The Nexus between Text Criticism and Linguistics: The Case of Leviticus 1:17

At the end of May I will deliver a paper on this topic at the annual meeting of the Canadian Society of Biblical Studies. The paper is finished, although I have it out to a couple trusted readers. When I feel it is complete, I will post it on the blog and note it here.

Below is my summary of the paper.  I welcome your thoughts, especially those of you more text-critically inclined, since I do not claim to be a text critic as such.


1. Introduction

There is an uncomfortable truth that linguists of ancient languages admit only rarely and even then with some reticence (and usually in a dark, empty room): we are, plain and simple, dependent on the paleographer, the epigrapher, and … (dare I admit it?), the text critic. For without those scholars who concern themselves with the decipherment of scripts, the first reading of texts, and the reconstruction of textual histories, the linguist would have nothing to analyze.

The difficulty of facing this truth is manifested in the modus operandi of ancient Hebrew linguists: rarely do scholars of biblical Hebrew question the wholesale acceptance of using the Masoretic text, dating to 1008 C.E., as representative of the linguistic system(s) of ancient Hebrew from 1500 years prior. They use the text of the standard printed critical edition, the Biblia Hebraica Stuttgartensia – or its electronic form from some computer program, without qualification, without reckoning with complex textual history represented by the data.

Ancient Hebrew linguists do not always seem to have learned the lessons taught by the Qumran texts, that while some scribes were quite passive transmitters, other scribes expanded, rearranged, and clarified the materials they were transmitting. It is thus clear that ancient language linguists rely on those scholars who investigate the features of an ancient text’s scribal history (using scribal here to refer to author and/or copyist). It is equally clear that, while we need not become experts in textual criticism ourselves, we must understand the issues involved.

Conversely, everything I have said indicates that if the linguist must be aware of text, the text critic must be aware of the linguistic systems. For if the scribe updates a text based on his native grammar and that grammar differs, in large or small ways, from the grammar of the text being copied, the text critic must also be aware of the diachronic changes in linguistic systems in order to understand properly the diachronic changes in the text. It is from this perspective, what a text critic may learn from a linguist, that I shall consider the case of a pronoun variant in Lev 1:17.

2. The Text Critical Problems: Lev 1.17

The text of Lev 1.17 as it is given B19a is provided in (1).

(1) Lev. 1:17 in MT (B19a) (LXX, SamP, Pesh, Targ.Onq)

‏ וְשִׁסַּע אֹתוֹ בִכְנָפָיו לֹא יַבְדִּיל וְהִקְטִיר אֹתוֹ הַכֹּהֵן הַמִּזְבֵּחָה עַל־הָעֵצִים אֲשֶׁר עַל־הָאֵשׁ עֹלָה הוּא אִשֵּׁה רֵיחַ נִיחֹחַ   לַיהוָה׃

‘And he shall tear [the bird] by its wings; he shall not split (it) in two. And the priest shall send it up in smoke at the altar, upon the wood that is on the fire. A burnt-offering is it, a fire-offering of soothing aroma to Yhwh.’

The Hebrew text of B19a is closely followed by the Septuagint, Samaritan Pentateuch, Peshitta, and Targum Onqelos. One text, given in (2), from Cave 4 at Qumran, however, provides some interesting departures.

(2) Lev 1.17 in 4QLevb (4Q24 fl_7:20-21)

ושסע אתיו בכנפיו ולא יבדיל [והקטיר אתו הכהן המזבחה על הע]צ֯יֹם אשר ע֯ל [ הא]ש עלה היא ריח ניחוח ליֹ[יהוה]

‘And he shall tear it by its wings; he shall not split (it) in two. [And the priest shall send it up in smoke at the altar, upon the wo]od that is on [the f]ire. A burnt-offering is it, (or: a burnt-offering is) a soothing aroma to Yhwh.’

The Qumran version exhibits three noticeable differences from B19a: two orthographic variants (the addition of the י in אתיו and the addition of the ו on ולא) and one lexical variation (the lack of the word אִשֵּׁה). Towards the end of the verse, though, there is a difference that, as far as I can tell, has not been given due consideration: the third person pronoun, underlined in the example, is the feminine pronoun in contrast to the the masculine pronoun in the B19a. This difference in the pronoun is also reflected in one Aramaic Targum, Pseudo-Jonathan, given in (3).

(3) Lev 1.17 in Targum Pseudo-Jonathan

ויתלע יתיה בגדפוי ולא יפריש גדפוי מיניה ויסיק יתיה כהנא למדבחא על קיסין דעל אישתא עלתא היא קרבן דמתקבל ברעוא קדם ייי

‘And he shall tear it by its wings. But he shall not separate its wings from it, and the priest shall take it up to the altar upon the wood that is upon the fire. A# The burnt-offering is it, (or: a burnt offering is) sacrifice that is received with pleasure before God.’

#Thanks to Pete Bekins for the correction

As with the Qumran text, the pronoun standing after the noun עלתא ‘burnt-offering’ is the feminine pronoun היא instead of the masculine הוּא of B19a. Thus, Pseudo-Jonathan witnesses the same pronoun difference we see in 4QLevb.

The central question is whether the pronoun differences between the B19a tradition and the non-agreeing texts reflect transmission errors, i.e., textual and linguistic “garbage,” or intentional changes. If the changes of the pronouns were intentional, what was the reason for each? As with many text critical issues, deciding which textual option is better is not simply a matter of prioritizing certain manuscripts and the traditions they represent or mechanically siding with the weight of manuscript evidence. In the case of Lev 1.17, while both the typically favored textual traditions and the number of manuscripts suggest that the 3ms הוּא is the older, better reading, we could argue in response that lectio difficilior potior (‘the more difficult reading is preferable’): the 3fs היא is the more difficult and thus older reading and the 3ms הוּא represents scribal correction.

It is precisely at such a point where the linguist might have the tools to break the impasse. In the case of the pronoun of Lev 1.17, the textual differences may highlight a feature of the grammatical system of Hebrew that is often overlooked and more often misunderstood.

3. The Linguistic Problem

At the heart of the ‘pronoun problem’ in both passages is the syntactic role of the independent personal pronouns in Hebrew. In most cases, the anaphoric pronouns, such as הוּא ‘he/it’ and הִיא ‘she/it’, are just that – pronouns referring back to a previously stated or assumed nominal referent. In some cases these pronouns are used demonstratively, as in הָאִישׁ הַהוּא ‘that man’. And in some debated cases the 3rd person pronouns may fulfill the role of the present tense copula ‘is’, linking a subject and its predicate. The first and last of these roles pertain to the possible function of the pronoun in Lev 1.17, which I have summarized in (4).

(4) Grammatical Options for Lev 1.17 in B19a (and parallel witnesses)

a) הוּא is an anaphoric subject pronoun, referring back to the sacrificial item (‏הָעוֹף from v. 14), with אִשֵּׁה רֵיחַ נִיחֹחַ לַיהוָה as appositional modification, ‘a burnt-offering is it, a fire-offering of soothing aroma to Yhwh’;

b) הוּא is a resumptive pronoun following עֹלָה as a casus pendens noun phrase with אִשֵּׁה רֵיחַ נִיחֹחַ לַיהוָה as predicative modification, i.e., ‘a burnt-offering – it (is) a fire-offering of pleasing aroma to Yhwh’;

c) הוּא is a copular pronoun serving to link עֹלָה with a classifying predicative phrase אִשֵּׁה רֵיחַ נִיחֹחַ לַיהוָה, ‘a burnt-offering is a fire-offering of pleasing aroma to Yhwh’.

The first option (4a), that the pronoun is not related to the immediately preceding fs noun עֹלָה but refers back to a ms noun in a preceding verse, is the way that most modern interpreters understand the clause. The second option (4b), as a casus pendens with the resumptive use of the pronoun, is grammatically problematic due to the lack of agreement between the fs עֹלָה and the ms הוּא. The third option (4c) also suffers from lack of gender agreement, but as we will soon see, this is not necessarily an issue for the pronominal copula.

Assuming that the change of pronoun in the texts of 4QLevb and Targum Pseudo-Jonathan does not reflect a scribal error in either text, there are three similar options for analyzing the grammar of the 3fs היא, which I have provided in (5).

(5) Grammatical Options for Lev 1.17 in 4QLevb (and Targum Pseudo-Jonathan)

a) היא is an anaphoric subject pronoun, referring back to some feminine singular referent within the discourse unit of Lev 1.14-17, with ריח ניחוח לייהוה as appositional modification, i.e., ‘a burnt-offering is it (?), a pleasing aroma to Yhwh’;

b) היא is a resumptive pronoun following עלה as a casus pendens noun phrase with ריח ניחוח לייהוה as predicative modification, i.e., ‘a burnt-offering – it (is) a pleasing aroma to Yhwh’;

c) היא is a copular pronoun serving to equate עלה with ריח ניחוח לייהוה, ‘a burnt-offering is a pleasing aroma to Yhwh’.

Whereas with the 3ms pronoun הוּא of B19a the second and third options are grammatically problematic, the case is exactly the reverse with the 3fs היא of 4QLevb: the anaphoric option (5a) suffers from the lack of any suitable feminine referent within the context of vv. 14-17; there is no noun for היא to point back to. In contrast, the second (5b) and third (5c) options are grammatical for 4QLevb.

To sort these options out and reconstruct a plausible history for the differences requires that we investigate the copular use of the pronoun in Hebrew. But this post is already lengthy enough, so I will post in the future on my argument for a pronominal copula in ancient Hebrew. For now, here is my solution to the variant problem in Lev 1:17.


4. Conclusion: Linguistic Insights for Text Critics

For the variant in Lev 1:17 we face manuscripts that diverge on the gender of a pronoun. Building on my case for the copular use of the pronoun in Hebrew, I suggest that it is possible to reconstruct a plausible history for both text critical issues.

For Lev 1.17, B19a and those versions agreeing with it represent the older text. The 3ms pronoun הוּא is anaphoric, but its referent is three verses back (in 1.14). Long-distance anaphora such as this is difficult to process. The long-distance anaphora as well as the known―even if rare―copular use of the pronoun allowed or provoked the scribe of 4QLevb to update the text to provide a grammatical smoother reading in the immediate context. In other words, by virtue of one consonantal change (ו to י) the scribe converted the 3ms anaphoric pronoun into a 3fs copular pronoun, intentionally or not. The 3fs pronoun in Targum Pseudo-Jonathan may be explained similarly, particularly in light of the greater use of the copular pronoun in Aramaic, or it may be that the Vorlage for this Targum lies in the tradition spawned by the change in the 4QLevb text.

I began this study by confessing linguists’ reliance on scholars who study texts and their histories. In a delightful case of turnabout, I have illustrated how the relationship works the other way as well, that those who work on texts and their histories must confess a reliance on those of us who work on languages and their histories.


9 Responses to “The Nexus between Text Criticism and Linguistics: The Case of Leviticus 1:17”

  1. Simon Holloway Says:

    Thank you for a stimulating article – I look forward to reading your post on the pronominal copula!

    In the meantime, I am a little unclear as to how you reached this particular conclusion. I am inclined to agree with the statement that you made in §2, that היא constitutes the older reading, but you seem to have resolved in your conclusion that the original text was הוא, and that היא represented a scribal “correction”. Does it not make more sense to suggest that the original was היא, but that it functioned as an epicene pronoun, updated later to the masculine form?

    (On a related note, I have tremendous difficulty even distinguishing a waw from a yodh in the Qumran material! My experience with it is limited, and I’ve not seen 4QLevb, so I shall take your word for it that this is what it says.)

    • robertholmstedt Says:


      Thank you for stopping by!

      To clarify what I say in §2 — I don’t assert that היא is the older reading; rather, the last paragraph of §2, I suggest that it would be *possible* to assert thus using the typical text critical principle of lectio dificilior. But then I go on to suggest a different way of approaching the problem and a different solution — that הוא of B19a is the older (even though it’s also transparently the “easier” reading) and those witnesses to a 3fs היא are later.

      By the way, I’m intrigued — do you have good evidence that the feminine pronoun, of any stripe (demonstrative, personal) functioned as an epicene pronoun in ancient Hebrew? If one could make an argument (thought I don’t, not in a formal way) that any pronoun functions this way, I would have thought it would be the masculine pronoun.

      On the waw and yod in the Qumran texts — this is a good observation. And you’re wise to trust me on this, since I’d be a pretty darned poor scholar if I hadn’t both read the plates myself for the *entire* text in question (to determine the scribe’s habit) and checked my observations with a true DSS scholar (a colleague in my department). In the DSS texts in question, the scribe consistently distinguishes the waw and yod, thankfully!

  2. Simon Holloway Says:

    You are correct: my apologies. For some strange reason, I remembered there being a number of instances within the Pentateuch where the ktiv is היא and the qere is הוא. Of course, it was the other way around. I have heard it suggested by Gary Rendsburg that Hebrew may have originally possessed a single, epicene pronoun, although I note that the only reference grammars at my immediate disposal to actually comment upon the הוא/היא phenomenon (Gesenius, and Joüon and Muraoka) dismiss the epicene theory. It did appeal to me, but it would have been the הוא and not the היא, as you suggest.

  3. robertholmstedt Says:

    Ha! I wondered if Gary was lurking behind your idea. He’s the only person I know of who has made this epicene suggestion.

  4. Pete Bekins Says:


    I agree with your analysis. Note that Psuedo-Jonathan is even more explicit in its reading the phrase as option c. Whereas MT and 4QLevb read indefinite ‏עלה, PsJ reads it as definite ‏עלתא while ‏קרבן, the predicate, is properly indefinite.

    ‏עלתא היא קרבן דמתקבל ברעוא קדם ייי
    “The offering is a sacrifice which….”

    Also, while I suppose a connection between PsJ and 4QLevb is technically a logical possibility based on the evidence you have gathered, it is so unlikely that I don’t know if I would even mention it.


    • robertholmstedt Says:

      Good catch — that was a poor translation on my part. (And I have now noted the change above in the post.)

      And yes, I agree that any direct connection is hard to make (and I had to read back through to make sure I didn’t mistakenly make the connection); rather, I was implying that both texts reflect a similar grammatical adjustment, which given the long-distant anaphora is not hard to believe of two different “authors.”

      • Pete Bekins Says:

        I was thinking of this statement particularly:

        “or it may be that the Vorlage for this Targum lies in the tradition spawned by the change in the 4QLevb text”

      • robertholmstedt Says:

        Hmm … I thought that this statement was sufficiently vague: that I allowed for the two to stand in relation within very large parameters, not that I was linking the two “texts” to each other directly. It is an “or” statement, after all. ;-)

  5. February 2011 Biblical Studies Carnival | A Fistful of Farthings Says:

    […] Holmstedt applied textual criticism and linguistic principles to Lev. 1:17 and then presented a solution for the textual variant in that verse. John Cook argued for the […]

Comments are closed.

%d bloggers like this: