Phrase Explorer

What this shows

Frequency analysis of ordered multi-word sequences (n-grams) in Finnic runosong verses. Unlike collocates (which measure unordered co-occurrence), phrases show adjacent word sequences — the actual stock formulas used by singers.

Bigrams vs Trigrams

• Bigrams — two-word sequences like "vanha väinämöinen" (old Väinämöinen) or "kuldane kaeketine" (golden necklace)
• Trigrams — three-word sequences like "ei või tõusta" (cannot rise) or "yks kaks kolme" (one two three)

Phrase types

• alit — alliterative phrase (all words start with the same letter)
• rep — repetitive (same word repeated, typical of refrains)

Difference from Collocates

Collocates measure words appearing anywhere in the same verse (unordered). Phrases show actual adjacent word sequences (ordered). "kuldane" and "hõbedane" are collocates (same verse). "kuldane kaeketine" is a phrase (adjacent words).