The Arabic Language

kaskasa
; according to Sībawayhi, this consisted 
in replacing 
-ki 
with 
-kis
. There may be a connection with the pronunciation of 
/k/ and /q/ before front vowels as /č/ and /j/ in contemporary Arabian dialects 
(Chapter 11, p. 194).
The preceding dialectal differences concerned phonetic or phonological differ-
ences between the dialects. There are some testimonies that refer to differences 

---

Arabic in the Pre-Islamic Period 
51
at the morphological or syntactic level. The existence of an undeclined dual in 
Ḥijāzī Arabic is sometimes inferred from the Qurʾānic verse 20/63: 
ʾinna hāḏāni 
la-sāḥirānī
‘verily, these two are sorcerers’, in which the particle 
ʾinna
seems to be 
construed with a nominative instead of the Classical accusative. This verse was 
a crux for the commentators. In the earliest period of Arabic grammar some of 
them even suggested that the nominative was a scribal error, which should be 
corrected by reading the accusative in the following noun. In connection with 
this, ʾAbū ʿAmr ibn al-ʿAlāʾ (d. 154/770) is reported to have transmitted the view 
of one of the Companions that ‘the codified text contains errors, but the Arabs 
will set it straight’ (
ʾinna fī l-muṣḥaf laḥnan wa-sa-tuqīmuhu l-ʿArab
) (al-Farrāʾ, 
Maʿānī 
l-Qurʾān 
II, ed. an-Najjār, Cairo, 1955–72, p. 183). 
An alternative was to change the particle 
ʾinna 
to 
ʾin
. According to the grammar
-
ians, 
ʾinna 
‘verily’
 
and 
ʾanna 
‘that’ had an abbreviated form 
ʾin
, 
ʾan 
(the so-called 
ʾin, ʾan muḫaffafa 
‘lightened 
ʾin
, 
ʾan
’), with the following noun in the nominative. 
These forms seem to have been more current in the Ḥijāz than in the East. Some 
examples occur in the 
Qurʾān
, for example, 
Q
36/32: 
wa-ʾin kullun la-mā jamīʿun 
ladaynā muḥḍarūna
‘verily, all will be brought together before Us’. To complicate 
matters, they can be followed by an accusative as well, for example, 
Q
11/111: 
wa-ʾin kullan la-mā yuwaffiyannahum rabbuka ʾaʿmālahum
‘verily, thy Lord will repay 
everyone their deeds’. Not surprisingly, the grammarians tried to correct such 
forms, either by changing the case ending of the following word, or by reading 
the full form 
ʾinna
. 
An oft-cited difference between Ḥijāz and Tamīm is the construction of 
mā
as 
a nominal negator. According to the grammarians, 
mā
could be construed in the 
same way as the verb 
laysa
‘to be not’, with an accusative in the predicate, for 
example, 
mā huwa kabīran
‘he is not big’. This so-called 
mā ḥijāziyya
did not occur 
in the Eastern dialects.
There are indications that the negation 
ʾin
, which occurs not infrequently in 
the 
Qurʾān
, for example, 
Q
11/51: 
ʾin ʾajriya ʾillā ʿalā llaḏī faṭaranī
‘my reward is not 
due except from Him who created me’, is characteristic of Ḥijāzī speech.
In some dialects, a relative pronoun 
ḏī
or 
ḏū
(the so-called 
ḏū ṭāʾiyya
, i.e., of the 
tribe Ṭayyiʾ) is attested; this relative does not occur in the 
Qurʾān
, but it probably 
occurs in the an-Namāra inscription (cf. above, p. 36), and it is found in some 
pre-Islamic poems, for example, in a line quoted in the 
Ḥamāsa
, a ninth-century 
anthology of poetry: 
li-hāḏā l-marʾi ḏū jāʾa sāʾiyan
‘to this man who has come to levy 
tax’ (cf. Reckendorff 1921: 431).
Apart from the possible, but unlikely, occurrence of an undeclined dual in one 
verse in the 
Qurʾān
, all these points concern relatively minor differences. There 
is, however, one grammatical issue that touches upon the core of Arabic syntax, 
the construction of verbal and nominal sentences. In Classical Arabic, when the 
verb precedes the agent in the so-called verbal sentence (cf. below, Chapter 6, 
pp. 97–9; Chapter 7, pp. 112–14), there is no agreement in number between verb 

---

52
The Arabic Language
and agent. According to the grammarians, some dialects in the 
Jāhiliyya
did allow 
agreement in this case. The stock example cited for this phenomenon is 
ʾakalūnī 
l-barāġīṯ
‘the fleas have bitten me’ (instead of 
ʾakalatnī l-barāġīṯ
). The evidentiary 
verses stem from Ḥijāzī poets exclusively. This is the only example of a syntactic 
feature ascribed to a pre-Islamic dialect that has a parallel in the modern dialects 
of Arabic. These do not exhibit the Classical Arabic difference between verbal and 
nominal sentences, and always have agreement between verb and agent. Yet the 
canonical word order in the modern dialects is Subject–Verb–Object, rather than 
the Classical Arabic word order Verb–Subject–Object in 
ʾakalūnī l-barāġīṯ
. It is not 
clear, therefore, whether this feature in Ḥijāzī Arabic should be interpreted as the 
first step towards a later development or represents an independent phenom-
enon. In the text of the 
Qurʾān
as we have it, this feature does not occur.
Whenever differences between Eastern and Western Arabic existed, the 
language of the 
Qurʾān
usually reflects the Eastern usage. As regards the pronun-
ciation of the glottal stop in the early Islamic period, it was felt to be more presti
-
gious and more fitting for the recitation of the Holy Book, although there seems 
to have been considerable opposition on the part of the early reciters to such a 
pronunciation, which they branded as affected. It is equally obvious, however, 
from the list of differences that the dialects were not very far apart from each 
other. Most of the features mentioned above concern phonetic or phonological 
phenomena. Apart from the 
ʾakalūnī l-barāġīṯ
syndrome, the sources mention a 
few syntactic differences, which we have not listed here, since their status is hard 
to determine. The various constructions with 
ʾillā 
‘except, unless’, for instance, 
for which one dialect is said to have used the nominative and the other the 
accusative, almost certainly represent theorising on the part of the grammar
-
ians. There is one thing that transpires from such syntactic 
luġāt
: if there is any 
reality to them, both dialect groups must have used case endings. The evidence 
for an undeclined dual mentioned above is too meagre to warrant any conclusion 
about a possible loss of case endings. In view of the central role of declension in 
the various theories about the linguistic situation in the pre-Islamic period, this 
absence of evidence for loss of declension in the grammatical literature is crucial 
to our understanding of the historical development of Arabic.