John S. Bell, exceptional theoretical particle physicist (1928-1990)

Bell is well known for his sharp analysis of quantum probability, which he showed can violate some theorems of probability theory (as understood in the classical sense). This work came after he joined CERN. He was appointed to CERN, perhaps after the deep results he obtained on the PCT theorem (at about the same time as R. Jost). His work on this has been ignored, mainly because it was after the work of Pauli and Luders; it was more general than this work, but less so than the work of Jost (1957); so Bell's work fell between two schools.

While at CERN, Bell (with Roman Jackiw) obtained evidence for the axial anomaly. My reading of the paper is that the authors thought that the anomaly was not really there, because it disappears if the theory is regularised. This is in contrast to Adler's paper, where the point is clearly made that the anomaly persists whatever renormalisation, consistent with gauge invariance and special relativity is made.

I remember Bell's giving the main talk at one of the annual Christmas meetings at Rutherford Lab (to which I was invited in those days). He was describing the "paradox" of EPR (Einstein, Podolski and Rosen), and I was very suspicious because he was trying to analyse a quantum problem with classical probability. I sensed also some hostility from the rest of the (hard-nosed) audience. Bell described how a measurement of the spin of the particle far to the right gives us information about the spin of the particle on the far left, without the need of measurement on the left. He explained this by asking us what was the colour of his socks. They could be one of several colours. But if he showed us one (he did so; it was red) then we could immediately conclude that the other one was also red (without the need to see it), UNLESS we thought him to be some sort of crank! We fell about laughing: this was exactly what we thought. What we did not appreciate was that without his deep analysis of the EPR experiment, we would not know that it is IMPOSSIBLE to describe the data using classical probability; see my article on quantum probability.

Bell saw the EPR "paradox" as a fundamental non-locality of quantum mechanics; see the beautiful essay by R. Jackiw and A. Shimony, The Depth and Breadth of John Bell's Physics. On the contrary, I argue that there is no non-locality, merely a correlation, which needs analysis by quantum probability and the theory of quantum games.