5 replies to this topic

[Robert Holmstedt]

In the syntax database, participles are treated as the complements of a (mostly null, rarely היה) copula.

 

This occurs in two different syntactic environments: 1) when the null copula is the main/matrix predicate, e.g., "the man [is] running"; and 2) when the null copula is within a relative clause (marked or unmarked by ה), e.g., "the man (who) [is] running". 

 

In the case of #2, an additional complication is that sometimes the head of the relative is also null (aka "a headless" or "independent" relative), e.g., "Ø[the one] (who) [is] running ... is a friend".

 

Searching for cases like this can be found with the searches attached below (one search without the ה relative marker and one search with the ה relative marker). Note that in the absence of the relative ה, there are 3 nulls in a row: "[he] [who] [he] is running". Therefore, it is necessary to click the Null item and chose "3 consecutive" nulls to get the correct hits.

 

Finally, one further complication is the use of the "participle" form (Qal=qotel; Hiphil=maqtel) as an agentive noun. These cases of "substantival participles" are not actually participles; they should be understood as agentive nouns that simply share the same form as the participle (i.e., homophony). But the homophony creates obvious confusion—is רֹעֶה simply "shepherd" or is it truly participial, "[one] (who) [is] shepherding"? We have taken a narrow approach to identifying this category the question and follow the criteria below for analyzing a given QOTEL/MAQTEL form as a substantive:

 

Participles are tagged as simple nominal items WHEN:

i. The Participle is one of the few recognized (mostly) agentive forms (e.g.,Qal שׁוֹפֵט‎, בּוֹנֶה‎, יוֹצֵר‎, עֹשֶׂה, ‎רוֹפֵא‎, and רוֹעֶה, and less commonly from derived בנינים such as מַצִּיל), 

 

and

 

ii. does NOT modify an overt NP head (relatively), and does NOT have any complements or non-possessive adjuncts, e.g., אֶבֶן מָאֲסוּ הַבּוֹנִים ‘the stone (that) the builders rejected’.

 

and

 

iii. is NOT bound to a following pronoun or noun, e.g., בְּמוֹת הַשּׁוֹפֵט ‘in the dying of the judge’. 

 

* If the pronoun or noun to which the participle is bound does NOT make sense as a complement of the participle (e.g., רֹעֵי גְרָר ‘shepherds of Gerar, in which Gerar makes no sense as the complement of shepherding; contrast with יושׁבי העיר in which the NP is an appropriate locative complement and so the participle must be tagged as a relative), but does make sense as a possessor, then the participle may also be tagged simply as a noun, e.g., שׁוֹלְמִי ‘my friend’. Note that such cases of a substantive participle bound to a possessor (and not a complement) are very rare. 

 

Searching for substantive uses of these morphological forms (QOTEL/MAQTEL) is a simple as searching for SUBJECT=VERB(Participle), COMPLEMENT=VERB(Participle), etc. 

 

 

 

Attached Files

•  NullHeadParticipleRelatives.zip   3.56KB   2 downloads

Edited by Robert Holmstedt, 22 November 2017 - 09:39 AM.

[Michel Gilbert]

Hi Robert, 

 

I find this detailed explanation helpful, especially the reminder about the number of null constituents.

 

But I find the contrast between רועי גרר and יושׁבי העיר a bit confusing. Since גרר is a עיר, then רועי גרר could be a locative complement, those who are shepherding near/around/in the vicinity of the city of Gerar, just like יושׁבי העיר , those who are living in/inhabiting the city. I'm not necessarily asking you to clarify, because I plan to go over all of the Manual, etc., again. If I'm still confused, I may ask then.

 

I also have a question about real participles - were you ever tempted to tag them as adjectives and nouns (agentive), and drop the term participle altogether?

 

Thanks, and regards,

 

Michel

[Robert Holmstedt]

Dear Michel,

 

The difference is that רעה does not take a locative complement, regardless what you do with גרר, whereas ישׁב does take a locative complement (often a PP, but not always). So it boils down to the valency of the verbal root. 

 

No, (non-agentive) participles are clearly eventive, which clearly distinguishes them from adjectives. We followed John Cook's analysis of participles, which he dealt with in a 2008 JNSL article and also in his 2012 verb book.

 

Regards,

Robert

[Michel Gilbert]

Hi again,

 

I was asking about tagging bc I borrowed a copy of your Biblical Hebrew: An Illustrated Introduction (Draft Copy), 2011, where you say, "Participles are adjectives that have בנינם distinctions . . . declined like adjectives." So it was just an out of curiosity question if anyone has considered subsuming participles under adjectives, in a grammar or database.

I wasn't challenging you about tagging, or nomenclature. It is just a fact that is your database, and we its users have to understand how you tagged it. Understanding more about theory helps, as well as finding out how you tagged something that we are interested in. I guess that my biggest question about methodology centers around valency of the verb and null constitutents. As I read through the Manual, etc. again, I want to gather clues about valency and then do some more reading on it. For instance, and again, this is not a challenge, Gen 26:20 seems to mean (to me) that the רועי גרר were shepherding a distance from גרר, came into contact with another group of shepherds, and claimed it was their territory. Such disputes occur all over the ANE among shepherds, etc., over land rights. Now my question is, whether you agree with me or not, am I taking valency of the verb too far? And what about lexicons or dictionaries or specialized word studies that mention these types of wider social and political contexts, not just for רעה? 

On another note, I'm really looking forward to Isaiah. I do have questions, and I hope you have the answers.

Regards,

Michel

[Robert Holmstedt]

Michel,

 

First, no worries. And if you have specific Isaiah questions, send them to me. I can always look ahead of where I'm tagging at the moment (Isaiah is going very slowly, but that's also because I'm using it to start an English translation that mirrors the syntax database, including representation of nulls -- it's a tedious process).

 

Second, yes, that's not quite what valency is about. What you're discussing is the semantics of the bound relationship, apart from the valency issue of the root. Whether you interpret the bound relationship רעי גרר as "shepherds of Gerar" or "shepherds around Gerar", the point is that גרר is not a complement in either case. When it is bivalent, the verb רעה takes a complement of [what is shepherded] (i.e., flocks, or metaphorically, people).

 

Third, on nulls and valency, John Cook has a couple articles coming out that discuss these things. You should message him for a copy.

 

Finally, on the participle, technically it represents the nominal takeover of event predication (Cook 2008). As such, yes, it is at some level an adjective. And for the database, the syntactic tagging between adjective and participle differs in one significant regard: the inclusion of a predication with the participle (due to it incorporating an event predication). So I could step back a half-step on my first explanation above: participles could be classified as "adjectives," producing two adjectives classes -- those with and those without their own predication. In fact our database would not differ if they were. But this morphological label is embedded in the Westminster tagging, not our syntax layer.

 

Regards,

Robert

[Michel Gilbert]

 

. . . on nulls and valency, John Cook has a couple articles coming out that discuss these things. You should message him for a copy.

 

 

Thanks Robert for the tip, and John for sending me the article, which helped a lot. It's starting to make a lot of sense.