Although digital humanities (DH) has received a lot of attention in recent years, its status as “a discipline in its own right” (Schreibman et al., A companion to digital humanities (pp. xxiii–xxvii). Blackwell; 2004) and its position in the overall academic landscape are still being negotiated. While there are countless essays and opinion pieces that debate the status of DH, little research has been dedicated to exploring the field in a systematic and empirical way (Poole, Journal of Documentation; 2017:73). This study aims to contribute to the existing research gap by comparing articles published over the past three decades in three established English-language DH journals (Computers and the Humanities, Literary and Linguistic Computing, Digital Humanities Quarterly) with research articles from journals in 15 other academic disciplines (corpus size: 34,041 articles; 299 million tokens). As a method of analysis, we use latent Dirichlet allocation topic modeling, combined with recent approaches that aggregate topic models by means of hierarchical agglomerative clustering. Our findings indicate that DH is simultaneously a discipline in its own right and a highly interdisciplinary field, with many connecting factors to neighboring disciplines—first and foremost, computational linguistics, and information science. Detailed descriptive analyses shed some light on the diachronic development of DH and also highlight topics that are characteristic for DH.

Although digital humanities (DH) has received a lot of attention in recentyears, its status as “a discipline in its own right” (Schreibman et al., A companion to digital humanities (pp. xxiii–xxvii). Blackwell; 2004) and its position inthe overall academic landscape are still being negotiated. While there arecountless essays and opinion pieces that debate the status of DH, little researchhas been dedicated to exploring the field in a systematic and empirical way (Poole, Journal of Documentation; 2017:73). This study aims to contribute tothe existing research gap by comparing articles published over the past threedecades in three established English-language DH journals (Computers andthe Humanities, Literary and Linguistic Computing, Digital Humanities Quar-terly) with research articles from journals in 15 other academic disciplines (corpus size: 34,041 articles; 299 million tokens). As a method of analysis, weuse latent Dirichlet allocation topic modeling, combined with recentapproaches that aggregate topic models by means of hierarchical agglomera-tive clustering. Our findings indicate that DH is simultaneously a discipline inits own right and a highly interdisciplinary field, with many connecting factorsto neighboring disciplines—first and foremost, computational linguistics, andinformation science. Detailed descriptive analyses shed some light on the dia-chronic development of DH and also highlight topics that are characteristicfor DH.