(1) Overview
Repository location
The dataset is openly available via Zenodo at the following DOI: https://zenodo.org/records/15408113
Context
This dataset was produced as part of the Horizon Europe MSCA Doctoral Network project CASCADE (101119511) (www.horizoncascade.net), which focuses on understanding how language expresses meaning in various contexts, particularly over time. The data has been used in the peer-reviewed paper Rewriting Tradition: Quantifying Change in Lady Gregory’s Irish Legends, which will be presented at the Digital Humanities Conference 2025, taking place from 14–18 July 2025 in Lisbon, Portugal. The paper explores computational methods for measuring textual transformation between original Irish folk tales and their adaptations. Supporting materials and analysis code are available via the accompanying GitHub repository: https://github.com/MSCAcascade/Gregory-vs-Original-Irish-Legends.
Lady Augusta Gregory was an Irish dramatist, folklorist, and key figure in the Irish Literary Revival. She co-founded the Abbey Theatre with W. B. Yeats and was deeply invested in preserving and reinterpreting Irish oral tradition for modern readers. Her 1903 collection Cuchulain of Muirthemne adapts tales from the Ulster Cycle – one of the major cycles of Irish mythology – rendering them into English prose that is accessible yet literary in tone. Gregory’s aim was to present Irish heroic legends in a unified, national form. The book was well received in her time and played a significant role in shaping perceptions of Irish cultural identity.
(2) Method
Steps
The dataset consists of 28 plaintext files: 14 English-language Irish legends and 14 English-language corresponding literary adaptations by Lady Augusta Gregory. The original legends span in date from the early 12th century to the late 14th century, and the earliest available reputable English translations were sourced, in part, from the Corpus of Electronic Texts (CELT) hosted by University College Cork (https://celt.ucc.ie) (Best, Bergin, O’Brien, & O’Sullivan, 1954–1983; Cross and Slover, 1996/1936; Henderson, 1899; Hull, 1898; Leahy, 1905; Meyer, 1883–1885; Muller, 1878; Stokes, ca. 1910; Van Hamel, 1933). The corresponding texts adapted by Lady Gregory were taken from her 1903 collection Cuchulain of Muirthemne (Gregory, 1903). Texts were curated manually to ensure accurate pairing between source material and adaptations. A metadata.xlsx file was created to map each Lady Gregory adaptation to its corresponding folk tale. All files were saved in UTF-8 encoded .txt format to ensure compatibility with standard text analysis tools. File naming conventions (lowercase for originals, uppercase for adaptations) were implemented to support automatic matching.
Sampling strategy
Fourteen Irish folk tales were selected based on the availability of both a traditional version and a corresponding adaptation by Lady Gregory. The selection aimed to capture a range of narrative types and themes representative of Irish oral tradition and Lady Gregory’s literary treatment. The dataset is not intended to be exhaustive but to offer a representative and analysable corpus.
Quality control
Text normalization steps included the removal of non-standard characters and encoding all files in UTF-8. Metadata was validated for consistency and completeness.
(3) Dataset Description
Repository name
Zenodo
Object name
Irish Folk Tales and Lady Gregory Adaptations (Plaintext Collection)
Format names and versions
Plaintext (UTF-8 encoded .txt files), Microsoft Excel (.xlsx)
Creation dates
2024-11-02 to 2024-11-11
Dataset creators
Rachel McCarthy – Primary curator and compiler, CASCADE researcher, University College Cork
Language
English
License
Creative Commons Attribution 4.0 International (CC BY 4.0)
Publication date
2025-05-14
(4) Reuse Potential
This dataset is designed for high reusability across several domains. In digital humanities, it can support stylometric analysis, authorship attribution, and computational literary studies. In folklore and cultural studies, it enables exploration of adaptation, oral tradition, and the transformation of narrative structures – highlighting how orally transmitted narratives were recorded, translated, and later adapted into literary form. This facilitates comparative analysis of the stylistic, structural, and thematic shifts that occur when oral material is reshaped to fit literary conventions and nationalistic or aesthetic agendas. By capturing both the original versions of oral legends and their literary reworkings, the dataset contributes to ongoing research into how literary works draw on and transform oral sources. It enables the tracing of intertextual relationships and cultural transmission across time, media, and modes of storytelling.
Natural language processing researchers may also use it for fine-tuning or evaluating models on literary or historical text corpora, including style transfer and text similarity tasks. The clear pairing of source and adaptation allows for aligned corpora analysis. The dataset can also serve pedagogical purposes in literature and DH courses, offering accessible primary materials for student assignments and projects.
Limitations include the relatively small size of the corpus (14 pairs of texts, 28 texts in total) and its focus on English-language materials, half of which were translated from Irish, which may limit linguistic diversity for certain computational applications. However, its compact and curated nature also enhances manageability for focused studies and use.
Acknowledgements
A selection of the original folk tales was kindly provided by the Corpus of Electronic Texts (CELT) project, University College Cork: https://celt.ucc.ie. We are grateful for their work in preserving and making available Irish textual heritage.
Competing Interests
The author has no competing interests to declare.
Author Contributions
Rachel McCarthy – Data Curation.