Generalists and the PhD
A PhD is somebody who gets to know more and more about less and less until he knows everything about nothing.
That bromide (a variant of others) gets passed along quite frequently about academics, and a new book by David Epstein seems to confirm the implication that super-specialization is not useful. As described in an excerpt in The Atlantic, when narrowly focused experts try to make predictions, they fail spectacularly in comparison to predictions made by generalists. One example is the conflicting forecasts of Paul Ehrlich’s “population bomb” versus the counter-prediction of continued economic improvement made by Julian Simon; both missed the mark in different ways, but both continued to double down on their forecasts. Following many others, Epstein compares the two groups to hedgehogs and foxes. So why on earth should we make hedgehog PhDs?
On its face, a PhD is generally trying to untie one small knot in our universe of knowledge. When did the Rio Grande Rift start extending? What is the power law exponent for sodic feldspar if deforming by dislocation creep? Just how many angels can dance on the head of a pin, anyways? If all we do is train somebody to continue, arrow-like, on that initial trajectory into some byzantine corner of human knowledge, then we have failed. So what then would be success?
Success should be learning how to identify problems worth solving that are solvable and then defining a course of action that will yield that solution. In short, a PhD should be an exercise in learning these skills and applying them in one place to demonstrate mastery. Why would this lead to deeply entrenched viewpoints seemingly unchangeable by evidence?
One reason is probably that we are failing to emphasize the right messages to our PhD students. Some mistake the value of the PhD, thinking it isn’t learning how to think at a high level, it is producing some nugget of information considered of value. And so if your self-worth is based on that nugget instead of your training, you are apt to fight to maintain the validity of the nugget. In a sense, this is the nut at the bottom of the Ehrlich-Simon conflict: In the natural world of Ehrlich, populations only are controlled by crashing; in the economic world of Simon, resources are infinitely elastic. Combining the two views can be hard as it might seem you are abandoning the insights you gained in your field.
But there are scales to this hedgehog and fox dichotomy. GG is a hedgehog in global sense, working on the origin of orogens and their geologic history. GG’s officemate in grad school was unquestionably a fox–he would read Science cover to cover; after being sufficiently tortured to complete a PhD, he went to work as a science advisor in the Senate (where he gloried in the week-to-week change in subjects) and then the Presidential Science Advisor’s office before hopscotching across the D.C. landscape in different guises.
Within earth science, though, GG feels a bit more like a fox. He has colleagues who are global experts in all the ins and outs of some facet of earth science: how to use surface waves, where water hides in the mantle, how a landscape responds to a change in climate. Yet GG wants to understand enough of all these things to apply them to problems of his interest. (GG is amused by his standing in a geophysical meeting as a geologist and as a geophysicist at a geological meeting; basically, he is nobody’s expert).
In some sense, the conflicts and failures of experts seems another facet of the divergence of belief in climate change between Democrats and Republicans as scientific literacy increases: the more you know, the easier it is to find your way around evidence contrary to your beliefs. Oddly, this can be an advantage for a research scientist. Consider Alfred Wegener’s proposal for continental drift: the original theory had a number of deep flaws, from lacking a credible motive force to miscalculating modern rates. Yet the essence of the proposal–oceans and continents were fundamentally different and the continents could move–was correct. Abandoning the theory at the start would have been a mistake. To some degree scientists are stuck advocating for their ideas because objections can reflect anything from a fatal flaw to an imperfect description of the idea to bad observations. It can take some perseverance to figure out which objections are fatal and which are not. But that skill–which is in some ways first acquired in gaining a PhD–can get in the way of making a judgement of the field overall.
How do we balance this? One way is asking our PhDs in exams, so what? What is the broader context? Maybe how would you approach this problem with a different technique? Another is in the limited classroom instruction we have. GG teaches a western U.S. tectonics class where a point is to confront different approaches to addressing tectonic problems, most of which are not the specialty of the students in the class.
We’ll surely continue to generate hedgehogs in academia; some people are really just that way. But if we are careful and emphasize the right things, maybe our hedgehogy tendencies can be balanced a bit with some foxiness.