Poetic Structure & Style Guide
About This Guide
This page is a reference guide to the poetic analysis tools in RunoVerse. Each section describes one explorer page, explains what data it covers, and links directly to the tool.
How to Navigate
Scroll through the sections below or jump to a tool that interests you. Each section ends with a green button that opens the corresponding explorer. The tools work independently — you can use them in any order.
What Is Kalevala Meter?
Kalevala meter is the traditional verse form shared by Finnish and Estonian runosong: trochaic octosyllabic lines with characteristic alliteration and semantic parallelism. The tools on this page let you examine these features quantitatively across over 292,000 poems.
Related Pages
Corpus Guide — poem collections and metadata
Dictionary Guide — lexicon and word lookup tools
Similarity Guide — poem and verse comparison algorithms
Languages Guide — cross-lingual features and cognates
Scansion — metrical analysis of individual verses
The Kalevala-meter tradition has distinctive poetic features — alliteration, semantic parallelism, formulaic phrases, and characteristic meters. These tools let you explore those features across the combined Finnish and Estonian corpus of over 292,000 poems and 4.35 million verse lines.
Alliteration Explorer
Alliteration — the repetition of initial sounds across words in a verse — is one of the defining features of Kalevala-meter poetry. The Alliteration Explorer lets you examine how initial sounds are distributed across 4.35 million verse lines and discover which sound combinations are most common in the tradition.
- Browse alliterative word pairs grouped by initial sound (a–a, k–k, s–s, etc.)
- See frequency distributions of initial sounds across Estonian and Finnish verse
- Find the most densely alliterative verses in the corpus
- Compare alliteration patterns between Estonian (ERAB) and Finnish (SKVR, JR) traditions
Parallelism Explorer
Semantic parallelism — expressing the same idea in two or more consecutive lines with systematic word substitutions — is perhaps the most characteristic structural device of runosong. The Parallelism Explorer offers two complementary ways to study this feature.
Consecutive parallelism
Covers 200,627 poems where adjacent verse lines express similar ideas through word substitutions (for example, "brother" and "sister", or "gold" and "silver" appearing in corresponding positions). Detection uses a 5-signal scoring system combining word embeddings, BERT similarity, morphological matching, cluster membership, and pair frequency.
Formulaic recurrence (substitution clusters)
Covers 102,065 poems with word substitution patterns organized into 2,206 clusters containing 8,708 words. These clusters reveal which words are interchangeable within poetic formulas — showing, for instance, that "gold", "silver", and "copper" regularly occupy the same slot in repeated verse patterns.
Open Parallelism Explorer →Verse Metrics
The Verse Metrics page provides quantitative prosodic analysis of 4.3 million verse lines. Kalevala-meter — trochaic octosyllabic verse — is the traditional form shared by both the Finnish and Estonian traditions, though actual performance shows considerable variation.
- Examine verse length distributions (syllable and word counts per line)
- Measure alliteration density across the corpus
- Review parallelism metrics for individual poems and across the tradition
- Compare prosodic patterns between Estonian and Finnish verse
Phrase Explorer
Oral-formulaic composition relies on recurring multi-word phrases — stock expressions that singers draw on to build verses in performance. The Phrase Explorer identifies frequently recurring bigrams and trigrams across 4.35 million verse lines.
- Browse the most common two-word and three-word phrases in the corpus
- Identify alliterative word combinations that function as poetic building blocks
- Discover stock phrases characteristic of specific poetic themes or regions
- Filter by language to compare Estonian and Finnish phraseology
Word Collocates
The Collocates tool uses Pointwise Mutual Information (PMI) to find words that appear together significantly more often than chance would predict. This reveals formulaic word pairs and deep poetic associations that are not obvious from simple frequency counts.
- Search for any word and see its strongest collocates ranked by PMI score
- Discover formulaic word pairs that form the backbone of oral poetic composition
- Identify poetic associations — words that the tradition regularly links together
- Explore co-occurrence patterns across 4.35 million verse lines
Emotion Vocabulary
The Emotion Vocabulary explorer organizes emotion-related words from the runosong corpus into semantic domains. A sidebar lets you navigate emotion word families — from joy and sorrow to anger and fear — to study the emotional landscape of folk poetry.
- Browse emotion words organized by semantic families (joy, sorrow, anger, fear, etc.)
- See how frequently each emotion word appears in the corpus
- Explore connections between emotion words and the poetic contexts where they occur
Emotion Map
The Emotion Map provides an alternative, visual approach to the same emotion vocabulary. Instead of a sidebar list, it displays emotion words as interactive chips, making it easier to see how emotion families cluster and relate to each other across the corpus.
- Visual chip-based layout showing emotion word distributions at a glance
- See relative sizes and relationships between emotion families
- Quickly identify the most prominent emotion domains in the tradition