This dissertation addresses the way a linguistic grammar can yield to poetic organization in a poetic text. To this end, two corpora are studied: the sung lyrics of country music singer Hank Williams, Sr. and the rapped lyrics of gansgta rap artist Snoop Dogg. Following a review of relevant literature, an account of the poetic grammar for each corpus is provided, including the manifestation of musical meter and grouping in the linguistic text, the reflection of metrical grouping in systematic rhyme, and rhyme fellow correspondence. In the Williams corpus, final cadences pattern much as in the English folk verse studied in Hayes and MacEachern (1998), but differ in that there are more, and therefore more degrees of saliency. Rhyme patterns reflect grouping structure and correlate to patterns in final cadences, and imperfect rhyme is limited to phonologically similar codas. In the Snoop Dogg corpus syllables do not always align with the metrical grid, metrical mapping and rhyme patterning often challenge grouping structure, and imperfect rhyme is more diverse, as has been shown to be the case for contemporary rap generally (Krims 2000, Katz 2008). Following Rice (1997), Golston (1998), Reindl and Franks (2001), Michael (2003), and Fitzgerald (2003, 2007), meter, grouping and rhyme are modeled as driving phonological, morphological and syntactic deviation in Optimality Theoretic terms. In the Hank Williams corpus, metrical mapping and grouping constraints are shown to drive a number of linguistically deviatory phenomena including stress shift, syllabic variation and allomorphy, while rhyme patterning constraints govern syntactic inversion. In the Snoop Dogg corpus, rhyme fellow correspondence and rhyme patterning constraints play a more significant role, driving enjambment, syllabic variation, and allomorphy. Some linguistically deviatory phenomena derive from ordinary language variation, e.g. (flawr)~(flaw.[schwa]r), and some do not, e.g. syllable insertion in insista. The latter is more common in the Snoop Dogg corpus. / text