Trying to understand pronominal suffixes

[Susan]

I’m looking at pronominal suffixes on nouns and trying to figure out how they’re tagged in the syntax module so that I can use them in searches.  Most of them are adjuncts, but some of them are complements.

 

I understand that a bunch of things tagged as nouns in the Westminster morphology are acting as prepositions (that’s my way of thinking about it anyway....  לפני, etc), and when these have pronominal suffixes they are consistently complements, which makes sense. What I’m not following is why some nouns that seem to me pretty normal (i.e. without special semantic requirements) have pronominal suffixes that are complements.  There does seem to be a pattern -- many of them are words for family relationships ( אב, אחות, אח etc), and others are body parts -- so I suspect that it is intentional, and I can maybe see how these categories do have special semantic requirements, but since most other instances of those words have suffixes that are adjuncts, I can’t quite figure it out.

 

E.g. Gen 12:19:

 

למה אמרת אחתי הוא ואקח אתה לי לאשה ועתה הנה אשתך קח ולך

 

The suffix on אחתי is a complement and on אשתך it's an adjunct. (Does that verse deserve another nested level of speech?)

 

But other places אשה takes a complement as suffix, e.g. Gen 3:17:

 

...כי שמעת לקול אשתך

 

Thanks.

[Robert Holmstedt]

Susan,

 

For all the tagging error routines I've run, this wasn't one I thought of. Thanks -- you've uncovered a set of inaccuracies (98x) that are fairly easy to fix. There is nothing special about אח, אחות, or אשׁה, etc. Simple mistakes. Thankfully, it's a tag change and not one requiring a lot of bracketing changes. 

 

Clitic pronouns on nouns are always (supposed to be) adjuncts, except for the cases you mention in which the noun has become part of prepositional complex (לפני, מפני, בעבור, בתוך, etc.). 

Edited November 27, 2015 by Robert Holmstedt

[Susan]

Thanks, Dr. Holmstedt, that's helpful.

 

I think there are a fair number of nomen rectum (not sure how that becomes plural....clitic hosts?) that are tagged as complements as well. I'm looking at אהל מועד in Num 4:3, a phrase elsewhere tagged with an adjunct. I'm curious what you think of these two searches (below) for finding them. I wasn't sure about the choice of lexemes to exclude in the bound column. With this search I have too many hits because the exclusion is missing some lexemes that seem to deserve exclusion only when connected to prepositions. I suppose that adding those (e.g. פה) plus another column immediately before to exclude hits with a preposition in that position may work, but the search would then exclude any construction preceded by a preposition, which doesn’t seem appropriate.

 

Some of them may have something more complicated going on -- Gen 19:28 is the first hit my search with the article brings up, and there it looks like (the second) ארץ is tagged as a verbal complement. Ex 22:12 looks to me like the morphology tagging is wrong (er... it disagrees with the syntax tagging), and עד is not in construct.

 

 

Edited November 28, 2015 by Susan

[Robert Holmstedt]

Yes, this is another good proofing search. A number of the hits using this are actually ok (compound PNs, the use of פי as a particle), but some are simply mis-tags and others show an interesting divergence from the Westminster tagging (e.g., I would take חשׁבי מחשׁבת in Exod 35.5 as a participle bound to its complement, but the Westminster tagging takes חושׁב as a noun, not a participle). I would estimate about 75 tagging errors, just looking through the results. 

 

I will have these and the other set fixed fairly soon. Thank you.

Edited December 6, 2015 by Robert Holmstedt

[Michel Gilbert]

Hi Robert,

 

Fyi, SESB BHS tags חשבי in in Exod 35.35 as "Qal, participle . . . "

 

Regards,

 

Michel

Edited November 29, 2015 by Michel Gilbert

[Robert Holmstedt]

Good to know! Thanks.

[Susan]

I'm still trying to get a feel for distinguishing adjuncts and complements. (Ruth handbook introduction re-read, re-appreciated.) Based on the above discussion, it seems that a participle may be bound to an accusative complement, but when nouns are bound to what I would call an "objective genitive" (realizing that this terminology is foreign to this discussion), these are necessarily adjuncts. So in Ex 35:35, חשב earlier in the verse without complement/adjunct is a noun? Or it's a participle with a different valency compared to the latter?

 

I take it that prepositional phrases may be adjuncts or (if adverbial) complements, but the object is always a complement to the preposition since it is required? For the search below I excluded the Aramaic portions (just because I can't read them). I get 113 hits (from 1581 without the lexical restrictions), most of which seem to be objects of prepositions mis-tagged as adjuncts unless I'm missing something (which is likely).

 

Interestingly, a few of these overlap with the first search in this thread for complement pronominal suffixes, i.e., the complement/adjunct seem to be flipped, e.g. Deut 18:15, 1 Sam 6:8 (Gen 47:19, unless that’s a complex preposition). There is also a group of qere readings, mostly Sam/Kings, that come up because of how the qere is diagrammed, as a series of adjuncts (e.g. 1 Sam 8:3, 10:21), which I suppose is intentional. 

 

[Robert Holmstedt]

A lot of these are still hits where the "noun" is used in a non-nominal way (e.g., פי), but some are simply errors and I will also fix these. 

 

The participle issue is complex, but yes, to simplify I would agree that that the first use in Exod 35:35 is still participial, but with a null complement, much the same way that a verb like "build" (bivalent) can sometimes be used without an overt complement, "He built and built and built, but didn't produce anything". John Cook will discuss these when he works up the valency lexicon that we've been using for the project. In the meantime, he has a few published works on valency in BH that might be helpful.

 

Thanks.

[Susan]

My lexical exclusions were an attempt to get rid of most of those non-nominal uses, including פה/פי and the other common ones. I think it captured most of them, though I see now Ps 72:7 אד בלי ירח  probably warrants adding בלה  to that list of nouns. One can throw out all construct forms, but that also loses some such as Deut 18:15 מאחיך that are real nouns with pronominal suffixes. 

 

Thank you very much for the explanation on valency. (I’m still thinking of going back to this search if/when I can understand the idea.) I’m very excited to hear that there is a “valency lexicon” in the pipeline. Having a reference for that would be very helpful. In the meantime, I will work through this article (link provided for other Accordance users perusing the syntax forums who might be interested), which seems to provide quite a bit of explanation.

[Λύχνις Δαν]

Thanx for the link to the valency article - looks really good.

 

Bit of a linguistics-heavy week this one

 

Thx

D

 

 

 

[Susan]

Thanx for the link to the valency article - looks really good.

 

Bit of a linguistics-heavy week this one

 

Right... I think I still have a fairly superficial understanding of how the syntax diagrams (and the linguistic realities that underlie them) work, and this is the biggest gap for me.  This seems to be a good primer. Maybe we could link it to the syntax documentation somehow. (Sorry if it already is and I missed it.)

 

Is there a good way to attach a construct search to posts here in a way that people can open in Accordance? I’ve just been doing screenshots, but that one above, for instance, is kind of obnoxious to reproduce because the number of lexical items that have to be entered.

 

Also, is there a way to bring up the word-select box within the enter-lexeme box in the construct window? Command-L that works in the regular search tab doesn’t seem to be working for me. The box does pop up from time to time, but I can’t seem to figure out how to command it, and entering plain text without specification about vowels (not to mention homonyms) is no good. (I expected it to exclude all of the options (from my negated term) just like the search box will find all of the options from unvocalized text, but it didn’t seem to be getting the one I needed.)

 

Thanks!

[Λύχνις Δαν]

Right... I think I still have a fairly superficial understanding of how the syntax diagrams (and the linguistic realities that underlie them) work, and this is the biggest gap for me.  This seems to be a good primer. Maybe we could link it to the syntax documentation somehow. (Sorry if it already is and I missed it.)

 

I'm working through Fromkin et al. (An Introduction to Language - recommended to me by Dr. Holmstedt in another post), at the moment and drawing Phrase Structure Diagrams like mad in chapter 4. I believe that a better understanding of the theory will help my use of the syntax modules.

 

 

Is there a good way to attach a construct search to posts here in a way that people can open in Accordance? I’ve just been doing screenshots, but that one above, for instance, is kind of obnoxious to reproduce because the number of lexical items that have to be entered.

 

Not that I know of - you can attach a workspace or saved tab but beyond that I don't know of a way. Actually I don't think you can attach the acc file types so you would have to link to DB or something like that.

 

Also, is there a way to bring up the word-select box within the enter-lexeme box in the construct window? Command-L that works in the regular search tab doesn’t seem to be working for me. The box does pop up from time to time, but I can’t seem to figure out how to command it, and entering plain text without specification about vowels (not to mention homonyms) is no good. (I expected it to exclude all of the options (from my negated term) just like the search box will find all of the options from unvocalized text, but it didn’t seem to be getting the one I needed.)

 

 

If I understand you correctly, double click on LEX of your LEX element and that will pop it up. For Greek you can just type a comma and then go on typing the next word and hit return when you have it partly done and the Select Lexical Forms box comes up. I can't type Hebrew well enough yet to make that test easily in a Hebrew construct.

 

Thx

D

[Susan]

I'm working through Fromkin et al. (An Introduction to Language - recommended to me by Dr. Holmstedt in another post), at the moment and drawing Phrase Structure Diagrams like mad in chapter 4. I believe that a better understanding of the theory will help my use of the syntax modules.

 

If I understand you correctly, double click on LEX of your LEX element and that will pop it up. For Greek you can just type a comma and then go on typing the next word and hit return when you have it partly done and the Select Lexical Forms box comes up. I can't type Hebrew well enough yet to make that test easily in a Hebrew construct.

 

 

Thanks for the reading recommendation!

 

Double clicking LEX does work. The problem is that in this case (and not infrequently in creating other construct searches), I was running the search repeatedly, looking through the results to find additional lexemes I wanted to exclude, and then returning to the construct hoping to add to the list. If I double click LEX, I have to start the entire list over again. If I only enter the new one, the rest of them disappear, which is not ideal.

[Λύχνις Δαν]

Yeah I wondered if that might be the problem. Sorry I don't know a good way around that. You can perhaps cut and paste the list of words from LEX and paste it into the lower right box in the Lexeme popup ? I don't really know. Hopefully someone else has an idea for this

 

Thx

D

[rwrobinson88]

I found this masters thesis done by a student under Dr. Cook. Thought it might be helpful also. 

 

Verbal Valency in Biblical Hebrew: An analysis of the valency of "br"

 

http://place.asburyseminary.edu/ecommonsatsdissertations/598/

Edited December 3, 2015 by rwrobinson88

[Robert Holmstedt]

-- update: I have now fixed all the tagging issues mentioned in this thread; I found and fixed @250 tagging errors. Many thanks to those using the syntax module for raising these issues.

[Susan]

-- update: I have now fixed all the tagging issues mentioned in this thread; I found and fixed @250 tagging errors. Many thanks to those using the syntax module for raising these issues.

 

Snappy fast, as expected. 

[Robert Holmstedt]

FYI -- there will soon be an update to the Hebrew syntax with all these corrections, as well as the last 4 chapters of Hosea.