There has been a long felt need to investigate what the study of poetic language contributes to an understanding of ordinary language, and how posing the question in this way may indeed shift some of the assumptions about the way ordinary language works. Metaphor remains an important case in point: How do we get from “He kicked the bucket?” to “He died”? In this metaphorical idiom, the process is unavailable, without going into etymological hypotheses, because the meaning is already given – it’s lexicalized, a “dead metaphor” -- no semantic construal is involved and the literal meaning doesn’t play a role in understanding the meaning of the expression. A poet might break up the idiom and say, “He kicked the bucket and broke it to pieces” – defamiliarizing the automatic reading of the idiom as “he died” and opening up suggestions, for example, of overcoming death; thus lending plasticity to and making present both the figurative and literal meaning. But true access to the process of metaphorical meaning-making is only available in cases of novel poetic metaphor, where the domains brought together are often disparate. I argue that there are good reasons for seeing poetry, fiction and literature in general not as speech acts but as representations, or semblances of speech acts (Searle’s notion of a pretend speech act doesn’t do this justice). Poems stage dramatic situations in which both poetic speaker (that is, the lyrical “I,”) and addressee, if any, are part of a fictional world, rather than the biographical poets themselves addressing a reader. The often ironic gaps and dialogical tensions between the norms and values projected by the poem itself and those expressed by any one of the voices in the poem – including the lyrical ‘I’s perspective – have a crucial expressive force unto themselves. As Searle realized, the indirect speech act model for poetry does not work; whereas “Could you pass the salt?” is a request masquerading as a question, there is no such one-to-one relationship between the poetic utterance and a single determinate speech act hiding behind it. Searle’s and Austin’s approaches couldn’t fully succeed, on my view, because there are key aspects of the language-game of poetry that distinguish poetry from ordinary language speech acts. Let’s say these are three aspects of a Gricean poetic contract between speaker and hearer: 1) the distinction between the lyrical ‘I’ and the biographical poet; 2) what I call the "quasipropositionality" of utterance in poetry (and fiction as well); and 3) poetic form and the ways in which it makes meaning through textual strategies of mediation, like point of view, parody, irony, enjambment, rhyme, meter, rhythm, sound-play, allusion, and metaphor. These forms of aesthetic mediation make it very problematic to isolate the serious or not-serious utterance of a single-line as a primary semantic unit.