The data behind RunoVerse
RunoVerse brings together roughly 290,000 Finnic runosong texts (runolaulu, regilaul) from four collections, and about 6,200 Estonian treasure legends. The texts remain under the terms their holders set. This page says where each collection comes from and on what terms it may be used, what RunoVerse has added to the texts, and how to cite them.
The source collections
| Collection | Held by | Poems | Years |
|---|---|---|---|
| SKVR Suomen Kansan Vanhat Runot | Finnish Literature Society (SKS) | 89,247 | 1564–1939 |
| JR Julkaisemattomat runot | Finnish Literature Society (SKS) | 85,607 | 1653–1971 |
| KR Kirjalliset Runot | Finnish Literature Society (SKS) | 10,544 | 1658–1908 |
| ERAB Eesti regilaulude andmebaas | Estonian Folklore Archives, Estonian Literary Museum | 108,969 | 1644–1957 |
The counts are of the collections as RunoVerse received them. The site itself shows slightly fewer poems, because some records, among them duplicates, were left out when it was built.
Terms of each collection
SKVR. The Finnish Literature Society publishes SKVR in its SKVR online service and releases the service's data under the Creative Commons Attribution licence (CC BY 4.0). The service asks to be cited as SKVR-verkkopalvelu 2024–. Toimittaneet Jukka Saarinen, Niklas Alén, Maria Niku, Kati Kallio ja Eija Stark. Helsinki: SKS. (terms and citation, in Finnish).
JR. The same online service holds the JR corpus, the runosong in the Finnish Literature Society's archives that the printed SKVR volumes left out, and asks for it to be cited as JR-korpus. SKVR-verkkopalvelu 2024–. Toimittaneet Jukka Saarinen, Niklas Alén, Maria Niku, Kati Kallio ja Eija Stark. Helsinki: SKS. Its licence sentence speaks of the service's data without naming JR, so for terms beyond citation, ask the Finnish Literature Society.
KR. Kirjalliset Runot, poems from written and printed sources such as the Kalevala, the Kanteletar and the Kalevipoeg, was compiled by Sakari Korpikallio at the Finnish Literature Society within the FILTER and REFOP projects. Its description at github.com/sks190/KR releases it under the Creative Commons Attribution licence (CC BY 4.0), on condition that the Finnish Literature Society and the FILTER and REFOP projects are named as its original compilers.
ERAB. The Database of Estonian Runosongs (Eesti regilaulude andmebaas) is kept by the Estonian Folklore Archives of the Estonian Literary Museum. The database's description page names no open licence. It says that anyone publishing songs from it, in print or online, must name the Estonian Folklore Archives as the source and give each song's archive reference, and it asks that the archives be contacted (era@folklore.ee) before a larger collection based on the database is published. It asks to be cited as Eesti regilaulude andmebaas. Eesti Kirjandusmuuseumi Eesti Rahvaluule Arhiiv. Tartu, 2003-2026.
Treasure legends. The Estonian treasure legends, about 6,200 of them, were written down between 1866 and 1994 (some are undated). Each shows its archive reference, and the texts remain under the terms of the archive that holds them. Their motif types, with their numbers (P 1, P 2 and so on) and names, follow Mare Kalda's type index of Estonian treasure legends: Mare Kalda. Jututüübid peidetud varandustest eesti rahvaluules. Aineloend. Tartu: Eesti Kirjandusmuuseumi Teaduskirjastus, 2024.
What RunoVerse adds
All annotation on the site, base forms included, is automatic and experimental: an aid to reading the texts, not a reference.
- AI annotation. About 290,000 poems were annotated with the DeepSeek-R1 language model by Lidia Pivovarova (University of Helsinki), following the method of Pivovarova et al. (2025): base forms, English translations, morphological descriptions and etymological roots for each word.
- Base forms (lemmas). RunoVerse's own automatic base forms: for Estonian texts with EstNLTK together with dialect and runosong word lists and expert annotations, for Finnish texts with Omorfi, Voikko and Stanza, supplemented by three dictionaries.
- Dictionaries. Word entries link to nine Estonian, Finnish and Karelian dictionaries, listed on the About page; their texts remain under each publisher's terms.
- Places. Place coordinates and historical parish and commune outlines come from registers and maps listed on the About page.
- RunoVerse's own work. The site itself, its indexes and visualizations, and the data made for it, such as the English names of song types, the clustering and the similarity calculations, are © Kaarel Veskis.
Reuse
For the texts and the AI annotation, the terms of their sources apply, as given above. Use of RunoVerse's own work in research and teaching, with attribution, is welcome; for other uses, please get in touch (veskis.kaarel@gmail.com). Bulk reuse or redistribution of RunoVerse's own data files needs the author's prior agreement. The full wording is in the Rights and reuse section of the About page.
How to cite
RunoVerse. Cite RunoVerse as follows (a BibTeX entry is on the About page):
Veskis, K. (2026). RunoVerse: A Suite of Tools for Exploring Finnic Runosong Corpora (Version 1.0) [Web application]. https://runoverse.org/
A single poem. The Cite button beside a poem's title in the Reader gives a citation that names the source collection or archive first, with RunoVerse as the place it was read. Links of the form reader.html?poem=... are stable.
The collections. As their holders ask, above.
Catalogue records
RunoVerse is also described in these catalogues:
- LINDAT/CLARIAH-CZ (CLARIN): https://hdl.handle.net/11372/LRT-6312
- Wikidata: https://www.wikidata.org/wiki/Q141343271
- SSH Open Marketplace: https://marketplace.sshopencloud.eu/tool-or-service/WI4970