Estonian vs Finnish Runosong Traditions Help

What this page shows

This page provides a side-by-side comparison of the Estonian and Finnish runosong (regilaul / runolaulut) traditions. Both traditions share the Kalevala meter — an ancient Finnic oral poetry form based on trochaic tetrameter with alliteration and parallelism — but developed independently over centuries. The data comes from lemmatized and AI-annotated texts across four major corpora totalling nearly 300,000 poems.

The four corpora

ERAB (Eesti Regilaulude Andmebaas) — the Estonian Runosong Database containing 108,969 Estonian poems collected from 131 parishes across Estonia and the Estonian diaspora. Managed by the Estonian Folklore Archives.
SKVR (Suomen Kansan Vanhat Runot) — the Old Poems of the Finnish People, a monumental 34-volume series containing roughly 180,000 Finnish and Karelian poems collected between the 1820s and 1940s.
JR (Julkaisemattomat Runot) — Unpublished Poems from the Finnish Literature Society archives, adding approximately 11,000 further Finnish poems not included in SKVR.
KR (Kirjalliset Runot) — approximately 10,500 literary Finnish poems from the FILTER project, including the Kalevala, Kanteletar, and Kalevipoeg.

Overview panel

The blue (Estonian) and orange (Finnish) overview boxes show the total number of poems, collection places, unique lemmas, and word tokens for each tradition. These counts are aggregated live from the site's lexicon and place data by language, so they are approximate and can differ from the official per-collection totals on the About page (Estonian: ERAB; Finnish: SKVR, JR, KR).

Part-of-Speech distribution

Each row shows one Universal Dependencies POS tag (NOUN, VERB, ADJ, etc.) with two bars representing the percentage of total word occurrences belonging to that category in each tradition. Click any bar to expand the top lemmas for that category. Differences here can reveal structural contrasts — for example, the Estonian material carries a noticeably higher share of adpositions (ADP) and adverbs, while the Finnish material leans more on nouns and verbs.

Emotional landscape

The emotion comparison draws on the emotion-place index, which maps emotion families (love, grief, anger, fear, etc.) to collection places. Values are normalized per 1,000 poems to allow fair comparison despite the different corpus sizes. The top emotion domains are shown; click a bar to see the contributing places. A higher bar means that emotion domain appears more frequently relative to the number of poems in that tradition.

Shared vocabulary

This section counts lemmas that appear in both Estonian and Finnish poems. The Finnish side covers the oral collections (SKVR and JR); the printed Kalevala-metre collection (KR) is deliberately excluded from this oral-tradition comparison. Despite the shared poetic meter and common Finnic ancestry, the overlap is remarkably small — typically around 1,200 shared lemmas, well under 1% of the total vocabulary. The table lists each shared lemma with its part of speech and per-tradition counts; it is searchable, sortable by any column, and paginated (50 per page). Click a lemma to search for it across the corpus, or a POS tag to open its category page.

Semantic Worlds

This section shows how 32 semantic categories (kinship, animals, plants, body parts, weapons, etc.) distribute across the two traditions, drawn from the thematic song-type bridge. Bars can be sorted by divergence, frequency, or name; click a category bar to see which song types contribute to it. Large divergences point to themes one tradition foregrounds more than the other.

Formulaic Density

This section measures how much of each tradition's verse material is shared, recurring formula, using FILTER verse clusters. Toggle between counting all verses or only those matched to a cluster, and apply poem-length floors to control for very short poems. Per-corpus bars (JR / SKVR / ERAB), a distribution histogram, and the most/least formulaic poems are shown; the "How this is measured" panel explains the method in full.

Key differences

The insight cards at the bottom highlight notable contrasts between the two traditions, such as differences in emotional intensity, vocabulary independence, geographic collection scale, and grammatical structure. These observations are derived from the same data shown in the charts above.

Related pages

Shared Vocabulary — full explorer for cross-tradition lemmas. Cognates Cognate Freq. Song Bridge — etymological cognate families across Finnic languages. Emotion Geography — map-based emotion distribution by place. Dialects — dialect-level vocabulary analysis. Linguistic Features — detailed morphological and syntactic patterns.