Intermediate Classical Hebrew I: Descriptive grammatical analysis

A. In this course, we will be dealing with three principal stages of Hebrew phonology:
1) Proto-Hebrew: a reconstructed stage that will be the object of Tuesday historical grammar sessions;
had begun evolving towards the second stage by the fourteenth century BCE (evidence from the texts
discovered at Amarna in Egypt that had been sent there from Canaan).
2) Biblical Hebrew: to the extent that the details can be known, this is the Hebrew that we may derive
from two sources: a) extra-biblical Hebrew inscriptions dating to the pre-exilic period, and b) the
consonantal text to which the Massoretes attached vowels, this consonantal text being typologically
early, representing Hebrew as set down in writing in the immediate post-exilic period, roughly 500-200
BCE.
3) Massoretic Hebrew: the tradition of pronouncing Hebrew codified by the Massoretes in the vocalic
system that had been invented in the last centuries of the first millennium CE (the oldest relatively
complete manuscripts date to the late tenth century CE).
We will, of course, read from the Massoretic tradition but the explanation of Biblical Hebrew
phonology will be based on the following hypothesis: vocalic length was phonemic, i.e., meaningful, in
Biblical Hebrew (/i/ and /ī/, for example, were distinct phonemes), but vocalic length was not encoded
in the Massoretic vocalic system which represented vowel quality only (e.g., ḥireq was used to represent
ָ ‫ ִא‬ʾiššāh, ‘woman’, and /ī/, as in ‫ ִאישׁ‬ʾīš , ‘man’); the only explicit length distinction is in
both /i/, as in ‫שּׁה‬
the ḥatef-vowels (hatef-pataḥ being a form ‘reduced’ in length of normal pataḥ, etc.). The hypothesis of
phonemicity of length in Biblical Hebrew is based on the distribution of matres lectionis in the extra-
biblical inscriptions and in the consonantal text to which the Massoretic vowels were attached: the use
of the matres lectionis {y} and {w} correlates at roughly 95% with vowels that were grammatically
long, i.e., words such as ‫שּׁה‬ ָ ‫ ִא‬ʾiššāh or ‫ ֻכֹּּלה‬kullōh, ‘its entirety’, are very rarely written with a mater—this
means that the writers were hearing and representing distinct phonemes when they were quite
consistently writing /i/ without mater /y/ and /u/ without mater /w/ but grammatically long /ī/ and /ū/
frequently with the mater.
Examples to this point have referred to the rather clear cases of /i/ī/ and /u/ū/. As we shall see in the
course of the year, the problems /a/ā/ and /o/ō/ are very much more complicated. This is because (1) the
two proto-vowels /a/ā/ developed on a different pattern from the other two sets and because (2) the
vocalic evolution of the /a/ā/ set between Biblical and Massoretic Hebrew bifurcated at some point into
the so-called Sephardic and Ashkenazi traditions. In the latter, some of the secondary outcomes of /a/
and /u/ fell together and were represented by the Massoretic qāmeṣ, a vowel quality intermediate
between /a/ and /o/, hence often represented in modern transcription by a distinct sign, such as ø, å, or a
backwards c. In the Sephardic tradition, on the other hand, the original phonemes developed along
independent lines, and the Sephardic reading tradition thus distinguishes between qāmeṣ gādōl,
pronounced as /a/, which represents the /ā/ of Biblical Hebrew (as in ‫ ָחָכם‬ḥākām, ‘wise’) and qāmeṣ
qāṭōn, pronounced as /o/, which represents the evolution of /u/ to /o/ in Biblical Hebrew (as in ‫ָחְכֶמה‬
ḥokmāh, ‘wisdom’). The problems surrounding the evolution of /a/, /ā/, and /u/ will therefore be an
important part of both descriptive and historical grammar—but can only really be understood from the
latter perspective.
B. The question of the parts of speech in any given Semitic language has been much discussed (for an
overview, see §4.4.2 in B. Waltke and M. O’Connor, Introduction to Biblical Hebrew Syntax — which
will be the required grammatical reading for third quarter). I have in recent years adopted the approach
of the traditional Arab grammarians, who work with three principal categories, preferred because the
grammatical features that are overtly marked show very different profiles in the three categories:
 (1) Substantives (nouns and adjectives) are marked for gender, number, state, case, definiteness, and in
Proto-Hebrew were marked for case;
(2) Verbs are marked for person, gender, number, tense-aspect-modality (TAM), and stem (Qal, Niphal,
etc.);
(3) Particles are basically unmarked for any of the above categories, though many particles have
developed from substantives and may have vestigial gender, number, or case markers.
Thus substantives are NOT marked for person, TAM, or stem, while verbs are NOT marked for state,
definiteness or case.
This view requires that mixed categories be allowed for:
independent personal pronouns are basically substantives marked lexically for person;
infinitives, i.e., productive verbal nouns, are nominal in morphology, productive within the
verbal system (e.g., every speaker would know to generate an infinitive of the kᵊtōb type
for a strong root), and in syntax may function either a noun (e.g., kᵊtōb hassēpher, ‘the
writing of the document’—kᵊtōb is in construct with hassēpher) or as a verb (e.g. liktōb
ʾet-hassēper, ‘in order to write the document’—ʾet marks the following word as the
direct object of the verbal noun);
participles, i.e., productive verbal adjectives, are adjectival in morphology, productive within
the verbal system (e.g., every speaker would know to generate an infinitive of the kōtēb
type) and in syntax may function either a noun (e.g., kōtēb hassēper, ‘the writer of the
document’—kōtēb is in construct with hassēper) or as a verb (e.g. hūʾ kōtēb ʾet-
hassēper, ‘he is writing the document’—ʾet marks the following word as the direct
object of the verbal adjective);
particles may be subdivided by function into the categories known from other languages, such
as prepositions, adverbs, conjunctions, etc., but these sub-categories are not consistently
distinguished, for example the basic prepositions (l, b, k) cannot function as
conjunctions, i.e., be followed by a verbal clause, but some of the particles that have
evolved from substantives can and do show both functions (e.g., ʿad, either ‘up to’
followed by a nominal phrase or ‘until’ followed by a verbal phrase).

C. Three principal analytical categories will be regularly mentioned, the first two are descriptive, the
third historical. Because the three terms are basically metaphors from physical phenomena, the first two
from plant terminology, their application is not transparent, with the result that they are not used
consistently by grammarians. These are the definitions for usage in this class:
Root: the consonantal substructure (skeleton) of any given word: the root of ‫ ָכַּתב‬kātab, ‘he wrote’, is
‫כתב‬/KTB, consonants only, and it may be expressed either in transcription or in Hebrew script.
In the case of “primitive” nouns, i.e., those that must have arisen first as nouns and not as verbs,
some argue that the “root” in fact includes the vowel, e.g., /kalb-/ ‘dog’. This is no doubt true,
but, because native speakers of Semitic languages create secondary forms from these old
nouns, creating a new vocalic structure on the three consonants, even in such cases we can
refer meaningfully to the ‘root’ of Hebrew ‫ ֶכֶּלב‬being ‫כלב‬/KLB.
Stem: the core form of a Hebrew word that appears in a given morpho-syntactic situation, e.g., the noun
dābār has three singular stems: the absolute-state form, just cited; the construct-state stem
dᵊbār, which also appears before the heavy suffixes, as in dᵊbarḵem, ‘your (pl.) word’; and the
nominal segment dᵊbār- which appears only before suffixes, as in dᵊbārōh, ‘his word’.
There is another traditional use of the term ‘stem’ that must be kept kept in mind: the verbal
binyanim are called stems in traditional terminology, as in ‘Qal-stem’, ‘Niphal’-stem, etc. Note
 that this is indeed a very distinct usage from the first because it describes a complete binyan,
and each form within each binyan will have its ownset of ‘stems’ according to the first
definition, e.g., the Qal suffix-conjugation of strong roots shows three stems : kātab in the 3m.s.
as well as in the 1c.s., 2m./f.s., and 1c.s. (kātabtā, etc.), katᵊb- in the 3f.s. and [Link]. (kātᵊbāh,
kātᵊbūw), and kᵊtab- in the 2m./[Link]. (kᵊtabtem, kᵊtabten).
N.B. “c” in these parsings stands for “common,” i.e., common gender, no distinction between
masculine and feminine.
Base: the reconstructed Proto-Hebrew form of a word expressed in terms of the root QTL (more on this
next week).

D. Descriptive grammatical analysis, then, must include all of the features mentioned above in §B, with
the exception of ‘case’ because, despite the presence of remnants of the case markers in Biblical
Hebrew, true case ceased functioning as a grammatical feature in late Proto-Hebrew.

Some examples, from Job 1:1 (the order of mention of the features used here should be used; the
syntactic analysis included here is not required if the request is only “to provide a descriptive analysis”
but it may be requested as a separate exercise):
‫ ִאישׁ‬substantive, noun, masculine, singular, absolute, indefinite, root ‫שׁא‬, root meaning ‘man’, meaning
in context ‘a man’ (may, and should, be abbreviated: n.m.s. abs. indef. ‫‘ שׁא‬man’/‘a man’)
syntactic analysis: subject of following verb
‫ ָהיָה‬verb, third, masculine, singular, suffix-conjugation, Qal, root meaning ‫‘ היה‬to be’, meaning in
context ‘he was’ (abbreviated as: vb. 3m.s. SC Qal ‫‘ היה‬to be’/’he was’)
syntactic analysis: verbal predicate of the main clause to be translated ‘(There) was a man’
‫ ְבֶאָרץ‬particle, preposition ‫‘ ב‬in’ + n.f.s. cstr. def. ‫‘ ארץ‬earth’/‘the earth’/‘in the land of’
syntactic analysis: first element (nomen regens) of a construct chain within a compound prepositional
phrase; the noun is definite because in construct with a definite noun
‫ עוּץ‬proper noun (place name), definite (all proper names are definite)—geographic identification and
etymology uncertain
syntactic analysis: second element of the construct chain (nomen rectum), final element of the
prepositional phrase modifying the main clause ‘in the land of ʿUṣ’
‫ איּוֹב‬proper noun (personal name), definite; the root is ‫ איב‬but the origin and meaning of the name are
debated
syntactic analysis: subject of nominal sentence
‫שׁמוֹ‬ ְ n.m.s. pron. def. ‫‘ שׁם‬name’ + [Link]. ‘his name’
N.B. we parse for three states, absolute, construct, pronominal, the third defined as a noun to which
a pronominal suffix is attached. Why? Because the ‘stem’ as defined above can vary from the
absolute and construct states, as it does here.
syntactic analysis: predicate of nominal sentence
‫ ְוָהיָה‬particle, coordinating conjunction + vb. 3m.s. SC Qal ‫‘ היה‬to be’/’he was’
syntactic analysis: predicate of verbal sentence, either ‘and he was’ or ‘and he always was’
N.B. the problem with this form is whether the conjunction is the ‘waw-retentive’ or the ‘waw-
conjunctive’ (see articles by Pardee and Garr that will be distributed)
‫ ָהִאישׁ‬particle, definite article + n.m.s. abs. def. ‫‘ שׁא‬man’/‘the man’)
syntactic analysis: subject of nominal sentence
‫ ַההוּא‬particle, definite article + particle, distal demonstrative adj. m.s. def. < 3m.s. indep. pron. ‫‘ ה‬that’
syntactic analysis: attributive adjective, agreeing in gender, number, and definiteness (as adjectives
functioning attributively must)
‫ ָתּם‬substantive, adjective, m.s. abs. indef. ‫‘ תמם‬to be complete, whole, proper’/‘(he was) everything he
should be’
syntactic analysis: predicate adjective, agreeing in gender, number, but not in definiteness (as
adjectives functioning predicatively must not)
‫שׁר‬ָ ָ‫ ְוי‬adj. m.s. abs. indef. ‫‘ ישׁר‬to be straight’/‘(he was) upright’
syntactic analysis: predicate adjective
‫ ִויֵרא‬particle, coordinating conjunction + stative verbal adj. (stative participle) Qal m.s. cstr. indef. ‫ירא‬
‘to be fearful’/‘(he was) a fearer of’
syntactic analysis: predicate verbal adjective
‫ ֱאֹלִהים‬[Link]. abs. indef. ‫‘ אלה‬god’/‘God’ (the common noun in the plural, both indefinite and definite,
often functions as a divine name, often simply translated ‘God’)
syntactic analysis: second element of construct chain, an ‘objective genitive’, i.e., ‘God’ is the object
of the fearing, not the subject
‫ ְוָסר‬particle, coordinating conjunction + active verbal adj. (active participle) Qal m.s. abs. indef. ‫‘ סור‬to
turn aside’/‘(he was) a turner aside’
syntactic analysis: predicate verbal adjective
‫ ֵמַרע‬particle, preposition ‫‘ מן‬from’ + n.m.s. abs. indef. ‫‘ רעע‬to be evil’/‘evil’
syntactic analysis: prepositional phrase modifing the predicate adjective and expressing an abstract
locational relationship

A possible English translation: ‘There was a man in the land of ʿUṣ whose name was Job; that man was
(perpetually) proper and upright, a God-fearer and someone who avoided evil.’