Pivotal Issues in Analyzing the Verbless Clause 7
Verbal sentences, however, were divided by the medieval grammarians into
two different parts: the fiºl (lit., ‘action’ or verb’) followed by the faºil (lit.,
‘agent’).
16
Within the category of the nominal sentences, the grammarians
also included sentences that contained verbs when the verb was not in initial
position in the sentence. Thus, in sentences with the order subject–verb, the
constituent that we think of as the subject was called by the grammarians the
mubtadaª, and the verb was called the habar.
The distinction between nominal and verbal sentences was based on the
grammarians’ observations of various kinds of sentences. In the first, illus-
trated by qama Zaydun (‘Zaydun got up’), the verb is initial in the sentence
and the agent follows. In the second, illustrated by Zaydun ªahuka (‘Zaydun is
your brother’), a predication is formed without a verb; Zaydun is the mub-
tadaª and ªa huka is the habar.
17
A third type, illustrated by Zaydun qa ma, pat-
terns like the second with respect to word order. More importantly, the first
and third types differ with respect to verbal agreement with the number of the
subject.
18
In the first, if the subject is plural, this fact is not indexed in the
16. In addition to these specific terms for subject and predicate in nominal and verbal
sentences, respectively, the grammarians from the 10th century onward used general terms
to denote subject (al-musnad ªilayhi) and predicate (al-musnad ) in both kinds of clauses
(Aryeh Levin, “The Grammatical Terms al-musnad , al-musnad ªilayhi and al-ªisna d,” JAOS
101 [1981] 145–65).
17. In Arabic, the nominative case is also the citation form of the noun; both nouns in a
verbless predication are in the nominative case, except that a locational predicate is in the
accusative case. From a comparative point of view, it is interesting to note languages in
which the citation form is not nominative and a nominal predicate is in the citation form.
Examples are attested in Cushitic (Dick Hayward, The Arbore Language: A First Investiga-
tion [Kuschitische Sprachstudien / Cushitic Language Studies; Hamburg: Buske, 1984]
114–15, 135–46), in Nilotic (A. N. Tucker and M. A. Bryan, Linguistic Analyses: The Non-
Bantu Languages of North-Eastern Africa [London: Oxford University Press, 1966] 431–
34, 483–85), and in Bantu languages (Patrick Bennett, personal communication; see also
Bernard Comrie, “The Typology of Predicate Case Marking,” Essays on Language Func-
tion and Language Type Dedicated to T. Givón [ed. Joan Bybee, John Haiman, and Sandra
A. Thompson; Amsterdam/Philadelphia: Benjamins, 1977] 39–50).
18. Agreement of person and gender is indexed on the verb regardless of whether the
sentence is verbal or nominal: (a) verbal qamat Hindu(n) (‘Hind got up [f.]’), (b) nominal
Hindu qamat. The Arab grammarians interpreted the verbal affixes indexing number and
Section of the Sentence,” in Style and Text: Studies Presented to Nils Erik Enkvist (ed. Hå-
kan Ringbom; Stockholm: Språkförlaget Skriptor AB, 1975) 317–34. For a connection of the terms topic and comment with mubtadaª and habar, respectively, see Diethelm Michel,
“Probleme des Nominalsatzes im biblischen Hebräisch,” ZAH 7 (1994) 215–24. For a
somewhat different understanding, see Georgine Ayoub and Georges Bohas, “Les gram- mairiens arabes, la phrase nominale et le bon sens,” in The History of Linguistics in the Near East (ed. Cornelis H. M. Versteegh, Konrad Koerner, and Hans-J. Niederehe; Amster-
dam Studies in the Theory and History of Linguistic Science 28; Amsterdam: Benjamins, 1983) 31–48.