The Arabic Language

6.3 Word structure
The most characteristic feature of all Semitic languages is the peculiar relation
-
ship between form and meaning. According to the traditional view, the lexical 
meaning is represented by root consonants (radicals), and the morphological 
meaning is added to these radicals in the form of a vowel pattern, sometimes 
with auxiliary consonants. In modern morphological theories, this is called root-
and-pattern morphology. In this type of morphology the syllable structure of 
the word is specified by a template that is applied to the CV skeleton. This is 
somewhat similar to the way in which Arabic grammarians represented the struc
-
ture of a word with the help of the 
f-ʿ-l 
notation (Chapter 7). Root-and-pattern, 
or template, morphology is also referred to as non-concatenation (or in French 
interdigitation
). This term underscores the fact that the vocalic melody is applied 
discontinuously to a consonantal root. 
The root-and-pattern analysis presupposes the linguistic reality of the root 
structure. Proponents of the root-based approach cite as evidence the Obliga-
tory Contour Principle (OCP), which states that similar adjacent elements are not 
permitted. This principle works from left to right at the level of the CV skeleton, 
and forbids the occurrence of roots like *
d-d-r
or *
m-m-d
. Homorganic consonants 
(i.e., consonants from the same articulatory class) in adjacent position are not 
allowed either, so that a root like *
b-m-f
is forbidden as well, because it contains 
two adjacent bilabials (Rosenthall 2008). It has been known for a long time 
that such co-occurrence constraints exist in Arabic, as well as in other Semitic 
languages (Greenberg 1950).
The fact that phonological restrictions like the OCP operate at the level of the 
root implies that root consonants are treated differently from additional conso-
nants. The linguistic reality of the root system is also demonstrated by psycho
-
linguistic research in speech errors and child language. In making a speech error 
speakers are known to switch root consonants, for example, 
matkab 
instead of 
maktab
, but apparently they never switch a root consonant with an affix, so that 
errors like *
tamkab
do not seem to occur. Similar evidence comes from patients 
with aphasia (Idrissi 
et al
. 2008). Studies in first language acquisition show that 
from an early age children are able to produce new forms on the basis of their 
knowledge of patterns and roots (Ravid 2003).
A different approach to the Arabic lexicon takes the stem as the point of depar-
ture of the morphological derivation. The stem-based approach assumes that the 
information in consonantal roots is insufficient and that the vowels are needed 
to explain the derivational forms. Take the forms of the prefix-conjugation (tradi-
tionally called ‘imperfect’) of the verb ‘to write’ in Table 6.1. In the stem-based 
approach, these are analysed as consisting of a stem CCvC, to which various affixes 
are added. In the case of the stem 
-ktub- 
the vowel 
-u- 
is lexically specified. Gafos 
(2002, 2003) points out that there is a phonological reason for the stem to have 

---

90
The Arabic Language
this form rather than, for instance, CvCC, because verbal forms receive consonant-
initial suffixes, as in 
ya-ktub-na
. With a hypothetical stem 
-kutb-
, this would result 
in a non-allowed consonant cluster 
*ya-kutb-na
. In nouns this constraint does not 
apply because nouns have vowel-initial suffixes, for example,