From corpus creation to formative discovery: the power of big-data-rhetoric teams and methods

ABSTRACT Rhetoric has been slow to adopt big-data techniques, but that is changing. In this article, we describe the formative work of our rhetoric-data science team on an ideographic analysis of state veteran laws. Our interdisciplinary approach enabled us to build a corpus of more than 7,000 files, segment that corpus into likely public and private laws, and develop dictionaries for discerning individual entitlements, such as waived fees for gun permits. Early results show state-level trends in the number of veteran laws, proportion of veteran laws concerning disabled veterans, and proportion of veteran/disability laws affording individual entitlements. While this article presents early findings, its broader purpose is to contribute to discussions of corpus building, data cleaning, formative analysis, and the value of big-data-rhetoric collaborations. Our experience provides five insights: (1) big-data collection methods can save a public rhetoric project when customary retrieval methods fail; (2) big-data-rhetoric work starts conceptually and becomes concretized; (3) formative big-data rhetoric work can problematize fundamental research assumptions, such as what should be included in a corpus; (4) big-data methods can produce interesting results early, yielding a roadmap for future work; and (5) big-data-rhetoric teams need more guidance from the field.