Jump to content


Photo

Syntax labels


  • Please log in to reply
17 replies to this topic

#1 Enoch

Enoch

    Gold

  • Active Members
  • PipPipPipPip
  • 486 posts

Posted 03 October 2010 - 04:37 PM

Is there a special section of this forum board to discuss Syntax?

On looking at this new purchase, I am wondering

1) Why Accordance did not just choose traditional diagramming (instead of trees) to display syntax?
2) Why the letter N was chosen as an abbreviation for "Independent, Non-Speech,"
3) Why the letter L was chosen for "Subordinate, Non-Speech" ???

I don't know if I posted this question at the best place. I searched for Syntax on this board & did not find a special discussion sub-forum for it.

#2 Robert Holmstedt

Robert Holmstedt

    Platinum

  • Accordance
  • 522 posts
  • Gender:Male
  • Accordance Version:10.x

Posted 03 October 2010 - 05:40 PM

Is there a special section of this forum board to discuss Syntax?

On looking at this new purchase, I am wondering

1) Why Accordance did not just choose traditional diagramming (instead of trees) to display syntax?
2) Why the letter N was chosen as an abbreviation for "Independent, Non-Speech,"
3) Why the letter L was chosen for "Subordinate, Non-Speech" ???

I don't know if I posted this question at the best place. I searched for Syntax on this board & did not find a special discussion sub-forum for it.


The syntactic labels do not represent theoretical claims. That is, they reflect the need for one-character labels in the tagging process. "N" (seNtence) was chosen because "S" was already taken for Subject; "L" (cLause) was chosen because "C" was already taken for Complement. It is entirely possible that at some point in the future the single character labels in the trees will be replaced by more typical labels, but that is not sure.

Traditional diagramming is inadequate structurally. The tree diagrams show the hierarchical relationship of the constituents in ways that traditional diagramming does not (that is, if we're talking about the same thing -- the diagramming I remember from grade 7 English).

** correction: since "adequacy" depends on the needs of the user in question, it is more accurate for me to say that the tree diagrams are the natural visual representation of the actual syntactic tagging. This would not have been true for traditional diagramming w.r.t. the syntactic tagging.**

For a brief explanation of some principles and the terms, see my paper from the Users' Conference.

Edited by Robert Holmstedt, 03 October 2010 - 06:48 PM.

Associate Professor, Ancient Hebrew and Northwest Semitic Languages
Dept. of Near and Middle Eastern Civilizations
The University of Toronto
blog: ancienthebrewgrammar.wordpress.com

#3 Helen Brown

Helen Brown

    Mithril

  • Admin
  • 8,209 posts
  • Gender:Female
  • Location:heart in Israel
  • Accordance Version:10.x

Posted 03 October 2010 - 05:52 PM

Enoch: We have indeed created this special section of the forum, so I split your question and the reply into a new topic here.
Helen Brown
OakTree Software

#4 Enoch

Enoch

    Gold

  • Active Members
  • PipPipPipPip
  • 486 posts

Posted 15 October 2010 - 07:42 PM

Thanks a lot for the response. I became acquainted with tree diagrams long ago when I took Linguistics at the U of Minnesota around 1970. In looking at the representations in the Syntax module, the trees do not altogether have face validity to me. But this could partly be because of my prejudice, having loved diagramming since the 8th grade. Perhaps I will post some of my reservations. But, I admit, I could be wrong.

I am thankful for now knowing the mnemonic cause of those abbreviations.

#5 Robert Holmstedt

Robert Holmstedt

    Platinum

  • Accordance
  • 522 posts
  • Gender:Male
  • Accordance Version:10.x

Posted 15 October 2010 - 09:59 PM

Thanks a lot for the response. I became acquainted with tree diagrams long ago when I took Linguistics at the U of Minnesota around 1970. In looking at the representations in the Syntax module, the trees do not altogether have face validity to me. But this could partly be because of my prejudice, having loved diagramming since the 8th grade. Perhaps I will post some of my reservations. But, I admit, I could be wrong.

I am thankful for now knowing the mnemonic cause of those abbreviations.


Trees are merely the visual representation of whatever syntactic analysis is presented. Trees representing a linear, non-hierarchical analysis will thus be quite different than trees representing a non-linear, hierarchical analysis. Ours is linear and hierarchical due to the positions we have taken w.r.t. data and theory, as well as the intended audience of the syntax modules.

Any judgment about the "validity" of the trees, by which one really must mean the validity of the syntactic analysis the trees represent, depends on the linguistic assumptions of the user. Although I personally knew at least one of the syntax profs at Minn in the late 90s (and she was a generative minimalist), I have no idea what they taught at Minn in the early 70s (and whole boatloads of theory has changed since then, as have entire departments). But if what you learned and still use as your linguistic frame of reference is somehow incompatible with what we've done, that is, you do not agree with my linguistic assumptions, then the syntax modules will be minimally useful to you. But that is true of any syntactic modules that you yourself don't create. If this is the case (and I hope it is not), you might like to use the syntax diagramming tool in Accordance to draw your own trees. From what I've seen on the Forum, this is very popular with syntactically interested users.
Associate Professor, Ancient Hebrew and Northwest Semitic Languages
Dept. of Near and Middle Eastern Civilizations
The University of Toronto
blog: ancienthebrewgrammar.wordpress.com

#6 Rod Decker

Rod Decker

    Gold

  • Active Members
  • PipPipPipPip
  • 251 posts
  • Gender:Male
  • Location:Pennsylvania
  • Accordance Version:10.x

Posted 17 October 2010 - 06:35 AM

you might like to use the syntax diagramming tool in Accordance to draw your own trees.


Is this really possible? Where and how? I must be missing something.

Edited by Rod Decker, 17 October 2010 - 07:57 AM.

Rodney J. Decker, ThD
Professor of NT & Greek
Baptist Bible Seminary
NTResources.com/blog/

#7 James Tucker

James Tucker

    Platinum

  • Active Members
  • PipPipPipPipPip
  • 644 posts
  • Gender:Not Telling
  • Accordance Version:10.x

Posted 17 October 2010 - 07:26 AM

Is this really possible? Where and how? I must be missing something.

Yes, it is possible. You would create your own trees in the Diagram Feature of Accordance. Granted, they might not have the rounded "limbs" as the charts do, but you can still successfully create your own diagrams. Moreover, your diagrams would not be the same as the Syntax Charts offered by Accordance, in that they are not searchable. The ability to search is based upon the data entry done by Dr. Holmstedt hence the need to recognize his linguistic assumptions. Of course, you could build your own database in conjunction with your diagrams and run searches with grep functionality. Nonetheless, what I think Dr. Holmstedt has in mind here is that the Accordance Diagram feature hosts the ability to create your own tree in the case that you don't agree with his linguistic presuppositions.



#8 Rod Decker

Rod Decker

    Gold

  • Active Members
  • PipPipPipPip
  • 251 posts
  • Gender:Male
  • Location:Pennsylvania
  • Accordance Version:10.x

Posted 17 October 2010 - 07:56 AM

Yes, it is possible. You would create your own trees in the Diagram Feature of Accordance. Granted, they might not have the rounded "limbs" as the charts do, but you can still successfully create your own diagrams. Moreover, your diagrams would not be the same as the Syntax Charts offered by Accordance, in that they are not searchable. The ability to search is based upon the data entry done by Dr. Holmstedt — hence the need to recognize his linguistic assumptions. Of course, you could build your own database in conjunction with your diagrams and run searches with grep functionality. Nonetheless, what I think Dr. Holmstedt has in mind here is that the Accordance Diagram feature hosts the ability to create your own tree in the case that you don't agree with his linguistic presuppositions.


Of course you could produce an approximation of the trees with the traditional Accord diagram tool, but that's really not suitable. The diagram tool uses the conventions of traditional grammatical analysis and each graphic form has its own meaning. The tree diagrams have, it seems, an entirely different structure and set of conventions and labeling for which the diagram tool is poorly suited. It could be adapted to function either way, but it would only be a poor work-around at present--unless there are hidden abilities that I haven't discovered in the new version. (I assumed that by "the syntax diagramming tool in Accordance" that there might have been a new tool added.)

Edited by Rod Decker, 17 October 2010 - 02:39 PM.

Rodney J. Decker, ThD
Professor of NT & Greek
Baptist Bible Seminary
NTResources.com/blog/

#9 Robert Holmstedt

Robert Holmstedt

    Platinum

  • Accordance
  • 522 posts
  • Gender:Male
  • Accordance Version:10.x

Posted 17 October 2010 - 08:37 AM

Of course you could produce an approximation of the trees with the traditional Accord diagram tool, but that's really not suitable. The diagram tool uses the conventions of traditional grammatical analysis and each graphic form has its own meaning. The tree diagrams have, it seems, and entirely different structure and set of conventions and labeling for which the diagram tool is poorly suited. It could be adapted to function either way, but it would only be a poor work around at present--unless there are hidden abilities that I haven't discovered in the new version. (I assumed that by "the syntax diagramming tool in Accordance" that there might have been a new tool added.)


Apologies is my comment was misleading.

Your comment regarding the traditional diagramming is quite apropos. Traditional diagramming as most of learned it (usually by the 19th c. Reed and Kellogg method) is built around the 8 parts of speech in English. The point of such diagramming is more about clarity re: the parts of speech and very basic syntactic relations than about the hierarchical constituent relations represented in our trees. The constituents overlap with some but definitely not all the parts of speech, as you can tell by the labels. By way of example, a fundamental difference between the two is the treatment of objects. In the Reed and Kellogg method, a transitive sentence is diagrammed as Subject | Verb | Object (i.e., all are on the same level). In our analysis, the object is actually contained within the predicate, Subject | Verb (Complement), a relationship that is shown in the trees by the position of the Complement node.

I hope this clarifies a bit of what we're doing versus traditionally diagramming, which I also think is fun, but would not have worked for our linguistically-grounded module.
Associate Professor, Ancient Hebrew and Northwest Semitic Languages
Dept. of Near and Middle Eastern Civilizations
The University of Toronto
blog: ancienthebrewgrammar.wordpress.com

#10 James Tucker

James Tucker

    Platinum

  • Active Members
  • PipPipPipPipPip
  • 644 posts
  • Gender:Not Telling
  • Accordance Version:10.x

Posted 17 October 2010 - 08:49 AM

Of course you could produce an approximation of the trees with the traditional Accord diagram tool, but that's really not suitable.


I have in mind here the traditional methods of transformational grammar. If you are trying to reduplicate the visuals of the Syntax Charts, then I concur, to a limited degree, that the Diagram module is not ideal. I say concur to a limited degree because I don't deem the current diagram tool as being "poorly suited" for transformational grammar. I am sure some additional features could be added to the predefined set making the module more amenable for transformational grammatical analysis. On another note, I have used, with ease, the current diagram tool to create discourse analysis charts of certain texts. I am very pleased with the results, and they can easily be made into charts or an appendix, if so needed or desired. My point is: I see more flexibility in the current diagram tool for completing various tasks.

Another element I think to recognize is what Dr. Holmstedt brings attention to above: the visuals are built upon an analysis of the text. Thus, we shouldn't devalue linear logical analysis. While it is nice to have the visual charts, I would suggest that if one doesn't have the means to read the chart, then egregious errors are looming. I'm not saying that you are doing any of these things; I am merely saying that we could even create our own user notes containing our own decisions on the syntactical structure of particular verse — we wouldn't have to have a chart per se. Posted Image

Edited by James T, 17 October 2010 - 08:55 AM.


#11 Rod Decker

Rod Decker

    Gold

  • Active Members
  • PipPipPipPip
  • 251 posts
  • Gender:Male
  • Location:Pennsylvania
  • Accordance Version:10.x

Posted 17 October 2010 - 02:57 PM

Apologies is my comment was misleading.

Your comment regarding the traditional diagramming is quite apropos. Traditional diagramming as most of learned it (usually by the 19th c. Reed and Kellogg method) is built around the 8 parts of speech in English. The point of such diagramming is more about clarity re: the parts of speech and very basic syntactic relations than about the hierarchical constituent relations represented in our trees. The constituents overlap with some but definitely not all the parts of speech, as you can tell by the labels. By way of example, a fundamental difference between the two is the treatment of objects. In the Reed and Kellogg method, a transitive sentence is diagrammed as Subject | Verb | Object (i.e., all are on the same level). In our analysis, the object is actually contained within the predicate, Subject | Verb (Complement), a relationship that is shown in the trees by the position of the Complement node.

I hope this clarifies a bit of what we're doing versus traditionally diagramming, which I also think is fun, but would not have worked for our linguistically-grounded module.


Your syntax work will certainly be an added bonus and enable a whole new vista for searches. I hope there is much more info coming in terms of the "hows and whys" of constructing searches--and that it will be equally representative of Greek (helpful to those of us whose Hebrew reading ability isn't as fast as our Greek! :)). I've now read your paper from the users conference and look forward to seeing more such info.

I would note that the traditional grammatical analysis isn't quite as limited as you suggest however. The subject/predicate distinction is quite clear, including the various types of components in the predicate with objects and complements distinguished from each other. It would be possible, I think (though perhaps not practical) to build a syntax module on the traditional diagrams. BibleWorks has been working on doing just that for some time, though I've not heard if they are continuing development of late. (I reviewed the Leedy diagrams for them several years ago.) To do so they developed the various graphical sigla far beyond the typical set. That's gets pretty complicated really fast since there are multiple sigla that appear to be identical. If you check a current ed. of BW you'll notice how complex the tool palette is in their diagramming module.

But my point is not to defend that approach. I'm glad to have a working syntax search capability of any sort--and I'll be even happier when you get the Mark module finished (that's "my book" and one for which I'm about 2/3 finished with a grammatical/syntactical analysis due to the publisher next summer).
Rodney J. Decker, ThD
Professor of NT & Greek
Baptist Bible Seminary
NTResources.com/blog/

#12 Enoch

Enoch

    Gold

  • Active Members
  • PipPipPipPip
  • 486 posts

Posted 27 November 2010 - 11:40 AM

You say "Subject | Verb | Object (i.e., all are on the same level)."

If you mean "level" as measured in inches (cm) from the bottom of a sheet of paper, you are correct. But that statement is not true if you mean level of grouping. In the system I was taught, between the Subject and Predicate (of which the verb is the first part) there is a longer verticle line which goes below the horizontal line equidistance as above. The Direct Object then (not necessarily "object") goes to the right of the verb (first element in the predicate). Between the verb and the direct object there is a vertical line that goes up from the horizontal (not below).

Thus the diagram indicates that there are 2 parts of the sentence; subject on the left, predicate on the right. Within the predicate there is a verb and a direct object. So far as the syntactical idea is concerned, a main level of division is indicated by the long Sub/Pred line. Then a sub-division within the predicate is indicated: verb & D.O.

To me the traditional diagramming gives the mind a much clearer gestalt of the syntax than does the tree system. And the diagramming indicates a mediated syntactical relationship between a modified word & the object of the prepostion. The preposition connects its object to the word modified. The preposition & its object are not coordinate elements related to a modified word.

#13 Robert Holmstedt

Robert Holmstedt

    Platinum

  • Accordance
  • 522 posts
  • Gender:Male
  • Accordance Version:10.x

Posted 27 November 2010 - 09:41 PM

You say "Subject | Verb | Object (i.e., all are on the same level)."

If you mean "level" as measured in inches (cm) from the bottom of a sheet of paper, you are correct. But that statement is not true if you mean level of grouping. In the system I was taught, between the Subject and Predicate (of which the verb is the first part) there is a longer verticle line which goes below the horizontal line equidistance as above. The Direct Object then (not necessarily "object") goes to the right of the verb (first element in the predicate). Between the verb and the direct object there is a vertical line that goes up from the horizontal (not below).

Thus the diagram indicates that there are 2 parts of the sentence; subject on the left, predicate on the right. Within the predicate there is a verb and a direct object. So far as the syntactical idea is concerned, a main level of division is indicated by the long Sub/Pred line. Then a sub-division within the predicate is indicated: verb & D.O.

To me the traditional diagramming gives the mind a much clearer gestalt of the syntax than does the tree system. And the diagramming indicates a mediated syntactical relationship between a modified word & the object of the prepostion. The preposition connects its object to the word modified. The preposition & its object are not coordinate elements related to a modified word.


Enoch,

It is wonderful that you like your diagramming method so much. Anything that helps one understand syntax better is useful. Although perhaps your method is unique, since a search on such diagramming turns up quite a few different approaches as well as numerous variations on the Reed and Kellogg method. And what you described does not match the description of the Reed and Kellogg method I read in many places.

As for our trees, again, if you don't like them, I encourage you not to use them. They are there because they represent the hierarchy represented in the syntax tagging and are for the aid of the user. If it's not an aid, then ignore them and use the user diagramming tool available in Accordance.
Associate Professor, Ancient Hebrew and Northwest Semitic Languages
Dept. of Near and Middle Eastern Civilizations
The University of Toronto
blog: ancienthebrewgrammar.wordpress.com

#14 Enoch

Enoch

    Gold

  • Active Members
  • PipPipPipPip
  • 486 posts

Posted 29 November 2010 - 09:54 PM

Actually, I think it is more profitable to discuss tree diagramming vs. traditional diagramming and how best to represent syntax; not to discuss what I like or don't like.

I refer you to the Wikipedia article on diagramming a la Kellogg at http://en.wikipedia....entence_diagram
where the vertical line between subject and predicate is as I said & not the same as the line between verb & D.O. I don't know anything unique about the diagramming system I was taught; however, I am aware of minor variations.

Iron sharpens iron.

#15 Robert Holmstedt

Robert Holmstedt

    Platinum

  • Accordance
  • 522 posts
  • Gender:Male
  • Accordance Version:10.x

Posted 30 November 2010 - 06:15 AM

Actually, I think it is more profitable to discuss tree diagramming vs. traditional diagramming and how best to represent syntax; not to discuss what I like or don't like.

I refer you to the Wikipedia article on diagramming a la Kellogg at http://en.wikipedia....entence_diagram
where the vertical line between subject and predicate is as I said & not the same as the line between verb & D.O. I don't know anything unique about the diagramming system I was taught; however, I am aware of minor variations.

Iron sharpens iron.


Thank you for the link. I read that quite a while ago when this discussion began, but there are, of course, better resources available, even through Google Books, that illustrate the significant variations out there.

But as a whole I think what you keep missing is that our trees are the *natural* representation of the hierarchy of the syntax in the modules. We won't change them because there is no way to change them to unless we change the tagging scheme, which won't happen.

Because we won't be changing how we approach syntactic relationships, I consider this line of discussion, in terms of the syntax trees used in the syntax modules, to be closed.

Edited by Robert Holmstedt, 30 November 2010 - 01:16 PM.

Associate Professor, Ancient Hebrew and Northwest Semitic Languages
Dept. of Near and Middle Eastern Civilizations
The University of Toronto
blog: ancienthebrewgrammar.wordpress.com

#16 danzac

danzac

    Gold

  • Active Members
  • PipPipPipPip
  • 275 posts
  • Gender:Male
  • Location:Acadia Divinity College
  • Interests:NT, historical Jesus, DSS, NT Backgrounds

Posted 02 December 2010 - 01:28 PM

I was about to start a thread called "users for syntax label form"

Basically, I don't like the labels. I know I can hover over "s" "c" "f" etc., but why not use a slightly longer and more descriptive abbreviation?
How about
I.C.
D.C.
pred.
art.

etc.?

#17 Robert Holmstedt

Robert Holmstedt

    Platinum

  • Accordance
  • 522 posts
  • Gender:Male
  • Accordance Version:10.x

Posted 02 December 2010 - 04:49 PM

I was about to start a thread called "users for syntax label form"

Basically, I don't like the labels. I know I can hover over "s" "c" "f" etc., but why not use a slightly longer and more descriptive abbreviation?
How about
I.C.
D.C.
pred.
art.

etc.?


I agree, although I wouldn't want to use any of the parts of speech labels, since (as I wrote in my "principles" paper, parts of speech don't work entirely for syntax purposes). I'd like to see labels like Cl, Subj, VP, Comp, Det, Voc, Disl, etc.

In any case, I also wrote in the "principles" post on my blog, the labels we're currently using were borne out of convenience -- it was easier in the programming to using single graph labels. Then later, out of necessity, had to use digraph labels for direct speech (e.g., "NA"). Eventually, I hope to have all the labels translated to more standard ones. But since the vast majority of our time is currently taken with tagging new texts and tweaking the search programming, such things like label changes will have to follow later.

We're on the same general page, I think.
Associate Professor, Ancient Hebrew and Northwest Semitic Languages
Dept. of Near and Middle Eastern Civilizations
The University of Toronto
blog: ancienthebrewgrammar.wordpress.com

#18 joelmadasu

joelmadasu

    Bronze

  • Active Members
  • PipPip
  • 50 posts
  • Gender:Male
  • Accordance Version:10.x

Posted 13 August 2013 - 08:29 AM

I was/am trying to find a "tree diagramming" program that I can use with Mac. While searching, I came to this thread, but this was posted in 2010. Are there any updates since then in the Sentence Diagramming tool, where we can draw tree diagrams? Or is there any other software that can do this?


Thank you
Joel




0 user(s) are reading this topic

0 members, 0 guests, 0 anonymous users