This is what I’ve been working on for the past week or so. I now have a basic draft complete, but it still has a lot of work left. I need to go through it checking for errors and adding missing annotations. If you want to give any feedback, feel free. I’ll take it and consider it – this is only a draft.
And again, if this shows anything, it shows that Carl’s point about the flow of these verses being terrible convoluted.
A Note on Labels…
Phrase & Clause Types:
I’d like to think that labels like NP, AdjP, AdvP, and PP are pretty obvious, so below are only the less-than-obvious ones.
Cl – This is a clause. I’m a little frustrated by it since sometimes these Cl’s were annotated as “IP,” others “CP,” and one or two more as S. But the program won’t display that annotation in the tree for some strange reason. Rule of thumb on determining which it is: If its a CP, then its immediate daughter is going to be an IP. If the Cl is an IP, then the immediate daughter is going to be an S. The Cl will be an “S” if it is a nominal and thus verbless clause. See below for the meaning of these labels.
IP – Short hand for a clause and literally stands for “Inflection Phrase.” I use it on the assumption that the Greek phrase is headed by the Verb, which fills the “I” position.” This stands in contrast to Opentext.org which has a completely flat structure, not recognizing any constituent as more important than another. Focused (hopefully consistently labeled “New” and “Prominent”) constituents can appear directly before the I in the IP.
CP – This is another clause type marking Complement clauses and it stands for “Complement Phrase.” These clauses will mark relative clauses and a number of different subordinate clause types such as ὄτι and ἴνα clauses as well clauses which begin with toplicalized material. Relative clauses always begin with a toplicalized relative pronoun. Toplicalized constituents are marked with “Prominent” but not “new” – more on that later.
S – This is phrase/clause that does not have a lexical head. It appears in two positions in the above diagram. It is the first “Cl” since there is not verb in the matrix clause of this convoluted sentence. “S” also appears as the daughter of the “IP” and sister of the “I.” Since “S” is an non-headed phrase, it allows for the “free” constituent order that we regularly see in Greek – though as Steve has pointed out to me, there are pragmatic reasons for the ordering of constituents here. Regardless, there is no clear lexical head in these post-verbal phrases. I’m working on a better way of marking these constituents by means of discourse/information functions rather than phrase structure, I think.
DP – This is a Determiner Phrase. It is a functional category in that it provides functional/grammatical information to the phrase it contains, whether it be an noun, adjective, or clause. Positing its existence is logical since Greek phrases tend to have a phrase initial head in their most unmarked order.
Different phrases are annotated with a few different grammatical functions/roles. Most of them, like Subj and Obj are self-explanatory. Others are not and are explained below.
Obj2 – This is the secondary or “indirect” object. By annotating them as Obj and Obj2, I allow for a more concise description. Were I do use “DObj” and “InObj,” I would also have to include “PrepObj” for the objects of prepositions. This way, I can get away with just two Object labels. There are other reasons too, but those are more complicated.
Adjnct – stands for Adjunct. This function marks extra modifiers in phrases as well as in clauses that are not necessary for the word they modify. Adjuncts mark adverbial meaning.
Oblq – stands for “Oblique” and marks phrases that are necessary for a particular verb, but are not “core” arguments like Subject and Object (2). For example, in the sentence, “I went to the store.” the prepositional phrase “to the store” is a phrase required by the verb. When you go, you have to go somewhere. The difference between Obliques and other arguments such as Subject or Object (2) is that Obliques can often be understood implicitly. Thus, when a small child tells his mother, “I have to go.” His mother knows where he has to go even though its not stated.
Comp – stands for “Complement” and is only used for Genitive Noun Phrases, which seem to have a different distribution in Greek than other types of Nominal Modifiers.
Predlink – is basically an unspecified argument in non-verbal predicate clauses. It denotes two things: 1) It marks the Predicate of the clause (hence “Pred”). 2) It denotes that the predicate is linked to the Subject, predicating something about the subject – i.e. locative, existential, attributive, etc.
I had a lot of trouble trying to figure out how to represent these in the tree and I’m not yet entirely satisfied with what I’ve done. Perhaps when I read Lambrecht, I’ll change my mind again. But for now, I’ve got the following scheme, which I think (at present) does a reasonably good job concisely describing a variety of functions.
|+New||contrastive / emphatic focus||completive / presentational focus|
|-New||(shifted) topic, link||continuing topic, tail|
I’ve used these in the tree above only sparsely, because this is not my area of expertise by any means. Also, the program I’m using for building the trees prevents me from using +/- symbols, so of what I have annotated, if something is marked both “Prominent” and “New,” then it should be read as “+Prominent, +New.” If something is marked with only “Prominent” or only “New,” then it should be understood that the missing annotation is “-New” or “-Prominent.” If both a missing from a given constituent, that means one of two things. Either I haven’t done anything (the majority of cases) or the information is is both “-Prominent, –New” (some of the cases). This scheme is basically taken from Hye-Won Choi, “Phrase Structure, Information Structure, and Resolution Mismatch” in Formal and Empirical Issues in Optimality Theoretic Syntax (Peter Sells ed.; Stanford, Calif.: CSLI Publications, 2001), 17-62. At the very least, I know that Choi has read Lambrecht.
Now I’ve only labeled a few phrases with Semantic roles thus far. They will likely change as this draft is improved upon and I will continue to add others to the rest of the phrases. I think all of the ones I’ve used are pretty self-explanatory, so I’ll leave it at that.