This dissertation brings together aspects of writing-in-the-disciplines research, reader-response theory, and empirical reading research in an investigation of literary scholars reading poems and constructing arguments. I begin with a review of literary criticism published over the past 70 years on Donne's "The Flea," Milton's "Song: On May Morning," Hopkins' "God's Grandeur," and Eliot's "Conversation Galante." This review suggests that certain New Critical interpretive conventions persist in scholarship. In particular, literary scholars continue to read lyrics as dramatic utterances and as organic wholes. I then present findings from a think-aloud study in which English professors read the aforementioned poems and planned a hypothetical conference talk about them for the MLA conference. Reader-response theorists have argued that readers activate certain text-making conventions in order to read literature as literature. In my study, participants' disciplinary reading conventions were so deeply ingrained that their initial processing of the four poems mirrored the interpretive patterns in published criticism of those poems. Next I analyze the think-aloud data and follow-up interviews from the perspective of writing-in-the-disciplines research. Previous researchers found that scholarly literary argument relies on a limited set of special topoi and is not always directed toward the accumulation of new knowledge. The scholars in my study relied more heavily on some topoi during initial interpretation of the poems, while other topoi were used more often during argument planning. The picture of literary argument that emerges is a hybrid of ceremonial rhetoric and communal knowledge building. Finally, I analyze the think-aloud data from the vantage-point of expert/novice research in cognitive psychology. Previous researchers have used the term "generic expertise" to describe expert knowledge that all members of an academic discipline possess. Despite the belief of some within literary studies that their discipline lacks a core, participants in my study demonstrated generic expertise both in their interpretations of poems and in their argument planning. I conclude by arguing that previous descriptions of scholarly literary argument need to be revised. Literary scholars relate to their objects of study in a unique way that ensures the distinctness of literary argument. / text