“In the long run, the methods are the important part of the course. It is not enough to know the theory; you should be able to apply it.”

Fifth Lecture, Applications in Statistics and the Theory of Errors, p. 141
Probability, Statistics And Truth - Second Revised English Edition - (1957)

On his plans to use the Nobel Prize money to help fund peace organisations he has worked with, quoted in "Ahtisaari wins Nobel Peace Prize" in BBC News (10 October 2008) http://news.bbc.co.uk/2/hi/europe/7662922.stm

Source: Science and Statistics (1976), p. 792

Source: 11-12 July 1941, quoted in Hitler's Table Talk, 1941–1944

Interview in The Hindu (2013)
Context: The improved understanding of the equations of hydrodynamics is general in nature; it applies to all quantum field theories, including those like quantum chromodynamics that are of interest to real world experiments. I think this is a good (though minor) example of the impact of string theory on experiments. At our current stage of understanding of string theory, we can effectively do calculations only in particularly simple — particularly symmetric — theories. But we are able to analyse these theories very completely; do the calculations completely correctly. We can then use these calculations to test various general predictions about the behaviour of all quantum field theories. These expectations sometimes turn out to be incorrect. With the string calculations to guide you can then correct these predictions. The corrected general expectations then apply to all quantum field theories, not just those very symmetric ones that string theory is able to analyse in detail.

Part 5 "On training in Jeet Kune Do"
Jeet Kune Do (1997)

Source: Aphorisms and Reflections (1901), p. 216

On Practice (1937)