Proceedings used to be useful

There has been a practice which the science librarian sometimes encounters, which, to the best of my knowledge, is now essentially abandoned, but that proves to be of inestimable value for epistemological purposes.

Archiving the discussions.

Take the 20th Solvay conference, for instance. The group photo I have is not good but you can still find a vivid picture of the prestigious attendants:

The chair of the event is Mandel, but Paul, not Leonard, although as you can see the latter is still there amongst the participants.

Now, in addition to a paper that formally captures the essence of the live presentation—as we are still used to—there are also so-called "discussion sessions", which are the Q&A post-talk segments, but that have been edited to capture or retain the content of this interesting, in fact, the most interesting part of the oral presentation. Here you can find, for instance, live-looking contributions such as this one from H. Carmichael to the talk (they called it 'lecture') of I. Prigogine, specifically targetting an earlier comment by Prigogine's co-author T. Petrovski (the Chairman is Sudarshan):

And here is the reply, first from Petrovski and then from Prigogine himself, in a tone which I leave to your apreciation:

Note that neither addressed the comment itself, on what seems to be Strong Coupling in cavity QED.

I will not comment at length on this particular excerpt, but this obviously captures something interesting, including at the pure scientific level. For instance, Prigogine (who was an underrated genius, possibly the biggest brain in these exchanges) describing non-Markovianity as a superposition of different timescales is a neat contrast of clarity and inspiration to the technical and mathematical equivalent description of his co-author.

Those contributions appear to be edited, in particular when they come to detailed derivations, such as this later comment by Scully:

I don't reproduce it in full, but clearly this was not simply a "standing question", probably it made use of the blackboard or involved a broad description of the main lines with a posteriori editing. This is Prigogine's reply to that:

Such reports are also recorded for other lectures; the one from C. Cohen-Tannoudji for instance is also very interesting as it reads like a dialogue, with short questions and short replies (except Scully at the end who comes back with «a rough cut at Professor Kimble’s question» and derives a result which he says «deserves, and will receive, a more careful treatment.»