search

biblicalhumanities / greek-new-testament

Greek New Testament
45 stars 18 forks source link

Missing morphological information from Nestle1904 trees #4

Open jonathanrobie opened 9 years ago

jonathanrobie commented 9 years ago

Morphological data is missing in the following words of Nestle1904, resulting in empty tags in the pruned version.

  1. Matthew: <w morphId="40021030013" class="adj" role="adv" lemma="ὕστερος" case="" gender="" number="">ὕστερον</w>
  2. Revelation: <w morphId="66009012002" class="noun" role="np" lemma="οὐαί" case="" gender="" number="">Οὐαὶ</w>
  3. Revelation: <w morphId="66009012010" class="noun" role="np" lemma="οὐαί" case="" gender="" number="">Οὐαὶ</w>
  4. Revelation: <w morphId="66011014002" class="noun" role="np" lemma="οὐαί" case="" gender="" number="">Οὐαὶ</w>
  5. Revelation: <w morphId="66011014008" class="noun" role="np" lemma="οὐαί" case="" gender="" number="">Οὐαὶ</w>

For instance, here is the Nestle1904 node that corresponds to the first of these:

Compare that to the equivalent node in SBLGNT:

rkjtan commented 9 years ago

This is a case of the different sets of morphology preferring to tag the words differently. The 3 οὐαί cases are tagged according to form as indeclinable by both the Nestle1904 & SBLGNT morphs (I'll just call these morphs Nestle1904 & SBLGNT for short in most of what follows as a matter of convenience). It normally functions as a particle, which SBLGNT treats it as here too. (Notice that GBI added a node to transform the particle to a noun to enable the syntax tree to form.) Nestle1904 treats it as a noun as it is treated as if it were a nominative feminine singular noun (as shown by case, gender, & number of the article in Rev9:12 & 11:14). If case, gender, & number exist for in the 3 οὐαί cases the SBLGNT version, it is a result of some underlying information GBI had put in there to make the syntax trees form (an indeclinable tag doesn't help trees to form; indicating that an indeclinable noun functions like a nominative feminine singular noun & hence is modified by the matching article is what allows the syntax tree to form).

With ὕστερος, the Nestle1904 morph tagged directly according to function--it is functioning adverbially. It also apparently chose to list the adverbial form as a separate form from the adjective ὕστερος, from which it is derived (though the UnicodeLemma should be corrected to ὕστερον in this case). The SBLGNT morph has apparently chosen to tag ὕστερον as just an adverbial use of the adjective ὕστερος, thus still providing case, gender, & number as is customary for adjectives. GBI then had to indicate that the adjective functions adverbially in the tree. In this particular case, it looks like we decided to keep the adjective categorization even for the Nestle1904 to preserve tree unity across versions, while showing in the attributes the different morphological tagging decision for the Nestle1904.