Phonology tab
Choose which tokens to analyse and (optionally) how to limit them, then pick a single view. Each token is normalised (diacritics stripped, final ς→σ) and syllabified by the maximal-onset principle; the segmental and syllabic views are derived from that single pass, so their numbers are mutually consistent.
- Segments: frequencies, positional distribution inside the word (including the final law read directly off the corpus), bigram phonotactics with over- and under-represented transitions (PMI), and the functional load of each segment contrast measured by minimal pairs.
- Syllable structure: shapes, cluster inventories, and the sonority contour of complex onsets, with the non-rising exceptions itemised.
- Prosody and sandhi: the views that go beyond orthography by reading the merged metrical record: weight by nature checked against weight by position (surfacing muta cum liquida and correption), the dichrona α ι υ resolved by how they scan, accent placement against the limitation laws, and elision and hiatus at word junctures in verse order.
1. What to analyze
Choose which tokens to analyse and (optionally) limit them, then run. For full control, open Advanced.
A read-only query selecting the tokens to analyse; it must return at least one text column of word forms. Ctrl/Cmd+Enter runs.
Waiting for corpus…
2. Phonological view
Note: the segmental and syllabic views are heuristic analyses over normalized orthography (letters as phoneme proxies, orthographic syllabification), not a reconstruction of Ancient Greek phonology. The views under “Prosody and sandhi” additionally read the corpus itself: the per-word metrical record, the accented forms, and word order in the line.