Yang was the chairman of the session and so the question was posed formally to the chairman. And Yang said, ‘Well, TD Lee and I have looked at that, but it doesn't lead anywhere, or we didn't find that it… we didn’t find that it led anywhere.’ That was it. So Feynman didn't push this much further.

Then in May in Moscow, at that meeting that I described, I mentioned that there were various hypotheses about the τ and θ and parity, and so they scheduled a lecture on the subject. And the hall was completely full, standing room only — and there were people standing – and there were many people who couldn't get in. And I described this idea, ascribing it to Marty Block and saying that Feynman had proposed it to the meeting and that I had endorsed it and that it was a possible idea that parity was simply not conserved in the weak interaction; or there was the idea of parity doubling which I had thought of and not published and which Yang and Lee had published. And the third hypothesis which I thought was very unlikely was Marshak's very ugly hypothesis – Marshak was right there – I said Marshak's very ugly hypothesis that it's a two plus state, and that the τ spectrum is really not what it seems to be, then parity is conserved.

[Q] You wouldn't have a flat spectrum of course?

No. So… well, the idea that parity was not conserved was very upsetting to some people, and they got up and said, ‘But parity has to be conserved! Are you suggesting that Lorentz invariance is wrong, too? What about rotational invariance? This is just a silly idea.’ And I said, ‘Well look, I didn't propose this, but it is a perfectly possible hypothesis. Parity conservation does not come from reflection of co-ordinates; it comes from the transformation properties of the Hamiltonian, of the interaction concerned. In the case of the weak interaction nobody has ever really tested whether parity is conserved.’ Well, actually I didn't know it, but there had been an experiment in 1928 that tested it and found that it wasn't conserved. It's still not known whether that experiment was really correct or whether the result was an artifact, but I didn't know that it had been tested back in 1928. What I should have done if I had been a real… a really serious person was to propose a test right there. I could have said, ‘Well, sigma.p is a… is a pseudo-scalar if… if the spin depends on the momentum… the spin direction depends on the momentum direction. Then yes, it would be… it would be not violated.’ I didn't say that unfortunately. Anyway, there was so much interest that I had to repeat the talk immediately afterwards for a whole other audience, which also had standing room only. But the reaction of the big shots was very hostile to the parity violation idea. The parity doubling idea intrigued them, but the parity violation idea they didn't like at all. Well, the more they resisted it the more I thought, well, maybe its true. Well, when I got back Yang and Lee had changed their opinion and put forward the idea that it was not conserved, and had some… had this very, very complicated test of it which was being carried out at low temperatures by Mrs Wu and Ambler… and Ambler instead of suggesting just that they look for polar and longitudinal polarization of beta rays which would have been much simpler. Well then, of course the result was announced and parity was violated and many people were very, very shocked. To me it was a very important development, but it was certainly not shocking. And to Feynman it wasn't shocking and to so some other people like Fermi and Dirac it wasn't shocking. I'm told that when they announced to Dirac that parity was violated in the weak interaction, he said, ‘Parity? Oh yes, parity conservation. Is that in my book?’ And of course it isn't. The… well, as a… as a consequence of that people began to think about possible theories. And Salam and Landau — who was finally converted, although he hadn't believed in the whole thing — and I think maybe one other person suggested this 'one plus gamma five' idea, which I had been thinking about for a brief period earlier; but it was actually one minus gamma five that they were talking about because the data supposedly indicated scalar and tensor interaction and therefore the parity, the sign of the parity violation was… was actually wrong. People thought the electrons were spinning in one direction when they were really spinning in the other.