If there's one thing I've thought weird about my particular field of expertise, it is the tendency of experts in computer science to engage in "holy wars," in which a particular practice is irrationally touted as "the one true way," and all opposing tendencies are denounced as evil.
Holy wars emerge on all kinds of matters. Long ago there were editor wars, in which people argued about the one-true-way(tm) to edit text. There were two competing programs in mainframes, vi and emacs. Vi was super-minimalistic, designed for low-bandwidth connections. Every byte counted, so it avoided sending extraneous information. Emacs was a highly detailed editor that allowed you to do all kinds of tasks to the text while you edited it. This was helpful because much of the text being edited was program-code, and emacs could compile it for you, show you the results, and bring you back to the code. With a keystroke. Emacs could also debug, spell check, and if you were feeling depressed about your code not working, it had "ELIZA," an imitation psychologist based on a school of psychology that rephrased your questions to give you a new perspective.
In the editor wars, the opposite tendencies of each side were routinely mocked. Vi fanatics ridiculed emacs's complexity ("Escape-Meta-Alt-Ctrl-Shift," because emacs editing often involved hitting chords of keys to trigger particular tasks), the code size, ("Eight Megabytes And Constantly Swapping". Eight megabytes was a lot of memory back then.) and the tendency to do everything within the editor. ("It's not an editor, it's an OPERATING SYSTEM.") Emacs fans found Vi oversimplified, and therefore stupid. But rather than see it as a matter of preference (which it was), the other faction was "evil" for not accepting your assumptions as true.
Celebrities in the field also promote holy wars. Let us take the case of Edsger Dijkstra. Mr. Dijkstra is a brilliant programmer who invented the shuttling yard algorithm, taught at Texas A&M, and seriously caused a revolution in networking. He also has an utter hatred of unconditional jumps as a programming technique. Most languages use the keyword "goto" to do unconditional jump, and Mr. Dijkstra wrote an essay denouncing it as "Goto considered harmful." He also despises COBOL, although to be fair, so do most people who have used it.
These little quibbles are called "Holy wars" after their resemblance to "religious" fights in the past -- the stakes are small, the fighting is vicious, and after all is said and done, little is gained. The practitioners are convinced of the superiority of their own way of doing things and offended by the very existence of alternatives.
Currently running holy wars in computer science include Windows vs. Macs, GPL vs. BSD, proper tabbing for programming (in which tabs often suggest which loops belong where, but how many spaces per tab is hotly contested), GUI vs CLI, KDE vs. Gnome, and scripting vs. compiling.
The strange thing is, I really don't see much of this in other fields. I don't see Jungian psychologists having a slap-fight against the Freudians. I don't see vicious, insult-riddled debates between architects, even among the many schools of design with conflicting ideas. I've yet to hear of the Cubist painters swearing that Pointillism was eviler than eating babies. Is this because I talk to more Computer Scientists than other majors, or does CS just inherently attract mostly insufferably picky people?