9 replies to this topic

[Harvey]

1. While hovering over an item in the Hebrew syntax pane highlights the corresponding words in the text pane, the reverse is not happening (as I understand it should). Is there any setting that I am missing?

2. Is there a way to specify in a search that an item must be at the beginning of a phrase or clause? I am trying to do a search for overt relative clauses consisting of a predicative prepositional phrase. I have tried anticipating and excluding various items, null and overt, that might precede the prepositional phrase in the relative clause (including subject and verb), but still get extraneous results. After a while it occurred to me that I can do a simple morphological search for אשר followed by a preposition; but since this is the type of problem that may not have such an easy solution in other contexts, is there a way to specify the search syntactically?

3. Is there a way to conduct a search within the results of a prior search, which it seems to me would often be an efficient alternative to building a more complicated search that accomplishes everything in one step?

Edited by Harvey, 27 June 2012 - 08:37 AM.

[Robert Holmstedt]

Harvey,

2. Yes, use the PLACE function and insert it in the first column of the phrase or clause. Since the אשׁר is the first item in the subordinate clause you're searching for, you'd want to use PLACE=2 to ensure that the next constituent is a preposition. But this is all missing the underlying structure of what you're looking for. If you mean by "predicative prepositional phrase" a clause in which a PP is the Complement of a null copula (or, in more common though less accurate terms, a verbless clause with a PP as the predicate), then the place position of the PP isn't the issue. Eventually (when the complex searching is further ironed out -- we're getting there!), you will need to build the search to mirror the clause structure: clause [dependent] = אשׁר (first column), Predicate Phrase (second column); within the Pred Phrase, Predicate = Null (first column), Complement = Particle [Preposition] (second column).

For now, there is a simpler search that also highlights the flexibility of Accordance syntax searching by mixing morph and syntax in a flat structure, like so:

 Screen Shot 2012-06-27 at 9.58.33 AM.png   13.01K   15 downloads

I checked these examples and there are a few dozen you'll need to sort out, but you can isolate these by adding an extra search element for a quicker scan:

 Screen Shot 2012-06-27 at 9.58.24 AM.png   16.91K   18 downloads

Does that get you what you want?

[Robert Holmstedt]

I should add that you can also include in the flat structure search a null Subject, in case you want to specify that the Subject is not overt. Alternatively, you include an overt Subject (which returns results like Gen 30.35 אֲשֶׁר־לָבָן בּוֹ.

 Screen Shot 2012-06-27 at 10.20.55 AM.png   13.76K   3 downloads

[Helen Brown]

The under the cursor highlighting does not apply to the Syntax tree pane, it just cross highlights to the text pane.

Currently you cannot directly search the contents of a tab, but you can open another tab and search it with [CONTENTS first tab] <AND> my new query.

I hope this helps to answer the other questions.

[Robert Holmstedt]

Since this question provoked us to some search testing, I will add two additional screen shots: both use the basic "flat" structure, but do add some depth (i.e., they are moving towards the clause structure search) and also add the Subject (which can be overt or made null by adding the NULL item). This is beginning to look just like what the clause search for this issue will eventually look like when the kinks are ironed out.

 Screen Shot 2012-06-29 at 11.04.11 AM.png   23.17K   15 downloads
 Screen Shot 2012-06-29 at 11.08.55 AM.png   23.66K   18 downloads

[Harvey]

(My first question resolved itself.)

My thanks to Robert and Helen for your responses. (I did not find them until just now, because I was under the evidently mistaken impression that I would receive notice of responses to my question. Is there a setting that I missed for this?) Once I have digested the responses, I'll post if I have further questions.

Harvey

[Helen Brown]

You can click Follow this Topic in the top right of the window, to be notified. There are other settings you can find in your member profile.

[Harvey]

Robert,

Do I understand from your responses that there are what would seem to be well-formed "complex" searches that the syntax module cannot yet perform? If so, is there any available description of what those constraints are? I have been experiencing a good deal of frustration running searches, involving a few levels of nesting, that come back with no results, even though they seem to perfectly match syntactic structures that I find in the syntax module; I wonder whether I am running up against those limits.

Thanks.

[Robert Holmstedt]

The syntax searching in the Hebrew construct window continues to improve, but does still face some complex programming obstacles (so I'm told). Frankly, it's way too complicated for me -- I see the syntax (indeed, sometimes I dream trees) and think like you do, that if I can see a structure, it should be searchable. And it eventually will, thanks entirely too the brilliance of the programmer working on it. But it will take a bit more time.

In the meantime, those using the Hebrew syntax module can look forward to more texts, as we continue to tag. The full Pentateuch is not far off!

[Helen Brown]

We'll be releasing 9.6.3 very soon, and I know that Syntax searching has been further refined in this rev.