What makes Caltech special?

As you may have heard, Caltech has been ranked #1 in the world by the Times Higher Education (THE) Rankings the last few years…

That’s nice and all, but Caltech is such an oddball of a place that to some extent, where it places in such rankings is more a function of how they normalize by size than anything else.  Caltech is simply an order-of-magnitude smaller than other places (we have only ~300 faculty institute-wide), so if you don’t normalize it suffers…

But, while rankings are always problematic, I think the feature that the THE put out about Caltech this week actually summarizes very well what makes Caltech such a special and unique place…and, of course, it comes from its size.

In particular, I like the following two quotes in the article from our Engineering division chair that make the point nicely:

•  “I always refer to Caltech’s small size as being very similar to the size effect that exists in materials – there are special properties that exist when you are extremely small”
• “I have 77 faculty in engineering and applied science. MIT  has 490. How can I compete with an excellent place like MIT? We have to have engineers interact with all of the sciences and vice versa – it is a matter of survival. We don’t have the breadth to do things in a big way unless they interact.”

Both of these points hit me almost immediately when I arrived at Caltech six years ago.  I came to Caltech after spending ten years at CMU (undergrad+grad), which prides itself on interdisciplinary education and research, and I thought I knew what to expect from the new interdisciplinary environment I was moving into.  It turned out that I had no idea. 

Interdisciplinary meant something completely different at Caltech.  In my case, it meant that within my first four months on campus, as a result of discussions that started at the faculty bar, I went from having no background in economics to co-teaching a course in the economics department! …and now nearly half of my research draws on such tools, and my students come from four different departments.  More broadly, it means that every faculty in CS at Caltech has active collaborations with someone outside of CS.

Maybe it’s too simplistic, but I really do think that the small size of Caltech is not just advantageous for creating its interdisciplinary environment, it is necessary.  As the quotes above point out, Caltech has created an environment where being interdisciplinary is, in a sense, unavoidable and required.  

In the local sense, if one thinks of faculty as interacting closely with a dozen or two other faculty members, then at a place like Caltech, this necessarily involves interactions with people in very different departments and areas.  In the more global sense, to have the size to have huge impact, Caltech faculty must pursue cross-cutting agendas.

But, of course, being small is a high-risk & high-reward strategy.  You can’t try to do everything, and you must make bets on people & areas successfully…  I think this is perhaps the biggest challenge in trying to mimic the Caltech phenomenon.

And, because of the risk involved, it is natural to try to let the number of faculty in an area, in a department, and institute-wide grow… However, the effect of this is, in the long run, a change in the culture away from interdisciplinary necessity.  Perhaps the most amazing thing about Caltech is that it has avoided this danger, and maintained its size, for so long.

If one thinks about growing a graph, there are two options: add nodes or add edges.  The Caltech approach is basically to avoid adding nodes and thus force the creation of edges.  The structure that emerges seems to be fundamentally different from that of other schools: more connectivity, more clustering, etc.  (Maybe there’s a strategic-network-formation paper in there somewhere…)