Searching for the loss of the dative

Post Reply
Stephen Hughes
Posts: 3323
Joined: February 26th, 2013, 7:12 am

Searching for the loss of the dative

Post by Stephen Hughes »

The loss of the final nu of the accusative in pronunciation would make the accusative and dative equivalent in some declensions.

Is it possible to use the trees programs to search for whether ἐν tends to be used to mark the dative for words where the loss of nu in the accusative would make the dative and accusative equivalent in the spoken idiom? In other words, do words that proparoxytone in the nominative and accusative tend not to use ἐν to mark the dative, while others do?

Do the trees programs contain metadata indicating what type of prepositional phrase a prepositional phrase is?
Γελᾷ δ' ὁ μωρός, κἄν τι μὴ γέλοιον ᾖ
(Menander, Γνῶμαι μονόστιχοι 108)
Jonathan Robie
Posts: 4165
Joined: May 5th, 2011, 5:34 pm
Location: Durham, NC
Contact:

Re: Searching for the loss of the dative

Post by Jonathan Robie »

Stephen Hughes wrote:Do the trees programs contain metadata indicating what type of prepositional phrase a prepositional phrase is?
We have metadata that indicates the role of a prepositional phrase. So in this example, class="pp" means it is a prepositional phrase, role="p" indicates that the prepositional phrase is functioning as the predicate of the verb.

Code: Select all

            <wg nodeId="400030120040040" class="pp" role="p">
               <w morphId="40003012004" class="prep" lemma="ἐν">ἐν</w>
               <wg nodeId="400030120050030" class="np">
                  <w morphId="40003012005" class="det" lemma="ὁ" case="dative" gender="feminine"
                     number="singular">τῇ</w>
                  <wg nodeId="400030120060020" class="np">
                     <w morphId="40003012006" class="noun" head="true" lemma="χείρ" case="dative"
                        gender="feminine"
                        number="singular">χειρὶ</w>
                     <w morphId="40003012007" class="pron" lemma="αὐτός" case="genitive"
                        gender="masculine"
                        number="singular">αὐτοῦ</w>
                     <pu>,</pu>
                  </wg>
               </wg>
            </wg>
There are different ways to analyze a text, so you may want to add your own labels - either in the text itself or in a separate file. This is possible too, but it's up to you to do that.
Stephen Hughes wrote:Is it possible to use the trees programs to search for whether ἐν tends to be used to mark the dative for words where the loss of nu in the accusative would make the dative and accusative equivalent in the spoken idiom? In other words, do words that proparoxytone in the nominative and accusative tend not to use ἐν to mark the dative, while others do?
To do this, you would need a list of words that proparoxytone in the nominative and accusative, you could generate that from the text. Then I imagine you would simply count the percent of times that they do / do not use ἐν to mark the dative in prepositional phrases. Is that what you are looking for? The second part would be easy.
ἐξίσταντο δὲ πάντες καὶ διηποροῦντο, ἄλλος πρὸς ἄλλον λέγοντες, τί θέλει τοῦτο εἶναι;
http://jonathanrobie.biblicalhumanities.org/
Stephen Hughes
Posts: 3323
Joined: February 26th, 2013, 7:12 am

Re: Searching for the loss of the dative

Post by Stephen Hughes »

Jonathan Robie wrote:There are different ways to analyze a text, so you may want to add your own labels - either in the text itself or in a separate file. This is possible too, but it's up to you to do that.
So does this means that catergorisations like Wallace's have NOT been included into the trees but could be user defined in individual installations?

In the code, the information that the preposition ἐν takes a dative is not there. Is that left to be implied by an analysis of the whole unit of code from <wg nodeId="400030120040040" class="pp" role="p"> to the end? That seems to come from a "let the text speak for itself" line of reasoning, which in a simple case like the preposition ἐν is going to work, but "letting the text speak for itself" (if I've correctly understood that particular aspect of the rationale behind the code) will produce unintelligent results in the case of attracted relative pronouns. Are verbs (), nouns βασιλεῦς (+gen.) and adjectives ἄξιος (+gen.) marked for what

In terms of general practice, I've never seen it done, but I think that convention for quoting relative phrases should be something like this;
Acts 7:17 full verse quoted wrote:Καθὼς δὲ ἤγγιζεν ὁ χρόνος τῆς ἐπαγγελίας ἧς ὤμοσεν ὁ θεὸς τῷ Ἀβραάμ, ηὔξησεν ὁ λαὸς καὶ ἐπληθύνθη ἐν Αἰγύπτῳ,
Acts 7:17 only a relative clause quoted wrote:ἧς (ἣν) ὤμοσεν ὁ θεὸς τῷ Ἀβραάμ
If just ἧς ὤμοσεν ὁ θεὸς τῷ Ἀβραάμ is quoted, it would seem to somebody who doesn't look into the broader context like ὀμνύειν "to swear" takes a genitive of the thing sworn (to be done).

Specifically for this discussion, does your trees program have what the case of the relative would be without the preceeding context, so that when people work from a nodeId of the relative clause itself, they get an intelligible result. In relative like Luke 1:77 ὅρκον ὃν ὤμοσεν πρὸς Ἀβραὰμ τὸν πατέρα ἡμῶν of course it doesn't matter.

[As a side note: According to LSJ, what is sworn by could have been in the genitive in Classical Greek, but in the New Testament it is in the dative as in Matthew 5:36 "μήτε ἐν τῇ κεφαλῇ σου ὀμόσῃς", where considering that if the final nu of the oxytone accusative τὴ κεφαλή were not used in pronunciation, it would be pronounced identically with the perispomenon dative τῇ κεφαλῇ that this would (could) have been in the classical idiom, but is in New Testament times the dative comes to be marked with the ἐν. cf. Revelations 10:6 ὤμοσεν τῷ ζῶντι according to the Byzantine text, where the (third) declension is unambiguously dative. But on the other hand Mark 6:23 "ὤμοσεν αὐτῇ" has an unmarked dative of the person to whom an oath (effectively a solemn promise) is made - making the whole hypothesis about the non-pronunciation of final nu -> the use of ἐν questionable. I would like to see more of the larger-picture pattern of ἐν used to mark the dative to get a better idea of it.]
Γελᾷ δ' ὁ μωρός, κἄν τι μὴ γέλοιον ᾖ
(Menander, Γνῶμαι μονόστιχοι 108)
Jonathan Robie
Posts: 4165
Joined: May 5th, 2011, 5:34 pm
Location: Durham, NC
Contact:

Re: Searching for the loss of the dative

Post by Jonathan Robie »

Stephen Hughes wrote:
Jonathan Robie wrote:There are different ways to analyze a text, so you may want to add your own labels - either in the text itself or in a separate file. This is possible too, but it's up to you to do that.
So does this means that catergorisations like Wallace's have NOT been included into the trees but could be user defined in individual installations?
The goal of the syntax trees is to be part of an open data infrastructure.

The trees are just data files that can be queried. They don't have Wallace's categories, but if someone wanted to provide Wallace's categories I would be happy to help them do so, perhaps in a separate file - and I would hope they would also make that public. Similarly, the trees don't have a lexicon, but we have also worked with others to create a lexicon that can be used together with the trees. The trees are based on morphology done by James Tauber and Ulrik Sandborg-Petersen.
Stephen Hughes wrote:In the code, the information that the preposition ἐν takes a dative is not there. Is that left to be implied by an analysis of the whole unit of code from <wg nodeId="400030120040040" class="pp" role="p"> to the end?
What it modifies is the head of the noun phrase contained in the prepositional phrase:

Code: Select all

<w morphId="40003012006" class="noun" head="true" lemma="χείρ" case="dative"
        gender="feminine"
        number="singular">χειρὶ</w>
Stephen Hughes wrote:That seems to come from a "let the text speak for itself" line of reasoning, which in a simple case like the preposition ἐν is going to work, but "letting the text speak for itself" (if I've correctly understood that particular aspect of the rationale behind the code) will produce unintelligent results in the case of attracted relative pronouns. Are verbs (), nouns βασιλεῦς (+gen.) and adjectives ἄξιος (+gen.) marked for what ...
This is a syntax tree, not a lexicon. You could derive the kind of lexicon you are looking for from the syntax tree, assuming the syntax tree is consistent. But for any given sentence, the goal is to describe that syntax as understood in context, not other ways that the same words could be used elsewhere.
Stephen Hughes wrote:Specifically for this discussion, does your trees program have what the case of the relative would be without the preceeding context, so that when people work from a nodeId of the relative clause itself, they get an intelligible result. In relative like Luke 1:77 ὅρκον ὃν ὤμοσεν πρὸς Ἀβραὰμ τὸν πατέρα ἡμῶν of course it doesn't matter.
No, they always interpret the sentence in context, explaining the interpretation they have in that context. That's what they are for.
Stephen Hughes wrote:[As a side note: According to LSJ, what is sworn by could have been in the genitive in Classical Greek, but in the New Testament it is in the dative as in Matthew 5:36 "μήτε ἐν τῇ κεφαλῇ σου ὀμόσῃς", where considering that if the final nu of the oxytone accusative τὴ κεφαλή were not used in pronunciation, it would be pronounced identically with the perispomenon dative τῇ κεφαλῇ that this would (could) have been in the classical idiom, but is in New Testament times the dative comes to be marked with the ἐν. cf. Revelations 10:6 ὤμοσεν τῷ ζῶντι according to the Byzantine text, where the (third) declension is unambiguously dative. But on the other hand Mark 6:23 "ὤμοσεν αὐτῇ" has an unmarked dative of the person to whom an oath (effectively a solemn promise) is made - making the whole hypothesis about the non-pronunciation of final nu -> the use of ἐν questionable. I would like to see more of the larger-picture pattern of ἐν used to mark the dative to get a better idea of it.]
The classics folks are creating syntax trees for classical literature, I see a day where we can query to get very useful information from a much broader corpus. But this is all in its infancy.
ἐξίσταντο δὲ πάντες καὶ διηποροῦντο, ἄλλος πρὸς ἄλλον λέγοντες, τί θέλει τοῦτο εἶναι;
http://jonathanrobie.biblicalhumanities.org/
Post Reply

Return to “Syntax and Grammar”