Description : We may observe a growing preoccupation in archival literature with characterizing digital archives as big data – a term that suitably captures both their scale and their potential for manipulation through the application of computational methods and techniques for the purposes of discovering new insights. The possibilities for working with digital archives as data, from supporting archival arrangement and description tasks to promoting the use of digital archives as data sets by researchers, are indeed encouraging. But what are digital archives becoming when they are reframed as data, big or otherwise? What consequences might such a conceptualization have for the ways archival professionals imagine their role and their work? To the four archival paradigms of evidence, memory, identity, and community theorized by Terry Cook, a fifth may now be poised to emerge: an archives-as-data paradigm. In this article, I begin to map out what an archives-as data paradigm could entail by exploring how the conceptual and practical dimensions of applying computational methods to digital archives may work conservatively to revivify notions of archival neutrality. For an archivesas- data paradigm to realize the more liberatory aims of which it is capable, an active and ongoing commitment to recognizing and calling out these tendencies is necessary.