Have you read that Michael Pollan article on nutritionism yet? You really should. Here’s a chunk:
Most nutritional science involves studying one nutrient at a time, an approach that even nutritionists who do it will tell you is deeply flawed. “The problem with nutrient-by-nutrient nutrition science,” points out Marion Nestle, the New York University nutritionist, “is that it takes the nutrient out of the context of food, the food out of the context of diet and the diet out of the context of lifestyle.”
If nutritional scientists know this, why do they do it anyway? Because a nutrient bias is built into the way science is done: scientists need individual variables they can isolate. Yet even the simplest food is a hopelessly complex thing to study, a virtual wilderness of chemical compounds, many of which exist in complex and dynamic relation to one another, and all of which together are in the process of changing from one state to another. So if you’re a nutritional scientist, you do the only thing you can do, given the tools at your disposal: break the thing down into its component parts and study those one by one, even if that means ignoring complex interactions and contexts, as well as the fact that the whole may be more than, or just different from, the sum of its parts. This is what we mean by reductionist science.
Does this remind you of anything? Macroeconomics, say? Economists like nothing more than to isolate different bits of the economy — housing starts, say, or durable goods orders, or initial jobless claims, or any one of hundreds of other statistical series — and try to work out how those numbers fit into what they invariably think of as the “economic cycle”. As far as I can make out, a large part of the reason why Alan Greenspan now thinks there’s a small chance of a recession in 2007 is simply that there hasn’t been one in 63 months, and therefore a recession is probably overdue. Which idea, of course, is utter codswallop, so he tries to dress it up by pointing to all manner of his beloved “indicators”.
The fact is that a manufacturing recession is a harbinger of a recession in much the same way as a diet high in monounsaturated fats is a harbinger of obesity. It might be, but, on the other hand, it might not be — and really, carving the economy up into little chunks and looking at the little chunks as indicators of what the whole insanely complex thing might do makes no more sense than carving a diet up into “nutrients” and looking at the nutrients as indicators of how the whole insanely complex body might react. Nutritionists and economists do this because it’s all they can do — no models have even come close to replicating the human body or the macroeconomy in all its chaos. But most of the time we’re all better off simply ignoring them.
Maybe the whole thing is so complicated that we’ll never understand it.
But that’s a pretty defeatist attitude. It’s a good thing they give up before they figured out what caused scurvy.
Maybe it would be better if we said what we know and what we don’t know.
We know that certain levels of some otherwise beneficial nutriants are poisonous. Even too much water can kill you.
So it’s not that we know NOTHING about nutrition, or the body, or other complex systems. It’s that in normal circumstances, we don’t know the exactly intricate interplay of forces. Same in the economy.
We do know that certain things are poisonous: Too much monopoly, too much inflation, too much unemployment, too little savings, frothy asset markets, etc.
It seems to me that our knowledge about how these forces play out is constantly improving. The improvement in monetary policy over the past decades could be one example. Why not take advantage of the advances in knowledge to improve our world?
Single-factor scientific experiments are still used by many. However they have been outdated in many applications since at least the 1920s when Ronald Fisher started publishing his works on agricultural experiments.
In manufacturing much multi-factor experimental work is based on derivatives of Fisher’s work created by Genichi Taguchi. Google analytics uses multi-factor techniques more closely based on Fisher’s work to help advertisers improve their returns.
In you posting you sound as though you are unaware of this work. Do I misread you?
Single-factor scientific experiments are still used by many. However they have been outdated in many applications since at least the 1920s when Ronald Fisher started publishing his works on agricultural experiments.
In manufacturing much multi-factor experimental work is based on derivatives of Fisher’s work created by Genichi Taguchi. Google analytics uses multi-factor techniques more closely based on Fisher’s work to help advertisers improve their returns.
In you posting you sound as though you are unaware of this work. Do I misread you?
No, you don’t misread me, but you read me very narrowly. I daresay that as you increase the number of factors in your model, the better your model will be. But things like the human body and the global economy are not reducible to even a large number of component factors.