Subject and Predicate in Arab Grammatical Tradition

Author(s): Gideon Goldenberg

Source: Zeitschrift der Deutschen Morgenländischen Gesellschaft , 1988, Vol. 138, No. 1
(1988), pp. 39-73
Published by: Harrassowitz Verlag

Stable URL: https://www.jstor.org/stable/43377734

JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide
range of content in a trusted digital archive. We use information technology and tools to increase productivity and
facilitate new forms of scholarship. For more information about JSTOR, please contact [email protected].

Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at
https://about.jstor.org/terms

Harrassowitz Verlag is collaborating with JSTOR to digitize, preserve and extend access to
Zeitschrift der Deutschen Morgenländischen Gesellschaft

This content downloaded from

79.123.184.37 on Sat, 01 Jan 2022 10:08:04 UTC
All use subject to https://about.jstor.org/terms
 Subject and Predicate in Arab Grammatical Tradition

By Gideon Goldenberg, Jerusalem

To Frithiof Rundgren
1 The basic notion of predication, common for the diverse sentence-
forms, was in Arab grammatical theory, as elsewhere, one of the funda-
mental ideas. The clear exposition of predicative relation is not really
impaired by the different specific terms used for each of the sentence-
types, viz. the nominal and the verbal. In Arab grammatical literature,
the treatment of 'nexus' as such, in nominal and verbal sentences alike,
is by no means marginal, and, except for the most elementary works,
has not usually been neglected. The analysis of nexal relation in Arab
grammatical tradition, and the technical terms involved, have been
expounded afresh in some meticulous studies published recently,1 but
the repeated criticism of the Arab grammarians as being unable to de-
velop the generalized conceptions of subject and predicate may call for
some further comments.
Such criticism was first expressed by Weil: Verständnis. In his view,
the Muslim scholar, for whom the grammatical sentence was not the
expression of judgement as for the Greeks but a group of words which
makes sense, was unaware of the generalized conception of 'subject'
borrowed from formal logic, but distinguished between different kinds
of subjects, or rather nominatives, according to the forms of their predi-
cates (v. especially ibid. 38518_23).2 Some thirty years ago Henry
Fleisch: Verbe stated that the Arab grammarians by employing for
'nexus' the term yisnãd had come near to a general theory of the sen-

1 Suffice it to mention here Levin: Musnad 151-164, Levin: Distinction (esp.

§ 1), al-Mayyāh: 7 snād (esp. ch. 1 "malnà l-isnād ").
2 Weil: Verständnis presents Islamic science (including linguistics) as origi-
nally based on an inductive principle opposite to the logical principle underlying
Greek thought. In his opinion Greek influence on Arabic grammatical termino-
logy came late and was adapted to the conceptual framework of Islamic schol-
arship (ibid. 38536-3863). It was Weil's idea that the development of Muslim
scientific thought from naql "tradition" to ť aql "reason" had given rise to the
excessive rationalization characterizing Arabic linguistics.

This content downloaded from

79.123.184.37 on Sat, 01 Jan 2022 10:08:04 UTC
All use subject to https://about.jstor.org/terms
40 Gideon Goldenberg

tence without reaching it; that musn

"terms of grammatical logic" for analys
- xabar and fã1 il - fťl were the only
respective structures of nominal and
suggested to generalize xabar for 'p
should be called xabar fi'lï) or nomin
muxbar lanhu stand for 'subject'.3
As a matter of fact, not only musnad a
and muxbar lanhu (and other express
Arab grammarians as accepted techni
ject' (cf. Levin: Musnad ch. VI, 161-1
ther below. Al-muxbar ť anhu will also
Bakr b. al-'Anbāri, al-Zaģģāģī, 'Abdal
>An§ārī). Xabar al-mubtada* and xabar
C 1 856) are equally termed xabar [the la
rnī 156)]. The question mooted in Ara
was not whether in a sentence like qãma
qãma "Zayd stood" the word qãma "s
xabar 'predicate'; the problem behind
question what analysis of the verbal p
the asymmetry observable in compar
Zaydüna qãmu.4 The preponderant op
and the one consentaneous to reason,
as Vb. Pred. + Noun Subj. and Noun Su
cal, or symmetrical, in their basic st
does by no manner of means imply a "d
failing to perceive the actual division co
bal sentences into subject and predic
made by Weil and Fleisch was si

3 See Fleisch: Verbe 153-155. Fleisch ibid

that the expression al-muxbar ť anhu is fou
explications"; his recommendation was in
making it a technical term.
4 An especially clear presentation of this
$id I 32717-3288.
5 Méhiri: öumlah 4010_i7 especially disapp
nal grammar of N + V sentences as nomi
pay proper attention to the semantic functi
an agent. Méhiri' s general criticism of
grammatical tradition is in the same vein

This content downloaded from

79.123.184.37 on Sat, 01 Jan 2022 10:08:04 UTC
All use subject to https://about.jstor.org/terms
 Subject and Predicate in Arab Grammatical Tradition 4 1

Mosel6 and Rundgren.7 D. Cohen: Formes duprédicat , while criticiz-

ing the Arab grammarians for having analysed the N + V sentence-form
as nominal,8 rejects the analysis of the extraposed mubtada* as subject,9
and expresses his agreement with Fleisch, stating (ibid. n. 5, p. 226)
that "la tradition grammaticale arabe n'a pas su dégager la notion de
sujet grammatical". Unlike Cohen, M. G. Carter would not accord to
the fā'il "agent", which follows its verb, the status of a subject, only
mubtada' s having in his view the right of being regarded, indiscriminate-
ly, as "true subjects" (Carter's note 1 to al-Širbini: Nūr§ 7.5, p. 159).
Such an understanding of ' subject' , which might seem influenced by the
traditional terminology, is in fact contrary to the teachings of the Arab
grammarians, who insisted on the analogy between fā'il and mubtada .
For better understanding the treatment of predicative relation in
medieval Arabic grammar, some evidence will be brought together and
examined in the following chapters concerning (1) the terminology of
predicative relation, (2) the functional definition of the parts of speech,
(3) kalām and ģumlah, (4) the verb as a nexus-complex, and (5) the syn-
tactical exercises of Hxbâr.

I. Terminology of predicative relation

2 Terms (and semi-terminological expressions) that have been
employed for characterizing predicative relation fall under two catego-

6 See Mosel: Term. Sib. esp. 2 l2s, rightly criticized by Beeston: Rev. Mosel
650-651.

7 Rundgren: Einfluß 134 speaks of Arabic "orthodox grammar" as surpri

singly still operating "mit einer doppelten Definition des Satzes". Like Fleisc
he thinks that the Arab grammarians could not make the last step needed fo
reaching a generalized idea of the primary parts of the sentence (ibid.).
8 Cohen comes close to the doctrine once attributed to the Kūfans (Ibn
Hišām: Muģni CII 379u = D 497!) in taking e.g. Zaydun for fā'il (subject) in
Zaydun qãma the same as in qãma Zaydun regardless of position (and agree-
ment) .
9 A mubtada* that is resumed by a lā*id in an oblique case is, in Cohen's opin-
ion, standing in extraposition, and cannot be regarded as a grammatical sub-
ject. The Arab grammarians held the view that al-fťlu wa-l-fãHlu ģumlatun bi-
manzilati l-mubtada'i wa-xabarihi (al-Zaģģāģī: Ū(ļāh 12019).
10 al-mubtada* naziru l-fā'ili fi l-ixbãri ť anhumã "the mubtada ' is the counter-
part of the fā%l in having both of them sth. predicated of them [i. e. in being both
of them subjects]" (IbnYaťiš: ŠarĶL 10714 = C I 8823). [The similar terminology
in al-Mubarrad: Muqtaxļab III 1276 refers to the special case where an agent is
transformed into a predicate: (ļarabtu - ► al-(ļāribu *anā &c., cf. infra ch. V].

This content downloaded from

79.123.184.37 on Sat, 01 Jan 2022 10:08:04 UTC
All use subject to https://about.jstor.org/terms
42 Gideon Goldenberg

ries: those referring to the relation

members of the predicative constr
functional meaning conveyed throu
Tarkīb, H(ļāfah, in'iqäd, iHilāf See.
grammatical literature for indicating
sophical tradition of describing the
xion" or "composition" (of subject an
sometimes speak of the nexus as iHï
Ū(ļāh I 932_6)12 or (ļamm "joining",13
Hmālatuhū * Hay hi wa-ģalluhū mut
attaching) (one) thing to (another) thin
connected and contiguous (to the ot
quoted in full below § 4); or else as
kalāmu "the composition by which
Šarķ L 236 = C I 2020).
The technical term in use for 'nex
writings throughout the ages is the
thing) against (another)".14 This ter
tion between subject and predicate imp
the sentence is made to lean upon the
cative construction are referred to as
infelicitous terms that have at times

1 1 For oúvúeoiç indicating predicative n

Arabic metaphrase gives tarklb ('Isfoaq b.
Fārābī: Herm. comment. 4424_27 explains
the predicate noun and the subject noun (
He would otherwise speak of irtibāt ( al-m
predicate with the subject)" (al-Fārābī ibid. 3
42 12, 4415, 457 13, 4622, icón). Cf. Boetiu
tio (Boetius: Herm. comment, maj. 433 A
B-C).
12 fa-l-ismu yaHalifu ma1 a l-ismi fa-yakü
fťlu ma1 a l-ismi fa-yakünu dālika (al-Fār
words of his teacher Ibn al-Sarrāģ: Mūģa
črurģānī: Muqtaßid I 936).
13 The noun is the subject of predicati
tatimmu bihi 1-fãHdatu "it is that you jo
'informative usefulness' is completed" (
for (ļamm in the same sense, v. al-May
14 The discussion here following shoul
Musnad, where the problems involved ar
information quoted at length.

This content downloaded from

79.123.184.37 on Sat, 01 Jan 2022 10:08:04 UTC
All use subject to https://about.jstor.org/terms
 Subject and Predicate in Arab Grammatical Tradition 43

confusion because of their grammatical ambiguity. The verbal construc-

tion is ' amodia B Hlà A "you lean(ed) B against A", "you made B lean
upon A". The mumid is then always the speaker, but " musnad " can be
construed both (I) substanti vally, to mean "a support upon which
something leans" (= sanad), and (II) adjectivally, as a passive participle
in the sense of "that which is made to lean (upon sth.)"; and " al-musnad
'ilayhi " can be either (I) a direct attribute with ('ilay)hi referring to the
other term mentioned before, or (II) an indirect attribute with (yilay)hi
resuming its own hidden pronominal mariūt.
The terms (1) al-musnad and (2) al-musnad * Hay hi (being thus derived
from >asnadta B yilà A ) would then yield two possible interpretations (I &
H):

A
I (1) almusnad [= alsanad] =A
B A
(2) almusnad * üay-hi [- almusnad B Hlà musnadin A] = B

B
II (1) almusnad [= almusnad B Çilà musnadin Hlayhi A)] =B
A (B) A
(2) al-musnad Hlay-hi [= al{A) -musnad(B) )

It is obvious that 1(1) = 11(2) = A, and

It has convincingly been argued15 that
stood the terms according to interpretation
musnad Hlayhi (= yilà A) = B. 16 So far as th
tence is concerned, Sîbawayh made his m
A, the support, and the predicate B is m
The only direct evidence for this identif
statement17 saying fa-l-mubtada'u musn
musnadun Hlayhi (Sîbawayh I P 21812 = B
come here just to correlate two sets of gram
of the unlearned, but to state that in the

15 See Praetorius: Rev . Jahn 710-711, Talmon: Problems 66, and esp.
Levin: Musnad 149-151.

16 Sibawayh's terms al-mubtadai and al-mabniyy <alayhi (sc. ialà l-mubt

have a similar grammatical structure (though a different meaning).
17 The general statement that follows appears in a section dealing with
special construction hãdã lAbdu-llãhi munļaliqan, not in the chapter whe
musnad and al-musnad Hlayhi are mainly discussed.

This content downloaded from

79.123.184.37 on Sat, 01 Jan 2022 10:08:04 UTC
All use subject to https://about.jstor.org/terms
44 Gideon Goldenberg

the part that is made to come first (w

the subject A (al-musnad) , and the
first) is the predicate B (al-musnad
not the case in the verbal sentence,
first is the predicate. The desired p
tion with word-order in the verbal se
who analyses qãma ' Abdu-llahi " 'A
maJbniyyun calà qãma w a- qãma hadl
qãma , and qãma is its predicate"18
also 2913.15; cf. id.: yĪJ$vl I 39jo-ii>
Sibawayh himself refrained from
directly identify subject and predi
stated explicitly which part is accor
and which is the B musnad Hlayhi (
inference (Mosel: Term. Sib. 19, 22
Levin: Musnad 148-149, cf. Praetorius: Rev. Jahn 710-711) that
Sibawayh' s musnad was not the subject but always the part in first posi-
tion (muhtadď or verb) is founded on Sibawayh' s statement that in V +
N it is the first part that "needs" the second as in N + N (v. Mosel ib.,
Levin: Musnad 148b). Yet that statement can also be understood
otherwise, as saying that the predicate B needs the subject A the same
as the subject A needs the predicate B. Does not Sibawayh say in the
heading to his ch. 3 about al-musnad (A) and al-musnad Hlayhi (B) that
none of them can do without the other (lã yastagni wahidun minhumã
ť an-i l-ãxar)Í The need for a second part to form a sentence is one
aspect of predication; the *isnãd of the predicate upon the subject is an-
other aspect. The difference between the two is made clear by Ibn al-
Sarrāģ in the passage quoted above. In fact what Sibawayh intended
when saying fa-l-mubtada*u musnadun wa-l-mabniyyu calayhi musnadun
'ilayhi (v. supra) closely resembles what was somewhat differently for-
mulated later by Ibn al-Sarrāģ: al-mubtada* yubtadďu fìhi bi-l-ismi l-
muhaddati ť anhu qabla l-ķadīti "in (the case of) the mubtada? one begins
with the subject before the predicate". The difference between mubtada?
and fã1 il, according to Ibn al-Sarrāģ, is that al-fãHl mubtada'un bi-l-
ķadīti qablahü "the fã1 il is preceded by the predicate" [lit. "(as to) the
fã% one begins with the predicate before it"] (Ibn al-Sarrāģ: 'Uçûll 637_
8; cf. ib. 817_8). As already mentioned above Sibawayh himself usually
(and maybe deliberately) evaded practising the known terms musnad

18 Ņadīļ for 'predicate' is discussed in the sequel.

This content downloaded from

79.123.184.37 on Sat, 01 Jan 2022 10:08:04 UTC
All use subject to https://about.jstor.org/terms
 Subject and Predicate in Arab Grammatical Tradition 45

and musnad y ilayhi , and consequently failed to specify their respective

functions.19
Of great interest to Sîbawayh's terminology is the occurrence in the
same sense of musnad Hlayhi in the Arabic presentation of Aristotelian
Hermeneutics by Ibn al-Muqaffať.20 R. Talmon has recently shown that
in the Logic of Ibn al-MuqafíV musnad * ilay hi is used in its verbal sense,
meaning "leaning upon it, predicated of it":21 Aristotle's definition of
the verb as being always an indication of what is said of something else
(àeì T<x>v xaů' èxépoi) Aeyofievcov orļļieīov) is rendered by Ibn al-Muqaffa*
lã yakünu Hllā mahmūlan ť olà ģayrihī musnadan y ilayhi "it is always pre-
dicated22 of something else; (it always) leans upon it" (Ibn al-Muqaffať:
Manfiq 2812; cf. ibid. 2815) ; musnadan y ilayhi is parallel to mahmūlan <alà
ģayrihī, musnadan glossing upon maĶmūlan and (*ilay)hi referring to
(calà) ģayri( hi): 23 The same text of Ibn al-Muqaffať contains much that is
important to the historical study of grammatical terminology which
does not come within the purview of the present inquiry. Relevant,
however, to the treatment of predicative relation in Arabic grammar is
the use here attested of musnad yilà ... in the sense of "predicated of
. . ." (maķmūValā. . .) and the fact that it is employed in connexion with
the verb characterized as being by definition predicate (see ch. II).
3 While the same doctrine which regarded the predicate B as being
made to "lean" upon i^usnida Hlà) the subject A persisted in Arab lin-
guistics, the terms al-musnad and al-musnad y ilayhi later came to be
understood according to interpretation II, viz. al-musnad = B, al-mus-
nad y ilayhi = A (see abové). The evidence adduced by A. Levin shows
this usage as beginning in the tenth century (Levin: Musnad 151-153).
19 Sibawayh's disinclination to use these terms, whatever his motivation
might be, is justified by their inconvenience and literal ambiguity. Their ambigu-
ity was explained above. Inconvenient they are because term B in Sibawayh's
usage is incapable of being mentioned standing by itself: " al-musnad Hlayhi" is
impossible when not following the term to which ( Hlay)hi refers; " al-musnad Hlà
l-musnad " would be clumsy and obscure.
20 Or rather the son of Ibn al-Muqafíať (2nd half 8th cent.)? See Zimmer-
mann: Log. trad. 537 n. 2.
21 See Talmon: Tafkir , id. Musnad. My thanks are due to R. Talmon for let-
ting me consult his unpublished paper (Talmon: Musnad) and for providing me
with a xerox copy of Ibn al-Muqaffa1: Manfiq. The occurrence itself of musnad
' ilayhi in this text was first mentioned by Zimmermann: Log. trad. 544 n.
22 Since mahmūl as a technical term for "predicate" is common to all sources,
we are allowed to translate it here as iņtilāh.
23 Similar wording of the same definition of the verb will be found in al-Fārisī:
'īiļāh I 7616: yamma l-fillu fa-mā kãna mustanidan Hlà ëayHn (cf. al-Zaģģāģī: U(ļāh
4 218_i9 quoted below § 3).

This content downloaded from

79.123.184.37 on Sat, 01 Jan 2022 10:08:04 UTC
All use subject to https://about.jstor.org/terms
 46 Gideon Goldenberg

Levin (ibid. 161-162) also shows

terms derived from the roots s-n-d
here together so far as they are i
predicate and predication.24 While
xion itself between subject and pre
of hadit belong to the other categor
veyed by the nexus, and represent
as the attribution of new inform
For terms derived from x-b-r , the
tahū (rarely: xabbartahü) bi-B 'an A.
muxbar is the interlocutor,25 al-mu
expresses the xabar (i. e. the new inf
lan-hu is the subject of whom (or
The verbal construction with derivatives of h-d-t is haddalta WA bi-
B. Al-muhadditis the speaker, al-muhaddat ť an-huis the subject, and al-
ķadīt (lan A) is the predicate.
The following excerpts from the grammatical literature are intended
to show how the terms of iisnādy yixbār and Ķadit are comparable and do
commonly interchange, and how deeply rooted was the idea of predica-
tion as the basis of sentence-structure (nominal and verbal alike) among
the Arab grammarians.27
In Sïbawayh' s explanation of the basic structure common to passive
and active verbal sentences, the verbal idea is said to be al-muhaddat
bihi , the subject nouns being referred to as al- asma* al-muhaddat ť anhã
(Sîbawayh I P 10ļ8_2o = B 1410_i2; see also ibid. I P 128ļ9 = B 1546 and
other instances). Even if such use of these expressions was not yét fully
technicalized, it perfectly corresponds with the established terminology
of later grammarians.

24 The terms fā*idah and Hfādah, or (kalām, or qawl) mufid , referring to the
"(informative) usefulness" of the utterance to the hearer (al-fā'idatu li-l-sāmiH fi
l-xabari [al-Mubarrad: Muqta4ab IV 1267]), just indicate generally that the con-
struction is communicative (it should then have the structure of a sentence).
These terms will not be discussed here for themselves.

Cf., e.g., al-muxãtab al-muxbar (Sïbawayh I P 10917 = B 13010, cf. ibid. I P

1718 = B 22 n). [Rarely muxbar is perhaps found to occur as abbreviated where
muxbar bihi or muxbar lanhu are the appropriate terms.]
26 For the modal meaning of xabar see below § 4; al-ixbãr 'anX in the sense of
"forming X into a predicate" will be discussed in ch. V.
27 The idea itself of the grammarians concerning predication has already been
made clear in previous studies, see refs. in fn. 1 above. Sïbawayh' s usage of the
expressions mentioned is not discussed in this context, because not all of them
were already employed in his book as established technical terms.

This content downloaded from

79.123.184.37 on Sat, 01 Jan 2022 10:08:04 UTC
All use subject to https://about.jstor.org/terms
 Subject and Predicate in Arab Grammatical Tradition 47

Ibn al-Sarrāģ used mostly terms of hadlt for predication in the verbal
sentence as in the nominal.28 He also expressed himself explicitly
saying that al-fàlU miuļāri'un li-l-mubtadaH min yaýli *annahumã
ģamīcan muķaddaģun ť anhumā "the fã1 il (i. e. the agent which follows the
verb in a verbal sentence) resembles the mubtada * (= the subject at the
beginning of a nominal sentence) , as both of them are subjects (parts of
the sentence of which the new information is predicated) " (Ibn al-Sar-
rāģ: 'U$ül 1 63 12- 13) ; the difference between them is in the position of the
subject: when it is a mubtada ' the subject comes first, and when it is a
fāHl it is preceded by the predicate (ibid. 637_8, quoted above).
Also al-Zaģģāģī speaks, with respect to the mubtada' about miuļā-
ra^atuhū li-1-fāHli "its resemblance to the fàliT (al-Zaģģāģī: öumal 486).
He elsewhere treated philosophically the division of the parts of speech
in three categories, which he considered universal. After stating that
the interlocutor and the speaker and the subject and the predicate are
substances and accidents represented in speech by their names
(nouns),29 he says that (being different from the nominal parts) al-xabar
huwa l-fťlu wa-ma ětuqqa minhu *aw tadammana ma'nāhu, wa-huwa l-
haditu li-lladi dakarnahu "the xabar is the verb and what is derived from
it or implies its meaning, and it (= the xabar) is predicate of what we
have mentioned" (al-Zaģģāģī: Ūfļāķ 42 18_19) . The full meaning of the pas-
sage as a whole and of the terms there used needs further study,30 but
the identification of xabar (and ķadīt) with verb, reflecting the philoso-
phical tradition, is another evidence of these terms being used regard-
less of sentence-type and word-order. The analysis of, e. g., (ļarabtu as a
predicative complex where the speaker himself (as represented in the
1st pers. afformative) is the subject (al-Zaģģāģī: Ū<ļāii 1367_8) will be
discussed in ch. IV below.
In Ibn öinnTs usage: yasnadtaB*ilàA = ķaddatta bi-B canA (= nasabta
B Hlà A ?); v. Ibn öinni: Imma ť 106_7 & 132_3.
'Abū ťAlí al-Fārisī makes it clear that any predicate, whether it is
nominal or verbal, following the subject or preceding it, is equally the
predicate (xabar) of its subject noun: 'Abdu-ltiihi muqbilun, qãmaBakrun
- fa-muqbilun xabarun 'an * Abdi-llahi , wa-qama xabarun can Bakrin (al-
Fārisī: 7 (ļah, I 69 10). His statement *amma l-fi'lu fa-mã Jcãna mustanidan
Hlà šayHn "as to the verb, (it is) what is predicated of (lit. leaning upon)

28 Ibn al-Sarrāģ: Mūģaz 2913_15 2i*> id.: *U$ūl I 3910_n.

29 al-muxāļab wa-l-muxā'ib wa-l-muxabbar ť anhu wa-l-muxabbar'bihī' * aģsa -
mun wa-aWā4un tanūbu fi l-ibārati lanhā iasmāiuhā (al-Zaģģāģī: ^Ifļāh 4214_15).
30 Cf. Versteegh: Greek Elements 145-148.

This content downloaded from

79.123.184.37 on Sat, 01 Jan 2022 10:08:04 UTC
All use subject to https://about.jstor.org/terms
 48 Gideon Goldenberg

something" (al-Fārisī: Ū<ļāķ I 7616,

ed by al-črurģānī: Muqtaçid I 7619_
maýrá l-ixbãri, fa-ka'annahü qãla : w
€an šayHn "you should know that iisn
if he said: as to the verb, (it is) that w
thing" (see also ch. II below). Al-čru
explains the principles of sentence-s
formed in Arabic only bayna l-is
*axūka , fa-Zaydun mubtadďun wa
minhuma smun, yaw bayna wa-l
wa-surra Bakrun wa-nfalaqa * Abdu-ll
kā muxbarun lanhu "[the nexus (iHi
noun and a noun, as you (may) say:
ther"), and then Zaydun is a mub
occupy initial position) and *axuka i
is a noun; or beteeen a verb and a
("Zayd went out") and surra Bakrun
du-llahi ("'Abdallah set off"), these a
a subject (muxbar ^anhu)" (al-0urģā
The use of Hxbãr and yisnād in the same context will also be found in
Ibn al-)Anbārī: yAsrãr, where the verb, e.g., is characterized as mã yux-
baru bihl wa-lã yuxbaru lanhu (36_7 = 2410_n) or mã 'umida Hlà šay*in wa-
lam yusnad Hlayhi šay*un (613 = 2117_8), both expressions saying that a
verb can be a predicate but not a subject (cf. ch. II below). As to the sub-
ject noun, he explains *anna l-mubtada'a muxbarun ť arām kamā ' anna l-
fā'ila muxbarun lanhu "the mubtada* is a subject the same as the /ā^/ is a
subject" (Ibn al-'Anbārī: *Asrãr 3022-23 = 2699_10). Here as in some other
places, the fact that both the mubtada* and the fā'il are subjects is indi-
cated for explaining why both of them are in the nominative.
In terms of 'isnãd, 'ixbãr and hadit, the basic constructions of nominal
and verbal sentences alike are equally analysed as being composed of a
subject and a predicate (the arrow indicates word-order):

A Subject B Predicate
nominal Noun -► any expression
sentence (al-mubtada>) (al-mabniyy ť alà l-mubtadď)
verbal Noun <- Verb
sentence (fā'il) (fťl)
any musnad Hlayhi musnad (Hlà A)
sentence muxbar lanhu xabar (or: al-muxbar bihi)
muhaddat ť anhu ķadīt (or: al-muhaxidat bihi)

This content downloaded from

79.123.184.37 on Sat, 01 Jan 2022 10:08:04 UTC
All use subject to https://about.jstor.org/terms
 Subject and Predicate in Arab Grammatical Tradition 49

If any additional evidence is needed at all to convince the disbelieving

that this terminology was common and well established and repeatedly
employed to convey the accepted idea that each sentence is to be analy-
sed, in the first place, as consisting of a subject and a predicate, then Ibn
Yaťiš can still be quoted analysing, e.g., Zaydun *axüka "Zayd is your
brother" as muxbar ť anhu + xabar , and, e. g., qãma Zaydun "Zayd stood"
with the verb as xabar and the following noun as muxbar lanhu (Ibn
Yaťiš: Šarķ L 236_8 = C 1 2O20-22) or the verb as xabar predicated (s-n-d)
of a mvhaddat 'anhu (ibid. L 237_8 = C I 202i_22> also L 2815_16 = C 1 25ļ_2).
Instances of referring to predication as yisnād (*ilà l-ism A) = Hxbãr ('an
A) = ķadlt ('an A) (like Ibn ťAqil: Šarķ to verse 10, al-Širbīni: Nūr 24ļ)
could easily be multiplied.
4 Though interchangeable as terms for subject, predicate and predica-
tion, the derivatives of s-n-d, x-b-r , and h-d-t differ in their implications,
as xabar also denotes "statement" or "proposition", thus contrasting
with the other, 'modal', kinds of sentences. About the number of these
classes of sentences there was no agreement between the scholars, sug-
gested categories ranging from two to six.31 The Aristotelian Àóyoç àno-
(pavtixóç, which was rendered into Arabic as qawl ģāzim32 or kalām
façl ,33 will be found in the grammatical literature marked as xabar ,
yixbār, ģumlah xabariyyah (e.g. Ibn öinni: Laima ť 75 10) or kalām xabari
(al-Širbīnī: Nūr 2809). Aristotle's definition of sentences of this category
as having in them truth or falsity (Aristotle: Herm. 1 7a3) - in Arabic $idq
or kidb 34 - was often repeated (sometimes incidentally) by grammar-
ians.

A few instances quoted from Arab grammatical writings will suffice

for illustrating the use and the meaning of this terminology: - wa-l-
xabaru mā ģāza lalà qā*ilihi l-ta$dïqu wa-l-takdïbu "and the xabar is that
whose sayer can be declared to be right or wrong" (al-Mubarrad: Muqta-
(ļab III 894); wa-bi-l-xabari yaqacu l-taçdïqu wa-l-takdïbu "and in the
xabar 35 lies the expression of truth and falsity" (Ibn al-Sarrāģ: >t Jçiïl I
67 14) ; in connexion with the interrogative V, the difference (farq) bayna

31 See Hasan Šādili Farhud in his note 14 (pp. 148-149) to Ibn Öinni:
^Uqud, Versteegh: Greek Elements 145-148.
32 Aristotle: Herm. 17a2-3 as translated by Tsfraq b. IJunayn. Incidentally
kalām ģāzim in Ibn al-Muqafiať: Manfiq 29 15.
33 Ibn al-Muqaffať: Manfiq 29 10. n & elsewhere.
34 So in Tsķāq b. IJunayn's translation; cf. Ibn al-Muqaffať Manfiq 2915_18
35 In this context, the xabar mentioned is the predicate in a declarative nomi-
nal sentence.

4 ZDMG 138/1

This content downloaded from

79.123.184.37 on Sat, 01 Jan 2022 10:08:04 UTC
All use subject to https://about.jstor.org/terms
 50 Gideon Goldenberg

l-istifhãmi wa-l-xabari "between quest

('Abū Bakr b. al-'Anbārī: *Alifãt 4564.
*arbalatun: Hxbãrun wa-stixbãrun wa-
of the sentence are four: statement, q
tion" (Ibn öinni: lUqūd 1405); a relativ
ģumlatun xabariyyatun tahtamilu l-çidqa
is susceptible of truth and falsity" (Ib
I 18612-i3); al-kalãmu qawlun muf
yinša*un (Ibn Hišām: Gãmi ť 24).37
Since the terms xabar and Hxbãr con
tion to syntactic analysis is restricted
analysis of other kinds of sentences the
that indicate neutrally "nexus" will
' anna l-isnãda maýrahu maýrá l-ixbãri (

fãyidatan laysat fi l-ixbãri, wa-hiya ^anna min

Hflãqu l-ixbãri lalayhi ka-fťli l-amri nahwi
l-amru lã yakunu min hay tu *anna l-xahara
l-kidbu, wa-ya$ihķu *an yuflaqa ialayhi l-isn
' isnãdi Hfļāfatu 1-šayH Hlà 1-šayH wa-imãlatuhu
çilan wa-mulãmisan "you should know that * is
(

there are verbs to which Hxbãr is inapplicable

e. g. U-ya4rib Zaydun ("let Zay d strike!"), since
for xabar is that which admits truth and falsity
to them, because the real meaning of ' isnãd
(another) thing, inclining it to it, and makin
guous (to the other)" (al-0urģānī: Muqtaçid I
was also expressed clearly by Ibn Yaťiš: al- isn
li-anna l-isnãda yaëmalu l-xabara wa-gayrahü
wa-l-istifhãmi, fa-kullu xabarin musnadun w
xabaran " * isnãd is more general than xabar ,

36 It is explained that in, e. g., *al-daharayni harram

he forbade (to eat)?" Qur'än vi 143 it is the prolonga
ence between the statement form * al - and the q
37 This classification implies an overall distinctio
fact (Hxbãr) and the expression of everything inv
attitude (Hnša*). About Hxbãr ( xabar) and Hnša* v.
I 779-780, III 541-542. The same distinction, with
identically expressed in Ottoman Turkish grammar
and in§aí siygalar.

This content downloaded from

79.123.184.37 on Sat, 01 Jan 2022 10:08:04 UTC
All use subject to https://about.jstor.org/terms
 Subject and Predicate in Arab Grammatical Tradition 5 1

("statement") as well as not- xabar: command and prohibition and ques-

tion; so every xabar is musnad but not every musnad is xabar" (Ibn
Ya'is: Šarķ L 232.3 = C I 2016_17).38
Ibtida* deserves to be studied apart, and will not be further discussed
in this connexion.

II. The functional definition of the parts of speech

5 The characterization of noun and verb in terms of their syntactical
function in predication has already been mentioned above, especially
the Arabic rendering of Aristotle's definition of verb as indicating what
is predicated of something else (v. supra § 2, cf. Versteegh: Greek Ele-
menta 71-72). The similar definition of noun as a subject of which
something can be predicated was studied by Versteegh: Greek Ele-
ments 57-59. 39 The definition of noun by referring to its function as a
subject was said to be the most useful (' anfa'u lalāmāti l-ismi) because it
could also be applied to pronominal bound morphemes which do not
show the other characteristics of nouns; see al-Širbīnī: Nūr 24ļ_5 (cf. § 10
below). Neither of these definitions will be discussed for itself in the
present chapter,40 nor the question of the priority of the noun, which
was connected with these definitions.41 This chapter is intended to
adduce some instances illustrating the method of defining systemati-
cally all parts of speech by their capability of occurring in the positions
of subject and predicate. This method, though not equally developed in

38 This reference is mentioned in Levin: Musnad 162 a. In al-Ra<Jī: Šarķ al -

Kãfiyah I 82_6 the ' isnãd "nexus" is said to be present either on the surface (fi l-
hal) or at the bottom (fi l-açl ), the latter category also embracing the forms that
are not statements, which would not fall under Hxbãr.
39 The definition mentioned there which is ascribed to al-'AxfaS al-'Awsat
says: al-ismu rrvā ģāzafihi nafa1 ani wa-<ļarranī "the noun is that which admits
rmfa^ani ("

'Itļāķ 49 12). This is a method of

mid-century American structu
morpheme-position) class". Mo
environment in general terms o
its susceptibility of being a m
40 Definition of the verb by it
^Uqüd 14015; for the noun defin
54, al-Zamaxšari: Mufaççal 419
Hiääm: öämi ť 110.
41 About the priority of the
Regarding the noun as having
confined to its privileges of oc
4«

This content downloaded from

79.123.184.37 on Sat, 01 Jan 2022 10:08:04 UTC
All use subject to https://about.jstor.org/terms
52 Gideon Goldenberg

all grammatical writings, was comm

ians, and will be found in scholarl
standard books. Besides being emplo
method clearly establishes their hi
tion of the method as generalized f
noun is still characterized as being
a subject; later it was usually defin
and a predicate.
fa-l-ismu taxu$$vhū yašya*u yuHaba
isma mā tjāza *an yuxbara lanhu, nah
qàma BaJcrun, wa-1-fťlu mā kãna xab
[

takūna xabaran [

one of which is to say that the noun is that of whic

dicate (sth.), as you say ! Amrun muntaliqun ("(Am
qãma Bakrun ("Bakr stood") ; and the verb is that w
cate and it is not possible for it to be a subject [

are those which can be neither subjects nor pred

yU$ūl I 398_14; cf. id.: Mūģaz 272-i4).
hādihi l-aqsâm al-talātah lahã talãtu marãtiba, f
bihï wa-yuxbaru ť anhu , wa-huwa l-ismu, nahw
minhã mã yuxbaru bihï wa-lã yuxbaru 'anhu, w
qãma Zaydun , wa-minhā mā lã yuxbaru bihï wa-l
huwa l-ķarfu, nahwu hal wa-bal "these three parts (
ranks: (one) of them is that which may be a predica
ject, and that is the noun, like Zaydun qâ'imun ("
them is that which can be a predicate and cannot
is the verb, like qãma Zaydun ("Zayd stood") ; and
which can be neither predicate nor subject, and
ticle), like hal and 6aZ" (Ibn al->Anbāri: yAsrār^s.% =
2i3-I9 = C 417- 54).
al-manfüq bihï Hmmā yan yadulla lalà matnan y
w a-1 anhu wa-huwa l-ismu [

ť anhu wa-huwa l-fi'lu [

lā bihï wa-huwa l-ķarfu "the pronounced [sc. word] wil

meaning proper for predication to be expressed by i
[= subj.], and that is noun [

expressed by it and not of it, and that is verb [

predication to be expressed neither of it nor by it, and

(al-Zawāwī: Fuņūl 313-43). Cf. al-Mutarrizī: Miçbãh
statement couched in terms derived from h-d-t.

This content downloaded from

79.123.184.37 on Sat, 01 Jan 2022 10:08:04 UTC
All use subject to https://about.jstor.org/terms
 Subject and Predicate in Arab Grammatical Tradition 53

For explaining why a minimal nexus-construction ought to consist of

either two nouns or a noun and a verb, Ibn Yaťiš mentioned the fact that
al-ism kamā yakünu muxbaran ť arihu fa-qad yakünu xabaran "the noun,
the same as it may be a subject, would also be a predicate", and al-fi'l
nafsuhü xabaran "the verb itself is a predicate" and therefore cannot
function as a subject, or without a subject noun, while al-harf ģā*a li-
malnanfi l-ismi wa-1-fťlifa-huwa ka-l-ģuzH minhumā, wa-ģuz*u 1-šayH lã
yan'aqidu mala ģayrihī kalāman "the particle comes to indicate some
meaning in the noun and the verb, and it is like a part of them;42 and a
part of a thing would not compose with the rest of it a sentence" (Ibn
Yaťiš: Šarķ L 237_10 = C I 2021_24).43
A later author similarly expressed the same idea saying that al-ismu
bi-hasabi l-wafii yaçluhu li- an yalcüna musnadan wa-musnadan Hlayhi,
wa-l-fťlu yaçluhu li-kawnihï musnadan lã musnadan Hlayhi, wa-l-harfu lã
yaçluhu li-ahadihimã "the noun, according to function, is fit to be predi-
cate and subject, the verb is fit to be predicate but not subject, and the
particle would not fit to be any of the two" (al-Raçlí: Šarķ al-Kãfiyah I
8apu-pu).
The functional definition of word-classes is for a certainty an integral
part of Arab linguistic heritage, and it cannot be separated from the
generally applicable notion of predicative relation.

III. Kalām and ģumlah

6 Kalām and ģumlah as grammatical terms overlap in their implica-

tions, and are comparable when denoting 'sentence'. In that sense they
are sometimes interchangeable, and often confused, but are not synony-
mous. Kalām means "speech" or "utterance", and thence a self-sufficing
unit of speech, an independent sentence; ģumlah "ensemble, complex"
is opposed to mufrad "a single word", and is employed as a technical
term meaning a nexal construction, a structured sentence or clause of
any form or type.44 The precise meanings of kalām and ģumlah (translat-
42 Wanna l-ķarfa lã ma'nà lahū fi nafsihi "because the particle does not have
a sense of its own" (Ibn Yaťiš: Šarķ L 288_9 = C I 2422_23).
43 Cf. ibid. L 283_10 = C I 2416_23, which is also quoted by Levin: Musnad
159 a-b.

44 The contrast between (a) simple expressions (single words) and (b) compo-
site expressions (subject - predicate combinations) goes probably back to Aris-
totle's Categories la16- 19; cf. Versteegh: Greek Elements 139. The grammati-
cal terms (a) mufrad and (b) ģumlah correspond to (a) mufrad and (b) muSallaf in
the later Arabic version by 'Afcīr al-Dīn al-'Abharī of Prophyry's Eisagogē (al-
'Abharī: Eisagogē Arab. 422-56 = 17-18).

This content downloaded from

79.123.184.37 on Sat, 01 Jan 2022 10:08:04 UTC
All use subject to https://about.jstor.org/terms
54 Gideon Goldenberg

ed "discours" and "proposition" re

32) will presently be explained by
literature. The reason of dealing her
they are connected with the defin
tence as primarily consisting of
The history of kalām and ģumlah a
died in this connnexion. Terminol
kalām for "sentence" (kalām mus
denoting a well-constituted sente
wayh. öumlah appears to have be
known to al-Mubarrad (v. esp. hi
Qur*ān attributed to al-Zaģģāģ it
Since ģumlah is a predicative cons
telligible purpose (mufidah) and may
to make some complete sense. But
sarily capable of standing alone,
compound. As ģumlah mufidah w
kalām , the two terms have not
In the usage of Ibn al-Sarrāģ kalā
dent sentence; if it is incomplete or
clause or a sentence taking part i
tences in general may be referred
(Ibn al-Sarrāģ: ' UģūI I 43 14, 445_6 1
dicate (xabar) of some mubtada* is ģ
mufrad , and can itself consist of f
ibtida* & xabar (= muķaddat ť avh
n ig; cf. ibid. 637_8). An important
saying explicitly: kullu ģumlatin ta'
H*rābihā ka-ķukmihā H(ļā lam yakun
that comes after a mubtada* [i. e. th
ing to it with respect to its Hlrāb is
when there is no mubtada* before i
fact indicated here, viz. that the str
the same if it stood alone, is the

45 See al-Mubarrad: Mugtcuļab III 127

wa-lã šakka yannaka qad * arafta l-ģum
ģumala ļn(at)āni: fi'liyyatun wasmiyy
(-type) s; do not you see that it has be
are two: verbal and nominar (al-Zaģģ
539„).

This content downloaded from

79.123.184.37 on Sat, 01 Jan 2022 10:08:04 UTC
All use subject to https://about.jstor.org/terms
 Subject and Predicate in Arab Grammatical Tradition 55

kalām and ģumlah could not plainly develop to correspond to the differ-
ence between a sentence complete in itself and a dependent clause.
7 A note should here be made of a difference in the use of "ģumlah" be-
tween earlier and later writers. A sentence like Zaydun yaqümu "Zayd is
standing" would be analysed by most of the grammarians as a construc-
tion of a mubtada * with its xabar in which the latter constituent, being a
finite verb, contains its fā'il in the form of a personal pronoun referring
back to the mubtada* . For Ibn al-Sarrāģ, such a verb-form which makes
the xabar , though complex by nature, is not a ģumlah : a predicate of a
mubtada* can either indicate (I) an entity identical with the subject,47 or
else (II) what is other than the subject, with a personal pronoun recal-
ling the subject;48 in the latter case it may either be (Ila) a verb contain-
ing a personal pronoun referring to the subject,49 or (lib) a ģumlah con-
taining such a pronoun.50 Examples for (Ila) are sentences like Zaydun
yaqümu, or al-Zaydãni yaqümäni, and for (lib) , inter alia, Zaydun (ļarab -
tuhū, or Bakrun qãma *abūhu (or with a xabar that is a nominal sen-
tence) . It will be noticed that Ibn al-Sarrāģ would not regard as ģumlah a
verb-form whose agent pron. resumes directly the subject, but in his
view a verbal xabar , like any predicative construction, is a ģumlah when
referring to the subject indirectly through any other actant.51 No wonder
he called his chapter concerning the annexion of nouns indicating time
to predicative complexes babu *i(ļāfati l- asma* i *ilà l- a fali wa-l-ģumali
"chapter on the annexion of nouns to verbs and ģumlah- constructions"
(Ibn al-Sarrāģ: *U$ūl II 9ļ),52 separating the verbs (that are in fact

47 That will then be xabaru l-mubtada?i lladi huwa l-awwalu fi l-malnà (Ibn al-
Sarrāģ: *U$ūll 689). The same idea (?) is also expressed otherwise: al-mubtada '
lã budda min *an yaküna l-mabniyyu lalayhi šay 'an huwa huwa (Sibawayh I P
2395_6 = B 2786). Cf. al-Mubarrad: Muqta<ļab III 1 1812, IV 127 10; Ibn al-'Anbārī:
Unçãf L 2825 = ^ ^^8-9*
48 xabaru l-mubtadaH wa-huwa lladi yakunu gay ra l-awwali wa-ya?haru fihi
(ļamiruhū (Ibn al-Sarrāģ: 'UqüI I 6915_16).
49 *an yaküna l-xabaru fi'lanßhi <ļamiru l-mubtadaH (ibid. I 6916_17). Note that
Ibn al-Sarrāģ would regard qãHmun in the sentence Zaydun qa*imun as a predi-
cate identical to the subject in its reference, whereas yaqümu in Zaydun yaqümu
is a verbal predicate distinct from the subject (yakünu gay ra l-awwal ), only its
pronominal agent resuming the mubtada'
50 . . ?aw yakuna äumlatan fihā (ļamiruhū (ibid. I 709).
51 Cf. al-Mubarrad: Muqtadab III 1774.
52 For such annexion in general, see Sibawayh I P 4095-4103 = B 4608-461n
(§ 260); also ibid. I P 447-457 = B 5418-552; al-Mubarrad: Muqta4ab III 1767-
1774. Ibn al-Sarrāģ seems well to recognize that verbs are possible in the posi-
tion of a muiļāf Hlayhi precisely as they are sentence-like (see his ^Uçûl II 912).

This content downloaded from

79.123.184.37 on Sat, 01 Jan 2022 10:08:04 UTC
All use subject to https://about.jstor.org/terms
 56 Gideon Goldenberg

nexus-complexes) from the ģumal. [C

rāģ, the predicate (a) qãma in, e. g.,
garded as a ģumlah , the same as (b) qa
the pers. pron. hidden in (a) qãma b
yabvhu. See Ibn Yaťiš: Šarķ L 107 18
read for the mistaken qãma yabūhu
Rather strange appears the classifi
Zaģģāģī in his grammar: the xabar-
xaraýa ìabvhu "Z ay d - his father wen
9axãka " 'Abdallah honoured your br
ma ttaçala bihī min fā'ilin wa-mafü
attached to it", and only a xabar in
defined as ģumlah (al-Zaģģāģī: Gum
however, would in other places rega
the same status as a nominal senten
bi-manzilati l-mubtadaH wa-xaharih
1 1919) .
Ibn öinni did not mean to refer directly to the question of the differ-
ence between kalām and ģumlah ; he nevertheless commented upon the
subject. In the first chapter of the Xaçâ'iç, Ibn öinni explains the differ-
ence between qawl and kalām:54 While qawl is any word pronounced
(Ibn öinni: Xa$ã'i$ I 1713), kalām is a (unit of) speech which is indepen-
dent, "useful" (mufid) for (communicating) its meaning (Ibn öinni:
Xaça'iç I 179), complete (ibid. 1713_14) and capable of standing alone
(ibid. 197). In some sense it matches "sentence". What is important is
Ibn öinni' s note including within the category of kalām both (a) nexus-
constructions (ģumal) and (b) interjections or exclamatory expressions
Caçwãt and ' asmã ì ál-afãl):ss wa-huwa [= al-kalãm ] (a) allodi

53 Four xaòar-constructions are enumerated there to be predicated of a sub-

ject: (1) ism huwa huwa "noun identical (in its reference) with the (subject) (2)
fťl wa-ma ttaçala bihī min fã1 Hin wa-mafülin, (3) zarf "adverbial expression",
and (4) ģumlah = a nominal clause, as in Zaydun * abühu qä^imun "Zayd - his
father is standing". This classification was sharply condemned in Batalyosi's
critical commentary precisely because constr. (2) is here separated from (4) the
ģumal (al-Batalyösi: 7 $lāķ al- x alai 150).
54 This lengthy argument is possibly directed against the logical tradition
where qawl was employed for rendering "sentence".
55 On these see esp. Rödiger: ^Asma* al-afāl and Canard: Fatāli (and the
bibliographical references there); Fleisch: Traité II 503 (§ 153 i-j). The most
comprehensive treatment of the ' asmã ' al-afãl and the ya$wãt in Arab grammati-
cal literature will be found in al-Ra<ļī: Šarh al-Kāfiyah 6529-84n (on Ibn al-IJā-
ģib: Kãfiyah 573-582).

This content downloaded from

79.123.184.37 on Sat, 01 Jan 2022 10:08:04 UTC
All use subject to https://about.jstor.org/terms
 Subject and Predicate in Arab Grammatical Tradition 57

yusammihi l-nahwiyyüna l-ģumcda, nahwu Zaydun ^axuka wa-qãma

Muhammadun, wa-<ļaraba Salïdun, wa-fi l-dāri *abuka, wa- (6) §ah, wa-
mah, wa-ruwayda, wa-ha*i wa-ā?i fi l-açwât, wa-hassi, wa-labbi wa-uff
wa-awwah; fa-kullu lafyin istaqalla bi-nafsihl wa-ģuniyat mirām tama-
ratu macnãhu, fa-huwa kalàmun "and it [= the kalām ] is (a) what the
grammarians call ģ umal, like Zaydun 'axuka "Zayd is your brother",
qāma Muhammadun "Muhammad stood", (ļaraba Sa'īdun "Sa'id struck",
or fi l-dāri *abūka "in the house is your father", and (b) $ah "hush!", mah
"let alone!", ruwayda "gently!", ķā*i or cãyi "go!" (in driving camels) of
the inteijections; also hassi "autsch!", labbi "at your service!", yuff
"ugh!", yawwah "ah, alas!"; thus any (unit of) speech independent by
itself from which one picks the fruit of its meaning, that is kalām " (Ibn
öinni: Xaçâ'iç I 179_12). It will be noted that only category (a) is identi-
fied with what the grammarians will call ģumal. The idea is repeated in
the same connexion, where Ibn öinni explains that, unlike qawl , which
may also be incomplete (nciqiç), the kalām is necessarily complete
( tāmm) , fa-l-tāmmu huwa l-mufídu, *alnï (a) al-ģumlata wa-(b)mā kārta fi
ma<nãhã min nahwi çahin wa-ïhin "now the complete is the 'useful' (or:
'informative'), I mean (a) the ģ umlah and (6) what is in its meaning, like
sah "hush!" and yih "well, proceed!"" (Ibn öinni: Xaçâ'iç I 1713_14).56 By
adopting the definition of kalām as any (unit of) speech that makes
some complete sense,57 inteijections and exclamations are made to be
subsumed under this category, which undoubtedly are not ģumal , as the
term ģumlah implies a structure with predicative relation between its
constituents: subject placed in front {mubtadď) with xabar that is predi-
cated ( musnad , muhaddat) of it (Ibn öinni: Laima ť 1015_i6 & 107) or verb
predicated ( musnad ) or attributed (mansvb) to its agent (ibid. 1016 &
1^2-3)-
Besides showing that not every kalām is a ģumlah , Ibn öinni' s exposi-
tion also makes it clear that for him kalām, being the generic term for
ģumal (and their likes), is the more comprehending: a ģumlah like qãma

56 In some contexts, inteijections and (nominal) exclamations are not expli-

citly mentioned among the sentence-types, for the sake of brevity (as in Ibn
Öinni: Xaçiïiç I 197, 328_10) ; in such cases one can get the wrong impression that
kalām = independent ģumal. [The reversely formulated statement in Ibn öinni:
Lama ť 10 15 is inaccurate.] öumal murakkabah are regarded as the normal forms
of kalām (Ibn öinni: Xaçâ'iç I 26), inteijections and independent exclamatory
expressions being just their analogues (cf. ibid. Ill 178).
57 On this conception in Sibawayh see Mosel: Term. Sib. 18. This is the idea
of kalām as often in Arabic grammar characterized by Hfādah and (optionally)
followed by sukūt.

This content downloaded from

79.123.184.37 on Sat, 01 Jan 2022 10:08:04 UTC
All use subject to https://about.jstor.org/terms
58 Gideon Goldenberg

Muhammadun "Muhammad stood

sequence qäma Muhammadun , wa-
and your brother is öa'far" is kalām
wa-axuka Galfarun, wa-fi l-dāri Sa'
is your brother, and in the house i
notion of kalām is not strictly gra
explicit distinction, communicative
speech (discourse), and a complete an
last mentioned sense especially appr
as a sentence.

A conception of kalām as a grammatical term, very similar to w

has often been considered sentence by many a modern linguist, is
of Ibn Fāris, a contemporary of Ibn öinni's. Ibn Fāris mentions
alternative definitions saying that kalām is an utterance, or a com
tion of words, which makes sense: za1 ama qawmun *anna l-kalām
sumi1 a wa-fuhima, wa-dalika qawlunã: qäma Zaydun wa-dahaba 1 Amr
wa-qãla qawmun: al-kalāmu ķurūfun mu'allafatun dāllatun 'alà maz
"some people maintain that kalām is what has been heard and un
stood, and that is [like] what we say: qäma Zaydun "Zayd sto
dahaba * Amrun "<Amr went"; others say: kalām is a combinatio
words which signifies some meaning" (Ibn Fāris: Sahibi 81).
examples for the signifies some meaning" (Ibn Fāris: $āķibi 81).
examples for the former definition, and hurüf mu9allafah of the latt
would equally characterize ģumlah.
8 Confusion of the terms kalām and ģumlah could arise from cont
where a kernel sentence forming a complete grammatical struct
standing alone, and making sense, is said to be kalām and also a ģu
From such cases one could draw the conclusion that the two terms were
taken there to be synonymous, as if a common referent should necessa-
rily involve an identical sense. For most writings such a conclusion
appears to be wrong. Al-0urģānī, e.g., in his practical grammar, opens
the bab al-mufrad wa-l- ģumlah "chapter on single word and (nexus-)con-
struction" saying: illam yanna l-wahida min al-ismi wa-1-fťli wa-l-ķarfi
yusammà kalimatan, fa- ida Halafa minha tnāni fa-afādā nahwa xaraýa
Zaydun summiya kalãman wa-summiya ģumlatan "you should know that
the singular of a noun or a verb or a particle is called kalimah ("word"),
and when two of them are combined and are "useful" [i. e. they are in
predicative relation], like xaraýa Zaydun "Zayd went out", it is called

58 Qabãwah: Tmò al-ģumal 3 17 is certainly right in reckoning Ibn öinni

among those who had not regarded kalām and ģumlah as equivalent terms.

This content downloaded from

79.123.184.37 on Sat, 01 Jan 2022 10:08:04 UTC
All use subject to https://about.jstor.org/terms
 Subject and Predicate in Arab Grammatical Tradition 59

kalām , and called ģ umlah" (al-0urģānī: Gumal 402_3). It is well made

clear that xaraýa Zaydun is kalām , and also a ģumlah. What comes next
in the same chapter is not less clear in showing that in no way was it
possible for the author to think that "kalām" is identical with " ģumlah ",
as he goes on specifying the cases in which al-ģumlatu taqa*u mawqťa
l-mufradi "the ģumlah may occur in the position of a single word" (ibid.
406), in which context the substitution of kalām for ģumlah is both
unthinkable and would never occur.
The same can be stated with respect to the similar (and better known)
passage in al-Zamaxšarfs Mufaççal, where in the first fa$l we read: wa-
l-kalāmu huwa l-murakkabu min kalimatayni iusnidat yihdāhumā Hlà l-
'uxrà wa-dMika lã yata*attà HUā fi smayni [

two words one of which is predicated of the other

ed to be realized if not with two nouns [

This passage means to tell that each of the basic

refers is kalām , i. e. (an) independent (unit of) spe
i. e. a predicative construction. The author here
of sentence structure, whose main condition is s
predication, the leaning (yisnād) of one word upon
tioned above, kalām and ģumlah refer to two aspec
sentence. For al-Zamaxšari, the two terms are b
those cases where the ģumlah is complete and in
cases, ģumlah alone is used, and it is very unlikely
terchange there with kalām ; in fact it never d
that are contrasted with single words (nouns) , occ
of predicate (ibid. 138 = 2412), ķāl (ibid. 297 = 647),
çilah (ib. 572-3. 5 = 1428.9, 143^, are always trea
(wäqi'ah) in one or another position.
To sum up the common conception of the diff
tween the two terms it was said that al-farqu
kalāmi y anna l-ģumlata mā tacļammana l-isnāda
maqçûdalan li-dātihā *aw lā, ka-l-ģumlati llatī h
See. [

çûdan li-dãtihi , fa-kullu kalāmin ģumlatun wa

ence between ģumlah and kalām is that ģumlah

59 Interjections and exclamations standing alone

Zamaxšari with relation to their analogy to sentences
ing with them as forms (al-Zamaxšari: Mufaççal 617
168ļ).

This content downloaded from

79.123.184.37 on Sat, 01 Jan 2022 10:08:04 UTC
All use subject to https://about.jstor.org/terms
 60 Gideon Goldenberg

nexus whether (that ģumlah) is intende

not, like the predicate clause &c. [

basic nexus and is intended for itself; thus ever

not vice versa" (al-Raçlî: ŠarĶ al-Kãfiyah I 822-
An accentuated discrimination between kalām
found in the writings of Ibn Hišām al- 'Andari. B
mon mistake made by considering the two te
Ibn Hišām expressly distinguished between the
its technical sense, al-malnà l-i$ļilāķī60) as a sayin
ful" ('communicative' or 'informative') by intentio
which denotes a meaning after (the expression
be silent,62 whereas ģumlah is a predicative const
capable of being intelligibly uttered by itself. For
relation, Ibn Hišām is, truth to say, dependent on
the sentence", as he refers (Ibn Hišām: Muģnī C
the constructions /¿7 -fā'il and mubtada y - xa
and fails to employ the readily available terms
ģumlah is also commonly used for dependent c
"protasis", ģumlat al-ģawāb "apodo sis", or ģu
clause", which do not stand the test of Hfādah, it
Ibn Hišām' s view, that while eveiy kalām is a ģum
is a kalām ; in his words: the term kalām is m
ģumlah more general (Varam) (IbnHišām:M^m
It will be noticed that Ibn Hišām' s notion of kalām is narrower than Ibn
öinni's (see above, § 7).
Relevant to the discrimination between kalām and ģumlah is the
enlarged predicate construction, where the predicate is itself a ģumlah
whose predicate is again a ģumlah.6* Ibn Hišām explains the division of
ģumal in such constructions into two types: ģumlah kubrà, being a com-
plex sentence (whose predicate is itself a ģumlah), and ģumlah çugrà,
being a pedicate clause (which is a part of the former) ; one ģumlah can
be kubrà with respect to a smaller one which it includes, and at the same
time it can be $uģm with respect to a larger ģumlah . This division can be
represented by the following bracketing of Ibn Hišām' s own example:

60 Ibn Hišām: Šarh Šudūr al-dahab 1810_i5-

61 al-kalāmu huwa l-qawlu l-mußdu bi-l-qa§di (Ibn Hišām: Muģnī C 3748 = D
49O5).
62 mā dalla lalà maínan yahsunu l-sukūtu ialayhi (Ibn Hišām: Muģnī C 3749 = D
4905_6). "Silence" as a sentence boundary marker had already been used by
Sïbawayh, v. Sîbawayh I P 22217 = B 2617.
63 This could also be described as a chain of extrapositions.

This content downloaded from

79.123.184.37 on Sat, 01 Jan 2022 10:08:04 UTC
All use subject to https://about.jstor.org/terms
 Subject and Predicate in Arab Grammatical Tradition 61

[Zaydun [' abūhu [ģulāmuhū muntcdiqun]]]

1 2 3

[1] is kubrà; [3] is ņuģr

respect to [1]. As [1] f
is also referred to as ģu
4974_14).
Of all Arab grammarians, Ibn Hišām was the most anxiuos to empha-
size the difference between kalām and ģumlah. He even criticized al-
Zamaxšari for having identified (in his Mufaççal § 1 , see above) the mini-
mal form of kalām as ģumlah , as if that identification should imply con-
fusion of the two terms (Ibn Hišām: Muģnī C 37413_15 = D 49O10.n),
despite the fact that a statement similar to al-ZamaxSari's was else-
where made by Ibn Hišām himself.66 As to the characterization of predi-
cative relation, Ibn Hišām failed to adopt the notion and terminology of
predication as such, which in the works of his predecessors had made an
integral part of the conception of ģumlah.

IV. The verb-form as a nexus-complex

9 As inflected verb-forms necessarily include besides the predicate
base some representation of the subject, predicative relation is also
implied within the finite verb, and its treatment in Arab grammatical
literature is relevant to the present study. The analysis of the finite verb
as a nexus-complex comprising (1) indication of person, (2) lexeme
expressing the predicate, and (3) predicative nexus, has recently been
discussed at some length in my disquisition on Verbal structure.61 If in
general a verb-form minus its personal marker is virtually nominal (v.
ibid. § 10), in Semitic languages especially this fact still appears dis-

64 [3] can also be a verbal sentence.

65 Another occurrence known to me of the expression ģumlatu l-kalām in the
grammatical literature is in Ibn al-Sarrāģ: *Uçûl II 3743; there, however, it refers
to an object-clause presenting speech (ķikāyah): "(said (that):)

66 al-kalāmu qawlun mußdun [

wa-smin, wa-yusammà ģumlatan ismiyyatan ^aw fťliyy

"kalām is a 'useful' ('informative' or so) saying [

tion is of two nouns, or a verb and a noun, and it is cal

bal ģumlah depending on its beginning" (Ibn Hišām: G
67 Goldenberg: Verbal Structure . Besides the historical account of the
various ideas concerning the nature and constitution of verb-forms, this inquiry
is mainly intended to show how the constituents of the verb-complex are autono-
mous in their syntactic behaviour in the cases of negation, extraposition, focus-
ing, nominalization, and adverbial complementation.

This content downloaded from

79.123.184.37 on Sat, 01 Jan 2022 10:08:04 UTC
All use subject to https://about.jstor.org/terms
 62 Gideon Goldenberg

cernible (with a little effort) on th

fact, Semitic verb-forms have long be
nal sentences: "The verbal forms o
forms, mostly in combination with p
so to say, a sentence, consisting of a n
ually been contracted or shrivelled
Lectures (1890) 16418_22); "What ar
morphology, derivable either directly
termed "nominal sentences"" (Hodge: Nominal Sentences (1975) 1
[= 69]). 68 To the Arab grammarians it looked evident that verbal forms
involved an indication of the agent (fãHl ) and that inflected verbs were
the outcome of agglutination; they too would say that the verbal base
and the agent marker were shrivelled up into a single word: Huna l-filla
lammã kãna lã yaxlü min al-fā'ili wa-lã yastagnï ť arihu (ļarūratan, ļumma
ttaņala bihi muiļmarun $āra ka-ba'iļi ķurūfihī, wa-$ārat-i l-ģumlatu kali-
matan wãkidatan "as the verb would not be devoid of the agent and,
necessarily, would not dispense with it, a personal pronoun was attach-
ed to it which became like one of its letters, and th(at) ģumlah became
one word" (al-Zaģģāģī: Ūfļāh 755_6).69 In this regard there will be found
some typical inconsistency in terminology: sometimes it is said that the
agent (fā'il) has the status of a part of the verb (fill), 70 sometimes the
same idea is expressed otherwise, by saying that the verb (/¿7) and the
agent (fā'il) have the status of one word,71 all that because 'fiT is alter-
nately used for verb-form and for verbal base. Somewhat complicated is
the question in what positions the 3rd pers. agent marker should be con-
sidered pronominal. This is left to be discussed in brief in an excursus to
the present chapter, as our main concern here is the predicative relation
within the verb-form as treated in Arabic grammar.

68 Some points relevant to the general comparison of syntactical and morpho-

logical constructions were studied in Buccellati: Stative.
This process should account for the fact that in the Imperfect the modal
endings follow immediately the suffixed part of the agent marker and not the
verbal base to which they naturally belong, also the fact that in the Perfect the
modal endings are omitted before the afformatives (ibid. & elsewhere).
70 waJlam ianna l-fāHla ka-l-ģuzH min al-fťli (al-6urģānī: Muqtaçid I 32717,
32819) ; al-fãHl yatanazzalu manzilata 1-ģuzH min al-fi'li (Ibn al->Anbāri: Luma ť al-
* adillah 6516; cf. id.: 'Asrār 3514_15 = 27914); fa-law lam yunazzilü (ļamīra l-fãHli
manzilata ķarfin min sinxi l-filli. . . "and if they had not put the agent pronoun in
the status of a letter of the verb's root . . (Ibn al-'Anbārī: ' Asrãr 35i8_i9 =
280s.6).
71 fa-law lam yatanazzal-i l-fťlu wa-l-fãHlu bi-manzilati l-kalimati l-wāhidati
(Ibn al-Anbārī: Luma ť al- adillah 6710, 67!).

This content downloaded from

79.123.184.37 on Sat, 01 Jan 2022 10:08:04 UTC
All use subject to https://about.jstor.org/terms
 Subject and Predicate in Arab Grammatical Tradition 63

10 An inflected verb-form is clearly analysed in the same terms as a

sentence: in (ļarabtu "I struck", the $arò "striking" is the predicate
(xabar) and the 1st sg. afformative - tu (which is regarded as nominal, v.
infra) is the subject (al-muxbar lanhu). In the following mas^alah, quoted
by al-Zaģģāģī in the name of 'Abu l-'Abbās (al-Mubarrad?), the parallel
treatment of a verb-form and a nominal sentence is especially explicit:
al-farqu bayna (ļarabtu Zaydan wa-Zaydun (ļarabtuhū, 'annaka yidā quita
(ļarabtu Zaydan , fa-innamā *aradta *an tuxbira lan nafsika , wa-tutbita
*ayna waqala fťluka. wa-idā quita Zaydun (ļarabtuhū , fa-innamā ^aradta
'an tuxbira can Zaydin "the difference between (ļarabtu Zaydan "I struck
Zayd" and Zaydun (ļarabtuhū "Zayd - I struck him" is that when you
say (ļarabtu Zaydan your intention is to predicate of yourself stating
where your deed took place, and when you say Zaydun (ļarabtuhū your
intention is to predicate (sth.) of Zayd" (al-Zaģģāģī: 7 <ļah 1367- 137 ļ).
The 1st sg. preformative of the Imperfect would also be termed 'alifu l-
muxbiri lan nafsihï "the 'alif of him who predicates of himself" ('Abū
Bakr b. al-'Anbārī: yAlifāt 2836.7, 4502. 6. 8. 11-12 = 27?6, 8I9. n. l2. 16. is)-
There is no need to repeat time and again that a verb-form is regarded
as including its agent's marker. Some further evidence, however, show-
ing that ordinary terms of predication are used for analysing verb-forms
may be quite apropos. Some such instances are here following. Even the
3rd sg. zero-marker of the Perfect is taken for a subject: Hn xabbara can
dakarin kānat calāmatuhū fi l-niyyati "if he [= the speaker] predicates
(sth.) of a male, his [= the male's] marker is understood [lit. in the in-
tention]" (al-Mubarrad: Muqta<ļab 1 2627 = 23975).72 The pronoun which
is in the verb-form is characterized as being equivalent to a subject
noun: al-fā'il [

nasabta dãlika l-ficla Hlà dãlika l-ismi [

baidahū fa-huwa mufļmarun fihi "the agent [

tion after a verb predicating and attributing tha

[

(Ibn öinni: Luma ť 13 2-7) ; wa-mu(ļmaruhū fi l-isnā

"and the pronominalized [form of the agent], wi
subject of predication, is like its overt [form]"
1 1 1 i_i2 (§ 21) = 18u; cf. Ibn Yaťiš: Šarh L 92ļ
The definition of noun by its capability of functi
predication (bi-1-ķadīti canhu or bi~lJisnādi Hlay
the most useful of all noun characterizations p
abstract and can be employed for recognizing th
72 This is a rephrasing of the statement made on the
P 330 = B 376.

This content downloaded from

79.123.184.37 on Sat, 01 Jan 2022 10:08:04 UTC
All use subject to https://about.jstor.org/terms
64 Gideon Goldenberg

which do not take the definite art

/-ti) Wannaka Ķaddatta lan-i 1-tãH
gay ru l-tā'i min al-<ļamāHri ka-cļ
struck" [or -ta/-ti], because you h
the act of striking, and it [= the a
the pronouns other than -tu [-ta /-t
bīnī: Nūr 244_6, & v. supra § 5).
The verb-form is thus analysed
relation between its constituents is
which are the general terms of p

11 Excursus: The status of inflexional affixes

The rather intricate rules of agreement which typify A

raise some questions regarding the status of inflexional af
ous verbal positions.
There are sentences (A) where the subject noun is explicitly
expressed; (B) where the subject is already known from the context, its
pronominal representation (3rd pers.) should suffice. What complicates
the matter is the fact that in the agreement required between subject
and predicate in Arabic, number and gender do not go together. For
examining the problems relevant to the status of verbal formatives, the
following partial comparison of cases A and B may be helpful ( qãm -
"stood"; Zayd mase. pers. name, susceptible (like all names) of being
made dual or plural; Hind fem. pers. name; -ā(ni) dual suff.; -ū(na)
mase. pl. suff.):
A B
qãma Zaydun qãma
qãma I- Z ay dāni qāmā
qãma l-Zaydüna qämü
qãmat Hindu(n) qãmat
qãma l-yawma Hinduin)1 3 qãmat

The grammarians would also consider in this connexion the fo

non-standard construction usually known as luģat yakalü
Those who used this form would say:

73 Cf. Sîbawayh I P 2023_4 = B 23512_i3; al-Mubarrad: Muqtcu

337 N2; al-Zaģģāģ: Trãb al-Qur*ān II 612-615. This variety is bro
show that also the agreement in gender, even for the natural femin
tain cases optional, thus not completely alien from the agreemen
where variation is not unknown (see below).

This content downloaded from

79.123.184.37 on Sat, 01 Jan 2022 10:08:04 UTC
All use subject to https://about.jstor.org/terms
 Subject and Predicate in Arab Grammatical Tradition 65

A B
qãmã (a) I- Z ay dāni qãmã
qämü ( a) I- Z ay dūna qämü
On the status of the inflexional suffixes in Arabic there was no com-
mon assent among the grammarians. Two opinions are mentioned by
Ibn Ya'īš ( Šarh L 40918_22 = C III 8720_24 & L 4103_5 = C III 883.5): One
was held by Sîbawayh, and the other attributed to al-Māzinī (For other
views v. al-Batalyõsi: lI$lāh al-xalal 151-152). In Sïbawayh's opinion,74
the verbal suffixes -ã and -ü &c. are sometimes personal pronouns
(nouns representing the suppressed) and sometimes particles indicat-
ing number and gender. The same applies to -ã- and -ü- in Imperfect
forms like yaqümäni and yaqümüna (v. infra) . Pronouns they are when
they have the same syntactical position as the overt nouns:
VNVN VNVN

qãm-ã || qãma l-Zaydãni, qäm-ü || qãma l-Zaydüna

-ã and -ü here represent nouns already mentioned and thus known from
the context; in the sentences qãm-ãt qām-ū the suffixes have the posi-
tion of the agent noun, which must follow its verb. Both qãma l-Zaydūna
and qämü are then to the recognized as sentences constructed as Vb.
pred. + agent noun (muçhar" overt, explicit", or muýmar" covert, kept in
the mind").
Since N + V sentences with dual/plural N require the same verbal suf-
fixes: al-Zaydãni qãmã , al-Zaydüna qämü , they are necessarily ana-
lysed as nominal sentences whose predicates are verbal sentences75 -
qãm-ã , qäm-ü - comprising Vb. -I- pron. agent. Both in qãma "he stood"
uttered by itself and in Zaydun qãma the verb-form qãma is accordingly
said to have a zero- suffix as agent, whereas in qãma Zaydun , qãma l-
Zaydãni , qãma I- Z ay dūna the verb-form is bare.
The parallel forms in the Imperfect are treated by Ibn al-Sarrāģ: *U$ūl
I 6915-709: yaqümu in Zaydun yaqümu is regarded as including a prono-
minal agent which has no explicit marker; in al-Zaydãni yaqümäni / al-
Zaydüna yaqümüna the agent pronoun finds its expression in the aflbr-
matives -ã- / -ü - resp.,76 while the same suffixes are just nominal dual/
plural markers (tatniyat al-ism / ģamc al-ism) when added to the predi-
cative participles.

74 Cf. Sïbawayh I P 20116-202! = B 2353_10. The presentation of Sïbawayh's

doctrine as brought here does not follow exactly the reasoning and the examples
in this passage of the Kitãb or in Ibn Ya'īš.
75 Or analogous to verbal sentences, see above S 7.
76 The Imperfect preformatives are not considered at all in this connexion.
5 ZDMG 138/1

This content downloaded from

79.123.184.37 on Sat, 01 Jan 2022 10:08:04 UTC
All use subject to https://about.jstor.org/terms
 66 Gideon Goldenberg

In Hindu(n) qämat , as in qãmat uttere

sidered pronominal, but in qämat Hin
follows, the inflexional suffix in qäm
agent (because agent there can be on
announcing in advance the gender o
(see above, fn. 73), this suffixed har
does not precede the agent noun im
The construction of yakalüni l-barāģ
verb-form with an inflexional suflix th
announcing the gender and number
12 According to the other view, a
grammarians, the suffixes -ã as in q
particles and not pronouns, that is to
harf and not to the category ism, un
sumed; al-alif fi qãmã wa-l-wãw fi qäm
wa-l-fãHlina l-mu(ļmarīna, wa-l-fãHlu f
Zaydun qãma fa- fi qãma (ļamirun fi
zahiratun. fa-idã tunniya *aw ģumi'a
gayra *anna lahü 1 olámat an "the yalif
mü are two particles (ķarfāni) marki
agents,77 and the agent is 'in the int
qãma there is in qãma a pronoun 'in t
indication. And when formed in the du
'in the intention' except that it has
4103_5 = C III 883_5). According to th
are implied, but not expressed, in the v
assuming the same suffixes to be
removed.
The idea that the inflected verb-form implies a "pronomen intentum"
while the inflexional affixes just allude to the agent by marking its gen-
der and number may well be recognized in the arguments of Ibn
Ma<Jā'78 who otherwise represents whatever is contrary to al-Māzinī's
attitudes.79 Ibn Maijā* would mostly regard the notion of agent as in-
herently present in the verb, like those of accident (ķadat) and time

77 muýmar is the word (here: the agent noun) which is "kept in the mind" sup-
pressed, and not explicitly expressed.
78 Compare especially Ibn Yaťiš: Šarh L 4104_5 = C III 884_5 with Ibn Maijā*:
Radd 1043_6 = 28213_16 (< ļamirun fi l-niyyati || fi nafsi l-mutakallimi <ļamir ).
79 Al-Māzinī was the great propagator of the technique of grammatical forma-
tion exercises ( tamarin , cf. infra ch. V), which Ibn Maijā* sharply opposed.

This content downloaded from

79.123.184.37 on Sat, 01 Jan 2022 10:08:04 UTC
All use subject to https://about.jstor.org/terms
 Subject and Predicate in Arab Grammatical Tradition 67

(, zamān ), and reject speculated pronominalization; see Ibn Magia1: Radd

103-106 = 28115-84, cf. ibid. 1006_12 = 27921-804.
Whether the agent is pronominally represented by the inflexional affi-
xes, or inherently present, though in a subauditur, within the verb-form,
the predicative bond embodied in the verb is not called in question.

V. 'Ixbar

13 A common syntactical exercise well developed in Arab

was the turning of an extracted part of a sentence into a
modern terminology this transformation could approximately
bed as one type of rhematization, or focusing, or "la mise en v
the formation of a cleft sentence. None of these terms is perf
others are not easy to form.81 The expression used for the tam
ing some word X into a predicate is: *axbir lan X "turn X
cate!".82 For putting yixbār into practice, the rest of the sent
be nominalized to be made the subject (mubtada?) ofthat se
actually the case in the formation of cleft sentences. The nom
can be morphological (-► e.g. al- + participle) or syntacti
allodi + vb.). qama Z ay dun "Zayd stood" (making Zayd pre
qa*imu Zaydun / allodi qãmo Zaydun ; (ļorabtu Zaydan "I s
(-tu -+ Pred.) - ► al-(ļāribu Zaydan yanā / allodi doraba Z
doraba * Abdullãhi 'axãka " 'Abdallah struck your brother" ('yo
- ► Pred.) -► al-(ļāribuhū 1 Abdullãhi yaxüka / allodi (ļarabah
yaxūka' Zaydun *axüka "Zayd is your brother" (Zayd -> Pr
huwa *axuka Zaydun &c.

80 "Rhematization", "focusing" and "la mise en vedette" are not

stricted to the formal syntactical function of the grammatical pre
sentence", which literally presupposes the cleaving of an underlyin
struction", is not the best term for a language like Arabic, where
fall perfectly into line with all normal nominal sentences, though
literal meaning this term is impeccable.
81 "Predication" is here improper, as it means something else, and
is Howell's (Grammar I 608-614, § 179) "enunciation" for Hxbãr ,
his "enunciative" (xabar). " Prédicat ( i v)ization" or the like would
exact or clumsy.
82 It will be noticed that the meaning of the construction 'axbara
usage [= ģa*ala X xabarari] is different from the outwardly identica
in its ordinary usage treated above [= "predicated sth. of X", wh
ject] . This double usage was perplexing enough to require some c
in fact will be found in Ibn 'Aqil's commentary on the Alfiyyah (Ib
308-31 1 = C 281-286 (297-301) = Ca IV 60-66 ad Ibn Mālik: Alfi
717-725).

5*

This content downloaded from

79.123.184.37 on Sat, 01 Jan 2022 10:08:04 UTC
All use subject to https://about.jstor.org/terms
 68 Gideon Goldenberg

Since the act of Hxbãr involves the nominalization and the ibtida*
"fronting" of the rest of the sentence, the transformation in question
was variously termed Hxbãr^an . . .), ibtidã' al-alif wa-l-lãm, al-ixbãr
bi'lladi wa-bi-( l-alif wa-) l-lām &c.83 The grammarians most interested
in the syntactical exercises of Hxbãr were al-Māzinī, al-Mubarrad, Ibn
al-Sarraģ, and al-Fārisī. Al-Māzinī, who also developed the morphologi-
cal excercises,84 was, so it seems, the initiator of the syntactical tamàrïn
of formal rhematization as a research model. Unfortunately, his Kitãb
al-alif wa-l-lãm has not been preserved, but his views are quoted at
length in the writings of al-Mubarrad (Muqta4ah III 1 14, 117, 127-129)
and Ibn al-Sarrāģ (yU$ül II 312, 330u-33116). Vzòâr-transformations
with their intricacies are discussed in detail in al-Mubarrad: Muqtajfab
III 89-132 ( bob al-ibtida* wa-huwa lladï yusammïhi l-nahwiyyūn al-alif
wa-l-lãm "chapter on fronting [= thematization] which the grammar-
ians call al-alifwa-1-lãm" and the following chapters), ibid. IV 133-135
(within the chapter on al-musnad wa-l-musnad Hlayhi ), ibid. IV 352-
353; in Ibn al-Sarrāģ: *U$ül II 310-381 (chapters on yixbār of various
parts followed by bab masa" il 'min] al-alif wa-l-lãm) ; and in al-Fārisī:
yī(ļāķ II 1145-1 166. 85 Also deserve being mentioned al-črurģānī's thor-
ough commentary on the *ixbār chapter in al-Fārisī' s Ū(ļāh' al-črurģānī:
Muqtaçid II 1145-1166, and al-Zamaxšarī: Mufaç$al § 179 5719- 5815 =
1443-14513 with Ibn Yaťiš: ŠarĶ L 4717-47523 = C III 156-160. For the
treatment of syntactical tamãrin in the later literature v. Ibn Mālik:
yAlfiyyah verses 717-725 with Ibn lAqīl: Šarķ L 308-3 1 1 = C II 281-286
(297-301) = Ca IV 60-66; of greater importance are Ibn TJçfur: Šarh al-
Gumal II 494-512 and al-Racļī: Šarķ al-Kãfiyah II 43apu-53 23.86

83 Ibn Hišām ( *Aw(ļah al-masālik III 209 ļ j ) also mentioned an alternative term
for bãb al-ixbār bi-lladī . . . wa-bi-l-alif wa-l-lãm, namely bab al-sabk (see also
al^Ušmūnī: Manhaý III 611 pu-u). Ibn Hišām, there, speaks of such exercises as
intended for practical training.
84 huwa lladï fataha baba l-tamārīni ģayri l-lamaliyyati fi l-çarfi lalà mu'âçirîhi
(Qayf: Madāris 1195).
85 The refernce given here, as all references to al-Fārisī' s Ū(ļāķ in_the present
article, is to al-(5urģānī: Muqtaçid, which includes the text of the Ū(ļāh with al-
črurģānī's commentary. Bab al-ixbār bi-lladī wa-bi-l-alif wa-l-lãm will be found in
l_al-Fārisī: Al-I(ļāh al-'afļudī. Ed. IJasan Šādilī Farhúd. Cairo 1969, pp. 57-62.
86 From among the Arabic grammars written by Westerners de Sac y' s is per-
haps the only one where the traditional practice of Hxbãr is really reflected (de
Sacy: Grammaire II 282-289, §§ 493-498 = 2349-356, §§ 604-609). Vernier's
meagre notes (Vernier: Grammaire 27 0-27 2 § 833, 422 § 977) are hardly com-
parable. The practice of syntactical rhematization-exercises is not known to
have met any strong criticism such as that which the morphological tamãrin had
to face (cf. al-0urģānī: Dalā'il 24, Ibn Macļā*: Radd 161-164 = 2135-137).

This content downloaded from

79.123.184.37 on Sat, 01 Jan 2022 10:08:04 UTC
All use subject to https://about.jstor.org/terms
 Subject and Predicate in Arab Grammatical Tradition 69

The syntactical tamārin were intended to examine the formal rhemati-

zation of any extracted word or morpheme. Even the Hxbãr of the inner-
object infinitive (mafül mutlaq) was considered, and (unless the inf. is
followed by an adjective) rejected, as it denotes nothing more than the
verb itself (see al-Mubarrad: Muqta4ab III 103-105, Ibn al-Sarrāģ: yU$ūl
II 310-312). As a matter of course, any Hxbãr of the verb could not be
contemplated, since nafsuhü xabarun (v. supra § 5). Many inter-
esting points of syntax were treated brilliantly in the chapters on al-
*ixbār bi-lladī wa-bi-l-alij wa-l-lãm ; some others suffered from the
grammarians' tendency to over- speculate on possible transformations
rather than seek those that are observable in actual use. It is, however,
certain that for the grammarians who dealt with Hxbâr as the syntacti-
cal exercise, the question of where in the sentence there lies the predica-
tive axis was of central importance: in <ļarab ' tu Zay dan the speaker, as
represented by - tu , is the subject; with Zayd topicalized the sentence is
transformed into Zaydun ' (ļarabtuhū (al-Zaģģāģī: Ū(ļāh 1367- 137 j ) , and
with Zayd rhematized - into allodi (ļarabtuhū | Zaydun or the like.

References

al-'Abhari: Eisagõgè Arab. - Arabic version by 'A£ir al-Dīn al-'Abharī

of Porphyry's (d. c. 303) Eioaytoyri [Introd. to Artistotle: Categories ].
ces are to (1) Îsâgûcî qerhi f eneri. Der Saadet 1310 [1892/3], and to
*Abī Yahyà Zakariyyã al- Armàri: ai-Muyala1 lalà matn Usāģūģī fi l-m
*Atir ai-Din al-Abhari. Cairo 1351/1933.
'Abū Bakr b. al-'Anbārī: Alifāt. - 'Abū Bakr Muhammad b. al-Qāsim b. Baššār
b. al-'Anbārī (885-940): Kitāb šarķ al- alifāt. Ed. jAbü Maģfū? al-Karīm
al-Ma'ģūmī. In: Maģallat al-Maģma' al-Tlmi al-'Arabī [Revue de l'Académie
arabe de Damas] 34 (1959), 273-290, 447-461; 2ü Zitāb muxta$ar fi ļikr aljali-
fãt. Ed. IJasan Šādilī Farhúd. In: Maģallat Kulliyyat al-'ādāb, črámiťat al-
Riyâçl [Journal of the College of Arts, University of Riyadh] 6 (1979), 69-94.
Aristotle: Categories. - Aristotle: Herrn. - Ilepi ep|ir|veíaç. - References to Aris
totle's text are to the Bekker edition oí Aristotelis apera omnia. Berlin 1831.
al-Batalyõsi: 7 çlâh al-xalal. - }Abū Muķammad 'Abdallah b. Muh. b. al-Sīd
al-Batalyosī [or: -yūs-/-yaws-] (1052-1127): Kitāb al-hulal fi 'içlãh al-xalal
min kitāb al-Gumal [Commentary on al-Zaģģāģī: öumal]. Ed. Sa'īd (Abdal-
karīm Sa"üd!. Baghdad 1980 [printed in Beirut].
Beeston: Rev. Mosel. - A. F. L. Beeston: [Review of] U. Mosel: Die syntak-
tische Terminologie bei Sïbawaïh. In: BSOAS 39 (1976), 648-653.
Boetius: Herrn, comment, maj. - Manlius Severinus Boetius (c. 480-524): In
librum Aristotelis de interpretation editio secunda , seu majora commentarla. In:
Patrologia Latina. Ed. J. -P. Migne. 64. Paris 1891, 393-640.
Boetius: Herm. comment, min. - Manlius Severinus Boetius: In librum Aristotelis
de interpretations editio prima, seu minora commentaria. In: Patrologia Latina.
Ed. J. -P. Migne. 64. Paris 1891, 293-392.

This content downloaded from

79.123.184.37 on Sat, 01 Jan 2022 10:08:04 UTC
All use subject to https://about.jstor.org/terms
 70 Gideon Goldenberg

Buccellati: Stative. - G. Buccellati: An

tive as a Nominal Sentence. In: JNES
Canard: Fatāli. - Marius Canard: La fo
35), 5-72.
Carter, M. G. v. al-Sirbinī: Nūr.
Cohen: Formes du prédicat. - David Cohen: Les formes du prédicat en arabe et la
théorie de la phrase chez les anciens grammairiens. In: Mélanges Marcel Cohen.
Ed. D. Cohen. The Hague - Paris 1970, 224-228.
Payf: Madãris. - Sawqï Çayf: Al-madãris al-nahwiyyah. Cairo 1968,
41979.
al-Fārābī: Herrn, comment. - Sarķ al-Fārābī [d. 950] li-kitāb Arisļūļālīs fi l-Vbā-
rah - Alf ar abi' s Commentary on Aristotle's Ilepi epfiTļveiac; (De Interpretations).
Ed. Wilhelm Kutsch and Stanley Marrow. Beirut 2 1971.
al-Fārisī: U<ļāh. - 'Abū ťAli al-Qasan b. 'Ahmad b. ťAbd al-Gafíar al-Fārisī (901-
987): Kitāb al-I<ļāh. (al-lA<ļktdī) fi l-nahw. Refs. apud al-0urģānī: Muqtaçid
(& cf. footnote 85).
Fleisch: Traité. - Henri Fleisch: Traité de philologie arabe. 2 vols. Beirut
1961-1979.
Fleisch: Verbe. - Henri Fleisch: Etudes sur le verbe arabe. In: Melanges Louis
Massignon. 2. Damascus 1957, 153-181.
Fleischer: Kleinere Schriften. - Heinrich Leberecht Fleischer: Kleinere
Schriften. 3 vols. Leipzig 1885-1888.
Goldenberg: Verbal Structure. - Gideon Goldenberg: On Verbal Structure
and the Hebrew Verb [in Hebrew]. In: Language Studies. 1. Ed. M. Bar-Asher.
Jerusalem 1985, 295-348.
al-0urģānī: DalãHl. - (Abd al-Qãhir al-0urģānī (d. 1078): Kitāb Dalā'il al-Hlģāz.
Ed. Muģammad 'Abduh & Muģammad Maģmūd al-Tirkizï. Cairo: al-
Manār Í1321/19031.
al-0urģānī: Gumal. - lAbd al-Qāhir al-0urģānī: Al-Gumal. Ed. (Alī Ha yd ar.
Damascus 1392/1972.
al-0urģānī: Muqtaçid. - ťAbd al-Qāhir al-0urģānī: Kitāb al-Muqta$id fi šarh a
U(ļāh [Comment, on al-Fārisī: *I<jāĶ'. Ed. Kā? im Baģr al-Maröän. 2 vols.
Baghdad 1982 [printed in Amman].
Hodge : Nominal Sentences. - Carleton T. Hodge : The Nominal Sentence in
Semitic. In: Afroasiatic Linguistics 2/4 = vol. 2 (1975), 69-75.
Howell: Grammar. - Mortimer Sloper Howell: A Grammar of the Arabic
Language translated and compiled from the Works of the most approved native or
naturalized authorities. 4 parts in 7 vols. Allahabad 1880-1911.
Ibn al-'Anbārī: Asrār. - 'Abu 1-Barakāt ťAbd al-Raķmān b. Muhammad b. al-
>Anbārī (1119-1181): 'Asrār al-Arabiyyah. Ed. C. F. Seybold. Leiden 1886;
2Ed. Muģammad Bahöah al-Bītār. Damascus 1337/1957.
Ibn al-'Anbāri: Unçãf. - 'Abu 1-Barakāt ťAbd al-Rahmān b. Mub. b. al-'Anbārī:
Kitāb al-Inßäf fi masa' il al-xilāf bayn al-naķwiyyīn al~Ba$riyyīn wa-l-Kü-
fiyyin. - L Die grammatischen Streitfragen der Basrer und Kufer. Ed. Gott-
hold Weil. Leiden 1913. - C Ed. Muņammad Muģyī al-Dīn 'Abd al-
Hamīd. Cairo 1364/1945.
Ibn al-'Anbārī: Luma ť al-adiūah. - 'Abu 1-Bakarāt ťAbd al-Rahmān b. Mufe. b.
al-'Anbārī: Luma ť al-adülah fi iu$ūl al-nahw. Ed. Attia Amer. Stockholm
1963.

This content downloaded from

79.123.184.37 on Sat, 01 Jan 2022 10:08:04 UTC
All use subject to https://about.jstor.org/terms
 Subject and Predicate in Arab Grammatical Tradition 7 1
Ibn ťAqil: Šarķ. - 'Abdallāh b. {Abd al-Raķmān b. 'Abdallāh Bahā' al-Dïn ibn
ťAqil (d. 1367): Commentary on Ibn Mālik: Alfiyyah. - L Alfijjah carmen
didacticum grammaticum auctore Ibn Mâlik et in Alfijjam commentarius quem
conscripsit Ibn ' Aķīl . Ed. Fr. Dietrici. Leipzig 1850-1851. - C Šarh Ibn lAqil
lalà Alfiyyat Ibn Malik. Ed. Muhammad Muģyī al-Dïn 'Ab d al-IJamīd.
Cairo 1931 (1935) [references marked C are to the Cairo editions of 1931 and
(in brackets) 19351; Ca by the same editor, 4 vols. Cairo 1386/1967.
Ibn Fāris: $aķibi. - 'Abu 1-IJusayn 'Aļimad Ibn Fāris (d. 1005): Al-$āhibi fi fiqh
al-luqah wa-sunan al- Arab. Ed. M. el-Chouémi. Beirut 1382/1963.
Ibn öinni: Lama1. - 'Abu 1-Fath 'Utmān Ibn öinni (915-1002) : Kitãb al-Lumalfi
l-naķw. Ed. Hadi M. Kechrida. Uppsala 1976.
Ibn öinni: lUqūd. - 'Abu 1-Fatķ cUtmān Ibn öinni: ť XJqüd al-Luma1 fi l-nahw. Ed.
IJasan Šādilī Farhüd. In: Maģallat Kulliyyat al-'ādāb, öämi'at al-Riyā<ļ
[Bulletin of the Faculty of Arts, Riyad University] 5 (1977-1978), 135-
154.
Ibn Ginnī: Xa$ãyiç. - 'Abu 1-Fath 'Utmān Ibn Ginnī: Al-Xaçâ'iç. Ed. Muģammad
(Alī al-Naööär. 3 vols. Cairo 1371-1376/1952-1956.
Ibn al-Hāģib: Kãfiyah. - Gamāl al-Dīn 'Abū ťAmr (Utmān b. (Umar Ibn
(1174-1249): Kitãb al-Kãfiyah fi l-nahw , ap. The Five Books on Arabic
mar. Ed. John Baillie. Vol. 3: Comprising theKafeea of Ebn-e-Hajeb <b
cutta 1805.
Ibn Hišām: Aivfļah al-masālik. - Gamāl al-Dīn Ibn Hišām al- ' An § āri ( 1308-
1360): *Aw(ļah al-masālik "ilà ' Alfiyyat Ibn Mālik. Ed. Muhammad Muhyî
al-Dīn 'Ab d al-Hamīd. 3 vols. 5Beirut 1966.
Ibn Hišām: õãmi' - öamäl al-Dīn Ibn Hišām al- 'Andari: Al- Gārni1 al-$aģīr fi Him
al-nahw. Ed. Muhammad Šarīf Sa'īd al-Zaybaq. Damascus 1388/1968.
Ibn Hišām: Muģnī. - öamäl al-Dīn Ibn Hišām al- 'Andari: Muģni l-labib lan kutub
al-alārib. - C Ed. Muhammad Muhyî al-Dïn 'Abd al-Hamīd. 2 vols. Cairo
n. d. - D Ed. Māzin al-Mubārak, Muģammad 'Alī IJamdallāh & Sa'īd
al- 'Afgānī. Damascus 1964.
Ibn Hišām: Sarh Sudūr al-dahab. - Gamāl al-Dīn Ibn Hišām al- ' An § āri: Sa
Šudūr al-dahab fi ma'rifat kalām al- Arab. Ed. Muģammad Mujjyî al-D
'Ab d al-IJamīd. 2 vols. Cairo 1364/1945.
Ibn Maçlâ': Radd. - 'Abu l-'Abbās 'Aļunad b. 'Abd al-Raķmān Ibn Maçlâ' al-Qur-
tubī (1119-1196): Kitāb al-Radd ťoZá l-nuķāh. Ed. Šawqī ī)ayf. Cairo 1366/
1947; 2Ed. Muhammad 'Ibrāhīm al-Bannā. Cairo 1399/1979.
Ibn Mālik: Alfiyyah. - Muhammad b. 'Abdallāh Ibn Mālik (1202-1274): (Al-
Xulā$ah) al- Alfiyyah (References by verse numbers), cf. Ibn ťAqil: Šarh.
Ibn al-Muqaffať: Mantiq. - Ibn al-Muqaffa' (8th cent., cf. footnote 20) : Al-Manfiq
(Logic) by Ibn Muqaffa' Iļudūd al-Mantiq (Definitions of Logic) by Ibn Bihriz.
Ed. Muņammad Taqî Dānešpažūh. Tehran 1978.
Ibn al-Sarrāģ: Mūģaz. - 'Abū Bakr Muhammad Ibn al-Sarrāģ (d. 928) : Al-Mūģaz
fi l-nahw. Ed. Muçtafà al-Šuwaymī (el-Chouémi) & Bensālem Dâmirôî
(Damerdji). Beirut 1965.
Ibn al-Sarraģ: 'Uçul. - 'Abū Bakr Muhammad Ibn al-Sarrāģ: Kitāb al-U$ūl fi l-
nahw. Ed. 'Abd al-Husayn al-Fatlī. 2 vols. 1: Naģaf 1972, 2: Baghdad
1973.
Ibn ťU$fur: Šarh al-Gumal. - 'Alī b. Mu'min b. Muķammad b. 'Alī Ibn 'U^fur al-
'Išbīlī (1200-1270): Šarķ Gumal al-Zaģģāģi (al-šarķ al-kabir). Ed. Çâhib 'Abü
Öanäh. 2 vols. Mosul 1400-1402/1980-1982.

This content downloaded from

79.123.184.37 on Sat, 01 Jan 2022 10:08:04 UTC
All use subject to https://about.jstor.org/terms
 72 Gideon Goldenberg

Ibn Yaťiš: Šarķ. - Muwaffaq al-Dīn Ya'ī

Commentary on al-Zamaxšari: Mufaçç
sarî's Mufaççal. Ed. Gustav Jahn. 2
MufaĢĢcd li-. . . Ibn Yaťiš. 10 parts.
'Ishäq b. Hunayn: Herrn. Arab. - Die Her
Übersetzung des Isķāķ ibnffonain [
references also given to the text of
Herm. comment.
Levin: Distinction. - Ar ye h Levin: The Distinction between Nominal and Verbal
Sentences according to the Arab Grammarians. In: Zeitschrift fur arabische Lin-
guistik 15 (1985), 118-127.
Levin: Musnad. - Ar ye h Levin: The Grammatical Terms al-musnad, al-musnad
Hlayhi and al-isnād. In: JAOS 101 (1981), 145-165.
al-Mayyāh: ' Isnād . - Rasmiyyah Muhammad al-Mayyah: "Isnād al-fťl.
Baghdad, date given variously as 1965, 1966, 1967.
Méhiri: Gumlah. - (Abd al-Qādir al-Mahīr ī:Al-ģumlahfi na%ar al-nuhãh al-
' Arab . In: Hawliyyãt al-öämi'ah al-Tūnisiyyah [Annales de l'Université de
Tunis] 3 (1966), 35-46.
Mosel: Term. Sīb. - Ulrike Mosel: Die syntaktische Terminologie bei Sibawaih.
Diss. München. 2 vols. München 1975.
al-Mubarrad: Muqta^ab. - Abu l^Abbās Muhammad b. Yazid al-Mubarrad
(825-898): Kitãb al-Muqta4ab. Ed. Muhammad ťABD al-Xāliq (Udaymah.
4 vols. Cairo 1385-1388 [c. 1965-1968]; I2 1399/1979 with pagination chan-
ged.
al-Mutarrizî: Miçbãh. - *Abu 1-Fatfr Nā^ir b. ťAbd al-Sayyid al-Mutarrizi (1144-
1213): Kitãb al-Miçbãh fi l-nahw, ap. The Five Books upon Arabic Grammar.
Ed. John Baillie. Vol. 1: Containing the Meeút Âmel, Shúrhú Meeút Âmel,
and the Mesbâh. Calcutta 1802.
Praetorius: Rev. Jahn. - Franz Praetorius: [Review of] Sîbawaihi's Buch
über die Grammatik . . . übersetzt und erklärt . . . vonG. Jahn. In: Göttingische
gelehrte Anzeigen 1894 (Nr. 9), 705-715.
Qabäwah: Trāb al-ģumal. - Faxr al-Dīn Qabäwah: Traò al-ģumal wa-aábâh
al-ģumal3 . Beirut 1401/1981.
al-Raçlï: Šarh al-Kāfiyah. - Kitãb al-Kãfiyah fi l-nahw ta^lïf. . . Ibn al-Ķāģib. . .
šaraķahū Ratļl al-Dīn Muhammad b. al-Ķasan al-Astarābādī (d. 1287). Istan-
bul 1310 [1893].
Rödiger: 5 Asma * al-afãl. - Ioannes Roediger: De nominibus verborum Arabi-
cis - J asmã 1 al-afãl. Halle 1870.
Rundgren: Einfluß. - Frithiof Rundgren: Über den griechischen Einfluß auf
die arabische Nationalgrammatik. In: Acta Societatis Linguisticae Upsaliensis
N.S. 2, 5 (1976), 119-144.
de Sacy: Anth. gramm. - A. I. Silvestre de Sacy: Anthologie grammaticale
arabe. Paris 1827-9.
de Sacy: Grammaire. - A. I. Silvestre de Sacy: Grammaire arabe. 2 vols.
Paris 1810, 21831.
Sîbawayh. - 'Abū Bišr (Amr b. ťUtmán Sïbawayhi (757?-c.796) : Al-Kitāb. - P
livre de Sîbawaihi: Traité de grammaire arabe. Ed. Hartwig Derenbourg.
vols. Paris 1881-1889. - B Kitãb Sïbawayhi. 2 vols. Bûlâq 1316-1318 [1898
1900].

This content downloaded from

79.123.184.37 on Sat, 01 Jan 2022 10:08:04 UTC
All use subject to https://about.jstor.org/terms
 Subject and Predicate in Arab Grammatical Tradition 73

al-Širbīnī: Nūr. - Šams al-Dīn Muhammad al-Širbīnī (d. 1570): Nūr al-saýiyyah
fi hall ì alfāz al-Āģurrūmiyyah. Arab Linguistics: An introductory classical text
with translation and notes. Ed. by M. G. Carter. Amsterdam 1981.
Talmon: Musnad. - Rafael Talmon: Notes on the Grammatical Terms musnad
and musnad Hlayhi: the historical perspective (unpublished) .
Talmon: Problems. - R. Talmon: Problems in the Syntax of Sentence Theory in
Sibawayhi's 'al-Kitab' [in Hebrew]. Diss. Tel-Aviv 1979.
Talmon: Tafkir. - R. Talmon: Al-Tafkïr al-nahwi qablaKitab Sìbawayh - dira
sah fi tãrix al-mu$'alah al-nahwï al- Arabi. In: Al-Karmil (Haifa) 5 (1984), 37-
53.
al-'Ušmūnī: Manhaģ. - 'Abu 1-IJasan 'Ali b. Muhammad al-'Ušmūnī (d. 1494):
Šarh al-Ušmūnī lalà Alf iy y at Ibn Mãlik al-musammà Manhaģ al-sãlik Hlà
3 Alf iy y at Ibn Mãlik. Ed. Muņammad Muçyî al-Dīn 'Abd al-Hamīd. 3 vols.
Beirut 1375/1955.
Vernier: Grammaire. - Donat Vernier: Grammaire arabe composée d'après
les sources primitives. 2 vols. Beirut 1891-1892.
Versteegh: Greek Elements. - C. H. M. Versteegh: Greek Elements m Arabic
Linguistic Thinking. Leiden 1977. (Studies in Semitic Languages and Lin
tics. 7.)
Weil: Verständnis. - Gotthold Weil: Zum Verständnis der Methode der mosle-
mischen Grammatiker. Ein Beitrag zur Geschichte der Wissenschaften im Islam.
In: Festschrift Eduard Sachau. Ed. G. Weil. Berlin 1915, 380-392.
Wright: Lectures. - William Wright: Lectures on the Comparative Grammar of
the Semitic Languages. Cambridge 1890.
al-Zaģģāģ: UWãb al-Quťán. - Träft al-Quryān al-mansub 'ila 1-Zaģģāģ [844-923].
Ed. Tbrāhīm al-'Abyârî. 2 vols. Cairo 1963-1964.
al-Zaģģāģī: öumal. - 'Abu 1-Qasim ťAbd al-Rafrmān b. 'Ishaq al-Zaģģaģ
949): Al-öumal. Ed. Mohammed Ben Cheneb [Ibn 'Abī Šanab]. A
1926-7.
al-Zaģģāģī: '/ (ļaJi. - 'Abu 1-Qāsim ťAbd al-Rahmān b. 'Ishaq al-Zaģģāgī: Al-I<ļa
fi Hlal al-nahw. Ed. Māzin al-Mubārak. Cairo 1959.
al-Zamaxšarī: Mufaççal. - 'Abu 1-Qāsim Mahmūd b. ťUmar al-Zamaxšarī ( 1075-
1144): Al-Mufa$$al, opus de re grammatica Arabicum. Ed. J. P. Broch. Chri-
stiania 2 1879; Kitãb «al-Mufa$$al» (fi Him al-Arabiyyah). Cairo 1323/1905.
al-Zawāwī: Fuçul. - Jaķjā ibn ťAbd el-Mu'tî ez-Zawâwî's [d. 1231] Kitâb el-
Fuçûl, Kap. I- II. Ed. Erik Sjögreen. Diss, Uppsala. Leipzig 1899.
Zimmermann: Log. Trad. - F. W. Zimmermann: Some Observations on al-
F arabi and Logical Tradition. In: Islamic Philosophy and the Classical Tradition.
Essays presented to Richard Walzer. Ed. S. M. Stern, A. Houraní & V.
Brown. Oxford 1972, 517-546.

This content downloaded from

79.123.184.37 on Sat, 01 Jan 2022 10:08:04 UTC
All use subject to https://about.jstor.org/terms