Reconstructing the Pronunciation of Biblical Hebrew Epigraphic Hebrew

Edition 2.0

12 December 2011

A Note on Epigraphic Hebrew

by David Steinberg

The recovery of the phonetic shapes of Hebrew words attested in inscriptions from the biblical period is often confounded by uncritical retrojection of Masoretic orthoepy onto ancient spellings.

Quoted from Andersen 1999 p. 5.

Epigraphic Hebrew (EH) is the term used for Hebrew inscriptions, largely from the pre-exilic Kingdom of Judah mostly dating from the late eighth century BCE until the end of the kingdom in 586 BCE. It is probable that Judahite Epigraphic Hebrew (JEH) represents a sub-literary governmental administrative register of Judahite Hebrew while CBH represents its literary register. Linguistically, JEH is rather closer to CBH than to PCBH though it has a number of features shared with neither[1]. Scribes trained in Jerusalem 730-586 BCE were likely the authors of the bulk of surviving JEH e.g. Siloam Inscription, Lachish ostraca, Arad ostraca. The same circles were likely the composers and/or transmitters of most of the pre-exilic biblical texts.

We are fortunate in having a fine grammar cum lexicon of the corpus of EH (JEH and IEH) inscriptions as well as a number of scholarly collections of the material and a reconstructed vocalization[2]. JEH documents have been preserved in their original language and orthography and, within limits, can serve as a guide to the original orthography of CBH.

On the negative side, the fact that scholars of the highest caliber such as (in alphabetical order) Anderson, Blau, Barr, Muraoka, Pardee, Rainey, Richter and Sarfatti are in contention about aspects of the vocalization of Epigraphic Hebrew makes it clear that the information required to make definitive decisions about areas of dispute is simply not there. On the one hand, recreating the phonetics of dead languages is impossible beyond a certain point. This is compounded by and the lack of almost any internal vowel letters in EH (see Matres Lectionis in Hebrew and Matres Lectionis in the Biblical Text). On the other hand, the extreme paucity of epigraphic materials found to date means that we are working with a miniscule basis of written evidence. This contrasts with, for example, Latin and Akkadian scholars who have mountains of vocalized epigraphic remains.

Two issues in the vocalization of JEH, and hence of EBHP, have occupied me:

· whether a case could be made that the historic diphthongs written in JEH <w> and <y> could have already contracted to [ô] and [ê] respectively? and,

· what, if any final vowels were not represented by vowel letters?

· the pronunciation of the 3ms. pronominal suffix written <h> in JEH.

a. See Did Word-Final Short Vowels Exist in EBHP and Were All Word-Final Vowels Marked by Vowel Letters?

b. See Heterogeneous Diphthong Contraction

Matres Lectionis in JEH
3. The system (of the use of vowel letters in JEH) was briefly as follows: (a) waw for û[3]. (b) yod for î. (c) he for â. In the final position ē and ō were also represented by he. While the case for these equations is reasonable, it has also been claimed that waw is used for ô and yod for ê. This argument is based entirely on the contention that the diphthongs aw and ay had been contracted with retention of the original consonants as vowel letters, i.e., historical spelling. Ultimately, contractions occurred but at different times in different dialects of NW Semitic and there is no unequivocal evidence for it in the early period, apart from Phoenician. 4. The system of internal vowel letters was more restricted since only waw and yod were pressed into service for such vowels: (a) waw represented û. (b) yod represented î. Naturally they could also represent consonants and diphthongs: the latter have to be regarded as consonants until and unless contracted, when they became long vowels. All remaining vowels were not indicated at this stage of development: ā was left unrepresented because it was not considered feasible to use he internally, so in effect ā was relegated to the zero-classification like the short vowels. As for ō and ē, their time would come, but so far we have no decisive evidence for their representation in this period. What is needed is the occurrence of non-etymological vowel letters that were not part of the consonantal structure of the word and whose vocalization is equally certain.... For Hebrew orthography... the official system called for the representation of all final vowels by appropriate vowel letters (h, w, y). All known Hebrew inscriptions from the 9th to the 6th centuries follow this pattern.... The dubious or questionable cases are the result of the misapplication of Masoretic vocalizations to the inscriptional material.... When it comes to internal vowel letters it is clear that they were used, but before the Exile the practice was sporadic at best. Such vowel letters are attested as early as the 8th century and their use may go back further. Unequivocal examples are found in proper nouns and in distinctive verbal (and nominal) forms where the function and probably the purpose are unmistakable. For example: o ʾrwr = ʾarūr, cursed, in the Shebna inscription (8th cent.) o hbqyd = hibqῑd, he appointed, from Arad (7th-6th cent.) But systematic use is unattested in any Hebrew inscriptions or those of closely related dialects. Quoted from Freedman 1992 pp. 6-7.

Matres Lectionis in JEH

3. The system (of the use of vowel letters in JEH) was briefly as follows:

(a) waw for û[3].

(b) yod for î.

In the final position ē and ō were also represented by he. While the case for these equations is reasonable, it has also been claimed that waw is used for ô and yod for ê. This argument is based entirely on the contention that the diphthongs aw and ay had been contracted with retention of the original consonants as vowel letters, i.e., historical spelling. Ultimately, contractions occurred but at different times in different dialects of NW Semitic and there is no unequivocal evidence for it in the early period, apart from Phoenician.

4. The system of internal vowel letters was more restricted since only waw and yod were pressed into service for such vowels:

(a) waw represented û.

(b) yod represented î.

Naturally they could also represent consonants and diphthongs: the latter have to be regarded as consonants until and unless contracted, when they became long vowels.

All remaining vowels were not indicated at this stage of development: ā was left unrepresented because it was not considered feasible to use he internally, so in effect ā was relegated to the zero-classification like the short vowels. As for ō and ē, their time would come, but so far we have no decisive evidence for their representation in this period. What is needed is the occurrence of non-etymological vowel letters that were not part of the consonantal structure of the word and whose vocalization is equally certain....

For Hebrew orthography... the official system called for the representation of all final vowels by appropriate vowel letters (h, w, y). All known Hebrew inscriptions from the 9th to the 6th centuries follow this pattern.... The dubious or questionable cases are the result of the misapplication of Masoretic vocalizations to the inscriptional material....

When it comes to internal vowel letters it is clear that they were used, but before the Exile the practice was sporadic at best. Such vowel letters are attested as early as the 8th century and their use may go back further. Unequivocal examples are found in proper nouns and in distinctive verbal (and nominal) forms where the function and probably the purpose are unmistakable. For example:

o ʾrwr = ʾarūr, cursed, in the Shebna inscription (8th cent.)

o hbqyd = hibqῑd, he appointed, from Arad (7th-6th cent.)

But systematic use is unattested in any Hebrew inscriptions or those of closely related dialects.

Quoted from Freedman 1992 pp. 6-7.

We should note the JEH: second person masculine singular suffix on singular noun <k>[4]; SC second person masculine singular suffix <t>[5] and, perhaps the second person ms. independent pronoun <ʾ[t]>[6]. See the discussion on anceps vowels. We can start with the reasonable assumption that JEP and PreExH are two closely related registers of Hebrew written in the same scribal community at the same period. This together with the evidence of BH_Qum[7] and TH renders it most probably that these three forms, in JEH end in unwritten vowels i.e. <k> = */ka(ː)/; <t> = */ta(ː)/; *<ʾt> = */ʾàta(ː)/.