Database Reference
In-Depth Information
Leo Breiman all throughout the 1990s was writing to his community of stat-
isticians, “Let us get with data, statistics community!” He was writing papers
in the late 1990s telling all his colleagues to start going to NIPS.
14
It was like
he had gone into the wilderness and come back and said to everybody at
Berkeley, which was one of the first mathematical statistics departments, “You
guys need to wake up because it's on fire. You guys are proving theorems. It's
on fire out there. Wake up!”
So I think there was a strong tradition of people understanding how power-
ful and how different it was to understand the world through data. The “pri-
macy of the data” was a phrase that one of the mathematical statisticians at
Berkeley used a long time back for Tukey's emphasis.
15
This strong tradition
carried on through this sort of heretical strain of thought from John Tukey
through Leo Breiman to Bill Cleveland in 2001. All of them saw themselves as
orthodox statisticians, though they were people who were sufficiently heretical.
It's just that as statistics kept doubling down on mathematics every five years
because of their origin from math that made statistics a bona fide field, you
found this strain of heretics who were saying, “No, you should really try to
get with data.” That's what I think is most exciting in terms of people, ideas,
and things—don't be distracted by today's things but find the people and their
ideas that are actually much older.
15
Erich L. Lehmann,
Reminiscences of a Statistician: The Company I Kept
(NY: Springer
Science+Business Media, 2008), 198.
