Literary Pattern Recognition: Modernism between Close Reading and Machine Learning

NovelTM team members Hoyt Long and  Richard Jean So have just published a new article in Critical Inquiry on how humans and machines think about poetic texts.

Literary Pattern Recognition: Modernism between Close Reading and Machine Learning

Hoyt Long and
Richard Jean So

The title of this essay announces its core ambition: to propose a model of reading literary texts that synthesizes familiar humanistic approaches with computational ones. In recent years, debates over the use of computers to interpret literature have been fierce. On one side, scholars such as Franco Moretti, Matthew Jockers, Matthew Wilkens, and Andrew Piper defend the deployment of sophisticated machine techniques, like topic modeling and network analysis, to expose macroscale patterns of language and form culled from massive digitized literary corpora.1 On the other side, scholars such as Alexander Galloway, David Golumbia, Tara McPherson, and Alan Liu, who work in the field of New Media Studies, have criticized machine techniques for reducing the complexity of literary texts to mere “data” or for being incommensurable with the goals of critical theory.2 Here we move beyond this impasse by modeling a form of literary analysis that, rather than leveraging one mode of reading against another, synthesizes humanistic and computational approaches into what we call literary pattern recognition.

The entire article as published in Critical Inquiry Vol. 42, No. 2 (Winter 2016), pp. 235-267 can be read  here