Wednesday, March 31, 2010

Reductionism, part III: Self-reinforcing idealization in science

We've pointed out the history of genetics as a highly empirical research program that has led to an unprecedented century of new knowledge of a fundamental nature, about genetic causation. But we pointed out that this has been brought about by clearly intentional tunneling through what was known from the beginning as the greater truth.

Even today, the same logic used by Mendel, and the same intentional tunneling, is the basis of genetic research. It is a form of self-reinforcing idealization. The method is strengthened by its discoveries, and the tunneling reinforced to keep out 'extraneous' aspects of the truth, even if they may be important to an overall understanding. That's why there are problems using this method to understand the genetic basis of complex traits.

Complex traits are the result of interactions among many functional aspects of DNA, in a highly orchestrated pattern of gene usage in space and time, as an embryo develops from a fertilized cell to an adult. This is intimately related to evolution since different species get that way by developing differences in their structures. The understanding of the latter is a field called EvoDevo.

Geneticists from Mendel to today have focused on clear-cut, transmissible genetic effects, stereotyping what genes are and what they do, to get a handle on various questions about genetic function. On their part, developmental biologists have done something similar. They have divided the continuously changing development of an organism into various stages, treating them as if, like genetic effects, they were distinct, discrete entities.

This is very similar to the stereotyping of genes and genetic effects that have characterized genetic research from Mendel to Morgan to today. Development and EvoDevo use similar tunneling through reality to try to understand the genetic basis for the origin of complex structures. This ignores the obvious and well-recognized fact that stages and structures are not independent. Evolution shows this in many ways, as there are developmental differences among closely related species, that are brought about by different genetic mechanisms (or their timing).

We were led to relate the two kinds of tunneling through reality by conversations Ken had with Alan Love on his recent visit to the University of Minnesota. Alan alone, and also in collaboration with Rudy Raff of Indiana University, have written interesting papers on this subject, and on the way that this kind of tunneling (our term for it, not his) has both helped and restricted the way we view development and its evolution.

Science may work today only because of self-reinforcing idealization--the ability to tunnel through complexity by focusing on simple aspects of more complex wholes. Whether a new thought process may eventually be called for, to broaden the tunnels, or whether tunneling is the best approach to understanding Nature, isn't clear. But it is clear that until a better way comes along, we will continue to tunnel.

1 comment:

Ken Weiss said...

We forgot to provide references to the papers of Alan Love's that we referred to, for those who might be interested:

Idealization in evolutionary developmental i nvestigation: a tension between phenotypic plasticity and normal stages. Phil Trans Royal Soc, B 365:679-690, 2010

From philosophy to science (to natural philosophy): evolutionary developmental perspectives. Quarterly Rvw of Biol, 83:65--, 2008

These papers contain references to Alan's work with Rudy Raff that we referred to in this post.