What about the graphical query?

Page 2 of 4 (69 items) < Previous 1 2 3 4 Next >
This post has 68 Replies | 5 Followers

Posts 687
Jon | Forum Activity | Replied: Tue, Nov 3 2009 9:57 PM

That is odd... it has disappeared from the packages list. I still have a Gramcord septuagint which I had already, but not NA27 with GRAMCORD, so if you had it already you won't lose it, but if you didn't have it you'll no longer get it.... Unless this is an error? :)

Posts 352
Mike & Rachel Aubrey | Forum Activity | Replied: Tue, Nov 3 2009 10:14 PM

David B Phillips:
I am most used to Opentext.org, but there is a new database in 4.0, Cascadia. Mike Aubrey (who is getting his PhD in all this stuff, or something like that. He's really smart and syntax savvy) is pretty excited about the Cascadia resource (http://evepheso.wordpress.com/2009/11/02/logos-4-is-finally-here/). I haven't used it much yet.

You're making me blush. I'm actually currently an MA student in linguistics right now (and have been for too long now...). I had considered aiming for beginning a PhD program fall 2010, but that's been put on hold for now -- though Greek syntax & syntax databases like these are very much the stuff that I'm specifically dealing with in my thesis.

@Samuel:

The strength of Lexham is more in it's Expansion and Annotations than it is in it's graphs with have very little structure. Generally, it's very, very Wallace-esque (which you may or may not think is a bad thing). The lack of structure in the graphs themselves can make it difficult to find a given syntactic construction.

Opentext.org has it's strengths & weakeness too. It's strength is probably it's simplicity. It uses only a few labels for everything. This is also a weakness if you want to find something more detailed - i.e. it makes no distinction between direct and indirect objects; they're all complements. And in terms of the structure of Greek, it often feels as if the labels aren't necessarily related to how Greek works in and of itself: e.g. the adjectival relations such as relator, specifier, qualifier, etc. have not semantic (meaningful) basis in the language itself, which makes it kind of arbitrary. Opentext is also rather inconsistent at times in it's annotations and some of it's inconsistencies are rather silly.

Cascadia is awesome. It's a much more precise database and it has a closer relationship to how Greek actually works as a language than the others. You can ask more specific questions of the Greek text than you could with Opentext, and yes, as Rick said, it's much more accessible in terms of terminology. What's also very cool is the fact that morphological data isn't only stored at the word level like it is in Opentext, which means that you can enter morphological preferences for entire phrases or even Grammatical relations (Subject, Object, etc.). That makes doing searches that require morphology significantly easier and faster to make than before. And like Rick, it has also become my first choice syntax database.

As for Hebrew, you'll want to look at the Andersen-Forbes Phrase Marker analysis. It's complex, but also very cool. And there's another syntax database for the OT in SESB collection, which I don't have...

Posts 8133
LogosEmployee

JeremyEllis:

Is GRAMCORD gone from Logos 4 too?

None of the "LE" collections (Logos 4 products) include the GRAMCORD™ morphology. (If you are upgrading from a LDLS3 collection, you will retain it, but it's not included if you buy Logos 4 new.)

Posts 707
Russ Quinn | Forum Activity | Replied: Tue, Nov 17 2009 7:07 AM

I am very excited about the potential with all the development with the syntactical databases.

However, it is a concern that search capabilities on good old fashioned morphologically tagged databases have been curtailed.

I'm not sure that these resources are a sufficient substitute for the functionality that the graphical query provided for searching morphologically tagged texts.

The biggest problem, in my view, is that the new databases are limited to the text of the New Testament.

What about the need to do complex searches on other texts like the LXX, the Apostolic Fathers, Josephus, Philo, etc.?

In these cases, the new tools don't help as much and the missing functionality that was only available through the graphical query is especially missed.

I understand that the number of users that would have this concern might be small but it seems that Logos has historically been concerned for the needs of those who are working with more complex issues in the original languages (as evidenced by the emphasis on developing the syntactical resources).

My friendly challenge to the thinking behind depreciating the ability to easily do the most complex searches on any morphologically tagged text and advocating searching syntactical databases that are limited to the corpus of the NT is that this cedes important functionality to competitive products and reduces the potential of Logos 4 being the only Bible study tool necessary for advanced work in the original languages.

I totally understand putting a low priority on the functionality of the graphical query in relation to other functionality that the broader user base prefers. I do hope, however, that our kind developers at Logos might reconsider the ultimate fate of this helpful tool.

Please accept this suggestion knowing it comes from a very supportive user who is extremely grateful to this company for the value they have added to our ability to study the Bible.

Posts 418
davidphillips | Forum Activity | Replied: Tue, Nov 17 2009 7:32 AM

Russ,

As a big advocate of graphical queries, I miss their power. However, morphological searches (and complex morphological searches at that) can still be performed on every tagged database that exists - LXX, BHS, Philo, Josephus, Apostolic Fathers, etc. etc. Syntactical searches are indeed only available in the OT and NT, but that's understandable. It would be quite a feat to syntactically analyze everything. That said, I still have not (and I believe this is consistent with others) been able to replicate the power of graphical queries using just the morphological queries, and I realize this is your concern. I would love to see a graphical query re-implemented in 4.0, though at the moment that seems unlikely.

Posts 5337
Kevin Becker | Forum Activity | Replied: Tue, Nov 17 2009 7:49 AM

I too want the Graphical Query in L4. I used it but I did not have my computers reporting statistics. So my use was not included in the statistics they consulted (not that one user would make all that big of a difference, but I doubt I was the only one).

I wonder, if it would be easier to take the Graphical search query tool in L3 and add an export function that would produce a text string that would let L4 give us the search results we want. Or port it to a stand-alone widget... just a thought.

Perhaps if when people with search questions come to the forum we could include graphical search queries that would work too and get L4 only people to realize they want what they don't have. Perhaps a grassroots effort for graphical queries Smile

Posts 352
Mike & Rachel Aubrey | Forum Activity | Replied: Tue, Nov 17 2009 9:54 AM

As I understand it, syntax searching isn't necessarily meant to replace the graphical query. Specifically, Logos believes (this is from the beta forum) that power of the new morphology search is enough to replace the graphical query editor.

With that said, it wouldn't to overwhelmingly difficult to create a database of other Greek texts as you might think. The new Cascadia database wasn't annotated verse by verse, but created with a computational grammar, which means that this grammar could be just as easily applied to other Greek texts. There would still be the need for some to work through the results of the grammar's work, but it would be far less effort than actually annotating things by hand. Now all of this does have the caveat that I honestly don't know what Logos' plans are for other texts; I simply know the power and requirements of computational grammars, because I've done a good chunk of work on syntax databases for my own purposes.

Posts 25282
Forum MVP
Dave Hooton | Forum Activity | Replied: Tue, Nov 17 2009 2:38 PM

Mike, Kevin, David & others

I gladly join in support of re-instating the Graphical Query, for all the reasons you have mentioned but specifically for its ability to compare two terms (ie Agree) at a complex morphological level.

There was a request for a fairly simple morphological query at http://community.logos.com/forums/t/4835.aspx which I could not produce in 4, nor could I produce a Syntax solution with sufficient confidence in its completeness. The fact is that the L4 Morph search is significantly lacking the power of L3 Morph Search, see http://community.logos.com/forums/p/4896/38584.aspx#38584 for my summary. It also re-inforced my support of the need for Verse List as this greatly helps in comparing the results of two searches.

Syntax Search still presents a formidably steep learning curve. In particular , I'm still struggling to get a reasonable solution in Cascadia for the Granville Sharp Rule. This is much simpler in OpenText, but another query proved to be simpler in Cascadia. The "problem" with Cascadia is that its grammatical structure forces one to thing about the validity of other solutions, in particular just how does the article (determiner) act in relation to a nominal phrase (np) eg. compare Eph 5.20 "God and Father" with 2 Pe 1:2 "God and Jesus".

Dave
===

Windows 10 & Android 8

Posts 418
davidphillips | Forum Activity | Replied: Tue, Nov 17 2009 4:06 PM

Dave Hooton:
The "problem" with Cascadia is that its grammatical structure forces one to thing about the validity of other solutions, in particular just how does the article (determiner) act in relation to a nominal phrase (np) eg. compare Eph 5.20 "God and Father" with 2 Pe 1:2 "God and Jesus".

Agreed. This is a problem in opentext.org as well and I would guess in Andersen-Forbes. It is the biggest drawback to syntax searching (and I really like syntax searching).

 

EDIT- Mike, a side question for you, as you're the most familiar with the Cascadia graphs. What do the asterisks imply in the Cascadia graphs? I looked through the glossary, but they weren't mentioned. Thanks!

Posts 418
davidphillips | Forum Activity | Replied: Tue, Nov 17 2009 4:11 PM

Mike Aubrey:
With that said, it wouldn't to overwhelmingly difficult to create a database of other Greek texts as you might think. The new Cascadia database wasn't annotated verse by verse, but created with a computational grammar, which means that this grammar could be just as easily applied to other Greek texts. There would still be the need for some to work through the results of the grammar's work, but it would be far less effort than actually annotating things by hand. Now all of this does have the caveat that I honestly don't know what Logos' plans are for other texts; I simply know the power and requirements of computational grammars, because I've done a good chunk of work on syntax databases for my own purposes.

That is fascinating. I had no idea!

Mike Aubrey:
As I understand it, syntax searching isn't necessarily meant to replace the graphical query. Specifically, Logos believes (this is from the beta forum) that power of the new morphology search is enough to replace the graphical query editor.

Yes. And I've gotten more used to using the morphological searches, and they are much better than the morphological searches in 3.0. I have become a fan of them Smile. The biggest drawbacks here are setting agreement (as Dave mentioned) as well as being forced to write long strings of expressions using parentheses. The best part of the graphical query was that you could drag things around. It was much easier to develop the relationships between items because you didn't have to think through how to order the search. You just dragged arrows between items.

Posts 25282
Forum MVP
Dave Hooton | Forum Activity | Replied: Tue, Nov 17 2009 4:53 PM

David B Phillips:
The biggest drawbacks here are setting agreement (as Dave mentioned) as well as being forced to write long strings of expressions using parentheses.

I don't see why we should when L3 has the power with properly implemented term lists and term negation to produce a compact query. L3 also has text query constraints similar to what you can do in Graphical Query eg. Agreement!  The Morph Search text entry box is not suited to typing anything more than a  lemma:A ANDEQUALS @N expression (cursor positioning is difficult + lemma suggestions and morph boxes keep getting in the way).

Dave
===

Windows 10 & Android 8

Posts 418
davidphillips | Forum Activity | Replied: Tue, Nov 17 2009 5:18 PM

Dave,

I didn't realize that was possible in L3! I never used text queries much because of the graphical query tool.

Posts 352
Mike & Rachel Aubrey | Forum Activity | Replied: Tue, Nov 17 2009 11:27 PM

David B Phillips:
Mike, a side question for you, as you're the most familiar with the Cascadia graphs. What do the asterisks imply in the Cascadia graphs? I looked through the glossary, but they weren't mentioned. Thanks!

They marks the heads of phrases. You can see it pretty clearly here in Matthew :

As you can see, the Verb is the head of the clause, and in the Subject noun phrase, the noun υἱὸς is the only one with asterisks all the way up to the highest NP level. Likewise, the Genitive NP has it's own head, but isn't a head any higher than that because it is dependent upon the ὁ υἱὸς.

With that said, I take issue with their head analysis of the prepositional phrase at the end of this and the gloss given for ὐπ', which appears to be following the Lexham Greek-English Interlinear. It should be: "in this way also the Son of Man is going to suffer by them." Syntactically, ὐπ' (by) should be treated as the head of its phrase because prepositions function partially like verbs in that they require an Object. Besides, that's why its called a prepositional phrase: Verbs are the heads of VPs, Nouns are the heads of NPs, and Prepositions are the heads of PPs.

Anyway, that's a side issue. The first part is the answer to your question.

Posts 707
Russ Quinn | Forum Activity | Replied: Fri, Nov 20 2009 9:27 AM

The discussion from this thread http://community.logos.com/forums/p/5280/41725.aspx has increased my concern that usage statistics from L3 might not be the best basis for making a decision concerning the future of advanced morphological queries.

I wonder how much the usage stats reflect the fact that people have found competitive products to be more effective for advanced morphological queries rather than reflecting the desire for Logos to improve its capabilities in this area.

It seems odd that Logos is so aggressively pushing the envelope when it comes to syntactical searching (which I love and appreciate) but backing up a bit when it comes to morphological searching.

I have a much larger investment in Logos than any competitive product and would love to see Logos be my only Bible software solution. I would prefer to spend my money on additional resources than have to buy Bible texts in multiple platforms.

I think improving the graphical query might bring more return on investment than the usage statistics imply.

Posts 4508
Robert Pavich | Forum Activity | Replied: Fri, Nov 20 2009 9:53 AM

Russ,

I share your concern.

there is a very small gap that I believe that Logos should fill...the gap left by the capabilities of the Graphical Query.

If there are things that it could do....that now our current tools cannot but another software competitor can do...then that's not a good thing.

Either modify the current tools to give that functionality, or bring back the Graphical Query...

Logos' strength is cutting edge searching...let's keep it that way...

Logos?

Robert Pavich

For help go to the Wiki: http://wiki.logos.com/Table_of_Contents__

Posts 5337
Kevin Becker | Forum Activity | Replied: Fri, Nov 20 2009 10:00 AM

Robert Pavich:

Russ,

I share your concern.

there is a very small gap that I believe that Logos should fill...the gap left by the capabilities of the Graphical Query.

If there are things that it could do....that now our current tools cannot but another software competitor can do...then that's not a good thing.

Either modify the current tools to give that functionality, or bring back the Graphical Query...

Logos' strength is cutting edge searching...let's keep it that way...

Logos?

Hear, hear!

Posts 418
davidphillips | Forum Activity | Replied: Fri, Nov 20 2009 10:23 AM

Kevin Becker:

Robert Pavich:

Russ,

I share your concern.

there is a very small gap that I believe that Logos should fill...the gap left by the capabilities of the Graphical Query.

If there are things that it could do....that now our current tools cannot but another software competitor can do...then that's not a good thing.

Either modify the current tools to give that functionality, or bring back the Graphical Query...

Logos' strength is cutting edge searching...let's keep it that way...

Logos?

Hear, hear!

 

Agreed as well. It would be great to have this back.

Posts 418
davidphillips | Forum Activity | Replied: Fri, Nov 20 2009 10:26 AM

Mike Aubrey:

They marks the heads of phrases. You can see it pretty clearly here in Matthew :

Yes. That is very helpful. Thanks! I'm starting to get used to looking at Cascadia after using opentext.org for so long. I like that the taggings actually sound like Greek grammar Geeked.

Posts 1692
LogosEmployee
Bob Pritchett | Forum Activity | Replied: Fri, Nov 20 2009 3:33 PM

At ETS we got to talk to a graphical query user in person, who made a compelling case for some improvements in our non-syntax morph searching. We'll work on improving this soon, though I can't yet say if it'll be the old graphical query or something newer (and better!).

Thanks for the feedback on this!

Posts 4508
Robert Pavich | Forum Activity | Replied: Fri, Nov 20 2009 3:48 PM

Bob,

that's good to hear....I'm sure it will be outstanding!

Robert Pavich

For help go to the Wiki: http://wiki.logos.com/Table_of_Contents__

Page 2 of 4 (69 items) < Previous 1 2 3 4 Next > | RSS