Last week, I posted an article about “Word Groups” in the OpenText.org Syntactic Annotation. I promised some follow-up; and now it’s time for that.
There are obvious uses for this level of annotation in the realm of searching, but what about in just reading the text? Or in working through a passage exegetically?
The good news is that the visualization (graph) supports most operations you’re used to performing from a standard morphologically tagged Greek NT in Logos Bible Software. This article is about some of those options.
First, the word group annotation can help in breaking the text down into units of meaning. Let’s recall some basic word group terminology:
- Word Group
- A word group consists of a single head term and any and all of its modifiers, though it will frequently consist of just a single word.
- Head Term
- Informally, a word that does not depend/modify any other word in its group.
- Any word contained in a word group that is not a head term is considered to be a modifier. These modify the head term either directly or by modifying words that modify the head term. Modifier groups may be nested infinitely (Ro 1.1-6 is an example of a long and heavily nested series of modifiers).
Basically, a word group consists of, at minimum, a head term. If there are words that modify the head term, then those modifiers are included in the word group. If there are no words that modify a head term, then the word group only consists of the head term.
- Word Group 1, ????? .. ?????????: This is a two-word word group. Note first of all that word groups need not be contiguous. Here, ????? modifies ????????? as a definer. The definer here is an adjective, and it further defines the head term. The head term is translated condemnation. The condemnation is defined further as none or nothing or no.
- Word Group 2, ???: This is a one-word word group, and it is simply the particle ???. It can be translated as then or, as in many English translations of this verse, therefore.
- Word Group 3, ???. This also is a one-word word group, an adverb typically translated as now.
- Word Group 4, ???? ?? ?????? ?????: The last four words combine to form a word group with nested modification. The head term is the article ????, a plural article in the dative functioning as a relative pronoun, those. It is modified by the relator ?? ?????? ?????. Recall again that a relator is a prepositional phrase acting to modify a head term (or modifying a further nested modifier). The relator consists of the term ?????? (Christ), which is modified in two different ways. First by the specifier (preposition) ??; secondly by the definer ?????.
So, without performing any searches but just using the graph data, one can walk through the text and gain a clearer picture of the relationships that exist (according to the editors of the OpenText.org annotation) within the Greek text. You can see what the minimal units of meaning are (word groups), you can start to develop ideas about the relationships posited between these words. And the words in the column are in the order of the Greek New Testament, so if you read the column, you are reading the text.
This brings up an important point about how to read using the graph visualization. There are two ways to read the graph:
- Read the structure from left to right
- Read the text from top to bottom
Stop and let that sink in. Look at the word group image again, and read the structure of the second word group: It is a head term modified by a relator. The relator is modified by a specifier. The relator is also modified by a definer. Then read the text itself, from top to bottom: ???? ?? ?????? ?????.
As you’re reading the text, you may have need to view a lexicon article or expand the morphology. You can do that with just a click or two:
The goal is for examination of syntax to begin to play a larger role in exegesis and study. This is done by directing the user to these resources in contexts where they may be handy to consult (e.g., from Exegetical Guide) and in allowing access to the same sorts of functionality and tools (e.g. keylinking, lemma report) that are available in the morphologically annotated resources one is accustomed to using.
And once you start to consider syntactic information, you start to ask yourself questions that you never would have asked before. For instance, in looking at this passage, I thought, “what other passages have an article (as a head term, functioning as a relative pronoun) modified by a prepositional phrase like this?” Specifically, I was wondering what sorts of other things might have to do with “those” who are “in” something, like this passage talks about “those who are in Christ Jesus”.
So, I decided to do some searching. I specified my search (don’t quite have the dialog ready, so I don’t have a screen capture of that as I keyed in the raw query search language) which ended up looking for:
- a word group containing:
- a head term that has part-of-speech of article and number of plural
- a modifier that is in the category of relator (a prepositional phrase).
- the modifier contains the word ??
Across the New Testament, I located 31 instances. Please note these haven’t been verified, and we’re still testing the engine. But I had results for the whole New Testament displayed on my screen within six seconds of pushing the search button. The graphic on the right is a screen capture of the search results. The big blank space is where we’ll have some sort of visual representation of the query, but we haven’t quite figured that out yet.
The search result display is worth noting. See the grey background in the middle column? That is the extent of the specified query, in this case, the specified word group. Sometimes the query matches the whole word group; other times it matches a portion of the word group. The red text? That is text that matches items specified to be “highlighted” in the query result.
Here, I specified that the head term (the plural article) and the relator (the prepositional phrase that modifies the article) are highlighted. This is customizeable for each query; you just mark a structure as “highlighted” in the dialog (we’ll get into the dialog in subsequent posts) and it shows up as highlighted within the structure searched.
Note also the second hit reported. The word group is discontiguous; broken up by a conjunction. But the search (and the extent of hit) really doesn’t care. It finds what you ask for because you’re searching for structure, you aren’t approximating structure with parts of speech within N words of some other part of speech. And the extent of the word group is also known. So further searches could be done based on this search. For example, perhaps I want to know how many of the prepositional phrases here have an adjective in them (e.g. the second hit for Luke 21.21; “those who are in the middle of the city”).
There are some refinements that can be made in the search. Jn 5.28 occurs twice in the search results due to a word group embedded in a word group; we could further restrict the search to remove embedded word groups from the hits … or perhaps only search for these hits within embedded groups.
The cool thing to note is that this search was based on the word group structure and also utilized morphological information. The search merges word level stuff and word-group level stuff into one query. This doesn’t replace morphology and all you’ve learned about how to search with morphology; it adds another layer allowing you to use morphology, lexical forms and now higher-level tagging to be more precise and flexible in how you specify and search the Greek New Testament.
We’ve talked about word groups for a few posts now. There’s more that could be said, but I think it is time to move to the clause annotation. We’ll go there next; perhaps later this week or early next week.