On Jan 7, 2009, at 3:53 AM, Ray Saintonge wrote:
WJhonson@aol.com wrote:
It isn't necessary to go so far back. A large part of the important mathematics of the 1980s and 1990s does not appear in textbooks, or does so only implicitly, because there is little incentive for anyone to rewrite it.>>
This is a contradiction. If work on Number Theory were "important" than surely my new book on Number Theory would include it. If editors are solely referring to old notebooks, than that's their own issue. That doesn't prevent the rest of us, from using only the newest textbooks if we so choose. The very definition of "important" is, that many people cite it. If no one cites it, it's not important.
This is a bizarre definition of "important"; it might work for "influential" or "popular", but that is not what makes something important. Many new ideas are tangential to a general education about a subject, but are no less important to the advancement of knowledge. Textbooks are instruments for parroting the party line of received wisdom. They do little to address controversial issues.
Well, and on top of that, publishing is a commercial enterprise. Even academic presses make decisions on what to publish in part based on what they think they can avoid completely losing their shirts on. So by relying too heavily on the question of what is published, we inject a really problematic commercial bias into what we do.
"Encyclopedia" and "record of only what has been published in reliable secondary sources" are not synonymous terms.
-Phil