Ed Felten of "Freedom to Tinker" notes that in a list of all-time "must read" books selected by a surveyed group of university presidents, science and technology are conspicuous by their absence. Commentators on his blog (and on Crooked Timber's offshoot posting), asked to supply the missing masterpieces of science, draw almost entirely from three categories: (1) works on the philosophy and foundations of science (Popper, Kuhn); (2) Pop-science digests of various interesting topics (Douglas Hofstadter, Jared Diamond); and (3) Books by and/or about the practices of computer geeks.
Why were all these suggested books (Origin of Species might be one exception) so far in stature from the books on the presidents' list (the Bible, Hamlet, the Odyssey, The Prince)? The obvious answer is that the Great Books of science--and they do exist: viz., Euclid's Elements, Newton's Principia--simply don't occupy the same place in the scientific world that the Great Books hold in the humanities and related disciplines. No half-decent undergraduate curriculum, for example, would allow its literature students to escape deep familiarity with Hamlet, its philosophy majors to avoid studying The Republic, its classicists to skimp on the Odyssey, its divinity students not to delve deeply into the Bible, or its budding political scientists to pass on The Prince.
But few geometers feel any need to familiarize themselves with geometry as Euclid explained it; nor do physicists feel incomplete without understanding Newton's original notation for his laws of motion. And I would much sooner encourage students to master a few basic college texts in math, sciences and engineering than push them to grapple with the same concepts by studying the original works that introduced them. The originals, after all, each represented one farsighted individual's brilliant-but-still-hazy insight, which has since been clarified and extended far beyond that first attempt at elucidation.
That's one of the key distinctions between the "two cultures": science seeks to separate its ideas from individual humans, and to place them in an abstract world where not only objectivity but also impersonality reigns. No serious student of literature would argue, like the hero of "Metropolitan", that reading critical essays about great literary works is better than reading the originals. The reason is that the ideas in literary works--or, for that matter, philosophical, historical or religious ones--are bound to the original thinker (and his or her original words) in a way that scientific ideas are not. Plato's Republic stands above any recapitulation of it--even one that many consider brilliant and original in its own right--because Plato's ideas come only from Plato, and understanding them requires as direct a link as possible to him.
I understand the scientist's temptation to line up great classics of science (or excellent popularizations) alongside the Great Books of other disciplines. But science simply differs from the arts in its relationship to its seminal works. And we science-minded folks would all do better to take note of that fact, and point our non-scientific colleagues--just as we would any starting student--at a good, readable, comprehensive textbook, rather than at an original document. After all, if a philosophy professor suggested that I wasn't up to studying his field the way his own students do, and recommended instead that I stick to my scientific ways by studying a "summary of philosophical ideas" textbook in place of the original Plato, I'd be rather offended.