Writing by numbers: case studies in digital art history

Bowman, Mike Charles (2022) Writing by numbers: case studies in digital art history. PhD thesis, Birkbeck, University of London.

Abstract

While art history claims to be among the most interdisciplinary of fields, the use of digital resources and methods to create new knowledge or rethink its traditional questions is less developed than other disciplines. My thesis contributes to narrowing that gap and promotes a kind of art writing where the numerical stands alongside the verbal and the visual as an essential part of the interpretations being offered. The approach I use is the case study. Each draws on different kinds of textual and numerical data, and statistical methods for processing that data, to address some of the areas where digital art history is underdeveloped. In the first I use statistical methods to analyse the structure and content of the catalogues and criticism of the nineteenth-century Paris Salon. The readings I develop show how that language was involved in practical, conceptual and institutional change in the nineteenth-century French art world. My readings shed new light on that art world, and extend or challenge existing scholarship. In the second I use linear regression to model auction sales with twelve contemporary artists. My models give an understanding of collectors' preferences with different characteristics of the artworks sold, and of some of the ways in which the contemporary art market has been changing. With traditional art historical methods, it is not possible to develop the kind of disaggregated perspective on collectors’ preferences I present in this case study. For my third case study I created a data set of the metadata for 59,000 artworks in the online collections of 35 modern and contemporary art museums. I use several techniques individually and in combination to identify trends in that metadata, to which I give art historical interpretations. These include correspondence analysis, topic modelling and parts-of-speech parsing. I set out a way of thinking about the history of modern and contemporary art in terms of ongoing concerns and interests which rise and fall in importance, and which cuts across conventional narratives of artist, period or movement. My case studies are illustrations of the art historical value of using statistical methods to address textual and numerical data. They serve as examples for other art historians interested in developing that approach within their own work. There are also lessons for those working across the digital humanities from my work. My readings show how drawing together different data sources or statistical methods can support richer interpretations than comes from using them in isolation.

Metadata

Statistics

DownloadsShow export options

Export as XMLJSONCSV

Activity Overview

6 month trend

341Downloads

6 month trend

274Hits

Additional statistics are available via IRStats2.

Archive Staff Only (login required)

Edit/View Item