Daejeon, July 27–31
Scholars have long argued that Western literary canons remain shaped by exclusions along gender, race, and class (e.g., Guillory 1993; on gender, e.g., Showalter 1971; Russ 1983; Whittle 2013; for an overview, see v. Heydebrand/Winko 1996; Rippl/Straub 2013). Not least to ground such criticisms in a solid empirical basis, canonicity has increasingly been examined through empirical, quantitative approaches (see Related Work). Following these strands, we examine the factor of gender in the current academic canon of German-language literature. We confine our analysis to the categories “female” and “male”, a choice driven by data availability (see Gender in the GND) and not by our conceptual understanding of gender. We examine the connection between gender, canonicity, and author time period
In the following, the term “author time period” refers to the time period in which authors lived, not the time for which canonicity is reconstructed. The latter is always the current time (2025).
In recent years, empirical reconstructions of canonicity have become increasingly common (Algee-Hewitt/McGurl 2015; Algee-Hewitt et al. 2016; Barré/Champs/Poibeau 2023; Brottrager/Stahl/Arslan 2021; Fredner/Porter 2024; González et al. 2021; Rohe et al. 2025). Several studies have already examined gender-related aspects, consistently showing that the share of female authors in canons remains far below 50%, yet generally increases across literary periods, albeit to varying degrees (e.g., Müller-Spitzer/Wortmann/Wolfer 2023; Fredner/Porter 2024; Schmidgall 2025; Teichmann 2026).
In our previous research, we argued that empirical reconstructions of canonicity must be grounded in a robust theoretical framework (Hilger/Eckardt/Kröncke 2026), and we reconstructed the academic canonicity of German-language authors (Eckardt/Hilger/Kröncke 2027). Our reconstruction advances previous research by drawing on a particularly broad range of empirical sources and by not being tied to a predefined text corpus. The present study builds on this earlier work by offering the first analysis of gender regarding this resource of canonicity reconstruction. It also advances methodology by contextualizing the observed gender distribution with the total population of authors.
The present section leans strongly on recent work by our group (Hilger/Eckardt/Kröncke 2026). We understand canonicity as a gradual phenomenon and define the canonicity of object O at time t in relation to a carrier group G as the degree to which O has been deemed worthy of attention within G up to and including t. In our case, O refers to German-language authors from c. 1600 onward, t is the year 2025, and G consists of German-language literary scholars.
To reconstruct canonicity, we draw on six indicator groups (Eckardt/Hilger/Kröncke 2027): (1) the number of hits in the central subject bibliography for German-language literary studies ( BDSL); (2) article length in a lexicon ( Killys Literaturlexikon); (3) the number of pages on which authors are mentioned in two literary histories ( Brenner 2011 and Beutin et al. 2013); (4) mentions in 42 university reading lists; (5) mentions in course catalogs from three universities; (6) the number of titles by the author in a widely distributed edition series ( Reclams Universal-Bibliothek).
We compute a canonicity score for each author mentioned in at least one of the indicators. We represent values for indicator groups by the median across their individual documents. The canonicity score for each author is the median across the six indicator groups. All indicator values have been z-scored to ensure comparability.
The resulting scores for 5,167 authors, along with a detailed documentation, are available at https://github.com/agihil/germancanon.
We use the central German-language authority file (GND)
https://www.dnb.de/DE/Professionell/Standardisierung/GND/gnd_node.html (last accessed 22.04.2026).
We argue that adequate baselines are needed for the contextualization of canonicity distributions. In addition to a 50% baseline, we compare our data with the overall population of authors. As an approximation of this population, we select all “authors,” “writers,” etc. associated with German-speaking countries and born after 1550 from the GND, which yields around 79,000 individuals. While this set is certainly not identical to the true population of authors and is part of the canon-making process itself, it should come far closer to the population than the body of people referenced by the canonicity indicators.
Within the scope of this article, we cannot determine with certainty to what extent the writers listed in the GND approximate the full population. Since 1913, the German National Library, which manages the GND, has operated under a collection mandate that includes media works either published in Germany or in the German language. Some studies discuss and explore the use of GND data to estimate the overall population (e.g., Fischer and Jäschke 2018; Herrmann et al. 2021). Unlike those projects, our study operates at the author level rather than the textual level. This simpler use case suggests broader coverage.
A glance at the upper end of the canonicity distribution (Figure 1) not only shows that it follows a power law, but also that it is overwhelmingly dominated by male authors. The five most canonical authors are Goethe, Schiller, Kleist, Brecht, and Th. Mann. The first non-male author is Bachmann (rank 23), followed by Chr. Wolf (rank 30) and Jelinek (rank 36).
Figure 1 already suggests that the more narrowly we focus on the top of the distribution, the higher the male dominance.
For the visualizations, we aimed to design a color map that does not reproduce gendered color stereotypes, yet is not counterintuitive; see Bavdaž et al. 2025.
As outlined above, we believe that there is a connection between canonicity, gender, and author time period. The proportion of female authors among all authors is likely to have increased as a result of, among other factors, improved access to education, reinforced by the feminist movements of the 19th and 20th centuries.
Figure 3 reveals the gender distribution in the overall population of writers. The 16th and 17th centuries show a very low proportion of female writers. From then on, the share of female authors gradually increases, eventually exceeding 50% around the year 1980.
To what extent do the canonicity values reflect this development? When examining the birth years of all authors mentioned in the canonicity indicators (Figure 4, first line), we observe trends broadly similar to those in the estimated population. The proportion of female writers is initially very low but gradually increases, eventually surpassing that of male writers in the 20th century.
However, this somewhat changes when we move further up the distribution (Figure 4, lines 2-–4). The closer we get to the top, the more strongly women born in the first half of the 20th century are represented. Well-known authors from earlier periods such as Droste-Hülshoff, von La Roche, or Lasker-Schüler do not “make it” into the top 50.
Our study shows that the current German-language academic canon is heavily dominated by male authors. However, the share of women rises noticeably among authors born in more recent periods. Compared with the overall population of writers, women remain underrepresented in the canon, particularly in the top ranks. At the same time, the canon’s gender distribution deviates far less from this population than from a 50% baseline. Overall, our findings underscore the importance of accounting for author time period and carefully choosing the baseline when interpreting results.
Since gender likely interacts with other variables such as class and race, future work should consider them as well. The same applies to a non-binary focus on gender.
Even though our approach is descriptive, it carries normative implications. While there are many fruitful ways to increase the share of non-male authors in the academic canon, we want to draw particular attention to the factor of time-period canonicity. Because female authors are far more common in later periods, both in the general population and in the academic canon, one possible strategy would be to further advance the canonization of these periods, especially post-1945 literature, in order to counterbalance the strong emphasis on male-dominated periods such as Classicism/Romanticism.