From the Inbox:
One of my good memories from grad school days comes from a time in 1968 when I loaded up some fellow students and drove to another campus to spend an afternoon with Leo Kadanoff and take in his colloquium talk. I was impressed by the clarity of his talk and by his easy going manner. He spoke of his work on critical phenomena and described how fluid properties near a critical point could be described by a single parameter – a correlation length. At one point he showed a slide with some experimental data which agreed spectacularly well with the theory over several decades of variation of correlation length, but missed by a little nearest the critical point. He admitted to being disappointed by that but said that something better than his theory was needed because “it is difficult to make ony a three percent conceptual error”. I can think of several areas of science in which some similar humility would be welcome these days.
A few months earlier a physics Nobel laureate had visited OU. His arrogance and Leo Kadanoff’s openness both made lasting impressions.
See also: Kadanoff: Information a primary topic of 21st C science. A tribute to his dissertation advisor Leo Kadanoff (1937–2015) from Bill Dembski