Tool

Contrary to what I thought, PS did make an important point: Granting agencies are more willing to give you the money if there is a statistician in your list of collaborator.

The reason is entirely legitimate. The data are often complicated for mega clinical trials and epidemiological surveys, and you need a professional to handle the problem. As Adam Smith advocated, it is the division of labour.

Nonetheless, I am not at ease. Statistics is a tool we design to help us seeing the truth; it seems putting the cart in front of the horse if you excel at using some sophisticated tool but refuse to look by your own eyes what's happening in reality. What good is it to have an excellent calculation on some data you put in carelessly and collect from trials with shaky methodology ?

And there are precedents in the history. Logic was designed to facilitate discussion in philosophy. With Bertrand Russell and Ludwig Wittgenstein, however, logic became symbols and arrows and abstract discussion that could be done without language - it lost touch with the reality.

PS. It is a romantic idea to bring together specialists of different fields in order to accomplish some grandiose project. I could only think of the Tower of Babel.

Remember, Plato, Newton, Spinoza (oh, my prejudice), Darwin, Einstein, Watson and Crick accomplish the work on their own. (Well, the latter two chatted with others in the pub; I have no objection to collaboration of this kind.)