Mova0.6 pre-release⌕
☰
home / docs / languages / grc / The lexicon: word formation, compounds, loans and dictionaries

The lexicon: word formation, compounds, loans and dictionaries

Greek has a large vocabulary and makes new words easily: by suffixes, by prefixes on verbs, and above all by compounds, which Latin used little and Greek freely. Through Latin and later directly, it gave the European languages most of their words for science, philosophy, medicine and the Church. This chapter gives:

Contents

  1. Suffixes: nouns, adjectives, verbs
  2. Compounds
  3. Verbs with prefixes
  4. Words Greek borrowed
  5. Greek in Latin and English
  6. The open dictionaries and lexical data
  7. How Mova handles it
  8. Sources

1. Suffixes: nouns, adjectives, verbs

(smyth-1920 §§838–868)

Nouns:

Adjectives:

Verbs:

2. Compounds

Greek compounds join two stems, the first usually with a linking vowel -ο- (smyth-1920 §§869–899):

The accent of compounds is in chapter 2. A compound's head is usually its second member. A lemmatizer meeting an unknown compound can split it at the linking vowel and look up the parts: φιλο-σοφ-ία.

3. Verbs with prefixes

Most Greek verbs take one or more prepositional prefixes, often with a spatial or completive sense (chapter 12, §3):

The prefix may change the meaning entirely. The augment stands between it and the verb (chapter 7, §4), so that a lemmatizer must strip the prefix, the augment and then find the stem: κατ-έ-λυσε → καταλύω.

4. Words Greek borrowed

5. Greek in Latin and English

Greek words came into Latin from the earliest contacts: everyday words (oleum), the learned words of philosophy and rhetoric, and the vocabulary of the Church (la chapter 11, §7). Through Latin and French, and from the Renaissance directly, they make much of the learned vocabulary of English:

6. The open dictionaries and lexical data

What the Scientist holds for Greek (grc-lexres), with each item's licence. The folder sa holds share-alike data, nc non-commercial:

resourcecoverslicenceuse in Mova
Abbott-Smith, Manual Greek Lexicon of the New Testament (1922), TEI editionthe New Testament, with the Septuagint and Hebrew equivalentstext public domain; the edition's licence to checkthe first candidate for a Koine lexicon
Dodson, Greek Lexiconthe New Testament, short glossesCC0can be built in
Strong's Greek dictionary (Open Scriptures)New Testament words by Strong's numberpublic domainthe link to the Bible tools and to STEPBible
STEPBible datamorphology, lexicon and alignment of the Greek BibleCC BY 4.0morphology and lemmas of the New Testament and the Septuagint
MorphGNT (SBLGNT morphology)every word of the New Testament parsedCC BY-SAmeasure and learn only
Morpheus (Perseus)a morphological analyser of classical Greekshare-alikemeasure and learn only
Middle Liddell, Autenrieth (Alpheios)classical; Homershare-alikemeasure and learn only
LSJ (Liddell–Scott–Jones), Unicode editionthe whole of classical Greekin nctests only

So for the Koine there is an open path: Abbott-Smith, Dodson, Strong's and STEPBible give lemmas, glosses and morphology under free licences. For classical Greek the main tools (LSJ, Morpheus) are share-alike or non-commercial. Mova may measure and learn with them, but not build them into a release (Data and model licences).

7. How Mova handles it

Mova has no Greek lexicon yet. Its Greek lemmas come only from the treebank it parses, and a word not in the treebank has none. The plan, in order of what is open:

  1. The Koine lexicon from Dodson (CC0), Strong's (public domain) and STEPBible (CC BY), linked by Strong's number to the English Bibles Mova already reads. This would put Greek words into the meta-world of the parallel Bibles.

  2. A lemmatizer by rule:

    • the contraction table (chapter 8);
    • augment and reduplication stripping (chapter 7);
    • the prefix list (§3) and the dialect alternatives (chapter 16);
    • the -μι verbs and the suppletive verbs as a closed list (chapter 9).

    Every rule is deterministic, and each can be tested against the treebank lemmas, error by error.

  3. Compound splitting at the linking vowel for unknown words (§2), as the English lexicon guesses from suffixes.

8. Sources

This page in Markdown → https://md.mova.elamur.ai/free/en/docs/languages/grc/18-lexicon/