This is not a bad book per se, just a disappointing one. I expected to learn much more about the structure and function of everyday conversation, but I found the work (and it does feel like work getting through it at times) to be less informative in that regard than standard works on hermeneutics or pragmatics. I am reminded of the quip about Wittgenstein's Tractatus: the journey is long and the distance covered is comparatively short.
But there is much to learn here nevertheless. Goffman in Forms of Talk showed an astute eye, or ear, for a textuality of language which falls off the radar of mainstream linguistics. Hence, another criticism I have is that the domain of language he covers is too marginal or peripheral to be of much use to a grammatically minded linguist. This is no book on grammar. He looks at phenomena ranging from the nearly pre-cultural act of saying 'oops' to oneself to the over-wrought asides of media professionals. But he manages to bring it all home by using these seemingly inconsequential examples as his laboratory on natural speech. He does this by returning time and again to his theories, based largely on the arch-metaphors of footing and frame. Using these concepts, he minutely examines the places where our common sense about how language works breaks down. Language use is more than a tit for tat of turntaking and information sharing. Language performance can be this back and forth, but it also entails shifts in the social dynamic (say, in a lecture or a radio announcement where the audience is bracketed off and not in a dialog in any strict sense). Or language use can be directed at no one, as in the case of semi-involuntary but socially witnessed response cries.
This book can also lead the reader to grasp the framing of discourse in novel ways. For example, the book made me think about media discourse as being the model for interpersonal discourse and vice versa. Political Correctness can then be seen as the model for not offending an unseen diversity on the receiving end of media discourse, and this in turn can serve as a sociolinguistic model for speakers to follow outside of the mediated context.
I did not like the lecture in the middle of the book that was about 'the lecture'. It seemed more tedious than reflexive. The best parts of the book were at the beginning, where responses and replies were given a micro-pragmatic analysis, and at the end where radio talk is shown to be highly relevant to the understanding of discourse mishaps and listener charity when the possibility of back-channel cues is present.
The book says a lot about a little, which makes it boring at times. I just wish it had said more.
The chapter and Footing was the most illuminating. Radio Talk is pretty long for an example. Responses and Replies dwells into linguistic analysis and can be tedious.