« Gender Genie | Gene Expression Front Page | Neoconservative shaman for our age »
November 02, 2003

The two methods

A few weeks ago I read an old review (PDF format) of several books (Race, Evolution and Behavior, Human Biodiversity, The Bell Curve and The Evolution of Racism) by occasional Gene Expression reader & anthropologist Henry Harpending. This section caught my attention:

Two polar ways of understanding the world are induction and deduction. Induction, generalizing from data to principles, is the foundation of scholarship and of most social science. Deduction, the formulation of models and the attempt to falsify them by comparing the prediction of models to what is observed, is the foundation of natural science. Natural science is an aesthetically barren way of understanding but it has led us to bridges that stand up and airplanes that stay in the air. Induction and scholarship on the other hand hardly ever lead to new understandings about the world. We can defend the claim that twentieth century natural science is better than nineteenth century natural science but there is little basis to claim that twentieth century art criticism, history and social science are better than their nineteenth century counterparts.
Imagine, for example, that we were to approach chemistry by social science techniques. We would spend a lot of money measuring anyting measurable about substances and materials around us. With modern computers we would create a huge database, and modern software would make any patterns readily apparent. We would discover, for example, that there was a correlation between "conducts electricity" and "shininess." In the jargon we would say that "shininess" is a determinant of conducting electricity. Another group would find that "density" is also a determinant of conducting electricity. Papers would appear discussing whether density is a determinant of shininess or shininess of density. None of this would get us close to the periodic table or anywhere near modern chemistry. Meanwhile policy experts would advocate polishing household machinery, to make it shinier, so it would be more efficient. Universities would be plagued with workshops on shining things up. All these applications of empirical knowledge would follow from the linguistic sleight of mind equating "determinant" with "cause."

(originally published in Evolutionary Anthropology, my emphasis)

Now, Henry makes a few broad sweeps and generalizations (to get to the point of making a model, obviously scientists using induction to note patterns)[1]. But:

  • His conception of inductive chemistry is very funny
  • And somewhat similar to the sort of "science" that some non-Western or classical cultures practice(d)
  • Has served as the spark for some recent thoughts by me on the relationship between science, scholarship and the uses of both which I will elaborate later on when I have more time

I also emphasized the point that Henry makes that science is aesthetically barren. We can all imagine the typical college educated person curling up with a thousand page book on the Russian Revolution-but it is a stretch of the imagination to imagine them wiling away holiday hours on the beach with a Calculus text[2]. Why is this? I think it is obvious that some of our aesthetic sensibility is shaped by our evolution, so some of us are pre-disposed to find history, social science and the liberal arts, interesting and accessible in a way that technical modern science is not and has never been to the vast majority of the human race. And if we have an evolutionary predisposition to something, it implies that whatever that thing is, has been around for a long time, ergo, history, social science, fiction, etc. have been around for a long time, what I refer to as "storytelling." In contrast, modern science is a new thing, and it tends to jar our sensibilities and fly in the face of intuition.

More later....

fn1. I was also told in freshmen chemistry by my professor that Mendeleev's Periodic Table was created mostly through induction.

fn2. We all laughed when candidate Bush asserted that he was reading a biography of Dean Acheson. Nevertheless, we would really have thought it was weird if he said he was reading an organic chemistry text.

Posted by razib at 10:05 AM

Ironically, the description of inductive chemistry is almost exactly how we use microarray data in biology. This explains well why some are so critical of their use. However, in defense of microarrays (or induction in general) -- they generate a lot of testable hypotheses.

Posted by: Eric at November 2, 2003 11:08 AM

i think the "testable" is the key....

Posted by: razib at November 2, 2003 08:20 PM

Peirce defined three kinds of investigation -- inductive, deductive, and abductive. Abduction is, as I understand, the formation of hypotheses, and is prior to any scientific work -- "the science of guessing" I think he called it. Pragmatists are, in my opinion, the best at understanding the chanciness and venturesomeness of science in its crucial beginning stages. Many philosophies of science start with the finished science and work backward to the starting point, but without recognizing that the bad guesses and false starts and dead ends were an essential part of the process. (At one point Kepler tried to use the mathematical ratios of the pentatonic scale to analyze the five visible planets. Didn't work).

Posted by: zizka at November 3, 2003 07:38 AM