In this series (see post #1, #2, #3, #4), I have argued that the study of Biblical Hebrew word order has lacked methodological rigor. In this, the penultimate post, I introduce the last criterion by which the word order data must be filtered.
I have been a bit slower putting up this post since I wanted to check and re-check my data, questioning my judgments as I went in order to produce the best possible results. My eyeballs now hurt more than ever. But, I still hope to finish off the last section by the end of the weekend. From my two or so readers, I covet input.
The Criterion of Pragmatics
The third, and final, criterion by which the raw frequency data is filtered concerns ‘pragmatic markedness’. Attention to the pragmatic features of a clauses is particularly significant for so-called ‘free-order’ languages like Hebrew, that is, languages exhibiting a great deal of word order variation. At the core of this approach is the recognition that the majority of language data contains pragmatically ‘marked’, or ‘non-neutral’, clauses. Even for languages that have a more rigid word order, such as English, pragmatics can produce extreme but grammatically acceptable examples, as with Into the room walked the Prime Minister, a VS clause with a fronted locative PP—certainly not basic order in English.
The operative pragmatic notions for Hebrew are Topic and Focus, both of which motivate the fronting of constituents, which in turn appears to motivate VS order (see Holmstedt 2009 for my model of information structure). In brief, Topic draws a constituent to the front of the clause to either 1) orient the reader/listener to which among previously established entities will now act or experience an event, or 2) set the scene with time or place adjunts (e.g., a temporal PP). Focus similarly draws a constituent to the front of a clause, but for a different reason: it is to contrast the fronted entity with other known or assumed (based on shared knowledge) entities with which it forms a contextually or logically established set. Importantly, whether or not a particularly entity has been previously established (and thus can be a Topic or makes sense as a Focus) is sensitive to the embedded discourse worlds (i.e., conversations) within the larger text. So, for instance, the fact that some person has been mentioned in the narrative does not necessarily establish that entity as available for Topic-status within a conversation embedded within the narrative.
(8) Distinguishing Discourse ‘Worlds’ within a Layered Text
Gen 38.22 וְגַם אַנְשֵׁי הַמָּקוֹם אָמְרוּ לֹא־הָיְתָה בָזֶה קְדֵשָׁה
In (8), the SV clause does not present any Topic or Focus on the Subject. Although the entity אַנְשֵׁי הַמָּקוֹם has been invoked previously the larger discourse (v. 21), there it was invoked by the narrator, whereas here in v. 22 (8) the entity is used within a conversation between Judah and his servant cannot necessarily be taken as an established entity (and thus, available to carry Topic marking). Some entities are assumed as a part of general knowledge (at least, between the narrator/speaker and reader/listener) and so carry Topic marking from the first use. This is not the case with אַנְשֵׁי הַמָּקוֹם in Gen 38.22, since the phrase is new to Judah and cannot be assumed. In light of such complexity, the linguist filtering the word order data by the pragmatic criterion must be sensitive to numerous strategies by which the information structure of a text unfolds.
I have previously separated out clauses with fronted constituents, as I did in Post #3 in example (2), where I separated out qatal and yiqtol clauses that have fronted Adjuncts (2b, 3b) or fronted Complements (2d, 3d). Similarly I pointed out that pronominal Subjects (as in (4b)) are not appropriate for basic word order clauses since such Subjects in Hebrew signal Topic or Focus marking. Using those examples will illustrate how Topic and Focus work as well as how these pragmatic features affect clausal word order, thereby rendering their clauses poor candidates for basic word order. Consider the examples in (9).
(9) Constituent Fronting
a. Adjunct-fronting: Gen 29:34 עַתָּה הַפַּעַם יִלָּוֶה אִישִׁי אֵלַי
b. Complement-fronting: Gen 31:42 אֶת־עָנְיִי וְאֶת־יְגִיעַ כַּפַּי רָאָה אֱלֹהִים
c. Subject-fronting: Gen 23:6 אִישׁ מִמֶּנּוּ אֶת־קִבְרוֹ לֹא־יִכְלֶה מִמְּךָ
In (9a), there are two fronted adjuncts, the temporal adverb עַתָּה ‘now’ and the adverbial NP הַפַּעַם ‘this time’. The first is fronted as a scene-setting (temporal) Topic, the second as a contrastive Focus: Leah thinks that this, third son, will finally endear Jacob to her, whereas apparently the first two sons did not gain her the favor she desired. The Topic-Focus order in (9a) illustrates that even the pragmatic functions have an order in Hebrew: Hebrew exhibits multiple Topics, multiple Foci, but when both a Topic and Focus are present, the order is always Topic-Focus.
Like הַפַּעַם in (9a), the fronted Complement in (9b) carries Focus marking. In Gen 31:42, Jacob finishes his blistering charge against Laban, which culminates in our example and the short clause that follows it, וַיּוֹכַח אָמֶשׁ ‘and he rebuked (you) last night!’. It is not clear if the NPs עֳנִי and יְגִיעַ כַּפַּיִם constitute the contrastive constituents or the 1cs pronouns attached to the NPs. Is Jacob asserting that his experience has been one of pain and suffering (presumably in contrast to how Laban would characterize it) or is he simply contrasting who God has favored: him (not Laban)? Both options are contextually felicitious and both may be intended, which is possible since the scope of the Focus is over the entire compound constituent.In any case, this fronting of the Complement communicates something like the following: Though you (Laban) have continually treated me unfairly, my oppression and my toil caught God’s attention.
Finally, two features in (9c) indicate that it is an unambiguous example of Subject-fronting. First, the Verb is negated, which I have suggested above is a feature associated with VS order. Thus, any constituent in front of the Verb can only be there due to a pragmatically-motivated fronting. Also, the presence of the Complement אֶת־קִבְרוֹ before the Verb is unarguably a case of Topic or Focus fronting. This necessarily points to the Subject that precedes the fronted Complement as a case of fronting as well. So, what pragmatic roles to the fronted Subject and Complement fill? In the context, taking the Subject אִישׁ מִמֶּנּוּ ‘a man from us’ as the Topic makes good sense, since the Hittite speaker(s) is orienting Abraham to the previously mentioned entities (whom Abraham had referenced as ‘you’ in the preceding verse) would act. Another way to think of this is as a choice that the Hittite speaker made in the response: among the obvious choices, rather than starting with ‘you, Abraham, …’, he started with ‘a man among us’. The key to understanding the force of the Focus-fonted Complement is understanding that the while the scope of the Focus lies over the entire NP, it can also be associated with one constituent within the phrase; in the case of אֶת־קִבְרוֹ the Focus is on the 3ms suffix, indicating that the force is ‘his (own) grave’. Thus, a paraphrastic translation of (9c) that highlights the pragmatics is ‘no man among us would withhold his own grave from you’.
There is, to my knowledge, complete agreement that examples like (9a) and (9b) illustrate the Topic- or Focus-fronting of a constituent that is normally positioned after the Verb. The numbers in (10) demonstrate that a fronted constituent is much more often followed by VS order than SV order. But given the complicating factor of the fronting itself, neither ‘X-VS’ or ‘X-SV’ (where ‘X’ means a fronted constistuent) can be used to isolate the basic order.
(10) Constituent Fronting
a. Adjunct-fronting: VS (101x) vs. SV (6x)
b. Complement-fronting: VS (25x) vs. SV (0x)
c. Subject-fronting: 192x
The Subject-fronting illustrated in (9c) and quantified in (10c), though, lies at the heart of the basic word order discussion. It is clear that many SV clauses are best understood in the context as non-basic, whether the Subject is a pronoun (which, in a null Subject language like Hebrew, always signals Topic or Focus) or is positioned before another fronted constituent, as in (9c). The SV order in such clauses reflects either a Topic or Focus-marked Subject. The same cannot be said, though, for a number of examples like (11).
(11) SV without Topic or Focus-marked Subject (47x)
Gen 37:20 חַיָּה רָעָה אֲכָלָתְהוּ
There is nothing in the discourse context of the example in (11) that even weakly suggests a Topic or Focus reading of the Subject. The plan by Joseph’s brothers to pass off Joseph’s disappearance as a wild animal attack does not build on any previously established or generally presumable discourse entity. As a newly introduced entity, חַיָּה רָעָה cannot be a Topic; moreover, a contrast created by Focus-marking on חַיָּה רָעָה makes no contextual sense—with what would the imaginary wild animal be contrasted?
Although the majority of main, indicative SV clauses reflect Topic or Focus marking on the Subject (9c, 10c), the existence of some SV clauses (11) that are in main, indicative clauses with no Topic or Focus marking—and so arguably basic—challenges the traditional VS classification of Biblical Hebrew. To add to this, if Biblical Hebrew were a strong VS language (Longacre 1995), then even with the omission of wayyiqtol clauses, we would expect to see numerous main, indicative VS clauses. But we do not: qualifying SV clauses (11) number almost twice as many as qualifying VS clauses (12).
(12) VS ‘basic word order’ (28x)
Gen 16:2 הִנֵּה־נָא עֲצָרַנִי יְהוָה מִלֶּדֶת
How do we deal with the paucity of simple VS clauses in Genesis? How can we account for the strong tendency towards VS order in subordinate clauses, modal clauses, and any clause with a fronted constituent (other than a fronted Subject)?
In the 6th (and last) post, I will briefly summarize the issues from the methodological perspective that I have been so strongly pushing (since the methodological component has been inexusably absent in previous discussions of Biblical Hebrew word order. I will also provide my ‘story’—how I account for the various word order facts in Biblical Hebrew.
Holmstedt, Robert D. 2009. Word Order and Information Structure in Ruth and Jonah: A Generative-Typological Analysis. Journal of Semitic Studies 54/1: 111-39.
Longacre, Robert E. 1995. Left Shifts in Strongly VSO Languages. Pp. 331-54 in Word Order in Discourse, ed. Pamela Downing and Michael Noonan. Amsterdam: John Benjamins.