Monday, February 25, 2008

Greek Syntax Search using Logos: Article & Prepostional Phrase

Rick Brannan just uploaded an interesting post on the Logos blog in response to the following question posted on the Logos newsgroup:

Someone has commented that there are 484 occurrences of the definite article occurring without a noun introducing a prepositional phrase, such as, "τα επι τοις ουρανοις." I wonder if someone would teach me how to search my GNT (N/A27) to confirm this statement?
(I wanted to reply to Rick on his blog, but I need to include pics, so I will do it here.)
I think the search requested is not quite what Rick ends up with. Rick correctly finds all the instances of articular prepostional phrases.
Rick's results, therefore, include Matthew 5:16: τὸν πατέρα ὑμῶν τὸν ἐν τοῖς οὐρανοῖς = "your father in the heavens" where the prepositional phrase modifies "father." The request, however, is for "occurrences of the definite article occurring without a noun introducing a prepositional phrase..." That is, if I understand it correctly, the person wants instances where the article with prepositional phrase is functioning as a substantive and not when it is functioning as an adjectival phrase.
So, to eliminate those, here is the Syntax Search I used in Logos:
First, as I looked at some of the anticipated phrases in OpenText, I noted that the article was always the Head Term. On my search, therefore, note that I define Word 1 as being an article using the morphology selector, but I also checked the box indicating that it "Must be an immediate child of the parent." This choice makes sure that it doesn't turn out to be functioning as an adjective.
Second, rather than trying to define the function of the preposition as Rick does using the Modifier/Relator identification, I simply used the morphology selector again to indicate that it has to be a preposition. (And note that it should not be an immediate child.)
Third, I don't think that I need to worry about word order in this instance. The article will always precede the preposition, and as I long as I don't specify the preposition as an immediate child, it should pick up instance where, e.g., a postpositive particle intervenes. Running the search, I come up with the following:Now, I am a bit suspicious that my 149 results are exactly half of Rick's 298, but you will see that we come up with quite different hits. Actually, however, I only have 147 hits (or less), because I am almost certain that the very first one, Matthew 5:12, reflects a mistake in the OpenText coding as well as Luke 6:42. I say "almost certain," because I am still trying to figure this stuff out. [Those of you who know better, I would appreciate confirmation that the articular prepostional phrases in Matt 5:12 and Luke 6:42 are indeed incorrectly coded and should instead be marked as a relator phrase. BTW, I also ran the search using "modifier=relator" instead of the preposition tag, and I came up with 150 results.]
Bottom line: This is really a powerful way of conducting syntactical searches. I'm not sure that I am confident of my results, however, both because I'm not sure I fully understand the coding and because I think there are errors in the OpenText coding.


  1. I have a feeling that the original number of 484, came from me, in this post:
    Predicative & Attributive Adjectives

    My focus was simply prepositional phrases that were introduced by the article. Though its interesting what different search have come up with.

  2. Aha! Now we know where the 484 comes from. Thanks, Mike. It looks like Rick understood your search correctly in that either substantival or adjectival articular prepositional phrases were being considered. I come up with around 300, so your 484 number does seem high.

  3. Actually, I've found more than that now...and I've documented it

    The Greek Article & Preposition

    I blame it on's tagging. So both your searches are correct. Rick represented my original statement. You answered the newsgroup user's question. But because of Opentext's tagging, my extras are not recognized by Opentext since my search technically wasn't a true syntax search.

  4. (I posted this on your blog too, Mike.)
    Hmmm.... I see now what you were doing, Mike. You weren't really asking about function at all. You are really just doing a string search for
    Article (conjunction) Preposition
    In the example you give of Eph 4:24, however, the OpenText coding does actually make sense to me. The participle is really the key term in this word group, and the article does simply specify it. With this coding, I can look for instances where a noun/participle is specified by an article with an intervening prepositional phrase.
    BTW, I used BibleWorks to search for every instance of
    Article (conjunction)Preposition
    and I came up with 521 hits in 481 verses. It does miss instances (like Matthew 2:16) where there are technically two occurrences in the verse.

  5. (I've posted this at mine too)

    I agree that the participle is the key term - it is what the article is substantizing to function as a noun. But I still don’t think Opentext’s analysis is correct. I simply see how you can split a Noun Phrase like that. The prepositional phrase is functioning in the traditional attributive position of the adjective - i.e. within the bounds of the article as well.