EcoCor-DE
EcoCor-DE Corpus Description The corpus contains 183 narrative fiction texts from 1754 to 1933. Genres are heterogeneous and include, e.g. novels, short stories, children's and youth literature, and travelogues. Texts stem from 94 authors. 101 texts are written by male authors,
EcoCor-DE
Corpus Description
The corpus contains 183 narrative fiction texts from 1754 to 1933. Genres are heterogeneous and include, e.g. novels, short stories, children’s and youth literature, and travelogues. Texts stem from 94 authors. 101 texts are written by male authors, 83 by female authors; all texts are out of copyright. Text lengths vary from 119 to 223.836 words. Text sources are d-prose (including texts from Gutenberg, Textgrid, and KoLiMo), and Projekt-Gutenberg. The corpus has been created at the Hackathon “Environments as Networks” at the University of Potsdam 15. – 17. April 2026.
Corpus Compilation Criteria and Workflow
EcoCor is designed as a sample of ecocritically relevant narrative texts. Ecocritical relevance is understood in its qualitative as well as quantitative meaning. Texts are ecocritically relevant if they have already been analysed from an ecocritical perspective, if they show an outstanding number, diversity, or relative frequency of ecological entities (i.e. animals and plants), or if they include important or interesting depictions of environmental aspects. The criteria of corpus creation accordingly have been:
- Reference in at least one source of ecocritical research
- Outstanding quantities of ecological entities in a reference corpus
- Important or interesting depictions of environmental aspects
The corpus was compiled using two approaches:
a) Distant Reading in the form of second-hand criticism (cf. Moretti 2013): Academic works from Ecocriticism have been read, and all primary sources noted. Sources out of copyright have been included in EcoCor.
b) Computational Analysis: Works identified as having a high or notably diverse frequency of animals and plants were included. A dictionary-based approach was used to identify ecological taxa in terms of their frequency and diversity. In order to reach authorial gender parity, we thus analysed two corpora: The corpus of German language fiction (Fischer and Stroetgen 2017) and all (320) female-authored texts from d-prose (Gius et al. 2021). From both corpora, the following quantitatively outstanding texts were included in EcoCor:
10 texts with the highest relative frequency of animals and plants; without lower quartile; before mean(year)
10 texts with the highest relative frequency of animals and plants; without lower quartile; after mean(year)
10 texts with highest diversity of animals; without lower quartile; before mean(year)
10 texts with highest diversity of animals; without lower quartile; after mean(year)
10 texts with highest diversity of plants; without lower quartile; brfore mean(year)
10 texts with highest diversity of plants; without lower quartile; after mean(year)
10 texts with highest frequency of animalst; with
📤 Share this page
Found this useful? Share it with your network.
Files are hosted on the source repository. Click download to access the full dataset.