This is about something I've noticed on and off over my years in the alternate universe that is the academy. A great many faculty and other people bemoan the formation of discipline silos, and yet nobody seems to have a grasp on how and why they form. I'm starting to think they are the inevitable result of a number of forces that aren't going to disappear soon.
Let's start with one that's near and dear to every economist's heart; specialization. There are huge gains to be made in almost every field of human endeavor by having different people focus on different things. Knowledge creation is no different than any other human endeavor in this regard. We can delve more deeply into things through specialization. As our body of knowledge increases, the number of things someone needs to know to be able to make a meaningful contribution increases. The number of academic articles published across disciplines every month is more than a life time's reading, never mind an entire discipline's worth of background material. So the silos are in part a natural part of the growing body of knowledge.
Specialized methods of analysis: An increasing number of disciplines have moved away from the ancient philosophy mode of figuring out how things work, namely words and thought experiments. The use of formal mathematic models can create a barrier to entry that is hard for many to overcome. This division between disciplines, even when studying the same problems, makes cross understanding difficult. A small, but growing, number of academics are able to translate complex mathematical concepts into something more accessible. It's a long between theoretical physics and postmodern art.
Resources: This is the dirty little secret of universities in developed countries. A growing share of a university's budget is not devoted to academic staff. Instead it is being spent on a variety of other things, some legitimate, others ... let's just say I'm not convinced yet. The resulting fights for resources tend to make interdisciplinary cooperation difficult. In order get new positions or often to even replace retiring faculty, a remarkable battle with highly uncertain rules and potential outcomes ensues. Often, but not consistently enough to make it truly workable, resource allocations come down to the number of students who have declared a discipline as their major.
Confirmation bias: Now we focus on students. People choose disciplines and view disciplines based on their own preconceptions. How could they not? Which discipline will someone who believes poverty is the result of an oppressive system choose? It likely isn't evolutionary biology. It isn't likely to be economics, though it does happen. By the same token someone who believes in self determination isn't likely to pursue a degree in sociology. This self selection reenforces the things that make the disciplines different. Thus a difference between disciplines that starts out small will be continually reenforced until the gulf between the disciplines becomes almost impossible to cross.
All of these processes combine to make the gulf between disciplines hard to transcend. Maybe, just maybe, the gains of these silos out weigh the costs.