece if Ehrlich had stopped to wonder even a little how two famous ecologists could have muddled things so thoroughly. But no, Ehrlich concludes:
We directly tested a theory of cultural evolution. Our work has helped to uncover a piece of the larger, more complex process of culture change and has shown that it is reasonable to think of that change as evolution.
How has this research shown that it is "reasonable to think of that change as evolution?"

At every stage of this inquiry it seems to me that Ehrlich has paid little attention to levels of abstraction. And then having produced a muddled mess shouts "Proof!" It seems nuts to me, really nuts.

It's too bad that Phil's bosses don't appreciate where he's coming from. He does a good service in pointing to a reason why. I'm lucky I'm not a geek;-) It is quite upsetting to read an article by an eminent scientist discussing inquires into the nature of things with another very eminent scientist, enlisting colleagues from a prestigious University in a terribly flawed study and that along the line nobody said this is BS.

I flunked out my first go round in college. Logic was one of those culling courses Spolsky refers to, and I got culled, twice. So I'm not so sure about my abilities in handling abstractions. What I am sure is I expected better from Paul Ehrlich!

2 comments:

Steve said...

I enjoy reading your work John, it challenges me to think.

It seems that what you describe here may a classical case of hasty generalization. Let me take some time to read the original article and see if I agree.

Daisy said...

I didn't like that Spolsky thing at all. Maybe its because I only understood about half of it.

Or it could be, I could feel his rather unapologetic contempt for people who only understand half of it.

He's the type of person, if I demanded he explain the postmodern interpretations of Biblical imagery in THE SOUND AND THE FURY, he'd sneer that literature is fucked, and he is doing IMPORTANT STUFF. Meanwhile, the English majors (and everyone else) regard geeks as the people doing the grunt work of the world these days: Just get'r done, so I can use the thing, okay?

Moral of the story: It would be nice if we all just respected each other!

(Afraid to look at Ehrlich, now that I read that one. :P )