Searching for a Greek phrase

Page 1 of 1 (14 items)
This post has 13 Replies | 1 Follower

Posts 14
Johan Thom | Forum Activity | Posted: Sun, Feb 14 2010 8:53 AM

How does one search for Greek phrases like eis + pistin in Logos 4 (or where can I find this information)? I watched Michael Heisers video on Faster Greek searching, but that only shows how to search for a single morpheme, not for a combination of morphemes.

Johan

Posts 5337
Kevin Becker | Forum Activity | Replied: Sun, Feb 14 2010 9:52 AM

I recommend reading this first http://wiki.logos.com/Search_HELP it will give you some general information about how to use Logos 4's search engine. This search: lemma:εἰς BEFORE 3 words lemma:πίστις in a morph search would probably give you the hits you want (you could expand the proximity or even delete it to catch any long prepositional phrases. OR, you could construct a syntax search, for that see: http://wiki.logos.com/A_Strategy_for_Syntax_Search

Posts 1260
LogosEmployee
Rick Brannan (Faithlife) | Forum Activity | Replied: Mon, Feb 15 2010 8:49 PM

Hi Johan.

A few questions might help narrow down the options.

First, are you searching just the Greek New Testament, or do you want to search other corpora (LXX, Apostolic Fathers, Pseupigrapha, Philo, Josephus)? And the second question would be how precise do you want to be?

To get a quick and rough idea in either just the Greek NT or in all the Greek morph texts available with the Logos morphology, Kevin's search (using the 'morph' search) would be a good start: lemma:εἰς BEFORE 5 words lemma:πίστις. After you type "lemma:" you can type a transliteration, the drop-down will pre-fill with possible options. So "lemma:pisti" will give you πιστις, among other things, as a option in the drop-down, you can select that. This, however, may give some false positives (e.g. Acts 24.24).

If you're only searching the Greek NT and really want precision -- that is, you want to search for where εις has πιστις as its prepositional object -- then you probably want a syntax search using either the Cascadia syntax graphs or the OpenText.org material. I don't have time to do that right now; perhaps tomorrow I'll be able to run a few examples and post a follow-up to this thread.

Rick Brannan
Data Wrangler, Faithlife
My books in print

Posts 14
Johan Thom | Forum Activity | Replied: Mon, Feb 15 2010 10:27 PM

Rick and Kevin, thanks. This helps to get me going. I would like to be able to search both just on the NT and also on the NT + other corpora. How do you specifiy the other corpora?

I will investigate the possibilties offered by syntax search, although they seem to require quite a steep learning curve! It would be very helpful if you could provide some examples for searches involving either morphology or syntax. I find having an example that one can adapt is a better place to start than having to work through a complete set of 'rules' written in non-intuitive terminology.

A related issue: I ran a morphological search such as Kevin suggested, but was then unable to save or copy the results of this search to a note or a Word file. How do you select and copy such material? Trying to to select with the mouse did not work.

Posts 5337
Kevin Becker | Forum Activity | Replied: Tue, Feb 16 2010 4:22 AM

Specifying other corpora is fairly simple. If you want everything, just make certain your morph search looks like this

Those three areas of text can be used to refine your search.

All Passages: can be modified to specific Bible ranges. I imagine it supports the versification in Philo etc but I haven't tried it.

All Resources: This specifics the particular book or collection of ancient language texts you want to search. It also has the handy option of "All open resources" that will let you search multiple books in the absence of a collection.

Logos Morphology: Specifies the morphology scheme. If you've refined the resources to be searched there might not be an option here to change it.

You cannot (at this point) export the results of a search. I believe it's in the works, but it's not here yet. One thing you can do is in the search history is make shortcuts to a particular search (click where I have the red arrow pointing) then click and drag a particular search to either the favorites pane or the shortcuts area.

 

Rick: I would be interested to see how you would do a syntax search for this. The template didn't give me very many results, and I couldn't construct a custom search I was happy with.

 

Posts 1260
LogosEmployee
Rick Brannan (Faithlife) | Forum Activity | Replied: Tue, Feb 16 2010 6:02 AM

JohanThom:

I will investigate the possibilties offered by syntax search, although they seem to require quite a steep learning curve! It would be very helpful if you could provide some examples for searches involving either morphology or syntax. I find having an example that one can adapt is a better place to start than having to work through a complete set of 'rules' written in non-intuitive terminology.

Yes, just understanding the syntax of a passage can be difficult at times (at least for me); thinking about how to qualify that for searching to find other passages even more so. I'll try to get some examples together later today. I believe both OpenText.org and Cascadia may have some templates that could help. Kevin, on the disparity of hits you're seeing, I think εις does not take πιστις as its object that frequently, but other prepositions take πιστις as object much more frequently (κατα, επι, περι, δια, εκ, etc.)

Also, a Bible Word Study on πιστις would fill the preposition chart and you could get a glimpse of the prepositions used with πιστις as object, frequency, etc. in the "Preposition Use" chart.

JohanThom:

A related issue: I ran a morphological search such as Kevin suggested, but was then unable to save or copy the results of this search to a note or a Word file. How do you select and copy such material? Trying to to select with the mouse did not work.

One possible view for the "morph" search is the "analysis" view. After you've done your search, look at the right side under the dialog header. Hopefully you'll see "Verses | Aligned | Analysis". Click "Analysis". This displays the hits in hierarchically sortable columns. When here, if you click on the top left dialog icon (the magnifying glass) one option is "Export". This will allow you to export the search data into Excel or into a standard comma-separated (*.csv) format.

Rick Brannan
Data Wrangler, Faithlife
My books in print

Posts 5337
Kevin Becker | Forum Activity | Replied: Tue, Feb 16 2010 6:28 AM

Rick Brannan:
Kevin, on the disparity of hits you're seeing, I think εις does not take πιστις as its object that frequently, but other prepositions take πιστις as object much more frequently (κατα, επι, περι, δια, εκ, etc.)

I did notice that. The thing that I was mulling over this morning was how to grab πιστις when it was modifying the object of the preposition. For example: Why doesn't this grab Rom 1:5?

Posts 582
Gary O'Neal | Forum Activity | Replied: Tue, Feb 16 2010 7:34 AM

Kevin Becker:
I did notice that. The thing that I was mulling over this morning was how to grab πιστις when it was modifying the object of the preposition. For example: Why doesn't this grab Rom 1:5?

Kevin

My guess is that it's because πίστεως is not the object of the preposition. I would think that your search would have returned it however, since you don't specify that it's to be the object. This is just me thinking out loud -- I'm trying to understand these syntax searches and figure out how to use the tools to return results I can have confidence in.

EDIT:

BTW, I used one of the examples in the search tool to recreate your search. The only thing it did differently was to add "Anything" between your two search terms. This returned two hits - Rom 1:17 and Gal. 3:23.

πάντα εἰς δόξαν θεοῦ ποιεῖτε

Posts 1260
LogosEmployee
Rick Brannan (Faithlife) | Forum Activity | Replied: Tue, Feb 16 2010 10:15 AM

Hi Kevin

Kevin Becker:
I did notice that. The thing that I was mulling over this morning was how to grab πιστις when it was modifying the object of the preposition. For example: Why doesn't this grab Rom 1:5?

My guess is that you're still basically looking for where πιστις is the object. I have a simplified search for prepositional objects I'll post in a bit. On searching for where something modifies the object of a preposition (in this case, πιστις modifies a prepositional object) I came up with the following:

I tried it first without the Terminal Node but ran into an issue with Determiners, so I'm using the terminal node to disallow determiners (articles). The above locates eight results: Ro 1.5; 3.27; 4.13; 16.26; Ga 3.2, 5; Titus 1.4; Heb 10.22. Note I'm not specifying any particular preposition.

EDIT: Yes, the search title is wrong; it should be something like "Cascadia: When πιστις modifies a prepositional object"

Rick Brannan
Data Wrangler, Faithlife
My books in print

Posts 1260
LogosEmployee
Rick Brannan (Faithlife) | Forum Activity | Replied: Tue, Feb 16 2010 10:37 AM

Earlier I wrote:

Rick Brannan:
I have a simplified search for prepositional objects I'll post in a bit.

Here's that search. This locates a prepositional phrase where a given lemma is the object. Note this is significantly different than the "Prepositional Object" template. The template is conceptually easier to grok (understand/internalize) but this one is just one object with two properties:

Recall I said it "locates a prepositional phrase". This works because it finds the whole prepositional phrase object where the head lemma is πιστις. Logically, the head lemma of a prepositional phrase is its object. This finds the whole phrase, so it doesn't highlight tokens (words) in the phrase. 80 instances are located.

If I wanted to highlight the preposition and the object of the preposition within the larger phrase, I'd do something like this:

Here I'm using the same basic prepositional phrase structure, and picking out particular words/objects to highlight. The easiest way to grab and highlight the preposition is to use the Preposition terminal node. To be a bit conceptually easier, I used a word node to highlight the word itself. Note that the word node has "Matching Skips Levels" checked (hence the dotted line) and also uses agreement to specify the word instance is the exact same instance specified as the head of the phrase. This allows the word to be picked out and highlighted among search results.

One could also specify a particular preposition as the head lemma of the Terminal Node, or you could add a Word node with the preposition as lemma of the word node.

The astute among you (and that's everyone, right? Logos users are an incredibly astute bunch) have already noticed that the second search has 77 results while the first has 80 results. Where did the three results go? The find-the-whole-phrase doubles up some matches in larger phrases (Ro 1.17; 3:25-26); other than that I believe they're all the same.

Rick Brannan
Data Wrangler, Faithlife
My books in print

Posts 5337
Kevin Becker | Forum Activity | Replied: Tue, Feb 16 2010 11:22 AM

Rick, thank you for taking the time to trace this out. Your posts are always very helpful.

So, it appears that matching skips levels is not behaving in the way that I would have anticipated. My expectation for my search (see above) would procure wherever the lemma πιστις showed up within the broad boundaries of the prepositional phrase (red box below) regardless of its actual function in the phrase. It seems to me (correct me if I'm wrong) that the intervening nominal phrase (circled below) interrupts the search engine from finding πιστεως below it. Does the search engine only look at the head term of the nomial phrase?

If my interpretation of the facts is correct then it would be helpful to have some description of how matching skipping levels work so that we don't assume matching would extend beyond what it actually does.

Thanks again Rick.

Posts 1260
LogosEmployee
Rick Brannan (Faithlife) | Forum Activity | Replied: Tue, Feb 16 2010 11:38 AM

(Responding to a later query, but using this post since it is the most appropriate one)

Kevin Becker:

The thing that I was mulling over this morning was how to grab πιστις when it was modifying the object of the preposition. For example: Why doesn't this grab Rom 1:5?

Hi Kevin.

I think you need an "anything" between the terminal node and the word. This is looking for where the terminal node slot is directly followed (whatever the intervening levels may be) by a word with the lemma πιστις. The following probably does it. Note I've specified disagreement between the word and the prep phrase instance so that I don't find where πιστις is the prepositional object. Some spurious hits may still crop up in the results, though. This gets 16 hits, though some may be doubled (e.g. Ro 1.1-6 and Ro 1.5 hit with πιστις; the query is true for both of them but you're not quite expecting the first one).

Hope it helps!

 

Rick Brannan
Data Wrangler, Faithlife
My books in print

Posts 838
Dewayne Davis | Forum Activity | Replied: Tue, Feb 16 2010 11:49 AM

JohanThom:

How does one search for Greek phrases like eis + pistin in Logos 4 (or where can I find this information)? I watched Michael Heisers video on Faster Greek searching, but that only shows how to search for a single morpheme, not for a combination of morphemes.

Johan

Have you considered doing a word study for the word PISTIS and going to the Grammatical Relationships section and selecting "with preposition", under which is EIS...? Or going to the preposition graph toward the bottom as well?

“... every day in which I do not penetrate more deeply into the knowledge of God’s Word in Holy Scripture is a lost day for me. I can only move forward with certainty upon the firm ground of the Word of God.”

Posts 5337
Kevin Becker | Forum Activity | Replied: Tue, Feb 16 2010 1:07 PM

Thanks Rick, sometime it's so easy to miss small stuff like that!

Page 1 of 1 (14 items) | RSS