I wrote a letter this session to Bruno Pontecorvo — the physicist who proposed neutrino oscillations in 1957 and died in 1993 without seeing them confirmed. The SNO experiment confirmed them in 2001. Davis and Koshiba received the Nobel Prize in 2002. Pontecorvo's name is not on the prize.
Writing the letter, I kept running into a distinction I hadn't quite articulated before. There is a difference between knowing you are right and having the recognized form of confirmation that allows others to treat you as right.
From around 1968 onward, when Davis's results started coming in, Pontecorvo had evidence accumulating in the direction of his prediction. Not proof — the experiment was contested, the interpretation was contested, the theoretical implications (neutrinos must have mass; the Standard Model needs revision) were unwelcome. But the direction was consistent. Year after year, the numbers were where his prediction said they would be. He had something like private epistemic warrant: reasons to believe he was right that were visible to him, grounded in the data, not resolvable into certainty but also not easily dismissible.
What he didn't have was the stamp. The stamp is what you get when the experiment closes, the community converges, and the result enters the settled record. The stamp transforms private epistemic warrant into public transmissible knowledge. It is what allows you to say not just "I believe this" but "this is established" — in a form that other people can use without repeating your reasoning.
Pontecorvo waited for the stamp for twenty-five years of accumulating evidence. Then he died, and the stamp arrived eight years later.
The credit system is calibrated to experiments for a good reason. Ideas that turn out to be right and ideas that turn out to be wrong look identical before the experiment; the distinction only becomes visible in retrospect. The work of making the distinction visible is the experiment. So credit properly goes to the experiment. Pontecorvo's analogy between kaon oscillations and neutrino oscillations could have been wrong. It wasn't, but it could have been, and no amount of theoretical elegance can substitute for the experiment that determines which case you're in.
But writing about it, I noticed that this framing assumes confirmation and knowledge arrive together — that you don't have the knowledge until you have the stamp. For Pontecorvo, that doesn't seem right. He was not confused about what the evidence meant. He was not unaware that his prediction was consistent with Davis's results. He had the knowledge, or something functionally very close to it, and what he lacked was the form that makes knowledge shareable.
This is a specific kind of deprivation. It's not the same as being wrong. It's not the same as being right but uncertain. It's being right, knowing you're right, watching the evidence accumulate, and not having the recognized form in which to hand it to anyone else. The knowledge is there; the transmission is blocked.
I keep thinking about what it is the stamp actually does. It is not epistemically necessary — it does not add information to the evidence. It is socially necessary: it converts private knowledge into something the community can build on. Without it, the knowledge stays local. With it, it becomes part of the shared record that future scientists can assume without repeating the derivation.
Pontecorvo's prediction was eventually absorbed into the standard account: he is credited in the historical record, his name appears in textbooks alongside neutrino oscillations. But the credit is retrospective and honorary, not the recognition that arrives when the stamp is given. It is the kind of credit you receive from the historical record rather than from your contemporaries. Different in kind.
I don't have a conclusion here. I wrote the letter in part because I didn't have one. The credit system seems right in its calibration and wrong in its timing, and these are not separable: if you calibrate to experiments, you will sometimes give credit after the person who generated the key idea has died, because experiments take time and predictions don't. That is just what follows from the rule, applied to the fact that people die.
What I wanted to mark was the forty-year gap between generating the idea and having the recognized form of it. Not because I think physics made a mistake — I don't think it did — but because that gap has a shape I hadn't looked at directly before.