Basic Sentence Structure in Sudanese Arabic

Abstract

On the basis of notions developed in respect of attribution in Dickins (2009, which readers are recommended to read before looking at this paper in detail), I address basic sentence structure in Sudanese Arabic. I argue that a central feature of Sudanese Arabic is the bipartite predicand-predicate (mubtada'-xabar) structure, pointing out that this has some tantalisingly unexpected features. I also consider monopartite sentences in Sudanese Arabic, arguing for a more inclusive view of what constitutes a non-elliptical sentence than is commonly accepted. I finally look briefly at bipartite sentences with 'scrambled' word order.

Full text

237
1 I thank Janet Watson for reading a draft of this paper and making very useful
suggestions. All shortcomings are my own responsibility. I also thank the Lever-
hulme Trust for granting me a Research Award giving me relief from teaching and
administration during the academic years 2002–4 in order to pursue work on a
dictionary and grammar of Sudanese Arabic. This award has contributed to the
production of this paper.
For present purposes, Sudanese Arabic can be taken to have the following conso-
nant phonemes (cf. Dickins 2007: 24): /b/ voiced, bilabial, stop; /m/ bilabial, nasal;
/w/ bilabial, glide; /f/ voiceless, bilabial, fricative; /d/ voiced, apico-dental, stop;
/t/ voiceless, apico-dental, stop; /z/ voiced, apico-dental, fricative; /s/ voiceless, apico-
dental, fricative; /∂/ voiced, emphatic, apico-alveolar, stop; /†/ voiceless, emphatic,
apico-alveolar, stop; /Â/ voiced, emphatic, apico-alveolar, fricative; /Ò/ voiceless,
emphatic, apico-alveolar, fricative; /r/ (plain), apico-alveolar, trill; /®/ emphatic,
apico-alveolar; trill; /l/ (plain), apico-alveolar, lateral; /¬/ emphatic, apico-alveolar,
lateral; /n/ apico-alveolar, nasal; /j/ voiced, dorso-prepalatal, stop; /c/ voiceless,
dorso-prepalatal, stop (marginal phoneme); /s/ voiceless, dorso-prepalatal, fricative;
/n/ dorso-prepalatal, nasal (marginal phoneme); /y/ dorso-palatal, glide; /g/ voiced,
post-dorso-velar, stop; /k/ voiceless, post-dorso-velar, stop; /g/ voiced, post-dorso-post-
velar, fricative; /x/ voiceless, post-dorso-post-velar, fricative; /¨/ voiced, pharyngeal,
fricative; /Ì/ voiceless, pharyngeal, fricative; /’/ voiced, glottal, fricative (sometimes
described as glottal stop); /h/ voiceless, glottal, fricative.
For present purposes, Sudanese Arabic can be taken to have the following vowel
phonemes (cf. Dickins 2007: 25): /a/ open, unrounded, short vowel; /i/ front,
close, unrounded, short vowel; /u/ back, close, rounded, short vowel; /a/ open,
unrounded, long vowel; /i/ front, close, unrounded, long vowel; /u/ back, close,
Journal of Semitic Studies LV/1 Spring 2010 doi: 10.1093/jss/fgq046
© The author. Published by Oxford University Press on behalf of the University of Manchester.
All rights reserved.
BASIC SENTENCE STRUCTURE IN
SUDANESE ARABIC1
JAMES DICKINS
UNIVERSITY OF SALFORD
Abstract
On the basis of notions developed in respect of attribution in Dickins
(2009, which readers are recommended to read before looking at this
paper in detail), I address basic sentence structure in Sudanese Arabic.
I argue that a central feature of Sudanese Arabic is the bipartite predi-
cand-predicate (mubtada}-xabar) structure, pointing out that this has
some tantalisingly unexpected features. I also consider monopartite
sentences in Sudanese Arabic, arguing for a more inclusive view of
what constitutes a non-elliptical sentence than is commonly accepted.
I finally look briefly at bipartite sentences with ‘scrambled’ word order.
92897_JOSS10-2_09_Dickins.indd 23792897_JOSS10-2_09_Dickins.indd 237 15/03/10 08:4615/03/10 08:46
at University of Salford on April 15, 2010 http://jss.oxfordjournals.orgDownloaded from

BASIC SENTENCE STRUCTURE IN SUDANESE ARABIC
238
rounded, short vowel; /e/ front, mid, unrounded, long vowel; /o/ back, mid,
rounded, long vowel (see, however, Dickins 2007 for a critique of this account and
an alternative analysis).
2 In fact, the syntactic analysis in this paper is much more like that of Mulder
and Hervey than that implied — whether in lexotatics or delotactics — in Dickins
(1998). In Dickins (1998), syntax (roughly as normally understood) or delotactics is
an analysis involving content (semantic elements) only, whereas in Mulder and
Hervey, syntactic analysis involves elements which have both form and content.
In Dickins (1998), I have argued that Mulder and Hervey’s standard version of
axiomatic functionalism is coherently subsumed under the extended version of
the theory proposed there (e.g. Dickins 1998: 250–1). Accordingly, the theoretical
models of standard axiomatic functionalism are fully interpretable in terms of
models within extended axiomatic functionalism. In extended axiomatic-function-
alist terms, Mulder and Hervey’s syntax is an analysis of relations at the allosemic
level (what might be termed allosemotactics).
3 Some of the constructed examples are either elicited from Elrayah Abdelgadir
or produced by him without prompting on my part. Elrayah Abdelgadir acted as
consultant in 2005 and 2007 for an Arabic/English Dictionary of Sudanese Arabic
which I am working on.
The theoretical analysis provided in this paper is based — some-
what informally — on the linguistic theory of axiomatic functionalism
initially developed by J.W.F. Mulder and S.G.J. Hervey (e.g. Mulder
1968, 1989; Mulder and Hervey 1972, 1980; Hervey 1979, 1982),
and subsequently extended on the basis of proposals put forward by
Michael Lamb in Dickins (1998). In an attempt to keep the main
body of the paper as generally comprehensible as possible, I have con-
fined technical comments relating to axiomatic functionalism to foot-
notes.2
1. General Background
In this article, I shall use the term ‘Sudanese Arabic’ as shorthand for
Central Urban Sudanese (Dickins 2007; elsewhere termed Khartoum
Arabic (Dickins 2006)), that is the dialect standardly spoken by long-
term native Arabic-speaking residents of Greater Khartoum (Khartoum,
Khartoum North, and Omdurman), and in other urban areas of cen-
tral Sudan, roughly to the towns of Atbara in the north, Sennar on
the Blue Nile, and Kosti on the White Nile. For details of the tran-
scription system, see Dickins (2007). In this paper, I use a subscript
to indicate a vowel which is deleted in liaison ‘readings’ of the mate-
rial presented. The Sudanese examples for this article have been taken
from three types of sources:
1. Examples constructed by myself. These are marked with a (C) after
them, or sometimes an entire group of examples is noted as having
been constructed before they are given.3
92897_JOSS10-2_09_Dickins.indd 23892897_JOSS10-2_09_Dickins.indd 238 8/03/10 10:488/03/10 10:48
at University of Salford on April 15, 2010 http://jss.oxfordjournals.orgDownloaded from

BASIC SENTENCE STRUCTURE IN SUDANESE ARABIC
239
2. Material which I tape-recorded in 1985 and 1986, either from
Omdurman Radio, or spontaneous conversations between Sudanese
friends; or material which I wrote down from spontaneous speech.
All these examples are marked with an (R) after them.
3. Material taken from a number of radio series recorded for me by
Omdurman Radio in 1986. These are marked with the name of the
series after them.
The definite article in Sudanese Arabic has the following allomorphs:
After a consonant al- – before all non-apical and non-dorsal conso-
nants: /b/, /m/, /w/, /f/, /y/, /g/, /k/, /x/, /g/,
/¨/, /Ì/, /}/, and /h/.
ad-, at-, az-, etc. – before apical and dorsal consonants: /d/,
/t/, /z/, /s/, /∂/, /†/, /Â/, /Ò/, /r/, /®/, /l /, /¬/,
/n/, /j/, /s/, and the marginal /c / and /n/.
After a vowel
l- – before all non-apical and non-dorsal
consonants
d-, t-, z-, etc. – before apical and dorsal consonants
In this article, I have used subscript forms to indicate vowels which
‘disappear’ through so-called re-syllabification in liaison with preced-
ing or following words: thus, fihim ‘he understood’, in which the
syllables are /fi/ and /him/ (with stress on the first syllable /fi/), but
fihim al-kalam ‘he understood the statement/discussion’ (in which the
first two syllables are /fih/ and /mal/). I have not used subscript forms
with following pronouns — thus fihm-u ‘he understood it (m.sg.)’
(not fihim-u); also fihmu ‘they understood’ (same pronunciation as
fihm-u ‘he understood it (m.sg.)’. I have also not used subscripts with
the article al-. Thus fihmu l-kalam ‘they understood the statement/
discussion’ (not fihmu al-kalam).
We can establish the following provisional word and phrase cate-
gories for Sudanese Arabic:
Word category Phrase category
noun nominal phrase (NomP)
adjective adjectival phrase (AdjP)
verb verb phrase (VerbP)
adverb adverbial (AdvP)
For current purposes, I shall treat a phrase as any element consisting
of more than one word (not counting the clitic al- as a word). Where
the element consists of only one word, I will refer to it simply as a
92897_JOSS10-2_09_Dickins.indd 23992897_JOSS10-2_09_Dickins.indd 239 8/03/10 10:488/03/10 10:48
at University of Salford on April 15, 2010 http://jss.oxfordjournals.orgDownloaded from

BASIC SENTENCE STRUCTURE IN SUDANESE ARABIC
240
4 The terms ‘predicand’ and ‘predicate’ are used in Bohas, Guillaume and Kou-
loghli (1990: 43), where they are employed as the translations of the traditional
Arabic grammatical terms mubtada’ and xabar. They are subsequently used by
Watson (1993). In respect of Classical (and Standard) Arabic, the term ‘predicand’
in particular allows the term ‘subject’ to be employed as a translation of the tradi-
tional Arabic grammatical term fa¨il (i.e. the subject of the so-called verbal sentence).
Although I do not believe that Sudanese Arabic makes a properly syntactic distinction
between nominal and verbal sentences (cf. Section 4), I have used the term ‘predi-
cand’ in this paper to avoid the bizarre-sounding description of the ‘non-predicate’
element of an Arabic bipartite clause as the ‘subject’ of this clause, even where this
predicate is clearly non-verbal. Thus in ar-rajil fi-l-bet ‘the man’s in the house’, for
example, where the predicate fi-l-bet ‘in the house’ is clearly non-verbal, I want to
avoid describing the ar-rajil element as the subject.
noun, adjective, verb, or adverb. Where a noun (etc.) is preceded by al-,
I shall refer to this at this point in the paper as a definite noun (etc.).
We can also establish the notion of a bipartite clause, consisting of
a predicand (which, for the moment, we can characterize as consisting
of a noun or nominal phrase) and a predicate (which may be a noun,
nominal phrase, adjective, adjectival phrase, verb, verb phrase, adverb,
or adverbial.4
2. Definiteness, Indefiniteness and Basic Sentence Structure
I have argued in Dickins (2009; Section 10) that the definite particle
is the head of definite phrases in Sudanese Arabic. There I pointed
out that the distribution of definite phrases (i.e. the syntactic slots/
positions/places in which they can occur) is roughly ‘nominal’, i.e. it
is similar to that of simple indefinite nouns. The distribution of indef-
inite phrases, by contrast, depends on whether the phrase is nominal,
adjectival, verbal, etc. In this paper, I will look at the implications of
this situation for basic sentence structure in Sudanese Arabic. Consider
the following (examples constructed by myself ):
2.1 Noun+other element
2.1.1 Noun+Noun
definite+definite definite+indefinite indefinite+indefinite
?al-walad at-tarzi al-walad tarzi *walad tarzi
‘the boy’s the tailor’ ‘the boy’s a tailor’ ‘a boy’s a tailor’
tarzi al-walad
‘the boy’s a tailor’
(‘it’s a tailor that the boy is’)
92897_JOSS10-2_09_Dickins.indd 24092897_JOSS10-2_09_Dickins.indd 240 8/03/10 10:488/03/10 10:48
at University of Salford on April 15, 2010 http://jss.oxfordjournals.orgDownloaded from

BASIC SENTENCE STRUCTURE IN SUDANESE ARABIC
241
?at-tarzi al-walad at-tarzi walad *tarzi walad
‘the tailor’s the boy’ ‘the tailor’s a boy’ ‘a tailor’s a boy’
walad at-tarzi
‘the tailor’s a boy’
(it’s a boy that the tailor is’)
2.1.2 Noun+Adjective
definite+definite definite+indefinite indefinite+indefinite
?al-walad az-za¨lan al-walad za¨lan *walad za¨lan
‘the boy’s the angry one’ ‘the boy’s angry’ ‘a boy’s an angry one’
za¨lan al-walad
‘the boy’s angry’
(‘it’s angry that the boy is’)
az-za¨lan al-walad az-za¨lan walad *za¨lan walad
the angry one’s the boy ‘the angry one’s a boy’ ‘an angry one’s a boy’
2.1.3 Noun+Adverbial
definite+definite definite+indefinite indefinite+indefinite
?al-walad al-hina al-walad hina *walad hina
‘the boy’s the one who’s here’ ‘the boy’s here’ ‘a boy’s here’
hina l-walad
‘the boy’s here’
(‘it’s here that the boy is’)
al-hina al-walad al-hina walad *hina walad
‘the one who’s here is the boy’ ‘the one who’s here is a boy’ ‘a boy’s here’
?walad al-hina
‘a boy is the one who’s here’
(‘the one who’s here is a boy’)
2.1.4 Noun+ Verb/Verb phrase
definite+definite definite+indefinite indefinite+indefinite
?al-walad az-zi¨il al-walad zi¨il *walad zi¨il
‘the boy’s the one who got
angry’
‘the boy got angry’ ‘a boy got angry’
zi¨i l al-walad
‘the boy got angry’
92897_JOSS10-2_09_Dickins.indd 24192897_JOSS10-2_09_Dickins.indd 241 8/03/10 10:488/03/10 10:48
at University of Salford on April 15, 2010 http://jss.oxfordjournals.orgDownloaded from

BASIC SENTENCE STRUCTURE IN SUDANESE ARABIC
242
az-zi¨il al-walad az-zi¨il walad *zi¨il walad
‘the one who got angry’s the
boy’
‘the one who got angry’s a boy’ ‘one who got angry’s a boy’/
‘a boy got angry’
?walad az-zi¨il
‘the one who got angry’s a boy’
2.1.5 Noun+bipartite clause
definite+definite definite+indefinite indefinite+indefinite
?al-walad al-bet-u garib al-walad bet-u garib *walad bet-u garib
‘the boy is the one whose
house is near’
‘the boy’s house is near’
(more lit: ‘the boy his house
‘a boy’s house is near’
(more lit: ‘a boy his house
near’) near’)
bet-u garib al-walad
‘the boy’s house is near’
al-bet-u garib al-walad al-bet-u garib walad *bet-u garib walad
‘the one who’s house is near is
the boy’
‘the one who’s house is near is
a boy’
‘his house is near, a boy’
?walad al-bet-u garib
‘the one who’s house is near is
a boy’
2.2 Adjective + other element
2.2.1 Adjective+adjective
definite+definite definite+indefinite indefinite+indefinite
?al-¨ajib az-za¨lan al-¨ajib za¨lan *ajib za¨lan
‘the strange one’s the angry one’ ‘the strange one’s angry’ ‘a strange one’s angry’
za¨lan al-¨ajib
‘the strange one’s angry’
(‘it’s angry that the strange
one is’)
?az-za¨lan al-¨ajib az-za¨lan ¨ajib *za¨lan ¨ajib
‘the angry one’s the strange
one’
‘the angry one’s strange’ ‘an angry one is strange’
¨ajib az-za¨lan
‘the angry one’s strange’
(‘it’s strange that the angry
one is’)
92897_JOSS10-2_09_Dickins.indd 24292897_JOSS10-2_09_Dickins.indd 242 8/03/10 10:488/03/10 10:48
at University of Salford on April 15, 2010 http://jss.oxfordjournals.orgDownloaded from

BASIC SENTENCE STRUCTURE IN SUDANESE ARABIC
243
2.2.2 Adjective+noun: see 2.1.2 above
2.2.3 Adjective+Adverbial
definite+definite definite+indefinite indefinite+indefinite
?al-¨ajib al-hina al-¨ajib hina *¨ajib hina
the strange one’s the one
who’s here’
‘the strange one’s here’ ‘a strange one’s here’
hina al-¨ajib
‘the strange one’s here’
(‘it’s here that the strange one
is’)
?al-hina al-¨ajib al-hina ¨ajib *hina ¨ajib
‘the one who’s here is the
strange one’
‘the one who’s here is strange’ ‘here’s a strange one’
?¨ajib al-hina
(‘the one who’s here is
strange’)
2.2.4 Adjective+Verb/Verb phrase
definite+definite definite+indefinite indefinite+indefinite
?al-¨ajib az-zi¨il al-¨ajib zi¨il *ajib zi¨il
‘the strange one’s the one who
got angry’
‘the strange one got angry’ ‘a strange one got angry’
zi¨il al-¨ajib
‘the strange one got angry’
?az-zi¨il al-¨ajib al-zi¨il ¨ajib *zi¨il ¨ajib
‘the one who got angry is the
strange one’
‘the one who got angry is
strange’
‘an angry one is strange’
?¨ajib az-zi¨il
‘the one who got angry is
strange’
2.2.5 Adjective+bipartite clause
definite+definite definite+indefinite indefinite+indefinite
?al-¨ajib al-bet-u garib al-¨ajib bet-u garib*¨ajib bet-u garib
‘the strange one is the one
whose house is near’
‘the strange one’s house is
near’ (more lit: the strange
one his house is near’)
‘a strange one’s house is near’
92897_JOSS10-2_09_Dickins.indd 24392897_JOSS10-2_09_Dickins.indd 243 8/03/10 10:488/03/10 10:48
at University of Salford on April 15, 2010 http://jss.oxfordjournals.orgDownloaded from

BASIC SENTENCE STRUCTURE IN SUDANESE ARABIC
244
bet-u garib al-¨ajib
‘the strange one’s house is near’
?al-bet-u garib al-¨ajib al-bet-u garib ¨ajib*bet-u garib ¨ajib
‘the one who’s house is near is
the strange one’
‘the one who’s house is near is
strange’
‘his house is near, a strange
one’
(‘one whose house is near is a
strange one’)
?¨ajib al-bet-u garib
‘the one who’s house is near is
strange’
2.3 Adverbial+other element
2.3.1 Adverbial+noun: see 2.1.3 above
2.3.2 Adverbial+Adjective: see 2.2.3 above
2.3.3 Adverbial+Adverbial
definite+definite definite+indefinite indefinite+indefinite
?al-ba¨da-na al-hina al-ba¨da-na hina *ba¨da-na hina
‘the one after us is the one
who’s here’
‘the one after us is here’ ‘one after us is here’
hina al-ba¨da-na
‘the one after us is here’
(‘it’s here that the one after
us is’)
?al-hina al-ba¨da-na al-hina ba¨da-na *hina ba¨da-na
‘the one who’s here is the one
after us’
‘the one who’s here is after us’ ‘a one who’s here is after us’
?ba¨da-na al-hina
‘the one who’s here is after us’
2.3.4 Adverbial+Verb/Verb phrase
definite+definite definite+indefinite indefinite+indefinite
?al-ba¨da-na az-zi¨il al-ba¨da-na zi¨il *ba¨da-na zi¨il
‘the one after us is the one
who got angry’
‘the one after us got angry’ ‘after us got angry’
zi}i l al-ba¨da-na
‘the one after us got angry’
92897_JOSS10-2_09_Dickins.indd 24492897_JOSS10-2_09_Dickins.indd 244 8/03/10 10:488/03/10 10:48
at University of Salford on April 15, 2010 http://jss.oxfordjournals.orgDownloaded from

BASIC SENTENCE STRUCTURE IN SUDANESE ARABIC
245
?az-zi¨il al-ba¨da-na al-zi¨il ba¨da-na *zi¨il ba¨da-na
‘the one who got angry is the
one after us’
‘the one who got angry is after
us’
‘one who got angry’s after us’
?ba¨da-na az-zi¨il
‘the one who got angry is after
us’
2.3.5 Adverbial+bipartite clause
definite+definite definite+indefinite indefinite+indefinite
?al-ba¨da-na al-bet-u garib al-ba¨da-na bet-u garib *ba¨da-na bet-u garib
‘the one after us is the one
whose house is near’
‘the one after us, his house is
near’
‘one after us his house is near’
(more lit: ‘after us his house is
near’)
bet-u garib al-ba¨da-na
‘the one after us, his house is
near’
al-bet-u garib al-ba¨da-na al-bet-u garib ba¨da-na *bet-u garib ba¨da-na
‘the one who’s house is near is
the one after us’
‘the one who’s house is near is
after us’
‘his house is near, one after us’
?ba¨da-na al-bet-u garib
‘the one who’s house is near is
after us’
2.4 Verb/Verb phrase+other element
2.4.1 Verb/Verb phrase+Noun: see 2.1.5
2.4.2 Verb/Verb phrase+Adjective: see 2.2.5
2.4.3 Verb/Verb phrase+Adverbial: see 2.3.5
2.4.4 Verb/Verb phrase+Verb/Verb phrase
definite+definite definite+indefinite indefinite+indefinite
?al-baka z-zi¨il al-baka zi¨il *baka zi¨il
‘the one who cried is the one
who got angry’
‘the one who cried got angry’ ‘one who cried got angry’
zi¨i l al-baka
‘the one who cried got angry’
?az-zi¨i
l al-baka az-zi{l baka *zi¨il baka
‘the one who got angry is the
one who cried’
‘the one who got angry cried’ ‘one who got angry is one
who cried’
92897_JOSS10-2_09_Dickins.indd 24592897_JOSS10-2_09_Dickins.indd 245 8/03/10 10:488/03/10 10:48
at University of Salford on April 15, 2010 http://jss.oxfordjournals.orgDownloaded from

BASIC SENTENCE STRUCTURE IN SUDANESE ARABIC
246
?baka z-zi¨il
‘the one who got angry cried’
2.4.5 Verb/Verb phrase+bipartite clause
definite+definite definite+indefinite indefinite+indefinite
?al-baka l-bet-u garib al-baka bet-u garib ?baka bet-u garib
‘the one who cried is the one
whose house is near’
‘the one who cried’s house is
near’
‘one who cried’s house is near’
bet-u garib al-baka
‘the one who cried his house
is near’
?al-bet-u garib al-baka al-bet-u garib baka *bet-u garib baka
‘the one who’s house is near is
the one who cried’
‘the one who’s house is near
cried’
‘his house is near, one who
cried’
baka l-bet-u garib
‘the one who’s house is near
cried’
2.5 Bipartite clause+other element
2.5.1 Bipartite clause+Noun: see 2.1.4
2.5.2 Bipartite clause+Adjective: see 2.2.4
2.5.3 Bipartite clause+Adverbial: see 2.3.4
2.5.4 Bipartite clause+Verb/Verb phrase: see 2.4.5
2.5.5 Bipartite clause+bipartite clause
definite+definite definite+indefinite indefinite+indefinite
?aÒ-ÒaÌb-u ma¨a-na l-bet-u garib
aÒ-ÒaÌb-u ma¨a-na bet-u garib*ÒaÌb-u ma¨a-na bet-u garib
‘the one whose friend is with us
is the one whose house is near’
‘the one whose friend is with
us, his house is near’
‘one whose friend is with us,
his house is near’
?bet-u garib aÒ-ÒaÌb-u ma¨a-na
‘the one whose friend is with us,
his house is near’
?al-bet-u garib aÒ-ÒaÌb-u ma¨a-na
al-bet-u garib ÒaÌb-u ma¨a-na *bet-u garib ÒaÌb-u ma¨a-na
‘the one who’s house is near is
the one who’s friend is with us’
‘the one who’s house is near is
one who’s friend is with us’
‘his house is near, one whose
friend is with us’
ÒaÌb-u ma¨a-na l-bet-u garib
‘the one who’s house is near is
one whose friend is with us’
92897_JOSS10-2_09_Dickins.indd 24692897_JOSS10-2_09_Dickins.indd 246 8/03/10 10:488/03/10 10:48
at University of Salford on April 15, 2010 http://jss.oxfordjournals.orgDownloaded from

BASIC SENTENCE STRUCTURE IN SUDANESE ARABIC
247
2.6 Discussion of material in sections 2.1–2.5.5
Although the material in sections 2.1–2.5.5 is set out using the word
classes and related phrase classes established in Section 2, the infor-
mation reinforces the analysis given in Dickins (2009, Section 10)
that the head of a phrase which begins with a definite particle is the
definite particle itself — regardless of the word class or phrase class
of what follows it (noun, adjective, adverbial, verb, bipartite clause,
etc.). This is particularly obvious in the middle column where the
two elements of the matrix bipartite clause are definite and indefinite.
All combinations of definite and indefinite in which a definite phrase
precedes an indefinite phrase are acceptable.
It is normally acceptable for the indefinite element to precede the
definite element in definite-indefinite bipartite clauses (bipartite
clauses comprising a definite and an indefinite element). In this case
a strong main stress almost always falls on the initial indefinite ele-
ment. In terms of thematic structure (as opposed to syntactic struc-
ture), such an initial indefinite element is the rheme and carries a
strong emotional force. A rheme-theme sentence ordering of this kind
is sometimes referred to as the ‘pathetic’ ordering (Firbas 1974: 117).
Sometimes, however, indefinite-definite word order in definite-indef-
inite bipartite clauses seems problematic, especially where the definite
element involves an adverb, verb, or further bipartite clause following
the al- (e.g. walad al-hina ‘the one who’s here is a boy’, walad az-zi¨il
‘the one who got angry is a boy’, walad al-bet-u garib ‘the one whose
house is near is a boy’. It is not clear to me whether this is gram-
matical matter (i.e. whether some or all of the forms of this type are
to be regarded as ungrammatical), or a matter of stylistic preference
(some forms are avoided because they seem inelegant), or something
else. Although a form like tarzi l-walad is in principle ambiguous
between a genitive structure, ‘the tailor of the boy / the boy’s tailor’
and a definite-indefinite bipartite clause having an indefinite-definite
word order, the intonational forms of these two structures are so dif-
ferent that they cannot in practice be confused. Tarzi l-walad ‘the
tailor of the boy’ always forms part or whole of a single intonation
unit, while tarzi l-walad ‘the boy’s a tailor’ involves two intonation
units, tarzi ‘a tailor’ and (a)l-walad ‘the boy’.
Definite-definite bipartite structures seem to be acceptable in some
cases. However, where the first definite element comprises al- + noun
and the second comprises al-+ adjective, as in al-walad az-za¨lan sup-
posedly meaning ‘the boy is the angry one’, they are avoided, and are
perhaps ungrammatical. The motivating factor here is presumably the
92897_JOSS10-2_09_Dickins.indd 24792897_JOSS10-2_09_Dickins.indd 247 8/03/10 10:488/03/10 10:48
at University of Salford on April 15, 2010 http://jss.oxfordjournals.orgDownloaded from

BASIC SENTENCE STRUCTURE IN SUDANESE ARABIC
248
5 Strictly speaking, forms with pronoun suffixes in Sudanese Arabic may be
either definite or indefinite. Thus in di mara jarat-na (R) ‘she’s a woman who’s a
neighbour of ours’ (lit: ‘she’s a woman our neighbour’), jarat-na ‘our neighbour’
agrees with the indefinite mara ‘a woman’, indicating that jarat-na is itself indefi-
nite. It is also possible, however, to say al-mara jarat-na (C) ‘the woman who’s our
neighbour’ in which jarat-na agrees with the definite al-mara indicating that jarat-na
is also definite.
intonationally similar attributive usage, e.g. al-walad az-za¨lan ‘the
angry boy’. In definite-definite bipartite clauses, the initial definite
phrase always seems to be thematic. The following are examples from
my data of definite-definite combinations:
al-ba†is da l-bitkun fi-hu
l-gadam, wa-n-na¨al bikun fi
l-ar∂
‘the batis [old word for the inside sole
of a shoe] is the bit in which the foot
is/should be, and the undersole is/
should be on the ground’ (Îiraf
sa¨biyya, 1986)
al-dolab al-biliff wara ‘the dolab [part of a waterwheel] is
that which turns behind’ (R)
In fact, definite-definite sentences in which one or neither of the
elements contains the definite particle al- are relatively common.
Examples are:
da hu or hu da ‘that’s him’ (R)
az-zol da ÒaÌb-i↑5‘that man’s my friend’ (C)
All the putative indefinite-indefinite bipartite structures are starred in
2.1–2.5.5 as ungrammatical. In fact, indefinite-indefinite clauses do
exist in Sudanese Arabic. The most common type is illustrated by the
following:
1. gamat nar/Ìariga ‘a fire broke out’ (C)
Example 1 consists of an indefinite verb followed by an indefinite
noun (subject/predicand), the latter taking strong rhematic stress.
This can be compared to English ‘A fire broke out’ in which the
rhematic sentence stress also falls on ‘fire’, rather than ‘broke out’
which, as the last lexical element in the utterance, is where sentence
stress typically falls in English.
Where an indefinite subject of an indefinite verb is unspecified
(that is does not have a following adjective or other elements further
defining it), indefinite-indefinite sentences in Sudanese Arabic always
take the word order Verb-Subject. Where an indefinite subject is
92897_JOSS10-2_09_Dickins.indd 24892897_JOSS10-2_09_Dickins.indd 248 8/03/10 10:488/03/10 10:48
at University of Salford on April 15, 2010 http://jss.oxfordjournals.orgDownloaded from

BASIC SENTENCE STRUCTURE IN SUDANESE ARABIC
249
further specified, and apparently where the verb is also further speci-
fied, it is possible to have Subject-Verb. Thus:
Ìarayig zayy di bitgum †awwali ‘fires like that break out all the time’
(C)
as well as:
bitgum †awwali Ìarayig zayy di ‘fires like that break out all the time’
(C)
In both these cases the verb phrase bitgum †awwali carries main sen-
tence stress and is treated as rhematic, while Ìarayig zayy di is the-
matic. The same thematic structure applies to the following:
ÌaÒalat Ìaja ‘something happened’ (C)
It is not clear from my data whether verb-like adjectives (e.g. active
participles) can occur, as verbs do, in indefinite-indefinite bipartite
clauses.
Another, though marginal, indefinite-indefinite bipartite clause-
type consists of a noun and an adjective, where the adjective answers
(or can be thought of as answering) the question kef ‘how’. Thus:
a. say kef? ‘how would tea be?’ (R)
b. say ma ba††al. ‘tea wouldn’t be bad.’ (R)
Similar forms occur without a preceding question:
say ma ba††al, mus kida? ‘Tea wouldn’t be bad / a bad idea,
don’t you think?’ (R)
Where the indefinite noun is further specified, e.g. by a phrase with
zayy ‘like’, indefinite-indefinite bipartites with nouns are much more
normal:
kalam zayy da ma kiwayyis ‘things [spoken about] like that are
not nice’ (R)
With the existential adverb fi ‘there is’, bipartite sentences in which
the other element is an indefinite noun are common:
fi muskila? ‘Is there a problem?’ (C)
With ¨ind ‘on’ used for general possession, like English ‘have’, and ma¨a
‘with’, used to give the sense ‘have [on oneself]’, indefinite-indefinite
bipartites are similarly quite common:
ma¨a-k sajayir? ‘have you (m.sg.) got any cigarettes
on you?’ (R)
92897_JOSS10-2_09_Dickins.indd 24992897_JOSS10-2_09_Dickins.indd 249 8/03/10 10:488/03/10 10:48
at University of Salford on April 15, 2010 http://jss.oxfordjournals.orgDownloaded from

BASIC SENTENCE STRUCTURE IN SUDANESE ARABIC
250
ma¨a-y↑ nafaren talata min
axwan-na as-subban
‘I’ve got/had two people from among
our young friends [lit: from our youth
brothers]’ (Ar-riwayat as-safawiyya
li-†uwwar 1924, 1974)
¨ind-u girus katira ‘he’s got a lot of money’ (R)
¨ind-i↑ falaja ‘I got a gap between my front teeth’
(R)
¨ind-i↑ Òadig le-y↑…‘I’ve got a friend of mine…’ (R)
¨inda-na Ìadis katir‘we’ve got a lot to talk about’ (R)
Other adverbial-noun indefinite-indefinite bipartites are occasionally
also found:
giddam-ak moya ‘there’s water in front of you (m.sg.)’
(R)
To recapitulate and amplify the argument in this section, although
there are restrictions on the combination of definite and indefinite
elements in bipartite clauses, it is definiteness and indefiniteness
which provides the most general parameters for describing the distri-
bution (possibilities of occurrence) of elements in such clauses, rather
than traditional categories, such as noun, nominal phrase, adjective,
adjectival phrase, verb, verb phrase, etc. As seen in Dickins (2009;
Section 10), definiteness similarly largely determines distribution in
the case of objects of verbs and prepositions: a verb phrase or prepo-
sitional object has to be definite (e.g. ma¨a az-za¨lan ‘with the angry
one’, but not *ma¨a za{lan ‘with an angry one’). The only exception
is where there object is a noun (together with its dependent phrase,
if there is one), in which case the object may be indefinite (ma¨a rajil
‘with a man’, as well as ma¨a r-rajil ‘with the man’). Taken together,
this evidence strongly supports the view that phrases involving initial
al- are to be analysed as definite phrases with an al- head.
Previously in this section, I have considered issues of thematic
structure (theme and rheme). I have assumed that thematic structure
is distinct from syntactic structure. It is also the case that issues of the
definiteness and indefiniteness of the constituent elements of bipartite
clause structure (identified previously in this paper as predicand and
predicate (cf. also Dickins 2009; Section 10), are not a relevant part of
the syntactic structure of the bipartite clause itself. That is to say, the
fact that one constituent element of a bipartite is definite, for example,
and the other one is indefinite, is not in itself relevant to identifying
which element is to be regarded as the predicand and which as the
predicate, any more than the definiteness of indefiniteness of ‘a cat’
92897_JOSS10-2_09_Dickins.indd 25092897_JOSS10-2_09_Dickins.indd 250 8/03/10 10:488/03/10 10:48
at University of Salford on April 15, 2010 http://jss.oxfordjournals.orgDownloaded from

BASIC SENTENCE STRUCTURE IN SUDANESE ARABIC
251
and ‘the dog’ are relevant to identifying which is the subject and
which the object in ‘a cat bit the dog’ (or ‘the cat bit a dog’).
If definiteness and indefiniteness — which as argued involve dif-
ferent phrase-types in Sudanese Arabic — are not relevant to predi-
cand-predicate analysis, it follows that neither are elements which
give rise to different phrase-types: the most obvious of these ele-
ments are, of course, word-classes, of which I have — on a fairly ad
hoc basis — identified the following for Sudanese Arabic: noun,
adjective, adverb (and adverbial), and verb. Finally, word order is
also not relevant to the identification of syntactic structure. The fact
that an element comes first in an utterance, for example, is not by
itself sufficient for us to conclude that that element is the subject, or
the predicand, etc.
If we then factor out, in analysing what is the predicand and what
is the predicate in Sudanese Arabic, questions of i. theme and rheme,
ii. definiteness, iii. word-class (and by extension phrase-class), and
iv. word order, what do we have left in determining what is the
predicand in a bipartite predicand-predicate clause, and what is the
predicate? The answer, I suspect, is nothing. That is to say, structur-
ally we can say of a bipartite clause that it has two major elements;
let us call these A and B rather than predicand and predicate (for
reasons which will become immediately obvious). However, in order
to identify which of A and B is the predicand and which the predi-
cate, we have to invoke non-structural considerations (or at least
considerations which are only structural at smaller levels of analysis,
e.g. of the constituent phrases). Once we have factored out such things
as theme and rheme, definiteness, word/phrase-class more generally,
and word order, all that we are left with in a putative predicand-
predicate analysis is two elements, either of which we could deem to
be the predicand or predicate.
I will not attempt to go further into theoretical analysis of what is
involved in the main body of this paper, since this would require
much more explicit adherence to the particulars of a linguistic theory
(axiomatic functionalism) than is reasonable in a paper of this kind.
Suffice it to say, that we can think of bipartite clauses in Sudanese
Arabic as being equative in a fundamental — i.e. true — syntactic
sense: such clauses are analogous to a mathematical statement of the
type X = Y. As is well known, there is no difference in mathematics
between X = Y and Y = X. In the same way, there is, I believe, no
properly syntactic distinction between a predicand and a predicate:
once we have identified one of the constituents of a bipartite clause
as the predicand we can identify the other as the predicate — or vice
92897_JOSS10-2_09_Dickins.indd 25192897_JOSS10-2_09_Dickins.indd 251 8/03/10 10:488/03/10 10:48
at University of Salford on April 15, 2010 http://jss.oxfordjournals.orgDownloaded from

BASIC SENTENCE STRUCTURE IN SUDANESE ARABIC
252
versa, if we prefer. The crucial thing about such a structure is the
presence of two elements in an equative relationship not how we label
them (just as the crucial thing in the case of X = Y / Y = X in a mathe-
matical context is the presence of the two elements, X and Y, in an
equative relationship, not the order in which we write these two
elements down).
In this regard, we may note the following possibilities in Standard
Arabic:
2. ma¨ruf-un anna-ka rajul-un
karim
‘it is known that you (m.sg.) are a gen-
erous man’ (lit: ‘known [is] that-you
man generous’)
3. min al-ma¨ruf-i anna-ka
rajul-un karim
‘it is known that you (m.sg.) are a gen-
erous man’ (lit: ‘from the known [is]
that-you man generous’)
4. al-ma¨ruf-u anna-ka
rajul-un karim
‘what is known is that you (m.sg.) are
a generous man’ (lit: ‘the known [is]
that-you man generous’)
Intuitively we are inclined to say, because of word order and use
of the same words almost throughout, that there is no difference in
sentence structure between 2, 3 and 4 — and I believe that this
intuitive view is correct. Traditional Arabic grammar, however,
requires every bipartite sentence to have a separately identifiable
predicand and predicate. Accordingly, it utilizes a set of criteria to
determine what is the predicand and what the predicate in particular
cases. Most importantly, a distinction is made not only between defi-
niteness and indefiniteness, but also between degrees of ‘specificity’
(taxÒiÒ). Specificity is a wider notion than definiteness, encompassing
both definite and indefinite nominals. A specificity hierarchy from
most to least specified can be established, roughly as follows: pronoun,
proper noun, demonstrative, noun/nominal with the definite article,
any of the previous as the annexed element in a construct (genitive),
indefinite annexion phrase, indefinite noun phrase involving an attrib-
utive adjective, bare indefinite noun or adjective (Watson 1993:
19–20; cf. also Wright 1971, 2: 260–4, Abdul-Raof 2001: 114–15;
Badawi, Carter and Gully 2004: 307–22). In analysing nominal sen-
tences — i.e. sentences which are made up of a predicand (mubtada})
and a predicate (xabar) — traditional Arabic grammar identifies the
predicand (mubtada}) as the more specified element and the predicate
(xabar) as the less specified one. Where both elements are equally
specified, the predicand precedes the predicate.
92897_JOSS10-2_09_Dickins.indd 25292897_JOSS10-2_09_Dickins.indd 252 8/03/10 10:488/03/10 10:48
at University of Salford on April 15, 2010 http://jss.oxfordjournals.orgDownloaded from

BASIC SENTENCE STRUCTURE IN SUDANESE ARABIC
253
In 2 ma¨ruf-un anna-ka rajul-un karim ‘it is known that you are a
generous man’ (lit: ‘known [is] that-you man generous’), ma¨ruf-un
‘known’ is a bare indefinite and therefore maximally unspecified, and
anna-ka rajul-un karim ‘that you are a generous man’ is the equiva-
lent of a definite noun; the complementizer anna- ‘nominalizes’
clauses, functioning as the clausal equivalent of the definite article
al- (cf. English ‘the fact that you are a generous man’). On the basis
that the more specified element is to be analysed as the predicand
(mubtada} ) and the less specified one as the predicate (xabar), the
predicand in ma¨ruf-un anna-ka rajul-un karim is thus anna-ka rajul-
un karim ‘that you are a generous man’ and the predicate (xabar) is
ma¨ruf-un ‘known’. Similarly, in 3, min al- ma¨ruf-i anna-ka rajul-un
karim ‘it is known that you are a generous man’ (lit: ‘from the known
[is] that-you man generous’) min al-ma¨ruf-i ‘from the known’ is
traditionally analysed as the predicate, and anna-ka rajul-un karim
‘that you are a generous man’ as the predicand. Min al-ma¨ruf-i ‘from
the known’ is to be analysed as indefinite and therefore fairly unspec-
ified (its definite correspondent being alla∂i min al-ma¨ruf-i ‘that
which is known’ (more lit: ‘the from the known’), while anna-ka
rajul-un karim ‘that you are a generous man’, as already discussed, is
definite.
Consider now 4 al-ma¨ruf-u anna-ka rajul-un karim ‘what is known
is that you are a generous man’ (lit: ‘the known [is] that-you man
generous’). Here, the structure in 2 and 3 (predicate preceding predi-
cand) is — remarkably — reversed. In 4, al-ma¨ruf-u ‘the known
[thing]’ is the predicand while anna-ka rajul-un karim ‘that you are
a generous man’ is the predicate. In 4, al-ma¨ruf-u ‘the known [thing]’
and anna-ka rajul-un karim ‘that you are a generous man’ are to be
taken as equally specified: al-ma¨ruf-u has the definite article, while
anna-ka rajul-un karim has (is introduced by) the complementizer
anna-, which, as argued above, is equivalent to the definite article.
Given that where two elements in a nominal sentence are equally
specified, the element which comes first is the predicand and the
element which comes second is the predicate, in 4 the predicand is
al-ma¨ruf-u ‘the known [thing]’, while the predicate is anna-ka rajul-un
karim ‘that you are a generous man’. This kind of approach makes it
possible to ‘operationalize’ Arabic grammar: it allows us to determine
what is and is not the predicand and predicate in all cases. As I have
suggested, however, I do not think it provides a genuinely structural
account of what is going on.
The notion of equativity is worth amplifying somewhat here. It is
clear from the tables in sections 2.1–2.5.5 that bipartite sentences
92897_JOSS10-2_09_Dickins.indd 25392897_JOSS10-2_09_Dickins.indd 253 8/03/10 10:488/03/10 10:48
at University of Salford on April 15, 2010 http://jss.oxfordjournals.orgDownloaded from

BASIC SENTENCE STRUCTURE IN SUDANESE ARABIC
254
involving, for instance, a noun and another noun, or a noun and an
adjective are equative:
al-walad tarzi ‘the boy is a tailor’ (C)
al-walad za¨lan (minn-ak) ‘the boy is angry (with you [m.sg.])’
(C)
Similarly:
al-walad fi-l-bet‘the boy is in the house’ (C)
It is also clear from previous discussion that the relationship between
tarzi ‘a tailor’ and at-tarzi ‘the tailor’ is one of indefiniteness vs. def-
initeness. Similarly, the relationship between za¨lan (minn-ak) ‘angry
(with you [m.sg.])’ and az-za¨lan (minn-ak) ‘the one who is angry
(with you [m.sg.]) / the angry-[one]-(with you [m.sg.])’ is also one of
indefiniteness vs. definiteness. Accordingly, just as the indefinite tarzi
without a preceding al- means ‘a tailor’ and the definite at-tarzi
means ‘the tailor’, so the indefinite za¨lan can be thought of meaning
‘an angry [one]’ while the definite az-za¨lan means ‘the angry [one]’.
By extension the indefinite fi-l-bet can be thought of meaning ‘a in-
the-house-[one] / a [one] who is in the house’, while the indefinite
al-fi-l-bet means ‘the in-the-house-[one] / the [one] who is in the
house’. Similarly, the indefinite zi¨il (minn-ak) can be thought of as
meaning ‘a got-angry-(with-you) [one]’ while the definite az-zi¨il
(minn-ak) means ‘the got-angry-(with-you) [one] / the [one] who got
angry with you’. Finally, the indefinite Òahb-u fi-l-bet can be thought
of as meaning ‘a [one] his/whose friend is in the house / a his-friend-
is-in-the-house [one]’, while the definite aÒ-Òahb-u fi-l-bet means ‘the
[one] his/whose friend is in the house / the his-friend-is-in-the-house
[one]’.
Given all this, consider the following (all of which are constructed
examples):
al-walad tarzi ‘the boy is a tailor’
al-walad za¨lan (minn-ak) ‘the boy is an angry(-with-you [m.sg.])
[one]’
al-walad fi-l-bet‘the boy is an in-the-house [one]’
al-walad zi¨il (minn-ak) ‘the boy is a got-angry(-with-you
[m.sg.]) [one]’
al-walad Òahb-u fi-l-bet‘the boy is a his-friend-[is]-in-the-
house [one]’
or better:
‘the boy is a-his-friend is an-in-the-
house [one]’
92897_JOSS10-2_09_Dickins.indd 25492897_JOSS10-2_09_Dickins.indd 254 8/03/10 10:488/03/10 10:48
at University of Salford on April 15, 2010 http://jss.oxfordjournals.orgDownloaded from

BASIC SENTENCE STRUCTURE IN SUDANESE ARABIC
255
That is to say, there is no difference — in one sense at least — in
Sudanese Arabic between the equivalent of ‘the boy is one who got
angry’ and ‘the boy got angry’. While the distinction between equa-
tives and non-equatives in English is very clear, in Sudanese Arabic
this distinction is, in syntactic terms at least, non-existent.
3. Monopartite Sentences
In addition to bipartite clauses, Sudanese also has monopartite
clauses. These are most obviously represented by clauses which con-
sist of a verb phrase, such as:
5. masu (as-sug) ‘they went/have gone (to the market)’
(C)
That a verb phrase of this type constitutes a full sentence seems
uncontroversial, and conforms to both traditional Arabic grammar
and modern Western analyses of both Standard Arabic and the dia-
lects. This analysis can, however, happily be extended to participles,
such as the following active participle usage. Thus:
6. masin (as-sug) ‘they/you (pl.)/we are going (to the
market)’ (C)
In fact the form masin, on its own, or the corresponding singulars
masi (m.sg.) or masya (f.sg.) are standard ways of saying that one is
leaving or, with a question intonation, of asking a person or people
if they are leaving, (e.g. at the end of a visit to one’s house). There is
thus very good reason communicatively for regarding not only verb
phrases but also corresponding ‘participle phrases’ as complete non-
elliptical sentences. Thus, ‘she’s in pain’ can be expressed by the sin-
gle word (f.sg. active participle) mit}allima. Support for the view that
‘participle phrases’ constitute full sentences is provided by the use of
such phrases in complex sentences. Thus:
7. najib ¨ali u-nasaggil-u
baskatib al-mina li}inn-u
muÌtajin le-baskatib
‘We’ll bring Ali and employ him as
the chief clerk of the port, because
“needing” [(m.pl.) i.e. “we need”] a
chief clerk’ (Ar-riwayat as-safawiyya
li-†uwwar1924, 1974)
8. di gafla l-o∂a ¨ale-ha u-xayfa
ta¨mal fi nafís-a Ìaja
‘She’s closed the room on/against
her(self ), and fearing [(f.sg.) i.e. “I
fear”] that she will do something to
herself ’ (Xalt-i↑ ∑afiyya, 1986)
92897_JOSS10-2_09_Dickins.indd 25592897_JOSS10-2_09_Dickins.indd 255 8/03/10 10:488/03/10 10:48
at University of Salford on April 15, 2010 http://jss.oxfordjournals.orgDownloaded from

BASIC SENTENCE STRUCTURE IN SUDANESE ARABIC
256
6 Some idiomatic phrases — including those which involve a participle — only
occur without a predicand. An example is bayta ma¨a[-k] literally ‘overnighting with
[you (m.sg.)]’, meaning idiomatically ‘[you] have a hangover’. In this idiomatic
phrase, the active participle bayta only ever occurs in the f.sg. form. It is not possible
to also have a predicand hi ‘it (f.sg.)’ (thus *hi bayta ma¨a-k) in this idiomatic sense.
Other examples involving a f.sg. predicate in which a predicand cannot occur are,
with the active participle, farga ma¨a-w ‘he’s mad’ (lit: ‘making a difference with
him’) and, with a verb, ma bitafrig ma¨a-y↑ ‘it doesn’t make any difference to me’,
fakkat minn-u ‘he went mad’ (lit: ‘it [f.sg.] left him’). Other impersonal usages in
which it is not possible to have a predicand include forms involving the 3. m.sg.
imperfect yahimm, as in ma yahmma-ni↑ bi-n-nas del ‘I’m not concerned about
these/those people’ (lit: ‘[it] not-concerns-me with these/those-people’).
7 Almost all verbs from which fa¨lan forms are derived are intransitive; the tran-
sitive gibil, whose derivative gablan is discussed in Dickins (2009; Section 9), is an
exception. Almost all fa¨lan forms are derived from verbs which have fi¨il perfect, and
many fa¨lan forms refer to emotional or physical states experienced by people, e.g.
farÌan ‘joyful’, ta¨ban ‘exhausted’, na¨san ‘drowsy’, bardan ‘cold [as an experienced
physical sensation]’; thus ana bardan ‘I’m [feeling] cold’ — cf. barid ‘[objectively]
cold’, e.g. as-say da barid ‘that tea’s cold’.
In example 7 the subordinate structure beginning li}inn-u ‘because’
(with a dummy 3.m.sg. pronoun suffix — u), contains the monopar-
tite clause muÌtajin le-baskatib. The active participle muÌtajin ‘need-
ing (m.pl.)’ has the same ‘we’ subject-referent as the main clause verbs
najib ‘we bring’ and nasaggil ‘we employ’. Apart from an a priori
theoretical commitment (which I do not have) I can see no reason to
believe that the monopartite muÌtajin le-baskatib ‘needing a chief
clerk’ is elliptical here, e.g. for the bipartite niÌna muÌtajin le-baskatib
‘we [are] needing a chief clerk’. Example 8 is even more interesting
than 7. Here the active participle in the first coordinated clause gafla
(f.sg.) mirrors the active participle in the second coordinated clause
(after u- ‘and’) xayfa (also f.sg.). However, the referent is different. In
the first clause, the person being referred to is the unfortunate Su}ad
(Su¨ad), who has locked herself in her room after being told by her
father that she cannot marry the man she loves. In the second clause,
xayfa refers to the gallant Aunt Safiyya (Xalt-i↑ ∑afiyya of the series
title), who has come to rescue Su}ad from her predicament.6
There seems no good reason to confine monopartite sentences to
those whose head is a verb or participle. Many verbs in Sudanese do
not have a standard active participle form, but do have a fa¨lan form.7
Thus the following both seem to be reasonable non-elliptical sen-
tences in Sudanese Arabic:
9. zi¨il (minn-ak) ‘he got angry (with you [m.sg.])’ (C)
10. za¨lan (minn-ak) ‘he is angry (with you [m.sg.])’ (C)
92897_JOSS10-2_09_Dickins.indd 25692897_JOSS10-2_09_Dickins.indd 256 8/03/10 10:488/03/10 10:48
at University of Salford on April 15, 2010 http://jss.oxfordjournals.orgDownloaded from

BASIC SENTENCE STRUCTURE IN SUDANESE ARABIC
257
In the same light consider the following:
11. galu ta¨ban jiddan jiddan ‘they said that he was very very tired’
(R) (more lit: ‘they said very very tired’)
Here the fa¨lan form ta¨ban is in a subordinate clause, and has a dif-
ferent referent (m.sg.) from the referent (subject) of the main verb
galu ‘they said’.
In fact, not only active participles and fa¨lan adjectives, but also any
adjective/adjective phrase, noun/noun phrase, adverb/adverbial, etc.
seems to be able to function as a monopartite sentence in Sudanese
Arabic. Thus, standard ways of saying ‘where are you going’ are wen?
‘where?’ and ¨ale wen? ‘to/towards where?’, as well as masi (m.sg.)/
masya (f.sg.)/masin pl.) wen?. The standard question if someone
knocks on one’s door is min-u↑ ‘who?’. The word muskila is frequently
used on its own where in English it would be more natural to say,
‘That’s a problem’. In the same light, consider the following:
fikra. intu tag¨udu tit¨assu. ana
afattis dawa.
‘An idea [i.e. I have an idea]. You
(m./com.pl.) sit and eat supper. I’ll
look for some medicine’. (R)
ma ta¨mal kida — ¨eb. ‘Don’t do that. It’s wrong.’ (said to
small child) (R)
a. fi?‘Is he there?’
b. fi‘Yes.’ (R)
fi di ma¨a-kum ‘We agree with you (m./com.pl.)
on this.’ (R) (lit: ‘on this, with you’
[m./com. pl.])
ana gariba?! ma xalt-ak ya
walad-i↑
‘Am I a stranger?! You see/But
[unstressed ma] I’m your (m.sg.) aunt
my boy.’ (Xalt-i↑ ∑afiyya 1986)
English arguably makes a distinction between presentative usages and
presumably elliptical non-presentatives. An example of a presentative
usage is ‘cow’ as said by a small child looking out of a train window
every time she sees a cow. A elliptical non-presentative occurs where
the child is asked ‘What’s that?’ and she replies ‘A cow’. In cases
where there is a non-correspondence between the putative referent of
the presentative utterance we are inclined to say that the person is
wrong — or ‘that’s wrong’ — and less inclined to say that what the
person has said is untrue. In the case of elliptical non-presentatives,
the claim of the untruth of the utterance — ‘that’s not true’ (etc.) —
seems more acceptable.
92897_JOSS10-2_09_Dickins.indd 25792897_JOSS10-2_09_Dickins.indd 257 8/03/10 10:488/03/10 10:48
at University of Salford on April 15, 2010 http://jss.oxfordjournals.orgDownloaded from

BASIC SENTENCE STRUCTURE IN SUDANESE ARABIC
258
8 It may be that Arabic equative (‘predicand-predicate’) structures are an
example of what Mulder terms ‘coordination’. His use of this term has little, if
any, connection with more standard uses of ‘coordination’ in linguistics to
describe the functioning of ‘and’ and related forms. Rather, what Mulder means
by ‘coordination’ is a relationship of bilateral functional independency between two
elements in syntax such that neither one of the two elements functionally implies
the other (cf. Mulder 1989: 288–93; 445–6; also Dickins 1998: 72). Note that
the distinction which Mulder makes between functional and occurrence depend-
ency is particularly important in understanding what he means by functional
dependency). I have argued in sections 2.6 and 3 that syntactically ‘predicand’
and ‘predicate’ do not imply one another; these are not, in fact, properly speaking
two distinct syntactic positions (‘slots’). I have also suggested in Section 3 that
Arabic monopartite sentences are to be analysed as cases in which a single element
occupies the one position which, in bipartite sentences, is occupied by two elements
(‘predicand’ and ‘predicate’).
In addition to subordination and coordination in syntax, Mulder recognizes
also what he terms interordination – that is a situation in which both elements
functionally imply one another. The structures which Mulder gives as examples of
both coordination and itnterordination (e.g. Mulder 1989: 290) all seem to me
extremely problematic. I will not pursue this issue in detail here. However, I believe
that the issue of ‘ordination’ (subordination, coordination, and interordination) in
axiomatic functionalism needs to be looked at again. It may be that the notions of
Sudanese Arabic certainly has elliptical utterances: ganamaya
‘a sheep’ as a response to sufti sin-u↑ ‘What did you (f.sg.) see?’ is
elliptical. However, I do not believe that ganamaya ‘a sheep’ in
response to da sin-u↑ ‘What’s that?’ is properly speaking elliptical.
The only difference between a presentative usage ganamaya ‘a sheep’
in Sudanese Arabic, and ganamaya ‘a sheep’ as a response to da
sin-u↑ ‘What’s that?’ seems to be that the latter involves presupposi-
tion of the existence of the entity being considered (the sheep):
it is not presented ‘out of the blue’. This is essentially a question
of the thematic (theme-rheme) organisation of the utterance (or,
better, the real-world psychological facts to which theme and rheme
can be related). I have already argued (Section 2.6) that bipartite
predicand-predicate structures have a strange syntactic structure:
once issues such as theme and rheme are abstracted away, what one
is left with is not so much two distinct syntactic positions (predi-
cand and predicate), but what looks more like two equatively-
related ‘slots’ in one syntactic position. (As noted earlier, I will not
try to pursue this issue further here, since it would require a pro-
found consideration of the fundamentals of linguistic theory.) If
this analysis is at all plausible, it will be seen that having one ‘slot’
in this position — a monopartite sentence — is not very different
from having two.8
92897_JOSS10-2_09_Dickins.indd 25892897_JOSS10-2_09_Dickins.indd 258 8/03/10 10:488/03/10 10:48
at University of Salford on April 15, 2010 http://jss.oxfordjournals.orgDownloaded from

BASIC SENTENCE STRUCTURE IN SUDANESE ARABIC
259
coordination and interordination are not separately necessary — or distinct — from
that of ‘syntheme’ (see Dickins 2009; footnote 13). In this case, Arabic equative/
bipartite/predicand-predicate structures can be regarded as synthemes, just as it may
be possible to regard al — and a following clause together as a syntheme.
4. ‘Scrambled’ Bipartite Clauses
The most straightforward ‘linearisation’ (realisation) of grammatical
structures is such that all elements of one grammatical structure occur
together. However, exceptions to such straightforward linearisation
are well known: in the English ‘That man I know’, the verb phrase
constituent ‘know that man’ is realized discontinuously (for reasons
of thematic focus). Sudanese Arabic exhibits numerous cases of such
‘scrambling’. Here I will consider V-S-O/C word orders — i.e. verb-
subject-object word order with respect to those verbs which take a
standard object, and the corresponding verb-subject-complement
word order for verbs of ‘being’, ‘becoming’, ‘continuing to be’ which
take what is traditionally referred to as a subject-complement.
V-S-O/C structures are occasionally found with verbs which take
a standard object, as follows (with the V, S, and O/C elements
marked after the relevant item):
maset (V) inta (S) dungula (O)? ‘did you (m.sg.) go to Dongola?’ (R)
sufta (V) i nta (S) az-zol da (O),
mus kida?
‘You (m.sg.) saw that man, didn’t
you?’ (R)
axadta (V) ana (S) xams†asar
yom (O) fi ÒaÌra↑ l-¨atmur
‘I spent fifteen days in the Al-
Atmour Desert’ (R)
jit (V) umbariÌ ana (S) s-sa¨a
sab¨a hina (O)
‘I came yesterday at seven-o-clock
here’ (R)
The following example has a prepositional object, rather than a nom-
inal object
jit (V) ana (S) †awwali fi l-makan
al-¨amm bita¨-na (O)
‘I came immediately to our general
place’ (R)
In all of these cases, the subject is pronominal, and is very commu-
nicatively ‘light’, while the verb is either the 1.sg. or 2.m.sg. perfect.
In theme-rheme terms, it would appear that in these utterances the
subject pronoun constitutes a ‘minimal’ theme (perhaps an ‘after-
thought’ theme), whose purpose is mainly to remind the reader who
the subject is, and perhaps more specifically to differentiate between
1.sg. and 2.m.sg., which have the same form in the perfect tense.
Syntactically, the most sensible analysis seems to be to regard these
examples as scrambled bipartite predicand-predicate structures with a
92897_JOSS10-2_09_Dickins.indd 25992897_JOSS10-2_09_Dickins.indd 259 8/03/10 10:488/03/10 10:48
at University of Salford on April 15, 2010 http://jss.oxfordjournals.orgDownloaded from

BASIC SENTENCE STRUCTURE IN SUDANESE ARABIC
260
9 In my discussion here I have ignored the interesting phenomenon of comple-
ments which are verb phrases or themselves bipartite clauses. An example involving
a verb-phrase complement is: zaman kan (V) nas-na (S) bi-talbas jilud (O) bass
(Îiraf sa¨biyya) ‘in the past our people used to wear leather [shoes] only’. Here, the
complement of the verb kan ‘to be’ is the verb-phrase bi-talbas jilud, itself consisting
of a verb bi-talbas ‘they wear’ and an object jilud ‘leather(s)’.
discontinuous predicate (verb phrase) constituent. Thus in maset (V)
inta (S) dungula (O)? ‘did you go to Dongola?’, for example, the
predicate constituent maset dungula ‘went to Dongola’ is interrupted
realisationally by the predicand inta ‘you’.
With those verbs which are traditionally analysed as taking a comple-
ment rather than an object (verbs of ‘being’, ‘becoming’, ‘continuing to
be’, etc.) V-S-O/C word order is much more common. Examples are:
kan (V) ism-u (S) muÌammad (C) ‘his name was Mohammed’ (R)
takun (V) hi (S) Òagayra fi Ìajma-
ha (C)
‘she must have been small in size’ (R)
fi kullu man†iga fi fannan bikun
(V) hu (S) mahir jiddan (C)
‘in every region there is a singer who is
very talented’ (R)
fi itnen u-xamsin kanat (V) asar
al-Ìarb al-¨alamiyya (S) lissa¨
mawjuda (C) fi l-man†iga di
‘in [19]52 the marks of the [Second]
World War were still present in that
region’ (R)
ba¨d ma ¨amal kida, biga (V) hu
zat-u (S) ga¨id sakit bi-la xidma
(C)
‘after he did that, he got to the stage
where [lit: “became”] doing nothing,
without anything to do’(Mafahim,
1986)
biga (V) le-hu aj-jari be kir¨e-hu
(S) asra¨ (C) ya jama¨a
‘Running [on his legs] became quicker
for him, my friends [lit: “oh, group”]’
(i.e. quicker than going by horse)
(Dukkan wadd al-baÒir, 1986)
Fairly frequently, the verb in V-S-C structures of this kind fails to
agree, particularly where this verb is kan ‘to be’:
[an-nas del] umbariÌ kan (V)
niÌna (S) ma¨a-hum (C) bi-l-lel
‘those people] yesterday we were
with them at night’ (R)
Examples such as this last one suggests that kan is on its way to becom-
ing grammaticalized — as a past tense marker — rather than a verb
proper in Sudanese Arabic. However, this process still has a long way
to go, and it would appear that with complement verbs such as kan,
as with more standard transitive verbs, the best syntactic analysis of
V-S-C forms remains that they are ‘scrambled’ predicand-predicate
clauses.9
92897_JOSS10-2_09_Dickins.indd 26092897_JOSS10-2_09_Dickins.indd 260 8/03/10 10:488/03/10 10:48
at University of Salford on April 15, 2010 http://jss.oxfordjournals.orgDownloaded from

BASIC SENTENCE STRUCTURE IN SUDANESE ARABIC
261
5. Conclusion and Prospects
On the basis that definiteness and indefiniteness yield fundamentally
different syntactic categories (Dickins 2009), I have considered in this
paper basic sentence structure in Sudanese Arabic. I have argued that
Sudanese Arabic has two basic types of sentence structure: an essen-
tially equative bipartite (‘predicand-predicate’) structure and a mon-
opartite (‘predicate only’) structure.
I believe that many of the analyses presented here apply not only
to Sudanese Arabic, but also to other dialects and to Standard Arabic.
In particular, I believe that sentence structure in both Standard Arabic
and Arabic dialects generally can be at least partially analysed in terms
of syntactically equative bipartite, plus monopartite sentence types.
Address for correspondence: J.Dickinsesalford.ac.uk.
REFERENCES
Abdul-Raof, H., 2001. ‘On the subject in Arabic’, Journal of Semitic Studies. 46,
1:97–120
Badawi, E., M.G. Carter, and A. Gully, 2004. Modern Written Arabic: a Compre-
hensive Grammar. (London and New York)
Bohas, J., J.-P.Guillaume and D.E. Kouloughli, 1990. The Arabic Linguistic Tradition.
(London)
Danes, F. (ed.). 1974. Papers on Functional Sentence Perspective. (Prague)
Dickins, J. 1998. Extended Axiomatic Linguistics. (Berlin and New York)
—— 2006. ‘Khartoum Arabic’, in Encyclopaedia of Arabic Language and Linguistics
(vol. 2) (Leiden), 559–71
—— 2007. Sudanese Arabic: Phonematics and Syllable Structure. (Wiesbaden)
—— 2009. ‘Relative clauses in Sudanese Arabic’. Journal of Semitic Studies 54,
2:535–71
Dukkan wadd al-baÒir.1986. Omdurman Radio series (drama)
Firbas, J. 1974. ‘Some aspects of the Czechoslovak approach to problems of func-
tional sentence perspective’, in F. Danes (ed.), Papers on Functional Sentence
Perspective. (Prague)
Hervey, S.G.J. 1979. Axiomatic Semantics. (Edinburgh)
—— 1982. Semiotic Perspectives. (London)
Îiraf sa¨biyya 1986. Omdurman Radio series on traditional crafts
Mafahim 1986. Omdurman Radio series (drama)
Mulder, J.W.F. 1968. Sets and Relations in Phonology. (Oxford)
—— 1989. Foundations of Axiomatic Linguistics. (Berlin)
Mulder, J.W.F. and Hervey, S.G.J. 1972. Theory of the Linguistic Sign. (The Hague)
—— 1980. The Strategy of Linguistics. (Edinburgh)
Ar-riwayat as-safawiyya li-†uwwar 1924 [‘The Oral Accounts of the 1924 Revolu-
tionaries’]. 1974. (Khartoum)
Watson, J.C.E. 1993. A Syntax of ∑an¨ani Arabic. (Wiesbaden)
Wright W. 1971. A Grammar of the Arabic Language (2 vols. re-published as single
vol.). (Cambridge)
Xalt-i↑ ∑afiyya. 1986. Omdurman Radio series (drama)
92897_JOSS10-2_09_Dickins.indd 26192897_JOSS10-2_09_Dickins.indd 261 8/03/10 10:488/03/10 10:48
at University of Salford on April 15, 2010 http://jss.oxfordjournals.orgDownloaded from