We discussed founding such an institute but there were many questions, of course. What should it be like? What would it study? How would it operate? Some people suggested a relatively narrow focus for it. Nick Metropolis, or Metropolos, I think wanted it to be an institute devoted to computat... parallel computation, and so on. But my point of view, which was supported by some others and eventually, I think, by everybody, was that it ought to be very broad and consider a great number of fields, roughly, in fact, I think the spectrum that it covers now: from mathematics and computer science through physics and chemistry to evolutionary biology, ecology, neurobiology, psychology, immunology, economics, anthropology including archaeology, political science, history, something like that–linguistics, also, that sort of range. And that the computation, while a very important part of the work of the institute would be in great part in the service of science; try to use computation in addition to ordinary analytical theory and to observation as a tool for understanding, and then to use the computers themselves and software on computers also as systems of… to be studied, along with biological systems, social systems and so on. And gradually that idea took hold and I think everybody supported it. But the question was still open as to how the institute would operate and what—in these many fields—what would we do. Well, my feeling and that of some others was that we should make it interdisciplinary in operation, that is people should- it should… it should be a place where people would have the opportunity to co-operate, would be encouraged to co-operate, work together among different fields, something which is largely lacking in our great universities and institutes of technology. And of course, for good reason, because most of the criteria of excellence are within the disciplines. We have systems of degrees and courses and departments and journals and professional societies and sections of granting agencies and so on that are all arranged by field. And we know that mediocre talents like to hide in the crevices among fields. So it's a challenge to figure out how to recognize excellence in interdisciplinary work, but gradually we came to the idea that we would try to recruit people who were really pretty good in what they had done before and were perhaps were still doing to some extent, so we would have experts in all these different subjects working together. That way, if a group of people were working on a problem in linguistics, say, and only one was a linguist, that person would be responsible for making sure that it wasn't phony linguistics, and so on. That principle has held and I think that's more or less the way we… we operate.
