Shannon didn't consider intersymbol interference.
Sure he did. That's why the bandwidth term appears in his channel capacity equation.
(stuff deleted) I was probably being unclear. According to my text, Shannon's capacity formula does not account for intersymbol interference. My text also states that Nyquist used intersymbol interference as a limit in his 2-bit rule. I'm considering my text to be correct, because the capacity formula doesn't account for the imperfections (like intersymbol interference) and other problems encountered in practice (envelope delay distortion, the need for echo cancell- ation, etc.). And in your words, it says that you can, but it doesn't say how. For those who just tuned in, intersymbol interference refers to when the residual tail or bounce of a dying pulse interferes with a subsequent pulse.