What a Difference a Year Makes: Can Biblical Texts Be Dated

Linguistically?

Ziony Zevit

Hebrew Studies, Volume 47, 2006, pp. 83-91 (Article)

Published by National Association of Professors of Hebrew

DOI: https://doi.org/10.1353/hbr.2006.0013

For additional information about this article

https://muse.jhu.edu/article/439264/summary

Access provided at 26 Oct 2019 04:02 GMT from Stockholms universitet

 WHAT A DIFFERENCE A YEAR MAKES:
CAN BIBLICAL TEXTS BE DATED LINGUISTICALLY?*

Ziony Zevit
University of Judaism

Language played almost no role in dating biblical texts in historical-

critical scholarship prior to the middle of the twentieth century. Texts such
as Esther, Ezra, Nehemiah, Amos, Isaiah of Jerusalem, Second Isaiah, and
the like were dated according to explicit references in their contents. Texts
lacking such references were dated according to their supposedly implicit
references to externally datable events. Others were dated by their theology
or attitude towards the cult as determined by a chronological scale linked to
some form of the documentary hypothesis as expounded by Julius
Wellhausen or Abraham Kuenen, both of whom accepted a late, post-exilic
date for the Priestly source as propounded by Karl Heinrich Graf. This scale
was based on Enlightenment notions about the evolution of religion in gen-
eral and Israelite religion in particular as a stage between primitive religion
and Christianity. When first propounded, these evolutionary ideas were
well-founded and tightly linked to historical theory of the late nineteenth
and early twentieth century. But that clear theory with its chronological im-
plications was confounded after the implications of discoveries in the an-
cient Near East were acknowledged by biblicists. The old critical consensus
about the chronological dating of developmental stages in Israelite religion
and of biblical books became vulnerable and challengeable. Some objective
method was required to date the literature.
The study of language history evolved in Europe during the nineteenth
century and was able to mark impressive advances by the beginning of the
twentieth century. The study of Hebrew lagged behind because there was no
concentrated body of scholars interested in it. A “tipping point” was reached
in 1925 with the establishment of the Hebrew University in Jerusalem.
Since then, the discipline has attracted a few hundred researchers, developed
a large body of competent, reliable literature, and has revealed unsurpris-
ingly that Hebrew is a language with a history, just like other languages.
One significant discovery of significance to our panel today was that

*
These introductory remarks were delivered at the NAPH session that met during the 2005 SBL meeting.
The collection of articles from the first symposium may be found in HS 46 (2005), pp. 321–376.
Hebrew Studies 47 (2006) 84 Zevit: What a Difference

Hebrew remained a spoken language and mother tongue until the second
century C.E.
In 2005, historians of Hebrew were aware that in addition to the Hebrew
of the Iron Age, Neo-Babylonian, Persian, and early Hellenistic periods at-
tested in the Bible and inscriptions, there was Hebrew of the Greco-Roman
periods, described variously as Mishnaic or Tannaitic Hebrew, within which
dialects have been discerned that are classified both diachronically and re-
gionally. The dialect of the sectarian Qumran scrolls belongs to this broad
chronological horizon as does that of the Samaritan Pentateuch. From later
periods we have Medieval Hebrew in various dialects characterized both by
date (early, middle, late) and geography (Palestine, Babylon, Syria and
Egypt, and Europe), Enlightenment (Haskalah) Hebrew, and Modern Israeli
Hebrew. Viewed diachronically, each definable later stage emerged from an
earlier stage, preserving much and changing some of the inherited features
as well as innovating on the basis of inner-Hebrew catalysts and external
linguistic factors. But what of pre-Mishnaic-Tannaitic-Dead Sea Scroll
Hebrew?
Proto-Hebrew evolved from one or more of the West-Semitic Canaanite
dialects of the late fourteenth century B.C.E. described by Anson Rainey in
his four-volume, 1996 study of the dialects underlying the Amarna tablets,
and may be of Transjordanian origin.1 Rainey’s work, provides us only
with a terminus ante quem. Randy Garr’s 1985 work helps locate and define
individuated Hebrew dialects within the Syria-Palestine dialect continuum
of the tenth-sixth centuries.2 Garr’s discussion of phonology may be ex-
panded somewhat in the light of Robert Woodhouse’s 2003 study of sibi-
lants and of Richard C. Steiner’s innovative 2005 study dating how
phonemes represented by the graphemes fiayin and h.et were rendered in
various Hebrew dialects of the Greco-Roman period.3
Dating biblical books is not quite the same as dating the Hebrew of the
biblical period in which these books, or parts thereof, were written. Early-
on, scholars learned that the books from Samuel through Kings were repre-

1
A. F. Rainey, Canaanite in the Amarna Tablets: A Linguistic Analysis of the Mixed Dialect Used by the
Scribes from Canaan (Leiden: Brill, 1996). See in particular vol. 2.
2
W. R. Garr, Dialect Geography of Syria-Palestine 1000–586 B.C.E. (Pennsylvania: University of
Pennsylvania Press, 1985. Repr., Winona Lake, Ind.: Eisenbrauns, 2004), pp. 205–240. See also, B.
Halpern, “Dialect Distribution in Canaan and the Deir Alla Inscriptions,” in Working With No Data:
Semitic and Egyptian Studies Presented to Thomas O. Lambdin, ed. D. M. Golomb (Winona Lake, Ind.:
Eisenbrauns, 1987), pp. 119–139.
3
R. Woodhouse, “The Biblical Shibboleth Story in the Light of Late Egyptian Perceptions of Semitic
Sibilants: Reconciling Divergent Views,” JAOS 123 (2003): 271–289; R. C. Steiner, “On the Dating of
Hebrew Sound Changes… and Greek Translations…,” JBL 124:2 (2005): 229–267.
Hebrew Studies 47 (2006) 85 Zevit: What a Difference

sentative of typical or Classical Biblical Hebrew of the First Temple, and

used them to describe the lexicon and grammar of Classical Hebrew of the
First Temple period. This determination was influenced both by the con-
tents of the books as well as the language of the Siloam and Mesha inscrip-
tions. When compared with this Hebrew, scholars easily identified
distinctive linguistic characteristics in books self-identified as products of
the Persian period: Ezra, Nehemiah, and Esther. Noticing the Aramaic in-
fluence on the vocabulary and grammar of these books, a linguistic profile
of Late Hebrew was developed, enabling scholars to date some books by
their language. Consequently, Chronicles, Ecclesiastes, and Canticles were
also assigned to the Persian period. Scholars familiar with Mishnaic/
Tannaitic Hebrew recognized a precursor to the later Hebrew of the first to
second centuries C.E. in the Hebrew of these Persian period sources. Noth-
ing in these particular determinations conflicted seriously with conclusions
of scholars who approached the Bible through its literature and theology.
Biblicists, however, particularly in Israel, were debating the validity of
the Wellhause-Kuenen-Graf hypothesis that understood the Pentateuchal
documents to have evolved in the following order: JED (ca. 622 B.C.E.) + P
(post-586, sometime). Following Yehezkel Kaufman who maintained a
strong, pre-Grafian form of the documentary hypothesis, many argued that
the proper order should be J, E, P, D (622 B.C.E.)—all pre-586 B.C.E.
Their argumentation was primarily literary and (historical-)theological, not
linguistic, lacking some sort of hard data to back it up. The question of P’s
date was crucial to their anti-consensus move. On the one hand, P’s Hebrew
was clearly not Mishnaic/Tannaitic and not Persian period Hebrew. But that
did not mean that it was not post-586.
In a series of researches over thirty years, one of today’s panelists Avi
Hurvitz, advancing techniques used by Yehezkel Kutscher and Haim Rabin
of the Hebrew University, sought linguistic criteria useful for distinguishing
different historical phases of Biblical Hebrew. He settled primarily on a
contrastive analysis of lexical items and some syntax in order to characterize
the language of some compositions as Late Biblical Hebrew = Persian
Period Hebrew = Post-exilic Hebrew; Transitional Hebrew = Neo-
Babylonian = Exilic. His work involved but was not limited to semantics.
Hurvitz has treated selected texts—Pss 103, 117, 119, 124, 125, 144,
145, and the doxologies at the end of 41, 72, 106—on an ad hoc basis cau-
tiously. His work, oriented more to the practical rather than the theoretical,
has always remained humble in its claims. He has not generalized beyond
the evidence that he painstakingly presents. The big pay-off of his many ar-
Hebrew Studies 47 (2006) 86 Zevit: What a Difference

ticles came when he used years of accumulated data to demonstrate that the
Hebrew in Ezekiel, a book dated internally to the early sixth century B.C.E.,
was transitional and that the Hebrew of P was earlier than that of Ezekiel.
This latter conclusion had literary support in that Ezekiel 40–48 polemicises
against many of P’s cultic prescriptions.
Complementing Hurvitz’s early lexicographical work was a single, am-
bitious study by Robert Polzin. Polzin proposed a profile of Late Biblical
Hebrew in narrative prose consisting of nineteen phonological, morphologi-
cal, and syntactic features of Chronicles, absent in Classical Pre-exilic
Hebrew, most of which could not be attributed to Aramaic influence.
Consequently, he concluded, they were manifestations of inner Hebrew de-
velopment. By comparing the frequency of these features in P, Ezra, and
Nehemiah, Polzin created a scale along which these different compositions
could be placed in terms of how congruent and incongruent they were with
Chronicles and with each other. Decreasing congruency was the mark of an
earlier text. He discovered that JE, Dtr, and the Court History, what he
called “Classical Hebrew,” were relatively congruent with each other but
distinctively different from Chronicles. Chronicles, Ezra, and one part of
Nehemiah were relatively free of archaisms while a different part of
Nehemiah, along with Esther and Daniel, were characterized by many
archaisms and deemed later than Chronicles. Pg, the basic P corpus minus
secondary additions (labeled by him Ps), was later than Classical Hebrew
but earlier than that of Chronicles.4
Although Polzin’s basic profile required corrections—a not insignificant
matter since it is used to translate degrees of congruence/incongruence into
degrees of relative chronological separation—it has proven useful.5
Whereas Haggai and Zechariah self-present as post-Exilic texts of the early
Persian period, about 520 B .C .E., Hurvitz’s method, with its focus on lexi-
con, seems unable to demonstrate their lateness. Applying Polzin’s method-
ology and a slightly adjusted profile, A. E. Hill argued that Haggai,
Zechariah 1–8, Zechariah 10–14, and Malachi are congruent statistically
with Polzin’s figures for the Pg corpus, but not with his figures for
Chronicles and Ezra.6

4
R. Polzin, Late Biblical Hebrew: Toward an Historical Typology of Biblical Hebrew Prose (Missoula,
Mont.: Scholars, 1976).
5
See the corrections introduced in Z. Zevit, “Converging Lines of Evidence Bearing on the Date of P,”
ZAW 94 (1982): 494–501. For the convergence of an analysis combining Hurvitz and Polzin (as corrected
by Zevit), see p. 501.
6
A. E. Hill, “Dating Second Zechariah: A Linguistic Reexamination,” HAR 6 (1982): 105–134. For the
chart, see p. 131; for the qualifications, see p. 114; A. E. Hill, “Dating the Book of Malachi: A Linguistic
Hebrew Studies 47 (2006) 87 Zevit: What a Difference

Ultimately, what should result from work conducted along the lines de-
veloped by Hurvitz and Polzin is a diachronic grammar and lexicon of the
Hebrew language during the Iron Age, Neo-Babylonian period, and Persian
Periods. But scholarship is far from there.
There are a few problems with this approach that have not been ad-
dressed in the past. One is reflected in Richard M. Wright’s recent study,
Linguistic Evidence for the Pre-Exilic Date of the Yahwistic Source, 2005.7
The book’s forthright methodology is based on Hurvitz’s work. It is prob-
lematic in that it sets out to argue for a pre-exilic J on linguistic grounds
when the data in J have always been reckoned as pre-exilic in earlier stud-
ies. Consequently, Wright, worked with conclusions from studies that as-
sumed exactly what he claimed to be proving.8 Although he studied new
data, his major argument was largely circular and his conclusions were
foreordained by his methodological and historical assumptions.
While Polzin’s work has been largely ignored, many students of the
history of Hebrew have attempted to tie loose ends in the Hebrew data not
addressed by Hurvitz that pose apparent, though not necessarily real prob-
lems to his conclusions. These have dealt with problems for diachronic
studies posed by (1) residues from regional dialects that might be repre-
sented in narratives, cited speech, or certain poems, (2) the deployment of
different linguistic registers and narrative styles in different compositions,
and (3) the linguistic conventions of particular genres. Each of these topics
has been the focus of both synchronic and diachronic investigation, some by
members of our panel today: Frank Polak and Gary Rendsberg.
Studies addressing these problems were edited by Ian Young in an im-
portant collection, Biblical Hebrew: Studies in Chronology and Typology,
2003.9 They pose theoretical and practical problems for the enterprise of
dating that must be addressed.
Other students have questioned both the interpretations of data examined
by Hurvitz and Polzin and their a priori assumption of what may be consid-

Reexamination,” in The Word of the Lord Shall Go Forth: Essays in Honor of David Noel Freedman in
Celebration of His Sixtieth Birthday, ed. C. L. Meyers and M. O’Connor (Winona Lake, Ind.:
Eisenbrauns, 1983), pp. 77–89. For the chart, see p. 85.
7
R. M. Wright, Linguistic Evidence for the Pre-Exilic Date of the Yahwistic Source (London: T & T
Clark, 2005).
8
An additional minor problem is that he accepts Third Isaiah, i.e., Isaiah 56–66, and Jonah as post-exilic
texts on non-linguistic grounds even though the books do not self-identify themselves as such (R. M.
Wright, Linguistic Evidence, pp. 11–12).
9
I. Young, ed., Biblical Hebrew: Studies in Chronology and Typology (JSOTSup 369; Edinburgh: T & T
Clark, 2003).
Hebrew Studies 47 (2006) 88 Zevit: What a Difference

ered pre-exilic, Iron Age. Studies along these lines were also published by
Young in his volume, including one by panel member Martin Ehrensvärd.
Much of the material and most of the ideas in Young’s volume have en-
joyed two major evaluations. One in my review of the book published in
both the online and print versions of RBL in 2004; and another by panelist
Jens Kofoed, in the third chapter of his Text & History, published in 2005.10
Despite the date of publication, Kofoed’s book was written before Young’s
book came out, but it engages all key issues on the basis of articles that
appeared earlier.
My review of Young’s book concluded with a positive assessment of the
challenges to the type of work carried on under the methodologies advanced
by Hurvitz and Polzin because they were probing not only methodologies,
but also criticizing many specific instances of their application and claiming
to have discerned errors. In my estimation, although the critiques did un-
dermine some specific conclusions, they did not undermine the historical
conclusions of the research.
Kofoed’s conclusions regarding the possibility of dating are the
following:

1) Aramaisms in Southern texts point to a date of composition after

Imperial Aramaic influenced the language of Palestine in the sixth
to fifth centuries B.C.E.

2) Although the absence of Aramaisms point to an earlier date of

composition before the sixth to fifth centuries, “we cannot say how
much earlier.”

3) There are no objective criteria that indicate when texts composed

in the Northern dialect of Hebrew were composed. Although texts
such as Hosea and Elijah/Elisha stories most likely ante-date the
composition of Kings, on objective, linguistic grounds, “we are
still unable to go further back than before the sixth century B.C.E.”

4) Many of the lexical items used to date texts on the basis of Hebrew
vocabulary “have proved inconclusive” though some appear
reliable.

5) A distinction may be made between the Hebrew of late books such

as Ezra and Nehemiah and of earlier books, but the time of the

10
J. B. Kofoed, Text and History: Historiography and the Study of the Biblical Text (Winona Lake, Ind.:
Eisenbrauns, 2005), pp. 113–163. See also pp. 79–80.
Hebrew Studies 47 (2006) 89 Zevit: What a Difference

transition from one to the other during the Persian period has not
been established.

He concludes that linguistic change is a relatively soft yardstick for dating

texts.11
None of the recent discussions have seriously engaged orthography as it
reflects historical phonology or phonological development in Hebrew. The
only study of which I am aware that addresses this aspect of historical lin-
guistics is by Saul Levin, “The Hebrew of the Pentateuch,” published in
1988. Although Levin’s article is in need of updating with regard to the
quantity—not quality—of the data, his conclusions remain valid within the
Hurvitz-Polzin paradigm but problematic for its detractors who must then
assume not only two styles and registers of Hebrew functioning in parallel
during the post-exilic period, but also different ways of pronouncing “off-
glides.”12
***

A number of publications scheduled to appear during the 2006–2007

academic year will present new data challenging the notion that what is
called Classical Biblical Hebrew ceased to be written in the middle of the
sixth century are the following:

1) An article, “The ‘Late’ Vocabulary of the Book of Chronicles,” by

Robert Rezetko is scheduled to appear in 2006 in Studies in Histo-
riographic Tradition: Essays in Honour of A. Graeme Auld, ed. B.
Aucker, T. Lim, and R. Rezetko (Leiden: Brill). This article will
expand on the vocabulary section of Rezetko’s article in the vol-
ume edited by Ian Young, arguing that what is considered ‘Late
Hebrew’ is not necessarily post-exilic.

2) Additionally, Rezetko’s dissertation, “Source and Revision in the

Narratives of David’s Transfer of the Ark: Text, Language and
Story in 2 Samuel 6 and 1 Chronicles 13” is scheduled for publi-
cation by T & T Clark as a JSOT Supplement, in 2006 or 2007.
Judging from the data that Rezetko cited from this work in his
contribution to the Ian Young volume, and from the praise lav-

11
J. B. Kofoed, Text & History, pp. 162–163.
12
S. Levin, “The Hebrew of the Pentateuch,” in Fucus: A Semitic Afrasian Gathering in Remembrance of
Albert Ehrman, ed. Y. Arbeitman (Philadelphia: J. Benjamins Publishing, 1988), pp. 291–323. His data
have to be reviewed in light of F. I. Andersen and A. Dean Forbes, Spelling in the Hebrew Bible (Rome:
PBI, 1986). Levin was unaware of L. A. Bange’s A Study in the Use of Vowel-Letters in Alphabetic
Consonantal Writing (Munich: Uni-Druk, 1971) which also worked with the notion of off-glides.
Hebrew Studies 47 (2006) 90 Zevit: What a Difference

ished on it by those who have read it, this will be a formidable

contribution to the discourse.

3) A dissertation by Robyn Vern, on the significance of what is con-

ventionally recognized as Archaic Biblical Hebrew for dating is
scheduled for completion sometime in 2006.

4) And last, but not least, Ian Young and Robert Rezetko are writing
a book entitled Linguistic Dating of Biblical Texts: An Introduc-
tion to Approaches and Problems, that will be published by Equi-
nox Publishing in 2007. As described to me by Ian Young in an
email on August 23, 2005, this will be a forceful marshalling of
both theoretical arguments and case studies advancing the idea that
conventional linguistic dating is incorrect. It will attempt to coor-
dinate the different arguments of the various scholars assuming the
contra position. Furthermore, by the time of its publication, this
book will engage the strictures and objections raised against this
position by some of the authors in the Young volume, by some of
those who reviewed the book and found it wanting, by the lecturers
in this program last year, by this years’ panel, and by recent
articles that have appeared in scholarly journals.

I await the publication of this book because I anticipate that it will be a

‘best shot’ type of publication, one that may well define the discourse,
should one ensue, through the end of this decade. Among the targets on
which it will have to focus are data and arguments based on them presented
in the proceedings of an international research group that met at the
Institute for Advanced Study of the Hebrew University of Jerusalem in
2001–2002. These now appear in a 2006 volume edited by Steven E.
Fassberg and Avi Hurvitz, Biblical Hebrew in Its Northwest Semitic Setting:
Typological and Historical Perspectives published by Eisenbrauns. (This
collection may be considered a companion volume to the 2003 collection of
articles edited by I. Young discussed above.) Additionally, it will have to
consider the more pointed disagreements and attacks by scholars whose arti-
cles appear in last year’s “Symposium: Can Biblical Texts Be Dated Lin-
guistically?” (HS 46 [2005]) and this year’s continuation in the studies that
follow.
Hebrew Studies 47 (2006) 91 Zevit: What a Difference

The Symposium includes:

1. Using Linguistic Difference in Relative Text Dating: Insights from

Other Historical Linguistic Case Studies (Jens Bruun Kofoed)
2. Sociolinguistics, A Key to the Typology and the Social Background
of Biblical Hebrew (Frank Polak)
3. Aramaic-Like Features in the Pentateuch (Gary Rendsburg)
4. Why Biblical Texts Cannot be Dated Linguistically (Martin
Ehrensvärd)
5. The Recent Debate on Late Biblical Hebrew: Solid Data, Experts’
Opinions, and Inconclusive Arguments (Avi Hurvitz)