From: David Kummerow <farmerjoeblo AT hotmail.com>
To: b-hebrew AT lists.ibiblio.org
Subject: [b-hebrew] Overview and comments on Furuli, A New Understanding of the Verbal System of Classical Hebrew
Date: Tue, 20 Mar 2007 14:55:01 +1100
Since Rolf's dissertation is figuring prominently in current discussions
on this forum, over the last two days I decided to actually read it.
Based on this reading, I offer some reflective comments for those who
don't currently have access.
The dissertation is organised basically into nine parts. Part one covers
previous research on the BH verbal system under two headings: "theories
viewing waw as more than a conjunction" and "theories viewing the waw as
a normal conjunction".
Part two covers "methodology, definitions, and text" where Rolf
introduces methodology of making a "scrupulous distinction between
semantic meaning (uncancellable meaning) and conversational pragmatic
implicature (cancellable meaning)" (p. 47; also 46,n.43, 49, 70, et
passim). Here he develops a model of tense and aspect, based upon
interaction with Olsen's monograph and English data, which he suggests
is applicable for the analysis of the BH verbal system.
In part three Rolf tackles the question of "when did classical Hebrew
get four conjunctions?" His discussion here moves from a discussion of a
diachronic approach vis-a-vis a synchronic approach, where he sides with
a synchronic approach based upon his assessment that he has detected no
change in meaning between early, classical, and late Hebrew (both that
of the Hebrew Bible as well as the DSS and Ben Sira), to questioning the
basis for the supposed basis for wayyiqtol being an old short prefix
verb. Here the discussion centres upon the relevance of Akkadian, Amarna
Canaanite, Ugaritic, Phoenician and Punic, Aramaic, and Proto-Semitic.
Rolf's method here is a) demonstrating how previous scholarship has not
made a "scrupulous distinction" between "past tense" and "past
reference" and "cancellable meaning" and "uncancellable meaning"; and b)
demonstrating inconsistent labels used by scholars and the debate
between them. The aim here seems to me to endeavour to muddy the waters
and attempt to paint a deliberately fuzzy picture so that it can be
claimed that BH wayyiqtol has no connection. The discussion then moves
to the Masoretes and here the contention is made that the Masoretes are
the inventors of wayyiqtol based upon a) the lack of evidence for
wayyiqtol in the DSS; and b) the lack of evidence for wayyiqtol in the
Hexapla. From here, Rolf addresses the issue of "a semantic difference
between wayyiqtols and weyiqtols in the MT", discussing the related
issues of Masoretic orthography, apocopation, and modals. Finally,
Masoretic pointing is discussed in relation to the Karaites and
Masoretes. I note here in passing, since I won't mention the point
below, that although Rolf presents evidence that apocopation is somewhat
inconsistent---which for him points to non-semantic meaning
distinctions---his own view of phonological factors is still unable to
account for the inconsistency.
Part four covers the use of the infinitives (construct and absolute),
with some discussion of the active and passive participle.
Part five is a discussion of yiqtol: non-sentence-initial with past
reference; sentence-initial; clusters in prose; common verb roots with
yiqtol and wayyiqtol as indicating "that yiqtol and wayyiqtol belong to
the same conjugation"; semelfactive yiqtols with past reference; telic
yiqtols with past reference (particularly in relation to adverbs); and,
finally, "pre-past and present completed events expressed by yiqtols."
Part six addresses wayyiqtol: the problem of non-consecutive wayyiqtol;
the way(y) prefix; imperfective wayyiqtol; and "the intersection of
event time by reference time in wayyiqtols with past reference". The
latter is achieved by referal to specific texts, viz. Gen 19, 41; Josh
5; 2 Sam 15; and the infinitive absolute in the Karatepe inscription.
Part seven moves to a discussion of qatal by way of an initial
comparison of qatals and yiqtols with past reference and qatals with
present reference. Part seven then moves to the so-called "prophetic
perfect" (qatals with future reference), which is dealt with
consecutively: future qatals in prose texts; future qatals in poetic
texts; and future qatals in prophetic texts. The analysis of qatal then
moves to weqatal, where the related issues of "waw-relative",
"waw-copulative", and stress patterns are dealt with. The conclusion
here is that qatal cannot be prised apart into two separate
grammaticalised verbs, qatal and weqatal, but are better taken to be a
single conjugation. Prophecies with qatal and weqatal are then discussed.
Part eight is a chapter on "linguistic convention and the use of verbs":
firstly, "the linguistic background for a similar use of different
forms"; secondly, "situations with similar or almost similar use of the
[verb] forms" (eg, poetic parallelism, etc); thirdly, "situations where
the use of the imperfective aspect is significant" ("resultative
situations where the action ended some time in the past"; "one event or
state intersected by another event"; "conative and ingressive events");
and finally, the "linguistic conventions and patterns", ie, "the broad
patterns of the finite [verb] forms" and "the logic of this model from a
linguistic point of view."
Part nine presents a summary of the dissertation's findings with some
further comments, viz. "the explanatory power of the [aspectual]
definitions", "the application of the conclusions in practical work",
and "the Hebrew verbal system and the cognate languages."
Now for a few of my personal reflections on some of the dissertation, in
no real particular order:
1. I found it quite interesting that Rolf admits to the basics of
prototype theory, even though he himself doesn't use this term. He says
in relation to lexical meaning: "It seems that psycholinguistics are
correct when they say that the central part, or the nucleus, of such a
concept tends to be relatively clear, but the concept becomes more dim
or fuzzy the further away from the nucleus we proceed" (p. 28). In other
words, Rolf admits to a) prototypical meaning on the one hand and b)
fuzzy meaning on the other. The consequence of this in general
linguistics is commonly seen in the admittance of continuums and
multifunctionality. Now my opinion is that the dissertation here has not
fully come to grips with what it had admitted to in principal. The
advanced methodology---i.e., the scrupulous distinction between
cancellable meaning and uncancellable meaning---in reality admits no
continuum or fuzziness: either something is "cancellable" or it is
"uncancellable"; something cannot exist between these opposites.
Further, the grammaticalisation approach is dismissed because
"diachronic data are ambiguous" (it is also dismissed because proponents
are claimed to not have differentiated between "past tense" and "past
reference") (p. 32)---yet this is not surprising if one admits to
fuzziness in language. As such, the advanced methodology is in reality
unable to cope with fuzziness, i.e. incomplete grammaticalisation,
because anything that is incompletely grammaticalised will be neither
"cancellable" nor "uncancellable"---yet the basic premise is something
can only be one or the other!
2. Further on grammaticalisation, the claim of the dissertation is that
after an exhaustive analysis of wayyiqtol that evidence supporting the
grammaticalisation of wayyiqtol is unable to be found. However,
crucially, the dissertation has failed to investigate avenues where such
grammaticalisation is possibly seen, viz. the inclusion of the suffix
verb in a paradigm originally dominated by prefix verbs.
3. An interesting point is made that "[t]o demonstrate that wayyiqtol is
a semantically independent conjunction, one has to show ... that the
widespread use of wayyiqtol with past reference is due to the semantic
meaning of the form, and not just a linguistic convention" (p.48). To my
mind, this is nonsense: linguistic convention (i.e., the regular choice
of speakers to use similar linguistic tokens) defines meaning. In other
words, the entrenchment of meaning is related to linguistic frequency.
The prising apart here of "semantic meaning" and "linguistic convention"
means that much ink is spilled in the dissertation upon a fruitless
quest for contexts in which "uncancellable semantic meaning" may be
ascertained.
4. Areas of research which quite potentially have much bearing on
aspects of discussion are neglected: a) the distribution of qatal and
yiqtol with temporal adverbs "yesterday" and "tomorrow"; b) in
conjunction with the present-tense uses of (we)qatal no investigation is
made of cosubordination, politeness, gnomics, and performatives; c) the
distribution of paragogic nun; d) the distribution of third-person
pronominal suffixes augmented with nun; e) the possibility of
exaggerated futures with qatal (presumably because this is in conflict
with the advanced methodology); and f) default use of qatal in
non-paratactic constructions being anterior. Now in one sense it does
not matter what we label a verb so long as we accurately describe its
range of use. However, it is useful for labelling to reflect
prototypical function as some sort of mnemonic. The debate, then, is
over what is seen to be the prototypical use of the different BH verb
conjugations. The trouble as I see it with the dissertation is that
because of its methodology of finding uncancellable meaning, it is
unable to convincingly demonstrate prototypical meaning. The reason is
that the BH verbs, as indeed verbs, words, etc in other languages, are
multifunctional, i.e. they have more than one function. This is
particularly so with qatal, which may be seen as an extreme case of
multifunctionality. Multifunctionality basically implies incomplete
grammaticalisation and fuzziness. However, what if that multifunctional
fuzziness is essentially, as is the case with qatal, the multifunction
of a verbal conjugation that can be (construnctionally?) used for
anterior, performatives, gnomics, hypothetical/conditional, exaggerated
futures, etc? That is, there is a fuzziness to the range of uses which a
methodology that starts with the premise that meaning solely falls into
either (the non-fuzzy categories) "cancellable meaning" and
"uncancellable meaning" is unable to describe or relate to. And a
related fault, then, is that linguistic convention is seen to have no
bearing on semantics. (I guess I should point out that the debate
concerning the prototypical function of the BH verbal conjugations could
similarly be had over other fuzzy areas of language: function words
[esp. when not completely grammaticalised], the meaning of lexical
items, etc etc etc [the list is basically endless due to the fact that
speakers are unable to have an exhaustive inventory of semantic and
pragmatic meanings upon which to draw: generalisation, economy,
polysemy, and conceptual grouping are central aspects of language]. As
such, I am unconvinced that the methodology rigorously followed in the
dissertation is able to produce fruitful results in existent linguistic
multifunctionality and if the area of investigation exhibits incomplete
grammaticalisation.
Hope this helps!
Sincerely,
David Kummerow.
[b-hebrew] Overview and comments on Furuli, A New Understanding of the Verbal System of Classical Hebrew,
David Kummerow, 03/19/2007