For cogsci, I have read these too much, or read responses to or summaries of them too much:
- Searle's Minds, Brains, and Programs (I get so tired of reading a faulty analogy--this is the one with the Chinese Room experiment)
- Turing's Computing Machinery and Intelligence (really, I know what the Turing test is by now)
- Nagel's What is it like to be a bat? (and I loathe it every time)
- Dennett's Where Am I? (and I love it every time and wish I had been at the "lecture")
- Berlin and Kay findings (linguistic universals in color names), along with the paint chips study by somebody-or-other (I have never read the originals of either!)
- Asimov's robot stories involving Susan Calvin (this is related because I really would love to be a robopsychologist)
- Juster's The Phantom Tollbooth (OK, I never read it for cogsci, but I could have! Anyway I've read it 7 times so it totally belongs on this list)
What have you read too much, particularly in your field?