Influence

I can see how "Influence" may have once been viewed as the definitive study of the "science" of influence and persuasion, but its dependence on a steady stream of questionable behavioral science experiments and dubious after-the-fact interpretations ("I offered my college students a bonus test point if they got their parents to complete a survey and the 92% response rate proves the principle of kinship unity!") makes it hard to take seriously. There's value and insight here, but a good summary of the book would probably be a better place to extract it.