The Arabic Language

lingua franca
in this area, for 
instance, Akkadian and Aramaic. Some of the common features shared by the 
languages of the region may have been introduced by the presence of such a 
lingua franca
.
Some scholars continue to feel that a genetic classification is possible provided 
that the right principles are used. Thus, for instance, Hetzron (1974, 1976) proposes 
basing the classification on the principles of archaic heterogeneity and shared 
morpholexical innovations. The former principle implies that a heterogeneous 
morphological system is more archaic than a homogeneous one; the latter principle 
states that morpholexical innovations are unlikely to be subject to borrowing. He 
illustrates his approach with two examples. The suffixes of the first- and second-
person singular of the perfect verb in Arabic are 
-tu
/
-ta
, as in 
katabtu
/
katabta
, ‘I/
you have written’. In Ethiopic they are 
-ku
/
-ka
, but in Akkadian the equivalent 
suffix form of nouns and verbs (the so-called stative or permansive) has a set of 
personal suffixes 
-(ā)ku
/
-(ā)ta
. Such a distribution may be explained as the result 
of a generalisation in Arabic and Ethiopic, which implies that the heterogeneous 
system of Akkadian is older. The tendency towards homogenisation was realised 
differently in Arabic (and Canaanite), on the one hand, and in Ethiopic (and South 
Arabian), on the other. Hebrew has 
kātabti
/
kātabta
and thus shares this innovation 
with Arabic, setting it apart from the South Semitic languages.
Hetzron’s second example has to do with the prefix vowel of the imperfect 
verb. In Akkadian, the prefixes of the third-person singular masculine, the third-
person plural and the first-person plural have 
-i-
, while all other persons have
 -a-
. 
In Classical Arabic all persons have 
-a-
, while in Ethiopic all persons have
 -ǝ-
(< 
-i-
). In this case, too, Hetzron regards the heterogeneous system of Akkadian as 
the older one, whereas the prefixes in the other languages are the result of a later 
generalisation. Actually, the situation in Arabic is somewhat more complicated, 
since in pre-Islamic Arabic some dialects had 
-i-
in all persons, whereas others had 
-a-
(cf. below, pp. 49f.). Possibly, there was an intermediate step in which 
-i-
was 
generalised for all persons in verbs with a stem vowel 
-a-
, and 
-a-
was generalised 
for all persons in verbs with a stem vowel 
-u-
/
-i-
. The pre-Islamic dialects differed 
with regard to the further generalisation, in which the correlation with the stem 
vowel was abandoned.

---

Arabic as a Semitic Language 
15
On the basis of these and similar examples, Hetzron posits a group of Central 
Semitic languages, separating Arabic from its position in the standard model in 
which it is grouped together with South Arabian and Ethiopic as South Semitic 
languages. The main force of Hetzron’s arguments is the fact that he does not base 
his subgrouping of the Semitic languages on common innovations in phonology, 
syntax or lexicon – in these domains, borrowing is always a distinct possibility – but 
concentrates instead on morpholexical innovations, which are much less prone to 
borrowing. We may add that he excludes from his classification arguments based 
on common retention of features (‘negative innovation’), since this may occur 
independently in several languages and does not imply any sustained contact 
between the languages involved (see Figure 2.2).
While Hetzron still accepts the family-tree model for the Semitic languages, 
others have questioned its usefulness. Ullendorff (1970) rejects out of hand the 
possibility of ever reaching a classificatory scheme reflecting genetic relation-
ship. Garbini (1984) claims that it is possible to trace the historical development 
of the Semitic languages, but without any genetic hierarchy, since the pattern 
of linguistic development in the area is crucially different from that in the Indo-
European area. In his view, the present distribution of the languages involved is 
not the result of sudden migrations of peoples, but rather of a gradual infiltra-
tion from different centres, which reached out towards the periphery of the area. 
Such an infiltration could transmit innovations in a wave-like fashion that most 
strongly affected the central area, whereas in the periphery older forms stood a 
better chance of maintaining themselves. In Garbini’s view, one area in particular 
played an essential role in the distribution of innovations, namely, the Syrian 
plain (rather than the coastal region or Palestine), which he regards as the core 
Figure 2.2 The genealogy of the Semitic languages (Hetzron 1974, 1976)
Proto-Semitic
West Semitic 
East Semitic
(Akkadian)
South 
Semitic 
Central Semitic
Arabo-Canaanite 
Aramaic
Ethiopian 
Epigraphic 
Modern
South Arabian South Arabian
Arabic Canaanite

---

16
The Arabic Language
area of the Semitic languages. The main characteristic of the Syrian region in 
which these innovations are supposed to have taken place is the contact between 
sedentary settlements on the desert fringe and nomads from the desert. In some 
cases, the nomads settled and became part of the sedentary population, but in 
many other cases groups of settlers separated themselves and became isolated as 
desert-dwelling nomads. Garbini regards this constant alternation as the origin 
of the linguistic pattern of innovations spreading from the Syrian area into other 
areas. Exactly which innovations were brought from Syria into the peninsula 
depended on the period in which a particular group of people took to the desert.
Garbini cites examples from Akkadian and Eblaite, showing how these languages 
were not involved in the migratory process and did not share in some of the later 
innovations in the Syrian area. The common features which Arabic shares with 
Aramaic and Amorite stem from the period in which the ancestors of the later 
Arabs still lived in the Syrian region. In his view, Arabic is the nomadic variety of 
the languages spoken in Syria in the first millennium bce
, which he calls collec
-
tively Amorite. He regards the South Arabian and Ethiopian languages as the 
product of an earlier migration from the same area. According to this theory, 
those common features between Arabic and South Arabian that are not shared by 
the languages in the Syrian area are the result of later convergence: the Arabian 
Bedouin influenced the sedentary languages/dialects in the south, and inversely 
through the caravan trade the South Arabian languages/dialects became known 
in the north of the peninsula. The Modern South Arabian languages (Mehri, 
Soqoṭri) do not derive directly from the Epigraphic South Arabian language. 
They probably belong to strata that had never been reached by Arabic influence 
because they were spoken in remote regions. In some respects, their structure is, 
therefore, more archaic than that of Epigraphic South Arabian.
We have seen above (p. 12) that Kienast (2001) also views the history of the 
Semitic languages as a succession of waves of immigrants. Whenever a group of 
nomads in such a wave settle down, their language develops further, and loses, 
for instance, the declensional endings. He explains the retention of declensional 
endings in Arabic as a result of the fact that their language was spoken by people 
who stayed away from the sedentary area in the Near East for a long time and, 
therefore, did not participate in the linguistic developments that took place there. 
Even more radically, Edzard (1998) questions the possibility, or even the desir
-
ability, of a historical–comparative approach, which is based on the notion of one 
proto-language that bifurcates repeatedly until it develops into the individual 
languages. His solution is to adopt a polygenetic model, in which there is an 
initial pool of linguistic variation. Contacts between the groups involved lead to 
convergence. Eventually, new language varieties emerge in which the variants are 
redistributed. In such a model, there is no room for the kind of reconstructions 
favoured by historical comparativists. At the most, the correspondences between 
the individual languages may be viewed as a convenient way of representing the 

---

Arabic as a Semitic Language 
17
relations between similar forms in individual languages, for instance, when it is 
observed that Arabic words with /ḍ/ correspond with words with /ṣ/ in Hebrew 
or Akkadian. Such formulas may be useful as long as they do not lead to claims 
about an historical development from a common ancestor
In spite of the hazards of historical–comparative analysis, research in the 
twentieth century has expanded the scope of Semitic languages even further by 
including another group of languages, the so-called Hamitic languages. The name 
itself is derived from the old classificatory scheme of the Book of Genesis (10:1ff.), 
which divides all mankind among the descendants of the three sons of Noah. 
This scheme was used by later scholars to divide all languages into those of the 
descendants of Shem, those of the descendants of Cham and those of the descen
-
dants of Japheth (p. 5). The label of Hamitic languages was originally applied to 
all languages of Africa, but in the modern period Hamitic has come to be used 
collectively for five specific language groups in Africa: the Berber languages of 
North Africa and their ancestor, Old Libyan; Old Egyptian and its offshoot, Coptic; 
Hausa; the Cushitic languages; and the Chadic languages. When common links 
between these language groups and the Semitic languages were discovered, they 
became collectively known as the Hamito-Semitic languages. Since the 1970s, the 
current name for this group has become the Afro-Asiatic languages.
In the reconstruction of Afro-Asiatic, too, Garbini applies his theory of the 
innovatory Syrian area. In his view, any attempt to trace the various groups of 
Semitic and Hamitic (Egyptian, Libyan/Berber, Cushitic and possibly Hausa) back 
to one ancestor is doomed to failure. It is true that even a cursory comparison of 
the various groups reveals the presence of related forms, but the fact that there 
are almost no firm phonetic correlations of the type found in the Indo-European 
languages shows that we are not dealing here with a language family with sibling 
languages descending from a common ancestor. In Garbini’s view, the Hamitic 
languages are African languages without genetic relationships to the Semitic 
languages. At one time or another, and to different degrees, they were Semiti-
cised by groups of people coming from the Syrian area. Old Egyptian, for instance, 
would have become a Semitic language if the contacts had continued. The basis 
in this model, as in the one proposed by Edzard, is diversity; the unity of the later 
Semitic languages and the varying degrees of resemblance between Hamitic and 
Semitic languages are the result of later convergence.
In spite of the problems involved in applying the family-tree model to the 
Semitic languages, comparative research has persisted in the application of the 
reconstructive paradigm to the Afro-Asiatic languages and to even higher group-
ings of languages. The interest in language relationships has led to the establish
-
ment of progressively higher-level hierarchies, such as the proto-language above 
the Indo-European and the Afro-Asiatic group. Various attempts have been made 
to connect the root structure and the phonological inventory of both groups. 
To some degree, these attempts were facilitated by two developments in Indo-

---

18
The Arabic Language
European studies: the laryngal theory and the theory of glottalised consonants 
in proto-Indo-European. Both theories brought Indo-European and Afro-Asiatic 
phonology closer to each other.
Even more audacious conjectures seek to incorporate both Indo-European and 
Afro-Asiatic languages in such constructs as the Nostratic macro-family, including 
the Kartvelian languages (e.g., Georgian), Uralo-Altaic (e.g., Hungarian) and 
Dravidian (e.g., Tamil). It is hard to say what the value of such conjectures is, since 
the time-span involved allows for a great deal of speculation about the changes 
that make it possible to find lexical parallels. Besides, it is debatable whether it 
is permissible to apply the results of Indo-European linguistics to all linguistic 
relationships in the world. It could very well be the case that the type of relation
-
ship in which a mother language generates daughter languages, as is commonly 
held to be the case in the Indo-European languages, is an exception.