I cannot imagine writing a math or physics textbook with someone else. I mean it would be so hard for me to reconcile my ideas with another persons. I mean I like collaborating on homework assignments and such. But in each of the subjects I have taken, I have developed a very individual system of logic that works for that subject. And I think everyone else has as well. If I tried to combine my system of logic with someone else's, there would just be so many arguments over like trivial exposition questions that we would get nowhere. Although we can arrive at the same solutions to problems, we all have very different "associations" and "mnemonic" devices for doing this. And that is the reason why it is difficult to coauthor even if both coauthors are totally masters of the material. I have also noticed that almost all of the great physics and math books that I know of are written by a single person: Rudin (Mathematical Analysis books), Spivak (Differential Geometry books), Griffiths (QM and EM books), Stewart (Calculus books), Munkres (Topology books), Jackson (E and M book), Shankar (QM book). The only exception I can think of right now is Marsden and Hoffman's Basic Complex Analysis and I think Hoffman was more of a sidekick than a coauthor in that case. I think coauthored books are a bad omen for the book and also maybe a sign of weakness for the author. For instance, consider Thornton and Marion's Classical Dynamics. That is a clear example of what goes wrong when you coauthor. It is in general a pretty decent book. It is understandable and consistent and complete. However, it does not have the classic "feel" like the books of the authors I listed above have. You can almost feel a split personality while reading it. There are no overarching themes in the book. The exposition is bland. There is just nothing special about it. Just some thoughts.