Concord Consortium Blog

The word "model" means a lot of different things to different people. A model airplane looks like a real airplane, only smaller; a paper airplane flies like a real airplane, only not as far or as fast. Both are models, neither is the kind of model I have in mind.

For the purposes of this discussion I'm defining a model as a description of a phenomenon in terms of things that can't be seen, felt, or heard, but that explain what's going on. Models may involve things that are too small to be seen, or too big; processes that take place too slowly or too fast. The plate tectonics model, for instance, informs us that the Himalayas are being formed, even as we speak, by the earth crumpling like a car fender, as India crashes (rather slowly, to be sure) into Asia.

Science is all about models of this kind, and an important goal of science education -- and of the Concord Consortium -- is to give students some examples of models and show them how to use those models to make predictions, to guide experimentation, and generally to make sense out of their own and other people's observations and experiments. Scientific models are constantly subject to revision as new experiments are performed, new data collected, and new interpretations advanced to explain existing data. It is important, therefore, that we teach our students about this process as well, giving them the sense that science is perpetually a work in progress, rather than a set of unchanging "facts."

Take, for example, the theory of evolution, a model that systematizes the description of an enormous body of data in terms of three simple propositions:

1. that organisms inherit many of their physical traits from their parents,
2. that differences in those traits can result in differences in organisms' ability to survive and procreate, and
3. that new traits can arise through random variation.

Note that the model does not depend at all on understanding the processes by which these things happen. In fact, when Darwin proposed his model for evolution in 1859, only the first two could be demonstrated at all, and only the second could be explained. Mendel's model of genetic inheritance  would not be published for another seven years (only to be ignored and rediscovered in 1900) and the discovery of mutations, which give rise to the random variation required by proposition three, was half a century away.

Genetics is only one example - evolution depends critically on models drawn from many other sciences. Geological models of the age of fossil-bearing rocks, critical to giving evolution theorists enough time for the hypothesized processes to take place, bear directly on the feasibility of the evolution model. So do models of mutation rates in different organisms, which themselves depend on models of the environment in which those organisms lived a long time ago. And the burgeoning new science of genomics, by comparing DNA sequences across present-day species, gives us the ability to trace the evolution of those species with a precision that Darwin could never have imagined.

All this interdependency makes evolution a particularly fragile model, as its opponents often point out. The discovery of a single errant fossil^,* like the human footprint supposedly found amid dinosaur tracks, could in principle bring the entire elaborate edifice crashing down. But its very fragility confers on the evolution model its extraordinary power. Evolution is fragile because it is so broadly applicable. In providing a model for how the multitude of species arose on this planet, it feeds into and constrains dozens of other models. It is powerful, in other words, for precisely the same reason that it is fragile.

 

^*For an in-depth discussion of this issue try http://paleo.cc/paluxy.htm.