Jer 27:18--again

Rolf Furuli furuli at
Fri Sep 24 11:41:20 EDT 1999

Alviero Niccacci wrote,

>Dear list-members,
>Forgive me for taking up again Jer 27:18, on which much has been
>said, and maybe also too much speculated. -- BTW the more I study,
>the more I appreciate the Masoretic Text. Even where one may have
>good reasons for disagreeing with it, still one should feel obliged
>to trying to understand what they actually understood. IMO this is
>true unrespective of possible divergent ancient versions or any
>modern scholarly emendations. The Masoretes knew Hebrew better than
>we do despite our wide knowledge of Semitic languages and literatures.
>Back to Jer 27:18, Keil-Delitzsch suggested what I think is the right
>*The extraordinary expression _lebiltî bo'û_  has probably come from
>the omission of Jod from the verb, which should be read _yabo'û_. As
>it stands, it can only be imperative, which is certainly not
>suitable. _Lebiltî_ is usually construed with the infinitive, but
>occasionally also with the temp. fin.;  with the imperf., which is
>what the sense here demands, in Ex. 20:20; with the perf., Jer.
>The principle of a shared consonant is being used by different
>authors; in this case, the final Yod of _lebiltî_ is shared with the
>following verbform.
>In my view, _lebiltî_ + a finite verb--yiqtol or qatal-- can be
>explained by assuming an ellipsis of _'asher_ before the finite verb.
>In this same way we can explain the cases of a finite verb used as
>nomen rectum. A famous, though not always recognized, example is
>_bere'$ît bara' 'elohîm_, lit. 'In the beginning of God-created,'
>i.e. 'When God began creating.' One may consult my _Syntax_ #18.
>Similar cases of a finite verb used as nomen rectum are Exod 4:13;
>6:28; 1Sam 25:15; Jer 6:15; 8:12; Psa 65:5; 81:6.
>For equivalent constructions with and without _'as$er_ and/or a
>demonstrative pronoun, one may compare different ways of asking,
>'What have you done?' as found in 2Sam 12:21; Judg 8:1; Gen 12:18,
>and Gen 4:10. One may consult my paper in C.L. Miller, ed., _The
>Verbless Clause..._ pages 242-243.

Dear Alviero,

It is always interesting to read your arguments because of your vast
knowledge of Biblical narrative texts and your consistency in the use of
data. For the benefit of all those who are confused by all the different
ways in which the passagae apparently can be understood, I would like to
point to the basis of our differences, namely, different assumptions and a
different approach.

 I approach the Biblical text as a textual unit, and I try to use as few
assumptions as possible. At the outset I refuse to use as an *assumption*
that the corpus can be divided into two parts, narrative and discourse and
that verbs behave differently in the two. This means that we cannot start
with narratives, as is done in most cases, and explain (away) the meaning
of verbs in other contexts. To do this is the same as to put the parts of a
puzzle together into a fine picture, but at the end to discover that we
have 5 superflous pieces. In practical work, this means that we find a
grammar that seems to fit narratives, and then we read this into other
genres and smooth away their problems. This was exactly what Sperber

Hebrew is a dead language with no informants, and a model of its verbal
system must account for *all* its verbs not just those of a part of the
corpus (some anomalities are of course allowed). Viewing at the outset the
biblical corpus as a unit, does not mean that I insist that the corpus is
uniform, but only that I do not assume that it is not uniform. Therefore I
view a diachronic study of the corpus as important, but refuse to assume
that verbs have one meaning in old texts and another meaning in younger
text; such a difference must be shown, not postulated. I find discourse
analysis to be a fine tool in linguistic study, but do not accept that
verbs have different meanings depending on the genre. If this is
demonstrated, I will accept it, but admit that I am very sceptical.

I see the following three  basic weaknesses in studies of Hebrew verbs:

1) The limited corpus of each study. To apply a particular model to a small
part of the corpus and then extrapolate the results to the whole corpus is
not very convincing.

2) The total lack of a systematic distinction between pragmatic and
semantic factors. I am not aware of a single study that have attempted to
differentiate between them.

3) All the unfounded assumptions that are made. Assumptions must be used
but they should be as few as possible.

To *try* to remedy this situation I use the following approach:

1) The corpus is the biblical text, the DSS, the inscriptions and Ben Sira.

2) In unpointed texts there are just two clearly differentiated groups of
finite verbs, the prefix-conjugation and the suffix conjugation. By making
lists of all the verbs on the basis of the relation between speech time,
event time, and reference time, I will have a good basis to see if there
are two or four conjugations and whether tense is grammaticalized in Hebrew.

3) It is clear that particular forms tend to occur in particular contexts
(.e.g. WAYYIQTOLs in narratives). The reason for this may either be
semantic, pragmatic, or linguistic/stylistic convention. By a scrupelous
distinction between "semantic meaning" and "conversational pragmatic
implicature", based on Grece's principle, I hope to be able to throw some
light on which factors are behind verbal use.  Any results in the study of
a dead language are only tentative, but we should use as "objective"
methods as possible.

Applying these points to your words above, I will say: It seems to me that
your grammatical understanding is put above the textual witness. It is of
course possible that there was an original yod before BW)W, but there is no
evidence for such a yod. Looking at "The Book of Jeremiah The Hebrew
University Bible", it seems that the DSS cannot help. Of Hebrew witnesses
we find the infinitive B) in the Mechilta and the QATAL  BW)W in the Geniza
fragment G-B Ebb 22. Theodotion has an infinitive aorist passive, Targum
Onkelos has a YIQTOL and the Vulgate has a present conjunctive.  So I see
no textual reason to emend the consonant text. If we smooth away those
passages that represent difficulties from the point of view of our grammar
by textual emendation, we will never arrive at a grammar that can account
for all the data of the corpus.

We all agree that the form BW)W in an unpointed text must be either a QATAL
or an imperative. Further do we agree that an imperative is syntactically
impossible. Therefore, if we only take the unpointed text into account and
refrain from textual emendation, the form *must* be a QATAL.  The Masoretic
pointing suggests that the form be an imperative, which is impossible;
therefore, either the Masoretes pointed it "wrongly" (a much smaller
"emendation" than to add a consonant), or the pointing is a variant of the
normal pointing of QATAL, which some grammarians believe. Behind this
reasoning is no grammatical assumption, only the assumption that normal
syntax is valid. I found point 5.2.4 "The Perfect Aspect and Modality" in
Galia's thesis interesting. And I would like to add that there are hundreds
of other QATALs with present or future meaning.


Rolf Furuli
University of Oslo

More information about the b-hebrew mailing list