ASPECTS OF FULFULDE SYNTAX AND

MORPHOLOGY

Mary Hollis McIntosh-

Submitted for the degree of Ph.D. ,

School of Oriental and African Studies,
University of London.
 ProQuest Number: 10672704

All rights reserved

INFORMATION TO ALL USERS

The quality of this reproduction is d e p e n d e n t upon the quality of the copy subm itted.

In the unlikely e v e n t that the a u thor did not send a c o m p le te m anuscript

and there are missing pages, these will be noted. Also, if m aterial had to be rem oved,
a n o te will ind ica te the deletion.

uest
ProQuest 10672704

Published by ProQuest LLC(2017). C opyright of the Dissertation is held by the Author.

All rights reserved.

This work is protected against unauthorized copying under Title 17, United States C o d e
M icroform Edition © ProQuest LLC.

ProQuest LLC.
789 East Eisenhower Parkway
P.O. Box 1346
Ann Arbor, Ml 4 8 1 0 6 - 1346
 Abstract

On the basis of evidence from the Kaceccereere dialect of Southern

Zaria in Northern Nigeria, this study aims to reveal some of the
principles that underly Fulfulde (or Fula) verbal morphology, and
thus to improve on the ad hoc nature of earlier descriptions* It
shows that many morphological issues may be resolved by reference
to the syntax and phonology of the language, and. that there is also
a need for morpho-syntactic and morpho-phonological levels of des­
cription. In thus accounting for the morphology, certain new cate­
gories and principles are established for Fulfulde. Most important
amongst these are the syntactic categories of topic and focus; and
the morpho-phonologicdl principle termed the fsuffixal precedence
hierarchy1, which determines the surface-form of certain verbal suf­
fixes when they co-occur. An attempt is also made first to identify
morpho-syntactic properties such- as active, middle and passive, posi­
tive and negative, which are terms in categories such as voice and
polarity; and then to explain the principles that underly the morph­
ological realisation of such properties in the various tenses, or
'conjugations’ of Fulfulde.
 2

Table of Contents

page
List of T a b l e s ....................................................... 7

Acknowledgements ..................................................... 8

Introduction 9
0.1. L a n g u a g e ................................................. 11
0.2. I n f o r m a n t s ..................................................13
0.3. Aims of this S t u d y ........................................ 14
0.4. Conventions a n d ' A b b r e v i a t i o n s ...................... . . 15
0.4.1. Presentation of Examples ............................. 15
0.4.2. Translation of Examples ........................ 16
0.4.3. U n a c c e p t a b i l i t y ........................................... 17
0.4.4. Underlining ...................................... 17
0.4.5. Abbreviations .......................................... 17
F o o t n o t e s .............................................................. 19

Chapter 1 : Linguistic Background Information: Phonology, Conson-

and Alternation and the Nominal Class System
1.0. I n t r o d u c t i o n ............................................... 20
1.1. P h o n o l o g y .................................................... 21
1.1.1. Vowels ................................................. 21
1.1.2. Consonants ...................................... 23
1.1.3. Syllable Structure ................................... 24
1.1.3.1. 1Regular1 Syllable S t r u c t u r e ......................... 24
1.1.3.2. 'Irregular1 Syllable Structure ..................... 28
1.1.3.3. Syllable Structure: Conclusions .................. 30
1.1.4. Prosodic Features ...................................... 30
1.1.4.1. Final G l o t t a l i t y .................................. 30
1.1.4.2. Word S t r e s s ............................................. 31
1.1.4.3. I n t o n a t i o n .............. 33
1.1.4.4. Hypertonic W o r d s .......................................36
1.1.4.5. Prosodic Features: C o n c l u s i o n s .......................37
1.1.5. Transcription ........................................... 37
1.1.5.1. Vowels in Word-final Position .................... 37
1.1.5.2. 'W e a k 1 C o n s o n a n t s ............................... 38
1.1.5.3. Transcription: G e n e r a l ............................... 40
1.2. Consonant Alternation .................................... 42
1.2.1. Consonant Alternation in Verbal Complexes ........... 43
1.2.2. Consonant Alternation in Nominals .................... 46
1.2.3. Consonant Alternation in Class-Suffixes .............. 47
1.2.3.1. Suffix-Grades ......................................... 48
1.2 .3.2. Which grade with a given S t e m ? ..................... 51
1.2. 3,3. Suffix-Grades: C o n c l u s i o n s ........................... 52
1.2.4. Consonant Alternation in Root-final Position . . . . 53
1.2.5. Consonant Alternation: Conclusions ............. 55
1.3. The Nominal Class S y s t e m .................................... 56
1.3.1. The C l a s s e s ................................................ 57
1.3.1.1. Singular and Plural Classes ......................... 58
1.3.1.2. The 'Meaning1 of N o m i n a l s .............................. 61
1.3.2. Nominals: Nounsf Adjectives, P a r t i c i p l e s ................ 62
1.3.3. Nominal Class System: Conclusions .................. 64
1.4. C o n c l u s i o n s ................................................ 65
F o o t n o t e s ................................................................ 66
 3

page
Chapter 2 : The Fulfulde Sentence, Verbal Complex and Noun Phrase
2.0. I n t r o d u c t i o n ................................................. 68
2.1; The Fulfulde S e n t e n c e ...................................... 69
2.1.1. Verbal Sentences ...................................... 71
2.1.1.1. D e l e t i o n ................................................. 71
2.1.1.2. Verbal P r o p e r t i e s ...................................... 72
2.1.1.3. Negation ......................................... 73
2.1.2. Non-Verbal Sentences .................................. 75
2.1.2.1. Non-Verbal Sentences: A t t r i b u t i o n ............... 75
2.1.2.2. Non-Verbal Sentences: Identification . . . . . . . 78
2.1.2.3. Non-Verbal Sentences: Location and Existence . . . 82
2.1.2.4. Non-Verbal Sentences: 'Rag-bag1 '85
2.1.2.5. Non-Verbal Sentences: C o n c l u s i o n s ............... 86
2.1.3. The Fulfulde Sentence: Conclusions .................. 86
2.2 The Verbal Complex ...................................... 86
2.2.1. The Verbal R o o t ........................................... 87
2.2.2. Other Appendages in Verbal Complexes . . .............. 91
2.2.2.1. Subject and Object P r o n o u n s ............................. 91
2.2.2.2. E x t e n s i o n s ............................................. 93
2 .2.2.3. The A n t e r i o r i t y - M a r k e r ................................ 96
2.2.2.4. J_E and c f o n ................... -.........................97
2.2.3. Intonation . . . . . 101
2.2.4. The Verbal Complex as a Morphological W o r d ........... 105
2.3. The Noun Phrase . . . ................................... 107
2.3.1. Topic . .................................................107
2.3.2. Focus ................................................... 110
2.3.3. Answer to a Q u e s t i o n ................................... Ill
2.3.4. Other Types of N P ........................................ 113
2.3.5. The N P : C o n c l u s i o n s ...................................115
2.4. The Fulfulde Sentence, Verbal Complex and NP:
Conclusions . 115
F o o t n o t e s .................................................. 116

Chapter 3 : Paradigms of Verbal Inflections: Syntax and Semantics

3.0. i n t r o d u c t i o n ............................................... 119
3.1. Derivational Processes .......................... ...119
3.1.1. Extensions . ............................... 120
3.1.2. N o m i n a l s ................................................. 121
3.2. Inflectional Processes .................................. 122
3.2.1. 14 Inflectional Paradigms: 14 'Conjugations' . . . . 122
3.3. Verbal Properties and Verbal Categories ............ 123
3.4. 'Conjugational' Properties .............................. 124
3.4.1. Indicative and Non-Indicative Mood ................... 124
3.4.2. Completive and Incompletive Aspect ................... 129
3.4.3. Positive and Negative Polarity ............... . . . 132
3.4.4. Relative and Non-relative (or General) .............. 133
3.4.5. Emphatic and Non-Emphatic .............................. 133
3.4.6. Stative and Progressive ................................ 133
3.4.7. Imperative and Exhortative ............................ 134
3.4.8. Subjunctive ............................................. 134
3.4.9. V a g u e .................................. 135
3.5. Properties that do not distinguish one conjugation
from another . . 135
3.5.1. Active, Middle and Passive Voices . . . . . .......... 135
3.5.1.1. Middle Voice: Reflexive arid E r g a t i v e ................136
3.5.1.2. Voice and Extensions ............................... 140
 4

page
3.5.2. Anteriority ............................................ 143
3.5.3. Properties of Person, Number and Case ................ 144
3 *5.3.1. 1st, 2nd and 3rd P e r s o n ............................... 144
3.5.3.2. Singular and Plural N u m b e r .......................... 147
3.5.3.3. Subject and Object C a s e ............................... 148
3.6. Indicative Mood: Conjugations in Neutral Sentences . . 149
3.6.1. Completive A s p e c t ...................................... 150
3.6.1.1. Stative and General C o m p l e t i v e ...................... 150
3.6.1.2. Negative Completive .................................. 153
3.6.2. I-ncompletive A s p e c t ......................................154
3.6.2.1. Progressive and General Incompletive .............. 154
3.6.2.2. Negative Incompletive . ............................... 157
3.6.2.3. Vague Incompletive .................................. 158
3.7. Indicative Mood: Conjugations in Non-neutral Sentences 159
3.7.1. Completive Aspect ..................................... 159
3.7.1.1. Relative C o m p l e t i v e ...................................159
3.7.1.2. Emphatic Completive . .............................. 162
3.7.2. Incompletive Aspect .................................... 163
3.7.2.1. Relative I n c o m p l e t i v e .................................163
3.7.3. Indicative Mood: Conclusions ........................ 165
3.8. Non-Indicative M o o d ........................................167
3.8.1. Imperatives .................. 167
3.8.2. Exhortative ............................................. 169
3.8.3. Subj.unctive.............................................. 169
3.9. Verbal Conjugations: Conclusions........................ 171
F o o t n o t e s ............................................................... 172

Chapter 4 : Paradigms of Verbal Inflections: Morphology

4.0. I n t r o d u c t i o n ............................................... 174
4.1. The Morphological Composition of Verbal Complexes . . . 176
4.1.1. SV O r d e r ............................................. 178
4.1.2. VS O r d e r ............................................. 181
4.1.3. Morphological Composition of Complexes: Conclusions 184
4.2. VAP-Suffixes in Complex-final Position ................. 185
4.2.1. Anaptyctic i i ........................................... 188
4.2.2. Vowel Harmony ............................ 190
4.3. VAP-Suffixes in non-final Position ...................... 192
4.3.1. VAP-Suffixes for the Negative Completive Active . . . 194
4.3.2 The Clitic/Suffix Distinction .................... 194
4.3.2.1. Pronouns ............................................ 195
4.3.2.2. The A n t e r i o r i t y - M a r k e r ........ .................. 198
4.3.2.3. The Usefulness of the Clitic/Suffix Distinction . . 199
4.3.3. The Suffixal Precedence Hierarchy .................... 201
4.3.4.. Alternation between short-vowelled VAP-Suf fix &Zero. 206
4.3.5. VAP-Suffixes of Incompletive Aspect ............. 209
4.3.6. VAP-Suffixes in non-final Position: Conclusions . . 212
4.4. Verbal Complexes: Some Finer Points .................... 212
4.4.1. SV and VS C o m p l e x e s .................................. 213
4.4.1.1. VS>SV because of the number of Appendages in a
Complex » 213
4.4.1.2. VS>SV with a lsg s u b j e c t ......................... 216
4.4.2. 1E and d b n ........................................... 217
4.4.2.1. M S ........................... ........................217
4.4.2.2. D o n ................................................. 220
4.4.2.3. _MS and cfon: C o n c l u s i o n s ........................ 222
4.5. The Morphological Classification of Conjugations . , . 222
F o o t n o t e s ............................................................... 224
 5

Chapter 5 : Types of Exponence page

5.0. Introduction ........................................225
5.1. Syntactic Exponence ...................................... 226
5.2. Cumulative E x p o n e n c e ..................................... 227
5.2.1. VAP-Suffixes .......................................... 227
5.2.2. P r o n o u n s ................................................ 229
5.3. Sensitive Exponence ...................................... 230
5.3.1. Pronouns ............................................... 230
5.3.2. Anteriority ............................................. 234
5.4. Extended Exponence ...................................... 235
5.4.1. Stative and P r o g r e s s i v e ............................... • 235
5.4.2. P r o n o u n s ................................................ 236
5.4.3. M o r p h - O r d e r .............................................. 237
5.4.4. Consonant Alternation .................................. 238
5.5. H o m o n y m y ................................................... 239
5.5.1. S y n c r e t i s m .............................................. 239
5.5.2. 'Accidental Homonymy' - ( i ) ............................. 242
5.5.3 'Accidental Homonymy' -(ii) ............................244
5.6. C o n c l u s i o n s .................................................247

Chapter 6 : Topic and Focus

6.0. I n t r o d u c t i o n .............................................. 248
6.1. T o p i c ....................................................... 248
6.1.1. Topic and Anaphoric R e f e r e n c e ......................... 251
6.1.1.1. Subject and Topic c o r e f e r e n t i a l ...................... 251
6.1.1.2. Object and Topic c o r e f e r e n t i a l ...................... 252
6.1.1.3. Genitive and Topic c o r e f e r e n t i a l .................... 253
6.1.1.4 . Anaphoric Reference: S u m m a r y ........................255
6.1.2. The Position of a Topic NP in a S e n t e n c e ...............258
6.1.2.1. After-thought Topic .................................. 258
6.1.2.2. Inversion of Topic and Question-word, Relative
Pronoun or Adverb . . 2 5 8
6.1.2.3. Complement Clauses ................................. 259
6.1.2.4. Two T o p i c s ............................................260
6.1.3. Topic-Markers ........................................... 260
6 .1. 3 .1. K a m .....................................................261
6.1.3.2. B o o .....................................................261
6.1.3.3. M a a .....................................................262
6.1.3.4. K a i m n a a ................................................ 262
6 .1.3.5. Topic-Markers: C o n c l u s i o n s .......................... 263
6.2. F o c u s ........................................................263
6.2.1. NP F o c u s ............ 264
6.2.1.1. Focus: Relative C o n j u g a t i o n s ........................ 265
6.2.1.2. F o c u s : no Anaphoric Pronouns ....................... 269
6.2.1.3. F o c u s : Pronouns .................................... 269
6.2.1.4. Focus with Topic .................................... 270
6.2.1.5. Question-words ...................................... 270
6.2.1.6. Focussed Adverbs .................................... 271
6.2.1.7. NP Focus: C o n c l u s i o n s .............................. 272
6.2.2. Verb F o c u s ............................................... 274
6.2.2.1. Completive Aspect . .................................. 275
6.2.2.2. Incompletive Aspect .................................. 277
6.2.3. Focus: Conclusions ................................... 278
F o o t n o t e s ............................................................... 280

C o n c l u s i o n ............................................................. 281
 6

Appendix A : Summary of the Morphology of the 14 Verbal Conjugations

Introduction .......................................... 285
SVO C o m p l e x e s ........................................... 287
VSO C o m p l e x e s ........................................... 293
VOS C o m p l e x e s ........................................... 296
F o o t n o t e s ............................................................... 297

Appendix B : Two Texts

Introduction ............................................ 298
Text 1 .................................................... 299
Text 2 .................................................... 300
F o o t n o t e s ............................................................. . 3 0 2

B i b l i o g r a p h y ........................................................... 303

Erratum

In the numbering of the examples, numbers 69,

137, 138 and 139 have been omitted.
 7

List of Tables

Page (s)

I Vowels 21
II Diphthongs 22
III Consonants 23 -
IV Examples of Commonest Syllable-Types 25
V Combinations of Syllable-Types 26
VI Stress in Nominals 32
VII Alternating Consonants in Verbal Roots 42
and Nominal Stems
VIII Alternants of w and 43
IX Alternating Consonants in Nominal-Class 48
Suffixes
X Representative Class-Suffixes 49
XI Representative Nominal-Class Suffixes 50
with representative Stems
XII A. Consonants commonly found in 53 -
root-final position
B. Consonants rarely found in
root-final position
XIII Phonology of commonest Verbal Roots 54
XIV Nominal Classes 57
XV One-to-one correlation between Singular 59
and Plural Classes
XVI Correlation between Singular and Plural 60
Classes
XVII Range of Classes occurring with nen- 'person' 62
XVIII Sample of Classes occurring with maw- 'big' 62
XIX Pronouns of the Nge class 65
XX Plural Pronouns 145
XXI VAP-Suffixes in Complex-final Position 186
XXII VAP-Suffixes 193
XXIII Clitic and Suffix Pronouns 197
XXIV The Suffixal Precedence Hierarchy - 1 203
XXV The Suffixal Precedence Hierarchy - 2 204
XXVI Variable Suffixes participating in the 204
Suffixal Precedence Hierarchy
XXVII Classification of Conjugations 222
XXVIII Distribution of Pronouns by Conjugation 231 -
XXIX Form of a 3rd person case-function if 256
it is coreferential with the topic
 8

Acknowle dgements

Three people in particular have contributed invaluably to this study.

Professor D.W. Arnott initiated me into the intricacies of Fdlfulde,
and since his retirement has always willingly given his advice.
Mallam Hassan Moturba has been my mentor, teacher and friend since
1976. He has shared with me his love of his mother tongue, and it
is through him and his family that I met the other Fulani people
who inspired this study, and to whom I also owe grateful thanks.
Professor Jack C a m o c h a n has guided me firmly and kindly, and has
patiently read and reread the various drafts of this manuscript. I
am deeply grateful to them all.

I should like to thank my family and many friends for their

encouragement of a project that often seemed incomprehensible and
interminable. For stimulating and helpful linguistic discussions,
I am particularly indebted to the following: Andrew Carstairs,
John Hutchison, Suzy Platiel, John Saeed and Lemuel Stigler. With­
out Andrew Carstairs' insights into inflectional morphology, this
study would have been even more unwieldy than it is.

X should like to acknowledge the financial support, of the Centre

for the Study of Nigerian Languages, Bayero University, Kano, Nigeria,
and of the Department of Education and Science, London, during my
years of field-work and writing up respectively; and of the Wilfred.
Whiteley Memorial Fund, School of Oriental and African Studies, for
paying for a visit by Mallam Hassan to London.

Finally, I am grateful to Michele Durban for her cheerful and

efficient typing of the manuscript.
 9

Introduction

The Fulani people, who call themselves Ful6e, inhabit an area stretching
from Senegambia in the West to Nigeria, Cameroon, and beyond, in the
East. Traditionally, they are a pastoral people, who herd their cattle
in the West African savannah. Even in those areas where they have
established hegemony, such as Northern Cameroon, and Fouta Toro
tin Guin e a } , the Fulani have retained their links with their pastoral
tradition. Most Fulani in fact live as a minority ethnic group amongst
sedentary agriculturalists. Some are 'pure* pastoralists, and
practise transhumance; whereas others have adopted a sedentary, or
semi-sedentary, life-style, and mix farming with pastoralism. The
Kaceccere'en, whose language is the subject of this study, belong
to the latter category of Fulani.

The Kaceccere'en live in the area indicated on the map below.

This area is generally known as 'Southern Zaria*.

Approximate KADUNA
extent of
Kaceccere'en
habitation

Roads

Railways
tCacli’a
Area of
• KANO
Nigeria rep­
resented on
Kafan chan
larger map.

q-QKivs, Ittmad
 10

M o s t of this area has been under the suzerainty of the Emir of Zaria
since the jihad of Usman cfan Fodio in the early nineteenth century -
hence its commonly-used name today. Part of it, however, was in the
emirate of Jema'are. Today Southern Zaria constitutes the extreme
south-eastern part of Kaduna State. The majority of its inhabitants
are christianized farmers of various tribes. These include the Kaje;
and the Katab, KachLchere and Kagoro, who speak mutually-intelligihle
dialects of the same language Ccf. Hansford et al. 1976). Kachichere
is the name of an area, as well as of a dialect and people? and the
1
name Kaceccere'en is probably derived form it.

Unlike most of their neighbours, the Kaceccere*en are Moslems;

they herd cattle; and they live in encampments Csome temporary,
some virtually permanent), rather than in villages. Due to the
latter facts, they call themselves Ful&e na'i ('cattle Fulani'),
or Ful&e ladde ('bush Fulani'), thus distinguishing themselves both
from their non-Fulani neighbours, and from such Fulani as live in
towns, or who no longer herd cattle.

The Kaceccere'en are Fulani of heterogeneous origins, who

define themselves by the area where they now live. They claim
various areas of origin, such as Kano, Borno and Bauchi, and within
the Kaceccere'en umbrella they identify various lineages. For
example, there are the Gayaaji (who came from Gaya, near Kano),
the Bornanko1en (from Borno) , and the Wuntanko'en (from the Bauchi
ar ea). The precise history of the influx of cattle Fulani into the
Southern Zaria area, and the gradual formation of the Kaceccere*en
clan, is not documented. However, the oral traditions of the
Kaceccere'en indicate that they began to arrive during the jihad,
ie. during the first half of the nineteenth century, and that there
2
has been a continuing, gradual, influx since then.

In this study, I shall refer to the Kaceccere'en as a 'clan*

of the Fulani people. I do not use the term 'clan' in a technical
sense; it is simply a convenient way of referring to a group of

people who regard themselves as a unit within the larger designation

of 'Fulani' or Ful&e.
 11

0.1. Language

Hie Fulani in Mali, and in areas east of Mali, call their language
Fulfulde; and this is the term X shall use here. Elsewhere in
Africa, the language is called Pulaar or Pular; and in the literature,
it has been called Peul, Fula and Ful. The Kaceccere1en call their
particular brand of Fulfulde Kaceccereere (with stress on the ree
thus [katjet^e,lre:rej , wher e" precedes the stressed syllable); and
I shall sometimes use this term too. Since the Kaceccere*en, and
other Fulani, regard Kaceccereere as a distinct type of Fulfulde,
it is convenient to label it a 'dialect*.

Fulfulde is a member of the West Atlantic branch of the Niger-

Congo language family Ccf. Greenberg 1970); it is intonational,
and has SVO order in 'neutral* sentences Ccf. Greenberg 19661 . It
is an inflecting language, with a highly developed nominal-class
system, which has received much attention from scholars. On the
whole, case-marking is effected by word— or morph-order, although
certain personal pronouns are morphologically marked for case.

In Northern Nigeria, the lingua franca, Hausa, is gaining

ground at the expense of other languages, such as Fulfulde. This
fact is very evident in Southern Zaria. Here Hausa is the lingua
franca of the market-place, for only a few Fulani speak a smattering
of Katab or Kaje and no non-Fulani speak Fulfulde. Moreover, Hausa
and English are used in the local schools. The result is that
amongst the Kaceccere1en, Fulfulde is the language of the home but
not of the wider world.

The impact of Hausa on the Kaceccere'en is most evident amongst

school children. They themselves admit, and their parents complain,
that their Fulfulde is ‘not d e e p 1. It is quite conceivable that the
children of today's Kaceccere'en school-children will speak Hausa
rather than Fulfulde. However amongst the uneducated Kaceccere'en,
most of whom are of the older generation, the impact of Hausa has
been much less. Although there is considerable Hausa vocabulary
in their Fulfulde, the Fulfulde syntax and morphology is intact.
 12

The Hausa numerals after ten are preferred to the Fulfulde ones
no doubt because the latter, with a combination of base five and
base ten, are so complicated Ccf. A m o t t 1970: ch.27). Examples
of other Hausa words that are used interchangeably with their
Fulfulde counterparts are aiki 'work', makaranta 'school', and
asbiti 'hospital'. The Hausa word may be imperfectly borrowed:
for example, klila is the Kaceccereere form of Hausa watakila
'perhaps'. A Hausa borrowed word is often inflected according to
the rules of Fulfulde morphology: for example, gecfaaji 'ground
nuts' is borrowed from Hausa gyacfa, but has a Fulfulde plural
suffix, -ji; and mi-good-ake 'thank you' is formed from Hausa
goode 'to thank', with the Fulfulde subject pronoun mi_ 'I', and the
Fulfulde middle voice suffix -ake. (A borrowed verbal root always
takes suffixes of the middle rather than the active, voice. See
§3.5.1. below.)

Despite the :inroads that Hausa vocabulary is making into

Kaceccereere, Kaceccereere is in many senses a conservative dialect
of Fulfulde, if one considers its morphological and syntactic structure.
Three characteristics of Fulfulde are inclined to be eroded, or simp­
lified, in what one might call more 'progressive' dialects of the lan­
guage, such as that of Adamawa (in North Eastern Nigeria and Northern
Cameroon) , where Fulfulde is a lingua franca, and much used by non­
native speakers. These three characteristics are: the system of con­
sonant alternation; the occurrence of verb-subject order; and the middle
voice. For evidence of the loss of these characteristics in Adamawa,
see, for example, Skinner and Pelletier (.1978: 175), where no consonant
alternation is shown in verbs; Noye (1974: §84), where the rarity of verb-
subject order is mentioned; and Noye (1 9 7 4 : §63), where the disappear­
ance of the Middle Voice is commented on. In Kaceccereere, however,
all three of these characteristics are very much in evidence, as we
shall see in the chapters that follow.

Kaceccereere also has certain characteristics that distinguish

it from other dialects that have been described in the literature.
Phonological, morphological, and syntactic features will be pointed
out during the course of the following chapters. However, I '.shall
 13

mention here certain items of vocabulary that are not recorded else­
where (though they may be noted in the de St. Croix dictionary
3
manuscript ). It is not yet clear whether any of these items are
borrowed from the other languages that the Kaceccere1en come into
contact with. They include sinndeyi 'knife*,
gee&ol 'stream*, wonngo ’perhaps* and inke *it seems that*.

0.2. Informants

This study is based on the Fulfulde of several informants of the

Kaceccere*en clan. M ost are still resident in the Kafanchan-Kagoro-
Zonkwa area, and those that are not regularly visit their family
there.

My chief informants were Hassan Moturba and his wife, Fatu.

Hassan is of the Gayanko'en lineage of Kaceccere*en,born and brought,
up near Zonkwa. He is bilingual in Fulfulde and Hausa, and also
speaks some English. As a result of attending adult education *

classes, he is able to read and write in Fulfulde, Hausa and English.

In adult life, Hassan has lived in Zaria and Ngaoundere (Northern
Cameroon); and at present, he lives and works in Kano. His wife,
Fatu, has a similar pattern of residence, though she was born in
the Kagoro area, and is of the Wuntanko'en lineage. Fatu is bilingual
in Fulfulde and Hausa, but she is not literate, and she speaks no
E nglish.

My other informants are still resident in Southern Zaria, except

for Isa Pashi, who sadly passed away shortly after he had so greatly
helped me. Isa lived in Kafanchan, and was of the Yillaa&e lineage.
He was the most highly educated of my informants, having a B.A.
degree in History from Ahmadu Bello University, Zaria. He was
equally at ease in Fulfulde, Hausa, and English, and would himself
have made an excellent research linguist, for he h a d many insights
into his mother t o ngue, Fulfulde. He was well versed in local lore
and traditions; b u t he also knew much about the documented history
of the area. His widow, Altine, is similarly of the Yillaa&e
lineage. She speaks no English, and only 'market Hausa*.
 14

Dawobe Abayin is of tbe Bornanko'en lineage, and lives near

Zonkwa. He is bilingual in Hausa and Fulfulde, speaks some English.,
and reads and writes Hausa. Sa*adatu, his youngest wife, is a
girl in her teens, who is bilingual in Fulfulde and Hausa, but
is not literate, and speaks no English.

Hajia Maryam is a woman of the Yaakanaaji lineage, bilingual

in Fulfulde and Hausa, but not literate. She and h e r family live
in Samaru Katab. Her two sons, Haruna and Adamu, are approximately
25 and 18 respectively. Haruna is just literate in Hausa, and
is bilingual in Fulfulde and Hausa. Adamu has been attending the
local Government Secondary School. He speaks and understands a
lot of English; he is most at ease speaking Hausa; and although
Fulfulde is his mother tongue, he did not feel capable of answering
all my questions, complaining that his Fulfulde was no.t *deep enough1.

Other Kaceccere*en Fulani also helped greatly with this study, either
by letting me record their conversations or their folk-tales, or simply
by being enthusiastic to teach me their language. Many of the data
presented here have been extracted from such conversations; other
data have been provided by the informants mentioned above.

0.3. Aims of this Study

The aims of this study are principally syntactic and morphological.

Chapter 1 provides some phonological and morphological background inform-
ation. Chapter 2 then considers Fulfulde sentences, verbal complexes
and noun-phrases. The following three chapters are concerned with
verbal morphology: first, the syntactic and semantic justification for
the establishment of verbal paradigms (chapter 3) , then the morphology
of those paradigms (chapter 4) , and finally the morphological realisation
of morph-syntactic properties (chapter 5). In chapter 6, an outline
is presented of the way in which Fulfulde focusses an NP and a verb;
and NP focus is contrasted with the topic construction. The conclusion
points out some areas of potentially fruitful research, that could be
carried out on Fulfulde in the future.
 15

0.4. Conventions and Abbreviations

Before proceeding any further, I shall outline here some of the

conventions and abbreviations that will be used in this study.

0.4.1. Presentation of Examples

The transcription used for Fulfulde words will be explained in §1.1.5.

below. The only point that need be mentioned at this stage is that
hyphens are used to provide a morphological analysis of verbal forms,
for example:

l.a. Nge-wub - ake.

it mis carry
VAP
'It (cow) miscarried.'

Here, a hyphen separates the verbal root, wub- from the subject
pronoun, n g e - , and from the Voice-Aspect-Polarity CVAP} —suffix, — ake.

Under each morph in a verbal complex or participle,, a gloss

is given, as in l.a. above. (The significance of the VAP-suffix

will be discussed in §2. 2. 1. below.) Sometimes, I shall also

include other information below the Fulfulde sentence. For example, ~
the accompanying discussion might require the naming of the 'tense'
of a v e r b , as has been done in 1 .b..

1 ,b. Nge-wub - ake.

it miscarry
VAP
GEN COMPL MID
'It (cow) miscarried.'

The abbreviations used in such labels will be explained in §0.4.5.

below: and the significance of the ‘tenses' will become apparent
in chapter 3.

0.4.2. Translation of Examples

Each Fulfulde example will be accompanied by a free translation, as

 16

in l.a. and l.b. above. However, translation is no simple matter.

Problems arise particularly with the translation of the pronouns
and of 1tenses1, and I shall therefore explain here the policy
adopted with regard to these matters.

Let us consider pronouns first. There is no gender in Fulfulde;

however, there are 19 singular nominal-classes, and 6 plural ones.
(The nominal class system is explained in §1.3. b e l o w ) . One sing­
ular class is reserved for persons, as is one plural class, so
pronouns referring to nominals of these classes may be translated by
'he/him' or 'she/her', and b y 'they/them1 respectively. However,
pronouns referring to nominals of the other 18 singular and 5 plural
classes cannot satisfactorily be translated b y 'it* or 'they*. For
4
example, in the Nge class, only the nominals referring to 'sun1,
'fire' and-'cow' are found. Since neither the 'sun' nor a *fire*
is likely to 'miscarry' , there is little doubt b u t that the pronoun.
nge in 1. refers to a 'cow*; for a different referent, such as
a sheep or a goat, would require a pronoun of a different class.
T h u s , a pronoun referring to a nominal class (and third-person pronouns
necessarily refer to a particular class), bears the semantic con­
notations of that class. In the translations that I give of Fulfulde
sentences, I shall attempt to mention, as succinctly as possible,
what those semantic connotations are.

The translation of Fulfulde so-called 'tenses' is even more

problematic. (In Chapter 3, I argue that 'tense* is in fact an
inappropriate term for the verbal inflectional paradigms in Fulfulde,
and I argue that the term 'conjugation* is a suitable alternative.
In anticipation of that section, I shall henceforth use 'conjugation'
to refer to what Arnott and others have called 'tenses': cf. Arnott
19 70: 179.) Now, the translation of a Fulfulde conjugation will
depend on context; for example, the General Completive will be
translated by the English preterite in one context, the present perfect
in another, the future perfect in another, and the simple present in

yet another; and vice-versa, the English preterite may be translated

 17

b y a Fulfulde General Completive, Relative Completive, Emphatic

Completive, General Incompletive, etc.. The point is, there is
no one-to-one correlation between the uses of the various verbal
paradigms in the two languages. Given this fact, the choice of
tense for an English translation is often rather arbitrary. For
example, the Fulfulde sentence in 1, without any context, could
just as well be translated by 'It has miscarried1 as by 'It mis­
carried' . On the whole, if the choice is not determined by con­
text, I choose the most 'neutral' or 'obvious* English tense to
translate the Fulfulde verb. However, the above points concern­
ing the lack of equivalence between the vexbal systems of the
two languages m ust be borne in mind when considering the trans­
lations given throughout this study.

0.4.3. Unacceptability

In discussing certain morphological and syntactic structures,

I sometimes give examples which are preceded by an asterisk (*) .
This indicates that the sentence concerned is unacceptable, usually
for morphological or syntactic reasons. Generally, the relevant,
unacceptable, item is underlined, and is explained in the accompany­

ing discussion.

0.4.4. Underlining,

In order to draw attention to a point under discussion, the relevant

part(s) of Fulfulde sentences may be underlined.

0.4.5. Abbreviations

The following abbreviations and symbols are used in this study.

Each is followed by a reference, indicating where that particular
item is first discussed or identified in the text.
 18

Abbreviation/Symbol Meaning Reference

ACT Active (Voice) §3.5.1.

CL Nominal-class-suffix §1.3.1.
COMPL Completive (Aspect) §3.4.2.
CONT IMPER Continuative Imperative §3.4.7.
DO Direct Object §2.2.2.1.
EMPH Emphatic Completive §3.4.5.
EXHORT Exhortative §3.4.7.
EXT Extension §2.2.2.2.
GEN COMPL General Completive §3.6.1.1.
GEN INCOMPL General Incompletive §3.6.2.1.
H Hausa §0.1.
ID Ideophone §1.1.4.4.
IMPER Imperative §3.4.7.
INCOMPL Incompletive (Aspect). §3.4.2.
INF Infinitive - suf fix -

10 Indirect Object §3.5.3.3.

MID Middle (Voice) §3.5.1.
NEG Negative §3.4.3.
NEG COMPL Negative Completive §3.6.1.2.
NEG INCOMPL Negative Incompletive §3.6.2.2.
NON REL Non-re1ative §3.4.4.
0 Object §2.2.2.1.
PASS Passive (Voice) §3.5.1.
PL Plural §3.5.3.2.
Pi Plural §3.5.3.2.
POS Positive §3.4.3.
PROG Progressive §3.6.2.1.
QU Marker of yes./no question §1.1.4.3.
REL Relative §3.4.4.
R EL COM^L Relative Completive §3.7.1.1.
REL INCOMPL Relative Incompletive §3.7.2.1.
S Subject §2.2.2.1.
SG Singular §3.5.3.2.
sg Singular §3.5.3.2.
STAT Stative §3.6.1.1.
SUBJ Subjunctive §3.8.3.
TM Topic-Marker §6.1.3.
V Vowel (phonolog,, context) §1-1.1.
V Verb father contexts) §2.2.2.1.
VA Voice-Aspect §1.3.2.
VAP Voice-Aspect-Polarity §2.2.1.
VAGUE INCOMPL Vague Incompletive §3.6.2.3.
0 Zero §2.1.1.1.

1 First Person §3-5.3.1.

lexcl First Person exclusive §3.5.3.1.
lincl First Person inclusive §3.5.3.1.
2 Second Person §3.5.3.1.
3 Third Person §3.5.3.1.

morph-boundary §0.4.1.
< derived/calque from §0.1.
* unacceptable §0.4.3.
syllable-boundary §1.1.3.
 19

Introduction: Footnotes

1. It is not certain whether the second ch of Kachichere represents a

geminate affricate {[ t t ^ )/ or not. The orthography I use for this
name follows that of Temple (1922) and Gunn (1956}. However, there
is no doubt but that the Kaceccere*en Fulani refer to themselves as
[katjettjere?8:n]. In other words, the second affricate in their
name is geminate, but the first is not.

2. For historical information given here (as well as for much linguistic
h e l p ) , I am grateful to the late Isa Pashi of Kafanchan.

3. The manuscript of a Fulfulde-English dictionary, covering 'Central

Nigerian' Fulfulde, written by F.W. de St. Croix, is at the moment
being edited at the Centre for the Study of Nigerian Languages,
Bayero University, Kano.

4. I capitalise the first letter of the names of nominal classes,

in order to distinguish them from class-pronouns: see §1.3,.
 20

Chapter 1

Linguistic Background Informatian:

Phonology, Consonant Alternation, and the Nominal

Class System

1.0. Introduction

By comparison with most other African languages, Fulfulde is extremely

well documented. The first grammars of the language were published over
lOO years ago (Faidherbe 1875 and Reichardt 1876). Hiey have been followed
by numerous others, in various languages: one in Fulfulde, b y the
Waziri of Sokoto, Alhaji Waziri Junaidu, and the others in. French, German,
and English.

Despite the proliferation of such works, Fulfulde studies are not as

advanced as Hausa studies, for example. There are at least two reasons, for
this fact. Firstly, the diffusion of the Fulfulde language has given rise
to various studies, each, on the whole, describing a different dialect
from the others. For this reason, most studies have 'started from scratch',
rather than building on a body of inherited knowledge- Secondly, since
Fulfulde is not of paramount political or educational significance in any
one m o d e m political state, there has been no need felt Cas there has
in the case of Hausa) for an in-depth study of the grammar of the language.

The present study deals with a dialect that has been hitherto undescribed.
However, it draws heavily on the work of Arnott in particular, for the Gombe
dialect that he describes bears comparison with that of the Kaceccere'en ^
and it draws more generally on the work of Noye, Gaden, and Klingenheben.
This study is primarily descriptive in nature, but it does assume the
existence of a deep-structure, and reference will sometimes be made to
the deep structure, at the phonological, morphological and syntactic levels.

In this chapter, I provide some background information concerning the

Fulfulde of the Kaceccere'e n , Kaceccereere. While the treatment of these
matters by Arnott and others, is very adequate, it is felt that their
inclusion here will considerably facilitate the reader's understanding of
 21

subsequent parts of this studyj and it will also highlight the ways
in which Kaceccereere. differs, however slightly, from other dialects.
Section §1.1. deals with phonology. Section §1.2. is concerned with
consonant alternation, and includes a discussion of such alternation in
suffix-initial and root-final position, as well as in stem- and
root-initial position. Finally, section §1.3. outlines the nominal
class system.

1.1. Phonology

In this section, certain phonological features of Kaceccereere Fulfulde

are explained, in order to justify the transcription used. In §1.1.1.

and §1.1.2., the vowels and consonants of Kaceccereere are listed and
discussed. Syllable-structure and prosodic features are outlined in
§1.1.3. and §1.1.4.; and finally, in §1.1.5., the transcription is
ejqplained.

1.1.1. Vowels and Diphthongs

There are five short vowels and five long vowels in Fulfulde. Vowel-
length is particularly important for the distinction between different
verbal roots, and the following minimal pairs, using the infinitive
form of the verb., ^ have been chosen for that reason:

Table 1

Vowels

Orthographic Phonetic Minimal Pairs

Form Value

a H hacfugo 'refuse (someone something)1

aa Ca:l haacfugo 1taste b i t t e r , sharp1
e [si fewugo 'tell a lie'
1®J
[— 1

ee feewugo 'be cold'

i [0 birugo 'milk'
ii 00 biirugo 'wipe clean (eg. a. calabash) 1
o DO horugo ’remain, be left'
oo 00 hoorugo 'take h o m e '
u H furugo 's n o r t '
uu 00 fuurugo ’be startled'
 22

Hie left hand column in table 1 shows the orthographic form of

the vowels, while the middle column shows their phonetic values.
Note that there is a difference of vowe1-quality that correlates
with the difference of vowel-length: for example, the quality
of aa is not £a:], but [q:J.

Eight phonetic diphthongs are found in Kaceccereere Fulfulde,

as follows:

Table II

Diphthongs

Orthographic Phonetic Examples

Form Value

ay [at] saydltugo 'to fade *

ey [elJ leydi 'country, ground'
oy koycfum 'easy? recovery'
uy [vtj wuy&e 'thieves'
aw M fawrude 'woman-s calabash'
ew rw i pewri 1cold(ness)'
iw diwnugo 'to tremble'
ow M cfowdi 'shade, shadow'

All diphthongs except uy occur in word-final, as well as in medial,

position. Before a vowel, the second element of a diphthong 'becomes*
its glide equivalent, [j] or [w].

Phonologically, diphthongs are best treated as a combination

of vowel plus glide, and this fact is reflected in the orthography
shown in table II, which is also the orthography adppted in this
s tudy.
 23

1.1.2. Consonants

In consonants as well as vowels, length is phonemic, and table III below

therefore shows minimal pairs, wherever possible, in order to illustrate
this fact. There are, however, no minimal pairs to prove certain contrasts.
Despite this fact, I shall always show consonantal length in the transcription
used in this study, for it is relevant to the system of consonant alter­
nation, as we shall see presently (in §1.2. below).

Table III has been arranged in such a. way as to display groups of

consonants according to their phonological and morph-phonological behaviour.
Those in group (i) do not alternate with any others; but those in group (ii)
do. This fact, and the choice of examples in table III (some with the
relevant consonant in initial position, some with the relevant consonant
in medial position), will be elucidated in section §1.2..

Table III

Consonants

Orthographic Phonetic Examples

Form Value

(i). Non-alternating cf M fucfugo 1 (sun) to rise'

Consonants dcC [dag fucftfugo 'to begin'
6 [6] ha6ugo ‘to fight’
66 [66] ha66ugo 'to tie'
1 t e 1ugo
[?1 'to marry*
*y M ta'yugo 'to cut *
” y tyy] ma*'yugo 'to blinkv
t [t] fotugo 'to m e e t '
tt [tt] fottugo 'to suit'
l [i] Mi-walaa-ngel 11 ha v e n 't got i t '
n [ii] Mi-wallaa-ngel 'I didn't help it'
n [n] wonugo 1to b e , live'
nn Cnrg wonnugo 'to spoil'
m Qm] tamugo 'to roll (dough) in
mm [ram] tammugo 'to think* fa b a l l 1
ny QfQ wanyugo 'to refuse, reject’
nny i m sannyugo 'to weave'
i] m fcanugo 'to migrate; marry'

Alternating r [r] seerugo ’to divorce'

Consonants: w CWJ tawugo ’to find1
Y [Y] maayugo 'to die'
a. Continuants: *—i
h LSI mahugo 'to build’
f [£J fofugo 1to happen'
s [s] wasugo ' to dig'
 24

Table III (cont)

Consonants

Orthographic Phonetic Examples

Form Value

(ii.)' Alternating d KL seedugo 'to separate'

Consonants dd Cdd] teddugo 'to be heavy'
g Eg] tagugo 1to create'
b. Plosives and gg Taggugo 'to roll, curl up'
Affricates: b [b] hubugo 'to throw'
bb Ebb] hubbugo 'to dress (hair)'
k M takaago 'to cross (a bridge)'
kk [kk] takkaago 'to become stuck'
P*' 0?3 dr-pusi 'they burst'
PP Ep p J hippugo 'to spill, overturn1
j Ldj] wujugo 'to rub, smear'
jj Edch] wuj jugo 'to steal'
cc feccugo 'to split open'

c. Prenasalised nd M ndiyam 'water'

consonants: ng [ng] ngaari 'b u l l '
mb [mb'] mbeewa 'g o a t '
nj [nj] njumri 'h o n e y '

The system of consonant alternation justifies the distinction between many

of the above consonants, and also explains the arrangement of the consonants
in the table. For the moment, however, let us turn to other, phonological
matters.

1.1.3. Syllable Structure

I shall divide the discussion of syllable structure into two parts.

First, I shall consider what we can term 'regular' syllable structure;
and then, I shall consider 'exceptional' syllable structure.

1.1.3.1. 'Regular' Syllable Structure

In Kaceccereere Fulfulde, every syllable must begin with a consonant; and

a long vowel occurring in a closed syllable is a rare, exceptional phenomenon
Thus, the following syllable types are most commonly found:

(a) CV
(b) C W
(c) CVC

Examples of these three types are found in table IV. A plus sign, '+ ', is
/

used to mark syllable-boundaries.

 25

Table IV

Examples of commonest syllable-types

(a) CV mi ‘I 1; 'he/she; ja+6i 'accepted'.

(b) C W moo 'him/her' (in certain 'tenses')?
'aa 'you (sg)' (in certain 'tenses');
njaa+fee 'forgive (pi)!'
(c) CVC mbal 'beer'; 'on 'you (pi)';
wal+ l a ml 'help (sg) me I'

Several points are worth noting about the exemplification in table IV.

(i) Syllable structure does not necessarily tally with morphological

structure. The disyllabic examples given above are presented in terms of
their syllable-structure/ not their morphological structure. Ja+6i and
njaa+fee in fact consist of a verbal root (ja&- 'accept' and njaaf-
'■forgive'), plus a voice-aspect-polarity (VAP)-suffix (-i and — eei see
§2.2.1. below), and wal+lam consists of a verbal root (wall- 'help')
plus object pronoun (-am 'me'). Thus, morphologically these disyllabic
examples would be presented as ja£-i , njaaf-ee and wall-am.

(ii) Geminate consonants, such as -11-, count as two different con­

sonants, in two different syllables. This fact justifies their identi­
fication in the phonology, even if minimal pairs are lacking (as is
true for d_ and dd) .

(iii) Prenasalised consonants count as single consonants, unless they

are geminate. Thus, nj_ of njaa+fee is a single consonant, as is mb of
mbal. The same is true of other consonants that are conventionally
written as digraphs: namely, ny ( |ji~]) , and J_y ([yj) .

(iv) A glottal stop may occur in word-initial position: cf. _[_o and
1aa in table IV. This fact will be reflected in the transcritpion
used here, though in many transcriptions, glottal stop is only shown
intervocalically (cf. UNESCO 1971).

(v) The distinction between short and long vowels is neutralised in

word-final unstressed syllables (cf. §1.1.5.1. below). Thus, although
I have quoted m o o , 1aa and njaa+fee in isolation for the purposes' of
illustration, it must be understood that these forms only have a (final)
CW syllable when they are followed by another morph in the same word.
 26

Moo and 1aa in fact only occur in a particular position in the verbal
complexes of certain 'tenses', and are always followed by another morph.
Njaa+fee would be phonetically [njccfe], with C W + CV syllable-
structure, if uttered in pre-pausal position, but [nja:fe:], with
CW + CW syllable structure, if followed, for example, by -5e 'them'.

The syllable types in table IV can combine with each other.

Here are some examples:

Table V

Combinations of the Commonest Syllable Types

(a) CV + CW wa+laa 'There is/are not.'

CV + CVC cCa+tal 'p a t h , w a y '
(b) C W + CV nyaa+mu 'Eat (sg)!'
C W + CVC baa+tal 1needle'
(c) CVC + CV 'ig+ga 'It would be better than..
CVC + CW mob+tee 'Gather together (pi)!'

I shall not deal here with the question of which vowels or consonants
may occur in which position in a syllable, although these matters will
be touched-on in the discussion of consonant alternation in §1.2. below.

Normally, both C W C and CVCC syllables are unacceptable in the

Fulfulde of the Kaceccere'en. The unacceptability of a C W C syllable
is evident from derivational and inflectional processes which involve
3
a consonant-initial suffix. For example, when nyaam- 'eat' is followed
by the -r-/-ir- extension (cf. §2.2.2.2. below), or by the -ndu class-
suffix (cf. §1.3.1. below), then its vowel is shortened to a, as in
2.b. and c . :

2.a . 'On-nyaam-ay? 'On+nyaa+may

you eat CVC C W CVC
VAP

'Will you (pi) eat?'

b. Raa no nyam-r - at-on! nyam+ra+ton

see how eat you C VC CV CVC
EXT VAP
'See how you (pi) e a t ! '

Nyam-ndu wal-aa. nyam+ndu

eat is not C V C CV
CL NEG
'There's no food.1
 27

The syllable structure of the words involving nyaam-/nyam- is shown

on the right hand side. If the long-vowelled form, nyaam-, had been
used in 2.b. and c., then unacceptable C W C syllables would have resulted:

2 .d. *Raa no nyaam-r - at-on. *nyaam+ra+ton

see how eat you C W C CV CVC
EXT VAP

*Nyaam-ndu wal-aa. *hyaam+ndu

eat is CWC CV
CL NEG

Similarly, if the causative extension, -n-/-in- is added to soof- 'be/get

wet* the vowel of soof- is shortened, as in 3.b.;

3.a. Sarla maako soof-ay. soo+fay

trousers his wet C W CVC
VAP
'His trousers get wet.'

b. 'O-sof-n - ay sarla maako. 'o+sof+nay

he wet trousers his CV CVC CVC
EXT VAP
'He wets his trousers.'

If the long vowel of soof- had been retained in 3 . b . f then an un­

acceptable C W C syllable would have resulted.

3.c. * 10-soof-n - ay sarla maako. 'o+*soof+nay

he wet trousers his CV C W C CVC
EXT VAP

The unacceptability of a syllable with CVCC structure can be shown

by the fact that a suffix that may be vowel- or consonant-initial is
always vowel-initial with a verbal root or nominal stem that has CVCC
structure, (Whether the vowel, concerned should be regarded as an anaptyctic
vowel is a moot point.) For example, the -r-/-ir- and -n-/-in- extensions
seen above both have their latter forms (-ir- and -in-) with timm- 'end 1
and janng- 'read', as can be seen in the b. and c. sentences of 4. and 5.:

4.a. Kuugal timm-ay jooni. timfmay

work end now CVC CVC
VAP
'Work will end n o w , '

4.b. Min-timm-in - ay kuugal 'amin. min+tim+mi+nay

we end work our CVC CVC CV CVC
EXT VAP
'We'll finish our w o r k . 1
 28
4.c. No kuugal ngal timm-ir - ta? tim+mir+ta
how work this end CVC CVC CV
EXT VAP
'How will this work end ? 1

5.a. Bi 6 6 e njanng-ay. nj an+ngay

children read C W CVC
VAP
1 (The) children read . 1

b. Mi-janng-in - ay 6 i 6 6 e. mi+j an+ngi+nay

I read children CV CVC C V CVC
EXT VAP
'I teach (the) children.'

c. Jawdum njanng-ir - $ - mi -cfi. n j an+ngir+mi+cfi

quickly read I them CVC C V C CV CV
EXT VAP
II quickly read them . 1

Forms such as the following are unacceptable, because they involve

syllables with CVCC structure, as shown on the right hand side:

4.d. *Min-timm-n-ay kuugal. mint* timm+nay

we end work CVC CVCC CVC

e. *No kuugal ngal timm-r-ta? *timm+*rta

how work this end CVCC CCV

5.d. *Mi-janng-n-ay 6 i 6 6 e. mi+* j anng+nay

I read children CV C V C C CVC

e. *Jawdum n j anng-r-mi-cfi. *n j anng+*rmi+cfi

quickly read I them CVC C CCV CV

4.e. and 5.e. contain two unacceptable syllables each - in fact, it is

very difficult to assign any syllable structure to them at all.

1.1.3.2. 'Irregular 1 syllable structure

The three syllable-types described above are certainly the commonest in

Kaceccereere Fulfulde, and the d. and e. sentences of 4. and 5. point to
the fact that these CVC, C W and CV syllable-structures need to be
maintained. Nevertheless, certain uninflected words do have the exceptional
CVCC or C W C structure. Those that come to mind are:

CVCC: koyg (diminutive plural class-suffix, etc. - see below)

minn 'me' (independent pronoun)

CWC; 1aan 'you (sg) ' (Independent pronoun).

Let us first consider and minn, and then 1aan.

 29

Koyq is a class-suffix for the diminutive plural nominal class.

It also 'represents' that class in various other functions: as a dependent
pronoun, as deictic pronoun and adjective ('this'), and as an anaphoric
determiner. (Concerning the nominal class system, see §1.3. below.)
It has the following cognates in various other dialects:

kon: Gombe ( A m o t t 1970)

kon: Northern Cameroon (Noye 1974)
kony: Sokoto (Westermann 1909)
koy: Nigerian Adamawa (Skinner and Pelletier 1979)

Each of these involves a CVC syllable. One might try to 'reanalyse' the
Kaceccereere koyi), in order to make it conform to the general rules of
syllable structure. Perhaps the diphthong, o y ((jjlJ ) could be analysed
as W rather than VC: but this would result in an equally unusual C W C
syllable I Alternatively, one could argue that the final nasal is not
segmentally realised, but that nasality is a supra-segmental feature of
the whole form: thus, koy ([kSi]). Although the Q>rJ diphthong is
indeed nasalised, there is no escaping the fact that this form does also
entail a final (j^] ; and I believe it must therefore simply be regarded
as an exception to the 'no-CVCC rule.

The Isg independent pronoun is represented in other dialects as min

or m i i n . In Kaceccereere Fulfulde, the vowel is , not [i:J ; and
the alveolar nasal is always lengthened - unlike the final nasal of the
lpl (excl) pronoun, min. Despite the fact that nn occurs nowhere else in
syllable-final position, and the fact that this gives us a CVCC consonant,
we have no option but to regard this pronoun as another exception to the
'no-CVCC' rule.

The 2sg independent pronoun in Kaceccereere Fulfulde is identical to

that found in Gombe, though in Adamawa, according to Noye (1974) and
Skinner and Pelletier (1979), it is 'an rather than 'a a n . There is little
doubt but that it has C W C structure, and it must therefore be regarded
as an exception to the 'no-CWC' rule.
 30

1.1.3.3. Syllable Structure; Conclusions

Despite the exceptions described in §1.1.3.2./ it is useful to consider

CVC, C W and CV syllables as the norm for Kaceccereere Fulfulde. In
particular, as we shall see in §1 .1 .4.2 . below, this analysis of
syllable structure provides an extremely neat way of accounting for
word-stress in nominals.

1.1.4. Prosodic Features

Fulfulde is an intonational language, and tone is n<bt phonemic. I

shall not be showing prosodic features in Fulfulde examples, except in

this section, and in §2.2.3. below.

I shall discuss four features of prosody here: 'final glottality'

(in §1 .1 .4.1.); word-stress (in §1 .1 .4.2 .); intonation (in §1 .1 .4.3.);
and hypertonic words {.in ..§1.1. 4. 4.).

1.1.4.1. Final Glottality

Certain types of words are characterised by a glottal closure in pre-

pausal position. This closure is not phonemic; it does not occur in
non-pausal position; and it does not consistently occur, even in the
speech of a single speaker. It is therefore best regarded as a feature
4
of prosody. It occurs after nominals and after certain types of verbal
complex. In the following examples, it is marked b y a single inverted
comma (as indeed is the glottal stop in initial and medial position):

6. Kanjum won-i hoore haala^_.

that is head speech
VAP
'That is the introduction to (this) speech .1

7. 1E-mi-wood-i wiige &aleere^_. ^

I have heiffer black
VAP
'I have a black heifer . 1

8. '0 -yar - ay ndiyamj_.

he drink water
VAP
'He drinks w a ter . 1

9. Ngam 'o-hokk-a- min habaru no Ful 6 e ngar-d- oy- i 1.

so that he give us news how Fulani come
VAP EXT EXT VAP
'So that he may give us news of how the Fulani came
(from a distance)- *
 31

10. a. 1E-rai-wuund-ii- 6e_[_.

I clasp them
VAP
'I'm clasping/embracing them.'

Examples 6 . - 8 . have a nominal in pre-pausal position/ and thus have

final glottality; and examples 9. and 1 0 .a. have the sort of verbal
complex in pre-pausal position that may have final glottality. The
nominals h o o r e , wiige, habaru and Ful&e in 6., 7. and 9. do not have
final glottality, because they are followed by other words in the sentence.
Similarly, the verbal complex in 7.? (1E-mi-wood-i) , although it is similar
to that in 10.a. ('E-mi-wuund-ii-&e'), does not have final glottality,
because it is not in pre-pausal position.

Final glottality does not occur with every verbal complex. Its
occurrence is limited to complexes of conjugations of types 2 and 3 in
table XXVII r chapter 4. In other words, it occurs with complexes

of the Stative and Progressive, the Relative Completive and Incompletive,

and the Subjunctive. Moreover, it does not occur in such complexes if
a 2sg pronoun occurs in final position. For example, if the complex
in 10.a., which is Stative, has maa 'you (sg) ' instead o f 6 e as its
object, then final glottality does not occur:

lO.b. 'E-mi-wuund-i - maa.

I clasp you
VAP
'I'm clasping/embracing you (sg).'

(The alternation between the long-vowelled - ii VAP-suffix and the short-

vowelled -i, in lO.a. and b. respectively, is conditioned by the following
pronoun: see §4.3.3..)

1.1.4.2. Word-Stress

The identification of the stressed syllable in a word is not a simple

matter. Since vowel-length is phonemic, I shall take pitch as the chief
criterion for identifying stress. However, there are in fact correlations
between vowel-length and stress, especially in nominals.

The occurrence of stress in verbal complexes is discussed in §2.2.3.

below. In nominals, it generally occurs on the last non-final CVC or C W
syllable. In the absence of such a syllable, the initial syllable of the
nominal is stressed, regardless of its structure. In the following examples,
 32

the stressed syllable is marked by preceeding inverted commas (")

Table VI

Stress in Nominals

(i) Kacecce" reejo Ka+cec+ce+ree+jo 'Fulani of the Kaceccere'en clan 1

CV CVC CV C W CV

jal a- j a l 11noowo j a+la+j al+noo+wo 'person who makes others laugh'

CV CV CVC C W CV

nawlil,raawo naw+li+raa+wo 1co-wife'

CVC CV C W CV

m a 6 6 i"tirgel mab+ 6 i+tir+ge 1 1k e y 1

CVC CV CVC CVC

hon"ndoore hon+ndoo+re 'basket'

CVC C W CV

joo"cTorki joo-kfor+ki chair 1

C W CVC CV

"koyngal koy+ngal leg1

CVC C V C

"nyiiri nyii+ri 'food *

C W CV

(ii) Ka"ceccere1en Ka+cec+ce+ret'en 'Fulanis of the Kaceccere'en clan 1

CV CVC C V CV CVC

Wun"tanko1en Wun+tan+ko+.'en 'Fulanis of the Wuntako'en lineage'

CVC CVC CV CVC

"njawdiri nj aw+di+ri ram1

C V C CV CV

"hoccere hoc+ce+re 'mountain'

CVC CV CV

"fajiri fa+ji+ri 'dawn'

CV CV CV

"pamaro pa+ma+ro 'small (person)'

CV CV CV

"demal de+mal 1farm'

CV CVC

" 1esam ’e+sam 'my in-law'

CV CVC
 33

In each of the examples in (i) above, the penultimate syllable has C W or

CVC structure, and is thus stressed. This is not true in the examples in
(ii), where the stress has therefore been 'pushed b a c k 1 to an earlier CVC
or C W syllable if possible (as in the first four examples) , or to the
first syllable (as in the last four examples). Nominals with more than
three syllables, but without a C W or CVC syllable in penultimate position
are very rare - most are proper names, as are Wuntanko1en and Kaceccere'en.

Names, and words of foreign origin, in fact ppovide regular exceptions

to the rules of stress in nominals. For example, the Arabic numerals,

borrowed via Hausa, generally have stress on their final syllable: tala"tin
'thirty', ham"sin 'fifty'. There are also other words that are possibly
or definitely borrowed, and which similarly have stress on the final
syllable. They include kabi"dow 'raincoat' (from Hausa), pogo"low
'bag', and gee" 6 ol 'stream'.

Word-stress will not be shown regularly in the examples given in

this study. However, from time to time, when focus is under discussion
(e.g. in i‘2.3.2. and chapter 6 ) a whole word will be underlined in a
Fulfulde sentence. This will indicate that the word concerned is the focus
of the sentence, and that its stressed syllable bears the intonation-peak
of the sentence.

1.1.4.3. Intonation

The intonation-contour of a sentence is determined by many factors.

Apart from such non-linguistic factors as the attitude of the speaker,
these include: whether the sentence is a statement of a question; whether
it has a topic or focus (cf. Chapter 6 below); and whether it is verbal
or non-verbal. For present purposes, I shall assume that 'neutral' state­
ments have gradually falling intonation, and I shall outline some of the
ways in which 'neutral' questions, and non-neutral sentences, deviate
from the norm.

Yes/no questions are often distinguishable from statements only

by their intonation. However, they may also have a 'question-marker',
n a , which I gloss in the examples as 'QU'. Yes/no questions have rising
intonation, whereas statements have falling intonation, as can be seen
from the graphic representation of the pitch of the following sentences.
 34

The two horizontal lines represent the normal high and low limits of
the voice.^

11. 'A-famt - ii.

you understand
VAP
'A+fam+ti(i).
'You (sg) have understood.'

12. a, 'A-famt - ii?

you understand
VAP
'A+fam+ti(i)?
'Have you understood?'

1 2 .b. 'A-famt - ii na?

you understand
VAP QU
'A+fam+ti (i) na?
'Have you understood?'

1 2 .a. and b. are alternative yes/no-question equivalents of the statement

in 11.. The rising intonation that is characteristic of yes/no questions
occurs on the final syllable of the utterance. Thus, in 12.a., it is
on t i i , but in 1 2 .b. it is on n a .

A question that begins with a question-word (equivalent to the

wh- questions in English), has falling intonation, like a statement.
For example:

13.a Ngeye rim - i?

which bear
VAP
Nge+ye ri+mi?
'Which (cow) has given birth?'

14.a Ndee ndi hirs - aa?

when it slaughter
VAP
NdeeS ndi+hir+sa (a)?
'When was it slaughtered?'

The question-word may or may not be contrastively stressed. If it is,

then the intonation pattern of the sentence will be as indicated in
13.b. and 14.b. below. If not, it will be as above.
 35

A sentence with a focussed constituent often has contrastive

9
stress (i.e. extra high pitch) on that constituent.

13.b. Ngeye rim-i?

which bear
VAP-------------------- *-- *— *■
Nge+ye ri+mi?
'Which- (.cow) has given birth?'

14.b. Ndee ndi-hirs - aa?

when it slaughter
VAP
NdeeS ndi+hir+sa(a)?
'When was it slaughtered?'

15. Ngetton rim-i. 5-------

that bear •
VAP *-*-
Nge+ton ri+mi
'That one over there has given b i r t h . '

16. Fajiri ndi-hirs - aa.

dawn/morning it slaughter
VAP
Fa+ji+ri ndi+hir+sa(a).
'It (.eg. bull) was slaughtered at dawn I'

As can be seen from the accompanying pitch-representations, contrastive

stress involves exceptionally high pitch on the relevant syllable of

the focussed item? thus, the first syllable of ngeye in 13.b. has
higher pitch than its counterpart in 13.a., for example. If the verb
of a sentence is focussed, its root may receive contrastive stress.

17. Ndi-hirs-a, naa ndi-waat-u.

It slau- not it die • 0 0
ghte r *--------------- *—
Ndi+hir+sa naa ndi+waa+tu.
VAP VAP

'It (eg. bull) was slaughtered, it didn't (simply) die.'

NP-, adverb-, and verb-focus will be more fully discussed in §6.2.

below.

If a sehtence begins with a topic (which. A m o t t calls a 'prelude':

see his 1970: §7.17., and §6.1. below), there is a potential for pause
between the topic and the rest of the sentence. The rest of the
sentence has intonation as if the topic were not there. This pause
is represented by a comma in the Fulfulde text:
 36

18. Kanko, 'o-famt - ii.

him he understand
TOPIC VAP
Kan+ko 'o+fam+ti(i)
1 (As for), him, he's understood.'

19. 'Onon, 'on-pamt - ii?-------- ~----

you you understand •
topic vap — r — t—
'On+on 'on +pam+ti(i)
'(As for) you Cpl) , have you
understood?'

Intonation will not be marked in the Fulfulde sentences of this

study, unless it is particularly relevant to the discussion in hand.

1.1.4.4, Hypertonic Words

Certain Fulfulde words are always hypertonic: in other words, they

always bear pitch that is higher than any other syllable in the sentence.
They thus 'break' the general rules of intonation contours that have
just been given. These hypertonic words include ideophones; and three
quantifiers: fu and pat (both meaning 'all' or 'each'), and tan
(meaning ‘o n l y ').

The hypertonic words in sentences 20. - 25. are underlined, and

their high pitch is indicated in the accompanying pitch-representations.
Cup in 20. emphasises the earliness of fajiri; similarly, tar emphasises
the whiteness of daneyel in 21.. In 22., det does not have an emphatic
function; rather, it adds a meaning to its own to the verb wad-'do*.
(Christiane Seydou, in personal communication and in Seydou (^J.) (1979)
calls words such as det 'ideophones', and those such as cup and tar
'intensifiers'.) Fh and pat in 23. and 24. are interchangeable, both
meaning 'all', 'each' or 'every*. These, and tan in 25., most commonly
occur in prepausal position, and especially at the end of a topic (cf.
chapter 6 ).

20. Fajiri cup 'umm-ii-mi -----------^--- > ----------

dawn rise I • •
• * *
ID VAP :--------- ---------- «-
fa+iji+ri cup um+mii+mi
’At the crack of dawn I got u p . '

21. Ngel daneyel tar)

it white
ID
ngel da+ne+yel tar
'It's gleaming white.'
 37

22. Dere wacf-i d e t .

Dere do silent
VAP ID
De+re wa+cfi det
'Dere fell silent.'

23. Laahooji pat, cfi - laah-ay.

dogs all they pant
TOPIC VAP
laa+hoo+ji pat cfi+laa+hay
'All dogs p a n t . '

24. Nagge fu, sey mi-tuf - i-nge.

cow each, then I vacc— it
inate
nag+ge fu sey mi+tu+fi+nge
TOPIC VAP

"Then I vaccinated every cow.'

25. Kanko t a n , 'o-juul-ay.

him only he pray
TOPIC VAP
kan+ko tan 'o+juu+lay
'Only he is a Moslem.'

1.1.4.5. Prosodic Features; Conclusions

Prosodic features have been mentioned here in order to give some idea
of what spoken Kaceccereere Fulfulde sounds like. Let us n o w return to
segmental phonology, and the orthographic conventions adopted in this study.

1.1.5. Transcription

There is no standard orthography for Fulfulde, although the re commendations

of the 1966 Bamako Conference have been implemented in some countries,

such as Niger. In Nigeria, when Fulfulde is written in Roman script

many Hausa orthographic conventions are adopted: for example, long and
short vowels are generally not distinguished. However, since this study
involves considerable discussion of morpho-phonology in particular, a
fairly precise, narrow, transcription is used. In this section, I
shall explain some of the orthographic conventions I use in order to
represent Kaceccereere Fulfulde as it is spoken - for it is rarely
written down.

1.1.5.1. Vowels in Word-final Position

In this dialect, there is no distinction between long and short vowels

in word-final unstressed syllables. For instance, in a sentence 'such
as:
 38

26.a. Bello yar - i i .

Bello drink
VAP

'Bello drank.'

the -ii is not noticeably longer than the -i of:

27.a. Sey Bello yar - i^.

then Bello drink
VAP
'Then Bello drank.'

However, the -ii vs. -i distinction is maintained in the transcription

used here, because there is an audible difference in the length of these
two vowels if they are followed by another morph in the same word. Thus

26.b. Bello yar - ii-ndi.

Bello drink it
VAP

'Bello drank it (eg. njvunri 'honey'

has a noticeably long-vowelled VAP-suffix, whereas:

27-b. Sey Bello yar - i_ - ndi

then Bello drink it
VAP
'Then Bello drank it {.eg. honey).'

has a noticeably short-vowelled VAP-suffix.

The choice of -ii or -i as the VAP-suffix in 26. and 27. is deter­

mined by the inflectional paradigm to which the verbal complex (yar-ii
or yar-i) belongs. The General Completive Active has the long-vowelled
-i i , whereas the Relative Completive Active has the short-vowelled -i
suffix. (The former is used in 'neutral sentences', the latter is
necessitated by the occurrence of sey 'then'.)

Vowel-length in word-final position is transcribed here according

to its potential, rather than its actual, length. This principle will
be applied to the transcription of aa, ee, oo and u u , as well as of ii.

1.1.5.2. 'Weak' Consonants

Certain consonants in root-final position are sometimes phonetically

realised as a vowel: I therefore term them 'weak consonants'. The
consonants concerned are h, (glottal stop), w and y.
 39

When the root-final -h_ in yah- 'go 1 is followed by a zero morph,

or b y a consonant-initial suffix, it is assimilated to the preceding &
v o wel? thus yah- is realised as £ya :3 • This fact is recorded in the
transcription. Thus, we may contrast 28.a. below with 28.b . , c. and d . :

28.a. M i n - n j a h . - a a to.
we go not yet
VAP
'We (excl)' haven't gone yet.'

b. Yaa-0 ton' . c. Yaa.-go d. Yaa^-du.

go there go go
VAP INF CL

'Go over thereJ' 'To go.' 'Journey.'

Similarly, when the root-final glottal stop in wi'- 'say' is followed

by a zero morph, by a consonant-initial suffix, or by ani-initial suffix
(cf. 33. on page 41 below) , then it is assimilated to the preceding i_, and
wi'- is realised as [wi:J. Again, the a. form below may be contrasted
with the b., c. and d. forms:

29. a. No 6 e - rabi_[_-etee?
how they say
VAP
'What are they called?'

b. Wii-^-mo'. c . Wii-go. d. Bii-^cfo

say him say say
VAP INF VA CL
'Tell h i m l ' 'To say . 1 'Person who said.'

In 28.a. and 29.a., the transcription . reflects the phonological and

phonetic reality. In the b., c. and d . , forms, however, the transcription /
is phonetically, but notphonologically, true.^°

The sequences -uw and -iy in root-final position are realised as [uf]
and (i:"] respectively when they are followed by zero morph or by a
consonant-initial suffix. Thus, in the a. sentences below, the root-final
consonant is realised as such? but in the b. and c. forms, it is realised
vocalically:
 40

30. a . ’Q-hUW-ay - n o .
he work
VAP ANT

'He used to wo r k . '

b. *Gn-kUU-d - ii. c. Kuu-gal

you work work
EXT VAP CL

'You (pi) worked to­ 'Work (noun) . '

gether. '

31.a. Mi-fiy-et-e.
I hit you
VAP
'I r11 hit you Csg). '

Mi-fii-r - ii-mo - ndu. c. Fii-go.

X hit with him it hit
EXT VAP INF

'I hit him with it. * •To h i t . '

In the transcription used here, the phonemes /w/ and /y/ are
written as y and i_ if they are preceded by an u or i and followed by
a consonant: c f . -kuu-d— in 30.b. and fii-r- in 31.b.. Otherwise,
they are written as w and y:c f . -huw-ay- in 30.a. and fiy-et- in
31.a., and the examples of diphthongs in table II on page 22 above.

1.1.5.3. Transcription: General

The transcription used here shows a certain amount of phonetic

information. For example, cficfafc- 'second' becomes cficTaw- before
certain consonants, and this fact is represented orthographically.
Thus, the transcription can be effectively used as a guide to
pronunciation, bearing in mind the rules of stress and intonation
mentioned in §1.1.4.2-3., and in §2.2.3..
 41

Much of this study, however, is concerned with morphology

and syntax - and in particular, the morphology and syntax of verbal
complexes. For this reason, a detailed morphological analysis is
presented for each verbal complex, with each morph identified in the
gloss below each example: cf. 30. and 31.a. and b. above. There
are instances where in quick speech, the form of the complex would
be considerably reduced, for example:

32. a. '0 -sood-u-nde.

he buy it
VAP
’He bought it.'

would become:

32.b. '0 -son-nde.

'He bought i t.'

In 32.b., only the pronoun subject, 'o- has retained its integrity,
and sonnde is morphologically difficult to analyse, without ref­
erence to rules of consonant assimilation and vowel reduction.
For this reason, when such a choice exists, the fuller, slow-speech
version, is always given. Sometimes, the fuller version in fact
no longer occurs at all, and in these cases, the morphologically-
analysable deep-structure form will be given, together with the
surface-structure form. Thus:

33. 10 n-ngii- 6 e .
(< 1O n - n g i '-ii- 6 e).
you see them
VAP
'You saw them . 1

With the root y i 1- 'see1, as with w i 1- 'say/call/tell', the final glott­

al stop (_!_) of the root is dropped when an i-initial suffix follows ;
and the i_ vowel of the root coalesces with the i_ or ii_ vowel of the suf­
fix. The form 1O n - n g i 1-ii be is not normal in Kaceccereere (though it is
 42

recorded by Arnott for the Gombe dialect), but since it facilitates a

morphological analysis of 1On-ngii- 6 e , it is represented in brackets with
the latter form.

The only nominals for which I sometimes give a morphological analysis

are participles, since they have many verbal characteristics. Other nominals
will only be morphologically analysed during the course of the discussion
of consonant alternation and the nominal class system, which are the subject
of the next two sections, §1.2. and .§1.3..

1.2. Consonant Alternation

If the reader refers back to the table of consonants {table III), he

will notice that the alternating consonants have been arranged into
three groups: continuants, plosives, and prenasalised consonants.
The justification for this arrangement will now become evident.

In verbal roots and nominal stems, there are in fact eight sets of
alternating consonants, as shown in the vertical columns of table VII.
Each member of each set can conveniently be classified as belonging to
a 'continuant', 'plosive' or 'prenasalised' series, as shown in the
horizontal lines in table VII.

Table VII

Alternating Consonants in Verbal Roots and Nominal Stems

Continuant series: w w y y r h f s
Plosive series: b g g j d k P c
Prenasalised series: mb ng ng nj nd k P c

In considering table VII, the following three points must be borne in mind: (

(i) The labels 'continuants', 'plosives' and 'prenasalised' are

morpho-phonologically useful, butthey do not always correspond
to phonetic reality: j_ and £ are affricates, not plosives,
and k, p and £ are not prenasalised.
 43

(ii) k, p. and do 'double duty' as plosive and prenasalised

alternants of h, f and s_ respectively.

(iii) w and y_ alternate with two different sets of plosive and

prenasalised consonants, the choice of set being to a large
extent phonologically determined, as shown in table VIII:

Table VIII

Alternants of w and y

y^ alternates with cj and ng, if followed by e^, ee_, i_ or i i ;

y alternates with j_ and nj_, if followed by aL, aa, o_,o o , u or u u .
w alternates with ,b and mb if followed by e_, ee, i_ or i i ;
i b and m b , or .
w alternates with if followed by a, aa, o, oo, u or uu.-
— - g ng — — — — — —

There is no way of predicting which, alternation takes place when w is

followed by a_, aa, o^, o o , ii or u u : one simply has to know the character
of the particular word concerned.

Consonant alternation most obviously takes place at the beginning

of verbal roots and nominal stems; but since these facts have been
very adequately described in the literature (eg. Arnott 1970 and Klingen-
heben 1963), they will only be briefly outlined here (sections §1.2.1 and
§1.2.2.)., However, alternation also takes place at the beginning of
nominal-class suffixes, and at the end of verbal roots; and since these
facts have been barely mentioned e l s e w h e r e , ^ they will be treated more
fully here, in sections §1.2.3. and §1.2.4..

1.2.1. Consonant Alternation in Verbal Complexes

The verbal complex will be defined and discussed in §2.2. below. Essen­
tially, it consists of a verbal root and a cluster of appendages. The
latter include at least a suffix denoting Voice, Aspect and Polarity
(which I call the 'VAP-suffix'), and often a pronominal subject and pro­
nominal object(s) too. Although participles also contain a verbal root,
they necessarily inflect for nominal-class, and they are syntactically NP's.
The distinction between participles and verbal complexes is further discussed
in § 2 .1 .2 .1 . below.
 44

If a verbal root begins with an alternating consonant, the continuant

alternant is used when the subject both precedes the root, and is singular.
For example.

34.a. Ndee 'o-war-i?

when he come
S V VAP

1When did he come ? 1

Here, the verbal root, w a r - , begins with the continuant w, since the
subject ('o-) both precedes the root, and is singular. However, the
prenasalised alternant, n g , is used if the subject is plural (as is 6 e-
in 34.b.), and/or if the subject follows the root (as does -cfaa in 34.c.

and - ’on in 34.d . ):

34.b. Ndee Be - ngar-i?

when they come
S V VAP
1When did they come ? 1

c. Ndee ngar-0 - cfaa? d. Ndee ngar-0 - 'on?

when come you when come you
V VAP S V VAP S
'When did you (sg) come?' 'When did you (pi) come?'

The prenasalised alternant, ng-, is 'triggered' by the plural subject,

Be- in 34.b . ; and b y the following subject, -cfaa in 34.c. and - 'on in
34.d . . Note that the prenasalised alternant is used in 34.c. despite the
fact that the subject is singular; thus, the position of the subject
overrides its number (ie. whether it is singular or plural) in determining
the choice of alternant.

Here are some further examples of the operation of consonant

alternation in verbal complexes. In each a. sentence, the subject is
singular, and a continuant alternant is therefore used; and in each
b. sentence, the subject is plural, and a prenasalised alternant is used.
(Note that whether the subject is pronominal or not is immaterial.)
Each c. and d. sentence has the subject following the verbal root:
the c. subject is singular, the d. one plural.
 45

35.a. Debbo 'am wuuw-ii-ndu. b. Rew&e 'am mbuuw-ii-ndu.

wife my sweep it wives my sweep it
S V VAP 0 V VAP 0

'My wife swept it (eg. 'My wives swept it (eg.

hut/room).' hut/room ) . 1

c. Ngim (fume mbuuw-u-mi-ndu? Ngim cfume mbuuw-u-'en-ndu?

for what sweep I it for what sweep we it
V 'VAB.'S 0 V VAP S 0
1Why did I sweep it?' 'Why did we (incl) sweep
it?
36.a. 'A - yar - ay kosam? b 'On-nj ar - ay kosam?
you drink milk you drink milk
S V VAP O S V VAP 0
'Will you (sg) drink some milk?' 'Will you (pi) drink some
milk?
Kosam njar - at-aa? Kosam njar - at-on?
milk drink you milk drink you
0 V VAP S O V VAP S
'Will you (sg) drink milk 'Will you (pi) drink milk
(not anything else ) ? 1 (not anything else)?'

Since we are only concerned with consonant-alternation at this stage,

the reader is referred to §4.1. below for an explanation of the factors
that determine whether a subject precedes or follows the verbal root.

By way of contrast with the examples given so far in this section,

here are some examples of a root whose initial consonant does not alternate.
Sentences 37.a. - d. are comparable to the a. - d. sentences of 34. - 36;
however, the verbal root in 37. consistently begins with cf, despite
variations in subject-position and -number:

37. a. Mi-cfoft - ay- 6 e . b. 'En-dbft - ay- 6 e.

I accompany them we accompany them
S V VAP 0 V VAP
'I'll accompany them,' 'We (incl) will accompany
them?

Har tokoye cfoft - ay-mi- 6 e? d. Har tokoye doft ay-'en- 6 e?

to where accom- I them to where accom- we them
pany pany
V VAP S 0 V VAP S 0
'How far shall I accompany them?' 'How far shall we (incl)
accompany them ? 1
 46

In 37., the verbal root begins with cf-, whether the subject is singular
(as in 37.a . ), or plural (as in 37.b.J; and whether the subject precedes
the root (as in 37.a. and b.), or follows it (as in 37.c. and d.).

Note that when a verbal root occurs in a verbal complex, only the
continuant and prenasalised alternants occur (cf. 34. - 36.). However,
when a verbal root is part of a participle, then a plosive alternant may
also be found. This will be discussed in §1.2.2..

1.2.2. Consonant Alternation in Nominals

For the time being, I shall subsume nouns, participles, and adjectives
under the term 'nominal', since they all necessarily display class-concord
in their class suffixes. This matter will be further discussed in §1.3.2.
below.

The Fulfulde nominal-class system is reminiscent of that found in

Bantu languages. Each of the 25 classes is 'associated with' only one
of the consonant series shown in table ^xi; and most classes also have
semantic connotations (cf. Arnott 1970: appendix 4). For example, if
a nominal stem begins with an alternating consonant, and occurs in the
Ngel class, its initial consonant must be a plosive, not a continuant
or prenasalised consonant; and it will indicate diminutive singular
meahing. However, if the same stem occurs in the Koyg class, its initial
consonant must be a prenasalised one, not a continuant or a plosive;:
and it will indicate diminutive plural meaning. These facts are ill­
ustrated in 38. and 39. In 38.a., the Ngel class determines the occurrence of
the stem-initial plosives, g_ and d^, and in 38.b. , the Koyrj class determines
the occurrence of the initial prenasalised consonants, ng- and n d - .
(The class of these nominals is indicated by the - e l , - y e l , -oyrj, and
-hoyij suffixes, which will be discussed in §1.2.3. below). The nominals
in 39. are unacceptable, because an inappropriate stem-initial consonant
has been used.
 47

38. a. Gapal-el dane - -yel. b. Ngapal-oyi) ndane-hoyq,

gown white gown white
CL CL CL CL
'Small white gown . 1 'Small white gowns.'

39.a. *Wapal-el rane - yel b. *Gapal-oyi} dane-hoy ij

gown white gown white
CL CL CL CL

*Ngapal-el ndane-yel d. *Wapal-oyr} rane-hoyi)

gown white gown white
CL CL CL CL

By way of contrast, consider the nominals in 40.a. and b . .

Although they are of the Ngel and Koyq classes (as are those of 38.a.
and b.), the initial consonants of the tuute- and Bale- stems are
constant. This is because jt and 6 ^ are not alternating consonants.

40. a. Tuute -yel Bale-yel. b. Tuute -hoy i} Bale-hoyi}.

flag black flag black
CL CL CL
'Small black flag.' 'Small black flags.'

Once the nominal-class system has been more full explained, then
a list of classes and their corresponding 1consonant-*series' will be
given tin table XIV below.)

1.2.3. Consonant Alternation in Class-suffixes

Each nominal class is characterised by up to 4 class-suffixes. For

example, the Ngel class has suffixes -el and -yel (seen in 38.a. above),
and also -gel and -ngel? the Koyg class has suffixes -oyq and -hoyg
(.seen in 38.b. above) , and also -koyrj. Further examplesjcf class-suffixes
are -ngal and -wal in 41.a. (both of the Ngal class) ,and -cfe and
- je in 41.b. (both of the De_ class).

41.a. D e m - ngal feere - w a l . b. Dem - feeree - j e .

tongue/ different tongue/ different
language language
CL CL CL CL
'A different language.' 'Different languages.'

(The length of the final vowel in the feere/feeree- stem is conditioned by

the initial consonant of the following suffix.)
 48

A particular nominal stem can only combine with one suffix of

any class. For example, it would be impossible to swop the class-
suffixes of the two stems in 41.a. or 41.b . , thus:

42.a. *Oem - wal feere - n g a l . b. *Dem - je feeree - cfe.

tongue/ different tongue/ different
language language
CL CL CL CL

42.a. is unacceptable, because cfem- cannot combine with -w a l , nor

feere- with -ngal. Similarly, 42.b. is unacceptable, because cfem-
cannot combine with - j e , nor feeree- with cfe.

Two important questions arise concerning the class-suffixes. How

can w e best describe the various suffixes for each nominal class? And
is there any way of predicting which type of suffix will combine with
a given nominal stem? I shall deal with the first of these questions
in some detail? and the second I shall deal with more cursorily.

1.2.3.1. Suffix-Grades

Arnott (1970: ch.17) establishes four 'grades' of class-suffixes, 'A1,

'B', 'C' and 'D'. I should like to suggest that his 'suffix-grades' are
in fact the manifestation of the consonant-alternation system operating
in class-suffixes.

The system of consonant alternation is not quite the same in

suffixes as it is in nominal stems and verbal roots. One difference is
the fact that there is what we can call a 'zero' series in the system, as
well as the continuant, plosive and prenasalised ones we have already seen
The 'zero' series of suffixes mostly have zero initial consonant, although
there is one exception (with, r-), which can be seen in table IX below.

Table IX

Alternating Consonants in Nominal Class Suffixes

Grade/Series: Zero Continuant Plosive Prenasalised

0 w g ng
0 y g ng
r r d nd
0 h k k
0 j cT cf
 49

Another difference is the fact that certain sets of consonants do not occur
in suffix-initial position, namely w/b/mb, y/j/nj, f/p/jP and s/c/c

(compare table VII, where these do occur). Furthermore, the j/d/cf

alternation found in suffixes does not occur in verbal roots and nominal
stems; and within this set, j_ is one of the continuant, not the plosive,
series (of table V I I ) .

Tables X and XI below illustrate the system of consonant 'alternation in

Fulfulde nominal-class suffixes. Table X gives the four suffixes for
certain representative nominal classes: the relevant consonants are underlined.

Table X

Representative Nominal Class Suffixes

Grade/Series Zero Continuant Plosive Prenasalised

Class

■Ngu u — wu - gu -ngu
Ngel - el - vel - gel -ngel
Nde - re - re - de -nde
Koyn - _?yr} ” 3a°yi} - koyg -koyij
De e - ge - cfe -cfe

The horizontal lines in table X show the four class-suffixes that characterise
the nominal class shown on the extreme left. Each vertical column shows
suffixes of a single grade (zero, continuant, etc.), but of different
classes. The underlined consonants in table X pattern with the consonants
shown in table XI.

Table XI shows the same suffixes as table X, but this time in morph­
ological context ie. with nominal stems. For each of the classes mentioned
in table X, table XI shows one example of each suffix-grade. (I shall
henceforth refer to 1suffix-grades', but 'continuant (etc.) series'
of consonants.) The factors that determine which grade of suffix occurs
with which stem will be discussed presently (in .§1 .2 .3 .2 . below).
 50

Table XI

Nominal S tems, with Nominal Class Suffixes

Grader Zero Continuant Plosive Prenasalised

Class
Ngu tatab- u bale-wu ngor-gu taarii-ngu
Ngel tatab- el bale-yel gor-gel taarii-ngel
Nde tatab-re balee-re wor-de taarii-nde
Koyg tatab- oyg bale-hoyg ngor-koyq taarii-koyg
De tatab-_e balee-je gor-de taarii-cfe

Meaning: 1third' 'black' 'male 1 'entwined'

(Note: The length of the final vowel of the bale-/ balee- stem is conditioned
b y the initial consonant of the following suffix.)

The horizontal lines in table XI show different stems in combination

with the four grades of class-suffix. Any one stem can only combine with
one grade of suffix. T a t a b - , for example, combines with zero grade,
bale- with continuant grade, and so on. The vertical columns, on the
other hand, show a single stem in combination with suffixes of the same
grade but different classes. For example, taarii- is shown with suffixes
of the N g u , N g e l , N d e , Koyq and *De_ classes. The meaning of stems plus
class-suffixes will be discussed in §1.3.1.2. below: That shown at the
foot of each column is in fact the meaning of the stem alone.

Table XI does not only show consonant alternation in suffixes: it

also shows such alternations in a stem, wor- 'male'. Non-alternating
stems, such as tatab-, bale- and taarii- are only subject to two variables:
class-suffix and suffix-grade. Alternating stems, such as w o r - , however,
are subject to a third variable: the consonant series 'associated with'
each nominal class. From table XI, we see that the Ngu class is associated
with the prenasalised consonant series (hence the nc[ of ngor-) , the Ngel class
with the plosive series (hence the g_ or gor-) , the Nde class with the
continuant series (hence the w or wor-) , and so on. The operation of
this third variable, in addition to the other two, often effectively
disguises the identity of a stem: the uninitiated might net guess, for
example, that ngor-gu and worde both mean 'male' - though, as we shall
see in §1.3.1.2. below, they in fact mean more than simply 'male*.
 51

1.2.3.2- Which grade with a given stem?

Which grade of suffix will combine with a given stem? There do not
appear to be any absolute rules concerning this. However, I shall
make some general observations here, which are similar to, though
not the same as, those made in Arnott (1970: ch. 17, and appendix 5).

Class-suffixes of the zero grade often occur with stems that

end in a plosive consonant, whether that consonant is single, geminate,
12
or part of a cluster. For example:

43.a. Goot-a. b. Ndu 6 6 - o. c. Kolt - e.

one fan-palm clothe
CL CL CL
'One (eg. donkey}.' 'Large fan-palm ® . 1 'Clothes.'

In all of the above, the class-suffix begins with 'zero-consonant'.

The final consonant of the nominal stem in 43.a. is -t, in 43.b. -6 6 ,
and in 43.c., the plosive jt is part of a consonant cluster, -It.

Ciass-suffixes of the continuant grade only occur with stems that

end in a vowel, for example:

44.a. Meere_-wol. b. Sirati-yel. c. Beydaa - ri.

useless Kanuri increase
CL CL CL
'Useless (eg. pen).' 'Child/small person 'Increase (e.g.
of the Kanuri tribe.' in salary ) . 1

Note that a stem-final vowel is always long before a suffix that begins with
j_- or _r- (cf. 41.b. , and 6 alee-re/ 6 alee-je in table XI) ; but that otherwise,
a stem-final vowel is short (cf. 41.a., and 6 ale-wu, 6 ale-yel and 6 ale~hoyg
in table XI ) . This fact explains the variation in the length of stem-
final vowels in 44. (e- and i- in 44.a. and b. , but aa- in 44.c.).

Class-suffixes of the 'plosive grade' always occur after a stem

that ends with the r^ of the instrumental or locative verbal extensions.

(The extensions are explained in §2.2.2.2.. See also Arnott 1970: chs.
57 - 59). For example:

45. a. Daar-or - gal. b. Joocf-oir- ki. c -Juul-ir - de-.

look sit pray
EXT CL EXT CL EXT CL

'Mirror.' 'Chair.' 'Mosque.'

 52

(Note that inasmuch as only certain, morphologically-defined, types of

-_r- final stems regularly combine with the plosive-grade suffixes, we
are dealing here with morpho-phonology, and not only phonology.)

Class-suffixes of the prenasalised grade often combine with stems

ending with a continuant (,cf. the examples given in 46.); and they
always combine with participal stems (cf. 47.):

46. a. 'Bow - .cfi. Yim - 6 e. Bun - ndu.

mosquito person well
CL CL CL

'Mosquitoes.' 'People.' 'Well.•

47..a. Gicf - aa-db. b. War - ay-nde. Sennd - u - 6 e.

love come separate
VA CL VA CL VA CL
'Beloved (person ) 1 'Coming/next ' (People) who have
(eg. week).' separated.'

The participial stems in 47'. in fact consist of a verbal root (gicf-, war-
13
and sennd-) , followed by a 'Voice-Aspect-Suffix' (VA-rsuffix) , and then
by a class-suffix.

1.2.3.3. Suffix-grades: Conclusions

By way of conclusion to this discussion of consonant alternation in

class-suffixes, I should point out that there are many features of
the class-suffix system that have yet to be explored. Most important is
the question of what really is the final consonant of a nominal stem:
in many instances, an underlying 'archiphoneme' has to be postulated,
whose surface-structure form is determined by the class-suffix. For
example, the stem meaning 'child' is 6 i C - , where 'C' represents a consonant
that is the same as the initial consonant of the class suffix. Thus:
Sicf-cfo 'child', 6 i£-£e 'children', Bin-nge 1 'little child', and
fcik-koyq 'little children'. Such assimilation we can term a type of
'forwards sensitivity', for the realisation of the one consonant is
'sensitive' to what follows. Yet the observations made above
concerning the correlation between nominal stem and suffix-grade involve
the opposite, i e . 'backwards sensitivity'. It is clear, then, that what
happens at the juncture of stem and suffix needs considerably more
investigation.
 53

1.2.4. Consonant Alternation in Root-final Position

C o n s o nant-altemation in root-final position is less obvious than

elsewhere. Moreover, it is not an active synchronic process.
Rather, it is found by analysis of the phonological structure of
verbal roots.

Most Fulfulde verbal roots have a basic CVC, C W C or CVCC

structure. Roots that are synchronically derived by the addition
of one or more verbal extensions (.cf. §2 .2 .2 .2.1 may be polysyllabic,
as indeed may roots of foreign origin. The observations made here
pertain only to the 'basic', monosyllabic, roots.

Taking 'basic' roots, we find that there is a certain pattern­

ing of consonants in root-final position, as shown in table XXI:

Table XII

A. Consonants most commonly found in

root-final p o s i t i o n ^

Type Examples

(i) Any non-alternating fib- 't i e ', ma&b- 'close 1

consonant (single or warn- 'dance', lamm- 'be bitter'
geminate) won_- 'live' hinn- 'greet'

(ii) Alternating Consonants:

a. Non-geminate cont- mar- 'own', day- 'be distant'

tinuants-^ heew- 'be full', laah— 'pant'
woof- 'err', wes- 'winnow'

b. Geminate Plosives wadd- 'bring', tigg- 'establish'

and Affricates cfabb- 'foment' , tikk- 'be angry'
sipp- 'hawk(wares)', majj- 'be lost'

c. Geminate Prenasalised n annd- 'resemble', jenng- 'be late'

Consonants 'oramb— 'cover*
 54

Table XII (cont.)

B. Consonants rarely found in root-

final p o s i t i o n ^

a. Non-geminate Plos- Some exceptions are found in

ives and Affricates: table III,* others are:
bad.- 'approach' , jog- 'hold*
sood- 'sell1/ wooj_- 'be red'

b. Non-geminate Pre­ Exceptions a r e :

nasalised consonants:
wuund- 'embrace, clutch'
foond- 'measure'.

What table XII shows us is that the following most commonly occur in
root-final position:

(i) non-alternating consonants:

15
(ii) a. non-geminate continuants;
b. geminate plosives and affricates?
c. geminate prenasalised consonants.

Of the alternating consonants, then, continuants on the one hand alter­

nate with geminate plosives, affricates, and prenasalised consonants on
the other. In other words, the phonology of verbal roots is such that
a root-final plosive, affricate or prenasalised consonant is likely to
be geminate (cf. A (ii) b. and c. in table X I I ) ; but a root-final
continuant is not. In view of the C V C , C W C or CVCC 'basic 1 structure
of verbal roots, the following are, therefore, the most commonly-occurr­
ing types:

Table XIII

Phonology of commonest Verbal Roots

(i) CV + Non-alternating Consonant

CW + Non-alternating Consonant

(ii) a. CV Continuant
CW + Continuant
b. CV + Geminate Plosive or Affricate
c. CV + Geminate Prenasalised Consonant

A long vowel is not possible with types (ii) b. or c . , because this

would create an unacceptable C W C C root. In this respect, Kaceccereere
Fulfulde differs from the Fulfulde described by Noye and Arnott, for
example. Both these writers quote roots with such a C W C C structure, for
 55

example, suumt- 'break one's fast* (Noye 1974: 342), and moobt- 'assemble'
( A m o t t 1970: 344) . The Kaceccereere equivalents of these words have
a short vowel, thus CVCC structure.

Although in root-final position there is no morph-phonologically-

conditioned alternation of consonants on a grand scale, there is never­
theless some slender evidence of such alternation in certain derivational
17
processes. The example that most easily springs to mind is the alter­
nation between heew- and hebb- as the.root for 'be full'. The former
is the 'basic' root. However, when the causative extension, -in,
is added, the 'basic' root becomes h e b b - , hnd the derived root is
thus hebb-in- 'fill'. Here, then, we see alternation between w and b b ,
continuant and geminate plosive. (The variation in vowel-length
(ee vs. ej could be explained, in Kaceccereere Fulfulde, in terms
of the unacceptability of C W C C roots. However, hebb-in- (not *heebb-in~)
is recorded in other dialects too (cf. Taylor 1932: 76, Gaden 1914: 95).
It is in fact more usefully explicable in terms of a fairly general
correlation between C W C and CVCC, especially with the causative extension.

For example wojj-in- 'cause to become red' is derived from wooj- 'be red';
cTucfcf-in- 'increase, make much/many' is derived from duucf- 'be much/many'.)

1.2.5. Consonant Alternation: Conclusions

When the list of consonants was presented in table III, pages >23 - 24>

it was pointed out that minimal pairs cannot be found for absolutely
every consonant that is distinguished in the transcription used here.
Some of these distinctions can only be justified on phonetic grounds: r-
vs. nq, and vs. J_y. Others can be justified on phonetic grounds,
and because they are useful for the analysis of syllable structure:
ny vs. rmy, _|_y vs. 1 'y, d vs. dd. Others, however, can be justified
in terms of morphophonology. Thus, while g_ cannot be established by
means of a minimal pair, it is useful to distinguish it because it
systematically alternates with f.

There are few minimal pairs to prove the phonemic status of the
prenasalised consonants. This is partly because they so rarely occur
in root-final position. One example is the following pair:

wuud- 'be intractable' vs. wuund- 'embrace'

 56

This minimal pair satisfactorily establishes the need to distinguish

d from n d . However, no equivalent contrasts exist for ng, mb or n j .
Nevertheless, the morpho-phonological status of these consonants in stem-
and root-initial position is indubitable.

There are no minimal pairs to establish the phonemic status of

the geminate prenasalised consonants either. Moreover, these do not
alternate in stem-r root-or suffix-initial position. It was pointed out
in §1.2.4. that they do alternate, in a non-active, ’fossilised1,
way, in root-final position. However, since the latter alternation is not
a significant morpho-phonological process in the language today, I have
not listed the geminate prenasalised consonants in table III.

Despite their lack of status, either phonemic or morpho-phonological,

I do distinguish geminate from non-geminate prenasalised consonants
in the transcription. This is because they are an undeniable phonetic
reality; and because the CVC vs. CVCC distinction is necessary for
the purposes of discussing syllable-structure.

1.3. The Nominal Class System

Fulfulde is a language with a highly developed inflectional morphology.

Two important types of 'content words' can be identified, according to
their inflectional paradigms and their syntactic behaviour. On the one
hand, there are what I call nominal stems; and on the other hand,
verbal roots. Cl shall consistently use 'stem' for nominals, and 'root'
for v e r b s , though this choice is quite arbitrary.) Nominal stems
necessarily inflect for nominal class: in other words, they must be
accompanied by a nominal-class suffix. Verbal roots necessarily inflect
for Voice, aspect and polarity: these are expressed in the 'Voice/
Aspect/Polarity-suffix' (VAP-suffix). They may also inflect for other cat­
egories too: these matters are discussed in chapters 3 - 4 . Participles
are hybrid forms: they consist of a verbal root and a *Voice/Aspect-suffix',
but they also take a nominal-class suffix, and they behave, syntact­
ically, like nominals. Since the major part of this study is devoted
to the morphology and syntax of verbs, I shall spend a little time
here on nominal morphology.
 57

1.3.1. The Classes

Each of the 25 classes in Fulfulde is 'associated with' the continuant,

plosive, or prenasalised series of consonants (cf. table VII, page 42),
and each is also characterised by a set of up to four class-suffixes
(cf. table XI, page 50). A nominal stem occurring with a particular
class must begin with a consonant of the series appropriate to that class;
and it will combine with one of the suffixes characteristic of that
class. These facts are illustrated in the examples of 38. and 41.,
page 47 above.

The class-system in Kaceccereere Fulfulde is essentially the same

as that described in Arnott (1970) for the Gombe dialect. Using
consonant alternation and class-suffixes as our criteria, we can dist­
inguish 24 classes, 19 singular and 5 plural. However, since one of the
classes so distinguished, the Nga class, corresponds to two different
plural ones — the choice depends on whether augmentative meaning is
present or not - it is convenient to split the (morphologically-defined)
Nga-class into two, each corresponding to one plural class. This gives
us 20 singular classes, and 5 plural ones, which I shall name as

indicated in table XIV below, according to the form of the dependent

subject pronoun that is characteristic of the class.

Table XIV

Nominal Classes

Initial Initial
Class 'Meaning' Consonant Class 'Meaning' Consonant
Name Series Name Series •

*0 Sg.Pers. Plosive Nge Continuant

Be Pl.Pers. Continuant Ngo Continuant
Ngel Dim.Sg. Plosive Ngu Prenasal
Koyn Dim.PI. Prenasal Ngal Plosive
Kal Dim.Qu. Prenasal Ngal Plosive
Ngum Dim.Pejor Prenasal Ka Prenasal
Nga Aug.Sg. Prenasal Ki Prenasal
Ko Aug.PI. Prenasal Ko Continuant
Nde Continuant Kol 'calf' Plosive
Ndi Prenasal Bam liquids Prenasal
Ndu Continuant Dum neutral Plosive
Nga Prenasal De non-human Plosive
plural
Di non-human Plosive
plural
 58

Notes:

(i) Table XXV is based on Appendix 3 of A m o t t (1970) , where further

details regarding the nominal classes and their morphology

will be found. The observations there are almost all pertinent
to the Kaceccereere dialect.

(ii) The following abbreviations have been used in table XIV:

sg singular dim. = diminutive

Pi plural aug. = augmentative

qu 'of quantity pejor. = pejorative
pers personal prenasal. = prenasalised consonant

(iii) If no 'meaning' is mentioned for a class, that indicates that the

connotations of the class are too diverse to be summarised.
The reader is referred to A m o t t (1970 Appendix 4) , and to
A m o t t (1967) for further information regarding the meaning or
connotations of the various classes. 'Calf is mentioned with
the Kol class, because this is its sole meaning.

(iv) The 'Initial Consonant Series' mentioned in the the third and
sixth columns of table XIV refers to the three series of

consonants indicated in table VII, page 42 above.

I shall now discuss two issues arising from table XIV. First,
the relationship between singular and plural classes (in §1 .3 .1 .1 .),
and then the 'meaning' of a nominal that is composed of a stem plus
class-suffix (in §1.3.1.2.).

1.3.1.1. Singular and Plural Classes

There is a one-to-one correlation between three of the singular and

three of the plural classes. This fact is illustrated in table XV:
 59

Table XV

One-to-one Correlation between

Singular and Plural Classes

Singular Plural

Class Example Meaning Class Example Meaning

'0 joocfii -cfo 'settled (person)' Be joocfii-£e ’settled (people)'

Ngel meere-yel 'useless (dimin.)' Koyi] meere-hoyi) 'useless (dimin.)'
Nga 6 in-nga 'large child' Ko 6 ik-ko 1large children 1

As well as being in one-to-one relationship with their singular or

plural counterparts, these six classes also have the most clearly
identifiable 'meaning' : J_0 and Be_ are the classes expressing person;
Ngel and Koyq express diminutives; Nga and Ko express augmentatives.
(This Nga class is one of the two Nga classes mentioned on page 57.)

The classes depicted in table XV are so predictable in their meaning

and in their singular-plural equivalents,, that one might want to argue
that they constitute only three classes, each divisible into singular and
plural. However, this would not be a practicable treatment of the other
classes; and for the sake of uniformity, it seems best to consider
singular and plural classes as distinct.

There is no simple correlation between the other seventeen classes

and their plural equivalents. The N d u , N g u , Kol and Ngol-classes, and
the non-augmentative Nga-class, always correlate with the 'Pi plural class.
(There are certain exceptions, but they need not concern us here: the
details given in A m o t t 1970: ch. 16 are relevant to the Kaceccereere
dialect, as well as to the Gombe dialect.) These correlations are
illustrated in table XVI:
 60

Table XVI

Correlation between Singular and Plural Classes

Singular Plural
Class Example Meaning Class Example Meaning

Ndu waa-ndu 'monkey * Di baa-cfi 1monkeys 1

Ngu ndane-wu 'white'(eg. horse ) 1 Di danee-j i 'white (eg. horses)'
Kol nyal-ol 1calf 1 Di nyal-i 'calves'
Ngol keh-ol ’b a m b o o ’ Di keh-i 1bamboos'
Nga nood-a 1crocodile' Di nood-i 'crocodiles'

Ngal gell-al 'guinea-fowl' De gell-e 'guinea-fowls'

Nde juulir-de 'm o s q u e ' De juulir-cfe 'mosques'
Ki lek-ki ’t ree' De lecf-cfe 'trees'
Ko feere-ho 'different (eg. soup)' De feeree-je 'different (eg.soups).
Dum bocfee-jura 'red (eg. thing) ' De bocfee-je 'red (eg. things)

As mentioned briefly in §1.2.3-3 above, various phonological and morpho-

phonological changes may occur at the juncture of nominal stem and class-
suffix. This- explains the variation in certain stems in table XVI:
l ek~ v s * led*-, ndane- vs. danee-, and feere- vs. feeree-.

Because two plural classes correlate with seventeen singular

classes (of which only ten are exemplified in table X V I ) , there is less
precision of reference in the plural than in the singular. This is
especially true of stems that may occur in many different classes, which
we can for the moment call ’adjectives'. For example, the -wu suffix of

— ^ -~wu restn c t s the reference of this word to 'nouns' that occur in the
Ngu class. Thus, the rules of class concord would prevent ndanewu

from being used with any of the other 'nouns' in table XVI. The -ji suffix
of daneeji similarly restricts the reference of the word to 'nouns' that
occur m the Di-class, However, such ’nouns' are extremely numerous. For
example, dan eeji could refer to any of the plural 'nouns’ in the upper
half of the table, since these are all of the D i class. The same

principles apply, mutatls m u t a n d i , to the referential properties of

feereho and feereeje, and to bocfee jum and bocfee j e , in the lower half
of the table.

(There is some doubt concerning the adjective/noun distinction

in Fulfulde, hence my use of inverted commas in the preceding paragraph.
This matter will be discussed in §1.3.2. below).
 61

1.3.1.2. The 'Meaning' of Nominals

Some nominal classes have a definable 'meaning1, such as was ascribed

to the classes featured in table XV. For example, a nominal stem
occurring in the 1Q-class necessarily refers to a person (either male
or female); a stem occurring in the Kol class necessarily refers to
a calf. Other classes have connotations with certain types of meaning,
but they are not so clearly definable. For example, a stem occurring
in the Ngol class is likely to refer to a long thin thing; a stem occurring
in the 'Darn glass is likely to refer to a liquid; and a stem occurring
in the Ki_ class is likely to refer either to a tree, or to a bladed
instrument. The class-suffix thus limits the reference of the stem.

The 'meaning' of a nominal is, then, composite. It is composed

at least of the meaning of the stem and the meaning of the suffix.
If it is a participle, it is also composed of the meaning of the
VA-suffix. This allows great flexibility of expression. For example,
there is a class-suffix meaning 'big' and a stem meaning 'big'; there
is a class-suffix meaning 'person' and a stem meaning 'person'. Thus,
both the following forms occur:

48.a. Nen - nga. b. Maw-cfo

person big big person
CL CL
'Big person . 1 ‘Big person.1

These Fulfulde words in fact have different connotations: nen-nga

is 'a huge person', whereas maw-cfo is more 'a chief, boss'. What
concerns us here, however, is the morphological composition of
Fulfulde nominals, rather than their precise semantic analysis.

The above examples do not only show that Fulfulde nominals are
morphologically and semantically composite. They also show the doubt­
fulness of the 'noun' vs. 'adjective* distinction in Fulfulde: for
'big' is expressed in the one instance by a suffix (-nga in 48.a . ) ,
and in the other by a stem (maw- in 48.b.); and 'person' is expressed
18
in the one instance by a stem (nen- in 48.a.), and in the other by
a suffix (-cfo in 48-b.). Let us now turn to the issue of nouns and
adjectives, which was brushed aside at the beginning of §1 .2 .2 ..
 62

1.3.2. Nominals: Nouns, Adjectives and Participles

A stem such as n e n - , seen in 48.a., occurs in few of the nominal classes.

Those that it does combine with are indicated in table XVII. Note that
its stem-final consonant assimilates to the initial consonant of the
class-suffix.

Table XVII

Range of classes occurring with, nen- ’person 1

Class Nouns Meaning

10 necf-cfo 'person*
Ngel nen-ngel 'little person'
Koyij nek-koyij 'little people’
Nga nen-nga 'big person'
No nek-ko 'big people'
Ngu nedcfaa-ku 'humanity, kindness

The plural of necf—cfo is yimbe , a suppletive form. Nedcfaa—ku is probably

a derived noun, for it is obviously based on necC-cfo, rather than the
'simple stem', nen-.

By way of contrast, a stem such as maw- 'big1, seen in 48.b.,

occurs in all 25 nominal classes. Those shown in table XVIII are only
a sample. Note, again, that assimilation often occurs between the
final consonant of the stem and the initial consonant of the suffix.

Table XVIII

Sample of classes occurring with maw- 'big , important 1

Class Adjectives Approximate Class Adjectives Approximate

Meaning Meaning

'0 maw-cfo 'important (person)' Be maw-be 'important (people)*

Ngel man-ngel ’important (little Ki mak-ki 'big (eg. tree)'
person or thing)’ Nde maw-nde 'big (eg. b o o k ) '
Koyi) mak-koyg 'important (.little Nge man-nge 'big (eg. cow ) 1
people or things)' Dum maw-dum 'big (thing)'
Nga man-nga 'huge (person or Ndi maw-ndi 'big (eg. b u l l ) '
thing)' ■Di maw-cfi 'big (eg. horses)'
Ko mak-ko 'huge (people or Be maw-cfe 'big (eg. beans)'
things)' Ka mak-ka 'big (eg. speech)'
Ngu man-ngu 'big (eg. h o r s e ) ' Kol mak-kol 'big (calf)’
 63

The classes on the left in table XVIII are the same as those in table
XVII; those on the right are a sample of other classes with which
maw- may combine.

Traditionally, forms based on a stem such as nen- are termed 'nouns',

and those based on a stem such as maw- 'adjectives'. This is the approach
adopted by Taylor (1921: ch.7), Noyz (1974: §296), and Skinner and
Pelletier (1979: 362), for example. Amott, although he uses both
terms, expresses considerable scepticism concerning the distinction (1970:
§15.1.). However, he defends it on two grounds. Morphologically,
'adjectives' occur in a wider range of classes than 'nouns'. This
is certainly borne out by the evidence of nen- and m a w - ; ;y e t could
it not be explained in terms of the semantic compatibility of stems and
class-suffixes? Syntactically, he claims that 'nouns' cannot qualify
the head of an NP, except in a genitival capacity. This point is worthy
of investigation, which I shall not pursue here. A m o t t could have
added a further difference: namely, the fact that 'adjectives' almost
always have verbal cognates. There is, for example, a verb mawn-

meaning 'to be or become big', but there is no parallel cognate verb

for n e n - .

I share Arnott's scepticism concerning the adjective - noun

distinction, despite the points made in the latter half of the above
paragraph. Morphologically, 'nouns’ and 'adjectives' occur in which­
ever classes semantic compatibility allows. The combination of rew-
'female 1 with -d a m (the class-suffix commonly referring to liquids)
is no more ridiculous than the combination of nag- 'cow* with -do
(the class-suffix reserved for 'person') - though both are conceivable
in a non-factive environment (cf. Lyons 1977: 795). Yet rew- is
commonly considered to be an adjectival stem, and nag- a noun one.
Moreover, the combination of rew- with the ^0_ class suffix (and certain
morpho-phonemic changes) gives the form debb-o 'woman' - presumably
a 'n o u n '.

D e b b - o 's membership of the rew- class paradigm, and its function

as the word for 'woman' highlights another point concerning 'nouns' and

'adjectives’. Every 'adjective* in Fulfulde can be used, syntactically.

 64

as an NP. In other words, a so-called adjective can be the subject

or object of a sentence; and it can also be the topic or focus of a
sentence - these matters will be further discussed in L§2.3. and
in chapter 6 below. An 'adjective' does not have to be accompanied
by the 'noun' that it supposedly qualifies - it can stand quite
independently.

This is not the place for the detailed syntactic and morphological
analysis that is needed in order to do justice to the noun/adjective
issue in Fulfulde. For present purposes, I shall ignore such differences
as there may be between stems, and concentrate on their similarities -
namely, t h e ,fact that both 'adjective' and 'noun'-stems necessarily
inflect for class; the fact that their initial consonant is sensitive
to class, if it is of the alternating type; and the fact that a word
consisting of an 'adjective'- ■ or 'noun'-stem plus class-suffix may
behave syntactically as an NP. These are also the characteristics of

participles, though the latter also inflect for Voice and Aspect:
for example, the -ii- in joocf-ii-cfo and joocf-ii-6 e in table XV expresses
19
Middle Voice and Completive Aspect. Given the characteristics that
nouns, adjectives and participles have in common, I have included
them all under the umbrella term 'nominals' for the purposes of this
discussion of the class-system in Fulfulde.

1.3.3. The Nominal Class System: Conclusions

Each nominal class has its own set of pronouns. These include independent
(or emphatic) pronouns, which are anaphoric in reference, and are used
as topic or focus of a sentence (cf. §2.3.1. - 2.); and dependent
pronouns, which are similarly anaphoric, but are used in verbal complexes
(ch. chapters 3 and 4). They also include deictic pronouns, of which
there are several types in Kaceccereere Fulfulde; and interrogative
20
pronouns. Here is the full paradigm of pronouns for a single class,
the Nge class.
 65

Table XIX

Pronouns of the Nge class

Independent Pronoun: kannge 'the one (eg. cow) referred to'

Dependent Pronoun: nge 'it (eg. cow)'
Near-deictic Pronouns: ngee 'this one (eg. cow)'
ngeecfoo 'this one (eg. cow)'
ngecfcfon 'this one (eg. cow) 1
Far-deictic Pronouns: ngeya 'that one (eg. cow)'
ngetton 'that one (eg. cow)'
Interrogative Pronoun: ngeye 'which one (eg. cow)?'

It is not clear how the various near- and far- deictic pronouns differ
in meaning and/or usage. Several of them are in fact morphologically
composite: doo means 'here1, don 'around/present', and ton 'there',
and forms such as ngetton obviously originate from the coalescence
of nge and ton.

In the discussion of verbal morphology, I shall frequently give

examples of dependent pronouns occurring in verbal complexes. In Fulfulde
sentences, I use pronouns from as many different classes as possible.
However, in tables (such as XXIII in chapter 4, and in Appendix A) ,
it would be impracticable to show the dependent pronoun for each and
every class, and I therefore show the Be class-pronoun as representative
of other classes. Usually, I place 'etc.' after ,. in order to indicate
its representative role.

There is one nominal class, however, which has different dependent

pronouns, according to conjugation Ctense') and case. This is the j_0
class. The dependent pronouns for this class are therefore presented
separately from others. In its range of pronoun-variation, the
class is similar to the 1 st and 2nd person pronouns, which are not,
of course, class-pronouns.

1.4. Conclusion

In this chapter, we have seen some of the essential linguistic features

of Kaceccereere Fulfulde: its phonology, the system of consonant
alternation, and the nominal class system. We are now in a position to
consider larger linguistic units: Fulfulde sentences, Noun Phrases
(NP's), and Verbal Complexes.
 66

Chapter 1: Footnotes

1. The infinitives given in table I are of the Active Voice. The ;u

vowel only occurs obligatorily after a C W C o r CVCC verbal root, and
I argue in §4.2.1. that it should be regarded as anaptyctic.
It is shown in all the examples here for the sake of uniformity.

2. Non-geminate p is never found in root-final position. This example

shows jd in root-initial position.

3. Certain common verbal roots and nominal stems, including nyaam-,

are identical. Which is derived from which, or indeed whether either
is more 'basic' than the other, is Immaterial to our concerns here.

4. Concerning the definition of 'nominal', see $1.3.2. below.

5. This analysis of stress in nominal* differs considerably from that

in A m o t t (1965: 78) , where it is claimed that stress occurs,
basically, on the first syllable.

6. The system of pitch-representation used here is very ad hoc in nature,

but the following principles have been adhered to:
(i) The syllable with highest pitch is marked by a large
dot, the others with a small dot.
(ii) Pitch-movement on a given syllables is indicated by a
tail, the length of the tail corresponding to the amount
of pitch-movement.

7. In word-final unstressed syllables, the difference between long and

short vowels is neutralised. Although my transcription consistently
shows long and short final vowels in this position (cf. §1 .1 .5 .1 .. below),
the former must be considered short for the purposes of syllable-
analysis, unless they are stressed. Where appropriate, I therefore
place brackets around the second part of final vowels in the syllable-
analysis that accompanies the pitch-representations.

8. Since this monsyllabic word is stressed, its final long vowel is

not shortened.

9. Concerning focus and contrastive stress, see footnote 24*, p. 118.

10. Another common verbal root that behaves like w i 1- is yi'-

'see'. It may be that yah- is the only root that drops' its final
li, as described here. Of the other, rare, h-final roots, mah-
1b u i l d ' , for example, retains its final h in all inflectional
and derivational paradigms.

11. Consonant-alternation in class-suffixes is mentioned in Labatut

(1973: §6.1.), and in Sapir (1971: 72-73).
 67

12. Professor Jack Carnochan has suggested that there may be a

correlation between non-alternating stem-final consonants and
zero grade nominal class suffixes. It may well be that further
investigation of this issue will enable such a correlation to
be established.

13. See note 19 below.

14. Table XII is based on a count of some 1,000 verbal roots. It

was found that for the number of roots ending in a continuant,
only approximately a quarter of that number end in its corres­
ponding non-geminate plosive. It is on the basis of these
observations that the terms 'most commonly 1 and 'rarely' are
used.

15. Of the alternating continuants shown in table VII (page 42),

only r_ has a geminate counterpart, found for example in
torrugo 'to bother, trouble'. This fact tallies with the
observations made in the discussion after table XII (page 54).
Non-alternating continuants, such as 1, n, and m, on the other
hand, usually do have geminate counterparts, found in Pullo
'a Fulani person', hinn- 'greet' and lamm- 'be bitter', for
example.

16. The following non-geminate consonants are never found in root-

final position in this dialect: p, ngy mb and nj.

17. Observations about consonant-alternation in root-final position,

both in Wolof and the Fulfulde (or Pulaar) of Senegal are to
be found in Sapir (1971: 76-77 and 104-106).

18. This stem is quoted as nen- for convenience. As can be seen

from table XVII, its final consonant is usually conditioned
by the following suffix.

19. Participles are by definition positive rather than negative.

It is therefore not possible to argue that the -ii- of
joocf-ii-cfo, for example, expresses positive polarity, since
negative polarity cannot occur in this environment. In this
respect, participles differ from verbal complexes, which do
express polarity,
20. Deictic and interrogative adjectives are morphologically ident­
ical to their pronominal counterparts.
 68

Chapter 2

The Fulfulde Sentence, Noun Phrase

and Verbal Complex

2.0. Introduction

In this chapter, I shall discuss the Fulfulde sentence, and give work­
ing definitions of the verbal complex and the noun phrase, since these
categories are crucial for the discussion in the following chapters.

2.1. The Fulfulde Sentence

I use the term 'sentence' here in the way in which it is used in trans­
formational grammar, i.e. to refer to what is called a clause, or a
one-clause sentence, in traditional grammar.

Fulfulde sentences can be divided into two main thpesr verbal and
non-verbal. I shall give some examples of each here, and then discuss
and define each in more detail. First, here are six examples of
verbal sentences;
1
49.a. 'E-mo-woow-a - cfum. b. 'E-be - mboow-a um.
he get it they get it
used used
S V VAP 0 S V VAP O
’He's getting used to it. 'They're getting used
to i t . 1
c. Dume raboow - at - aa?
what get used you
O V VAP S
'What are you (sg) getting used to?'

50.a. Mi-hul - ay - ngel. b. Min-kul - ay - n g e l .

I fear it we fear it
S V VAP O S V VAP O
*I fear i t . ' 'We (excl) fear it.'

c. Ndee kul - ay- mi-ngel?

when fear I it
V VAP S O
'When do I fear it?
 69

These sentences include a verbal complex, ie. a verbal root (woow-/

2
mboow- and hul-/k u l - ) , plus various 'appendages'. Amongst the latter
are pronouns (mo- -cfum, - 6 e - , -cCum, -aa, m i - , -ngel) ; and VAP-suffixes
{-a-, -at-, - a y - ) , which express voice, Aspect and polarity (see §2.2.1.
be low).

Two features distinguish sentences such as the above from non- ^

verbal sentences. Firstly, the initial consonant of the verbal root
is 'sensitive' to the position and number of its subject: w- and h- ^
occur in the a. sentences, where the subject (-mo-, mi-) is singular and
precedes the root; but mb- and k- occur in the b. sentences, where the
subject (6 e - , m i n - ) is plural and precedes the root, and in the c.
sentences, where the subject (-aa, -mi-) follows the root. This patt­
erning of alternating consonants is characteristic of verbal complexes:
see §1 .2 .1 . above.

Secondly, where the subject is 3rd person, there is no nominal-

class concord between the subject and any other constituent of the
sentence. The subject, - m o - , of 49.a. is of the class, and that
of 4 9 .b., 6 e-, is of the Be^ class: but neither of these classes is
represented elsewhere in the sentence. (In each sentence, -cfum is
an object pronoun, of the 13um class.)

Verbal sentences by definition contain a verbal complex, or may

indeed consist exclusively of such, a complex, as do the a. and b.
sentences above. Non-verbal sentences, on the other hand, by definition
do not contain a verbal complex. Here are some examples of non-verbal
sentences:

51. a . Koyij ndane-hoyi) .3

they white
CL
'They (dim.pl.) are white.'

52.a. '0 boow-u - cfo cfum.3

he get it
used
VA CL
'He's (a person who has got) used to it.'
 70

53.a. Be woow-u - 6e cfum.

they get it
used
V A CL
'They're (people who have got) used to it . 1

54.a. Nyaako 'am Sarkin Fulaani Naaraayi.4

father my Sarkin Fulani Narayi
1My father is Sarkin Fulani N a r a y i .'

55.a. '0 Sarkin Fulaani Naaraayi.

he Sarkin Fulani Narayi
'He's Sarkin Fulani Narayi.'

5 6. a . Labi woncfon.
knife is
'It's a k n ife.'

57.a. B u m la 6 i.
it knife
'It's a k n i f e . '

58.a. 'E-ngel cfon cfoo.

it here
'I t 's h e r e .'

59.a. Yillaa&e 'e cfon.

"The YillaBe (lineage) exist (s).'/'There*s the Yilla&e
(lineage).

These sentences represent three types of non-verbal sentences: those

expressing attribution (51.a. - 53.a . ) ? identification (54.a. - 57.a.);
and location and/or existence (57.a. - 59.a.). These three types will
be dealt with in sections §2 .1 .2 .1 . - §2 .1 .2 .3. respectively

Although 52.a. and 53.a. each contains a participle, none of the

sentences in 51.a. - 59.a. contains a verbal complex. The difference
between participles and verbal complexes will be elaborated on in greater
detail in §2.1.2.1. below. For the moment, however, it is worth noting
two points concerning sentences 51.a. - 53.a.. Firstly, each of the
nominals in these sentences has a nominal-class suffix, which I have
 71

separated from the nominal stems: the -ho y q , - *o and - e suffixes

represent the Koyq, <3 and Be_ classes respectively; and these suffixes
display class-concord with the subject of these sentences. Secondly/
the initial consonant of each nominal is one of the consonant-series
required by the class represented by the class-suffix. For example,
the Koyi} class requires a prenasaiised consonant: thus ndane-hoyq
in 51.a. begins with nd-; the jO_ class requires a plosive consonant,
hoow-u-cfo in 52.a. thus begins with b-. These points are highly
relevant to the distinction between verbal and non-verbal sentences.

2.1.1. Verbal Sentences

We have already noted the following points about verbal sentences:

(i) They necessarily contain a verbal complex;

(ii) If the initial consonant of a verbal root in a complex

is an alternating one, then it is sensitive to the
number and position of its subject.

(iii) Although the verbal complex may contain class-pronouns

(such as 5e and cfum) , there is n o class-concord between
the subject and any suffixes of the verbal root.

2.1.1.1. Deletion

Let us now consider some other characteristics of verbal sentences.

One of these is the fact that under identity conditions, a subject
or object noun-phrase, or the verbal complex itself, may be deleted
from the surface-structure of the sentence. In sentence 60., for instance,
the verbal complex koot-i has aero subject in surface structure,
because its deep-structure subject is the same as that of the preceding
verbal complex,'umm-ake:

60. Bi65e 'umm - ake, sey 0-koot-i

children arise then go
home
S V VAP S V VAP

’ (The) children got up and went h o m e . 1

Sentence 61., on the other hand, has zero object as well as zero subject;
again this is caused by the fact that the subject and object of the
second verbal complex, wacf-a, are in deep structure the same as those
of the first verb, wadd-a.
 72

61. Sey nedcfo wadd - a nebbam na'i, sey 0 wacf-a 0.

then person bring oil cows then put
S V VAP 0 S V VAP 0

"Then one should bring butter, and put it in.'

We see an example of zero verbal complex in sentence 62.b., where the

njar-u-mi of 62.a. has been deleted. Such deletion is possible despite
the fact that the subject of the two verbal complexes ('o-yar-i and njar-u-mi)
is different? the reason for this is no doubt the presence of a topic
noun-phrase at the beginning of each sentence: kanko and minn in
the first and second sentences respectively.

62.a. Kanko, mboyri 'o - yar - i; minn bo, kosam njar-u - mi.
him gruel he drink me milk drink I
TOPIC 0 S V VAP TOPIC TM 0 V VAP S
' (As for) him, he drank gruel, (as for) me, I drank milk.'
1He drank gruel, whereas I drank m i l k . 1

b. Kanko, mboyri 'o - yar - i; minn bo, kosam 0.

him gruel he drink me milk _
TOPIC 0 S V VAP TOPIC TM 0 VS
* He drank g ruel, I m i l k . 1
'He drank gruel, whereas I drank milk.'

Zero subject is also found with affirmative imperative verbal

complexes. The fact that the subject is present in deep-structure
but deleted in surface-structure is suggested by the consonant alter­
nation that can be seen operating between singular and plural imperatives
Thus, for instance, in the sentences of 63., we see that the verbal
root has its continuant/plosive form, joocf- for the. singular imperative,
and its prenasalised form, njoocf-, for the plural imperative. (This
verbal root has no *yoocf- form, and joocf- thus doubles as the continuant
and the plosive form.)

63. a. Joocf- a I b. Nj oocf-ee I

sit sit
VAP VAP
'Sit down! (sg. ) 1 1Sit down! (pi.)'

2.1.1.2. Verbal Categories

Another characteristic of verbal sentences is the fact that, by virtue

of the verbal complex they necessarily contain, they can express
properties of verbal categories such as:
 73

(i) mood (indicative versus non-indicative)

(ii) voice (active, middle, passive)

(iii) polarity (affirmative versus negative)
(iv) stativity
(v) continousne ss
(vi) aspect (completive versus incompletive).

There is also a distinction between the form of the verbal complex

found in relative clauses, sentences with a focussed/contrasted NP, and
sentences with a question-word on the one hand; and sentences without
any of these features, on the other. Rather unsatisfactorily, one could
g
call it the 'relative - general' distinction. All of these categories

will be further discussed and defined in the p.ext chapter; for the
moment, it is sufficient to mention that non-verbal sentences are much
more restricted than verbal sentences in the properties they can express.

2.1.1.3. Negation

Another characteristic of verbal sentences is that most of them form

their negative equivalents by means of a special negative suffix,
which replaces an affirmative suffix within the verbal complex. Changes
of pronominal form within the verbal complex may also accompany negation.
For instance, the negative sentences equivalents to 49.a. and-b. are
64.a. and b . :

49. a . 'E-mo-woow-a-cfum. 64. a . 'O-woow - ataa-cfum.

he get it he get not it
used used
S V VAP O VAP
'He's getting used to it. 'He's not getting used
to it/won't get used to
it. '

49.b. 'E-&e - mboow-a-cfum. 64.b. Be - mboow-ataa-cfum,

they got it they get not it
used used
S V VAP Q VAP
'They're getting used to it.' 'They're not getting used
to it/won't get used to
it. '
 74

If we compare these sentences with their affirmative equivalents,

we notice that the -a. VAP suffix has been replaced by -at a a , and
also that the J_e morph before the subject has been dropped. We also
notice that -mo- has been replaced by 1o- as the subject pronoun.
In fact, two negative conjugations, or '•tenses', are equivalent to six
positive conjugations, and this means that various distinctions
found in positive verbal sentences are neutralised in negative
verbal sentences (cf. §3.6 .1.2. and §3. 6 .2.2.). That is why I
have given alternative translations to sentences 64.a. and b..

There are, however, three conjugations whose negative counter­

parts involve a special sentence-initial negative particle, rather
than a negative conjugation. The Imperative and the Subjunctive
7
form their negative with the sentence-initial taa or t o , as in
65.b. and 6 6 .b. below; and the Emphatic Completive forms its negative
with n a a , as in 67.b. In each instance, the rest of the sentence
is identical to its positive counterpart.

65.a. Hu6b - u hiite. b. Taa hu 6 6 -u hiite.

light fire light fire
VAP NEG VAP
'Light a fire.'' 'Don't light a firei'

6 6 .a. De - kebb-in - ee. b. To cfe - kebb-in - ee.

they full they full
EXT VAP NEG EXT VAP
'They (eg. calabashes) 'They (eg. calabashes) should
should be filled.' not be filled.'

67.a. Mi-sur - al Naa mi-sur - al

I prevent I prevent
VAP NEG VAP
11 was prevented/held u p l 1 'I wasn't prevented/held u p l'

The naa seen in 67.b. is also used to form the negative of various
types of non-verbal sentences, and for this reason, negation is
not a tight criterion for the distinction between verbal and non­
verbal sentences.
 75

2.1.2. Non-Verbal Sentences

Non-verbal sentences do not contain a verbal complex, though they

are often equivalent to sentences containing a copular verb. In
this section, I shall discuss attributional, identificational, and
locative/existential sentences, and I shall also mention certain
types of non-verbal sentences that do not fit any of these categories.

2.1.2.1. Non-Verbal Sentences: Attribution

Sentences involving the attribution of a quality to the subject

usually consist of the subject, followed by a nominal expressing that
quality. The subject and the nominal display class-concord. Thus,
koyg and -hoyr> in 51.a. are of the same class, as are 1o and -do
in 52.a. and 6 e and - 6 e in 53.a..

51. a. Koyi; ndane-h o y q .

they white
CL
'They (dim.) are w h i t e . 1

52.a. _|_0 boow-u - do cfum.

he get it
used
VA CL
'He's (a person who has got) used to i t . 1

53. a. Be woow-u - 6 e cfum.

they get it
used
VA CL
'They're (people who have got) used to it.'

Sentences involving attribution have counterparts using the copular

, .8
verb wonx :

9
51.b. Koyq-ngon-i ndane-hoyq.
they are white
VAP CL
'They are white .'
 76

52.b. 10-won-i boow- u- cfo cfum.

he is get it
used
VAP VA CL
'He's (a person who has got) used to it.'

53.b. B e - ngon-i woow-u - 6 e cfum.

they are get it
used
VAP VA CL
'They are (people who have got) used to it.'

It seems likely, in fact, that the a. sentences are derived from

their b. equivalents. However, the former are much more commonly
used than the latter.

The negative counterparts of attributional sentences are formed

by placing naa in front of the non-verbal positive sentence, thus:

51. c. Naa koyi} ndane-hoyg.

not they white
CL
'They (.dim.) are not white.'

52.c. Naa 'o boow-u-cfo cfum.

not he get it
used
VA CL
'He's not (a person who is) used to i t . 1

53.c. Naa 6 e woow-u - 6 e cfum.

not they get it
used
VA CL
'They're not (people who are) used to it.'

However, naa cannot be placed in front of attributional sentences

which include woni (such as 51.b. - 53.b.). Thus, the c. sentences
above are the only negative counterparts of both the a. and the b.
sentences. (This naa is the same naa as was seen in 67.b. on page 74
above.)
 77

Attributional sentences generally have interrogative counter­

parts involving the use of noy 'how ? 1 and w o n i . For example, 51.a,
would be a suitable answer to 51.d . :

51.d. Noy koyq. ngon-i? 51. a-. Koyr^ ndane-hoyg.

how they are .they white
CL
’What are they like?'
"They (dim.) are white.'

Sentences 52.a. and 53.a. provide interesting test-cases for

the definition of verbal and non-verbal sentences, and especially
if they are compared with 49.a. and b.. The sentences concerned
are juxtaposed below, and the suffixes and consonants under discussion
are underlined:

49.a. 'E-mo-woow-a - cfum 49 .b. 'E-'6 e - mboow-a - cfum.

he get it they get it
used used
VAP VAP
'He's getting used to it.' ’T h e y ’re getting used to it.'

10
52. a. '0 boow-u - db cfum. 53.a. Be woow-u - 6 e_ cfum.
he get it they get it
VA CL VA CL
'He's (a person who's got) 'They're (people who've got)
used to it.' used to it.'

Sentences 49.a. and 52.a. are similar, inasmuch as they both have a
singular subject pronoun of the jdD class; similarly, the subject of
both 4 9 ,b. and 53.a. is b e . Moreover, all these sentences end with a
pronoun of the Dum class.

There are, however, three respects in which 49.a. and b. on

the one hand differ from 52.a. and 53.a. on the other:

(i) Whereas in 52.a. and 53.a., the nominal class of the subject
is picked up later in the sentence (by the -do and - 6 e
suffixes), this is not the case in 49.a. and 49.b., where the only
other class-like item is cfum.

(ii) In 49.a. and 49.b . , the verbal root displays sensitivity to

the preceding subject: cf. § 1 .2 .1 . above, and the following further
examples:
 78

68. a. Mi-woow-ii - mo. b. Min -mb o ow- i i - mo.

I get her we get her
used (excl) used
VAP VAP
"I've got used to h e r . 1 'Wei'ive gat used to h e r . ’

It is characteristic of this 'sensitivity' that a preceding singular

subject correlates with a continuant consonant (such as w ~ ) , whereas
a preceding plural subject correlates with a prenasalised consonant
(such as m b - ) .

(iii) In 52.a., although the subject precedes and is pingular,'.the

root begins with b_- (not w-} ; and in 52.b . , although the subject
precedes and is plural, the root begins with w- (not m b - ) . A plosive
initial consonant in fact characterises the j_0 class, and a continuant
the Be class (cf. table X I V ) .

Given these facts, we can but conclude that 49.a. and b. are
verbal sentences, ie. include a verbal complex; but that 52.a. and
53.a. are non-verbal sentences, ie. consist of a subject and jux­
taposed nominal, but include no verbal complex ,^1

2.1.2.2. Non-Verbal Sentences; Identification

I shall divide sentences involving identification into two types: those

that involve the use of a dummy subject, cfum (as in 56.a.) or a clefting
particle wondbn (as in 57.a.); and those that do not (such as
54.a. and 55.a.). I shall consider the latter first.

Sentences 54.a. and 55.a. are in many respects similar to 51.a.

and 53. a. on page 75 above. Although it is not immediately obvious
in 54.a . , there is class concord between the subject and the jux­
taposed NP in each sentence. In 54.a . , two NP's are juxtaposed:.

54.a. Nyaako 1am Sarkin Fulaani Naaraayi.

father ray , Sarkin Fulani Narayi,
NP ’ ' NP"

'My father is Sarkin Fulani Narayi.'

 79

If the intonation peak of 54.a. falls within Sarkin Fulaani Naaraayi,

then Sarkin Fulaani Naaraayi identifies the preceding NP, nyaako 1a m .

(With an alternative intonation pattern, 54.a. might be analysed
as two NP's, the second in apposition to the first.) Although
proper names do not have a class-suffix, their nominal class becomes
evident when they are referred to anaphorically. Sarkin Fulaani Naar­
aayi would be referred to by J_o, as indeed would nyaako 'am; thus
there is, in effect, class-concord between the two NP's in 54.a..
In 55.a., the subject is obviously of the class; and there is thus
agreement between the subject NP and the 'predicating N P 1 there too;

55.a. 'O Sakkin Fulaani Naaraayi.

he Sarkin Fulani Narayi
'He's Sarkin Fulani Narayi.'

Woni may intervene in 54.a. and 55.a. (cf. 51.b. - 53.b. on

pages 75 - 76 ):

54.b. Nyaako 'am won-i Sarkin Fulaani Naaraayi.

father my is Sarkin Fulani Narayi
VAP
'My father is Sarkin Fulani Narayi.'

55.b. 'O-won-i Sarkin Fulaani Naaraayi.

he is Sarkin Fulani Narayi.'
VAP

'H e 's Sarkin Fulani Narayi.'

Moreover, the negative counterparts of sentences 54.a. and 55.a.

involve the use of n a a , thus:

54.c. N y a a k o 'am naa Sarkin Fulaani Naaraayi.

father my not Sarkin Fulani Narayi.

'My father is not Sarkin Fulani Narayi.'

55.c. Naa *o Sarkin Fulaani Naaraayi.

not he Sarkin Fulani Narayi
'He's not Sarkin Fulani Narayi.'

As can be seen from 54.c., naa occurs after an NP, but before a
pronoun subject in such sentences. The implications of this fact
are considerable - chiefly because there is seme doubt as to whether
 80

nyaako '-am is the subject or the topic of 54.c., and indeed of

12
54.a., Thxs matter will not be discussed any further

Although in the above respects, identificational sentences are

identical to attributional sentences, there is one respect in which
they differ. The interrogative counterparts of 54.a. and 55.a. are
formed not with noy (cf. 51.d. on page 7 7 ) but with an interrogative
class pronoun, such as m o y e :

54.d. Moye won-i nyaako maada? 55.d. Moye 'o-won-i?

who is father your who he is
VAP , VAP
’Who is your father?’ 'Who is he?'

The fact that nyaako maada follows, whereas ^_o precedes, won-i
in 54.d. and 55.d. respectively is attributable to the fact that
'o is a dependent pronoun, whereas nyaako maada is an NP.

Let us now turn to identificational sentences involving

woncCon or cfum:

56.a. La&i wondon. 57.a. D u m la 6 i

knife is it knife
'It's a knife.' ’It's a knife.’

56.a. and 57.a. are in fact synonymous. Syntactically, however, they

are very different: for woncfon is a clefting particle of verbal
origin, and with certain verbal characteristics, whereas dum is a class
pronoun of the 'neutral' class (cf. table XIV). Both wondon and
dum are used to help focus an NP: see §6 .2.1.7. of chapter 6 .

Wondon is derived from the copular verb woni plus the adverb don
'present, around': in fact, wonidon (with -i_-) is in free variation
with wondon (without **i“ ) • The initial consonant of wond’
on
alternates according to the number of its preceding subject,
t hus:
 81

56. a. Lafci woncfon. b. Labe ngoncfon.

knife is knives are
'It's a knife.' 'They're knives.'

Despite this verbal characteristic, woncfon does not take any verbal
inflections, except the -i_- seen in wonicfon - if indeed that
13
is an inflection. Nor can a dependent pronoun subject be appended
to it: contrast 56.c. with 55.d. above, where 'o- is appended to
wo n - i :

56. c . *Ki - woncfon.

it
(eg •
knife)

Woncfon thus hardly qualifies to be considered a verbal complex, and

I treat sentences involving woncfon as non-verbal.

Whereas woncfon always follows, cfum always precedes the NP that it

identifies. There is no evidence to suggest that sentences with
cfum are derived from verbal sentences, for a construction such
as 57.b. is unacceptable:

57.b. *Dum won— i labi

it is knife

Like the other identificational and attributional sentences

discussed in this section, and in §2 .1 .2 .1 ., sentences with woncfon
or cfum have negative equivalents with n a a :

56.d. Naa labi woncfon. 57.c. Naa cfum labi

not knife is not it knife
'It's not a knife.' 'It's not a knife.'

The interrogative counterparts of identificational sentences with

woncfon or cfum involve the use either of cfume 'what?', or of an inter­
rogative class pronoun of another, non-neutral, class, which will
convey 'which?1. Thus:
 82

56. e. Dume woncfon? f. Kiye woncfon?

what is which is
1What is it ? 1 'Which is it?'

57.d. D u m cfume? e. D u m kiye?

it what it which
'What is it ? 1 'Which is it?'

The factors that determine the choice between cfume and a pronoun of
a non-neutral class are not easy to describe, and it is not, entirely
certain whether the ’what*/.‘w h i c h * dichotomoy just mentioned
correlates exactly with the Fulfulde facts. Consider, for example,
sentences 55.a. and 57.a. on p a g e s 79 and 80. In the former,
we see the juxtaposition of a class-pronoun, 1o , and an NP,
nyaako 'am. Syntactically, this is the same as 57.a., where we see
the juxtaposition of cfum and an N P , labi. The difference is that
whereas j_o and nyaako 'am display class-concord, cfum and labi
do not.

For the moment, we can say that if the speaker knows that the
item to be identified is likely to be of a particular nominal-class
(for example, most bladed instruments are in the ki_ class) ,
then the class—pronoun is likely to be used; but if the probable
class of the item to be identified is not known, then the neutral
cfum is more likely to be used. This distinction is not watertight,
however, for cfum is often used where a dummy 'it' occurs in English,
and even where the item concerned is bound to be personal. For
instance, D u m moye? 'It's who?' will be said when a person is
heard arriving at a compound (homestead) — not *0 moye? 'He/she
is who?' It may be that questions of reference determine such usage:
in any case, a lot more research is needed to clarify these issues.

2.1.2.3. Non-Verbal Sentences: Location and Existence

Under this heading, I shall consider sentences such as the following:

14
58.a. 'E ngel cfon cfoo. 59.a. Yillaabe le cfon.
it here Yillabe
It (dim.) is h e r e . 1 "The Yillabe (lineage)
exist(s). ‘/'There are the
Yillabe (lineage).'
 83

These both involve a subject' (ng e l , Yillaabe) with 1e and cfon; though
58.a. also involves dpp ’here'.
Moreover, they both have negative
15
equivalents using the defective verb walaa:

58.b. Ngel wal-aa cfoo. 59.b. Yillaabe ngal-aa.

it is here Yilla e are
NEG NEG
'It (dim) is not here.' 'The Yillabe don't exist.'
'There aren't any Yilla e . '

The essential difference between such sentences is the fact that a

sentence including a locative adverb, such as cfoo in 58.a., expresses
location; whereas a sentence without such an adverb, such as 59.a.
'only' expresses existence.

J_E will be discussed again in $2.2.2.4. below. It is uncertain

whether this 'e , which I shall term ’predicating 'e 1 , is cognate with
the preposition 'e meaning 'with'/'by1, and with the conjunction 'e
meaning 'and' (between NP's only). For the moment I shall disregard
the existence of the other types of 'e, and regard the latter as
homophones of the f o rmer.

Predicating occurs optionally in locative and existential

sentences: both 58.a. and 59.a. would be acceptable without 'e,
for example:

58.c. Ngel cfon cfoo. 59.c. Yillaabe cfon.

it here Yillafce
'It (dim) is here.' 'The Yilla e exist . 1

Nevertheless, Kaceccere'en speakers prefer to use |_e in such sent­

ences: in some sense, it seems to 'smooth over' the predication of
existence or location to the subject - which is why I call it
'predicating 'e'. Mote that when 'e does occur, it precedes a proncun
subject, (cf. 58.a . ) , but follows an NP subject (cf. 59.a . ) . The
reverse order is unacceptable:

58 .d. *Ngel 'e cfon cfoo. 59.d. *'E Yillaabe cfon.

In the glosses of Fulfulde examples, I do not translate 'e, since it

is very difficult to assign a definable 'meaning' to it.
 84

1)011 is, morphologically, a member of a paradigm of locative

adverbs, which include ton 'there 1 and cfoo 'here1. D o n can be

assigned the meaning 'present', 'around', or even 'existant': it
does not have such deictic force as ton and cfoo.

In locative sentences, cfon is optional? though if it is deleted,

'e must be retained. Thus, an alternative to 58.a. and c. is 58.e.;
but not 58.f . :

58.e. 'E ngel cfoo. 58.f. *Ngel cfoo.

it here it here
'It Cdim) is h e r e .'

58.e. is slightly less emphatic than its counterparts with cfon: for
example, it is less suitable as a response to a question regarding
the whereabouts of 'it'. In existential sentences, however, cfon occurs
obligatorily. Thus 5 9 .e. and f. are not alternatives to 59.a. and c.:

59.e. *Yillaa&e 'e. f. *Yillaa 6 e.

Both _|_e_ and cfon are used in the formation of the Progressive
and Stative verbal conjugations, as can be seen in:

70. 'E-mo-cfon-nokk - a . hiite.

he pick up fire
VAP
PROGRESSIVE ACTIVE
'He's picking up a burning twig from one fire to light
another.'

71. a. Musuuru 'e-cfon-cfaan - ii,

cat • sleep
VAP
STATIVE MIDDLE
'The cat's asleep.'

I have argued elsewhere that the Progressive and Stative have arisen,
diachronically, from the conflation of existential sentences and a
verbal form: 70. and 71.a. might have meant, at an earlier stage in
the language 'He is around (1e-mo-cfon), picking up fire (nokk-a h i i t e ) 1,
and "Ihe cat's around (musuuru 'e-cfon) , asleep (cfaan-ii) ' . Regarding
this historical hypothesis, see McIntosh (1980) regarding the
morphology of the Stative and Progressive, see §2.2.2.4. below.
 85

2.1.2.4. Non-Verbal Sentences: 'Rag-bag 1

There are two particles in Fulfulde that are probably of verbal origin,
but that do not inflect, and whose use in effect produces sentences
without a verbal complex, ie. non-verbal sentences. There is
also another particle, h i n , which has no obvious verbal connections,
and which is used as a complete sentence in itself

The first mentioned above are raa/ndaa a n d 'igga. Raa/Ndaa is

presumably cognate with raarugo 'to see'. Although the initial-
consonant alternation between r_- and nd- suggests a singular/plural
subject correlation, raa and ndaa are used interchangeably, regardless
of the number of the subject. They are very similar to French voi^i.
Here are some examples of their usage:

72. Raa pade maacfa. 73. Ndaa cTe.

see shoes your see them
'Here are your (sg) shoes.' 'Here they (eg. shoes) are.'

'Igga means 'It would be b e t t e r — '. It may be followed either

by an NP (cf. 74.a. and b.} , or by a sentence with a Subjunctive-verbal
17
complex (cf. 75.a. and b.). '

74.a. 'Igga nditton. 7 4 .b. 'Igga jaango.

better that better tomorrow

'That one over there 'Tomorrow would be b e t t e r . 1

would be better.'

75. a. 'Igga njah-0 - aa. 75.b. 'Igga Hass an joocf-oo.

better go you better Hassan sit
VAP VAP
'You (eg.) had better go.' 'Hassan h a d better sit down.'

Although 'igga does not inflect for voice, aspect, polarity or other
verbal categories, it may be followed by the Anteriority Marker ('ANT'),
as in 7 5 . c .:

75.c. 'Igga - no njah-0 - aa.

better go you
ANT VAP
'It would have been better if you (sg.) had gone.'

The Anteriority Marker is usually appended to verbal roots, and its

occurrence with 1igga suggest that the latter is of verbal origin.
 86

Hin is used to mean 'Here you are', 'Here it is' or 'Take

it Cie. what I'm'offering to you) ', but does not combine with an
NP or any other item.

2.1.2.5. Non-Verbal Sentences; Conclusions

On the basis of their interrogative equivalents, I have separated

attributional and identificational sentences - the former having
an interrogative equivalent with noy? 'how', the latter with an
interrogative class-pronoun. On the basis of their negative equiva­
lents, however, these types of sentences are similar, since they
both employ n a a . More research is needed in order to establish
whether the attributional/identificational distinction is syntacti­
cally useful. Locational and existential sentences, on the other hand,
are syntactially clearly distinguishable from attributional and identi­
ficational sentences: for they (locational and existential sentences)
both use J_e_ and cfon, and both have negative equivalents with walaa.
It may be that this two-way distinction within non-verbal sentences
is more useful than the three-way distinction that has been described
here. (For the moment, I am ignoring the 'rag bag' non-verbal sent­
ences .)

2.1.3. The Fulfulde Sentence: Conclusions

The ultimate purpose of this section on the Fulfulde sentence has been
to identify verbal complexes, and their occurrence in sentences. Let
us therefore now consider more precisely what characterises a verbal
complex.

2.2.0. The Verbal Complex

We have already noted various characteristics of verbal complexes

(cf. §2.1.). These include the following:
 87

(i) They consist of a verbal root plus a VAP-suffix,

and sometimes other appendages too;

(ii) The initial consonant of the verbal root, if it is

an alternating one, is sensitive to the number of
a preceding subject, and to the position of the
subject.

In this section, I shall define the verbal root and the verbal complex,
and elaborate on some of the characteristics mentioned above, in
preparation for the following three chapters, which deal with the
internal syntax and semantics of the verbal complex, its morphology,
and with types of exponence of verbal properties (such as active voice,
completive aspect, etc.).

2.2.1. The Verbal Root

Verbal roots can be distinguished from nominal stems b y the range

of inflections they can take. Verbal roots, whether part of a verbal
complex or of a deverbal nominal, are usually followed by a suffix
18
expressing voice and aspect; whereas nominal stems are followed
by a nominal-class suffix. This does not mean to say, however, that
a stem cannot become a root, or vice-versa. In fact, some nominal
stems can be 'converted' into verbal roots by the addition of the
-cf-/-icf- suffix; and verbal roots are commonly 'converted 1 into nominal
stems b y the addition of a VA -suffix plus nominal-class suffix. In
this section, I shall explore some of these facts.

Most verbal roots combine with a range of VAP suffixes each of

which expresses a group of properties that characterise a particular
conjugation. These groupings will be the subject of the next chapter.
For the moment, let us consider, by way of example, the VAP-suffix
-ii in 71.a. which is repeated here for convenience;

71.a. Musuuru 1 e-cfon-cfaan- i i .

cat sleep
VAP
'The cat's asleep.'

This -ii VAP-suffix contrasts with -oo:

 88

71.b. Musuuru *e-cfon-daan - oo.

cat sleep
VAP
'The cat is falling asleep . 1

Sentences 71.a. and b. form a minimal pair, since they are in every
respect identical except for the “ii and -oo VAP-suffixes: the former
expresses middle voice, completive aspect, and positive polarity;
whereas the latter expresses middle voice, incompletive aspect, and
positive polarity. They differ, then, with respect to aspect. This
same verbal root, cfaan- can also take other VAP-suffixes, such as
-ake and - aaki in 76.a. and b . :

7 6 .a. Hawwa cfaan - a k e. b. Hawwa cfaan - aaki.

Hawwa sleep Hawwa sleep
VAP VAP
'Hawwa's gone to sleep.' 'Hawwa's not gone to sleep.'

Although both these suffixes, -ake and -aaki, express middle voice

and completive aspect, they differ with respect to polarity: -ake

19
xs positive, b u t - aaki is negative.

As well as a range of VAP-suffixes, a verbal root can also take

one or more verbal extensions. These in some sense-.modify the
meaning of the root, creating what we can call an 'extended r o o t 1.
An extended root has the same inflectional potential as an unextended
root, and will be treated indiscriminately in the next two chapters.
However, at this point, it is worth giving a couple of examples of
extensions, since they are a characteristic of verbal roots, but

not of nominal stems. They will also be further discussed in §2.2.2.2.

of this chapter.

Consider, for example, sentences 77.a., b. and c . . In 77.a.,

we see a verbal root, hokk- in its 'unexteneded', or simple, form.
In 77.b. the extension -it has been added, and in 77.c., the extension
-ir:

77. a. Mi-hokk-0 - e secfcfa?

I give you some
VAP

'May/shall I give you some ? 1

 89

77.b. Mi-hokk-it - 0 - e secfcfa?

I give you some
EXT VAP
'May/shall I give you some back?'

c. Non kokk-ir - 0 - cfaa-mi-nga.

thus give you me it
EXT VAP
"That's how you gave it to me.'

The effect of -it in 77.b. is to add the meaning 'back' to 'give1.

The -ir- extension in 77.c. has a less definable meaning: it
always occurs with a verbal root in a complex, if the complex is
preceded (and sometimes if it is followed) by an expression denoting
time, manner or place. Non expresses manner,, and -ir- therefore occurs.

Another characteristic of verbal roots is the fact that parti­

ciples can be formed from them. Participles consist of a root plus
VA-suffix plus nominal class-suffix, and were exemplified by 52.a.
and 53.a. on page 75 above. Here is a further example:

78. Ndu cfaan - ii-ndu.

it sleep
VA CL
'It's (one, eg. a cat, that has) gone to sleep.'

Here, the VA-suffix, - i i - , expresses middle voice and completive

aspect. The participle would be incomplete without the class-suffix,
-n d u . This is a suffix of the prenasalised consonant grade, ndu
class.

Nominal stems, and their inflectional morphology, have been

amply discussed in the preceding chapter. Here, I shall simple
highlight some respects in which they obviously differ from verbal
roots.

Nominal stems must be followed by a nominal-class suffix. For

example, musuu- is the stem meaning 'cat', and -ru is the continuant
 90

grade class-suffix for the Ndu class: musuu- could not occur alone.
Similarly, the stem bun- 'well' could not occur without a class-
suffix, such as -ndu or -cfi. These are the prenasalised grade
suffixes of the Ndu and Pi nominal classes. Thus: bun-ndu 'well1,
bun-cfi 'wells’.

Unlike verbal roots, nominal stems may not, normally, be followed

by a VAP-suffix. For example, if the VAP-suffixes -ake and -aaki
(seen in 76.a. and b.) were added to bun- nonsense words would result:
*bun-ake, *bun-aaki. However, there are certain stems and roots that
are identical (cf. footnote 3, chapter 1 ); and some nominal stems may
be converted into verbal-roots by means of the verb-forming suffix,
20
-icf-/~cf- . By way of example, consider the stem yam- 'healthy'.
This can be made verbal b y the addition, of -cf-, thus jam- in 79.a.
is a nominal stem; b u t yam-cf- in 80. a. is a verbal root:

79.a. Ngal jam - al.

it healthy
CL
'It (.eg. chicken) is healthy.*

80. a. Ngal yam - cf-ay.

it healthy
VAP

'It (eg. chicken) will recover/become healthy.'

The - al suffix in 79.a. is the zero grade suffix for the Ngal nominal
class. The -cf- suffix in 80.a. is the verb-forming suffix mentioned
above; and once this has been added to y a m - , then the composite yam-cf-
becomes a verbal root, thus capable of taking a VAP-suffix such as
-ay. The fact that a VAP-suffix such as - ay cannot be suffixed to
a nominal stem (cf. 79.b.), and the fact that a nominal-class suffix
such as - al cannot be suffixed to a verbal root (cf. 80.b.), are
both facets of the inflectional distinction between nominal stems
and verbal roots.

79.b. *Ngal yam - ay.

it healthy
VAP
 91

80.b. *Ngal jam - cf-al.

it healthy
CL

There is also, of course, a morpho-phonological distinction

between verbal roots and nominal stems. The nominal stem j a m - 'in
79. a. has an initial j_- because the Ngal class recmi'res a consonant
of the plosive consonant series (cf. table XIV); the verbal root
yam-cT- in 80.a., on the other hand, has an initial y- because it is
preceded by its singular sub j ect, n g a l . Jam- and yam-cf- thus" respond to
different morpho-phonological rules.

Now that we have established the identity of verbal roots,

which constitute the core of verbal complexes, let us consider other
characteristics of verbal complexes.

2.2.2. Other Appendages in Verbal Complexes

Whereas every verbal complex necessarily contains a verbal root and a

VAP-suffix, the occurrence of other appendages in the complex is not
obligatory, and is determined b y semantic and syntactic factors. In
this section, I shall identify ahd discuss these other appendages,
which are: subject and object pronouns? extensions; the Anteriority
Marker; and j_e_ and cfon.

2.2.2.1. Subject and Object Pronouns

If a subject or object NP is pronominal!sed, and is not focussed, then

the resultant pronoun is appended to the verbal root. For example:

81.a. Ceede 6 eyd - ake.

money increase
VAP
'The money/price has increased.'

21
b. Tie -6eyd - ake.
they increase
VAP
'It has increased.'

c. "Kanje beyd - i.
they increase
VAP
'That's what has increased.1/'It's that that has increased.'
 92

Sectence 81.a. consists of a subject NP, and a verbal complex composed

of a verbal root (£eyd-) and a VAP-suffix (-ake). In 81.b., however,
the subject is an unfocussed pronoun, cfe. Such a pronoun cannot stand
alone, as answer to a question, for example (see £2,3,3. below); and
it is therefore 'appended' to the verbal root. In §4.3.2. I in fact
argue that class-pronouns like cfe— are clitics within the verbal complex.
In 81.c., the subject is pronominal., but focussed. Rather than cfe-,
kanje is therefore used. Kanje is an anaphoric pronoun of the de_ class,
as is cfe; but whereas cfe is a dependent pronoun, kanje is independent,
for it can stand alone, asanswer to a question. I argue below that
the independent pronouns are NP's (see §2.3. ).The use of the -i_
VAP suffix in place of - ake is caused by the change of conjugation
('tense') which is necessitated by the presence of a focussed NP in
the sentence.

The sentences of 82. illustrate the pronominalisation of an object

NP:

82.a. Mi-£il - ii Hassan.

I trouble Hassan
VAP
'I troubled Hassan.'/'I was a nuisance to Hassan.*

b. Mi- 6 il - ii -mo.
I trouble him
VAP

'I troubled him.'/'I was a nuisance to him.'

c. "Kanko £il - u - mi.

him trouble I
VAP

'It's him I troubled/was a nuisance to.'

In 82., the same relationship holds between Hassan, -mo, and k a n k o , as

holds between ceede, cfe and kanje in 81., except that Hassan, -mo and
kanko are each the direct oject of their respective sentences. In 82.a.,
Hassan is the NP object; in 82.b . , -mo is a dependent pronoun object -
like cfe- it cannot stand alone, and is therefore part of the verbal
complex; and in 82.c . , kanko is an independent pronoun object, which
is focussed and therefore fronted. Because of the focussing of the
object, the conjugation of the verbal complex in 82.c. is different
from that in 82.a. and b., hence the change from mi- 6 il-ii to £il-u-mi.
 93

However, in 82., the verbal complex does not only consist of the
verbal root, VAP-suffix, plus or minus an object pronoun. The subject
pronoun is appended to the root, and thus part of the complex too. In
fact, 1 st and 2 nd person pronouns are necessarily part of a complex,
unless they are focussed - this is because, unlike 3rd person pronouns,
they are not anaphoric, and do not refer back to an NP. Thus, the
only way in which the subject pronoun in 82. could not be part of the
complex would be if it were focussed, as in 82.d . :

82.d. "Minn 6 il - i Hassan.

me trouble
VAP
'It's me who troubled Hassan/was a nuisance to Hassan.'
1 1^ troubled Hassan/was a nuisance to Hassan.*

Minn is the independent first person pronoun, and functions as an NP:

see §2.3. below.

2.2.2.2 Extensions

Verbal extensions, mentioned on page 88 above, may occur between the

'simple', or unextended root, and the VAP-suffix. They constitute part
of the derivational, rather than the inflectional, morphology of the
language, and a derived root has the same inflectional potential as a
simple root, syntactic and semantic compatibility allowing. An
extension may affect the valency or voice of a root: for example, it
may render an intransitive root transitive, it may allow a root to
take two objects rather than simply one, or it may render a normally
active root middle. In this section, I shall review a few of these
extensions, in order to show that they must be considered part of the
verbal complex in which they are found.

Some extensions combine with a simple verbal root to produce a

root with a different, though related, meaning. For instance, -t-/-it~
can contribute reversive meaning, as can be seen by comparing the a. and
b. sentences of 83. and 84.. (The occurrence of -i- in this, and other,
extensions, is phonologically determined.)

83. a. 'O-war-ii b. 'O-war- t_ - ii.

he come he come back
VAP EXT VAP
'He came . 1 'He came back.'
 94

84.a. Laahooru cfomcf - ii. b. Laahooru cfomcf - it - ii.

dog become dog become ’back*
thirsty thirsty
VAP EXT VAP
’The dog has become 'The dog has quenched its
thirsty . 1 thirst.'

Others have prepositional or case-marking functions: -r-/-ir- accompanies

certain adverbial case-functions (cf. 77.c. on page >89!above and
85.b. below); and -an- accompanies a dative object (cf. 8 6 .b. below):

85.a. 'Be - ka 6 6 - i i - m o . b. *Be - kafifo-ir - ii - mo froggol

they tie him they tie with him rope
VAP EXT VAP
"They tied him up.' ’They tied him up with a rope.

8 6 .a . ’O-sood-ii - n d e . b. 'O-sood-an - ii - mo - n d e .
he buy it he buy for him it
VAP EXT VAP
'He bought it (eg. donkey).' 'He bought it for him.'

Others include, for instance, -indir- indicating reciprocity, and -n-/-in-

with causative meaning:

87.a. '0 - wall-i-yam.

he help me
VAP
'He helped me.'

b. Min-mb all-indir-i i .
we help
EXT VAP
'We helped each other.'

8 8 .a. 'O - jal - ii.

he laugh
VAP

'He laughed.‘

b. 1O - j al - n - ii - rao.
he laugh him
EXT VAP
'He made him laugh.'

On the whole, the same extensions exist in the Fulfulde of the

Kaceccere'en as are listed in chapters 57 - 59 of A m o t t (1970) for
Gombe Fulfulde. The only noticeable difference is in the form and
meaning of Arnott's 'celerative1. In the dialect of the Kaceccere'en,
 95

'quickly' is usually indicated by an adverb, yawke-yawke; and the

extension that is presumably cognate with Arnott's Celerative ‘ means
'early' or 'a long time a g o 1;' it has the form -iloy- before i_ and
consonants, and -ilow-before other vowels:

89.a. ' O - w a r - i i yawke-yawke.

he come quickly
VAP
'He came quickly.'

b. *0 - w a r - iloy-i i .
he come
EXT VAP
'He came early/a long time ago.'

c. '0 - war - iloy-taa.

he come
EXT VAP
'He won't come early.'

d. 10 - war - ilow-aayi.
he come
EXT VAP
'He d i d n 't come early.'

Here are examples of several extensions co-occurring with the same

verbal root; the b. sentences are ungrammatical, because the order
of the extensions is wrong:

90.a. 'O - war - t - iloy-i i .

he come
EXT EXT VAP
'He came back early.'

b. * 10 - war - iloy-t - ii.

91.a. 'O - d i l l - i d - an - ii-ma 'e maakoyij.

he go+ for you with them
away
EXT EXT VAP
'He went away with them for you (sg).'

b. * 'O - dill-an-id-ii-ma e maakoyn.

92.a. 10 - j i p p - i n - a n - i- yam donngal am.

he put+ for me load my
down
EXT EXT VAP
'He put down my load for m e . '
 96

92.b. * 10 - jipp-an-in-i-yam donngal am.

Hie order of co-occurring extensions is as given b y Arnott (1970:

§57.2. and §59.4.).

It is perhaps worth noting in passing that an extended root has not

only the same inflectional potential, but also the same derivational
potential as does a simple root. Thus, the extended root war-t-iloy-,
seen in 90.a. above, may form the basis of a participle:

93. Gar - t - iloy - 0 - cfo.

come back early
EXT EXT VA CL
'Person who has come back early.'

Here, the nominal-class suffix, -do, has been added to the extended
verbal root, gar-t-iloy-, just as the pronoun 'o- and the VAP suffix
-ii have been added to the same root in 90.a.. (The w-/g- alternation
is of course determined by the subject in 90.a . , and the class-suffix
in 93«)

2.2.2.3. The Anteriority Marker

The 'anteriority marker' will be discussed in greater detail in the

next chapters. Essentially, it sets the action depicted b y the verbal
complex one step back in tin®, using the moment of utterance or a
time determined within the discourse as its point of reference. Arnott
(1970) calls it the 'preterite particle',* its form is no or n o o ,
depending on the particular constellation of verbal properties that
occur within the verbal complex. Here are some examples of its
occurrence:

94. a. '0 - h u w - a y .
he work
VAP
'He works/will w o r k . '

b. '0 - huw - ay - n o .
he work
VAP ANT
'He used to w o r k . 1

95. a. 'O - suucf-ake-ma.

he hide you
VAP
'He hid/has hidden (himself) from you (sg).'
 97

95.b. *0 - suucT-ake-no - ma.

he hide you
VAP ANT
'He had hidden (himself) from you.'

96.a. Ngim maajum koll - u - mi- 6e - cfum.

because that show X them it
VAP
'That's why I showed it to them . 1

b. Ngim maajum koll-u - noo-mi- 6 e - cfum.

because that show I them it
VAP ANT
1T h a t 1s why I had shown it to them . 1

From the b. sentences of 94. - 96., we see that no follows immediately

after the VAP-suffix, but that it precedes pronouns within the complex.
Thus, if dependent pronouns (such, as -mi-, - 6 e- and -cfum in 96.b.)
are morphologically part of the verbal complex, no must be too.

2.2.2.4. 'E and Don

We saw in §2.1.2.3. above that these two 'particles' occur in sentences

expressing existence and/or location; and that they also occur in
Stative and Progressive complexes (cf. 70. and 71.a. on page 84.). Here
are some further examples of such complexes:

9 7. a- 'E-mo-cfon- joocf-ii. b. 1E-mo-cfon- joocf-oo.

he sit he sit
VAP VAP
'He is seated.' ‘He is sitting down.'

'E and cfon do not occur in complexes of any other conjugation, and they
often morphologically distinguish, the Stative from the Relative
Completive, and the Progressive from the Subjunctive, since these paiirs
22
of conjugations have the same 'basic' VAP-suffix (see §4.2.).

Whether occurring in an existential/locative sentence (cf. 58.

and 59. a. and d . ), or in a Stative or Progressive verbal complex,
1e precedes a dependent pronoun subject, but fallows an NP subject,
as can be seen in 98.a. and b.:

98.a. Hawwa 'e-cfon-ronnd-ii.

Hawwa carry
VAP

'Hawwa is carrying a head-load.'

 98

98.b . 'E-mo - cfon-ronnd-ii

she carry
VAP
S h e ’s carrying a head-load.'

The reverse order is unacceptable:

99.a. * 'E-Hawwa-cfon-ronnd-ii.

b . *Mo - 1e-cfon-ronnd-ii.

The presence of _|_e in a Stative or Progressive complex is not

obligatory. Thus, the b. sentences below, where _|_e_ does not occur,
are just as acceptable as the a. sentences, where does occur:

100.a . *E-mi-cfon-sood-a nagge.

I buy cow
VAP
'I’m buying a cow.'

b. Mi-cfon-sood-a nagge.
I buy cow
VAP
'I’m buying a cow.'

101.a. Muusa 'e-cfon-janng — a to Kano.

Musa study in Kano
VAP
'Musa is studying in Kano.'

b. Muusa cfon-janng-a to Kano.

Musa study in Kano
VAP
'Musa is studying in Kano.'

However, _^_e never occurs with 1en (.'you and I') or 'on ('you', pi.);
and in the pronoun 1aa ('you', sg.), it seems likely that ^e_ and 'a
have coalesced, for this would explain the long vowel of this pronoun,
which only occurs in Stative and Progressive complexes. Here are some
examples: the b. sentences are unacceptable, because co-occurs
with 1e n , 'on or 'aa:

102. a. 1En-cfon-n jah-a. b. *'E- 'en-cfon-njah-a.

we go we go
(incl) VAP
V
'W e 're g o i n g .'
 99

103. a. 'On-cfon-moft-oo. b . * *E-'on-cfon-mof t-oo.

you meet you meet
(pi) VAP

'You (pi) are gathering together.'

104.a. 'Aa-cfon-furt - in-*- a gite.

you protrude eyes
EXT VAP
'You're making your eyes 'pop
o u t '/protrude.'

1 04.b. * ’E - ’aa-dbn-furt-in-a gite.

you prot- eyes
•rude

It was noted in section §2.1.2.3. above that cCon looks like a

locative adverb. In its use in Stative and Progressive complexes,
cTon has no overtly locative meaning. However, interestingly enough,
it can be replaced b y cToo 'here' , ton ' Cover) there* , or nii 'just,

simply'. The latter necessarily interrupt the verbal complex in

which they occur, and they add a meaning that is not present with
23
cfon. Unlike don, they can be placed in an alternative position
in the sentence; thus both the a. and b. sentences of 106. - 108.
are acceptable:

105.a. 'E-mo-don-def -a nyiiri. b. *'E-mo-def-a nyiiri cfon.

she cook food
VAP
'She's cooking food . 1

106.a. 'E-6 e - cfoo-taar-a mbabba b. 'E-joe - taar-a mbabba cfoo.

they here tie donkey they tie donkey here
VAP VAP
'They're here, tying up the 'They're tying up the
donkey.' donkey h e r e .'

107.a. 'E-mi-nii - hecfit - oo. 'E-mi-hecfi t-oo n i i .

I just listen I listen just
VAP VAP
'I'm just listening.' 'I'm just listening.'

108.a. 'E-mo-ton - dur - a na'i. b. 'E-mo-dur - a na'i ton.

he there graze cows he graze cows there
VAP VAP
'He's over there grazing 'He's grazing cows over
cows.' there.'
 100

Sentence 105.b. is unacceptable because cfon cannot be placed outside

the verbal complex.

It is in fact possible for cfon and nii to co-occur within the

same complex; however, cfon cannot co-occur with ton or cfoo. Since
the latter are locative in meaning, whereas nii is not, this fact
suggests that, despite its lack of overt locative meaning, cfon does
still retain locative connotations, which disallow its co-occurrence
with locative adverbs. Sentences 105.c. - 108.c. illustrate these
co-occurrence possibilities:

105. c . ' E-mo - cfon-nii - def - a nyiiri.

she just cook food.
VAP
'She's just cooking food.'

106.c. *'E-&e - cfon-cfoo- taar-a mbabba.

they here tie donkey

107. c. 'E-mi-cfon-nii - hecfit - oo.

I just listen
VAP
'I'm just listening.'

108.c. * 1E-mo-cfon-ton - dur - a na'i

he there graze cows

One informant in fact compared sentence 106.c. to saying 'utmost b e s t 1

in English: b y implication, he felt that cfon cfoo was a repetition of
a locative idea, just as 'utmost best' in English would involve a
repetition of the superlative idea.

The sentences of 105. - 108. show us that cfon is different from

cfoo, nii and t o n , and thus confirm the status of cfon as part of
Stative and Progressive verbal complexes. Nevertheless, sentences
such as those of 105. - 108. are also exceptional, because they have
a deictic adverb within a verbal complex. This interruption of the
verbal complex is only possible with Stative and Progressive complexes.

For instance, if one were to insert cfoo or ton or nii between a

subject pronoun and the verbal root in a complex of any other conjugation,
an ungrammatical sentence would result. This is demonstrated in
sentences 1 0 9 .b. - lll.b.. The a. sentences are acceptable because
the deicitic adverb occurs outside the verbal complex. \
/
 101

109.a. *Be - njoocf-ake cfoo.

they sit here
VAP

"They sat down here . 1

b. * "Be-dbo-n joocf-ake.
VAP

HO.a. Min-kawt-ii ton.

we meet there
VAP
'We met (over) there . 1

b . *Min-ton-kawt-ii.
VAP

111.a. *Be - 6 ai]i)-ii nii.

they move thus/
house simply
VAP
'They simply moved.'

b. *Be - nil-bann-ii.
VAP

From the discussion in this section, we conclude that _^e_ and

cfon are dependent particles that contribute towards the marking of
stativity and progressivity, in thejpresence of suitable VAP-suffixes
in the same verbal complex. Moreover, we have seen that Stative and
Progressive verbal complexes are the only ones that allow any deictic
adverbs to interrupt the verbal complex. The Stative and Progressive
are therefore in some way 'special'; this impression is confirmed
by the fact that no other conjugations have any sort of marker (other

than subject pronoun) occurring before the verbal root.

2.2.3. Intonation

In listening to verbal complexes, it is tempting to search for a stressed

syllable, which can then be neatly marked In the transcription. In
the event, such a search proves difficult, and, I suspect, futile. For
the fact is that since vowel length Is phonemic in Fulfulde, a long-
vowelled syllable often sounds 'stressed' to the ear - or at least
to the English ear. Yet closer examination proves that there is always
one, or more than one, high-pitched syllable in the complex, which often
does not sound 'stressed* at all. It is this feature of high pitch, and
 102

its placing in the complex, that I want to discuss here.

This line of investigations is at a very preliminary stage,

and what is important is the fact that patterns are being found,
rather than the precise nature of those patterns. Taking into
account such a proviso, a fairly cursory consideration of the data
available suggests that in verbal complexes that are not in relative
clauses or in yes/no questions, high pitch generally falls on the
penultimate syllable of the complex. One thus finds sentences such
as the following, where the syllable bearing high pitch varies
according to the number of appendages in the complex (I have marked
the relevant syllable by preceding inverted commas,"):

112.a. 'Aa-don-hel"l-a-mmi.
you slap me
VAP

'You (sg) are slapping m e . r

b. *Aa-dbn-,,hell“a minyaa.
you slap your younger sibling
VAP
'You Csg) are slapping your younger sibling.'

113.a. *Igga mbal"l" 0 - a a - m i n .

better help you us
VAP
'You Csg) had better help us (exel) .'

b. 'Igga mball-an - 0 - aa-"min - Be.

better help for you us them
EXT VAP
'You (.sg) h a d better help them on our (excl) behalf.'

In sentence 112.a., the high pitch occurs on the penultimate syllable

of the complex, which is l a m ; and in 112.b . , since the NP minyaa
does not belong to the verbal complex, and since nothing follows the
conjugation-suffix, the high pitch is one syllable earlier than in
112.a., namely, it occurs on the hel syllable. In sentence 113.a.,
the high pitch again occurs on the penultimate syllable, laa; and
in 113.b., it occurs on the following syllable, m i n , since a
suffixed pronoun, - 6 e, follows min.

A variation of the 'penultimate high pitch' rule sometimes occurs.

When the penultimate syllable has CV structure, then the preceding CV
 103

syllable can also take high pitch. One thus finds, for example:

114. a. M o -yaa"f- et-e.

he forgive you
VAP
'He'll forgive you (sg).'

b. Mo-yaa"f - o"to-'en.
he forgive us
VAP
'He'll forgive us (incl).'

In 114.a., the penultimate syllable bears high pitch, as usual; in

sentence 1 1 4 .b., however, both the penultimate syllable, t o , and the
preceding syllable fo, bear high pitch. This does not happen if
either the penultimate or the pnceding syllable has C W or CVC
structure:

115.a. Kanjum wad*-i njanng - a"t-aa - mi footo.

that cause take you me photo
VAP VAP
'That's why you (sg) will take my photo.'

b.*Kanjum wacf - i njan"ng-a"t-aa - mi footo

116.a. Ndee ngubit-otoo-"mi-'on sauru?

when throw I you stick
VAP
'When shall I throw you (pi) the/a stick?*

b. *Ndee ngubit-o"too“ "mo-' on sauru?

Sentence 1 1 5 .b. is unacceptable, because the penultimate syllable

in the verbal complex has C W structure, so the preceding syllable
cannot also bear high pitch; and sentence 116.b. is unacceptable,
because the syllable before the penultimate has C W structure, and
so cannot bear high pitch.

The above variation could be reformulated to read 'Two penultimate

light syllables, or one penultimate heavy syllable, may recieve high
pitch in the verbal complex'. Another variation of the 'penultimate
high pitch' rule is that generally speaking, a syllable containing
an anaptyctic u (cf. §4.2.1.) may not take high pitch. Verbal
complexes whose penultimate syllable does contain such a ia have
high pitch on the second syllable before last. For example:
 104

117. a. ,E-mo-cfon"suucf-u-maa pacfe.

he hide you shoes
VAP

1H e ’s hiding (your) shoes from you (sg).*

b. * 1E-mo-cfon-suu',d-u-maa pacfe.

In 117.a . , the verbal complex has an anaptyctic -u- between the verbal
root and the suffixed pronoun -maa; the high pitch therefore falls
on the suu syllable, not on the cfu syllable (as in 117.b.).

In relative clauses, it is the final syllable, rather than the

penultimate one, which generally receives high pitch. For example:

118. 'O-yaaf - ataa"ko-mo boofi cfi 'o-wacf-an -i-"mo.

he forgive her mistakes which she do to him
VAP EXT VAP
'He won't forgive her the mistakes she made against h i m . 1

119. Ko njah - u - noo-"mi ngii-0 - raoo-mi.

when go I see him I
VAP ANT VAP
' (It's) when I had gone (that) I saw him.'

In sentence 118., the verbal complex 'o-wacf-an-i-mo has high pitch on

the final syllable mo, rather than on the penultimate syllable, n i ;
however, the complex which is not in a relative clause, 1o-yaaf-ataako-mo,

has the high pitch on the 'normal', penultimate syllable, ie. k o .

The first clause in sentence 119. is a relative clause introduced by
ko 'when', and the hi^h pitch falls here on the final syllable (mi)
of the verbal complex; in the second clause, the high pitch falls
on thejpenultimate syllable (moo) of the complex, since the clause is
not a relative one ( - the relative conjugation here is necessitated
by the focussing of the first clause).

It cannot be emphasised too much that the generalisations being

made here concerning the occurrence of high pitch are very tentative;
and it must be borne in mind that many of the sentences quoted (eg.
112. - 117., but not 118. and 119.) were elicited in an 'interview
situation', rather than recorded in a 'natural1 speech-situation.
They may therefore be unrepresentative. What is important, however,
is the fact that even in citation forms, patterns emerge; and the
existence of such patterns lends weight to the argument that the verbal
 105

complex is a single unit in terms of phonological analysis.

Further lines of investigation that could be followed in this

connection include the relationship of high-pitch in the verbal
complex to the general intonation-contours of Fulfulde sentences;
and the possible correlation between the occurrence of high pitch
and certain types of syllables (eg. light and heavy syllables).

2.2.4, The Verbal Complex, as a. morphological word

The preceding discussion has pointed to the fact that on the

phonological level, the verbal complex is a unit - let us call it a
'phonological w o r d ' . There is also considerable evidence, in fact,
for considereing the verbal complex as a 'morphological word' too.

The morphological evidence will be more fully developed in the

next two chapters. Essentially, it consists of the fact that there
is morphological interdependence between the component morphs of a
complex. For example, the person and/or number of the subject may
determine the shape of other morphs, as in 118..:

118.a. Ndee bad - i. - cfaa Kano?

when approach you Kano
VAP

'When did you (sg) approach Kano?'

b. Ndee bad - ii-mi Kano?

when approach I Kano
VAP
'When did I approach. Kano?'

Here, the length of the vowel of the VAP-suffix is determined by the

following subject pronoun. Similarly, the conjugation of the complex
may influence the shape of pronouns, as in 119.:

119. a. 'Aa-cfon-hoot-a na?

you go
home
VAP QU

'Are you (sg) going home?'

b. _VA-hoot-ii na?
you go
home
VAP QU
'Did you (sg) go home?'
 106
119. c . Ngim cfume koot-u - cfaa?
for what go you
home
VAP
'Why did you return hom? '

Here, the form of the 2sg subject is determined b y the conjugation in which
it occurs: the complex in 119.a. is of the Progressive, in 119.b. of the
General Completive, and in 119.c. of the Relative Completive.

Lyons (1968: 202-4), in discussing the word as a morphological unit,

mentions two criteria: internal stability, and positional mobility. The
Fulfulde verbal complex is internally stable to the extent that its form
is pre-determined, given a particular set of morpho-syntactic properties
to be realised (cf. chapter 5, which, deals with exponence). Since
Fulfulde is a language with fairly fixed order, it is difficult to dis­
cuss the positional mobility of the complex. The only constituents of
a sentence that do have a degree of positional mobility are in fact
unfocussed adverbial phrases. In.relation to these, one can show the
fact that a verbal complex moves as a unit. For example:

120.a. Fa£ 6 ijaango ’e n - ’umm-oto.

d a y after .we arise
tomorrow
VAP
"The day after tomorrow we (incl) shall arise (and
depart). 1

b. 1E n - 1umm-oto fa£ 6 i jaango.

we arise day after
tomorrow
VAP
'We (incl) shall arise (and depart) the day after
tomorrow . 1

It would be impossible for the adverbial phrase, fa££ijaango to inter­

rupt the verbal complex, 1e n - 1umm-oto:

120. c. * ’En-fal5&ijaango-* umm-oto.

The case for treating appendages before the verbal root as part
of the ’verbal complex word* is much weaker,both on phonological and
morphological grounds, than the case for treating appendages after the
root as part of such a ‘w o r d 1. Given the essential, minimal, ‘root +
VAP~suffix‘ morphological structure of the complex, and its essential
 107

'final or penultimate high-pitch' prosodic structure, there is no

way of testing the occurrence of high pitch on items that precede
the root. Moreover, the fact that n i i , cfoo and ton can intervene
between the subject pronoun and the verbal root in Stative and
Progressive complexes (cf. 105. - 108.), weakens the argument for
considering such complexes as one morphological word. Nevertheless,
it is descriptively convenient to treat the Stative and Progressive
as inflectional paradigms, and I shall not investigate the issue
of the word-status of the verbal complex any further.

Another issue is the syntactic status of the verbal complex. In

many senses, it is the 'verbal counterpart' of the NP, which will be
discussed in the next section. Nevertheless, I hesitate to call it
a VP. This question will have to be left for future research.

2.3.0. The Noun Phrase

While the term 'nominal' was discussed and provisionally defined

in the last chapter, the term 'NP' has been used so far without
definition. Moreover, it is useful to identify NP's, in order to °
clearly separate them from verbal complexes, and in order to establish
the essential difference between the type of pronouns that may occur
in verbal complexes ('dependent pronouns'), and the types of pronouns
that may not (cf. the list of types of pronouns given in §1.3.3.).

'Nominal' is a morphologically-defined category. 'NP', however,

is syntactically defined. I shall use three criteria to establish
the identity of NP's. These are the ability of a constituent to
stand as:

topic
focus
answer to a question.

I shall consider these three criteria in § 2.3 .1 . - §2.3.3. below,

and particularly with reference to different types of pronouns.
Then in §2.3.4., I shall mention other types of NP.

2.3.1. Topic

In this and the next section, I anticipate the establishment of

categories that are more fully discussed in chapter 6 .
 108

A topic is a constitutent in sentence-initial position/ which

the rest of the .sentence in some sense comments on. Examples are
nagge ngee and keddiraawo 'am in these two sentences:

121.a. Nagge ngee, nge-rim - ii.

cow this it give
birth.
TOPIC VAP
fThis cow, it's ^iven birth.'

122.a. Keddiraawo 'am, 'e-mo-wood-i 'encfam.

neighbour my she has kindness
T O P I C VAP

' (As for) my neighbour, she's kind.'

The commas in the transcription indicate that these constituents

are potentially followed b y pause.

Let us take the 'slot' filled by nagge ngee and keddiraawo 'am,

and see what other items may occur in their place. First, let us
test the Various types of class-pronouns mentioned in §1.3.3..
We find that if we place a dependent class-pronoun in this position,
unacceptable sentences result:

121.b. *N g e , nge-rim - ii. 122.b. *'0, 'e-mo-wood-i 'encfam.

it it give she she has kindness
birth

On the other hand, if we take the other types of class-pronouns, we

find that they may all occur in this slot. (I take ngee and 'oo as
representative of all the deictic pronouns.)

121.c. N g e e , nge-rim - ii.

this it give
one birth
TOPIC VAP
'This one (cow), it's given birth.'

N g e y e , nge-rim - ii?
which it give
birth
TOPIC VAP
'Which one (cow) has given birth?'
 109

121.e. Kann g e , nge-rim - ii.

the one it give
mentioned birth
TOPIC VAP

’The one (cow) mentioned, it's given birth.'

1 2 2 .c. 'Oo, 'e-mo-wood-i 'encfam.

this she has kindness
one
TOPIC VAP
’This one (.person) , she's kind.'

d. M o y e , 'e-mo-wood-i 'encfam?
who she has kindness
TOPIC VAP
'Who is kind?'

e. Kanko, 'e-mo-wood-i 'encfam.

the person she has kindness
mentioned
TOPIC VAP
"The one (person) mentioned, she's kind.'

I shall consider those pronouns which may fill the nagge ngee and
keddiraawo *am slots to be NP's. Thus, comparison of the b. sentences
and the c. - e. sentences above suggests that while dependent pronouns
are not NP's, deictic, interrogative, and independent pronouns are.

The above are all third person pronouns referring to particular

nominal classes. There are only two types of 1st and 2nd person
pronouns: dependent and independent. As might be expected, dep­
endent 1 st and 2 nd person pronouns may not be topic of a sentence,
but independent ones may be. For example:

123.a. *Mi, to London ngon-0-mi.

I in London live I

b. Minn,to London ngon-0-mi.

me in London live I
TOPIC VAP
1 (As for) me, I live in London . 1

124.a. * 'En, 'en Kaceccere 1en.

we we Kaceccere'en
 110

124.b. ‘E n e n , 'en Kaceccere'en.

us we Kaceccere1en
TOPIC
'(As for) us (incl), w e ’re Kaceccere'e n . 1

123.a. and 124.a. have a dependent pronoun (m i , 'en) in topic position,

and are unacceptable; 123.b. and 124.b., on the other hand, have
their independent pronoun equivalents (minn, 1enen), and are acceptable.

2.3.2. Focus

The focus of a sentence is a constituent that represents new, or high­

lighted information, and which often, though not necessarily, receives
24
contrastive stress. Examples are Dere and Ngaari in 125.a. and 126.a..
(The syllable bearing contrastive stress is preceded by inverted commas.)

125-a. "Dere te' - ay-mi. 126.a. "Ngaari sood-u - cfaa?

Dere marry I bull buy you
FOCUS VAP FOCUS VAP

'I'll marry D e r e .' 'Did you (sg) buy a b u l l ? '

Both Dere and Ngaari contrast with other items that implicity (cf. 125.a.
and 126.a.) or explicitly (cf. 125.b. and 126.b.) might have filled
the same 'slot* :

125.b. "Dere te 1 - ay-mi, naa Facfa.

Dere marry I not Facfa
FOCUS VAP
'I’ll marry D e r e , not Facfa.'

126.b. "Ngaari sood-u - cfaa, naa nagge?

bull buy you not cow
FOCUS VAP
'Did you buy a b u l l / not a cow?'

Let us do now as we did in the discussion of topic: test which

other types of 'item' may fill the Dere and ngaari slots. In fact,
we find the same as in 1 2 1 . and 1 2 2 .: a dependent pronoun may not
occur (hence the unacceptability of the c. sentences below); but
deictic, interrogative, and independent pronouns may occur (hence the
acceptability of the d . , e. and f. sentences below). (This time,
I use 'otton and nditton as representative of all deictic pronouns.)
 Ill

125.c. *"Mo te' - ay-mi. 126.c. *"NcLL sood-u-cfaa.

him marry I it buy you (sg)

d. 1Ot"ton te' - ay-mi. Ndif'ton sood-u-cfaa.

that marry I that buy you (sg)
one one
FOCUS VAP FOCUS VAP

'(It's) that (person) I'll '(It's) that (eg. bull)

marry.' y o u '11 b u y . '

125.e. "Moye te' - ay- mi? 126.e. "Ndiye sood-u-cfaa?

whom marry I which buy you (sg)
FOCUS VAP FOCUS VAP

'Who is it I'll marry?' 'Which is it you bought?'

f. __
"Kankot e ' - ay-mi. "Kayri sood-u-cfaa.
him/her marry I the one buy you (sg)
mentioned mentioned
FOCUS VAP FOCUS VAP
'(It's) him/her I'll marry.' '(It's) the one mentioned
(eg. bull) I bought.'

With 1st and 2nd person pronouns, we find the same as before: the
independent pronouns may be focus of a sentence (cf. 127.b. and 128.b.);
but the dependent pronouns may not (cf. 127.a. and 128.a.):

127.a. *"Mi winnd-ata. 128.a. *"' 0 n Be - mball-i.

I write you they help
(pi)

b. "Minn winnd-ata. b. " ‘Onon Be - mball-i.

me write you they help
FOCUS VAP FOCUS VAP
'It's me who'll write.' 'It's you (pi) they
helped.'

2.3.3. Answer to a Question

In this section, I shall consider one-word, non-verbal, answers to

'wh-questions'. These are necessarily incomplete sentences, though
they constitute complete utterances (cf. Lyons 1977: §1.6.). For
example, if a question such as 129.a. is posed, what sort of one-
word response may be given?

129. a. *Dume ngii-0 - cfaa?

what see you
VAP
'What did you (sg) see?'
 112

The answer may be a complete sentence in itself, such as 129.b . ;

or one word, such as 129.c.:

129.b. Puccu ngii-0 - mi. 129.c. Puccu.

horse see I horse
VAP
'I saw a h o r s e . ' 'A hors e . '

Let us now investigate what items may occur in the puccu 'slot'
in 129.c., on the same lines as above. Again, we find that a
dependent pronoun, such as ngu in 130. a., is unacceptable, but
that a deictic or independent pronoun (such as in 130.b. - c.) is
acceptable:

130.a. *Ngu. 130.b. Nguya.

that one
'That one (eg. horse ) . 1

c. Kanngu.
that one
mentioned
'The one (eg. horse) that
we've been discussing.'

(It is difficult to test interrogative pronouns in this 'slot',

since they are semantically unsuitable to be given as answer to a
question.)

Yet again, the same observations may be made concerning 1st and 2nd
person dependent and independent pronouns: the former may not stand
alone in answer to a question (,cf. 131.b. and 132.b . ) , but the latter
may tcf. 131.c. and 132.c.):

131.a. Moye won-i ndottiijo maacfa? b. **A. c. 1Aan

who is husband your you you
VAP
'You (sg).'
'Who is your husband?'

132. a. Moye yan - i? b. *Mi. c. Minn.

who fall I me
VAP
1M e . '
'Who fell?'
 113

2.3.4. Other types of NP

From the evidence of examples 121. - 131., I shall consider that

nominals, deictic pronouns, interrogative pronouns, and independent
pronouns are NP's, but that dependent pronouns are not.

However, the syntactic category of NP is by no means limited to

forms such as the above. Xn this section, I shall mention two other
25
types of NP: relative clauses, and genitival complexes. I shall
take the same types of 'slots' as in §2.3.1. - §2.3.3. above, and
supply relative clauses and genitival complexes to 'fill' them.
Thus, in the a. sentences below, we see a relative clause as topic;
and in the b. sentences, a relative clause as focus. Cln the
sentences with focus, I have marked one of several possible syllables
as having contrastive stress. For example, in 1 3 3 .b., the contr­
astive stress could occur on soo- instead of on w a a - .)

26
133.a. Nge sood-u - mi rowaani, nge-rim - ii.
which buy I last year it give
VAP birth
T 0 P I C VAP
"Ihe one (cow) i bought last year, it's given birth.'

26 27
b. Nge sood-u - mi rol,waani rim - i
which buy I last year give
VAP birth
F O C U S VAP
* (It’s) the one I bought last year (that) has given birth.'

26
134. a. Ko min-nan-i, cfum-woocf-aa.
what we hear it good not
VAP VAP
T O P I C
'What we've heard is not good.'

b. K o ^ min-"nan-i woocf-aa.
what we hear good not
VAP VAP
F O C U S
1 (It's) what we've heard (that) is not good.'

Similarly, the a. and b. sentences of 135. - 136. contain genitival

complexes as topic and focus respectively;
 135.a. Nge maacfa, nge-rim - ii.
it your it give
birth
T O P I C VAP
'Yours (ie. your cow) has given b i r t h . 1

2.7
b. Nge "maacfa rim — i ' , naa nge 'am.
it your give not it me
birth
F O C U S VAP
' (It's) yours (ie. your ccw) (that) has given birth
not mine.

136.a. Mo naane, 'innde maako 0 Yaa'i.

he former days name his Ya'i
T O P I C V
'He of former days (eg. the former Emir) , his name
was Ya* i . '

Mo "nanne mbolw-an-0 - mi.

him former days speak to I
F O C U S EXT VAP
' (It's) to the former one (eg. Emir) (that) I spoke

Finally, in the d. and e. sentences below, we see that relative

clauses and genitival complexes may stand as answer to the questions
posed in the c. sentences;

133.c. Ngeye rim - i?

which give
birth
VAP
'Which (cow) has given b i r t h ? 1'

d. Nge sood-u - mi rowaani.

which buy I last year
VAP
'The one I bought last year.'

e. Nge maacfa.
it your (sg)
'Yours (sg.).'

134.c. Dume wel - i - 'on?

what sweet you (pi)
VAP
'What pleases you/is sweet to your ears?'
 115

134.d. Ko min-nan-ii.
what we hear
VAP
'What we've heard.'

135.c. Moy 6 ur - i woocf-ugo?

who exceed good
VAP INF
'Who's better?'

d. Mo naane.
he former days
"Ihe former one (eg. Emir).'

According to all three of our criteria, then, relative clauses

and genitival complexes are, syntactically, NP's.

2.3.5. The NP: Conclusions

In this section, we have established that the following are syntacti­

cally N P 's :

deictic pronouns
interrogative pronouns
independent pronouns
relative clauses
genitival complexes.

There are, of course, other sorts of NP, such as nominals. However,

the structure of the NP is not so important to us at present as is
the structure of the verbal complex. We shall therefore leave the
more detailed study of Fulfulde NP's to a future date.

2.4. The Fulfulde Sentence, Verbal Complex and NP: Conclusions

In this chapter, we have identified verbal and non-verbal sentences,

the former containing a verbal complex, the latter not. We have
also shown some characteristics of verbal complexes, and identified
some types of NP. Let us now turn to the more detailed analysis
of the verbal complex: first, its syntax and semantics, and then its
morphology.
 116

Chapter 2: Footnotes

1. Hyphens are used to link the constituent morphs of verbal complexes

and participles.

2. I use 'appendage 1 as a cover-all term for suffixes and clitics that

occur with the verbal root in a verbal complex. The morphological
status of these various appendages is discussed in chapter 4 .

3. Although I link a dependent pronominal subject to the verbal complex

with which it occurs, I do not so link a dependent pronominal subject
to the nominal with, which it occurs. Thus, X have placed no hyphen
between koyg and ndane-hoyn in 51. a. , or between j_0 and boow-u-cfo
in 52. a.. Nor have I linked boow-u-cfo and its object pronoun, cfum,
b y a hyphen. This is because I do not want to pre-judge these
morphological issues at this point: the morphological status of
verbal complexes, and to a lesser extent, of participles, are what
concern us here.

4. In order to exemplify what is unequivocally an identificational

sentence, I give this example of an equative sentence, where the two
NP's could occur in reverse order, without changing the 'basic' mean­
ing. (Intonation and stress could change its information structure
what is new, and what is given.) In fact, there is evidence to
suggest that the first NP in such non-verbal sentences has topic,
rather than subject, status. This evidence includes the fact that
pause and/or what I call a ftopic marker 1 may occur between the two
NP's (cf. ch. 6 ); and the fact that the negative counterpart of such
sentences has naa between the two NP's (cf. 54.c., p. 79), not in
sentence-initial position. This matter is peripheral to our central
concern here - the identification of sentences with or without a
verbal complex - and will therefore not be pursued further.

5. Since the verbal complex that has been deleted contains the subject
as well as the verbal root (cf. 62.a.) , I have glossed 0 as *VS',
not only as ' V .

6. A comparable distinction is found in Hausa, and in several other

African languages. It is discussed, for example, in Schachter (.1973) .

7. Taa is reduced from to + 1a , where to is the negating particle, as in

66 .b. , and Jja. is the 2sg subject pronoun. In Kaceccereere, to 1a-hu£ 6 -u
hiite is a less common alternative to 65.b . .

8. Woni is based on the root won-, which means 'live', in the locative,
rather than the existential, sense. As 'live', won- takes a normal
range of inflections; but as 'be*, it only occurs as won-i (with
singular subject), and ngoni (with, plural subject); and as won-do/
ngon-do 'used to be', where the -do is a variant of the Anteriority
Marker, n o .

9. There is no obvious way of testing the length of this vowel, for it

is never followed by a dependent pronoun and is therefore usually
in word-final position. (Concerning the word-status of verbal complexes,
see §2.2.4. below. Concerning word-final vowels, see §1.1.5.1. above.)
However, I mark the vowel as short, because when won-i is followed by
no, the -i is dropped. If the -i were long, one would expect it to
 117

be reduced to a short vowel, but not to disappear altogether. I

make this judgement in anticipation of the discussion in §4 .3 .4 ..

10. The precise difference in meaning between a sentence such as this,

and a sentence with a Stative verbal complex:
'E-mo-woow-i - cfum.
he get
used
VAP
’He's used to it.'

is not clear. The use of brackets din the translation of 52.a. is

an attempt to convey the fact that 52.a. contains a participle.

11. The non-verbal nature of 52.a. and 53.a. is further confirmed by

the fact that a verbal complex with, woni can be added to these
sentences (cf. 52.b. and 53.b . w h i c h - is not possible with sent­
ences already containing a complex, such as 49.a. and b . .

12. Cf. footnote 4 above, concerning the possible topic status of the
first NP in 54.a. and c..

13. The fact that the -i_- of wonicfon is short is possible corroborating
evidence for considering the -i_ of woni to be short too: c f .. the
discussion in footnote 9. above.

14. Here again (cf. footnotes 1. and 3. above), the problem arises of
whether these morphs (.'e, n g e l , don and cfoo) should be linked by
hyphens; and yet again, I do not confront the issue. For the moment,
I do link J_e and cfon to each other and/or to other parts of a verbal
complex, when they constitute part of a Progressive or Stative complex
(cf. 70. and 71.a./ p.84)* but I do not so link them in existential
or locative sentences.

15. The -aa of walaa is suspiciously like the VAP-suffix of the Negative
Completive Active (cf. table XXIl). However, the 1root* wal- occurs
with no other inflections, and I therefore treat it as an abnormal,
defective verb.

16. There is another particle o f verbal origin, h a a n i , which is similar

in usage to ’igga, but has the meaning 'ought'. However, haani does
sometimes take a subject pronoun, and is therefore less qualified
than 'igga to be mentioned in this section.

17. If 1igga is a defective verbal form (as is suggested by its ability

to take n o , cf. 75.c . ) , then it may be that we should analyse
sentences such as 74.a. and 75.a. as:
74.a. S 75.a. S

0 1igga nditton 0 'igga njah-aa

 118

Since it can be followed by an NP that does not govern S, it is

unlikely that 1igga is 'simply 1 a sentence-initial modality-marker.
If the above analysis is correct, then some doubt is cast on the
idea tht sentences with. *igga are non-verbal.

18. Amongst deverbal nominals, participles always contain a VA-suffix;

but abstract nominals, such as gar-ol ‘coming, arrival', dem-al
'farming', and laam-u 'reign1, 'leadership', contain only a root-
become-stem, plus a class-suffix.

19. Although voice, aspect and polarity are expressed in the VAP-suffix,
other distinctions, such as Stative vs. Non-stative, are made in
other ways, such as b y the presence or absence of ^e_ and cfon in
the complex.

20. The occurrence of -i_-in -id-/-cf- is phonologic ally determined by

the shape of the stem to which it is added.

21. Ceede literally means 'cowrie-shells', and is plural.

22. With a 3rd person subject, _^e_ and/or cfon do crucially distinguish
these pairs of conjugations; however, with a 1 st or 2 nd person
subject, morph-order also comes into play.

23. This sentence assumes that cfon is within the verbal complex, as does
my use of hyphens in Stative and Progressive complexes. The inclusion
of cfon (and b y extension, _|_e which precedes it) within the verbal
complex might be disputed, especially given the fact that cfon
can be replaced b y cfoo, ton and n i i ; this matter will be taken up
again in §4.4.2. 3. . One of the reasons for including cfon and _£e_
in Stative and Progressive complexes is the fact that in other
dialects, their respective cognates have coalesced with other
parts of the complex (of. Arnott 1970: 195, footnote 3).

24. Recent research (e.g. Dik et al, 1980) distinguishes different types
of focus. For present purposes, I am simply concerned with an NP or
adverb that is fronted, and that potentially bears contrastive stress.

25. By 'genitival complex', I mean a constituent consisting of a head

plus genitival modifier. When the head is a nominal, such as
keddiraawo in 122.a . , then the NP status of the whole constituent
is fairly obvious. However, when the head is a dependent pronoun,
such as nge in 135. or mo in 136., then the NP status of the whole
is less obvious, and is therefore discussed in this section.

26. For the moment, the issue of whether these relative clauses are
'headless' or not does not concern us.

27. When in a sentence with a focussed NP, this complex is necessarily

of a Relative (rather than General) conjugation Csee §6 .2.1.1.),
and therefore has a short-vowelled VAP-suffix - contrast 133. a.
and 135 .a. .
 119

Chapter 3

Paradigms of Verbal Inflections:

Syntax and Semantics

3.0. Introduction

Fulfulde is an inflecting language, with an intricate verbal and nominal

morphology. The nominal morphology has already been briefly described
in §1.3., and is only of further interest in this study inasmuch as it
affects participles and dependent class-pronouns. The verbal morphology,
on the other hand, will now be treated in some detail. The aim of this,
chapter is to present the syntactic and semantic grounds for the estab­
lishment of inflectional paradigms based on verbal roots. The morphology
of these paradigms is explained in the next chapter.

I shall first mention certain derivational processes that operate

on verbal roots C§3.1.) , in order to distinguish them from inflectional
processes t§3.2.). Then, in §3.3., the difference between verbal cate­
gories and verbal properties is established. I go on, in §3.4., to
Identify those properties which, in various combinations, establish the
synatactic and/or semantic distinction between the fourteen inflectional
paradigms. In §3.5., those properties are discussed which may vary within
an inflectional paradigm, and which do not crucially distinguish one
paradigm from another. Finally, in §3.6. - §3.9., I review the various
paradigms, the properties they realise, and how they differ in usage and
meaning.

3.1. Derivational Processes

It is useful to distinguish two types of derivation process (cf. Robins

1964: 258): ’class-maintaining 1 and ’class-changing' processes. The
former effect no change on the syntactic potential of the root or stem
on which they operate, whereas the latter do. As regards verbal roots,
verbal extensions fall into the former category, whereas participles and
other deverbal nominals fall into the latter.
 120

3.1.1. Extensions

As was observed in §2.2.2.2., verbal extensions effect multifarious

modifications on verbal roots. What concerns us here is the fact

that extended verbal roots have the same potential for inflection
as do unextended roots. This can be exemplified b y comparison of
the a. and b. sentences below: the a. sentences contain an unextended
(or 'simple') root, war- 'come', whereas the b. sentences contain an
extended root war-t- 'come back.':

140.a . '0-war— i i . b. 10 -war - t_ - ii *

he come he come .back
VAP EXT VAP
'He came . 1 'He came back.'

1
141.a. Ngim cfume 'o-war-i ? b. Ngim cfume 'o-war - t_ - i ?
for what he come for what he com back
VAP EXT VAP
'Why did he come ? 1 'Why did he come back?'

142.a. Daga tokoye ngar-u-cfaa? b. Daga tokoye ngar-_t - u-cfaa?

from where come you from where come back you
EXT VAP
'Where did you (sg) come 'Where did you (sg) come
from?' back from?'

143.a . Ngol-war-aayi, b. Ngol-war - t_ - aayi.

it come not it come back not
VAP EXT VAP
'It (eg. letter) didn't 'It (eg. letter) didn't
arrive. come b a c k .'

We see that, apart from the presence or absence of the extension -t - ,

the verbal complex in each pair of sentences is identical.

The verbal extensions in fact form part of the syntagm of units in

the sentence; and, unlike the inflectional suffixes, they are not mutually
exclusive. Another extension can, for example, be added to the already-
extended roots in the b. sentences of 140.-143., thus:

140. c. 'O-war-t. — iloy-ii .

he come back
EXT EXT VAP
'He came back early/a long time ago.'
 121

141. c. Ngim cCume 1o-war - t^ - iloy-i?

for what he come back
EXT EXT VAP
‘Why did he come back early/a long time ago?

142.c. Daga tokoye ngar-t_ - iloy-0-cfaa?

from where come back you
EXT EXT VAP
'Where did you (sg) come back from Cso) early/a
long time ago? *

143.c. Ngol-war - _t - ilow-aayi.

it come back not
EXT EXT VAP
'It (eg. letter) didn't come back early/a long time ago.*

In these c. sentences, we see that two extensions co-occur in the same

complex. (The alternation between -iloy- and -ilow- is phonologically
determined: - iloy- before consonants and — i_ or -ii; - ilow— elsewhere.)
From such evidence, then, we know that the extensions are not in paradigm­
atic contrast with each other. I shall therefore establish the inflect­
ional paradigms of Fulfulde without regard to the extensions, since

extended roots equally share in the inflectional processes that I

describe below.

3.1.2. Nominals

Nominals can be derived from verbal roots by the addition of a nominal-

class-suffix. For example, the deverbal nominal gar-ol 1coming/arrival *
is derived from the verbal root war- ’come' via the nominal-class-
suffix - o l . Similarly, 1annd-al 'knowledge1, for example, is derived
from 'annd- 'know', via the nominal-class-suffix - a l .

Participles are derived from a verbal root via a nominal-class-suffix

plus an additional, intervening, suffix. The latter suffix expresses pro­
perties of voice and aspect. Thus, in joocf-ii-6e ‘settled (people)', -ii-
expresses middle voice and completive aspect, while —be is a nominal-
class-suffix. Likewise, in war-ay-nde 'coming/next1, -ay- expresses
active voice and incompletive aspect, while nde is a nominal-class-suffix.

By virtue of their two suffixes, participles straddle the nominal

and verbal systems of the language. The nominal-class-suffix is a
 122

member of a paradigm of such suffixes, which, as we saw in §1-3.1.,

characterise nominals. The intervening suffix, on the other hand,
characterises verbal roots, and is homonymous with the VAP-suffix
of verbal complexes. For example, -ii suffix characterises complexes
of the Relative Completive Middle and of the Stative Middle, as
well as Completive Middle participles; and -ay suffix characterises
complexes of the General Incompletive Active, as well as Incompletive
Active participles.

3.2. Inflectional Processes

Unlike the derivational morphology just described, inflectional

morphology determines the relationship between a verbal root and other
items in the syntagm in which the root occurs. It is the identification
of paradigms of verbal inflections that will be the concern of the rest
of this chapter, after we have defined more precisely what properties
may be expressed in these inflectional paradigms.

3.2.1. 14 Inflectional Paradigms: 14 'Conjugations'

Taking morphological, syntactic and semantic factors into consideration,

fourteen paradigms of verbal inflections are identifiable in Kaceccereere
Fulfulde. Each paradigm represents the intersection of particular
morpho-syntactic properties. I shall refer to the various paradigms as
follows:

General Completive
Relative Completive
Stative
Negative Completive
Emphatic Completive
General Incompletive
Relative Incompletive
Progressive
Negative Incompletive
Vague Incompletive
Subjunctive
Imperative
Continuative Imperative
Exhortative

Each of these names includes the minimal properties whereby the relevant
paradigm can be identified. Traditionalists would in fact call each of
these paradigms a ’tense’ (cf. Arnott 1970: 179); and in some ways, it
would be convenient to continue with this term, perhaps modifying its
meaning according to need. However, the factors that crucially distinguish
 123

one paradigm from another are not 'tense' (with its implications of
time), but aspect/ mood/ polarity and stativity, for example. I
should therefore prefer to avoid the term 1tenser.

This poses a problem/ which could be solved by simply continuing

to use 'paradigm'. Yet this term is surely too useful for descriptions
of nominal morphology/ as well as of verbal morphology/ to be limited
in this way. I therefore resolve the issue by making rather idiosyno—
ratio use of 'conjugation'. M y use of this term is substantially
different from its application to Latin verbs/ where a 'conjugation'
represents an arbitrary verb-class. There are no such morphologically-
defined verb-classes in Fulfulde. The term 'conjugation* is therefore
available; and I shall henceforth use it in this study/ without inverted
commas, to refer to an inflectional paradigm of verbal forms.

A verbal complex is any one member of a conjugation. The complexes

of a particular conjugation express a particular group of verbal prop-
2
erties. For example, all and only the complexes of the Negative In-
'completive conjugation express the combination of negative polarity,
incompletive aspect, and indicative mood. Any one complex of that
conjugation (and indeed of any conjugation) must also be marked for
voice; and may additionally be marked for anteriority, person, number
and case.

There is rarely a one-to-one correlation between verbal properties

and morphs. (Verbal properties are, of course, abstract, and might have
been termed 'morphemes'. I use 'morph' concretely.) A morph may express
several properties (as does a VAP-suffix, for example); and a single
property may be expressed over several mor p h s , Ihese matters will be
the concern of chapter 5.

3.3. Verbal Properties and Verbal Categories

In considering the various conjugations, their syntactic and semantic

functions, and their morphology, I shall use the general approach of
Matthews (1972 and 1974) and of Carstairs (1981). I shall identify
certain verbal 'properties', some of which (such as active voice) can
be seen as terms in larger, verbal 'categories' (such as voice). The
properties in any one category are mutually exclusive. For example,
 124

active voice and middle voice cannot co-occur in the same complex.
However/ the properties of different categories do co—occur: for
example/ active voice and positive polarity.

In the following sections/ I shall first identify 'conjugational

properties 1 w h i c h i n : various .combinations crucially distinguish one
conjugation from another; and I shall then identify those properties
that may or may not be expressed in any one complex of a conjugation/
and which are not crucial to the identification of the conjugation.

3.4. Conjugational Properties

The following properties come under this bracket:

Indicative and Non-Indicative

Completive and Incompletive
Positive and Negative
'Relative' and 'Non-Relative* (or'General1]
Emphatic and Non-Emphatic
Stative and Progressive
Imperative
Exhortative
Subjunctive
'Vague'.

Those properties mentioned in the first three lines are terms in the
categories of mood,, aspect and polarity, respectively and will be
treated in some detail here. The other properties cannot be so easily
pigeon-holed in terms of categories; and they will be more fully
treated in the discussion of the various conjugations C3.6. - 3.9.
below).

3.4.1. Indicative and Non-indicative Mood

The properties indicative and non-indicative are not formally realised

in the inflectional morphology of Fulfulde; and there is even some
doubt as to whether a certain conjugation, the Subjunctive, realises
indicative or non-indicative (see belo w ) . Nevertheless, it is useful
to establish the existence of these properties, both for semantic and
other reasons.

Complexes of conjugations that express indicative mood generally

express statements or questions; whereas those of non-indicative mood
 125

generally express commands, exhortations, obligations, requests

for permission, or the granting of permission - what Lyons (1977:

745ff) calls 'mands'. Consider the sentences of 144. and 145., for
example. The former contain complexes of indicative mood, whereas the
latter contain complexes of non-'indicative mood:

144.a. Faatu natt-ay suudu mum. 145.a. Faatu natt-a suudu mum!
Fatu enter hut her Fatu enter hut her
VAP VAP
'Fatu enters/will enter 'Fatu should enter her
her h u t . 1 hut!1

b. Be - soor-ii tew n g u u . b. Soor— u tew nguuI

they sell meat this sell meat this
VAP VAP
'They sold this meat.' 'Sell (sg) this meat!'

c. 'A - huw - ay jemma. c . Huw - atay!

you work night work
VAP VAP
'You work/will work at 'Keep on (sg) working!’
night.'

d. Na'i 'am cficfi majj-ii. d. Na'i maacfa majj-ai

cows my two lose cows your lose
VAP VAP
'Two of my cows have got ‘May your cows get lost*.
lost.'

e. 'O-wii nge-tuf - aama. e. lO-wii nge-tuf - ee.

(<wi'-ii) (<wi'-ii)
he say it vaccinate he say it vaccinate
VAP VAP VAP VAP
'He said it (cow) was vaccin­ 'He said it (cow) should
ated. ' be vaccinated.'

Sentence 144.a. expresses a statement of what Fatu does habitually, or

is going to do; whereas 145.a. expresses a command. A similar contrast
between statement and command holds between the b., c. and e. sentences.
While 144.d. is also a statement, 145.d. expresses a wish, almost a curse.

Apart from the 'statement vs. m a n d 1 contrast, two other features

clearly distinguish the two moods. Firstly, indicative can combine with
anteriority (see §3.5.2. below), whereas non-indicative cannot. Thus,
the sentences of 146. (which are the same as those of 144., except for
the addition of no, the marker of anteriority) are acceptable, whereas
 126

those of 147. are not:

146.a. Faatu natt-ay- no suudu mum.

Fatu enter hut her
VAP ANT
fFatu was going to/used to enter her h u t . 1

147.a. *Faatu natt-a-no suudu mum!

ANT

1 4 6 .h. Be - soor-ii - no_ tew nguu.

they sell meat this
VAP ANT
'They h a d sold this meat.'

147.b. *Soor-u-no tew nguu.

ANT

146.c. 'A - huw - ay— no jemma.

you work night
VAP ANT
'You (sg] used to work at night . 1

14 7.c . *Huw-atay- n o .
ANT

146.d. Na'i. 'am cficTi majj-ii-no.

cows my two lose
VAP ANT
'Two of my cows had got lost.'

147.d. * N a ’i maacfa majj-a-no.

ANT

3
146.e. 'O-wii nge-tuf - aa-no - ma .
(<wi'-ii) vacc-
he say it inate
VAP VAP ANT VAP
'He said it had been vaccinated.'

14 7. e. *'O-wii-nge-tuf-ee-no_.
ANT

In most instances of non-indicative mood, its incompatibility with

anteriority can probably be explained in terms of the fact that commands,
requests, exhortations, etc., are necessarily interpreted as referring
to the moment of utterance, and cannot be 'pushed back a step in time*.
 127

Secondly/ while both positive and negative polarity can combine

with indicative mood in a verbal complex, this is not true of non-
indicative mood. A sentence containing a non—indicative complex
is negated by means of a sentence—initial particle, to^, which in
the Imperative Singular is realised as taa. We can see these two
types of negation in the sentences of 148. and 149. respectively:

148.a. Faatu natt-ataa suudu mum.

Fatu enter not hut her
VAP
'Fatu doesn't/won't enter her h u t . 1

b. B e - soor-aayi tew ngua.

they sell not meat this
VAP
"They didn't sell this meat.'

c. 'A - h u w - a taa jemma.

you .work not night
VAP
'You don't/won't work at night.'

d. Na'i 'am cCicfi majj-aayi.

cows m y two lose not
VAP
'Two of my cows haven't got lost.1

e. '0 —1
w ii nge-tuf - aaka.
C<wi'-ii) vacc-
he say it inate not
VAP VAP
'He said it had not been vaccinated.'

149.a. To Faatu natt-a suudu mumI

not Fatu enter hut her
NEG VAP
'Fatu should not enter her hutI1

b. Taa soor-u tew nguu!

not sell meat this
NEG VAP
'Don't sell this meat!'

c. Taa huw-atayi
not work
NEG VAP
'Don't keep on working!1
 128

149.d. To^ na'i maacfa majj-a!

not cows your lose
NEG VAP

'May your cows not get lost!

e. 'O-wii to nagge ngee tuf - ee.

fie say not cow this vacc­
inate
NEG VAP
'He said this cow should not be vaccinated./

In the sentences of 148., negative is expressed within the verbal

complex, by means of the VAP— suffix, which isdifferent from the
VAP-suffix in the comparable positive complexes in 144.. In the sent-
ences of 149., however, negative is not expressed within the verbal
complex at all - in fact, the verbal complexes have the same VAP-
suffix astheir positive counterparts in 145.. Rather, theproperty
negative is expressed by the to_ or taa at the beginning of each sent­
ence of 149..

Although the distinction between indicative and non—indicative

looks neat and tidy as presented in 144. - 149., the picture is not
in fact quite so perfect; for the Subjunctive conjugation sometimes
realises indicative, somtimes non-indicative mood. This fact seems
to be a function of the relationship between incompletive aspect,
non-indicative mood, and non-factive modality (cf. Givon 1973). I
shall give here one example of the problem posed by the Subjunctive.
Consider the following;

150. Min-njonng-in - a ndiyam?

we boil water
EXT VAP
'May we boil the water?'

151. To min-ndef-ay sogora, sey min-kett-a doyaaje,

if we cook pounded then we peel yams
VAP yam VAP
min-njonng-in-a ndiyam,....
we boil water
EXT VAP
‘IfyVhen we cook pounded yam, Cthen) we peel yams,
boil water,....

In 150., the verbal complex, min-njonng—in— a is of the Subjunctive

conjugation, and expresses either a request for permission, or the
 129

granting of permission- This is tire 'typical' use of the Subjunctive,

and clearly, the Subjunctive is here expressing non-indicative mood.
In 151., we have the beginning of a recipe. It is Introduced by a
conditional clause C.To min-ndef— ay sogora] , followed by a series
of clauses containing a Subjunctive verbal complex (min-kett— a ....,
min— njonng-inw a ...] . Here, the Subjunctive is not expressing a
'mand'; rather, it is used in a series of clauses expressing in-
completive aspect. As we shall see in §3.7.2.1- and 53.8.3. below,
this use of the Subjunctive is comparable to the use of the Relative
Completive in a series of clauses expressing ^completive aspect; and
it is tempting to aay that it is here not expressing non-indicative
mood, but rather, indicative mood.

3.4.2. Completive and Incompletive Aspect

Each conjugation expresses one or the other aspect, as we shall see

in the discussion of the individual conjugations in §3.6. - §3 .9 .
below. A t this stage, it is perhaps appropriate to explain why I
establish the category of 'aspect' rather thafi 'tense*. (For discussion
of the theoretical issues involved in the aspect - tense distinction,
the reader is referred to Comrie 1976, and references therein. For
discussion of the formal differences between the two aspects the reader
is referred to sections §4.2. and §4.3. of the next chapter.) Let
us consider the meaning of sentence 144.a., which is repeated here
for convenience:

144.a. Faatu natt-ay suudu mum.

Fatu enter hut her
VAP
'Fatu enters/will enter her h u t . '

Here, natt-ay may express a habit (eg. 'enters regularly', 'always

enters1), or a future action ('will enter', 'is going to enter') -

in other words, it expresses an action that hasn't been completed. The

action may have started, and be continuing through the present into
the future; or it may not have started yet. The fact that we are
dealing with incompletive aspect, rather than future tense, is further
corroborated by the fact that a complex expressing incompletive (’future1)
may be used for a past habit. In such circumstances, the anteriority-
mark a r , no (cf. 146.a.— e.) explicitly tells us that the habit has
ceased. Consider 152.a.:
 130

152.a. Nde 'o—war - 0 - no fu, min—mhel—mbel— t - ay - n o .

when he come all we glad glad
VAP ANT EXT VAP ANT.
'Whenever he came, we rejoiced.'

Here, 1o-war-0-no is a complex of the Relative Completive Active, with

the anteriority-marker; and min-mbel-mbel-t-ay-no is a complex of

the General Incompletive Active, with an extension that involves

4
reduplication of the root (m bel-) , and with the anteriority-marker.
Sentence 152.a. would be correct, m y informants asserted, whether the
person who used to come were dead or alive at the time of utterance;
and the use of the completive rather than of the incompletive to
express the habitual rejoicing would be unacceptable:

152.b- *Nde 'o-war-0-no fu, min-mbel-mhel-t-ii.

Comrie (1976: 5) tells us that 'Aspect is... concerned with the

internal temporal constituency of ... one situation'. This statement
is surely borne out by the internal temporal constituency of sentence
152.a.: for the habitual coming expressed in the first, relative,
clause of the sentence is realised by a complex of Completive Aspect
('Past tense1) , because the coming was completed on every occasion
before the rejoicing started. The 'internal temporal constituency 1
of the coming in relation to the rejoicing thus determines the choice
of conjugation. It is the use of the Anteriority-marker, no, that
tells us that the habit of rejoicing (and indeed, of his coming) has
ceased, and thus relates the actions to the 'zero-point of utterance'
(cf. Lyons 1977: 682). (Concerning anteriority, see §3.5.2. below.)

The use of what I shall henceforth call completive aspect in

the first clause of 152.a. tells us that the action depicted by the
verbal complex, although habitual, was completed before the onset

of the action depicted by the next complex.. I should: now like to

give two further pieces of evidence for the fact that completive
and incompletive are aspects rather than tenses.

Firstly, the completive is used to express an action that is

completed in intention, though not in effect. If two people are
bargaining, the seller says Mi-soor-ii-ngel naira gootel: 'I'll sell
it for one naira', as he offers his 'last' (lowest) price for his
 131

goods. The buyer may not accept, the deal may never come off, but
the completive aspect (with -ii VAP-suffix) is used, rather than
the incompletive aspect (with -ay VAP-suffix) . Similarly, if a
person decides to go for a walk, he may call as he goes Mi-dill-ii
wane - ugo 'I'm going for a w a l k 1, using the completive, with -ii,
rather than the incompletive with- -ay.

Secondly, the completive aspect is used for performative verbs.

Here are some examples:

153. Mi-ha 6 6 -ii-ma duu 6 i tati.

I tie you years three
VAP
II (hereby) sentence you (sg) to three y e a r s 1
imprisonment . 1

154. Mi-nodd-an - ii binngel ngel Ahmadu.

I call for child this Ahmadu
VAP
'I (hereby) name this child Ahmadu.'

155 . Mi-wii - m a ngacf-0 - aa kuugal n g a l .

(<wi*-ii)
I say you do you work this
VAP VAP
'I (hereby) order you (sg) to do this work.'

156. Mi-wacf-ii 'alkawal mi-hokk~it - ay ceede den.

I make promise I give back money the
VAP EXT VAP
*1 (hereby) promise to give back the m o n e y . 1

157. Mi-yett — ii.

I thank
VAP
'Thank y o u . 1

As with the examples in the previous paragraph, the action in these

sentences is completed in intention, and the completive aspect is
therefore used: in fact, as the use of 'hereby1 in the English glosses
indicates, the intention is put into effect by the very utterance
itself - hence the term ‘performative verbs'.
 132

It must be pointed out that the non-tense nature of the completive

v s . incompletive distinction in Fulfulde was noticed as long
ago as 1931 b y Gaden:

Les formes du parfait ont pour caractere coimnun d'envisager

le proces sous son aspect acheve, tandis que celles de I'im-
parfait l'envisagent sous son aspect inacheve. L*opposition
n'est pas seulement dans le sens, elle est apparente dans
les desinences, justifiant ainsi la classification adoptee.
CGaden 1931: xii.)

Arnott similarly characterises his 'general past* and the other

'past tenses' as indicating 'completed action, or a completed process'.
He continues:

In dialogue it Cthe general past) very often refers to past

time in relation to the time of utterance, but it also
frequently refers to time that is past in relation to some
other time implied in the context or indicated by a time—
word, or to the time of some action indicated by other verbs.
(Arnott 1970: 262)

Referring to his 'general future/habitual tense*, Arnott says that it

would be possible to assume its two main meanings 'under one generalised
meaning of 'non-completeness ' 1 (Arnott 1970: 27b).

While Arnott continues to speak of 'past and future/habitual

tenses', in spite of his above observations, Gaden on the other hand
proceeds to speak of the 'parfait* and 'imparfait' as 'aspects'. Gaden
also points out the patterns of inflections that can be seen within the
two aspects: the characteristic i/ii/aa of the completive aspect, and
the a/oo/ee of the incompletive, which I shall be discussing in the
next chapter. It seems, then, that Gaden 'got there first'I

3.4.3. Positive and Negative Polarity

. 5
Positive and negative are properties in the category polarity. As was

observed in the discussion of indicative and non-indicative above,

negative is sometimes expressed within a verbal complex, by a VAP-suffix
such as the - ataa of 148.a. and c . ; but it is sometimes expressed by
a sentence-initial particle, such as the to and taa of 149..

There can be little doubt, in fact, that positive polarity is

morphologically unmarked, whereas negative polarity is marked: for
in the absence of a negative marker, positive polarity is assumed.
 133

This is evident not only in the fact that the VAP-suffixes for the negative
conjugations are morphologically composite (see §4.2. of chapter 4
below) ; but also in the fact that, in the use of sentence-initial
negative-markers, such as to^ and t a a , negative polarity is 'imposed'
on an otherwise positive sentence.

3-4.4. Relative and Non-Relative (.General)

These properties do not fall into a neatly-named category. They

refer to a syntactically-determined, morphologically-realised
distinction that is found in African languages of several families:
for example, Kanuri (of the Nilo-Saharan family) and Hausa £of Afro-
Asiatic) , as well as Fulfulde Cof Niger-Congo).

In certain types of sentence, the property ’relative* is obligatorily

realised. Such are: relative clauses (hence the term 'relative');
sentences with a focussed NP or adverb; wh— questions; and narrative
sequences. The conjugations that express relative are the Relative
Completive and Incompletive, and (in incompletive narrative sequences),
the Subjunctive. In other types-of sentence, the-property non-relative
is realised. The conjugations that express non-relative are the General
Completive, the Stative, the General Incompletive and the Progressive.
These facts will be illustrated in §3.6. - §3.7. below.

3.4.5. Emphatic and Non-Emphatic

The property emphatic is realised when completive aspect and verb-

focus coincide: this will be further discussed in §3 .7.1 .2 . below.
Emphatic is expressed in the Emphatic Completive conjugation: other
conjugations express non-emphatic by default.

3.4.6. Stative and Progressive

The properties stative and progressive are minimally distinguished in

their morphological realisation; they occur in very similar syntactic
contexts; and they are mutually exclusive, in the sense that they
may not co-occur in the same verbal complex. Yet they cannot be
claimed to be terms in a single category for they are not so much
in contrast with each other, as with, rather, their own absence- Thus,
stative contrasts with non-stative, when the properties completive and
 134

non-relative co-occur. In terras of conjugations, the Stative con­

jugation contrasts with the General Completive (cf. 158.a. and b.
respectively}, and the Progressive conjugation contrasts with the
General Incompletive Ccf. 159.a* and b. respectively}:

158.a. 'E—mo— 6 o r n - ii.. b. *0 - 6 orn - a k e .

he dress he dress
VAP VAP
'Ee is dressed/has got 'He got dressed/dressed
dressed.' himself . 1

159 .a . 1E—mo-£>orn - o o . b. ‘0 -fborn - o t o .

he dress he dress
VAP VAP
'He's getting dressed/ 'He'll get dressed/he does
dressing h imse l f. ' get dressed.'

Stative is expressed in 158.a., but not in 158.b.; and progressive

is expressed in 159.a., but not in 1 5 9 .b..

These properties, and the conjugations that bear their names, will
be further discussed in §3.6 .1.1. and §3.6 .2.1. below.

3.4.7. Imperative and Exhortative

The property imperative always co-incides with non-indicative mood,

and may co-occur both with completive and incompletive aspect. The
intersection of imperative and aompletive is expressed in the ’Ordinary
Imperative' conjugation, used for what we can call 'ordinary' commands:
whereas the intersection of imperative and incompletive is expressed
in the 'Continuative Imperative 1 conjugation, which is used for commands
to continue doing something. Whereas imperative always combines with
a 2nd person subject, the property exhortative, which is expressed
in the Exhortative conjugation, combines either with a 1st person plural
(inclusive) subject or with 1Allah 'God’.

3.4.8. Subjunctive

The property subjunctive, and the conjugation that bears its name,
have already been discussed in §3.4.1., with regard to their ambivalent
position vis-a-vis indicative and non-indicative mood. They will be
further discussed in §3.8.3. below.
 135

3.4.9. Vague

This property sometimes co— incides with, incompletive aspect, and

is expressed in the Vague Incompletive conjugation. See §3.6 .2.3.
below.

3.5. Properties that do not distinguish one conjugation from another

In this section, I shall first consider properties of voice, one of

which is necessarily expressed in every verbal complex of every
conjugation (§3.5.1.}. I shall then consider properties that are not
crucially distinctive, and that do not necessarily occur in a complex:
first, anteriority (§3.5.2.), and then the properties of person,
number, and case (§3.5.3.).

3.5.1. Active, Middle and Passive Voices

Each conjugation has a different VAP-suffix for each voice: in

fact, paradigms of complexes can be established within each con­
jugation, one paradigm, for each voice. The Imperative conjugations (both
'Ordinary* and Continuative} and the Exhortative only occur in the
active and middle, but all other conjugations occur in all three
voices.

Verbal roots can be classified according to which voice or voices

they occur in: some occur in all three; some in only two; and
others in only one. Each voice has certain semantic connotations,
not all of which are compatible with the semantics of every root.
Essentially, the active voice is associated either with transitivity,
where the action is performed by an agent (the subject} on another
participant (the object), or with intransitivity; the middle voice is
associated with reflexivity, and with borrowed verbal roots that
would be semantically more compatible with active voice; and the
passive voice is associated with actions performed on the subject, by
an unspecified agent.

Let us now consider some data that illustrate the most typical uses
of the three voices, and which also establish the integrity of the voice
paradigms. Examples of voice in the various conjugations will be given
incidentally in the ensuing discussion. However, here to begin with
 136

are examples of the three voices of the General Completive conj­

ugation, with the roots ronnd- 'put a head-load o n ’ and suud-
'h i d e 1 :

160.a. Mi-ronnd-ii - ma. 161.a. Mi - suucf - ii - ma.

I load you I hide you
VAP VAP
'I put a head-load on 'I hid you (sg) . 1
you (sgl.*

b. 1A - ronnd-a k e . b. *A - suua-a k e .
you load you hide
VAP VAP
'You (sg) put your head- 'You (sg) h i d (yourself).'
load on yourself . 1

c. 1A — ronnd-a a m a . a . 1A - suucf— aama.

you load you hide
VAP VAP
’You (sg) have had your . - ‘You (sg) have been
head-load put on ( y o u r s e l f ) h i d d e n . 1

The a. sentences consist of a verbal complex that depicts an action

performed by the subject on to the direct object. The b. sentences
depict a similar- action, but performed reflexively by the subject
on to himself/herself. (Note that there are no reflexive pronouns
in Pulfulde.) The c, sentences depict the result of such an action,
with the direct object of the a. sentences ‘promoted 1 to subject.
It is the - i i , - ake and -aama VAP-suffixes that typically distinguish
active, middle and passive voice respectively, in the General Compl­
etive conjugation.

3.5.1.1. Middle Voice: Reflexive and Ergative

While the active v s . middle voice distinction correlates to transitive

vs. reflexive meaning with some verbal roots, with other roots, the
relationship between active and middle voices is an ergative one:

in other words, the object of an active complex (e.g. ceede maaje in

162.a. below) is the subject of the equivalent middle complex (as is

ceede maaje in 162.b.),
 137

162.a . 1E—min-dbn— Beyd - a_ ceede m a a j e .

we increase money tiieir
VAP
PROGRESSIVE ACTIVE

'We’re increasing their p r i c e . 1

b. Ceede maaje 'e-cfon-6 eyd - o o .

money their increase
VAP
PROGRESSIVE MIDDLE
1Their price is going u p .'

c. Ceede maaje 'e-cfon-6 eyd - e e .

money their increase
VAP
PROGRESSIVE PASSIVE

'Their price is being increased.'

In the sentences of 162., we see the paradigm of active, middle and

passive Voices of the Progressive conjugation, with the a/oo/ee
VAP-suffixes. While the relationship between active and middle voice
meaning is different here from that which we saw in sentences 160. -
161., the relationship of the passive to the active is exactly the
same in all cases: the object of the transitive, active, complex
is 'promoted' to become subject of the passive complex.

On the whole, the verbal roots which have reflexive meaning in

the middle voice have an animate object in the active voice: cf. 160.a.
and b., and 161.a. and b.; whereas the roots which have an ergative
relationship between active and middle voices have an inanimate object
in the active voice: cf. 162.a. and b.. Certain conjugations of
the latter types of root have a slightly special meaning in the middle
voice: namely, they can express potentiality or impossibility. The
conjugations concerned are the General and Relative Incompletive, and
the Negative Incompletive. For example, if we change the conjugation
of 162.b. from the Progressive to the General Incompletive (whilst
retaining middle voice] we get the following:

162.d. Ceede maaje foeyd - oto.

money they increase
VAP
'Their price increases/will i n c r e a s e T h e i r
prices can be increased.'
 138

Hie negative equivalent of this sentence can express impossibility:

162.e. Ceede maaje 6 eyd - ataako.

money their increase not
VAP
'Their price doesn 1t/won *t increase.V ’Their
price can't be increased . 1

Certain types of verbs have two middle-voice equivalents to one

active-voice sentence. This is true, for instance, of verbs which
take two objects in the active voice, one of which is animate, and the
other inanimate. With an animate subject, the middle voice has
reflexive meaning, as in 163.b. and c. below; but with an inanimate

subject, the middle voice has the potentiality or impossibility

meaning, as in 163.d. and e. below:

163.a. 'O - 6 orn-ay- 6 e toggooje cfee.

she dress them gowns these.
VAP
'She dresses/will dress them in these gowns . 1

b. Be - 6 orn-oto toggooje cfee.

they dress gowns these
VAP
'They twill) dress in these go w n s . 1

c. Be - 6 orn-ataako toggooje cfee.

they dress not gowns these
VAP
'They won't/don't dress in these gowns.'

d. Toggooje cfee 6 orn-oto.

gowns these dress
VAP
'These gowns are wearable.'

e. Toggooje cfee &orn-athako.

gowns these dress not
VAP
'These gowns are unwearable . 1

It may be that these two types of meaning associated with the middle
voice would be better described in terms of the 'deep' cases of the
participants involved with the verb, cf. Fillmore 1968 and 1971. In
spite of subsequent revisions of his model, Fillmore's 1968 definitions
 139

of the agentive, dative and objective 'deep' cases seem pertinent

to the present discussion. Let us then first consider the ‘deep*
cases of the active voice/ before we compare these with, the cases
of the middle v o i c e .

It is reasonable to consider the surface-structure subject of

the active verb (J_o in 163.a.) to be the instigator of the action,
ie. Fillmore's agentive case. The animate object (6e in 163.a.)
is the patient, Fillmore's dative. And the inanimate object rep­
resents Fillmore's objective.

In the reflexive-middle sentences (163.b. and c.i, the ’deep*

agentive and dative are identical, and the latter has been deleted in
the surface-structure. Otherwise, the same syntactic and semantic
relations hold between the participants in the three sentences,
163.a. - c.. This is not so in what I shall call the ‘ergative-
middle* sentences, 163.d. and e.. Here, we see that the 'deep1
objective has been promoted to surface-structure subject; and the
other two cases are not, indeed cannot, be expressed at all.
(Semantically, the 'ergative-middle' is very similar to a passive
construction, but with the potentiality/impossibility idea added;
and passive constructions in Fulfulde likewise cannot express an
agentive 'deep1 case, for there is no equivalent to the 'by ...'
prepositional phrase in English.)

Before leaving this subject, here are a few more examples of

the 'ergative-middle' construction:

164 .a . Dum-wacf-ataako.
it do not
VAP
'It can't be done.'

165.a. Wal-aa ko wacf-otoo.

is not what do
VAP VAP
’There is nothing that can be done.'

166. a. Kolte cfee loon-ataako.

clothes these wash not
VAP
'These cloths cannot be wash e d . '
 140

It may be useful to compare these, both morphologically and semanti­

cally, with equivalent passive constructions:

164.b. 'Dum-wad’-athake.
it do
VAP
rIt isn't/won't be done.’

165 .b. Wal-aa ko wacf-etee.

is not what do
VAP VAP
’There is nothing being/going to be done.1

166 .b. Kolte cfee loon-ataake.

clothes these wash not
VAP
'These clothes aren't/won't be w a s h e d . 1

3.5.1.2. Voice and Extensions

At the beginning of section §3.5.1., it was mentioned that verbal

roots could be classified according to their 'voice-potential'. (As
far as I am aware, no such classification has ever been undertaken.)
Such a classification would have to take into account both extended
and simple roots, for the voice-potential of an extended root is not
necessarily the same as that of the simple root from which it is
derived. For example, the 'causative extension', -i n - , accompanies
the introduction of a new participant role to the root, and always
occurs with active or passive, never middle, voice; and the reflexive
extension, - i t - , always occurs with the middle voice. Since our
concern here is voice, I shall simply give a few examples of how ext­
ensions may affect voice.

An intransitive root such as war- 'come' may take the dative

extension, -an-:

167.a. Hassan war-aayi.

Hassan come not
VAP
'Hassan didn't come.1
 141

167.h. Hassan war — an — a a - kam.

Hassan come for not me.
EXT VAP
'Hassan didn't come for me.'/'Hassan didn't come to
fetch. me.'/'Hassan didn't come on my behalf.'

Although the intransitive verbal complex in 167.a. cannot be passivised,

the complex with the extended verbal root can. Thus, from 167.b.,
we can derive 167.c.:

16 7. c . Mi-war - an - a a k a .
I come for not
EXT VAP

'I haven't been fetched.*/'No one's come for me.'/'No

one's called for me.'/'No one's come on my beha l f . 1

The suffixation of the -an- extension on to the essentially intrans­

itive verbal root war— has thus affected the voice—potential of the root.
This fact contrasts, for instance, with the -iloy— ‘early/a long time
ago' extension, which does not introduce a new participant role, and
therefore does not allow an essentially intransitive verb to be
passivised:

168.a. B e - 6ar}i}- aayi.

they move not
VAP
'They didn't move (away) . '

b. Be — 5ary; — ilow - aayi.

they move early not
EXT VAP
'They didn't move (away) early.'

c. *Be - frarji} - ilow - a aka.

they move early not
VAP

The extension that has a reflexive effect on the verbal root

has the form -t_- or -it-, and is accompanied by the middle voice.
Thus, from femmbugo 'to shave* (active voice), we have femmbitaaqo
'to shave oneself' (middle voice):

169.a. Pemmhoowo femmb-ii Muusa.

barber shave Musa
VAP
'The barber shaved Musa ('s head).1
 142

169 .b . Muusa femmb— it - a k e .

Musa shave
EXT VAP

'Musa shaved himself.'

Sentences 169.a. and b. are in fact part of a four-way distinction,

based on active vs. middle voice, and presence or absence of a
-t-/-it- extension; for femmhaago (middle voice, b u t without an
extension) means 'to get shaved', and femmbitugo Cactive voice,
with the homonymous repetitive extension] means 'to shave again':^

169.c . Muusa femmb-ake.

Musa shave
VAP

'Musa got shaved.1

d. Pemmboowo femmb-it- ii Muusa.

barber shave Musa
EXT VAP
'The barber shaved Musa again.'

From the sentences of 169., then, we see that the interaction of

voice and of extensions with the verbal root are of paramount importance.

The simulative extension, -kin- occurs with the middle voice,

although it has no obviously reflexive meaning. Could it conceivably
have a 'deep' ergative-middle meaning? - the subject of the verb
causing him/herself to appear to be something he/she is not? Here
are two examples of the extension:

170. Naa 'o — w u m - 0 , 1o—wum — kin — 0.

not he blind he blind
VAP EXT VAP
'It's not that he became blind, (it's that) he
pretended to become blind.'

7
171. Naa ’o - maay-u, 'o - may-kin-0 .
not she die she die
VAP EXT VAP
'She didn't (really) die, she (simply) pretended to die.'

In this section, I have tried to show the extent to which the

meaning of a verbal complex depends on which of the three voices is
expressed in it. Let us now move on to other properties which, unlike
those of voice, are not obligatorily realised in a verbal complex.
 143

3,5.2. Anteriority

Anteriority is not a property of a category: it contrasts simply

with, its own absence. Its morphological realisation is usually
no; and I shall therefore refer to it as no . However, its final
vowel may be lengthened (resulting in noo) under certain conditions,
as we shall see in the next chapter.

No has already arisen in the discussion of mood (§3.4.1.} and

aspect (§3.4.2.). Essentially, it is a marker of past time. It
can occur in verbless sentences, such as 172.b.; in participles,
such as 173.b.; and, of course, in verbal complexes, such as 174.b.:

0
172.a. 'E mo cfon to Zaariya.
he present in Zaria
'He's in Zaria.1

Q
b. *E mo cfon no^ to Zaariya.
he present in Zaria
ANT
'He used to be in Zaria.'

173.a. Mi kokk-etee-cfo ceede.

I give money
VAP CL
'I will be given money.'/'I am to be given money.'

b. Mi kokk-ete-noo -do ceede.

I give money
VAP ANT CL

'I was to receive money.'/'I was to be given money.'

174. a . 1E-mo - cfon-war - a .

she come
VAP
'She's coming.'

b . 'E-mo - cfon-war - ay - no, sey 'o - 'al - it - i .

she come then she leave
VAP ANT EXT VAP
'She was going to come, then gave up the idea/changed
her m i n d . 1
 144

In all of the b. sentences above, we see that no_ not only Imposes
past time on the clause in which, it occurs, but that it also indicates
a change in the situation depicted in the equivalent a. sentences.
These features were also noted in sentence 152.a. above, and in the
ensuing discussion.

3.5.3. Properties of Person, Number and C a s e .

A verbal complex m a y or may not contain dependent pronouns (cf. §4.1.

below). If it does, then properties of person, number and case will
be realised in the complex. Even if it does not, however, such
properties may be realised by other means: for example, the very fact
that no subject pronoun occurs in a complex usually indicates
g
that the subject is 3rd person; and the singular or plural number
of the subject may be realised by the initial consonant of the root
(cf. §1.2.1.). In this section, I shall mainly be concerned with
the properties of these categories as they are realised in pronouns,
and I shall explain why it is necessary to establish their existence
for the purpose of adequately describing the morphology of the verbal
conjugations.

3.5.3.1. 1st, 2nd and 3rd Person

On the semantic level, Fulfulde distinguishes between speaker Cist

person), addressee (2nd person), and 3rd person. However, the morph­
ological representation of those persons depends on a combination of
factors, amongst which are number, case, and nominal class. Of these,
the first two are to be discussed in the next two sections, and the
last (which only concerns 3rd person) was discussed in §1.3.. A
further factor will, however, be considered here. This is the combin­
ation of persons that may occur, resulting in a 1st person plural
pronoun that explicitly includes the addressee, 1e n , and a 1st person
plural pronoun that explicitly excludes the addressee, m i n . As a
result of these combinations, Fulfulde has a rather unusual 7-way
distinction, which is illustrated by the subject pronouns (underlined)
in the complexes of 175.. These complexes constitute part of the
Relative Completive conjugation, active voice. Note that variation in
the person of the subject correlates in this conjugation with variation
in morph-order (VOSO, VSOO, S V O O ) , and variation in morph— shape (—moo-
v s . -mo— , and -u- vs. — i - ) .
 145

175.a. Ndee suujJ—u - moo-mi-foe? d. Ndee min-suxicC-l- mo - £e?

when hide her X them when we hide her them
V VAP O S O S V VAP O O
'When did X hide them from 'When did we (excl) hide
her?' them from h e r ? '

b. Ndee suixf-u-cfaa—mo — 6e? e. Ndee suud—u— 1en-mo — 6e?

when hide yon her them when hide we her them
V VAP S O O V VAP S O O
'When did yon (sgl hide 'When did we (incl) hide
them from her? ' them from her? '

c. Ndee 'o— suucf— i — mo — &e? f. Ndee suuef—u ~ 'on—mo — Be?

when he hide her them when hide you her them
S V O O V VAP S O O
'When did he hide them 'When did you (pi) hide
from h e r ? ' them from h e r ? '

g. Ndee - suucf-i - mo - 6e?

when they hide her them
S V VAP 0 O
'When did they hide her
from them?'

Let us now consider the plural pronouns and the persons they
represent, as summarised in table XX:

Table XX

10
Plural Pronouns

Ipl. e x c l ., min: 'I and he/she/they' 1st and 3rd persons;

'I and y o u 1, or (1st and 2nd- persons, or.
l p l . i n c l .,

2pl.,
'e n :
1o n :
11 , you and he/she/they
'you'
■} : (1st, 2nd and 3rd persons
2nd person;
3pl. , 6e: 1they 1 3rd person.

As can be seen from the above, the first person plural inclusive, and
the second person plural have similar morphological realisations. In
fact, as well as the 'en - 1on similarity, verbal complexes with
either of these person as subject have VS order in certain conjugations;
whereas complexes with min as subject always have SV order.’'"'*' For
these reasons, Arnott has suggested that the two plural persons con­
cerned should be called *2 p l . inclusive' and 12pl. exclusive1, depending
on whether they semantically include the 1st person (cf. Arnott 1970: 134).
 146

I do not: follow Airnott's suggestion, for consistency would

demand tiiat if:

'on is considered: 2pl. excluding 1st person, and

1en is considered: 2pl. including 1st person,

then:

min should be considered: 3pl. including 1st person.

This would leave Fulfulde without a 1st person plural - and this
would clearly be a misrepresentation of the facts, since the speaker,
i e . 1st person, is semantically included in two of the plural pronouns.
In fact, we have a more useful and consistent description of the pronouns
if we take the addressee(s), ie. the second person, as the variable.
Thus:

min is best considered: lpl. excluding 2nd person?

*en is best considered: lpl. including 2nd person? and
1on is best considered: 2pl..

The similarities which Arnott wanted to capture can at least be part­

ially covered by a statement such as 'VS order may occur in verbal
complexes whose subject expresses 1st, or 2nd, or 1st and 2nd, person*.

Although the sentences of 175. exemplify the full range of persons

in the subject, they all have third person objects Cmo/moo and B e ) .
In fact, object pronouns of all persons can occur in the verbal complex.
Here, then, are some sentences with an object pronoun of the 1st or
2nd person (underlined) :

176.a. Ndee suucf-u-cfaa-mi-Be? b. Ndee suucT-u-cfaa-min-Be?

w hen hide you me them when hide you us them
V VAP S 0 O V VAP S O 0
'When did you hide them 'When did you hide them
from me? ' from us (excl) ? '

c. Ndee 'o - suucf-i-1en-£e?

when she hide us them
S V 0 0
'When did she hide them from u s ? 1

177. a. Ndee suucf-u-maa — mi-&e? b. Ndee suucf-u-mi-1on-£e?

when hide you I them when hide I you them
V 0 S 0 V S O O
'When did I hide them from 'When did I hide them from
you (sg)?' you (pi)?1
 147

In 176.a., b. and c., the first object is lsg. (-mi-), lpl. excl.
(-m i n - ) and lpl. incl. (-'en-) respectively. In 177.a. and b . ,
the first object is 2sg. (-maa-J and 2pl. (.-1on-) respectively.
The second object in all of these sentences is consistently -Be.

By way of conclusion to this discussion of person, it should be

pointed out that all conjugations except the Imperatives and the
Exhortative resemble the Relative Completive (which was illustrated
in 175. and 176.) in that they may take a subject or object of any
person. However, the Imperatives can only take a 2nd person subject
(cf. 178. a. and b.), which is realised as zero unless the negative-marker-,
t o , is present; and the Exhortative can only take either a 1st person
plural inclusive subject (.cf. 179,a.), or 'Allah 'God' tcf. 179.b.)i:

178. a. 0pdakkit-a.' To 1on-cCakkit-ee!

move not you move
away away
S V VAP S V VAP
'Move away (sg)' 'Don't move away (pi) ! '

179.a . 1En-mballut-u - m o 1 b. 'Allah wallut-u - 'en!

we help him Allah, help us
VAP VAP
'Let's help him!' 'May God help us (incl) !

3.5.3.2. Singular and Plural Number

Fulfulde distinguishes singular and plural: there is no dual, or

other number distinction. Singular and plural are realised in dependent

pronouns such as have just been discussed. They are sometimes
also realised, though less consistently or obviously, by the initial
consonant of a verbal root, if it is preceded by its subject (cf. §1.2.1.)
In addition, singular and plural are always realised, perhaps rather
surprisingly, in the VAP-suffix of Imperative complexes - surprisingly,
because no other VAP-suffixes vary according to the number of the
subject. Thus, -a in 180.a. expresses singular, -ee in 180.b. plural:

180. a. Daan - ad b. T)aan - ee!

sleep sleep
VAP VAP
'Go to sleep (sg) ! ' 'Go to sleep (pi) I'

The realisation of number will be discussed in greater detail in the

chapter on exponence.
 148

3.5.3.3. Subject and Object Case

Subject and object case are realised b y both the morphology and the
syntax of verbal complexes. Consider sentence 176.a. and 177.a.
again:

176.a. Ndee suueC—u — cfaa-mi—5e?

when hide you me them
V VAP S 0 0
'When did you hide them from me?'

177.a. Ndee suucf-u — maa-mi-6e?

when hide you I them .
V VAP O S 0
'When did I hide them from you (sg) ? 1

Here, we have a minimal pair of complexes, distinguished only by the

initial consonant of the 2sg pronoun: dy (of -cfaa-] in 176.a., and
m- (of -m a a - ) in 177.a . . D a a is subject, whereas maa is object:
only their respective forms distinguish the one from the other.

Here, on the other hand, is a pair of sentences where the

relative order of pronouns distinguishes subject from object:

181.a. 131 - tokk. - ii- ' e n .

they follow us
VAP
"They (eg. sheep) followed us (incl).'

b. 'En-tokk - ii-cfi.
we follow them
VAP
'We (incl) followed them (eg. sheep).'

Whichever pronoun precedes the verbal root in the above sentences is

interpreted as realising subject; and whichever follows the root is
interpreted as realising object.

An indirect object is distinguished from the direct object by

order: the former precedes the latter, whether both, are within a
12
verbal complex, or not. Consider the sentences of 175. on p. 145..
In each, the first object pronoun is interpreted as the indirect
object, and the second as direct object. In order to effect a different
interpretation, the order of the object pronouns has to be switched.
 149

For example, 182.a. and b. correspond to 175.c. and d. respectively!

182.a. Ndee *o— suucC- i — Be - mo?

when he hide them her
S V VAP 10 DO
'When did he hide her from them?'

b. Ndee min-suncC- i - 6e_ - mo?

when we hide them her
S V YAP XO DO
'When did we texcl) hide her from them?'

The same principle of indirect object before direct object applies

in complexes with VS Cverb-subject) order as in SV complexes: witness
sentences' 175.a . , b., e. and f .. However, such complexes are verging
on theborderline of acceptability, or maybe we should say,inter-
pretability, for they involve three pronouns in a row, whose case-
roles have to be interpreted.

One further point concerning the properties of case in the verbal

complex. Our examples thus far have had a maximum of two objects.
When the complex contains one of certain extended verbal roots,
three objects may co-occur in the complex, for example:

183. '0-hokk-an - t - e_ - mo - cfum.

he give for you him it
S V EXT V A P O 0 0
'He'll give it tohim on your behalf.'

Such a series of objects is not possible in VS complexes, where the

subject pronoun, too, would be vying for interpretation.

These matters will be elaborated further in the next two chapters.

Let us now turn to a consideration of the various conjugations.

3.6. Indicative Mood: Conjugations in Neutral Sentences

The above title of course begs the question: what is a neutral sentence?
I shall in fact assume that a neutral sentence is a statement or yes/no
question which is not within a relative clause, a focus construction, a
wh- question, or a narrative sequence. These latter constructions will
be dealt with under non-neutral sentences, in §3.7. below.
 150

Five positive and two negative conjugations may occur in

neutral sentences of indicative mood. They are:

General Completive General Incompletive

Stative Progressive
Negative Completive Negative Incompletive
Vague In comp le ti v e .

I shall first discuss the conjugations in the left-hand column, all

of which express Completive Aspect; and I shall then discuss those
in the right-hand column, which express Incompletive Aspect.

3.6.1. Completive Aspect

3.6.1.1. The Stative and the General Completive

In order to establish the semantic distinction between these two

conjugations, let us consider the following:

184.a. 1E-cfum-he1y - i Muusa. b. Oum-he'y - ii-mo.

it suffice Musa it suffice him
VAP VAP
13
'It is enough for M u s a . 1 'It became/has become
enough for M u s a . 1

185.a. ’E-mo-waaw - i Fulfulde. b. 'O-waaw - ii-nde.

he be able Fulfulde he be able it
VAP VAP
13
'He can speak Fulfulde.' 'He became/has become pro
ficient in it.

186. a. ' E - m o - r o n n d - ii« b. ‘O - r o n n d — a k e .

she put on she put on
headload headload
VAP VAP
13
'She is carrying a 'She (has) put her head-
headload.' load o n . 1

The a. sentences above are of the Stative conjugation, whereas the

b. sentences are of the General Completive. Those in 184. and 185.
express active voice, whereas those of 186. express middle voice. As

the translation of these sentences suggests, the General Completive

depicts a completed process or action, the Stative a state.

Stativity is a property that is not compatible with every verbal

root, and the number of roots that may occur in the Stative conjugation
 151

is fairly limited. Certain roots in fact have a propensity to occur

in the Stative, and occur in few other conjugations. Such is waaw-
above, and others that express qualities, processes or states, for
example wooj- 'be/become r e d 1 , *annd- 'know*, and four- 'exceed'.
These all occur w i t h active Crather than middle] voice; and they
display a morphological quirk in their negative equivalents Ccf.
§3.6.1.2. b e l o w ] .

Stativity is particularly compatible with verbal roots that

express a process enacted on the body, and that results in a state.
Such is, for example, ronnd— in 186.. The middle voice of the Stative
conjugation very often occurs with such roots, as exemplified in 186.
above, and in the following:

187.a. Bikkoyi] 1e-foorn-ii. 188.a. 'E-mi-facf - ii.

children dress I wear
VAP shoes
VAP
1 (The)children are dressed.'
'I'm wearing Cray) shoes.1

In sentences such as .these, it is not specified whether the action

that led to the state was performed reflexively or not. However,hthe
middle voice of the General Completive of such roots only has ref­
lexive meaning:

187.b. Bikkoyi] foorn-ake. 188.b. Mi-fad - ake.

children dress I wear
VAP shoes
VAP
' (The) children dressed
themselves.‘ 1I put on (my) shoes.1

Arnott (1970: ch.52) distinguishes two Stative conjugations, on

both morphological and syntactic grounds. For Kaceccereere Fulfulde,
there is not enough evidence to justify two such conjugations. True,
complexes of the Stative may or may not contain cTon, which is one of
the morphological distinctions found in the Gombe dialect described by
14
Arnott; however, the occurrence of don in such complexes is expli­
cable in terms of the emphasis and/or the locative implications that
it lends to a complex. For example, in Kaceccereere Fulfulde, if a
Stative complex is used in reply to a yes/no question, then don is
likely to occur in the complex, for it emphasises the affirmative
15
nature of the reply:
 152

189.a. Loonde ndee 'e-heew-i na?

pot this fill
VAP QU

'Is this cooking-pot full?r

b. 1X i , ie-nde-cfon-heew-i.
yes it fill
VAP
'Yes, it is full.'

?c. 'I i / *e-nde-heew-i.

yes it fill

Sentence 189.c. is perfectly grammatical, but it is not ’strong1

enough to be used in reply to the question in 189.a.. The locative
implications of don are illustrated by the fact that a Stative complex
with don, b u t not one without, may be used in reply to a question
asking tokoye? 'where 1:

190.a. Tokoye Asabe?

where Asabe
’Where is Asabe?'

b. 'E-mo- cfon-j oocf-ii.

she sit
VAP
'She's seated.'/'S h e 1s around, seated.'

?c. 'E—mo — joocf—ii.

she sit
VAP

Again, the c. sentence here is perfectly well-formed; but it is not

suitable as a reply to 190.a.. (I have given a second,- alternative,
translation of 190.b., reflecting the locative meaning of cfon. The
possible origin of such forms was mentioned in section §2.1.2.3.,
page 84, above, and is further discussed in McIntosh 1980.) From
the above evidence, I believe that it would not be useful to establish
two Stative conjugations in Kaceccereere Fulfulde; rather, it is
better to regard the occurrence of don in Stative complexes as deter­
mined by discourse considerations.

The General Completive is possibly the most 'neutral' conjugation.

It is the conjugation most often used for citation purposes, and it is
 153

semantically compatible with, the vast majority of verbal roots.

Moreover, it is morphologically -uncomplicated: unlike many otfier
conjugations, it only has a single VAP-suffix for each voice, regard­
less of the other morphs which occur in the complex; it always has
SV order; and unlike the Stative, for instance, it does not make
17
use of periphrasis.

The 'meaning1 of the General Completive has already been discussed,

both with regard to the establishment of the category of aspect (cf.
§3.4.2.), and with regard to the way the Stative differs from it.

3.6.1.2. Negative Completive

The negative equivalent of the Stative and the General Completive is

18
the Negative Completive: the semantic distinction between the two
positive conjugations is thus neutralised under negation. The
following are the negative counterparts of 184.a. and b. - 186.a.
and b. (page 150 above) :

184.c. Dum-he'y - aayi Muusa. d. Dum-he'y - a a - m o .

it suffice not Musa it suffice not him
VAP VAP
'It isn't/hasn't become 'It isn't/hasn't become
enough for M u s a . ' enough for h i m . '

185.C. '0-waaw - aa Fulfulde. d. '0-waaw - aa - n d e .

he be able not Fulfulde he be able not it
VAP VAP
'He can't speak/hasn't be­ 'He can't speak/hasn't be
come proficient in Fulfulde. come proficient in it.

186.c. 'O - ronnd — aaki.

she put on headload not
VAP
'She isn't carrying/hasn't puit
on her headload.

The complexes of 184.c.-d. and 185.c.-d. express active voice (as do

their a. and b. counterparts), but that of 186.C. expresses middle voice

This voice distinction accounts for the occurrence of -aa and - aayi
in 184. and 185. on the one hand, and — aaki in 186. on the other.
However, how are we to account for the -aa/— aayi alternation in the
active?
 154

In Kaceccereere Fulfulde, — aayi generally occurs If the VAP-

suffix is in complex—final position, and — aa occurs elsewhere. In
184. c . , Muusa is an ME? object (which is not part of the verbal
complex] , the VAP-suffix is complex-final, and it therefore has the
form - aayi; but in 184.d., the object is a dependent pronoun, -mo
(which is part of the complex], the VAP-suffix is not in final
position, and it therefore has the form -aa. However, there are
certain roots that have the -aa rather than -aayi suffix in all
circumstances: these are the same roots as were mentioned on page
151 above, in other words, roots that express qualities, processes
or states. Waaw- is one of these roots, and this fact accounts for
the -aa (rather than - aayi] VAP-suffix in 185.c..

For the Gombe dialect, Arnott (1970: §53.13-17) establishes a

separate 'Negative of Quality* conjugation, which occurs only in the
active voice, has -aa VAP-suffix, and-.occurs with the same type of
roots as have just been mentioned. However, since, in t^ie Kaceccereere
dialect, there is no evidence of a semantic and/or syntactic contrast
between the - aa and - aayi VAP-suffixes, there is not enough justi­
fication for the establishment of a separate negative conjugation.
Rather, it would seem more prudent simply to consider such roots as
taking an exceptional, -aa, suffix instead of the more normal - aayi
suffix.

3.6.2. Incompletive A s p e c t .

3.6.2.1. Progressive and General Incompletive

The semantic distinction between the Progressive and the General

Incompletive is not as clear-cut as is that between the Stative and
the General Completive. In principle, the General Incompletive is
used for incompleted actions or processes, which are usually habits,
or in the future; whereas the Progressive is used for actions or
processes that are or were happening, and that are or were interrupted.
In practice, the Progressive is also increasingly used for habits, as
we shall see presently.

The sentences of 191. and 19 2. exemplify the distinction between

the two conjugations we are considering here:
 155

191.a. D i ~ n j a r ~ a y ndiyam to mheela ton.

they drink water at lake there
VAP

'They (eg. cowsl drink/will drink water at the

lake over there.1

b. 'E-cfi - njar-a ndiyam to mbeela ton.

they drink water at lake there
VAP
’They (eg. cows) are drinking water at the lake
over there.1

192.a. Yim&e cCucf&e m o 6 - t - oto.

people many gather
EXT VAP
'Many people gather/will gather together.1

b. Yim&e cfucf&e 'e-mob - t - oo.

people many gather
EXT VAP
'Many people are gathering together.'

The a. sentences above contain a complex of the General Incompletive,

whereas the b. sentences contain a complex of the Progressive. The
complexes in 191. are of the active voice, whereas those of 192. are
of the middle. The translations given are intended to cover the range
of meanings that may be conveyed by these conjugations. Sentence
191.a. expresses a habit, which may have started in the past and
still continue in the present. (However, if the habit has been discont­
inued, then the anteriority-marker, no will usually be needed in order
to make this fact explicit, thus: Pi-njar-ay-no ... 'They used to
drink ...*.) It may also express an action that will begin in the
future, whether as a habit or not. Sentence 191.b. depicts an action
that has started and is still continuing at the moment of utterance, or
that started and was interrupted in the past. The fact that the action
ceased in the past will usually be made explicit by the use of no
(1E-cfi-njar-ay-no ... 1They were drinking ...') , though contextual clues
in the discourse-situation may be sufficient. Sometimes, younger speakers
of Kaceccereere might use 191.h. to express a habit, in the place of
191.a.. It is probable that in this respect, they are copying the
frequent use of the Hausa Continuative aspect for habits (see Kraft
and Kirk-Greene 1973: 157, footnote 1}.
 156

The pattern of meaning between 192.a. and b. is parallel to

that which holds between 191.a. and b.: the first expresses a
habit or a future action? the second an action that is, or was,
interrupted.

The Progressive conjugation expresses a point in the progress

of an action or process, rather than the state reached at the end
of a process. The contrast between the Progressive and the Stative
may be usefully brought out b y the a. and b. sentences of the
following:

193.a. Dere 'e-wakk - ii sawru.

Dere put across staff
shoulders
VAP
'Dere is carrying a staff across his shoulders.'

b. Dere ’e-wakk - oo sawru.

Dere put across staff
shoulders
VAP
'Dere is putting a staff across his shoulders.'

The Stative complex in 193.a., 1e-wakk-ii, shows that the action of

putting across the shoulders has been completed; and the resultant
state is thus described. The Progressive complex in 193.b., 'e-wakk-oo,
shows that the action of putting across the shoulders is continuing;
and this continuing, or progressing, action is what is described.

Unlike the Stative, the Progressive is semantically compatible

with the vast majority of verbal roots, and there is no body of roots
that are 'inherently progressive' to parallel those that are ’inherently
stative'.

The place of cTon in Progressive complexes is exactly parallel

to its place in Stative complexes: cfon is used for emphasis, and has
locative connotations. For example, the b. sentences below would be
suitable as responses to the questions posed in the a. sentences;
but the c. sentences would be unsuitable:
 157

194.a. Musuuru 'e-cCon— fij - a na?

cat play
VAP QU

'Is tlie cat playing? 1

b. 'Ii, 1e-ndu-cfon-f i j-a.

yes- it play
VAP
'Yes, it is playing.1

?c. 'Ii, 1e-ndu-fij-a.

yes it play

19 5.a . Tokoye Faatu?

where Fatu

'Where is Fatu?'

b. 1E-mo - cfon-def - a nyiiri.

she cook food
VAP
'She's cooking food.'

?c . 'E-mo-def - a nyiiri.
she cook food

The Progressive without cfon, as in 194.c. and 195.c. is not 'strong1

enough to act as a reply to a yes/no question, and does not have

locative connotations that can serve as a reply to alocative question.

Thus, for the same reasons as were explained with reference to the
Stative, I do not establish two Progressive conjugations (see pp. 151-2).

3.6.2.2. Negative Incompletive

The negative counterpart of the General Incompletive and of the

19
Progressive is the Negative Incompletive: the interrupted vs.
uninterrupted distinction is thus neutralised under negation:

19I.e. Di - njar - ataa ndiyam to mbeela ton.

they drink not water at lake there
VAP
'They Ceg. cows) don't/won't drink water at the lake over
there.1
'They Ceg. cows) aren't drinking water at the lake over
there. 1
 158

192.c. Yimbe ducfbe mob — t - ataako.

people many gather not
EXT VAP
'Many people don't/won't gather together.'
‘Many people aren't gathering together.'

The complex .in 19I.e., with -ataa VAP-suffix, expresses Active Voice;
whereas that in 192.c., with -ataako VAP-suffix, expresses Middle Voice.

3.6.2.3. Vague Incompletive

When a Kaceccereejo Fulani takes his leave, he will often say:

196. a . Kinn - an - 0 - aa - mi w u r o .
greet for you me home
VAP
'Greet (.the people at) home on my behalf.'

The standard, formulaic, response to this is:

19 6 -b. Be - nan - ma.

they hear
VAP
'They hear/accept (your greetings).'

The latter, and its counterpart with a singular subject (‘Q-nan-ma

'He/she h e a r s ' ) , are the only forms of the Vague Incompletive that
most Kaceccere'sn know; and on such evidence, it would seem best
to regard these as fossilised forms, and to consider the Vague Incomp­
letive to be no longer in productive use in this dialect.

However, certain older speakers do use forms such as the following,

also of the Vague Incompletive:

19 7 .a . Ngel-yoor-uma cfoo .
it dry here
VAP
’It may dry h e r e .'/‘Maybe it will dry here.'

Younger speakers would more likely use the General Incompletive with
wonngo 1maybe 1:

197.b. Wonngo ngel-yoor-ay cfoo.

maybe it dry here
VAP
'Maybe it will dry here.'/'It may dry h e r e . '
 159

Both, of these sentences contrast with:

19 7 .c . Ngel—yoor— ay cfoo.
it dry here
VAP
'It will dry here. '

Sentences 197.a. and b. differ from 197.c. in the degree of certainty

that they express concerning the realisation of the action or process
depicted by the verbal root. This fact casts some doubt on the
suitability of the Vague Incompletive for inclusion in Indicative
Mood, which is primarily concerned with statements. Given this un­
certainty, and the rare use of the Vague Incompletive, I shall
not include it in further discussions.

3.7. Indicative Mood: Conjugations in Non-neutral Sentences

A t the beginning of section §3.6., I explicitly excluded certain

constructions from consideration. These were: relative clauses,
focus constructions, wh- questions, and narrative sequences. It
is these that use the three conjugations of Indicative Mood that
have not yet been discussed, namely:

Relative Completive Relative Incompletive

Emphatic Completive

Let us deal with these conjugations along aspectual lines: first,

the Relative Completive and Emphatic Completive, and then the Relative
Incompletive.

3.7.1. Completive Aspect

3.7.1.1. Relative Completive

The Relative Completive expresses Indicative Mood, Completive Aspect,

and Positive Polarity, as do the General Completive and the Stative.
However, it is distinct from the latter two conjugations in its

syntactic distribution.

First, the Relative Completive is used in relative clauses, where

the General Completive and the Stative may not occur. Thus, the c.
sentence of 198. is correct, because it contains a Relative Completive
 160

complex in a relative clause, whereas the d. and e. sentences are

not, because they contain either a General Completive or Stative
complex in a relative clause:

19 8. a. Min— njoocT— a k e . b. 1E-min-cfon-n joocf-ii .

we settle we settle
VAP VAP
GEN COMPL MID STATIVE MIDDLE
'We Cexcl] settled down.' 'We are settled.*

c. Raa to min—njoocT-ii.
see where we settle
VAP
REL COMPL MID
'Here's where we settled down/where we are settled.1

d. *Raa to min-njoocC-ake . e. *Raa to 1e-min-cfon-n joocf-ii.

see where we settle see where we settle

Secondly, the Relative Completive is used in sentences containing

a focussed NP or adverbial expression. This matter will be treated
in more detail in §6.2.1.1. below, but here is one example. The
c. sentence of 199. is correct, because it contains a Relative Compl­
etive complex in the presence of a focussed NP, k a n k o , whereas the d.
and e. sentences are unacceptable, because they contain a General
Completive or a Stative complex in the presence of a focussed NP:

199.a. Mi-bur - ii—mo semmbe. b. 'E-mi—bur - i_- mo semmbe.

I exceed him strength I exceed him strength
VAP VAP
GEN COMPL ACT STATIVE ACTIVE

'I. became stronger than . 'I'm stronger than him.'

him. '

c. Kanko bur - 0 - mi semmbe.

him exceed I strength
VAP
REL COMPL ACT
'I became stronger than h i m . 1/'I'm stronger than h i m .1

d. *Kanko mi-bur - hi semmbe.

him I exceed strength

e . *Kanko 1e-mi- bir - jl semmbe .

him I exceed strength
 161

Thirdly, the Relative Completive is used in w h — questions,

such, as 200.c.. Again, the General Completive and the Stative
(c£. 200.d. and e. respectively! are unacceptable:

200.a. D e — loon— aama. ’E-de — cTon—loon—a a .

they wash they wash
VAP VAP
GEN COMPL PASS STATIVE PASSIVE
‘They Ceg. clothes) have "They Ceg. clothes) are
been wash e d . 1 washed.1

c. Deye loon-aa?
which wash
VAP
REL COMPL PASS

'Which Ceg. clothes] have been/are washed?'

d. *Deye loon— aama? e. *Deye 1e-cTon-loon-aa?

which wash which wash

Fourthly, the Relative Completive is used in a n a r r ative.sequence

of verbal complexes expressing Completive Aspect. Such a sequence
is particularly often found in folk-tales. The General Completive
is unacceptable in such circumstances; and since we are dealing
w ith a series of events, rather than states, the Stative conjugation
would be semantically inappropriate. CNote that subsequent complexes
in such a sequence are frequently introduced by sey ’and then', but
the Relative Completive remains, with or without s e y ,) Here is an
example of such a sequence:

201. ’Aa -yii, dowroori 'e gawri 'e 'alhama 'e tammbaari, 20
y o u ( < y i 1-i) millet and guinea- and wheat and (a certain
see c° m cereal)
VAP

se^ — ■ ~ nan - i godcfo paanyo ’e-cfon, cfum bodeejo coy,

then they hear certain youth around he light- very
skinned
REL COMPL ACT
sey 6e - 1umm-ii, fre - njah-ay.
then they arise they go
VAP VAP
REL COMPL MID

’You see, millet, guinea-com, wheat and 'tammbari', they

heard that there was a certain very light-skinned young
man around, so they got up, and were going to go.*
 162

Xn 201., tile protagonists Call personified cereals) are introduced

by 1aa-yii X a complex of the Stative Active1 , and once the scene
has thus been set, subsequent actions C6e—nan-i, Be-'umm-ii] are in
the Relative Completive. Xn this particular instance, the sequence
of events is interrupted by the use of the General Incompletive,
Be -njah-ay, which expresses an unrealised intention.

An important point to note about the Relative Completive is that

its use neutralises the semantic distinction between the General
Completive and the Stative. This is brought out b y comparison of the a.,
b. and c. sentences in 198. - 200.. The a. and b. sentences are
semantically distinct, but in the c. sentence, two interpretations are
possible, one corresponding to the a. sentence, and the other to the
b . sentence. There is no evidence in this dialect of Fulfulde to
suggest that the General Completive and the Relative Completive realise
a semantic contrast in neutral sentences. This dialect thus differs
from those of Futa Toro and Futa Djallon, where such a contrast does
exist Ccf. Diallo 1979) .

3.7.1.2. The Emphatic Completive

The Emphatic Completive differs from the other conjugations that

realise Indicative, Completive and Polarity, in that it alone is
used in a complex whose root is focussed. Verb— focus will be more
fully treated in §6.2.2. below, but here is an example to illustrate
this conjugation. In 202.a., the two roots, hokk— and rams— , •are com-
bined with completive aspect, necessitating}the Emphatic Completive.

The use of the General Completive (as in 202.b.) r or of the Relative

Completive (as in 202.c.) is unacceptable in such a context:

202. a. 1A - hokk-0 - am ceede cfen, naa 'a — rams-0 - am-cfe.

you give me money the not you lend me them
VAP VAP
EMPH COMPL ACT EMPH COMPL MID
'You gave me the money, you didn't lend it to m e . 1

b. * 1A - hokk—i -yam ceede cfen, naa 'a—rams—ake— 1am-cfe .

you give me money the not you lend me them
GEN COMPL ACT GEN COMPL MID
 163

202.c. *Kokk-u - cfaa-rmi ceede cfen, naa ndams-i - cfaa-mi-cfe.

give you me money the not lend you me them
REL COMPL ACT REL COMPL MID

The Emphatic Completive is not used in relative clauses, in sentences

with a focussed NP or adverbial expression,, or in wh-.. questions -
as noted above, the Relative Completive is normal in such circumstances
(unless the complex expresses incompletive aspect: see the next
section}.

3.7.2. Incompletive Aspect

3.7.2.1. Relative Incompletive

The uses of the Relative Incompletive almost exactly parallel those

of the Relative Completive: for it is likewise used in relative
clauses, in sentences with a focussed NP or adverbial expression,
and in wh- questions, as can be seen from examples 203. - 205..
Neither the General Incompletive nor the Progressive may occur
in these circumstances.

203.a. Min-sipp-ay fciraacfam. b. 1E-min-cfon-sipp-a Mraacfam.

we hawk fresh milk we hawk milk':
VAP VAP
GEN INCOMPL ACT PROGRESSIVE ACTIVE
'We hawk fresh milk.' 'We're hawking fresh milk.'

c. 'Biraacfam min- v sipp - ata ' e - w e l - i masin.

fresh milk we hawk sweet very
VAP VAP
REL INCOMPL ACT
'The fresh milk that we hawk/are hawking tastes very good.'

d.*Biraacfam min-sipp-ay 'e-wel-i masin.

fresh milk we hawk sweet very
VAP VAP
GEN INCOMPL ACT

e .*'Biraacfam 1e-min-cCon-sipp-a 'e-wel - i masin.

fresh milk we hawk sweet very
VAP VAP
PROGRESSIVE ACTIVE

Sentence 203.c. contains a relative clause, with the requisite Relative

conjugation. Sentences 203.d. and e. are unacceptable because they
have an unacceptable conjugation CGeneral Incompletive, and Progressive
 164

respectively) in a relative clause.

As regards sentences with, a focussed NP or adverhial expression,

consider 204.c., where to maayo is focussed, and a Relative Incompl­
etive complex is used:

204.a. Na'i leer - ay_ to maayo.

cows wander to river
VAP
GEN INCOMPL
ACT
1 (The] cows wander towards (the) river.'

b. N a 1i 1e-cfon— 1eer — _a to maayo.

cows wander to river
VAP
PROG ACT
1 (The} cows are wandering towards (the) river.*

c. To maayo na ii• .eer - d

^ - ata.-21
to river cows wander
EXT VAP
REL INCOMPL ACT
1(The) cows wander/are wandering towards (the) river.1

d. *To maayo na'i 'eer - d - ay .

to river cows wander
EXT VAP

e. *To maayo na'i 1e-cfon- 'eer- d-a_.

to river cows wander
EXT VAP

Sentences 204.d. and e. are unacceptable because they contain a focussed

constituent together with a complex of the General Incompletive and
the Progressive respectively.

In sentence 205.c . , we see the occurrence of the Relative Incomp­

letive in a wh- question:

205.a . NgeI-woy-ay. b. 1E-nge1-woy-a.

it cry it cry
VAP VAP
GEN INCOMPL ACT PROG ACT
'It (eg. child) cries.' It (eg. child) is crying.'
 165

205.c. Ngim cfume ngel-woy-ata?

for what it cry
VAP
REL INCOMPL ACT
'Why does it (eg. child) cry?'/'Why is it crying?'

d. *Ngim cfume ngel-woy-ay? e. *Ngim cfume 'e-ngel-woy-a?

for what it cry for what it cry
VAP VAP

As with the other sets of sentences, the d. and e. sentneces of 205. are
unacceptable because they contain a complex of the General Incompletive or
the Progressive, in syntactic circumstances that require a relative con­

jugation.

In the three syntactic circumstances that have been discussed above,

the semantic distinction between the General Incompletive and the Progr­
essive is neutralised. This is why I have given two interpretations for
the c. sentences above, one corresponding to the a. sentaecace of the same
set, and '.one corresponding to the b. sentence of the same set. This neut­
ralisation is comparable to the neutralisation of the General Completive
vs. Stative distinction in Completive complexes.

There is one respect, however, in which the Relative Incompletive does

not parallel the Relative Completive. This is when a sequence of complexes
express Incompletive Aspect in a narrative (eg. a recipe). Here, rather
than the Relative Incompletive, which one might have expected (since the
Relative Completive is used for complexes expressing Completive A s p ect),
the Subjunctive conjugation is used. This use of the Subjunctive will be
discussed in §3.8.3. below.

3.7.3. Indicative Mood: Conclusions

The relationship between the conjugations and the verbal properties des­
cribed so far may be usefully depicted in the form of the tree on page 166 .
At the foot of the tree are the various conjugations (underlined), and
above each conjugation, the properties expressed in complexes of that con­
jugation may be traced. For example, the Relative Completive expresses
relative, positive, non-emphatic, completive and indicative; the negative
completive expresses negative, non-emphatic, completive and indicative; and
so on. The system of branches shows that certain semantic/syntactic dist-
22
inctions are neutralised, once a particular 'choice' has been made. For
example, the 'choice' of negative precludes the relative vs. non-relative

distinction; and the choice of relative precludes the stative vs. non-
 166

stative distinction (in completive aspect) , and the progressive vs. non­
progressive distinction (in incompletive aspect).

iH

•H
o
S
$
+j
d
u !> T -4
•H *rl I— i
-P
g <d O Cn -H
H u
■H
'd
o g
H
to
e
*H
■ rl

ft >
a -h
ft o
CP
g
-n
g
o
o
'd
I—I
§
w
0)
•H
-P
U
d) ■H i—i
f
o
t
p rH
fit

i—I

X have labelled the properties in the above tree according to the

terminology established in sections §3.4. - §3.5. above. However, an
alternative terminology could have been adopted, whereby each non-final
node in the tree would divide into '+' (plus) or (minus) one of the
properties in a category. For example, indicative could have divided
 167

into 4-completive and -completive, non-emphatic could have divided

into 4-positive and -positive, and so on. However, since the pro­
perties of some categories have well established names, such as
completive and incompletive, positive and negative, I have used
those names where they exist. Where no such name exists, the term-
4*
inology I use is essentially a - (plus-or-minus) system: emphatic
and non-emphatic, stative and non-stative, and so on.

Although the tree neatly captures certain characteristics of

Fulfulde conjugations, it is deficient in a least two respects.
First, the repetition of positive under non-emphatic and under emph-.
atic is descriptively uneconomical. However, a more economical ver­
sion of this part of the tree, such as:

Completive

Non-Emphatic Emphatic

would be inaccurate, since the negative completive is not the negative

counterpart of the emphatic completive. (A sentence with a complex of
the latter conjugation is negated by means of a sentence-initial 'particle',
naa: compare example 202.a..) Secondly, there is some evidence for inc­
lusion of the Subjunctive within indicative mood. (If this idea seems
like a contradiction, then either two morphologically identical conjuga­
tions would have to be established, one in indicative mood, the other in
non-indicative - and this, would be descriptively uneconomical; or, it
may be that the definition of mood needs revision.) The evidence concerning
the Subjunctive will be discussed below, in §3.8.3. - it involves the
use of the Subjunctive in narrative sequences expressing Incompletive
Aspect.

Let us now proceed to the discussion of the conjugations that

express non-indicative mood.

3.8. Non-indicative Mood

3.8.1. The Imperatives

The two Imperative conjugations express commands with a 2nd person

23
subject (which is often realised as zero). There is what we may
 168

call an aspectual difference between them, for tbe 'ordinary'

Imperative expresses a peremptory command to perform an action

that may not have started yet, and is to be completed; whereas tbe
Continuative Imperative expresses a command to continue an action
(.or indeed a state: see below) that has already started, and may
never be completed. The complexes in 206.a. and 207.a. exemplify
the 'ordinary' Imperative, and those in 2 0 6 ,b. and 2 0 7 .b. exemplify
the Continuative Imperative:

206. a. W a r -01 b. War - atay I

come come
VAP VAP
'Come (sg) 1' 'Keep on coming (sg)I'

207. a. Njoocf-eeJ b. Njoocf-eteei

sit sit
VAP VAP
1Sit down Cpl) I * 'Stay seated Cpl)I'

Because the ordinary Imperative anticipates the completion of an

action, it may be described as expressing completive aspect; and
similarly, because the Continuative Imperative anticipates the
continuation of an action, it may be described as expressing incom­
pletive aspect. There is also morphological evidence for treating
the distinction between the two Imperative conjugations as aspectual,
for the t_ found in the Continuative Imperative (cf. the b. complexes
above) is characteristic of many Incompletive conjugations, as we
shall see in the next chapter.

Through its two Imperative conjugations, Fulfulde has an

inflectional means of establishing aspectual distinctions that are
expressed periphrastically in the more familiar Indo-European languages
cf. the use of 'keep on and of 'stay in 206.b. and 207.b.
above. Moreover, the Continuative Imperative may be used not only
to express the continuation of an action or a proc e s s , as in 2 0 6 .b.
(which may mean 'Keep on coming (many times)1', or 'Continue on your
way here (once)!'); but also to express the continuation of a state
24
as in 207.b.. It has been argued (cf. Lakoff 1970: 104) that the
Imperative is incompatible with the notion of stativity; but this
idea is not borne out by the Fulfulde facts.
 169

3.8.2. . The Exhortative

The status of the Exhortative as a conjugation needs discussion. It

is complementary to the 'ordinary1 Imperative, inasmuch as the latter
occurs only with a 2nd person subject, and the Exhortative occurs
only with 'Allah C'Allah, G o d 1} or 'en D p i incl) as subject. Given
the semantic similarity between commands and exhortations, it is
arguable that the Exhortative Is a 1st and 3rd person Imperative.
Here are some examples of its use: sentence 208.a. may be compared
with the equivalent Imperative complexes for the same root, joocf-
1s it'.

208. a. 'En-njoocC-ob'. 209. ‘En-njah-ul

we sit we go
VAP VAP
‘Let's sit down I* 'Let's g o ! 1

210. 'Allah reen - 0 - el

Allah protect you
VAP
'May God protect you 11

The contrast between the -o<?, -u and 0 VAP—suffixes in the above complexes
is a matter of voice: oo is used for the middle voice, 0 or -u_for
the active v oice.

Despite the similarities, the Exhortative differs morphologically

from the ‘ordinary1 Imperative in certain crucial ways: unlike the
Imperative, its VAP-suffix is not sensitive to the number of its
subject (cf. the -0 and -ee of 206.a. and 207.a.}; unlike the
Imperative, the subject of an Exhortative complex is never realised
as zero (cf. all the complexes of 206. and 207.); and the Exhortative
VAP-suffixes, although they bear comparison with the ‘ordinary*
Imperative ones, are not the same. Given these facts, it seems advisable
to maintain the Exhortative as a separate conjugation.

3.8.3. The Subjunctive

The Subjunctive may express indirect (.reported) commands and exhortations,

as in 206.c. and 2 0 8 .b., which are comparable to 206.a. and 208.a.:
 170

25
206.c . *0-wii ngar - 0 - a a .
(wi'-ii]
lie say come you
VAP VAP
‘He said you (sg] should come.1

208.b. He - robii njoocf- oo - 'en.

Cmbi‘-ii]
they say sit we
VAP VAP
'They said we Cincl] should sit down.‘

Here, njar-0-aa is a Subjunctive Active complex, njoocf—oo— 'en a

Subjunctive Middle complex. The Subjunctive may also express purpose
(as in 211.], permission, or the request for permission (as in 212.),
and it may combine with sentence-initial 'particles' (cf. §2.1.2.4.]
to express obligation or necessity (as in 213. and 214.]:

211. Be - ndill-ii to 6e - ngi'-ee.

they leave lest they see
VAP VAP
'They left, lest they be seen.'

212. Mi-wolw-a?
I speak
VAP
'May X speak?'

213. Haani cfam-cTaan — in-ee .

ought it sleep
EXT VAP
'It (milk] ought to be left overnight (to sour].'

214. 'Igga mi-winnd-an-a-mo.

better I write to her
VAP
'I had better write to her.'

Although the above uses of the Subjunctive clearly express Non-

Indicative mood, the Subjunctive in fact straddles the two moods,
for it is used for a narrative sequence of actions that depict
habits - a usage that is comparable to that of the Relative Compl­
etive for a sequence of actions that are completed. Dam-war— t— a
in the following exemplifies this use of the Subjunctive.
 171

215. Daga Biraacfam, dam-war - t — a tordej daga torde

from fresh, milk it come butter—milk from butter­
milk
EXT VAP
dam-war - t — a daaniidam.
it come sour—milk
EXT VAP
'From Obeingi fresh milk, it becomes butter-milk;
from butter-milk, it becomes sour-milk.'

In the above usage, the Subjunctive expresses non-factive modality

(cf. Givon 1973]; but inasmuch as it expresses a statement of
habitual actions, rather than commands, obligations, or other sorts
of 'mends' (cf. Lyons 1977: 745], It surely expresses indicative,
not non-indicative, mood. Moreover, the Subjunctive VAP-suffixes
are most conveniently described under indicative mood, as we shall
see in the next chapter, for they are very characteristic of incompl­
etive aspect. It seems, then, that the category of mood may be un­
suitable for an analysis of the Fulfulde verbal system. It is to be
hoped that further research will solve some of the problems posed by
the Subjunctive.

3.9. Verbal Conjugations: Conclusions

The purpose of this chapter has been to establish the essential semantic
and syntactic features that characterise the various conjugations in
Kaceccereere Fulfulde. Considerations of space have not permitted a
detailed description of the use of all^the conjugations: for example,
I have not mentioned the use of the Stative and Progressive in dep­
endent clauses (Mi-yii-mo 1e-mo-joocf-ii *1 saw her seated'] . However,
I hope to have given enough details in order to establish the identity
of each conjugation on the syntactic and semantic levels. We shall
now see that they are also morphologically distinct from one another.
 172
Chapter 3: Footnotes

1. This short—vowelled suffix is of the Relative Completive Active

(necessitated b y ngim cfume?) , whereas the long—vowelled suffix
in 140. is of the General Completive Active.

2. ‘Morphosyntactic properties1 is the term used b y Matthews (1972) ,

in his discussion of Latin inflectional morphology. Since this
discussion is solely concerned with such properties in verbal
forms, I call them 'verbal properties'.

3. The VAP-suffix for the General Completive Passive is exceptional,

in that it may be interrupted b y no: see §4.1.1. below.

4. Such extensions are not discussed in this study. See Arnott

(19 70: chs. 57-59).

5. I use 'negative1 rather than 'negation', because the latter

implies a process of negating that only sometimes takes place
in Fulfulde. Moreover, 'negative' and 'positive* are more exactly
equivalent to each other than 'negation* and 'positive*.

6. In Kaceccereere Fulfulde, it is possible for two -it— extensions

to co-occur, the one with reflexive meaning, th^bther with
iterative meaning, thus: Muusa femmb-it-it-ake 'Musa shaved
himself again'. See also Arnott (1970: §59'.2.),.

7. The vowel of maay- is shortened because of -k i n - . See §1.1.3.1.

above.

8. See §2.1.2.3. concerning the problematic morph-syntactic status

of _|_e_ and cTon in non-verbal sentences. The same problems arise
with n o .

9. However, a focussed 1st and/or 2nd person subject is realised by

an independent pronoun outside the verbal complex (cf. p. Ill).

10. *Be here represents all 3rd person plural dependent pronouns.

11. Concerning SV and VS order, see §4.1..

12. This rule means that an unfocussed pronominal object necessarily

precedes an unfocussed nominal object (since the former is in
the verbal complex), and can only be interpreted as indirect,
not direct, object.

13. The Stative may refer to a past or future state, as well as to

a present state. Without contextual clues as to time— reference
(such as time adverbials), and without the Anteriority-marker,
no, present time will be understood, as indicated by the trans-
laton of 184.a. - 186.a.

14. The other morphological distinction which Arnott mentions is the

presence or absence of *e. In Kaceccereere Fulfulde, the dist­
ribution of j_e in Stative (and in Progressive) complexes is as
described in §2.2.2.4., page 98 , above.
 173

Chapter 3: Footnotes (.continued)

15. The emphasis described here is not the same as 'focus* (see
chapter 6). Focus refers to the contrast between different
items that might fill the same slot in the syntagm: for
example/. Biauba .ngii-mi 'It's Bub a I s a w ' , where Buuba is
contrasted with another person, or other people. The emph­
asis effected by cfon in 189 .b. is more a matter of the contrast
between affirmative (positive) and negative: don emphasises
the affirmative nature of 1 8 9 .b.. Focus correlates seman­
tically with wh- questions, emphasis with yes/no questions.

16. I am grateful to Isa Pashi for drawing my attention to this

use of d o n .

17. By 'periphrasis’, I mean the use of and/or cfon. Concerning

the VAP-suffix for the General Completive Active, see footnote
8 on page 224 below.

18. The Negative Completive is also the negative counterpart of

the Relative Completive: see example 360.b., page 268.

19. The Negative Incompletive is also the negative counterpart

of the Relative Incompletive: s.ee example 361.b., page 268.

20. Tammbaari is borrowed from Hausa tamba, which refers to the

cereal, Eleusine corocana. The 'pure1 Fulfulde term is
cargaari.

21. The -d- extension is necessitated by the preceding focussed

adverbial expression, to m a a y o .

22. I use the term 'choice1 advisedly. The tree is intended as

a convenient way of representing the relationship between
conjugations and properties, and I do not want to commit ny-
self to the idea that it has any psychological reality.

23. The subject is realised in surface-structure only if the

Imperative complex is preceded b y the negating particle, t o .

24. As well as expressing a command to remain seated, 207.b.

may also express a command to repeat the action of sitting down.

25. In indirect commands, it is not possible to express a command

to continue an action, state or process. In other words, the
Subjunctive is equivalent to the 'ordinary1 Imperative, but not
to the Continuative Imperative.
 174
Ch.apt.er 4

Paradigms of Verbal Inflections:

Morphology

4.0. Introduction

This chapter is concerned with the morphology of the paradigms that

were established on syntactic and semantic grounds in the last chapter:
namely, the fourteen verbal conjugations, or ’tenses', each express­
ing a particular set of verbal properties.

The morphology of the conjugations is complicated, as may be

confirmed by reference to Appendix A (pp. 285 - 297 below) , or to
Arnott (1970: Part V) . Let us take the Relative Incomplete.ve as an
example. This conjugation (or more precisely, verbal complexes of this
conjugation) expresses the properties relative, positive, incompletive
and indicative (cf. §3.2.1., and the tree on page 166 above). If these,
plus active, middle or passive, are the only properties to be expressed
in a complex, then the complex consists quite simply of a verbal root
plus VAP-suffix, such as jog- and -otoo in 215 a . . -Otoo is characteristic
of the middle voice of this particular conjugation. (In this and the
following examples the subject, verbal root and object are marked 'S',
'V' and 'O' respectively.)

215 .a. Ngim cfume Muusa jog-otoo leggal?

for what Musa hold stick
S V VAP 0
'Why is Musa holding a stick?'

In this sentence, both the subject, M u u s a , and the object, leggal, are
NP's, and they are therefore not part of the verbal complex (cf. the
discussion in chapter 2). However, any unfocussed anaphoric pronouns
in the sentence are appended to the verbal root, and are therefore part
of the complex. Thus, by virtue of the pronouns 'o- and -ngal in 235 .b. ,

the verbal complex 1o-jog-otoo-ngal expresses 3rd person, Singular,

Subject and Object, as well as the properties mentioned above:'*'
 175

215.b. Ngim cfume 'o-jog-otoo-ngal?

for what he h old it
S V VAP 0
'VJhy is he holding it Ceg. stick)?'

'O is of the 1O nominal class, ngal of the Ngal class. However, pronouns
of other classes, such as ndu and ngel in 215.c. and d . , may also occur
in verbal complexes, if their referents are semantically suitable to
2
be subject or object of the root in the complex. (The root jog-, for
example, is unlikely to have a subject of a class with inanimate conno­
tations .)

215.c. Ngim cfume ndu-jog-otoo-ngal? d. Ngim cfume ndu-jog-otoo-ngel?

for what it hold it for what it hold it
S V VAP O S V VAP 0
'Why is it (eg. dog) holding 'Why is it (eg. dog) hold-
it (eg. stick)?' ing it (dim.)?'

In fact, thenumber of verbal complexes of a single conjugation based

on a particular verbal root may be as many as several hundred: the
total will depend on factors such as the semantic compatibility of
class—pronouns with the root, whether or not the root is transitive
(and thus whether the complex may-contain object pronoun(s) as well as
a subject pronoun), and whether the root combines with active, middle
and passive voices or only with one or two of these.

Verbal complexes containing a 3rd person pronominal

subject and object are fairly easy to describe: SVO (subject-verb-object)
order always occurs (as in all the above examples),and the VAP-suffix
has a standard shape, in this, instance -otoo. This order is also
characteristic of complexes with NP subject and object. Complexes
involving a subject that expresses 1st and/or 2nd person, or involving
the Anteriority-Marker, however, are considerably more difficult to
describe, for SVO, VSO or VOS order may be found, and the shape of the
VAP-suffix may be different. Consider the following, for example,
which are of the Relative Incompletive conjugation, middle voice, just
as are 21 5.a. - d. :

2l5.>.e. Ngim cfume njog-oto-cfaa-ngal?

for what hold you it
V VAP S 0
'Why are you (sg) holding it (eg. a stick)?'
 176

215. f. Ngim cfume njog-oto-maa-mi?

for what hold you I
V VAP 0 S
'Why am I holding you (sg)?'

215.g. Ngim cfume 'o-jog-oto-noo-ngal?

for what he hold it
S V VAP ANT 0
'Why was he going to hold it?'

Unlike in the earlier examples, the VAP— suffix in 215.e.-f. is -o t o .

Moreover, although 2l5.g. has SVO order, 215.e. and f. have VSO and
VOS order respectively.

The first aim of this chapter will be to establish the morphological

integrity of each conjugation* In order to do this, I shall show
which VAP-suffixes are peculiar to each conjugation, and account for
the types of variation in morph-shape and morph-order that have been
exemplified in the complexes of 215.. Once these crucial features
have been explained, I shall then point out various morphological,
morpho-phonological and morpho-syntactic patterns that characterise
either all complexes, or particular types of complex.

Section §4.1. will consider the morphological composition of

verbal complexes, paying particular attention to the order of morphs
within the complex. Section §4.2. describes VAP-suffixes in complex-
final position, and morphological variation that is conditioned b y the
verbal root. Section §4.3. then describes VAP-suffixes in non-final
position, and the principles that control the shape of the pronominal
and other morphs that follow the root in a complex. Section §4.4.
is concerned with some of the finer points of the morphology of
verbal complexes. Finally, in section §4.5., a classification of
conjugations is presented, taking into account the various patterns
that have been brought to light in the rest of the chapter.

4.1. The Morphological Composition of Verbal Complexes

In this section, I shall give a general outline of the morphology of

verbal complexes. Some more complicated issues arising from this
 177

outline will be deferred until after the discussion of the various

conjugations, ie. until section §4.4- b e l o w .

The verbal complex is essentially a syntactic structure, whose

core is the verbal root plus VAP-suffix. This core may constitute
a sentence in itself (as is Ngar-ee in 216.a . ) , or it may occur as
part of a sentence (as does rim-ii in 217.a.):

216.a . Ngar—eeI 217.a. Nagge r i m — ii.

come cow give
V VAP birth
S V VAP
'Come (pi) I '
*The/a cow gave b i r t h . 1

The verbal root may be extended (as is ngar—t— in 216 .b.) , and the
VAP-suffix may be zero (as in 2 1 7 ,b.):

216-b. Ngar-t_ - eel 217.b. Nagge rim - 0!

come cow give
V EXT VAP birth
S V VAP
'Come back (pi) I *
'The/a cow gave b i r t h 1 *

As we have already seen, the verbal complex often, however, contains

more than the root plus VAP-suffix: it may- contain unfocussed,
dependent pronouns (such as 'on- in 216.c. and nge- in 217.c . ) ;
it may contain the Anteriority Marker, no (as in 216.d . ) ; and it may
contain "e and/or cfon (as in 217.d.) :

216.c. To 1on-ngar-ee1 217.c. Nge—rim — ii.

you come give
NEG S V VAP birth
S V VAP
'Don't come (pi)I'
'It (cow) gave birth.'

216.d . 1On-ngar-ii - n o . 217.d. 'E-nge-cfon-rim - a.

you come it give
S V VAP ANT birth
S V VAP
‘You (pi) had come.1
'It (cow) is giving birth.'

Most of the examples given so far have SV order: in other words, the
subject, whether pronominal or not, precedes the verbal root. This
 178

is in fact the order found in neutral sentences. I shall therefore

first consider complexes where the subject precedes the root (see
section §4.1.1.), and then proceed to consider complexes where the
subject follows the root (see section §4.1.2.).

4.1.1.. SV Order

Not only do all neutral sentences have SV order (cf. 218.a. and b „ ) :
non-neutral sentences with a 3rd person subject also always have SV
order (cf. 219.a. and b., which contain a focussed NP and a question—
word respectively):

218.a. Hassan weetor - ay. 2 1 8 .b. 'O-weetor - ay.

Hassan arise late he arise late
S V VAP S V VAP
'Hassan gets up late.' ‘He gets up late.*

219.a. Wunngere 6e - must-ata. b. "Dume 6e - must-ata?

fruit of they suck what they suck
deleb-palm O S V VAP
O S V VAP
'What are they sucking?'
1 (It's) the fruit of deleb-
palm (that) they're sucking.'

Non-neutral sentences with min 'I and he/she/they' as subject also

always have SV order:

219.c. Wunngere min-must-ata.

fruit of we suck
deleb-palm
0 S V VAP
'It's the fruit of deleb-palm that we're sucking.'

However, non-neutral sentences with a 1st or 2nd person subject (incl­

uding *en 'I and y o u 1) often have VS order: this will be exemplified
and explained in §4.1.2. below.

With SV order, the object always follows the verbal root plus
VAP-suffix, unless it is focussed (as in 219.a . ) , or a question-word
(as in 219 .b.) . This is true regardless of whether the object is a
dependent pronoun within the verbal complex (as in 220.b.) or an NP
outside the complex (as in 220.a . ) :
 179

220.a. Sey godcfo wes - a dowroori.

then one riddle millet
S V VAP 0
“Then one must riddle millet.*

b. Sey godcfo wes - a - n d i .

then one riddle it
S V VAP 0
*1hen one must riddle it.*

If there are two objects, the indirect object (marked *10* in 221.)
always precedes the direct object (marked 'DO* in 221 1 :

221.a. Mi-hokk-ii £inngel r.am del ko mo yicf-i.

I give child ray girl what she like
S V VAP IO DO
*1 gave my daughter what she liked.1

b. Mi-hokk-ii- mo - cfum.
I give her it
S V VAP 10 DO
*1 gave it to her.'

The Anteriority-Marker generally follows the VAP-suffix, but

precedes any object pronouns or NP's:

222.a. Ndee min-nja6 - ay - no_ ceede?

when we receive money
S V VAP ANT 0
'When were we going to receive (the) money?*

b. Ndee min-nja£ - ay - no^ - de?

when we receive it
S V VAP ANT 0
’When were we going to receive it?'

However, the Anteriority-Marker follows, rather than precedes, the

suffix-pronouns - am ('me') and -e_ ('you (sg)'), as can be seen by
the order - am-no and -e-no in 223. and 224.. (The use of these
pronouns, and their suffixal status, will be discussed in §4.3.2. and
4.3.2.1- b e l o w ) .
 180

223. 1O-hoos-an -0 - am-no -nda .

lie take from me it
S V EXT VAP IO ANT DO
'He had taken it (eg. book! from me.'

224. Mi-'ard - in - t — e_ — no.

I precede you
S V EXT VAP 0 ANT
'I was going to guide you Csg). *

Moreover, the Anteriority-Marker does not follow the -aama VAP-suffix

for the General Completive conjugation, passive voice; rather, it
intervenes between the - aa and the -ma of this suffix, as can be
seen in 225.:

225. Gawri facc-it - aa - no - ma

corn spill
S V EXT VAP ANT VAP
'Corn had been spilt.1

Arnott (19 7 0 : 218, footnote 4) observes that this 'intrusion* of no_

into the VAP-suffix happens with other conjugations in certain
dialects. However, in Kaceccereere, it only happens with the
General Completive.

The position of _[_e and d o n was first mentioned in §2.2.2.4.. Quite

simply, 2®. j-s always the first morph in a complex; and cfon always
immediately precedes the verbal root:

226. a. Penndo 1e-cfon-das-oo. b. 'E-mo — cfon— das—o o .

Penndo drag she drag
S V VAP S V VAP
'Penndo drags herself 'She drags herself
along/crawls.' along/crawls.'

Thus, follows an NP subject (as in 226.a.) , but precedes a pronominal

subject (as in 226 .b.) . The position and status of Jjs and cfon in
complexes will be further considered in section §4.4. below.

Sometimes, the subject of a sentence with SV order is realised

as zero. The presence of the subject in deep-structure, and its position
before the verbal root, can be deduced from the fact that the initial
consonant of the root, if it is an alternating one, will vary between
 181

its continuant and prenasalised forms (cf. $1.2.1.) • For example,

there is no overt subject in the complexes of 216.a. and b., but the plural
nature of that subject is evident from the fact that the initial con­
sonant is prenasalised. If the deep-structure subject is singular,
then the initial consonant is a continuant, as in 227.:

216;. a . 0-Ngar - e e 1 b. 0-Ngar-t — e e I

come come
S V VAP S V EXT VAP
'Come Cpl) I 1 'Come back Cpl) 1*

3
227.a. -(A »^
0-War-0 b. 0-War-t - ui
come come
S V VAP S V EXT VAP
'Come (sg)!' 'Come back (sg) 1'

Zero subject is also found in sentences with a topic N P , where the

topic is co-referential with the deep-structure subject. Again, the
presence and the position of the subject are evident from the
initial consonant of the root, if it is of the alternating variety:
thus, 228.a. has a preceding singular subject, but 228.b. has
a preceding plural subject - though neither appears in surface—
structure:

228.a. Muusa, sey 0-'umm-ii, 0-wurt - ii, 0-yah-i luumo.

Musa then arise go out go market
TOPIC S V VAP S V VAP S V VAP
'Muusa then arose, went out, and went to market.'

b. Bi66e, sey 0— 'umm—i i , 0—mburt— i i , 0—nj ah —i^ luumo.

children then arise go out go market
TOPIC S V VAP S V VAP S V VAP
'The children then arose, went out, and went to market.'

I have marked zero subject in these three pairs of examples (216., 227.
and 228.), b u t I shall not normally do so> unless it is necessary
for the discussion at hand.

4.1.2. VS Order

VS order never occurs with a 3rd person subject, or with 1st person
(excl) subject. However it regularly occurs in complexes of the
 182

Relative Completive and Incompletive conjugations if they have a 1st

5
person singular, 1st person (incl), or 2nd person subject; and in
complexes of the Subjunctive, if they have a 1st person plural
(inclusive), or a 2nd person subject. These are exemplified in
229. - 231.:

229 .a . Kanko kinn - u - m i . Do o nj ipp—i i - 1e n .

him greet I here alight we
0 V VAP S V VAP S
'It's him I greeted.' 'It's here we (incl)
alighted.*

c. Rowaani te' - u - cfaa. d. Ndee ngar-0 - .'on?

last year marry you when come you
V VAP S V VAP S
'It's last year you (sg) 'When did you (pi) come?'
got married.'

230.a. Maawuri suB - ay-mi-nga. Ngele njaB - et - en?

next year choose I it which accept we
V VAP S 0 O V VAP S
'It's next year I'll choose 'Which shall we (incl)
it (eg. car).' accept?'

c. Moye njanng-in - t - aa? d. t>ume ngacf-ot - on?

whom read you what do you
0 V EXT VAP S 0 V VAP S
'Whom do you (sg) teach?' 'What will you (pi) do?'

231. a. 'Igga luggicf-in-0 - en Inglisiire meecfen.

better deep we English our
V EXT VAP S 0
'We h a d better 'deepen'/improve our English.'

Sey muny-0 - aa secfda.

must wait you a little
V VAP S

'You (sg) must wait a little.'/'You have to be a little

patient.1

c. Haani njah-0 - o n .
ought go you
V VAP S
'You (pi) ought to go.'
 183

The complexes in 229. are all of the Relative Completive conjugation,

those in 230. are of the Relative Incompletive, and those in 231. are
of the Subjunctive. Note that a Subjunctive complex with 1st person
singular subject has SV order, unlike the other two conjugations; thus,
231.d. below may be contrasted with 229.a. and 230.a.:

231.d. Haa mi-hum - a to.

must r think yet
S V VAP
’I must think a while.'/'Let me think.'

One of the characteristics of pronouns that express 1st and/or

2nd person subject is the fact that they may not be deleted - the
only exception is in Imperative complexes Ccf. 216. and 227. on page
181 above). Thus, since VS complexes only occur with 1st and/or 2nd
person subject, zero subject never occurs in such complexes. For
example, a sentence with a 1st and/or 2nd person topic NP (such as
minn and 1aan in 232.a. and b.) that is coreferential with the subject
of a following clause or clauses will not have zero subject; and
the occurrence of -mi in 232.a. and of -cfaa in 232.b. may be contrasted
with the occurrence of 0 in 228. on page 181:

232. a. Minn, sey sipp-u - mi_ kosam, sey soor-id - 0 - mi~tfam.

me then hawk I milk then sell I it
TOPIC V VAP S O V EXT VAP S 0
* (As for) me, then I hawked milk and sold it all.'

b. 'Aan, sey nguj j- u - cfaa geroogal, sey ndogg-u - cfaa .

you then steal you chicken then run you
TOPIC V VAP S O V VAP S
' (As for) you (sg), you then stole a chicken and ran
away. 1

Normally, the object of a VS complex follows the subject. If it

is pronominal (such as -dam in 232.a . ) , it is the last morph in the
complex; if it is an NP (such as kosam in 232.a . ) , it immediately
follows the complex. Here are some further examples.

233.a. Keenya mball-u - mi- b e .

yesterday help I them
V VAP S 0
'It's yesterday I helped them.'
 184

233.b. Keenya mball-u - mi paanyo maacTa.

yesterday help I young man your
V VAP S 0
'It's yesterday I helped your boyfriend.1

As with SV order, if there are two objects, the indirect object ('10'
below) precedes the direct object {'DO1 below):

234.a. Sey nyett-an - 0 - m i - 'on kosam s i s i .

then scoop for I you milk 5 kobos
V EXT VAP S 10 DO
'Then I scooped up 5 kobos' worth of milk for you (pi).'

b. Sey nyett-an - 0 - m i - 'on-cCam.

then scoop for I you it
V EXT VAP S 10 DO
'Then I scooped it up for you.'

However, if a complex of the Relative Completive or Relative Incompletive

has a 1st person singular subject together with an object that is either
2nd person singular or 3rd person singular (personal class), then,
exceptionally, VOS order occurs.^ This may be seen in the complexes
of 233.c. and 234.c.:

233.c. Keenya mball-u - moo-mi.

yesterday help him I
V VAP 0 S
'It's yesterday I helped him.*

234.c. Sey nyett-an - 0 - maa-mi kosam sisi

then scoop for you I milk 5 kobos
V EXT VAP 10 S DO
"Then I scooped up 5 kobos' worth of milk for you {sg).'

The VOS order in these complexes contrasts with the more normal VSO
order found in 233. and 234.a. and b..

4.1.3. The Morphological Composition of Verbal Complexes: Conclusions

We have seen in this section that verbal complexes are syntactic structures,
and that around the core of a complex a cluster of pronouns and other
morphs may be found, whose order is determined by the internal syntax
 185

of the complex. We shall return to these matters in §4.4 below.

Let us now turn to a consideration of the issues that determine the
form of the morphs within a complex.

4.2. VAP-Suffixes in Complex-final Position

Each voice of each conjugation has a characteristic VAP-suffix.

However, the form of the suffix.is in many instances not completely
stable, for it may be influenced by the verbal root, and also by
the morphs that follow it in the complex. In this section, I shall
give an account of the basic, or full, forms of VAP-suffixes,
and show such variation as is due to ‘sensitivity1 to the verbal
root. In the following section (§4.3.), I shall consider the forms
of VAP-suffixes that are not i n complex-final position, and that may
7
therefore be 'sensitive' to following morphs.

The VAP-suffixes of the various conjugations are presented in

table XXI. The suffixes have been arranged in such a=;way $s to
highlight the patterns of similarities between conjugations. Thus,
the staggering of the suffixes in the column under Completive Aspect,
Active Voice, brings out the fact that most of these suffixes share
the feature i^, which is sometimes preceded by aa, and sometimes
lengthened to ii_ or ^i. Similarly, the staggering of the suffixes
in the column under Incompletive Aspect, Passive Voice, brings out
the fact that these suffixes share the feature e e . These patterns
will be further discussed after the table. It should also be noted
that since the two Imperative conjugations have different VAP-suffixes
for singular and plural subject, they are presented separately. This
is not necessary for other conjugations, where the basic VAP-suffix is
the same, regardless of the number of the subject.
 186

Table XXI

VAP-Suffixes in Complex-final Position

Completive Aspect Incompletive Aspect

Act Mid Pass Act Mid Pass

Emphatic (u) i a Subjunctive a oo ee

Stative i ii aa Progressive a oo ee
Relative i ii aa Relative (a) ta (o) too (e) tee
Negative aayi aaki aaka Negative (a)taa (a)taako (a)taake
General ii ake aama General ay (o) to (e) te
Exhortative (u) oo - Vague (u)ma ooma eema
1Ordinary1) Continuative
Imperative;* Imperative
Singular (u> a - Singular (a).tay (a) tay
Plural ee ee Plural (e) tee (e) tee
'

Several points need to be made concerning table XXI- First, the

occurence of the vowels in brackets will be explained in sections- §4.2.1.
and §4-2.2. below. Secondly, the placing of the Exhortative under Completive
Aspect is not entirely satisfactory: the -oo suffix of Exhortative Middle
is particularly anomalous, for it belongs, morphologically, with the Middle
VAP-suffixes for Incompletive Aspect. However, the. Exhortative expresses
a peremptory exhortation much like the 'Ordinary' Imperative expresses a
peremptory command (cf. §3.8.1.), and there are therefore no semantic or
syntactic reasons for placing it under Incompletive Aspect. Thirdly, I
have placed the Subjunctive under Incompletive Aspect because of its
use in narrative sequences (cf- §3.8.3.): morphologically, it also fits
well into Incompletive Aspect. Fourthly, where there is a small dash ('-')
in the table, this indicates that the conjugation concerned does not
occur in the passive. Fifthly, although I have shown a - (,e)te suffix
for the General Incompletive Passive, this form is in fact rarely used,
probably because it is homophonous with the VAP-suffix for the General
Incompletive Active (- (.e) t ) when it is followed by a 2sg pronoun (-e^:
see example 114. a. on page lQ3r and table XXIII on page 197 below. This
matter will be further discussed in §5.5'.2..
 187

Let us now consider the morphological patterns that emerge

from table XXI. Within Completive Aspect, I shall consider the top
five rows of suffixes, since the place of the Exhortative is question­
able, and since the Imperative suffixes (with the exception of (u)) do
not bear much comparison with the other suffixes of Completive Aspect.
Within these top five r ows, the suffixes are mostly 'variations on
a theme* of i_, ii_ and aa for active, middle and passive voices
respectively. For example, the suffixes for the Emphatic Completive
can be derived from those for the Stative and Relative Completive
by means of a vowel-shortening rule (which in the active results
in 0 or an anaptyctic la - see section §4.2.1. below); and those for
the Negative Completive can be phonologically derived from jL, i i ,
aa by means of a preceding negative marker, aa, and by replacing the first
element of the long vowels ii and aa by k - thus *aa-ii and *aa-aa
give aa-ki and aa-ka respectively. The only suffix that is not relatable
to this i_, ii_, an theme is the - ake suffix of the General Completive,
middle voice.

Within Incompletive Aspect, the suffixal patterns are slightly

more complicated. The suffixes can be described as ‘variations on a
theme1 of ja, oo and ee for active, middle and passive, but in addition,
there is a recurring t, which is replaced by y in final position
(compare the suffix for the General Incompletive, active voice). The
t/y. precedes the a^, o o , ee 'theme* in the Relative and Negative
Incompletive suffixes? and in the latter, we again see an aa
negative marker, with a following k and a short vowel, as was
observed for Completive Aspect. The only suffixes in Incompletive
that do not conform to the a^, oo, ee pattern are the (u)ma suffix
for the Vague Incompletive, active voice, where one might have
expected (a)ma? and the (a)tay and (e)tee suffixes for the Continuative
Imperative.

In spite of the exceptions mentioned in the last paragraphs,

I believe that table XXI offers interesting and useful corroboration
of the classification of conjugations that was established in the

last chapter. In particular, the morphological evidence of table XXI

 188

confirms the Completive vs. Incompletive distinction. Thus, all

the conjugations under Completive Aspect in the tree on page 166
have the i_, ii_, aa theme; and all the conjugations under Incompletive
Aspect have the a, oo, ee theme.

Let us now turn to the issue that was mentioned in the first
point after table XXI, namely the distribution of the vowels shown
in brackets in the table. The occurrence of these vowels depends on the
phonology and/or the morphology of the root to which the VAP-suffix is
appended.

4.2.1. Anaptyctic u .

The 1 (u)' shown at various points in table XXI represents an

alternation between u and zero. In these circumstances, it is con­
venient to regard zero as basic, and u as an anaptyctic vowel that
occurs in order to prevent the formation of an unacceptable syllable.
Exceptionally, u, also occurs in slow, deliberate speech, even though
it is not strictly necessary. Two types of unacceptable syllable
are in danger of arising in these circumstances; either a syllable
with C W C or CVCC structure (cf. §1.1.3.1.}, or a syllable ending
in a consonant that cannot be in syllable-final position. Let us
consider these matters in turn.

It was pointed out in §1.1.3.1. that regular syllable structure

in this dialect of Fulfulde is CVC, C W or CV. If the juxtaposition
of a root and VAP-suffix would result in a syllable of any other
structure, then the anaptyctic u occurs obligatorily. For example,
the b. sentences below are unacceptable because of the syllable
structure of the complex; whereas the a. sentences are acceptable,
because of the presence of u:

235. a. 'O - B o m - u 6iiko. 'o+6or+nu

she dress her child CV CVC CV
VAP

'She dressed her child (in a gown). 1

 189

235.b. *'O-&om-0 Biiko 1o+6orn

she dress her child CV *CVCC

236. a. Sari a yoor-uma cfoo. yoo+ru+ma

trousers dry here C W CV CV
VAP
'The trousers may dry here.'

b. *Sarla yoor-0ma cfoo. yoor+ma

trousers dry here * C W C CV

On the whole, the consonants that can be followed by a zero

morph, or oy a morph that begins with zero, are voiced continuants,
This fact is exemplified by the root-final -r and -m which I have
underlined in the following examples:

237.a. 'Be - ngar-0. 6e+ngar

they come CV C V C
VAP
‘They simply came.1

238.a. 10 - y i m - 0ma. ’o+yim+ma

he sing CV CVC CV
VAP
’He may sing.'

However, the following, where _u does occur, are also acceptable:

237.b. Be - ngar-u. 6e+nga+ru

they come CV C V CV
VAP

’They simply c ame.1

238.b. 'O-yim-uma. 1o+yi+mu+ma

he sing CV CV CV CV
VAP
'He may sing,'

These latter forms are more likely to occur in slow, deliberate speech,
If a root ends in a voiceless continuant (such as s_ of fus- in 239.),
or in a non-continuant consonant (such as & of fi6- in 240.), then
zero is unacceptable, and u is obligatory:
 190

239.a. *Taya fus^-0 nil.

tyre burst simply
VAP

b. Taya fus^-u nii!

tyre burst simply
VAP
‘The tyre simply b u r s t ! '

240.a. *Fi&- 0 boggol ngol.

tie rope the
VAP

b. Fib - u boggol ngol!

tie rope the
VAP
'Tie tsg) the rope!'

4.2.2. Vowel Harmony

Where a vowel other than u_ is shown in brackets in table X X I , this

indicates that that vowel alternates with zero in suffix-initial
position. Since the vowel concerned is always identical to the next
vowel in the suffix, we can say that a type of 'vowel harmony' is
operating. In the following discussion, I term this vowel the
'harmoni zing vowe1 1.

The harmonizing vowel, which only in fact occurs before the t_

that characterises many Incompletive VAP-suffixes, necessarily
occurs unless the verbal root is an extended root that ends in a

voiced continuant. For example, it occurs in the complexes of the

a. sentences below, but not of the b. sentences (I have underlined
the root-final consonant and the harmonizing vowel):
 191

241.a. Na'i n j a r - a t a a cfoo. b. Non min-ndef-ir-0ta-ndi,

cows drink not here thus we cook it
VAP EXT VAP
'Cows don't drink he r e . ’ "That's how we cook it
{eg. rice).1

242.a. Ndi-'un - ataako. b. Ndi- Vun - an - 0taake-ma.

it pound it pound for you
VAP EXT VAP
'It can't be pounded.' 1It w o n 't be pounded for
you (sg).'

243. a. Kanjum wacf - i 'o-wo^-uta.. b. Be - ngar - iloy_ - 0taa.

that cause he cry they come early not
VAP VAP EXT VAP
"That's why he cries.' "They don't come early.'

The harmonizing vowel is not needed in the b.sentences, because the

verbal root is extended, and the extension ends in a voiced continuant.
In fact, the occurrence of a harmonizing vowel in this position is
unacceptable:

241.c. *Non min-ndef-ir - ata-ndi.

thus we cook it
EXT

242.c. *Ndi-'un - an - ataake-ma.

it pound for you
VAP

243.C. *Be - n g a r - iloy - ataa.

they come early not
EXT VAP

However, the harmonizing vowel is needed in the a. sentences because,

although their respective verbal roots have a final voiced continuant
(.-£/ -n, -y) , those continuants are not part of an extension. The
harmonizing vowel is similarly needed in the following sentences
because the verbal root in each complex ends in a non-continuant
consonant (£> of he£- in 241.) or in a voiceless continuant (f_ of
tuf- in 242.):

241. Maaroori he6-oto to Kagoro.

rice get in Kagoro
VAP
'Rice is obtainable in Kagoro.'
 192

242. Ndee cfi - tuf - etee?

when they vaccinate
VAP
'When will they (eg. cows) be vaccinated?1

Since the distribution of the harmonizing vowels is sensitive

to the morphological structure of the verbal root, we are faced
here with morpho-phonological conditioning, and not simply phono­
logical conditioning as was the case with anaptyctic u.

Now that the various suffixes shown in table XXI on page 1|86 have
been explained, let us turn to VAP-suffixes that are not in complex-
final position.

4.3. VAP-Suffixes in Non-final Position

The forms shown in table XXI can be considered to be the 'basic' forms of
these suffixes. However, some of them are sensitive to verbal properties
that are realised more peripherally in the same verbal complex* In other
words, they display what could be called 'forwards sensitivity1, since
their shape may be modified, or influenced, by following appendages in the
same complex. In this section, I shall first present an outline of
the suffix-alternants that are found (cf. table XXII), and I shall then
discuss the factors that affect the choice of a particular alternant for
a particular verbal complex. This discussion will raise two important
issues: the division of 'appendages' into clitics and suffixes (see
§4.3.2.), and the suffixal precedence hierarchy, which determines
the shape of certain suffixes when, they co-occur (see §4.3.3.).

Table XXII below shows the full range of VAP-suffix alternants.

Those on the left of the Active, Middle and Passive columns are the 'basic'
or 'full' suffixes that were shown in table XXI; and those on the
right are the 'modified* or 'reduced' suffixes that occur when part­
icular pronominal or other appendages follow the VAP-suffix in the same
complex. (Although ’full' and 'reduced* might seem the most suitable
terms for these sets of suffixes, I hesitate to use them because the
'reduced' form of the Progressive VAP-suffixes is sometimes longer
than the 'full' form. Henceforth, I shall therefore use the terms
'basic' and 'modified'.)
 193

Table XXII

VAP-Suffixes

Active Middle Passive

Basic Modi­ Basic Modi­ Basic Modi
fied fied fied

Completive
Aspect

Emphatic (u) i 0 a
Stative i (u) ii i aa a
Relative i (u) ii i aa a
Negative aayi aa aaki - aaka _

General ii - ake -
aama
Exhortative (u) oo o —
1Ordinary'\ _
"
Imperative/*
Singular (u) - a 0 — _

Plural ee ee

Incompletive
Aspect

Subjunctive a (u) oo o ee e
Progressive a (u) oo o ee e
ay (o) to (e) te
Relative (a) ta (a) t (o)too (o) to (e)tee (e) te
(e) t
(o) t
ay
Negative (a)taa - (a)taako - (a)taake —

General ay (a) t (o) to - (e) te -

(e) t
Vague (u) ma ooma eema
Continuative
Imperative / ‘
Singular (a)tay (a) t (a)tay (a) t - _

Plural (e)tee - (e)tee - - -

Certain points concerning the presentation of table XXII need to be

explained. First, the forms found under the ’basic' columns are the same
as those presented in table XXI, page 186. Second, the use of brackets
around a vowel in the 'modified’ columns has the same significance as
their use in table XXI and in the 'basic' columns here. Third, the use
of means either that no form exists for that particular combination
of properties (e.g. passive and imperative), or that there is no modified
alternant to the basic suffix (as is true of the VAP-suffix for the Gen-
g
eral Completive, for example ).
 194

Let us now turn to the factors that determine the occurrence of

the basic and modified suffixes.

4.3.1. VAP-Suffixes for the Negative Completive Active

If the reader refers to S3.6.1.2. (pp. 153-154 above), and to table

XXI/. he will notice that - aayi is the VAP-suffix for the Negative
Completive Active when the VAP-suffix is in complex-final position.
However/ when a dependent pronoun or the Anteriority-Marker occurs
in the complex, the VAP-suffix generally has the form -aa. For example:

243. a. Nde-dacf - aayi. b. Nde-dacC - a a - no.

it escape not it escape not
VAP VAP ANT

'It (eg. hare) didn't escape.' 'It hadn't escaped.'

244.a. 'A - l a y 6 -aayi wowru. b. 'A -lay6-aa- ndu.

you wipe not mortar you wipe not it
VAP VAP
'You didn't wipe the mortar.' 'You didn't wipe it.'

In the a. sentences of 243. and 244., the VAP-suffix is in complex-

final position, and has the form -aayi. In 243.b. it is followed by
the Anteriority-Marker, and in 244.b. by the pronoun -ndu; and in each
of these, the VAP-suffix has the form -aa rather than -aayi.^

4.3.2. The Clitic/Suffix Distinction

The terms 'clitic* and 'affix' and their variants (proclitic, en­
clitic, prefix, infix and suffix) are well established in linguistic
literature: see Matthews (.1974: 124 ff and 168 ff) , for example.
However, there is a certain amount of confusion surrounding the use­
fulness of the distinction, and the exact boundary between the two
concepts. One point that is agreed on isthat clitics are more inde­
pendent, more word-like, than affixes. Indeed, clitics are often
considered to be phonologically-reduced forms of 'words': for example,
t' and m' in French are reduced from te_ and m e . Clitics are more pos­
itionally mobile than affixes: for example, Latin -que 'and' may be
appended to any syntactically-appropriate constituent. Affixes, on the
other hand, in no sense constitute a word, for they are not positionally
mobile, and they are usually bound to a particular type of stem or root:
 195

for example, the tense-suffixes in Latin can only be appended to verbal

roots.

Now, while the extreme instances of clitics and affixes are

clearly identifiable, there are large grey areas: one of these has
recently been exposed in Harris (.1980) , which concerns the status of
Romance dependent 'pronouns. In the Fulfulde verbal system, there is
a certain grey area of items that straddle the clitic/suffix border;
but on either side, a distinct type of appendage may be identified,
one of which can usefully be labelled a clitic, and the other a suffix.
(I shall henceforth dispense with the term 'affix', since Fulfulde has
no infixes or prefixes. However, I shall continue to use ’clitic',
since Fulfulde has both proclitics, .and enclitics.)^

Two questions may now be posed. Of the appendages that occur in

a Fulfulde verbal complex, which are suffixes and which are clitics?
And how does the clitic/suffix distinction help the description of
Fulfulde verbal morphology? The answer to-the latter question will
in fact emerge during the discussion of the former.

The status of VAP-suffixes has been anticipated from the start: they
are bound to a verbal root, and cannot be appended to a nominal stem (cf.
§2.2.1.). They thus have no existence independent of the verbal root,
and cannot be considered clitics. As regards _|_e and cfon, they also
occur in non-verbal sentences (cf. §2.1.2.3.), without a verbal root,
and thus cannot be considered to be prefixed to the root; and there
can therefore be little doubt but that they are clitics.

The two other types of appendage within the verbal complex are more
problematic. I shall deal first with pronouns, and then with the Anter­
iority-Marker.

4.3.2.1. Pronouns

Certain pronouns that occur in the verbal complex also occur in other
syntactic environments: for example, in non-verbal sentences (cf.
246.a. belo w ) , and as head of an NP containing a genitival modifier or
a relative clause (cf. examples 133. - 136. on pp. 113 - 114 above).
 196

I shall consider such pronouns to be clitics, since they display a

certain amount of independence- Other pronouns, on the other hand,
only ever occur within a complex; and since they are in effect bound
forms, I shall consider them to be suffixes.

The contrast between clitic and suffix pronouns may be demon­

strated by comparing the distribution of ma and e_, each of which
realises the properties 2nd person, singular, and object. In 245.a.
below, ma is object within a verbal complex of the General Completive
Active; and in 245.b., e_ is object within a complex of the General
Incompletive Active. Unlike the sentences of 245., those of 246. are
non-verbal; and here we see that although ma may occur in a non-verbal
sentence (as object of the participle, janng-in-0-do in 246.a.), this is
not true of e.

245.a. Mi-janng-in - ii - ma. b. Mi-janng-in - t - e_.

I read you I read you
EXT VAP EXT VAP
'I taught you (sg).' 'I'll teach you (sg).'

246.a. Mi janng-in - 0 - cfoma. b. *Mi janng-in-ay-do e_.

I read you I read you
EXT VA CL
'I'm the person who taught
you (sg).1

Sentence 246.b. is unacceptable because e_ does not occur as the object

of a participle; in fact, in this environment, ma would be correct,
thus:

246.c. Mi janng-in -ay-do ma.

I read you
EXT VA CL

'I'm the person who will teach you (sg).'

If we apply the same criteria to other pronouns, we find that depend­

ent pronouns may be classified as shown in table XXIII.
 197

Table XXIII

Clitic and Suffix Pronouns

Subject Person/ Obj ect

Number

Clitics , Suffixes Clitics Suffixes

mi — 1 sg 'am/yam kam
mmi
am
1a cfaa 2 sg ma maa
aa e
'o, - 3 sg mo, moo
ngel etc ngel, etc.

1en en 1 pi incl 'en en

min - 1 pi excl min -

'on on 2 pi '■on on
be etc - 3 pi b e , etc. -

It should be noted that 'etc.* under 3sg and 3pl indicates that the
other dependent pronouns referring to other nominal classes are clitic
pronouns, just as ngel and 6e_ are: space does not allow them all to
be shown here. (In fact, the only class pronoun that is a suffix is
m o o .) It should also be noted that the 1am/yam alternative shown as the
1 sg object clitic pronoun indicates that 1am and yam are virtually in
free variation. The only occasion when there is preference for one
over the other is if the preceding VAP-suffix in the same complex ends
in -i^: in such circumstances, yam rather than 1am invariably occurs.

The most obvious issue that arises from table XXIII concerns
the distribution of the various pronouns in table XXIII where more than
one is shown for the same combination of properties of person, number
and case. However, dicussion of this issue will be deferred until the
next chapter (see in particular table XXVIII on pages 231 - 233).

The clitic/suffix distinction in pronouns is important because

suffix pronouns interact morpho-phonologically with the VAP-suffix of
the verbal complex, whereas clitic pronouns do not. The presence of one
of the suffix pronouns shown in table XXIII can, for example, influence
the shape of the VAP-suffix, and thus account for the distribution of
 198

some of the VAP-suffix alternants shown in table XXII on page 193 above.
This will be explained in §4.3.3. - §4.3.6. below. However, since the
presence or absence of the Anteriority-Marker in a complex can also
influence the shape of the VAP-suffix and of suffix-pronouns, it is now
time to discuss the clitic or suffix status of the Anteriority Marker
itself.

4.3.2.2. The Anteriority-Marker

The Anteriority-Marker is in fact another 'grey area1. It is not limited

to occurrence in the verbal complex and has considerable freedom of
occurrence (cf. §3.5.2.); by these accounts, it should therefore be
considered a clitic. Yet when it does occur in a verbal complex, it
often interacts morpho-phonologically with the VAP-suffix and with
suffix pronouns in a way that clitic pronouns do not; and by these
accounts, it should be considered a suffix. In this section, I shall
give some examples that illustrate the ambivalent status of the Anter­
ior! ty-Makrer .

The Anteriority-Marker occurs in non-verbal sentneces, as can be

seen in the following:

247. Koyij cfon no to ladde.

they present in bush
ANT

'They (small things/people) used to be in the bush.'

248. Joocf-ii-6e fu no haabe.

sit all non-Fulanis
VA CL ANT
'All the people who settled here used to be non-Fulanis.'
(ie. 'All the inhabitants (of this area) used to be
non-Fulanis.')

Sentence 247. is a non-verbal sentence expressing location, and sentence

248. is a non-verbal sentence expressing identification. Neither sentence
contains a verbal complex (joocf-ii-be. in 248. is a participle), and it
cannot be claimed that the Anteriority-Marker no is suffixed to a verbal
root in either sentence. No^ must therefore be considered to have at least
clitic status, if not word status.

By way of contrast, consider the following:

 199

249.a. Thame t a ’y-at-aa? b. T>ume ta'y-ay-no - cfaa?

what cut you what cut you
VAP VAP ANT
'What are you (sg) cutting?' 'What were you (sg) going
to cut?'

The verbal complex in 249.a. consists of a root ta'y- followed by a VAP-

suffix -at/ and the suffixed subject -aa. The complex in 249.b. is of
the same conjugation (Relative Incompletive/ Active voice), but in
addition/ the Anteriority-Marker is present. However/ the shape of
the verbal complex in 249.b. is very different from that in 249.a.. In
249.b. , we find -ay instead of -at, -cTaa instead of -aa, and no_ intrudes
between the VAP-suffix and the subject. The presence of the Anteriority-
Marker thus correlates with changes in the shape of other appendages
in the complex. It might even be argued that these changes in the other
appendages are secondary exponents of anteriority.

Since suffixes are integral parts of the word to which they are
appended/ whereas clitics have a measure of independence, we would not
expect to find a suffix in a more peripheral position than a clitic; and
since cTaa in 249.b. is a suffix rather than a clitic (cf. table XXIII),
we must conclude that no in 249.b. is also a suffix. Yet we have just
established the fact th.at.nc3 in non-verbal sentences is a clitic rather
than a suffix. It seems, then, that no straddles the clitic/suffix
distinction, and is part of the 'grey area' mentioned above.

It is now time to show why it is useful to consider no and certain

pronouns as suffixes within a verbal complex.

4.3.2.3. The Usefulness of the Clitic/Suffix.Distinction

In the verbal complexes of 249., we witnessed the fact that the real­
isation of certain verbal properties is 'sensitive' to the presence of
other properties in the same complex; for in this way, we can account
for the at/ay and aa/cfaa alternations in those complexes. In fact,
the realisation of all verbal properties (such as active voice, 1st
person, etc.) is sensitive to the presence of other properties in the
same complex, since there is no way of knowing exactly how one property
will be morphologically realised without knowing which other properties
are to be concurrently realised. (The subject of types of realisation is
broached in chapter 5.) Now, two types of variable are subsumed under

sensitive realisation: morph-shape and morph-order.

 200

Some morphs are variable in shape, whereas others are not. The
invariable ones are the verbal root, d o n , and all the morphs for which
only one form is shown in tables XXII and XXIII (pages 19 3 and 197).
'E is also invariable, except when it coalesces with a 2sg subject to
form 1a a . The variable morphs are the Anteriority-Marker, and all
those for which more than one form is shown in tables XXII and XXIII.
Note that not-all- morphs that vary in shape also vary in form, or vice-
versa: for example, although the VAP-suffix is constant in position,
it is variable in shape; and although the lpl (excl) morph is constant
in shape, it is variable in position.

Variation in morph-shape is explicable by reference to morph-

order, and to the 'suffixal precedence hierarchy'. The latter will
be discussed in greater detail in §4.3.3. below. In this discussion,
I shall be principally concerned with morph-shape as determined by
morph-prder, and how this relates to the clitic/suffix distinction:
or in other words, I shall be concerned with how the internal syntax
of a complex determines its morphology.

As an example of the influence of morph-order on morph-shhpe,

let us consider the realisation of the properties 1st person, singular
and object in a complex of the Negative Completive Active. Here, we
find that this morph is realised as kam if it is the first pronoun after
the root (as in 250.a. and 251.b . ), but that it is realised as *am if
it is not the first (.as in 250.b.). These facts contrast with the
properties 3rd person (personal class), plural and object, which are
consistently realised as &e in 250. and 251., regardless of position.

250.a. 'On-mball-aa - k a m . b. 'On-mball-an - aa - 6e - 'am.

you help not me you help for not them me
VAP EXT VAP
'You (pi) didn't help me.' 'You (pi) didn't help me on
their behalf.'

251.a. 'On-mball-aa - be. b. 'On-mball-an - aa - kam-b e .

you help not them you help for not me them
VAP EXT VAP
'You (pi) didn't help them.1 'You (pi) didn't help them
on my behalf.1

An indirect object precedes a direct object (cf. §4.1.1.), which accounts

for the order of object pronouns in 250.b. and 251.b . . What is significant
about these sentences is the fact that the Isg object has the form kam if
 201

it is the first object pronoun, regardless of whether it is a direct

object (as in 250.a.) or an indirect object (as in 251.b.). However,
when the lsg. object is the second object pronoun (as in 250.b.), it
has the form *a m . It is thus raorph-order, rather than direct or in­
direct object case, that determines the shape of a morph.

Where there is a choice of pronouns for the same set of properties

(see table XXIII), a suffix is more likely to occur as first rather
than as second pronominal morph after the root. This does not preclude
the occurrence of a clitic pronoun in first position - witness
in 250.b. and 251.b.. What it does mean is that a pronominal morph
in this position may be sensitive to bther properties realised more
'centrally' in the complex, namely the 1conjugational properties'
realised in the VAP-suffix, and the Anteriority-Marker. It is now
appropriate to describe the nature of this morpho-phonological inter­
action between suffixes that co-occur in the same verbal complex.

4.3.3. The Suffixal Precedence Hierarchy.

In Kaceccereere Fulfulde, the Suffixal Precedence Hierarchy Cor SP

hierarchy for short) operates in complexes of the following conjugations:

Relative Completive
Relative Incompletive
Stative
Progressive
Subjunctive

I shall initially illustrate the operation of the hierarchy by reference

to middle and passive complexes of the above conjugations. Its oper­
ation in active complexes of these conjugations will be dealt with in
§4.3.4..

In middle and passive complexes of the above conjugations, either

one long vowel or one geminate consonant may follow the verbal root.
This rule holds regardless of the phonlogical or morphological structure
of the verbal root. Since both these features involve length, either
of a consonant or of a vowel, I shall henceforth refer to the ’feature
of length' occurring in verbal complexes.

By way of example, consider the verbal complexes in 252. and 253..

The a. and b- sentences are well-formed, for they involve one feature
 202

of length after the root; whereas the c. sentences are not well-
formed, for they involve two features of length after the root.

2 52. a. Doo n job-i-cfaa-m o . b. Ooo n j og-ii-mi-cfe.

here hold you him here hold I them
VAP VAP
'It ';s here you (sg) held 'It's here I held them.'
him. '

c. *Doo n job -ii -cfaa-m o .

here hold you him
VAP

253. a. Jaango 1o-hokk-etee-cfe.

tomorrow he give them
VAP
'It's tomorrow he'll be given them.'

b. Jaango 1o-hokk-ete-noo-cfe.
tomorrow he give them
VAP ANT
'It's tomorrow he was going to be given them.'

c. *Jaango 'o-hakk-etee-noo-d e .
tomorrow he give them
VAP ANT

From examples 252. a. and b. , we see that the VAP-suffix for these complexes
of the Relative Completive Middle is sometimes -i- and sometimes -ii-:
this fact is also shown in table XXII on page 193 above. The short-
vowelled -t- occurs with the pronominal suffix -cTaa; but the long-
vowelled -ii- occurs with the pronominal clitic, -mi. 252.c. shows
that the long-vowelled -ii- and -daa cannot co-occur. From 253.a. and
b., we see that the VAP-suffix for these complexes of the Relative
Ihcompletive Passive is sometimes -etee and sometimes -e t e ; the former
occurs with the clitic pronoun -cfe; but the latter occurs with the
Anteriority-Marker, n o o . 252.c. shows that the long-vowelled -etee
and noo cannot co-occur.

Why is the VAP-suffix in 252. long-vowelled with cfaa but short-

vowelled with mi? And why, in 25 3., is the VAP-suffix long-vowelled
with de_ but short-vowelled with noo? The answer to these questions
is to be found by reference to the suffixal precedence hierarchy. This
hierarchy determines which suffix will win out when a potential clash of
 203

features of length occurs. From the sentences of 252., it is already

apparent that the 2sg subject cfaa wins out over the VAP-suffix: a
complex such as *njog-ii-cfa-mo does not occur. From the sentences of
253., it is apparent that the Anteriority-Marker wins out over a
VAP-suffix: a complex such as * 1o-hokk-etee-no-cfe does not occur.
Thus far we know, then, that VAP-suffixes are lower than a 2sg suff­
ixed subject and the Anteriority-Marker in the SP hierarchy. Further­
more, from 254.", we can establish the fact that the 2sg pronoun is
above the Anteriority-Marker in the hierarchy:

254.a. TJoo njog-i - no - cfaa-mo.

here hold you him
VAP ANT
'It's here you tsg) had held him.'

b. *Doo njog-i - noo-cfa - mo.

here hold you him
VAP ANT

We can therefore set up a preliminary version of the hierarchy as

follows:

Table XXIV

The Suffixal Precedence Hierarchy - 1

(.i) 2sg suffixed subject pronoun

(ii) Anteriority-Marker
(iii) VAP-suffix

By using the same criterion of which suffix cedes its feature of

length in the presence of another, we can establish the full version of
the hierarchy. In doing this, it must be remembered that certain
suffixes are mutually exclusive. For example, only one VAP-suffix may
occur per complex; and only one 2nd person pronoun may occur per
complex. (There are no reflexive pronouns in Fulfulde.) Certain
suffixes therefore occur at the same level in the hierarchy. It must
also be remembered that clitic pronouns do not participate in the
hierarchy, and that we are therefore only concerned with the pronouns
marked in the 'Suffixes' columns of table XXIII.
 204

Table XXV

Suffixal Precedence Hierarchy - 2

(i) Suffixed 2sg pronoun: m a a , cfaa, aa

(ii) Suffixed 3sg pronoun: moo
(iii) Anteriority-Marker: - noo
(iv) VAP-suffix: i/ ii/ f!^/ — • ee'
(o)too, (e)tee
(v) Suffixed Isg object pronoun: mmi

For the sake of completeness, VAP-suffixes expressing active voice

have been included in the above, although they will not be discussed
until the next section.

All the suffixes in table XXV have been cited in their basic forms.
The modified forms of the VAP-suffixes have already been given on the
right hand side of each column in table XXII (page 193) t but are
repeated in table XXVI below for convenience. The 2sg pronouns shown
above do not have modified forms, since they never cede their length
to another suffix; and it is debatable whether moo has a modified
11
form. Table XXVI, then, displays the basic and modified forms of
all suffixes that participate in the suffixal precedence hierarchy.

Table XXVI

Variable Suffixes participating in the

Suffixal Precedence Hierarchy

Basic Forms Modified Forms

(i) 2sg pronoun maa, aa, cfaa

11
(ii) 3sg pronoun moo^ mo
(iii) Anteriority-Marker noo no
12
(iv) VAP-Suffixes i (u)
ii i
12
a (u)
aa a
oo o
ee e
(o)too (o) to
(e)tee (e) te
(v) lsg pronoun mmi mi
 205

Before we go on to consider the operation of the SP hierarchy

with basically short-vowelled VAP-suffixes, here are some further
examples of the way it oeprates in middle and passive complexes of
the conjugations mentioned at the beginning of this section. The b.
sentences below are unacceptable because the SP hierarchy has been
violated.

255.a. Keenya ndams-i - moo-mi ceede.

yesterday lend her I money
VAP
REL COMPL MID
'It's yesterday I lent her money.'

b. *Keenya ndams-ii - mo - mi ceede.

yesterday lend her I money
VAP

256. a. Tummbude 'e-cfon-hipp - i_ - n o o .

calabash invert
VAP ANT
STATIVE MIDDLE
*The calabash was inverted/upside d o w n .'

b. *Tummbude 'e-dbn-hipp - ii - n o .
calabash- invert
VAP ANT

257. a. 'E-6e - cfon-suudr-ee-mi.

they hide
VAP
PROGRESSIVE PASSIVE
'They are (±n the process of) being hidden from m e . '

b. *' E-&e - cfon-suucf-e - m m i .

they hide me
VAP

258. a. Ton n jawcf-oto-cfaa.

there hasten you
VAP
REL INCOMPL MID
'It's over there that you (sg) will go fast,1

b. *Ton n j awcf-otoo-cfa.
there hasten you
VAP
 206

259. a. Dakkit - a, ngim to ngi'-e_ - cfaa!

get away so that not see you
VAP VAP
SUBJ PASS
'Get away (from here), lest you (sg) be seen!1

b. *Dakkit - a, ngim to ngi*-ee-cfaal

get away so that not see you
VAP VAP

The VAP-suffix in all the a. sentences except 257.a. is short-vowelled,

because m o o , noo and cfaa are higher in the hierarchy. In 257.a.,
however, the VAP-suffix has not ceded its long vowel, because mmi

is lower in the hierarchy. (See table XXVI).

4.3.4. Alternation between short-vowelled VAP-suffix and Zero VAP-suffix

In active complexes of the following conjugations, the SP hierarchy

operates in a slightly different manner from that just described for
middle and passive voices:

Subjunctive
Relative Completive
Stative
Progressive.

The 'basic' VAP-suffix for active complexes of these conjugations is

a short vowel (-i or -ja: see table XXII on page 193) ; and when that
suffix is followed by a suffix that is higher in the hierarchy, then the
former is reduced to 0, despite the fact that there is no danger of
two 'features of length' co-occurring. The alternation between the
short-vowelled VAP-suffix and 0 is exemplified in the a. and b. sen­
tences of 260. - 262. below. In the b. sentences, 0 occurs because
of a following 2sg pronoun or the Anteriority-Marker.

260.a. 'Igga 'o-wi'-a - mmi. b. *Igga 1o-wii-0 - m a a .

better he tell me better he tell you
VAP VAP
SUBJ ACT SUBJ ACT
'He had better tell me.' 'He had better tell you (sg).

261. a. 1E-koyq-mbe 1 - a - m m i . b. 1E-koyij-mbel - 0 - m a a .

they sweet me they sweet you
VAP VAP
PROGRESSIVE ACTIVE PROGRESSIVE ACTIVE
'They taste sweet to me-' 'They taste sweet to you (sg)
 207

262.a. 1E-mo-yicT-i_ - m m i . b. 1E-mo-yicC-0 - noo-mi.

he like me he like me
VAP VAP ANT
STATIVE ACTIVE STATIVE ACTIVE
'He likes me. 'He used to like me.*

If two basically long-vowelled suffixes co-occur, then the SP hierarchy

operates as normal, as .can be seen in 262.b. above, where the Isg pro­
noun is short (-mi) in the presence of the Anteriority-Marker (which
is higher in the hierarchy). Similarly, in 263., the Anteriority-
Marker is short (-no) because it is lower in the hierarchy than the
2sg pronoun:

263. Kan jum wacf - i min-'yam-0 - no - maa

that cause wre ask you
VAP► VAP ANT
REL COMPL ACT
'That's why we (excl) had asked you (sg).'

However, in active complexes of t h e .Subjunctive and Relative

Completive, zero VAP-suffix does not only occur in the presence of a
suffix that is higher in the hierarchy: it also occurs in any VS
complex, regardless of what follows the subject. Here are some
examples:

264. Haani taw - 0 - en agoogoyel maako.

ought find we watch his
V VAP S
SUBJ ACT
'We (incl) ought to find his wrist-watch.1

265. Doo ke6-0 - mi-ngel.

here get I it
V VAP S
REL COMPL ACT
'It's here I got it (dim.).’

266 . D u m e ngacf-0_ - 'on?

what do you
V VAP S
REL COMPL ACT
What did you (pi) do?'

It should be noted that as elsewhere in the verbal system of Ful-

fulde, anaptyctic u occurs as an alternant of 0, when syllable-structure
so requires. Thus, if the verbal root has CVCC or C W C structure, u
 208

necessarily occurs

267.a. 'E - m o - 1annd-u - m a a . b. * 'E-mo-1annd-0_- maa

he know you he know you
VAP VAP
STATIVE ACTIVE
'He knows you (s:g)

268. a. Ndee koot - u - 'on? b. *Ndee koot - 'on,

when go home you when go home you
VAP VAP
REL COMPL ACT
When did you (pi) go home?

The b. sentences of 267. and 268. are unacceptable, because the presence
of a zero VAP-suffix causes an unacceptable CVCC or C W C syllable (1annd
and koot respectively).

Two other conjugations display an alternation between a short-

vowelled basic VAP-suffix and zero, but in the middle rather than the
active voice. These are the Ordinary Imperative (singular only) and
the Emphatic Completive, where the short-vowelled suffix is -a and -i_
respectively. In these conjugations, although the alternation bears
comparison with that just described (hence its inclusion her e ) , it is
not in fact conditioned by the SP hierarchy. RAther, it is a matter
of phonology. If the VAP-suffix is to be followed by a vowel-initial
pronominal suffix, such, as -am 'me' or -e_ 'you' (sg), then zero always
occurs, as in the b. sentences below; otherwise, -a or -i_ occurs, as in
the a. sentences.

269. a. Yaaf - a.- b. Yaaf - 0 - ami

forgive them forgive me
VAP VAP
ORD IMPER MID ORD IMPER MID

'Forgive (sg) themI' Fogive (sg) me I'

270.a. Naa 'o - jog - an-i - *on-mo?

not she hold for you him
VAP
EMPH COMPL MID
'Did she not hold him for you (pi)?'

b. Naa 'o - jog - an-0 - e_ - mo?

not she hold for you him
VAP
EMPH COMPL MID
'Did she not hold him for you (sg)?1
 209

Regarding the distribution of vowel-initial pronominal suffixes, the

reader is referred to table XXVIII and the accompanying discussion
(pp. 231 - 233 below).

It should be noted that since the zero VAP-suffix alternant only

occurs with a vowel-initial suffix in these two conjugations, there
is never any 'danger' of an unacceptable syllable, and anaptyctic u
never occurs.

4.3.5. VAP-Suffixes of Incompletive Aspect

Of the alternations shown in table XXIX, most have been accounted for,
either in terms of vowel harmony, or in terms of the SP hierarchy.
Those that have yet to be accounted for involve a t or a y,.

In complexes of the General Incompletive Active y occurs in syllable-

final position, and t elsewhere. Thus, if the VAP-suffix is in complex-
final (and therefore syllable-final) position, or if it is followed
by a consonant-initial appendage, the -ay alternant occurs. This is
illustrated in 271.a. and 272.a.:

271.a. Sukaafoe ngeew-ay Hassan.

children miss Hassan
VAP
'Hassan will miss (the) children.'

272. a. B e - nyaam-ay - cfe.

they spend it
VAP

'They spend it (money).'

In 271.a., the VAP-suffix is followed in the sentence by an NP object,

which is outside the verbal complex; and -ay therefore occurs. Although
the object in 272.a. is pronominal, and therefore within the complex,
it is consonant-initial (~cfe), and therefore -ay still occurs. However,

if the VAP-suffix is followed by a vowel-initial suffix, then the

- (a)t or - (e)t alternant occurs. (See §4.2.2. above concerning the
distribution of the harmonizing vowels, a_ and e_.) This is exemplified
in the following:
 210

271.b. Sukaa&e ngeew-et - e.

children miss you
VAP

'You (sg) will miss (the) children.'

272.b. Be - nyaam-at - am.

they eat me
VAP
'They will eat me.'

In fact, the only vowel-initial suffixes that occur in this conjugation

are the ~e and -am seen in 271.b. and 272.b . .

The same principle operates in the Continuative Imperative Sing­

ular. Here the only vowel-initial suffix that occurs is -am 'me1 -
a second person object cannot occur in an Imperative complex. Thus,
we see the basic (a)tay VAP-suffix in the a. sentences below, but
the modified (a)t one in the b. sentences:

2 7 3.a . Raar-atay-mini
look us
VAP

'Keep on looking (sg) at us (excl) ! *

b. Raar-at - ami
look me
VAP
'Keep on looking (sg) at me!'

2 74.a. Jog-atay 6iyam!

hold my child
VAP
'Keep on holding (sg) may child!'

b. Jog-at - am!1
hold me
VAP
'Keep on holding (sg) m e ! '

The distribution of the VAP-suffix alternants for the Relative

Incompletive Active depends on whether the complex is an SV or VS

one (cf. §4.1.1. and §4.1.2.). In SV complexes with the Anteriority-

Marker, the VAP-suffix is -ay; otherwise it is -(a)ta.
 211

275.a. Raa no min-ndef-ir-ay-noo-ndi.

see how we cook it
S V EXT VAP ANT
'See how we (excl) used to cook it (eg.' rice). 1

276. a. Walaa to cfi - ndur-ata.

is not where they graze
S V VAP
'There is nowhere for them (eg. sheep) to graze.'

In VS complexes/ the same principle operates as has just been described

for the General Incompletive Active: y occurs in syllable-final
position, but t. (with or without the harmonizing vowel) occurs in
syllable-initial position. Thus:

275.b. Raa no ndef-ir - a £ - mi-ndi.

see how cook I it
V EXT VAP S
’See how I cook it (eg. rice).'

276.b. Walaa to ndur-o^t - on - cfi.

is not where graze you them
V VAP S
'There is nowhere for you (pi) to graze them
(eg. sheep).1

In 275.b . , the VAP-suffix is followed by the consonant-initial - m i ,

and -ay therefore occurs; whereas in 276.b., the vowel-inital -on
follows the VAP-suffix, and -ot therefore occurs. (The ^ of -ot
'harmonizes with1 the o_ of -on.)

As mentioned in §4.3.3. above, the SP hierarchy operates in complexes

of the Progressive. This accounts for the alternation between a_ and (u)
in the active, oo and o^ in the middle, and ee_ and e_ in the passive
tcf. the complexes in 257. and 261. on pages 205 and 206 above, for
example). However, the SP hierarchy cannot account for the occurrence
of the other type of modified VAP-suffix found in Progressive complexes,
namely -ay, - (o)to and (e)te (cf. table XXII on page 193). The latter
three VAP-suffixes in fact occur in active, middle and passive complexes
of the Progressive if the Anteriority-Marker is also present. Thus
the basic VAP-suffix occurs in the a. sentences below, but the second
type of modified VAP-suffix occurs in the b. sentences:
 212

211, a.. 'E-mo-semt - a_ - mmi. b. 'E-mo-semt - ay - noo-m i .

he respect me he respect me
VAP VAP ANT
PROGRESSIVE ACTIVE PROGRESSIVE ACTIVE
'He respects me/feels shy 'He used to respect me/
in my presence.' feel shy in ray presence
(but doesnlt any more).'

278.a. 'E-mo-hecf - oo 'esiiko. b. 'E-mo-hecf - oto-noo 'esiiko.

she await her in­ she await her in­
law law
VAP VAP ANT
PROG MID PROGRESSIVE MIDDLE
'S h e 's awaiting her in­ 'She was awaiting her in­
law. ' law. '

279. a. Faandu 'e-cfon-hebbin-e e . b. Faandu 'e—cfon-hebbin-ete-noo,

pot fill pot fill
VAP VAP ANT
PROG PASS PROGRESSIVE PASSIVE
'The cooking-pot is being 'The cooking-pot was being
filled.1 filled.'

In the Gombe dialect of Fulfulde described by A m o t t , these 'y/ti

alternants only occur in the Relative and General Incompletive, and not
in the Progressive (cf. his 1970: chapter 38). Complexes with -a-noo,
ro-noo or -e-noo would in fact be phonologically acceptable in Kaceccereere
Fulfulde, as indeed they are in Gombe. We m u s t ,therefore conclude that
the conditioning that causes the occurrence of the *y/t_T alternants in
the Progressive is morphological rather than phonological.

4.3.6. VAP-suffixes in non-final Position: Conclusions

From this discussion of VAP-suffixes in non-final position, it has

become apparent that both phonological and morphological conditioning
operate in the verbal complex. Some of this conditioning is effected
by pronominal appendages that occur in the complex, and these will be
more fully discussed under 'sensitive exponence' in the next chapter.
For the moment, let us now turn to some finer points concerning the
morphological composition of the verbal complex.

4.4. Verbal Complexes: Some Finer Points

In this section, I shall first show some instances where one finds SV
order instead of the VS order that might have been expected (see §4.4.1.);
 213

and I shall then appraise the place of _^e_ and cfon in the verbal complex

(see §4.4.2.}.

4.4.1. SV and VS Complexes

It will be remembered that in §4.1., I showed that certain conjugations

have VS order with a 1st or 2nd person subject, whereas others do
not. Now, although the VS order described there is normal in this
dialect, there are in fact occasions when the rule is broken. This
happens when there are so many appendages after the verbal root that
the complex is in danger of becoming uniterpretable; and it also
frequently happens with a lsg subject, regardless of the number of
appendages after the root.

4.4.1.1. VS>SV because of the number of appendages in the complex

On the whole, the maximum number of appendages that may occur after an
unextended verbal root is five. In everyday usage, it is probably
rare to find more than 3 or 4; and a Fulani wag might use as many
as 6 or 7; but beyond 5, the complex becomes difficult to interpret
for normal purposes. As many as three of the five may be object
pmouns (cf. 280.a.), and the same number may be extensions (cf. 281.a.):

280. a. Ngim cfume danyooBe maacfa suucf-an - ta-maa-mo - cfe?

because what parents your hide for you her them
S V EXT VAP O 0 0
1 2 3
'Why will your parents hide them from her on your (sg)
behalf?1

281.a. Ndee min-ngar-t_ - id - an - ay - noo Muusa 'e maakoyrj?

when we come back with, for Musa with them
S V EXT EXT EXT VAP ANT
1 2 3
'When were we going to come back with them for Musa?'

The above sentences contain SV complexes with 5 appendages after the

root. In VS complexes, the presence of the subject pronoun necessarily
reduces the number of other appendages that may occur. For example,
the following are equivalent to 280.a. and 281.a.; but they have a
lsg and 2pl subject respectively, with VS order, and are unacceptable
in normal usage (I have therefore marked them '?*'):
 214

282.a. ?*Ngim cfume suucf-an - ay-maa-mi-mo - cfe?

because* what hide for you I her them
of
V EXT VAP O S O O
'Why was I going to hide them from her on your (sg)
behalf?'

283. a. ?*Ndee ngar-t - id - an - ay - noo-'on Muusa 'e maakoyi)?

when come back with for you Musa with them
V EXT EXT EXT VAP ANT S
'When were you (.pi) going to come back With them for
Musa?'

When a Fulfulde-speaker 'senses' that an unacceptable complex may

occur/ two types of solution present themselves. Firstly, he may use

a prepositional phrase in the place of an indirect object (cf. 282.b.),
or in the place of an extension plus indirect object (cf. 283.b.):

282.b. Ngim cfume suucf-an - ay -maa-mi-cfe daga to maako?

because* what hide for you I them from at her
of
V EXT VAP O S 0
'Why was I going to hide them from her on your (sg)
behalf?'

283.b. Ndee ngar-t - id - a y - n o o - 1on 'e maakoyi} ngim Muusa?

when come back with you with, them for Musa
V EXT EXT VAP ANT S
'When were you (pi) going to come back with them for
Musa?'

In 282.b.,daga to maako has replaced the indirect object, m o ; and in

283.b., ngim Muusa has replaced the extension -an- plus its object,
Muusa (- compare 282.a. and 283.a. respectively). Such prepositional
phrases may well be used with SV complexes as well as with VS complexes;
for example, 280.b. and 281.b. are possible variants of 280.a. and 281.a.:

280.b. Ngim cfume danyoolbe maacfa suucf-an - ta - maa-de daga to maako?

for what parents your hide foryou them from at her
S V EXT VAP 0 0

'Why will your (sg) parents hide them from her on your
behalf?'

281.b. Ndee min-ngar-t - id - ay -noo 'e maakoyi}ngim Muusa?

when we come back with with them for Musa
S V EXT EXT VAP ANT
'When were we (excl) going to come back with them for
Musa.?'
 215

However, while the difference between the a. and b. sentences of

282. and 283. is one of interpretability (and therefore acceptability),
the difference between the a., and b. sentences of 280. and 281. is
one of style and idiolectal preference.

Secondly, where there is 'danger' of an uniitherpretable VS complex,

the Fulfulde speaker may choose to use an SV complex instead. Thus,
282.c. and 283.c. are variants of 282.b. and 283.b . :

282. c. Ngim cfume mi-suucf-an -ta -maa-cfe daga to maako?

because+ what I hide for you them from at her
of
S V EXT VAP O O
'Why was I going to hide them from her on your (sg)
behalf?'

283. c. Ndee 1on-ngar-t - id-an - ay -noo Muusa 'e maakoyi)?

when you come back for Musa with them
S V EXT EXT EXT VAP ANT
'When were you (sg) going to come back with them for
Musa?'

Since in such SV complexes, only four appendages occur after the root,
the prepositional phrase may be dispensed with, and a complex with five
appendages after the root may be used:

282.d. Ngim cfume mi-suucf-an - ta-maa-mo - cfe?

because* what I hide you her them
of
S V EXT VAP 0 0 0
'Why was I going to hide them from her on your (sg)
behalf?'

283.d. Ndee 'on-ngar-t - id - an - a y — noo Muusa 'e maakoyi)?

when you come back with for Musa with them
S V EXT EXT EXT VAP ANT 0

'When were you (sg) going to come back with them for
Musa?'

(In 283.d . , M u u s a , as the object associated with the extension -an-,

must precede the prepositional phrase 'e maakoyi).) Again, the difference
between the c. and d. sentences of 282. and 283. is one of style and
idiolectal preference.
 216

4.4.1.2. VS>SV with a lsg Subject

Although with a lsg subject, VS complexes do occur in relative conjugations,

SV complexes are in fact a frequent alternative. This is true regardless
of the number of appendages in the complex. Thus, both the a. and b.
sentences of 284. and 285. are acceptable:

284.a. 'E zaman Aliyu mi-dany-aa.

at time Aliyu I bear
S V VAP
*I was b o m at the time of A l i y u .1

b. 'E zaman Aliyu dany-aa-mi.

at time Aliyu bear I
V VAP S
11 was b o m at the time of Aliyu. '

285.a. Ko mi-'annd-i wondbn.

what I know is
S V VAP
'It's what I know.'

b. Ko 1annd - u - mi woncfon.
what know I is
V VAP S
'It's what I know.'

It seems, then, that a lsg subject 'tolerates' either VS or SV order

in a relative conjugation.

Relative complexes with a 2nd person subject are much less 'tolerant'
of SV order, and VS order is the norm. Thus, 286.b. and 287.b. are
perfectly understandable in the Fulfulde of the Kaceccere'e n , although
286.a. and 287.a. are more normal and more correct:

286. a. 'E zaman Aliyu dany - a - cfaa.

at time Aliyu bear you
V VAP S
'You were b o m at the time of Aliyu. '

b. ?'E zaman Aliyu 'a-dany-aa.

at time Aliyu you bear
S V VAP
'You were born at the time of Aliyu.'
 217

2 87. a . Ko ’annd - u - cfaa woncfon.

what know you is
V VAP S
1It'ss what you know-'

b . ?Ko 1a - 'annd - i woncfon.

what you know is
S V VAP
'It's what you know.'

It is worth noting that in some dialects, such as that of Adamawa, VS

order is never found. It is recognized and understood, but considered
laughable and old-fashioned. Maybe VS order will eventually disappear
from the Fulfulde of the Kaceccere'en too. In any event, the evidence
of their Fulfulde today shows that the greatest 'resistance' to SV
order is found in the presence of a 2nd person subject.

4.4.2. *E and cfon

In this section, I shall briefly defend my inclusion of J_e_ and cfon

within the verbal complexes of the Stative and Progressive conjugations.
In order to do this, I shall review the functions of _]_e_ and then cfon,
both outside and inside verbal complexes.

4.4.2.1. _|_E

Both in discussion and in morphological analysis, I have so far treated

1e as an identifiable morph in Kaceccereere Fulfulde. Other writers,
describing other dialects of Fulfulde, treat _|_e as part of the pronouns
which are typical of the Stative and Progressive conjugations. (As
far as I am aware, j_e_ does not occur in other dialects with locative
and existential sentences.) Amott, for example, speaks of 'subject
elements series 3', having the forms ’e-mi, 1e-' a , 'e - m o , etc. (Amott
1970: 194) ; and Gaden (1931: xviii) speaks of 'modification du prorom-
sujet'. In this section, I shall justify my treatment of ^_e_ in Kaceccereere
Fulfulde not as part of a subject pronoun series, but rather as a morph in
itself; and I shall then show why j_e should be considered part of the
verbal complex when it occurs with the Stative Progressive.

The reason why it is convenient, and indeed necessary, to consider

1e as a morph that is independent of the subject pronoun is that it does
not only occur witn a subject pronoun. Consider, for example, the following
 218

288. a. Mi-yii-mo 1e-mo-cfaan-ii.

I see him he sleep
VAP
II saw him asleep.'

b. Mi-yii Hawwa 1e-cfaan-ii.

I see Hawwa sleep
VAP
'I saw Hawwa asleep.1

289.a. *E-&e - 1annd - i - m m i ,

they know me
VAP
’They know m e . ’

b. Bi&Be 'e - 1annd - i - m m i .

children know me
VAP
1 (The) children know me.'

From these sentences, we see that if a Stative dr Progressive complex

has a pronoun subject, j_e_ precedes the subject (cf. ’e-mo in 288.a. and
1E-£e in 289.a. ); but if the complex has an NP subject, then _|_e_ occurs
alone. If we say that j_e_ is part of a series of subject pronouns,
then there is no way of accounting for the occurrence of without
a pronoun, as in 288.b. and 2 8 9 .b. above.

There are two reasons why it is necessary to consider jjs as part

of Stative and Progressive complexes, rather than simply as a 'floating1

particle that 'happens' to accompany such complexes. Firstly, J_e_
has very evidently coalesced with the 2sg subject pronoun j_a, in
Kaceccereere Fulfulde, giving 1a a ;

290.a. 'Aa-waaw-i Bir - ki na'i na?

you milk cows
VAP INF QU
'Can you milk cows?'

b. * 'E-'a waaw-i Bir - ki na'i na?

you milk cows.

Although 290-b. would be acceptable in the Gombe dialect (cf. the

siob ject pronouns given in A m o t t (1970: 194} ) r it is not in Kaceccereere
Fulfulde* Since the subject pronoun is part of the verbal complex, a
morph that has coalesced with the subject pronoun must also be part of the
 219

complex. And if an underlying j_e_, as in 290. a. is thus part of a complex,

it is reasonable, in terms of the Stative and Progressive morphological
paradigms, to consider J_e_ elsewhere as part of the verbal complex with
which it occurs.

Secondly, the position of j_e in relation to the subject of a

complex is variable. If the subject is an NP, the ^e_ follows the
subject:

291.a. Yaa'u 'e-n o k k - a nyiiri.

Ya'u scoop 'nyiiri'
S VAP
'Ya'u is 'scooping up' 'nyiiri' (food).'

However, if the subject is pronominal, then J_e_ precedes the subject:

291.b. *E-mo-nokk - a nyiiri.

he scoop 'nyiiri'
S VAP
'He's scooping up 'nyiiri' (food).'

This fact suggests that cannot be separated from the verbal complex
by an NP; and that must either immediately precede a verbal complex,
or be considered to be the intial morph in a verbal complex. Given the
cl o s e 'relationship between _^_e_ and subject pronouns (compare *aa in
290.a. above), the latter suggestion seems the more reasonable.

Another reason for considering J|_e to be part of the verbal complex

is the fact that without ^e_, the complexes in 291. would in fact be
interpreted as being of the Subjunctive:

292.a. Yaa'u nokk - a nyiiri.

Yaa'u scoop nyiiri
VAP
'Ya'u should scoop up 'nyiiri'.'

b. Mo-nokk - a nyiiri.
he scoop nyiiri
VAP
'He should scoop up 'nyiiri'.'

1E is, then, an essential part of the Progressive morphological paradigm.

It is similarly essential to the Stative paradigm:
 220

29 3. a. 'E-cfi - cfaan-ii.
they sleep
VAP
'Theylre asleep. 1

b. Sey cfi - cfaan-ii.

then they sleep
VAP
'Then they slept.1

The complexes in 293.a. and b. are distinguished by the presence or

absence of J^e: while that in 293.a. is of the Stative Middle, that in
293.b. is of the Relative Completive Middle.

Given its crucial role in identifying the conjugation of a complex,

there can be little doubt but that 'e , when it occurs in a verbal sent­
ence, is morphologically part of the complex.

4.4.2.2. Don

Just as I have so far treated J_e as a separate morph within a verbal

complex, so have I also treated cfon. Again other writers vary in their
treatment of cfon. A m o t t considers it to be part of a 'subject element
series 2' ( A m o t t 1970: 194-5); Noye considers cfon to be an adverb
which helps to form the 'Progressif (Noye 1974: 58); and Gaden, like
Arnott, considers that the 'Duratif1 uses a special modification of the
pronoun subject, which sometimes includes do or cfon (cf. Gaden 1931:
xviii). in this section, X shall show why it is not possible to consider
^on as part of a composite subject pronoun in Kaceccereere Fulfulde; and
I shall then show why it is useful to consider cfon as a separate morph
within Stative and Progressive complexes.

The arguments concerning the first point are parallel to those that
were used for J_e above. When the subject of a Stative or Progressive
verbal complex is an NP, then no pronoun subject occurs with cfon (cf.
294.b . ):

294. a. '£-6e - cfon-njar-a njumri.

they drink honey
VAP
'They are drinking honey.'
 221

294.b. Yim£e 1e-cTon-n j ar-a njumri.

people drink honey
VAP
1People are drinking honey.'

If cfon is simply part of a subject pronoun# how can it be accounted for

in 294.b, , where it (.and indeed 1e ) occurs without a subject pronoun?
In the Fulfulde of the Kaceccere'en at least# it is necessary to consider
cfon, like ^_e, as. a separate morph.

It is less easy to argue for the inclusion of cfon in the verbal

complex than it was for ‘'e. As Noye's analysis points out, cfon is
essentially an adverb. It means 'around1, 'present', 'in existence',
and it belongs to a paradigm of deictic adverbs that also includes cfoo
'here' and ton 'there' (cf. §2.2-2.4. of chapter 2). In fact, the
latter two adverbs can occur in the place of cfon with a verbal complex,
thus:

295. a. Tummbucfe 1e-cfon-keew-i.

calabashes fill
'The calabashes are full.'

b. Tumnibucfe 'e ton keew-i.

calabashes there fill

'The calabashes are there, full.*

c. Tummbucfe le cfoo keew-i.

calabashes here fill
'The calabashes are here, full.'

I have rather prejudged the issue under discussion, by hyphenating cfon

to ^e_ and the verbal root, but by not so hyphenating ton and cfoo. The
fact is that while ton and cfoo are deictic in 295.b. and c. , cfon is not
in 295.a.. D o n , although maybe at one time deictic, is now 'neutral'
in meaning, and has become intergrated into the morphology of the
Stative and Progressive verbal conjugations; whereas ton and cfoo
have retained their deictic meaning. The latter in fact seem to
'intrude' in the verbal complexes in 295.b. and c . , to the extent that
one could question the unity of the complexes; and this is why I have
not hyphenated ton and cfoo to j_e and keew-i. It may be that in the light
of evidence such as in 295., the definition of the verbal complex needs

to be revised. Or it may be that we simply have to accept that sentences

 222

such as 295.a- contain a periphrastic verbal form? and that such forms
have variants such as 295.b. and c. Cat least in this dialect), which
are a throw-back to the origin of the periphrastic verbal forms. These
matters are further discussed in McIntosh (1980).

4.4.2.3. 'E and ‘Don: Conclusions

In conclusion, while there is every reason to include within the

verbal complex, as a separate morph, the issue is not so simple in the
case of cfon. Nevertheless, since cfon is more "neutral1 than the other
deictic adverbs mentioned above, it is possible to regard the Stative
and Progressive verbal complexes with cfon as periphrastic verbal forms
including, rather than interrupted by, cfon.

4.5. The Morphological Classification of Conjugations

By way of conclusion to this discussion of verbal morphology, it will

be useful to classify the 14 conjugations according to some of the
features that have been discussed here. The features I choose are
the occurrence of VS complexes (with. 1st or 2nd person subject only),
and the occurrence of _|_e_ and cfon. Using these two criteria, the
classification given in table XXVII results:

Table XXVII

Classification of Conjugations

VS Order "E and cfon

Type 1 Vague Incompletive

General Completive - -
General Incompletive -
Negative Completive - -
Negative Incompletive - -
Emphatic Completive
Exhortative - -
Imperative -■ -
Continuative Imperative -
Type 2 Stative — +
Progressive — +
Type 3 Subjunctive +
Relative Incompletive + -
Relative Completive +
 223

Although the above classification has been established on the

basis of only two criteria, it correlates to some extent with, other
features of the various conjugations: for the suffixal precedence
hierarchy operates in complexes of conjugations of types 2 and 3, but
with the exception of the Exhortative, it does not operate in type 1
conjugations; and broadly speaking, conjugations of types 1 and 2
occur in neutral sentences, whereas those of type 3 occur in non­
neutral sentences (cf. §3.6. - §3.8. of the preceding chapter).

A corollary of the observations that have just been made is the

fact that conjugations of type 1 are morpho-syntactically and morpho-
phonologically simpler than those of types 2 and 3. Type 1 conjugations
always have SV order, and there is rarely any morpho-phonological inter­
action between the VAP-suffix and other suffixes. Since the SP hier­
archy and VS order do not occur in type 1 conjugations, it is much
easier to 'compose1 a complex of one of these conjugations than of
type 2 or 3 conjugations.

Let us now turn from the detailed description of verbal complexes

to a consideration of the ways in which they realise the verbal prop­
erties that were established in chapter 3.
 224

Chapter 4: Footnotes

1. The complex in 215.b. in fact expresses 3rd person and singular

twice, once each in the subject (1o) and the object (ng a l ) .
2. Concerning the connotations or meanings of the various nominal
classes, see $1.3.1..
3. The alternation between -0_ and -u_ as the VAP-suffix for the Imper­
ative Active is conditioned by the final consonant of the verbal
r oot. See f4.2.1. b e l o w .
4. The short-vowelled -i that accompanies yah- characterises the act­
ive voice of the Relative Completive, whereas the long-vowelled
-ii that accompanies 1umm- and wurt- characterises the middle
voice of the same conjugation.
5. Henceforth, I shall use the phrase '1st and/or 2nd person1 to
refer to 1st person singular, 1st person plural (inclusive), 2nd
person singular and 2nd person plural.
6. in using the abbreviations ’S V O 1, 'VSO' and 'VOS', I disregard the
distinction between direct and indirect objects, since it is immat­
erial to the discussion.
7. My use of 'suffix1 is rather ambivalent. Inasmuch as I use it to
refer to a surface-structure phonetic form, it is concrete; but
inasmuch as I use it to refer to an item with alternative realisa­
tions, it is abstract.
8. The basic VAP-suffix for the General Completive Active is -ii. How­
ever, when this - ii is followed by the lsg object pronoun, the final
element of the long vowel becomes a glide, resulting in the sequence
-iy-am. Alternatively, the y_ could be analysed as the initial con­
sonant of the pronoun, and the VAP-suffix as a short -i_, thus:
-i-y a m . I use the latter orthographic represenation here, but I do
not believe there is sufficient reason to claim that there is a mod­
ified VAP-suffix for the General Completive Active.
9. Exceptionally -aayi occurs before the 2sg pronoun -ma. See $5.5.2..
10. It should be noted that I do not use 'clitic1 to refer to a phonolog-
ically-reduced form of a word.
Mo° only occurs as first object in a VOS complex (see example 233.c.,
p. 184), i.e. with a lsg subject in a Relative conjugation. If it
co-occurs with the Anteriority-Marker in such a complex, it is short,
for example: Keenya mball-u-no-moo-mi 'It's yesterday I had helped
h i m 1. I therefore place moo above noo in the SP hierarchy. If a 2sg
and 3sg (personal) object pronoun co-occur in a VOS complex, which­
ever is the indirect object occurs first, with a long vowel, and the
other follows the subject, with a short vowel: Keenya mball-an-maa-
mi-mo 'It's yesterday I helped him for you', and Keenya mball-an-moo-
mi-ma 'It's yesterday I helped you for him.' Whether complex-final
-mo and -ma are modified forms of their pre-subject counterparts is a
moot point.
12. ' (u) ' represents an alternation between ii and 0_. See pp. 207-8.
 225

Chapter 5

Types, of Exponence

5.0. Introduction

This chapter will address the question: how are the verbal properties
that were identified in chapter 3 realised in a verbal complex? In
other words, what types of exponenoa, realisation of properties,
are found in verbal complexes?

The identification of types of exponence is no simple matter,

and I shall confine this brief discussion to the most obvious types.
Moreover, I shall ignore the distinction made b y Carstairs (1981)
between a type of exponence that is displayed b y all properties be­
longing to one category, and a type only displayed by certain prop­
erties of a category. (It will be remembered (cf. §3.3.) that
most properties are terms in one of several categories: for example,
active, middle and passive are terms in the category of voice.)

In the discussion that follows, I shall use the terms 'exponence'

and 'realisation' synonymously, though the latter is more readily
qualifiable (eg. in the phrase 'realisation of properties') than is
the former. Similarly, I use 'express' and 'realise' synonymously.

There are two broad types of exponence in Fulfulde verbal complexes.

The first, which I shall term syntactic exponence, is not identified
by Matthews or Carstairs. Its existence in Fulfulde is a function
of the fact that the Fulfulde verbal complex may be a syntactic stru­
cture, with subject and/or objectCs). Syntactic exponence is effected
by means of morph-order, and is discussed in §5.1..

The second broad type of exponence is morphological in nature,

ie. it depends on morph— shape rather than order; and it comprises
several of the types of exponence identified by Carstairs and Matthews.
It includes cumulative exponence, where two or more properties are
 226

consistently realised together (.see §5.2,); sensitive exponence/

where the realisation of a property is 'sensitive' to the presence
of other properties in the complex (see §5.3.); and extended exponencef
where a property is realised by more than one feature in the complex
(see §5.4.). What we can term a fifth type of exponence, homonymy,
will be discussed in §5.5..

5.1. Syntactic Exponence

The internal syntax of verbal complexes was discussed in §4.1. above.

The order of pronominal morphs within a complex is in fact the most
important exponent of case, though, morph-shape may also contribute
(see §4.3.2. . and §5.3.1.).

Three cases may be realised in a complex: subject, direct object,

and indirect object. Normally, a subject pronoun precedes any direct
or indirect objects, and an indirect object precedes a direct object.
These facts are illustrated in the following by the position of nga
and nth in the a. sentences, the underlined pronoun is subject, in
the b. sentences, it is direct object ('DO') and in the c. sentences,
indirect object (’I 0 ‘).

2 9 6 *a. Nga-nyaam-ataa.
it eat not
S VAP
'It (eg. donkey) w o n 1t/doesn't eat.*

b. Tokoye soor-at - aa - nga?

where sell you it
VAP S DO
'Where will you (sg) sell it?1

c. 'O-hokk-ii - nga-ndu.
he give it it
S VAP 10 DO
'He gave it (eg. donkey) it (eg. food).'

297.a. Nduu suudu Bad - ii - mi

this house approach I
VAP S

'It's this house that I approached.'

 227

297.b. Seese-seese ’o-£>ad - ii - mi.

gradually he approach me
S VAP DO
'It's gradually that he approached me.'

c. Ngim raaajum 'o- 6 ad - an - ii - m i - m o .

for that he approach for me him
S VAP 10 DO

’That's why he approached him on my behalf.'

Note that a subject precedes any objects in the clause regardless of

whether the complex is an SV one or a VS one. (The only exception to
this rule was mentioned in §4.1.2., and exemplified by sentences 233.c.
and 234.c. on page 184 above.) Note also that the verbal root and
VAP-suffix may or may not intervene between the subject and the object (s),
depending on the conjugation of the complex, and the person of the
subject (cf. §4.1.).

5.2. Cumulative Exponence

The properties of voice, aspect and polarity, and sometimes other

properties besides, are realised cumulatively in the VAP-suffix;
and the properties of person, number, and sometimes of case, are
normally realised cumulatively in dependent pronouns within the complex.
(Number may also be realised in other ways, such as by initial consonant
alternation, and b y the VAP-suffix of Imperative complexes; and case is
realised by morph-order (.see above) as well as by the form of pronouns.)

5.2.].. VAP-Suffixes

The fact that a VAP-suffix realises properties of voice can be demon­

strated by the contrast between the -ii, -ate and -aama suffixes in
298. a. - c., which express active, middle and passive respectively:

298. a. Mi-suucf-ii - koyij.

I hide them
VAP
'I hid them (.eg. kittens).'

b. Koy 13- suucf— ake .

they hide
VAP
'They hid (themselves).'
 228

298 .c. Koyj} - suucf-aama.

they hide
VAP
1They have been hidden.'

The fact that a VAP-suffix realises properties of aspect can be demon­

strated by contrasting the -a y and -oto suffixes of 299.a. and b.,
which realise incompletive aspect, with the -ii and -ake suffixes
of 289.a. and b . , which realise completive aspect. (The General In-
completive Passive is rarely, if ever, used, for reasons that are
explained in §5.5.2.. The commpnly used incompletive equivalent to
298.c. is Koyg suucf-etee-koyn, which is a non-verbal sentence con­
sisting of a subject pronoun koyg and a participle suucf-etee-koyq,
and means 'They are (little things that are) to be h i d d e n . ’.).

299. a . M i -suucf-ay-koyg.
I hide them
VAP
'I’ll hide them.’

b. Koy i)- suucf-o t o .

they hide
VAP
'They’ll hide (themselves).'

The fact that a VAP-suffix realises properties of polarity can be shown

by contrasting the VAP-suffixes in the following with those in 298.:

300* a • Mi-suucC-aa - k o y g .
I hide not them
VAP
'I didn't hide them.'

b. Koy i}- su ucf-aaki.

they hide not
VAP
'They didn't hide (themselves).'

c. Koyg - suucf-aaka.
they hide not
VAP
'They haven't been hidden.'

In 300., the -aa, -aaki and -aaka VAP-suffixes which express negative

polarity, contrast with the - i i , - ake and -aama VAP-suffixes in 298- <
■
which express positive polarity.
 229

Sometimes, properties, of other categories are also realised in the

VAP-suffix. For example, with a 3rd person subject, the only morpholo­
gical (ie. non-syntactic and non-intonational). exponent of the prop­
erties relative and non—relative is the VAP-suffix. (.In complexes
with a 1st or 2nd person subject, on the other hand, the properties
relative and non-relative are partially realised by the distinction
between SV and VS morph-order.i Here are some examples of VAP-suffixes
as exponents of these properties. {.Inverted commas, ", have been
placed before the syllable bearing contrastive stress in 301.b..)

3 0 1 -a. Asabe hadd-ake-n d e . b. Asa"be hadd—ii - n d e .

Asabe gird it Asabe gird it
VAP VAP
GEN COMPL MID FOCUS REL COMPL MID
Asabe put it (skirt/ ‘I t ’s Asabe who put it
wrapper) o n . 1 (skirt/wrapper) o n . '

The only morphological difference between these two sentences lies in the
fact that the one. has -a k e , the other ii_, as its VAP-suffix, and it
is this distinction that realises the properties non-relative (or gen­
eral) and relative respectively. (.The relative conjugation in 301.b.
is necessitated by the fact that the subject of the sentence, A s a b e ,
is focussed.)

5.2.2. Pronouns

Dependent pronouns always cumulatively realise properties of person and

number. These pronouns are morphologically indivisible, for it would
be impossible to say that one part realises person, and another part
number. Thus, though ngel and mi both, realise singular, and although
ngel and koyg both realise 3rd person, the members of each pair share
no morphological feature that could be identified as realising singular
or 3rd person respectively. (,3rd person pronouns such as ngel and
koyg also realise nominal class; but since we are here concerned with
verbal properties, not nominal ones, I shall not discuss this fact any
further.)

Where the realisation of properties of person and number varies

according to case (cf. 2sg in table XXVIII below), then case is also
cumulatively realised in a pronoun. For example, -cfaa always realises
 230

subject case, and -maa object case, though both also realise person and
number.

5.3. Sensitive Exponence

Given the fact that Fulfulde is an inflecting language, it could be argued

that the realisation of every property is sensitive to the presence of
other properties in the same complex. The preceding discussion of cumula­
tive exponence, for instance, presupposed the fact that the properties of
voice, aspect and polarity are sensitive to each other, as are those of
person and number.

Sensitive exponence is not only involved in the morphology, but also

in the syntax of verbal complexes. By way of example, in type 2 conjuga­
tions (cf. table XXVII on page 222), VS order occurs with 1st or 2nd person
subject, but SV order with 3rd person subject. Since each conjugation
realises a particular group of properties, we must conclude that the real­
isation of those properties is sensitive to person and subject.

I do not propose to discuss all the intricacies of sensitive expon­

ence. There are, however, two areas of sensitive exponence that do merit
elaboration here. These are the realisation of pronouns (cf. §5.3.1.) and
of anteriority (cf. §5.3.2.).

5.3.1. Pronouns

The realisation of properties of person, number and case is 'sensitive' to

other properties realised in the same complex; and it is this fact that
explains why in some instance^ several pronouns representing the inter­
section of the same group of properties are to be found in table XXIII on
page 197 above.

The most essential criterion for the distribution of the pronouns

shown in table XXIII is conjugation. For example, a 2pl subject pronoun in
the Subjunctive Active is always -on;but a 2pl subject pronoun in the Gen
eral Completive Active is always 'o n . However, in certain conjugations the
realisation of object case and of 1st and/or 2nd person is sensitive to the
occurrence of another set of such 'pronominal' properties in the same com­
plex. For example, in the Emphatic Completive Active, a 2sg object is real
ised as -e_ if it is the sole object:
 231

302. a. '0- loot-0 - e_.

she wash you
VAP
'She washed you (sg).. '

But if the syntax and semantics demand that the 2sg object be preceded
by another object, then the 2sg object is realised as -m a rather than
as -<a:

302 ,b. 'O - loot— an - 0 - am-ma. c. *' O - loot-an - 0 - am-e.

she wash for me you she wash for me you
EXT VAP
'She washed you (sg) on
my behalf.'

From 302.a. and b . , we see that a group of properties of person, number

and case may have various morphological realisations, even in complexes
of a single conjugation.

Table XXVIII below shows the distribution of basic pronouns over

the various conjugations. (It should be remembered that -mmi may be
reduced to -mi because of the SP hierarchy: cf. table XXVI, p. 204.)

Table XXVIII

Distribution of Pronouns by Conjugation

Per­ Num­ Case Pronoun Conjugations of Occurrence Restrictions

son ber

1 sg sub j mi All conjugations

ob j am Emphatic Completive Act/Mid Only occurs as

Relative Incompletive Act first object
General Incompletive Act in a complex.
Exhortative Act
mmi Relative Completive Only occurs as
Stative first object; and
Progressive never after a lpl
Subjunctive (incl) or -2pl sub­
Exhortative ject in VS complex
1am/yam All conjugations Occurs in all cir­
cumstances except
as mentioned above
 232

Table XXVIII (continued)

Per­ Num­ Case Pronoun Conjugations of Occurrence Restrictions

son ber

2 sg subj ’aa Progressive -

Stative
cfaa Relative Incompletive In Rel.lncompl.
Relative Completive A c t. only occurs
Subjunctive Mid/Pass in complexes
expressing anter­
iority .
aa Relative Incompletive Act In Rel.Incompl.
Subjunctive Act Act., only occurs
in complexes not
expressing anter­
iority.
1a All other conjugations -

ob j e Subjunctive Only occurs as

General Incompletive Act first object in a
Progressive Act complex.
Emphatic Completive Act
Exhortative
maa Progressive Only occurs as
Relative Incompletive first object in a
Relative Completive complex.
Subjunctive
Exhortative.
ma All conjugations. In circumstances
mentioned with e
and maa above, only
occurs as second
object in a complex.

sg subj, mo Stative -
pers Progressive
class
’o/mo All other conjugations

non­ ngel, All conjugations

pars etc.
class
obj, moo Relative Completive Only occurs with VOS
pers Relative Incompletive complex and lsg sub­
class ject .
mo All conjugations In conjugations men­
tioned with moo, only
occurs as 2nd object.
non- ngel, All conjugations
pers etc. !
j class
 233

Table XXVI IE (.continued)

Per­ Num­ Case Pronoun Conjugations of Occurrence Restrictions

son ber

1 mcl pi subj en Subjunctive Act In Rel-Incompl. Act.

Relative Incompletive Act only occurs in com­
plexes not expre­
ssing anteriority.
1en All conjugations except In Rel.Incompl. Act
Subjunctive Act only occurs in com­
plexes expressing
anteriority.

obj en Subjunctive Act. Only occurs as

Relative Incompletive Act first object in
a complex.
1en All conjugations In conjugations
mentioned with en,
only occurs as
second object in
a complex.

1 excl pl subj min All conjugations _

obj min All conjugations -

2 Pi subj on Subjunctive Act In Rellncompl.Act.

Relative Incompletive Act only occurs in com­
plexes not express­
ing anteriority.

1on All conjugations except In Rel .Incompl. Act.

Subjunctive Act only occurs in com­
plexes expressing
anteriority.

obj on Subjunctive Act Only occurs as

Relative Incompletive Act first object in
a complex.
1on All conjugations In conjugations
mentioned with on,
| i only occurs as
1 second object in
! a complex.
i
i
i
i i
3 jP-l | subj £>e, cfi, All conjugations !

1
; etc. 1 i
i j |
i | obj ; £>e , Cfi | All conjugations
i 1 etc - 1
i
 234
\

In considering table XXVIII the following points should be borne

in mind* Firstly/ - am and —mmi occur in free variation as lsg object in
Exhortative Active complexes, as do and -'maa as 2sg object in Progressive
Active, Subjunctive Active and all Exhortative complexes. Secondly, the lsg

obgects -1am and -yam are in free variation, except that -yam always
occurs after an i-final VAP-suffix.

From the sensitive realisation of properties of person, number

and case, let us now turn to anteriority, which is also sensitively

realised.

5.3.2. Anteriority

Hie realisation of anteriority is sensitive to conjugation. In con­

jugations of type 1 (cf. table X X V H o n page 222), it is always realised
as n o ; but in conjugations of types 2 and 3, its basic form is n o o ,
which may be modified to no. These facts are exemplified in the foll­
owing :

3 0 3 -a. Rawaandu nyaam-ay - no - ngel.

lion eat it
VAP ANT
'The lion was going to eat it.'

b. *Rawaandu nya am-ay-noo-ngel.

3 0 4 -a. Jooni-jooni ndu-nyaam-ay -noo-n q e l .

now now it eat it
F O C U S VAP ANT
1It was going to eat it immediately . 1

b. *Jooni-jooni ndu nyaam-ay -no-ngel.

The complex in 303*a. is of the General Incompletive Active conjugation,

with short no; but that in 304-a. is of the Relative Incompletive Active,
requiring long n o o . The b. sentences are unacceptable because the
Anteriority-Marker has been incorrectly realised for that particular
conjugation.

Since the suffixal precedence hierarchy operates in conjugations

of types 2 and 3, noo may be shortened to no in the presence of a 2sg
pronoun:
 235

304.c . Jooni-jooni ndu-nyaam-ay -no - m a a .

n ow now i t e at you
VAP ANT

‘It was going to eat you (sg) immediately.'

d. *Jooni-jooni ndu-nyaam-ay -noo-maa.

In 304-c. , no is shortened because.of the presence of maa in the complex;

3 0 4 -d. is unacceptable, because noo has not been thus shortened.

5.4. Extended Exponence

The most striking example of extended exponence is found in the Stative

and Progressive: this is dealt with in section §5.4.1.. However,
there are also instances of this type of exponence that are less regular
but nevertheless worth noting. For example, the choice of a particular
pronoun may help to realise the properties of a conjugation (see §;5.4.2.) .
Often, extended exponence involves two or more of the types of exponence
that have already been identified. For instance, raorph-shape and morph-
order may together realise the properties emphatic and relative, or
the properties of case (see §5.4.3.),. Finally, when morpho-phonology
and morphology help to realise a property, then this must surely be
regarded as extended exponence (see §5.4.4.).

5.4.1. Stative and Progressive

The exponents of stative and progressive are J_e_ and/or don, plus one
of a particular set of VAP-suffixes in a verbal complex. Here are
some examples:

3 0 5 -a. *E-cfi - cTon-sankit-ii 305 *a. Nyalahol 1e-cTon-cfaan-oo.

they scatter calf sleep
VAP VAP
'They are scattered-' "The calf is falling
asleep . 1

Without 1e/cfon, (cf. 3 0 5 -b. and 3 0 5 -b.), or with an incorrect VAP-suffix

(cf. 305.c. and 306* c.), the stative or progressive properties are not
realised:

3 0 5 .b. Di - sankit-ii 3 0 5 -b. Nyalahol cfaan-oo.

they scatter calf sleefyAP
VAP
'They scattered, ’The calf should sleep . 1
 236

305 .c . * 1E-cfi-cfon-sankit-ataake.
they scatter

306.c. *Nyalahol 1e-cfon-daan-aama.

calf sleep

Although 305.b- and 306 .b. are grammatical, their complexes are of
the Relative Completive Middle and Subjunctive Middle respectively,
rather than of the Stative or Progressive. Sentences 305.c . and 3 0 6 .Cm f
on the other hand, have no possible interpretation, for the -ataake
and -aama VAP-suffixes can never co-occur with 1e/dbn: they are
characteristic of the Negative Incompletive Passive and the General
Completive Passive respectively.

5.4.2. Pronouns

Where a choice of pronouns for the same person, number and case is ind­
icated in table XXVIII then the choice of one pronoun rather than another
implies a particular property. For example, since -mmi only ever occurs
in active conjugations, it necessarily helps to realise active voice.
(In the middle and passive voices, -mmi is always reduced to -mi,
since mmi is at the bottom of the suffixal precedence hierarchy:
cf. tables XXV and XXVI.) Moreover, since -mmi only ever occurs in the
active voice of certain conjugations, the choice of mmi also helps to
realise the properties of those conjugations. True, the choice of a
pronoun rarely, if ever, absolutely distinguishes one conjugation from
another - the VAP-suffix is the most important and consistent marker

of conjugation. But in a certain sense, one cannot avoid the fact that
the choice of pronoun is a secondary marker of conjugation, and that a
particular kind of extended exponence therefore operates in this way.
Here is an example of the contrast between -mmi and -k a m , both lsg
object pronouns.

307.a. "Kosam he'y - i - m m i . b. Ndiyam he'y - aa-k a m .

milk suffice me water suffice me
VAP VAP
'There’s enough milk for me ’There's not enough
(but not enough of some­ water for m e .
thing else). '

The complex in 307.a. is of the Relative Completive conjugation (nec­

essitated by the focussing of kosam 'milk’) , whereas that in 307.b. is
of the Negative Completive conjugation. Now, is there any reason for us
 237

to say that the -i-/-aa- distinction between 307. a, and b . is any

more important than the -mmi-/-kam distinction, for the establishment
of these respective conjugations? Where the pronouns are so distinct,
I believe that they must be considered to be exponents of conjugation,
together with the VAP-suffix; and we must say that we have here an
instance of extended exponence,

(I do not believe that the above argument is invalidated by the

fact that -i and -mmi in 307.a., and -aa and -kam in 307.b . , are
contiguous suffixes; for these are separate, identifiable, morphs,
each being replaceable by another member of the same pronominal or
VAP-suffix paradigm. For example: "Kosam h e fy-i-6e 'There's enough
milk for t h em.1 and 'Igga kosam he 'y-a-mmi 'There'd better be enough
milk for m e ’ . In the former, -mmi has been replaced by 6e_ 'them',
and in the latter, -i has been replaced by the Subjunctive Active
VAP-suffix, -ei.)

5.4.3. Morph-Order

We have already seen, in section §5.1., that morph-order alone can

determine case. However, where more than one pronoun realises the
same person and number, then case may be determined by both order
and shape: and we thus have an instance of extended exponence. For
example, neither -mmi nor -kam in 307.a. and b. above can be a subject
pronoun; and object case is realised by their shape, as well as
by the fact that they occur after the verbal root, in a complex which
is preceded by its subject (kosam and ndiyam in 307.a. and b. resp­
ectively) .

Here are two further examples of the realisation of subject and

object case by both the shape of a pronoun, and by its position. In
308,a., the subject is j_a, which precedes the verbal root, winnd-an-
'write to'; and in 308.b. , the object is ma, which follows the verbal
root. Both these pronouns realise 2nd person and singular.

308-3-* 1A - winnd-an - ii-mo? b. 10-winnd-an - ii-ma?

you write to him he write to you
EXT VAP EXT VAP

'Have you (sg) written to h i m ? 1 'Has he written to

you (sg)?'
 238

If the shape of these pronouns is retained, but their position ex­

changed, unacceptable complexes result:

309. a. *Ma - winnd-an-ii-mo? b. *'O-winnd-an-ii-'a?

you write to him he write to you

Nevertheless, if a learner of Fulfulde, or a Fulani child, were to

say these unacceptable sentences, there is a good chance that
they would be interpreted as their equivalents in 308 : namely, 309 -a.
would be interpreted as meaning 'Have you written to h i m ? ', and 309 .b.
as 'Has he written to you?' This suggests that order in fact overrides
shape in the interpretation of Case. This whole question needs much
further research- For the moment, what is important is the fact that
case may be realised by both and shape, and that we therefore have
another instance of extended exponence.

There is another area where morph-shape and morph-order combine

to realise properties. Consider the complexes of 310 . : that in
310. a . has 0 VAP-suffix and VS order; whereas that in 310 .b. has
- ii VAP-suffix and SV order. The former is of the Relative Compl­
etive Active, and the latter of the General Completive Active.

310.a. Keenya ngar-jtf- 'on. b. 'On—ngar—ii keenya.

' yesterday come you you come yesterday
V VAP S S V VAP
'It's yesterday you (pi) 'You came yesterday.'
came. '

Is there any reason why morph-order should not be considered as much

an exponent of the relative and non-relative properties in these complexes
as are the 0 and - ii suffixes? If the answer to this question is 'no',
as I believe it is, then we must conclude that morph-order and morph-
shape may both be exponents of relative and non-relative, and constitute
an instance of extended exponence.

5.4.4. Consonant Alternation

One further type of extended exponence is worth mentioning. This concerns

the realisation of the properties singular and plural in a complex
based on a root whose initial consonant alternates. This consonant
is sensitive to the number of a preceding subject (cf. §1.2.1.);
 239

and if that subject is in fact a pronoun within the complex (rather

than an NP) , then it is arguable that that complex realises number
both in the pronoun and in the initial ocnsonant of the root. Con­
sider the complexes of 311. :

311.a . Nge1-woy-ataa. b. Koyrj-mboy- a t a a .

it cry not they cry not
VAP VAP
'It (.diminutive) doesn't: 'They (diminutive)
cry. 1 don't c r y .'

Is not the number of the subject of these complexes partially realised

by the initial consonant of the verbal root (w- and m b - ) , as well as
by the form of the ngejl and koyq pronouns? If yes, then this is
surely an example of extended exponence.

5.5. Homonymy

Homonymous exponence is fairly rare in Fulfulde verbal morphology. It

can be divided into two main types. Firstly, there are instances where
the morphological distinction between two properties is consistently
and systematically neutralised in a particular conjugation. I shall
call this 'syncretism', and it will be discussed in §5.5.1.. Then,
there are instances of what can be regarded as 'accidental' (as opposed
to 'systematic') homonymy, where the morphological distinction between
two or more properties is only sometimes lost in a particular morpho-
syntactic environment. Some instances of 'accidental, homonymy' are
deliberately avoided or disambiguated by Fulfulde speakers, and these
will be discussed in §5.5.2.. Others are automatically disambiguated
by context and they will be dealt with in §5.5.3..

5.5.1. Syncretism

Consider the following Imperative Singular complexes:

312. a. SuucT-va- mo! b. Suucf-a_- mo!

hide him hide him
VAP VAP
'Hide (sg) h i m ! ’ Hide (sg) from h i m ! '

Here, the distinction between active and middle voice is maintained by

the -u and -_a conjugation— suffixes, (There is no passive voice in
Imperative conjugations.) However, there is only one plural equivalent
 240

of 312.a. and b . :

313. a . Suucf-ee-mo I
hide him
VAP
'Hide h i m ! ' / 1Hide from h i m ! 1

In other w o r d s , Active and Middle Imperative complexes with a Plural

Subject both have -ee VAP-suffix. The ambiguity of 313.a. has to be
resolved either by context, or by the addition of a further phrase
or clauser as in 313-h. and c.:

313.b. Suucf-ee-mo, to 'o-yi'-ee!

hide him lest he see
VAP VAP

’Hide him, lest he be seen ! 1

c. Suucf-ee-mo, to 'o-yi'-a- 'on I

hide him lest he see you
VAP VAP

'Hide from him, lest he see you (pi)!'

In the Confirmative Imperative, the active - middle distinction

is realised neither with a singular nor with a plural subject. Thus,
the - atay and - etee conjugation-suffixes in 3 1 4 .a. and 3 1 5 .a. express

either active or middle voice:

314. a. B o m -atay-ngel!
dress it
VAP
'Keep on dressing him/her (eg. a child)!*/
'Keep on putting it on (eg. a little gown ) ! 1

315. a. Born ~ etee—koyiji

dress them
VAP 4

'Keep on dressing them (eg. children)!'/

'Keep on putting them on (eg. little gowns)!'

(I believe that it would not be useful to say that voice is not realised
in such complexes: if this were so, then surely no ambiguity would

arise.)

The neutralisation of the active - middle voice distinction in

Imperative complexes (except those with a singular subject in the
 241

'Ordinary' Imperative) is regular and systematic - rather!like- the

neutralisation of dative and ablative case in the plural of Latin
nouns. Despite this fact/ however, ambiguity rarely arises,, for

usually there are contextual clues as to the voice realised in a complex.

For example there are few pronouns that can refer either to a personal
or a non-personal referent. Ngel and koyi) in 314.a. and 315.a. are
amongst those few, and the choice of a different pronoun (as in 314.b.
and 315.b.), , or indeed of an NP object (as- in 314.c. and 315. c.), implicitly
realises the voice of Imperative complexes:

314.b. B o m — at ay—m o ! 315.b. Born — etee-cfe I

dress him/her dress them
VAP VAP
ACT MID
'Keep on (sg) dressing 'Keep on Cpl) putting
him/her!' (gowns) o n !'

314. c . B o m - atay binngel m a a ! 315. c , Born - etee toggooje d e e !

dress child your dress gowns these
VAP VAP
ACT MID
'Keep on (sg) dressing your 'Keep on (pi) putting
child!' on these gowns!'

Moreover, there are few verbal roots that sometimes take active
voice and sometimes middle. Suucf- and B o m - are amongst those few.
Most verbal roots consistently take either active or middle voice, but
cannot take both. (I am regarding extended verbal roots as different
roots, for the purpose of ascertaining the 'voice potential* of a root;
and since Imperatives do not combine with the passive voice, I am not
concerned here with whether a root can combine with the passive.) Thus,

since raar- 'look at' never takes middle voice, sentence 316. will be
interpreted as having a complex of active voice; and since Bad- 'approach*
never takes active voice, sentence 317. will be interpreted as consist­
ing of a complex of middle voice:

316. Raar-atay maaroori! 317. Bad - eel

look rice approach
VAP VAP
'Keep on looking (sg) 'Approach (pi)!'/
at the rice ! 1 'Come closer (.pl)!‘
 242

In complexes such as raar-atay and 6 ad-ee in 316. and 317., the voice
of the complex is absolutely obvious - to the Fulfulde speaker
from the choice of verbal root, and there is, therefore, no possibility
of ambiguity.

5.5.2. 'Accidental Homonymy 1 - (i)

I shall deal here with two instances of homonymy that involve 2sg
object pronouns, whose forms 'happen to' be homonymous with part of
a VAP-suffix.

If the reader refers to tables XXII and XXIII, he will note

the following:

VAP-suffix for VAP-suffix for

Negative Completive Active: aa General Incompletive Active: (e)t
2sg Object pronoun for 2sg Object pronoun for
Negative Completive Acti v e : ma General Incompletive Active:
VAP-suffix for VAP-suffix for
General Completive Passive: aama. General Incompletive Passive: Oe)t e ,

Now, the combination of the first two forms in each column (aa + ma, and
(e)jt + eO would lead to a form homonymous with the last form in each
column (aama and (e)te).

These potential instances of homonymy are recognised by Fulfulde-

speakers, and, on the whole, they are deliberately avoided. Usually,
the -aa suffix becomes -aayi in front of raa^cf. 318.b., but remains as
- aa in front of all other pronouns, cf. 318.b.. Thus:

318. a. Min-kokk-aayi-ma ko ngicf-0-cfaa.

we give you what want you
VAP VAP
'We didn't give you (sg) what you wanted.'

b. Min-kokk-aa-mo ko ngicf-0-cfaa.
we give him what wand you
VAP VAP
'We didn't give him what you (sg) wanted.'

The - aama suffix, on the other hand, is retained for the General
Completive Passive, as in 318.c . :
 243

3.18. c. Min-kokk-aama ko ngicf-0-cCaa.

we give what want you
VAP VAP

'We've been given what you (sg) wanted.'

However, Fulfulde-speakers do recognise — aa—ma aa a possible alternative

to - aayi-ma in 318 .a . , and it seems that a form such as min-kokk-aa-ma
in 318. d. could be used in the sense of 318 .a ., provided that there

were enough contextual clues (in the discourse situation, etc.) to

disambiguate it:

318 .d. Min-kokk-aa-ma ko ngicf-0-cfaa.

we give you what want you
VAP VAP

'We didn't give you what you wanted.'

Ambiguity in the General Incompletive is normally avoided by using

the General Incompletive Passive participle instead of a verbal complex.
Thus 313a. would normally be used, but not 319.b . :

319.a. M i - k o o w - etee-cfo £jj Yaamuusa.

I marry child Yamusa.
CL

'I'm to be married to the child of Yamusa.' (literally,

'I'm the/a person to be married to the child of Yamusa.')

?b. Mi-hoow-ete 6 ii Yaamuusa

I marry child Yamusa.

The -et-e form is normally reserved for the General Incompletive Active
with a 2sg Object:

319. c. Mi-hoow-et-e fbii Yaamuusa.

I marry you child Yamusa
'I'll marry you to the child of Yamusa.'

Nevertheless, Fulfulde-speakers do recognise -ete as a possible suffix

for the General Incompletive Passive (cf. 319.b.); and it is accepted

by some speakers in unambiguous contexts. For example, winnd- 'write*

takes a direct object referring to what has been written; and any
other object would have to be accompanied by an extension (such as
-an 'to/for'). Thus the -e_ part of -ete in 319.d. could only be
interpreted as being part of the General Incompletive Passive suffix,-
and not as the 2 sg object pronoun found in 319.c . :
 244

319. d. Dereeji cfii nibinnd-ete.

letters these write
VAP
'These letters will be written.'

5*5.3. 'Accidential 1 homonymy - (li).

I shall deal here with four instances of morphological homonymy that

are automatically disambiguated by context.

Firstly, with, a subject that semantically includes 3rd person, com­

plexes of the Relative Completive Middle and the General Completive Active

may be identical, for they have SV order, and the basic suffix is -ii
(cf. table XXII). Here are some examples of sentences containing
homonymous complexes:

320.a. Sey min-kadd-ii gude 'amin.

then we gird wrappers our
VAP
"Then we (women) put on our wrappers (= skirts).'

b. Min-kadd-ii minyiimin.
we gird our younger sibling
VAP

'We dressed our younger sister in her wrapper.'

Now, sey must be followed by a Relative (rather than Non-Relative)

complex; and the non-personal object (gude 'amin) in 320. a. indi­
cates that the action of ’wrapping around’ (hadd-) must be being per­

formed refiexively (ie. on the person indicated by the subject),

rather than on another person. For these two reasons, the complex in
320.a. will be interpreted as realising the properties relative and
middle. As for 320.b., since the object is personal (minyiimin) ,
the action denoted by the verb will be interpreted as being performed
on the object, rather than reflexively; and the voice of min-kadd-ii
in 320.b. will therefore be interpreted as active rather than middle.
(There is, however, no conjunction in 320.b. (like sey in 320 -a.) which
requires a Relative or Non-relative complex; and so the syntax of 3 2 0 *b.
gives no clues as to whether the complex is Relative or Non-relative.)
A Relative Completive Active complex would have -i^ rather than —ii
suffix; and therefore, the complex in 320,b. must be interpreted as
being of the General Completive Active.
 245

With a Subject that includes 1st and/or 2nd Person, complexes

of the Relative Completive Middle and of the General Completive
Active are morphologically distinct: for the former have VS order,
and participate in the Suffixal Precedence Hierarchy, whereas the latter
do not. These morpho-phonological and morpho-syntactic factors
help to reduce the number of homonymous complexes of these two
conjugations.

Secondly, it is possible to find complexes of the Negative

Completive Active and of the Relative Completive Passive that are
homonymous. For example:

321.a. Nyaako hoow - aa - mo.

father marry not her
VAP
'Father didn't marry her (off).'

b. Ndee Maaroori hoow - aa - mo?

when Maaroori marry him
VAP

'When was Maaroori married to h i m ? '

In such pairs of sentences, several factors help to disambiguate the

homonymous verbal complexes. Firstly, the semantics of the verbal root
will often dictate the interpretation of the voice of the complex.
In the case of hoow- 'marry', for example, a father marries off his
child, rather than is married off himself; and 321. a. is therefore likely
to be interpreted as realising active voice. However, a girl is married
off; and so 321.b. is likely to be interpreted asrealising passive
voice. (In any case, there are very few verbal roots that can have
a direct object with the passive voice.) Secondly, the syntax of a
sentence with a question-word, such as 321.b. requires a. Relative
conjugation; whereas there is no obvious syntactic need for a Relative
conjugation in 321.a.. This fact will help 321.at to be interpreted
as Negative Completive Active, and 321.b. as Relative Completive
Passive. Lastly, of course, the general discourse-context will provide
clues as to the meaning of such sentences. All in all, there is in
fact little danger of misinterpretation.

Thirdly, there is sometimes homonymous realisation of a 2sg

Subject in the Subjunctive and of the VAP-suffix of the Negative
 246

Completive Active and of the Relative Completive Passive,, which have

just been discussed. For example, the verbal complexes in the foll­
owing three sentences are in fact identical in surface form; but
the - aa in 322.a. is a 2sg pronoun, whereas in 322.b. and c . , it is
a VAP-suffix.

322. a. 'Igga suucf-0-aa - be.

better hide you them
VAP
*You had better hide them.*

b. Muusa suutf-aa-6 e .
Musa hide them
VAP
'Musa didn't hide them.'

c. Tokoye Muusa suucf-aa-£e?

where Musa hide them
VAP
'Where was Musa hidden from them? *

The disambiguating factors for sentences such as 322.b. and c. have

just been discussed: the complex in 322.b. is of the Negative Completive
Active (as is that in 321.a.), and the complex in 322.c. is Relative
Completive Passive (as is that in 321.b.). As for 322.a., the presence
of the sentence-initial modality-marker, 1igga, demands a following

Subjunctive complex; and since no NP or pronoun precedes the verbal

root in this sentence, the - aa suffix will be interpreted as a suffixed
2sg Subject, rather than as a VAP-suffix (as it is in 322.b. and 322..c.).
(.With a verbal root beginning with an alternating consonant, the form
of that consonant would help to identify the VS nature of a complex
such as in 322.a. from the SV nature of complexes such as in 322.b.
and c. . With the root suucf-, however, s_- does not alternate, and
the particular examples given in 322. do not provide this morpho-
phonological clue.)

The last instance of 'accidental homonymy' again involves a 2sg

Subject in certain VS complexes. This -aa Subject pronoun combines with
the VAP-suffix of the Relative Incompletive Active, resulting in the
sequence -at-aa:
 247

323.a. Ndee pudcC - at-aa?

when start you
VAP

'When will you start?'

However/ - ataa is in fact the conjugation-suffix for the Negative

Incompletive Active:

232 *b. Husseini'en pucfcf - ataa.

Husseini's start
people
VAP

'Husseini's people won't start.'

The complexes in 323* are homonymous, phonologically, although their

morphological structure is different (as reflected by the extra hyphen
in 323.a.). There is in fact no danger of ambiguity here. In 323.a.,
there is no subject before the root, so aa must necessarily be interpreted
as subject (as indeed we observed happened with. 322.a. too); and in
any case, ndee requires a Relative conjugation, which often results
in VS order. In 323.b . , Husseini'en is. obviously the subject, so there
is no danger of the aa of - ataa being interpreted as subject.

5.6. Conclusion

This chapter has been an attempt to look at the morphology of Fulfulde

verbal complexes from the theoretical rather than the descriptive point
of view. With the development of a more adequate theory of morphology,
perhaps it will eventually be possible to provide a more systematic
account of the types of exponence displayed in an inflecting language
like Fulfulde.
 248

Chapter 6

Topic and Focus

6.0. Introduction

The aim of this chapter is to establish the categories _o f topic and

focus, and to highlight some of their syntactic and morphological
exponents.

These two categories are in fact recognised in A m o t t (1970:

§7.10 - §7.17 and §7.7. respectively), though under a different guise.
Arnott calls topic 'prelude1, and focus 'emphasis'. However, since
the morphological and syntactic evidence o f Fulfulde bears comparison
with that of other languages where these categories have been est­
ablished, it seems advisable to use the more commonly-accepted term­
inology.

A topic is necessarily an NP; a focussed item may, however,

be an UP, a verb, or an adverb.'1' A focussed NP resembles a topic
NP inasmuch as it occurs at the beginning of a sentence. Although
the resemblance ends there, there is considerable room for confusion
between a topic NP and a focussed NP, and for this reason, focussed
NP's will be allocated considerably more attention here than other
focussed constituents.

After some initial exemplification, the discussion of topic

will consider topic and anaphoric reference (§6 .1 .1 .); variation
in topic-position within the sentence (§6 .1 .2 .); and topic markers
(§6.1.3.). The discussion of focus will be divided into a consid­
eration of NP focus (§6.1.2.) and verb focus (§6.2.2.). Finally,
some concluding remarks about focus will be made (§6 .2 .3 .).

6.1. Topic

Topic NP's were first discussed and defined in section §2.3.1. above.
Many types of sentence may be preceded by such an NP. However, in
 249

this outline, I shall concentrate on the occurrence of a topic with

the types of sentence that are discussed in various parts of this
study, and only brief mention will be made of topic in relative
clauses and other subordinate clauses.

Here, then, are some examples of topic NP's. 324 . and 325. are
verbal sentences, the former 'neutral*, the latter with a focussed
2
constituent; 326. is a non-verbal sentence.

324. Kanje, mi-'annd-aa ~ cfe.

them I know not them
TOPIC VAP
1 (As for) them (eg. books), I don't know them.'

325. 'Innde m a a k o , Gaya 1o-wi *-etee.

name his Gaya he call
T O P I C FOCUS VAP
' (As for) his name, he's called Gaya.'

326. mo nan-ataa Fulfulde fu, 'o kaadb.

who hear not Fulfulde all he non-Fulani
VAP
T O P I C
'Whoever does not understand Fulfulde is a non-Fulani.'

Note that the above examples involve various types of topic NP's.
Kanje in 324. is an independent pronoun. 'Innde maako in 325.
consists of a head ('innde) plus genitival modifier (maako). Mo nan-
ataa Fulfulde fu in 326. consists of a headless relative clause which
includes the quantifier, f u . (Fu 'quantifies' m o , not Fulfulde, and
conveys the '-ever 1 of 'whoever'.) Various other types of NP structure
will be exemplified in the topic NP's shown during the following
discussion; and in the section on topic and anaphora (§6 .1 .), the
importance of the nature of the topic NP will become apparent.

Note also that either part or all of the topic NP in the above
sentences is referred to anaphorically in what we may call the 'comment’
part of the sentence (,cf. Gruber 1969 and Gundel 1977) . The anaphoric
pronouns in the above examples are -de, 'o- and _|_o_ in 324.-326.
respectively. Because of the presence of such pronouns in the comment,

it has been argued that a topic NP is extraposed from within the com­
ment bv a orocess of 'topicalisation1. Such an analysis is made of
Hausa topics, for example, in Jaggar (1978).
 250

While sentences such as 324 . and 325 . may be satisfactorily

explained by a process of topicalisation, this is not true of 325 ..
Here, although maako and _|_o are coreferential, and although there
is obviously a close semantic relationship between topic and comment,
it is not possible to claim that 'innde (which is the head of the
topic NP) has been extraposed from within the comment. Such a
syntactic process is similarly improbable in a sentence such as the

following:

32 7. Sobbal kam, ko min-ngacf-ata, sey min'-6 o h — a dowroori.

'flour- what we do then we thresh millet
balls'
TM VAP VAP
' (As for) flour-balls,- what we do (is) we then thresh
millet.*

(Topic-markers such as kam will be discussed presently.) Here, the

comment does not even contain a genitival pronoun that is coreferential
with the topic NP, sobbal.

Given sentences such as 325 . and 327. t it seems to me that a

structure such as the following underlies Fulfulde sentences with a
topic:

topic

This allows the topic NP to be generated regardless of whether it

has a coreferent in the comment. Further research into sentences
with a topic will no doubt show whether this suggestion is valid.

Sentence 327. illustrates another characteristic of topic NP's.

This is the fact that they are frequently followed by a topic-marker
(glossed 'TM* in the examples). Unlike in Japanese, for example
(.see Gundel 1977: 17), the distribution of topic-markers is determined
in Fulfulde by discourse considerations rather than b y syntactic
and referential ones. For example, the topic-marker kam in 32 7.
marks sobbal as the topic, and could be used without regard to whether
sobbal has a coreferent in the comment or not. If boo were substi­
tuted for kam in the same context, a contrast with a preceding topic
 251

and comment would be implied. These matters will be discussed

further in §6.1.3.. For the moment, however, it should be noted
that when using the term 'topic', I include the topic-marker;
but 'topic NP' excludes the topic-marker.

The structure of this section on topic is as follows. In

§6.1.1., I shall discuss anaphora within topic-comment sentences.
Then, in §6.1.2., I shall turn to the position of a topic in a
sentence. Finally, in §6.1.3., I shall discuss the distribution of

topic-markers such as boo and k a m .

6.1.1. Topic and anaphoric reference

In this section, I shall be concerned with the way in which the topic
is anaphorically referred to by pronouns, or b y zero, in the comment.
The discussion can usefully be divided according to the case-function
of the anaphoric pronouns or zero.

6 .1.1.1. Subject and Topic coreferential

Let us compare the following three sentences:

328.a. K a m b e , no be - mbi'-etee?
them h o w they call
TOPIC S VAP
' (As for) the what are they called?’

329 .a. S e e , no 0 mbi'-etee?

these how call
TOPIC S VAP
’ (As for), these (people), what are they called?'

3 3 0 -a. Haabe cfoo, no 0 mbi'-etee?

non-Fulanis here how call
TOPIC S VAP
' (As for) the non-Fulanis h e r e , what are they called?'

These sentences are similar, yet crucially different. All have a

topic-comment structure; all contain the question-word no followed
by averbal complex of the Relative Incompletive Passive conjugation?
and the verbal root of each has the prenasalised form,m b - , which is
characteristic of a preceding plural subject. However, 3 2 8 -a. has
an independent pronoun, k a m b e , as topic, and be- as subject; 3 2 9 *a.
 252

has a deictic pronoun, Bee, as topic and zero subject; and 328.a. has
a nominal topic, Haabe (qualified by cfoo) , and zero subject. How
can we account for these differences in subject-form, given the fact
that K a m b e , See and Haabe all belong to, or refer to, the personal
plural n o m i n a l class?

Essentially, when the topic and subject of a sentence are

coreferential, then the form of the subject is determined by the nature

of the topic NP. The following rule in fact operates:

If the topic and subject of a sentence are coreferential,

then zero subject normally occurs, although a dependent
pronoun (such as be- in 328. a.) may optionally occur.
However, if the topic is an independent pronoun (such
.as kambe in 328-a - ) r then the dependent pronoun subject
is obligatory.

This rule precludes 328. b. as a variant of 328 .a., but allows 329 .b.
and 330.b. as variants of 329. a. and 330. a.:

32S.b. *K a m b e , no 0 m b i ’-etee?
them now call

329 .b. B e e , no be - mbi'-etee?

these how they call
' (As for) these (people), what are they called?'

3 3 0 -b. Haabe cfoo, no be_ - mbi'-etee?

non-Fulanis here how they call

' (As for) the non-Fulanis here, what are they called?'

In other words, a dependent subject pronoun such as be- is obligatory

with an independent pronoun topic such as kambe; but it is optional
with a topic of any other form, such as the deictic pronoun bee in
329., and the nominal Haabe in 330..

6.1.1.2. Object and Topic coreferential

When the topic is coreferential with the object, or one of the objects,
3
zero object selaom occurs. This is true regardless of whether the
object is a direct object (cf. 331 .) or an indirect object (cf. 332.) ,
and regardless of whether the topic is an independent pronoun (cf. 332.
331 ,b. and 332,b.) or not (cf. 331 .a. 332 -a.). The 'non-zero' objects
in the following examples are underlined:
 253

331. a. Dereeji cfii, sey min-njanng - a - cfi.

letters these must we read them
TOPIC VAP
1 (As for) these letters, we must read them.'

b. Kanji, sey min-n j anng - a - cfi.

them must we read them
TOPIC VAP
' (As for) them, we must read t h e m . 1

332.a. Banndam, 1a - hoow - aa - mo 6 iyaa.

my kinsman you marry not him your child
TOPIC VAP
1 (As for) my kinsman, you didn't marry your child to him.'

b. Kanko, 'a - hoow - aa - mo £>iyaa.

him you marry not him your child
TOPIC VAP

' (As for) him, you didn't marry your child to h i m . 1

Zero never occurs as the object of a preposition:

333. Lafbi 'am kesi, mi-ta'y-ir - ii kusel *e maaki.

knife my new I cut meat with it
T O P I C EXT VAP
' (As for) my new knife, I cut meat with it.'

6 .1.1.3. Genitive and Topic Coreferential

When the topic is coreferential with a genitive, then the situation is

slightly more complicated than that just described. Generally, the
genitive pronoun of the relevant nominal class is used (cf. 334.a.
and b . ) , or of the relevant 1st or 2nd person (cf. 335.a. and b . ) :

334.a. Kanko, na'i maako fu mbaat-id-ii.

him cows his all die
TOPIC EXT VAP
' (As for) him, all his cows have died.'

b. Puccu 'am, mi-he 6 -aayi maagani m a a g u .

horse ray I get medicine its
TOPIC VAP

' (As for) my horse, I haven't got its medicine.'

335.a. Minn, 6 i& 6 e 1am nan - ataa Fulfulde.

me children my hear not Fulfulde
TOPIC VAP

' (As for) me, my children don't understand Fulfulde.'

 254

3 3 5 .b. 'Qnon, tokoye wuro moocfon?

you where home your
TOPIC
' (As for) you (pi) , where is your h o m e .'

The sentences of 334. contain a 3rd person topic (kanko and puccu ;^am) ,
and the appropriate genitive pronoun in the comment (maako and maagu);
while those of 335. contain a 1st or 2nd person topic (minn and 1onon) /
and the appropriate genitive pronoun (1am and moocfon) .

However, if the topic is a nominal of the singular personal class

C' O) , then the genitive pronoun muucfum is used instead of maako.
Thus if we replace kanko of 334.a. by the nominal nyaako ‘father',
maako will have to be replaced by muucfum:

336.a. Nyaako, na'i muucfumfu m b a a t - i d - i i .

father cows his all die
EXT VAP

' (As for) father, all his cows have died.'

b. *Nyaako, na'i maako fu mbaat - id - ii.

father cows his all die
EXT VAP

336.b. is unacceptable because maako cannot refer back to a nominal. A

variant of muucfum is m u m , thus:

336.c. Nyaako, na'i mum fu m b a a t - i d - i i .

father cows his all die
EXT VAP
' (As for) father, all his cows have died.'

Muucfum and mum are interchangeable, and 336 .a. and c. have exactly
the same meaning.

There are two types of topic that seem to waver between maako
and muucfum (or m u m ) in their anaphoric requirements. A topic of
the singular personal class, and which includes a deictic adjective
or pronoun, may be referred to either by maako or muucfum (cf.
337.a, and b,),- and this is also true of a topic of the singular
personal class if it includes a participle (cf-338.a. and b.). For
example, both of the following were accepted by informants:
 255

337.a. 'Otton debbo, m i - 1annd-aa 'innde m a a k o .

that woman I know not name her
VAP
1 (As for) that woman, I don't know her name.'

b. 'Otton debbo, m i - ‘annd-aa 'innde m u m ,

that woman I know not name her
VAP
' (As for) that woman, I don't know her name.*

And both of the following were also accepted, though there was a

certain amount of doubt about the second, which I have therefore

marked (?):

338. a. Gacf-cfo kuugal, ngal, yaadu maako farit-ii.

do SG work the journey his postpone
PERS VAP
CL
PARTICIPLE

' (As for) the person who did the work, his journey
has been postponed.'

(?)b. Gacf-cfo kuugal ngal, yaadu muucfum farit-ii.

do SG work the journey his postpone
PERS VAP
CL
PARTICIPLE
1 (As for) the person who did the work, his journey
has been postponed . 1

Further, deeper, research will perhaps clarify some of the issues

involved here.

6 .1.1.4. Anaphoric Reference: Summary

Before discussing the more general implications of the observations

made in this section, let me summarise them in tabular form. Table

XXIX shows how a third person case-function is realised if
it is coreferential with the topic of the same sentence. In con­
sidering the table, the following points should be borne in mind:

(i) First and second persons have not been shown,

since their form does not depend on the nature
of a preceding coreferential topic.

(ii) The Nde class has been arbitrarily chosen to

represent non-personal classes.
 256

(iii) Participles based on war- 'come 1 have been arbit­

rarily chosen to represent all participles.
Pullo/Fulbe *Fulani/Fulanis1, and asaweere 'week'
have been arbitrarily chosen to represent all
other NP's.

Table XXIX

Form of a 3rd Person Case-function if it is coref­

erential with the Topic of the same sentence

Nature of Topic NP Nature of Coreferent of Topic in Comment Clas

Subject Object Genitive
Inde­ kanko 'o mo maako '0
pendent kambe be be mabbe Be
Pronoun kayre nde nde mayre Nde

NP incl­ 'oo 0/ o mo maako/muudhm '0

uding bee 0/be be mabbe Be
deictic ndee 0/nde nde mayre Nde

NP incl­ gar ay cfo 0/ o mo maako/muucftim '0

uding p a ­ waraybe 0/be be mabbe Be
rticiple waraynde 0/nde nde mayre Nde

Other Pullo 0/o mo muudhm '0

NP Fulbe 0/be be mabbe Be
Asaweere 0/nde nde mayre Nde

What are the implications of the observations made in this

section? First of all, I think that they should be studied in the

context of the system of reference in the language as a whole: for

the way in which a subject, object, or genitive is realised if it
is coreferential with the topic generally correlates with the way
each of these cases is realised if it is coreferential with any item
in the same horizontal line in table XXIX. So, for example, while
maako may have a kanko topic as its antecedent (cf. sentence 334.a.,
page 253), it may also have a kanko subject as its antecedent, or
a mo object. These latter facts are exemplified in 339.a. and b . :

339.a. Kanko n y o 1-i ngafalewol ma a k o .

he sew gown his
VAP
'It's he who sewed his (own) g o w n . 1
 257

b. Mi-yigg-an — ii mo baawo maako.

I rub for him back his
VAP
'I rubbed his back for him.'

Similarly, while mum/muudhm may have a noun topic asitsantecedent,

it may also have a noun subject (cf.340.a.) or object (cf. 340.b.)
as its antecedent:

340.a. Pullo semt - ay 'af o mum.

Pullo be shy first-born his
VAP
'A Fulani is shy of/shows respect to his first-born
child . 1
b. 'On-ngi-ii Dawoobe 'e danyoobe mum?
you see Dawobe and parents his
VAP
'Did you see Dawobe and his parents?'

The only caveat that needs to be made concerning the 'mutual

referability' of the items in the horizontal lines of table XXIX
is that if or mo has been used, then it insist be referred to by
1o, mo or m a a k o , not 0_ or muudhm/mum. Thus 341.a. is correct, but
341.b. is not:

341.a. Raa mo 'e-mo-wuund-ii binngel m a a k o .

see him he hug child his
VAP
1Here he i s , hugging his child . 1
b. *Raa mo 'e-0-wuund-ii binngel mum.^
see him hug child his

In summary, then, table XXIX is not only useful for determining

the form of an anaphoric subject, object or genitive that refers
back to a topic; it is also useful for determining the form of such
a subject, object or genitive referring back to any case-function
though with the caveat mentioned, in the preceding paragraph.

Since the main concern of this section is topic NP's, rather

than reference in general, I shall leave this matter here. However,

I believe that it could lead to fruitful research in the future.
 258

6.1.2. The Position of a Topic NP in a sentence

The examples given so far in this section have the topic NP in

sentence-initial position. This is to be expected, for the tppic
represents 'given' information, about which comment is made, and
it therefore more naturally occurs before the comment. In Fulfulde,
it is possible, however, for a topic to be in non-initial position.
In this section, I shall deal with the principal exceptions to the
sentence'-initial rule.

6.1.2.1. After-thought Topic

First of all, one finds 'after-thought topics' in Fulfulde. (Con­

cerning this construction, see Givon 1976, for example.) It is
arguable that these should be treated separately from other, 'ordin­
ary* , topics. However, for the purposes of this study, I shall treat
them as variations of the same construction. The after-thought topic
in 342'. is underlined:

342 . Dume 'o-wacf-ata, lcanko boo?

what he do him TM
'What's he do/What's he doing, him, on the other hand?'
'He, on the other hand, what's he do/what's he doing?*

An after-thought topic may in fact co-occur :v£Lth an *ordinary' topic,

for instance:

343 . *Qo, no 0 wi'-etee, mo jooni?

this how call him now
'This person, what's he called, him of today?'
'What's he called, the present one (eg. leader)?'

Here, 1oo is an 'ordinary' topic, and mo jooni an after-thought

topic. Alternatively, 'oo and mo jooni could be analysed as a
’discontinuous topic'.

6.1.2.2. Inversion of Topic and Question-Word, Relative Pronoun or Adverb

Some speakers of the Kaceccereere dialect accept the inversion of

a topic NP and a question-word (cf. 344 .) or a relative pronoun or
adverb (cf. 345.) . Thus, the a. sentences below have the topic in
initial position, but inversion has occurred in the b. sentences:
 259

344. a. 1A a n , moye ngicf-0 - cfaa wolw - an -go?

you whom want you speak to
TOPIC VAP EXT INF
' (As for) you (sg), whom do you want to speak to?'

b. Moye 1aan/ ngicf-0 - cfaa wolw - an - go?

whom you want you speak to
TOPIC VAP EXT INF

* (As for) you (sg), whom do you want to speak to?'

345. a. 13ee kammaa, nde 0 senndir-i, 0 ngart-i Kaceccere'en

these when. separate become Kaceccere'en
TM VAP VAP
T O P I C S S
' (As for) these (people), when they separated (off),
they became Kaceccere'en.'

b. Nde 5ee kammaa, 0 senndir-i, 0 ngart-i Kaceccere'en

when these separate become Kaceccere'en
TM VAP VAP
T O P I C S S

'When these (people) separated (off), they became

Kaceccere'en.’

It must be emphasised that 344.b. and 345.b. were not accepted by all
ray informants, and I therefore suspect that this type of inversion
is only marginally acceptable. It is also worth noticing that such
inversion occurs most easily, perhaps only, when the topic is core­
ferential with the subject.

6 .1.2.3. Complement Clauses

Another circumstance when a topic does not always occur in initial

position is when its comment is a complement clause, ie. the object
of a verb of thinking, knowing, etc.. Consider sentence 346.a.:

346.a. Mi-tamm-ii, foikkoyg wojjere, fowru looh-ay-koyi}.

I think children hare hyena hunt them
VAP T O P I C VAP
'I think the hyena hunts leverets.'

Here, the topic (Bikkoyg wojjere) intervenes between the main clause
and its complement. However, a variant of 346.a. is 346.b., where
the topic is in the more normal, sentence-initial position:
 260

346.b. Bikkoyn wojjere, mi-tamm-ii fowru looh-ay-koyi}.

children hare I think h^rena hunt them
T O P I C VAP VAP
' (As for) leverets, I think the hyena hunts them.'

6 .1.2.4. Two Topics

A final type of sentence where a topic NP is not in initial position

is where there are two topic NP's - they cannot both be simultane­
ously in initial position. Consider sentence 347., for example, where
the first topic is an independent pronoun ('aan), and the second
an NP that includes a relative clause (ko ngicf-cfaa fu) :

347. 'Aan, ko ngicP-0-cfaa fu , sey be - kokk-u-maa-cfum.

you what want you all must they give you it
VAP VAP
TOPIC T O P I C
' (As for) you (sg), whatever you want, then they
should give it y o u .'

Here, the first topic is coreferential with the indirect object (-maa-)
and the second one is coreferential with the direct object (-cfum).
The second topic contains a quantifier, fu 'all'. (Fu and p a t , both
meaning 'all' /'every' , and t a n , meaning 'only* , often occur in this
topic-final position. Their co-occurrence possibilities with the
topic-markers that are discussed in the next section, have yet to
be investigated.)

6,1.3. Topic-Markers

Certain 'particles' are sometimes used to mark off a topic from the
rest of a sentence, ie. from the comment. Such is for instance,
kam (found in 327. above). Other such topic-markers are boo,
maa, (which is almost certainly borrowed from Hausa), and kammaa
(which is sometimes used in other ways as well as a topic-marker) .

Topic-markers 'oil the wheels' of discourse, but they are not

syntactically obligatory. In fact, many of the examples given so far
are not accompanied by a topic-marker. However, since these items
frequently occur, I shll offer an explanation of how a sample of
them are used.
 261

6.1.3.1. Kam

Kam is the most 'neutral' of the topic-markers, which is perhaps

why it is sometimes combined with one of the less 'neutral' ones.
It may be used to mark the first topic in a series of sentences on
5
a related theme in the discourse. For example:

348. Altine kam, ‘e-mo—yicf-i maaroori.

Altine she like rice
TOPIC TM VAP
^ ' (As for) Altine, she likes rice.'

Let us assume that this is the opening sentence in a conversation.

Gradually, we shall follow through various possible responses to such
an opening. Kam may simply be used to mark the first topic in a
discourse (eg. Altine in 348.) , or it may be used to relate the
next topic (eg. Muusa in 349. a.) to the previous sentence. In the
latter function, it is similar to English 'and'. Thus:

349 .a. Muusa kam, 'e-mo-yicf-i masaari.

Musa he like maize
TOPIC TM VAP
'And (as for) Musa, he likes maize.'

Although there is a contrast both in the topic and comment of 348 . and
349 .a., the contrast is not brought out, not emphasised.

Kam is rarely collocated with the 3rd person independent pronouns,

which all consist, morphologically, of kaN- (where N_ represents a
nasal that is homorganic with the subsequent consonant) plus a nominal
class-suffix. A collocation such, as kanko kam is, perhaps, not euph­
onious to the Fulani ear. Alternatively, I wonder whether kam and
kaN- are cognates, and whether they are not easily collocated for that
reason?

6 .1.3.2. Boo

Boo implies a contrast of both topic and comment. Thus, if a speaker

wanted to emphasise the difference between the culinary tastes of
Altine and Musa, he would use boo Instead of kam, thus:
 262

349 .b.Muusa boo, masaraari fo-yid—i

Musa maize he like
TOPIC TM VAP

'Musa, on the other hand, it's maize he likes.'

The discourse-function of boo is thus roughly equivalent to 'on

the other h a n d 1, 'but1, or 'however' in English. Here are some
further examples of the use of b o o , the first of which involves
kam too:

350. Sisil&e k a m , to Zaariya 0 ngon-i; Kaceccere'en b o o ,

Sule&awa in Zaria live Kaceccere'en
TOPIC TM S VAP T O P I C i>m

, cfoo Kaceccere 0 ngon-i.

here Kaceccere live
S VAP
'The Sule&awa (Fulani clan) live in Zaria;
however, the Kaceccere'en live here in Kaceccere.'

351. Muusa, duuBi mum joweetati; Habu b o o , duu&i mum sappo.

Musa years his eight Habu years his ten
TOPIC TOPIC TM

'Musa's eight (years old), whereas Hahu's t e n . 1

6 .1.3.3. Maa

Maa means 'also', 'too' or 'even'. It is often used to mark a new

topic, whose comment is similar to the preceding one. Thus, if we

return to our examples expressing culinary taste, a speaker might
say after sentence 349.a.:

352- Minn m a a , 'e-mi-yicf- i masaraari.

me I like maize
TOPIC TM VAP

'Me too, I like maize.'/'Even I like mai z e . ’

The comment in 352- is similar to that in 349.a. (both concern the

liking of m a i z e ) , but the topic is different (minn rather than Muusa).

6 .1.3.4. Kammaa

The meaning of kammaa is. roughly equivalent to English 'moreover'

or 'whereas'. It is often placed after a topic (cf. 353.), though
it also occurs elsewhere in a sentence (cf. 354.):
 263

353 . Bee, sey 0 l a a t - i moodiB&e; Betton kammaa, sey

these then become teachers those then
TOPIC S VAP TOPIC TM

0 ngadd-an - aa garmaaje.
bring for hoes
S EXT VAP
'These (.people)., they then became teachers; whereas
those, they were brought hoes (ie. became farmers).'

354 . Asali na'i woncfon, kammaa, ko Bur - i.

lineage cows is moreover which exceed
VAP
'It's the lineage of cattle (Fulani), moreover, that
is b e s t . '

Since kammaa follows the topic, Bet t o n , in 353.-, I have glossed it 'TM'
ie. 'topic-marker'. However, since kammaa in 354. does not occur
after a topic, but rather after a focussed NP (asali na'i - see
section 6 .2 .1 . below) and its accompanying clefting particle (wondbn -
see 6 .2.1.7. below), it cannot be glossed *TM' there. It may be,
then, that kammaa should simply be regarded as an adverb that can,
sometimes, be placed after a topic.

It seems possible, though it has yet to be proven, that kammaa is

derived from a combination of kam and m a a , which have just been
described. The fact that one informant mentioned the existence of
another 1particle'/topic-marker, kamboo lends weight to this idea.
However, the latter is not used b y most of my informants, and the
investigation of these possibly compound forms will have to be cont­
inued at a later stage.

6 .1.3.5. Topic-Markers : Conclusions

Since topic-markers constitute a link between one sentence and another,

they should really be considered in their discourse context, rather
than within isolated sentences. The remarks made here are intended,
then, simply as a pointer to further research in this field.

6.2. Focus

I use 'focus' to refer to a constituent that introduces new or high­

lighted information into the discourse, and that potentially receives
contrastive stress (cf. §1.1.4.3.). I shall concentrate here on some of
 264

the morphological and syntactic exponents of focus, which were touched

on in §2.3.2. and §3.7.1.1. above. I deal first with NP focus (in §6.2.1.),
and then with verb focus (in §6.2.2.). These two typesof focus are mutually
exclusive in the same clause, as will be illustrated in §6 .2 .3 ..^

6.2.1. NP Focus

Foucssed N P ’s, like topic NP's, were first mentioned in section §2.3.1..
Like topic NP's, they usually occur in sentence-initial position. However,
if a topic and focussed NP co-occur, the former precedes the latter. Con­
sider the following three sentences, for example (here, as throughout this
section, I shall underline a focussed constituent and its English equivalent)

355.a. 'A - y o 6 -ii-mo ceede (fee.

you pay her money this
VAP
'You (sg) paid her this money.'

b. Ceede cfee kam, ’a - yo&-ii-mo-cfe.

money this you pay her it
TM VAP
T O P I C
1 (As for) this money, you paid her it. '

c. Ceede cfee njofo-u-daa-mo.

money this pay you her
F O C U S VAP
' (It's) this money you paid h e r . '

d. Ceede cfee kam, kanko njo 6 -u-cfaa-cfe.

money this her pay you it
TM VAP
T O P I C FOCUS
' (As for) this money, (it's) her you paid it to.'

Sentence 355.a. is a 'neutral sentence', consisting of a verbal complex of

the General Completive Active conjugation. Sentence 355.b. begins with a
topic ceede cfee k a m , which is potentially followed by a pause (indicated
by the comma), and which is referred to anaphorically by the object
pronoun, -cfe - compare the discussion in §6.1.1. above. Sentence
355, c, starts with the NP ceede cfee: however, here theee is no pot­
ential pause, and no anaphoric pronoun referring to ceede cfee;
 265

and moreover, the verbal complex is of the Relative Completive

Active conjugation. Sentence 355.d. also has ceede cfee kam in

initial position. This time, ceede dee kam is followed by potential

pause, and is referred to anaphorically by cfe in the comment; the
comment begins with an independent pronoun, k a n k o , and mo is no
longer present in the sentence; and the verbal complex is of the
Relative Completive Active. What we have in sentence 355.d. is in
fact a topic followed b y a comment which begins with a focussed
NP.

There are several features of focussed NP's that are illust­

rated in 355.e. - d. . These are:

(i) A focussed NP usually requires a relative conju­

gation in the same clause.

(ii) A focussed NP is not referred to anaphorically in

the same clause.

(iii) If a focussed NP is an anaphoric pronoun, then it

has the form of an independent p r o n o u n , rather than
of ,a dependent pronoun.

(iv) If a topic NP and a focussed NP co-occur, the former

precedes the latter.

In view of (ii), I assume that a focussed NP, unlike a topic NP,

(_cf. §6 .1 . above) is part of the matrix sentence, and that it has been
fronted by a process of 'focalisation'. This is why I speak of a
focussed Crather than focus) NP.

In sections §6 .2.1,1. to §6 .2.1.4., I shall deal with each of

the above features in turn.

6 .2.1.1. Focus: Relative Conjugations

A positive sentence containing a focussed NP usually has a relative

rather than a non-relative complex. Thus, a General Completive
complex, for example, is unacceptable:

355-e. *Ceede cfee 'a - yo 6 -ii-mo.

money this you pay him
VAP
 266

^55 .f . *Ceede cfee, kanko 1 a - yo&-ii-cfe.

money this her you pay it
VAP

As the the tree on page 166. demonstrates, there are only

two relative conjugations, one each for Completive and Incompletive

aspect. This means that the conjugations in the left-hand column
below are 'neutralised* in favour of those in the right-hand column:

Non-Relative Relative

General Completive 1 „ . ^.
„ , T Relative Completive
Stative J

General Incompletive1 „ , .. _ ...

„ . V Relative Incompletive
Progressive J

The occurrence of a Relative Completive instead of a General Compl­

etive complex in clauses with a focus has already been exemplified in
sentences 355. c. and d . . Here are some further examples of the non­
occurrence of the left-hand conjugations in clauses with a focus­
sed NP:

356.a. 1E - m i - 'annd-i hiigomum.

I know friend his
VAP
STATIVE
*I know his friend.*

b. Hiigo mum ' a n n d - u - m i .

friend his know I
VAP
REL COMPL
1 (It's), his friend I know.'

c. *Hiigo mum 'e-mi-*annd-i.

friend his I know

3 5 7 ,a. 1A - tuf - ay na'i?

you vaccinate cattle
VAP
GEN INCOMPL
'Do you (sg) vaccinate cattle?'

k• Na' i tuf - at - aa?

cattle vaccinate you
VAP
REL INCOMPL
'Is it cattle you vaccinate?'
 267

357.c. *Na*i 1 a - tuf - ay.

cattle you vaccinate

358-a. 1On-cfon - mbaal - an - a - 6 e?

you spend for them
night
EXT VAP
P R O G R E S S I V E

'Are you (pi) spending the night at their place?'

b. Kamfie mbaal - an - t - on?

them spend for you
night
EXT VAP
REL INCOMPL

'Is it at their place that you're spending the night?'

c. *Kam&e 'on-cfon-mbaal—an-a?
them you spend for
night

In 356.-358., each a. sentence is 'neutral' and contains a non-

relative complex; each b. sentence contains a focussed NP, and a
complex of the requisite relative conjugation (completive in 356.b.,
inconpletive in 357.b. and 358.b.); each c. sentence is unacceptable,
because a non-relative conjugation has been used (Stative Active
in 356.c,, General Incompletive Active in 357.c . , and Progressive
Active in 358.c.).

(Only a sentence with a Subjunctive verbal complex does not

undergo a change of conjugation if one of its NP's is focussed.
For example:

3 5 9 .a. ’O-weet - an - a bandiiko.

he visit for his kinsman
early
EXT VAP
'He should visit his kinsman in the early morning.'

b. Bandiiko 1o-weet - an - a.
his kinsman he visit for
early
FOCUS EXT VAP
‘Its his kinsman he should visit early in the morning.'
 268

Here, despite the focussing of the direct object, bandiiko, the

Subjunctive has been retained in 359.b.. If the Relative Incompl­
etive were used instead, the jussive force of the sentence would
be lost, and a sentence With. a different meaning would result:

359. c. Bandiiko 'o-weet - an - ta.

his kinsman he visit for
early
FOCUS EXT VAP
'It's his kinsman he'll visit early in the morning.'

This is equivalent to the neutral sentence:

359. d. 'O-weet - an - ay bandiiko.

he visit for his kinsman
early
EXT VAP

'He’ll visit his kinsman early in the morning.')

There are only two negative conjugations in Fulfulde, and

these are used in sentences with or without a focussed NP. For
example:

360. a. Nde-yool - aaki.

it drown not
VAP
'It (eg. hare) hasn't drowned.'

b. Kayre yool - aaki.

it drown not
FOCUS VAP
'This is the one that hasn't drowned,'/
'It's this one that hasn't drowned.'

361. a. Min-Bust - ataa ceede maaje.

we reduce not cost their
VAP

'We won't reduce their price.*

b. Ceede maaje min- 6 ust - ataa.

cost their we reduce not
F O C U S VAP

'It's their cost that we won't reduce.'

Here, focussing is effected by fronting, and in 360.b. by the use of an

independent pronoun, k a y r e ; but the same conjugation is used in the a.
sentences, which are neutral, as in the b. sentences, which are not.
 269

6 .2.1.2. Focus: no Anaphoric Pronouns.

In the discussion of sentences with a topic, it was noted that the

topic is frequently referred to anaphorically in the comment Csee

section §6.1.1.). However, a focussed NP is not normally referred

to anaphorically in the same sentence. Consider, for example:

362. a . Koynton, tni-loohj-ii-koyg.

those I hunt them
TOPIC VAP
"Those (.little ones), I hunted them.'

b. Koygton looh - u - m i .
those hunt I
FOCUS VAP
'It's those (.little ones) that I hunted.'

In 362.a., koynton is a topic, which is referred to anaphorically by

koyg in the comment. However, in 362.b . , koyuton is the focussed
object, and it is not referred to anaphorically in the rest of the
sentence. In fact, if koyg were added to 362.b . , then the focussing
effect would be lost, and an unacceptable sentence would result:

36 2. c . *Koyr)ton looh-u-mi-koyg.
those hunt I them

6 .2.1.3. Focus: Pronouns

A dependent anaphoric pronoun never occurs as the focus, of a sentence;

rather, an independent pronoun must be used in its place:

363 .a. 1Aa-cfon-jill-an - a - mmi-cfi?

you mix for me them
VAP
'Are you mixing them for me?'

b. Kanji nj ill—an - t — aa - mi?

these mix for you me
FOCUS VAP
'Is it these you're mixing for me?'

c. *Di njill-an - t-aa - mi?

them/ mix for you me

363.a. is a neutral sentence. In 363.b . , the direct object pronoun

has been focussed and taken the form kanji; and the verbal complex
 270

is of the Relative Incompletive conjugation. In 363.c./ all the

requirements for focussing have been met, except that dl_ has been
used instead of k a n j i ; 363-c. is therefore unacceptable.

Deictic pronouns and adverbs are often focussed. For example:

364-a. M i — soor-ii ndetton. b. Ndetton soor-u-mi,

I sell that
that one
on that one sell I
FOCUS VAP
*I sold that o n e . 1
'It’s that one I sold.'

However, unlike anaphoric pronouns, deictic pronouns and adverbs only

have one form, regardless of whether they are focussed (as in 364.b.)
or not (as in 364.a.).

6 .2.1.4. Focus with Topic

Sentence 355.d. above has provided us with an example of a topic

preceding a focussed NP. Here is a further example:

365.a. 'Onon, bilkiire mbolw - ot - on.

you pidgin speak you
Fulfulde
TOPIC F O C U S VAP
'You, (pi), speak pidgin Fulfulde. 1

If the order of the two constituents is reversed, then an unacceptable

sentence results:

365.b. *Bilkiire 'onon mbolw-ot-on.

pidgin you speak you

The only occasion when a topic does not precede a focus is faith, an
after-thought topic, (cf.§6 .1.2.1. above) such as 1onon in 363.c . :

365.c. Bilkiire mbolw - ot - on, 'onon.

pidgin speak you you
F O C U S VAP
'You speak pidgin, y o u . 1

6.2.1.5. Question-Words

Questions involving a question-word are syntactically very similar to

sentences involving a focussed NP. Compare, for example, the following:
 271

366-a. Moye meel - a n — t o o - ' e n ?

whom put on for we
turban
EXT VAP
'In whose honour are we (incl) putting on our turbans?'

b. Laamiidb meel - a n — too - 'en.

Emir put on for we
turban
FOCUS EXT VAP
1We *re putting on our turbans in honour of the E m i r . 1

367-a. Dume dany-aa? b. Ko natt-i dany-aa.

what bear what enter bear
VAP VAP VAP
F O C U S
1What has she had? 1
'S h e 's given birth to a b o y . 1

The verbal complexes of the a. sentences in 366. and 367. are each of
a relative conjugation, just as are those of the b. sentences; but
whereas the b. sentences contain a focussed NP, the a. sentences
contain a question-word. In a sense, question-words and focussed NP's
fill the same syntactic slot in such sentences,and so constitute
part of the same syntactic paradigm; and I shall therefore henceforth
consider question-words in Fulfulde to be a particular type of focus.

This is in line with Welmers (1973; 416—7) , for example.

6 .2.1.6. Focussed Adverbs

Focussed adverbs and adverbial phrases behave syntactically like

focussed NP's. Consider the following, for example:

368. a. Tokoye cfi - ndur-etee?

where they graze
VAP
'Where are they grazed?'

369.a. Non ndef - ir - ay - mi-ngal.

thus cook I it
EXT VAP
'That's how I cook it.'

In view of the remarks that have just been made concerning question-words,
I have underlined tokoye, and consider it as a focussed adverbial
question-word. In both 368. and 369.a, the focussed adverb has been
fronted, and is accompanied by a relative verbal complex - just as
 272

happens with, a focussed NP. Moreover, a variation of the above sentences

involves a pseudo-cleft construction! with, a copular verb linking the

focussed adverb and a relative clause:

368.b. Tokoye woni to cfi - ndur-etee?

where is where they graze
VAP

’Where is where they are grazed ? 1

'Where is it that they are grazed?'

369.b. Non woni no ndef - ir - ay - mi-ngal.

thus is h ow cook X it
EXT VAP
'Thus is how I cook i t . '/'T h a t 's how I cook i t . 1

Whether sentences such as 368*b. and 369*b. are identifying sentences,

with an underlying 'NP be NP' deep—structure is a moot point. (For
discussion, see Pinkham and. Hankamer 1975, and Gundel 1977, for example.)
Given the syntactic similarities between sentences with a focussed
adverb, and those with, a focussed NP, it seems likely that they are in
fact identical in deep-structure. It is to be hoped that future res­

earch. will clarify this issue.

6.2.1.7. NP Focus: Conclusions

This section, has been concerned with NP focus and two of its most
important exponents: fronting, and the use of a relative conjugation-
By way of conclusion, I should like to show the similarity between
focus constructions and relative clauses. Consider the underlined
verbal complexes in the following, each, of which is of a relative

conj ugation:

370. a. Mo 1annd-in - 0 - cfaa winnd-an - aa - 'en.

whom know you write to not us
EXT VAP EXT VAP
'The person you informed has not written to us.’

b. Minn 1annd-in - 0 - cfaa.

me know you
EXT VAP
'You informed me.'
 273

370 c. Minn woni mo 1annd-in - 0 - cfaa.

me is whom know you
EXT VAP
'The person you informed is m e , '

371.a. 'Agoogoyel ngel sood - u - mi sacC-'aa,

watch which, buy I dear not
VAP VAP
'The watch I bought is not expensive.'

b. 'Agoogoyel sood - u - m i .
watch buy I
VAP
'I bought a watch. ‘

c. 'Agoogoyel woni ko sood - u - m i .

watch is what buy I
VAP
'A watch is what I bought.'

The underlined complexes in the a. and c. sentences are each in a

relative clause whose antecedent (mo, n g e l , or k o ) immediately precedes
it. If the relative clauses contain a non-relative complex, unacceptable
sentences would result, for example:

370. d. *Mo 'a - 1annd-in-ii winnd-an - aa - 'en.

whom you know write to not us
GEN COMPL ACT

e. *Minn woni mo 'a - *annd-in - i i .

me is whom you know
GEN COMPL ACT

The underlined complexes in 3 7 0 -b. and 3 7 1 -b. are the 'out of focus 1
parts of sentences beginning with a focussed NP (minn and 'agoogoyel
respectively). Like their counterparts in the a. and c. sentences,
they cannot be replaced by a complex of the General Completive con­
jugation:

370. f. *Minn 'a - annd-in - i i .

me you know
GEN COMPL ACT

However, unlike their a, and c. counterparts, they are not in a

relative clause.

One way of explaining the occurrence of a relative conjugation

with a focussed NP is to postulate that sentences such as 3 7 0 -b. and

 274

371.b. are derived from sentences such, as their c. counterparts. The

latter are in fact pseudo-cleft sentences. Further research will, it
is hoped, prove whether this suggestion is viable.

Before leaving this matter, I should like to point out that

sentences like 370.b. and 371.b. also have, cleft counterparts. These
involve a topic NP containing a relative clause, followed b y the
focussed NP and either a dummy subject, cfum 'it' (cf. 370-5- and 3 7 1 *d.),
or a clefting particle wondbn (cf. '370.h. and 371•e.):

370-g- Mo 'annd-in - 0 - cfaa, cTum m i n n .

whom know you it me
EXT VAP
‘Whom you (Sg) informed, i t ’s. me. ‘/ ' I t ’s me you
informed.*

h. Mo 1annd-in - 0 - cfaa, minn woncfon.

whom know you me is
EXT VAP
'Whom you informed,, it's m e . '/tit's me you. informed. *

371 .d. Ko sood-u-mi, cfum agoogoyel.

w h a t buy I it watch
VAP
'What I bought, i t ’s., a watch.'/.'It* s a watch I bought.*

e. Ko sood-u-mi, agoogoyel wondbn.

w h a t buy I watch is
VAP
'What I bought,, it's a watch. '/'It's a watch I bought.'

These also need to be considered in the investigation of sentences with

NP focus.

6.2.2. Verb-focus

By verb-focus, I mean the focussing of a verbal root in a complex. This

is realised b y several methods. The most obvious and frequent is
contrastive stress on the root, which. I mark in the following examples
by means of underlining. However, m y main concern here will in fact
be the syntactic and morphological, rather than the prosodic, exponents
of verb-focus.
 275

6.2.2.1. Completive Aspect

In completive aspect, the Emphatic Completive conjugation is normally

used in a positive clause with verb focus:

372.a* 'A - hokk-a ngaa mbabba na?

you give this donkey
VAP QU
'Were you given this donkey?'

373.b. ©6 - ngar - 0
they come
VAP
'They came. '

Often, such a clause is accompanied by another, constrasting, clause:

372.b. 'A - hokk-a ngaa mbabba na, koo 1a - wuj j-u-nga?

you give this donkey or you steal it
VAP QU VAP
‘Were you given this donkey, or did you steal it?'

373.b. Naa cfoo be - ndany-aa, be - ngar-0.

not here they bear they come
VAP VAP
'It's not here they were b o m , they came '/
'They weren't b o m here, they're immigrants.'

The contrasting clause may similarly contain a focussed verbal root (such
as wuj j- in 372.b.); but the focus may be elsewhere - in 373.b.,
it is on the adverb, cfoo.

An optional accompaniment to the Emphatic Completive in completive

aspect is the clef ting particle wond b n , which was mentioned at the end
of the last section (§6 .2.1.7.), and illustrated in 370.h. and 371.d . .
It is placed after a clause whose verb is focussed, if the verbal
complex is in clause-final position. For example:

373. c. Naa cfoo be — ndany-a, be - ngar-0 wondbn.

not here they bear they come is
VAP VAP
'It's not here they were born, (it's that) they came . 1
 276

As the translation of 373.c. suggests, woncfon in a sense focusses

the whole verbal complex. However, &ince within a complex the verbal
root is the only constituent that is eligible for focus, the effect
of woncfon is to focus the verbal root. (& focussed pronoun is nec­
essarily fronted,, and cannot stay within the complex: cf. the
discussion in §6 .2 .1 .2 . above.)

A few speakers accept the use of the General Completive con­

jugation in place of the Emphatic Completive in sentences such as
those of 372. - 373.. Thus, 372.c. is marginally acceptable:

?372.c. 'A - hokfc-aama ngaa mbabba na, ko *a — wujj— ii-nga?

you give this donkey or you steal it
VAP QU VAP
GEN COMPL GEN COMPL ACT
PASS
'Were you given this donkey, or did ..you steal it?'

However, any other positive conjugation is totally unacceptable. For

example, neither the Relative Completive, nor the Stative can be used-
(Since sood- does not occur with the Stative, I have used a different
root in 37'4. .)

372-d. *Kokk-a-cfaa ngaa mbabba na, ko ngujj—u-cfaa-nga?

give you this donkey or steal you it
REL COMPL QU REL COMPL ACT
PASS

3 7 4 . a. Naa mi-cfaan - i , mi-joocf—i nii.

not I sleep X sit just
VAP VAP
'I'm not asleep, I'm just sitting . 1

b. *Naa 1e-mi-cfaan-il, 'e-mi-joocf-ii nii.

not I sleep I sit just
STAT MIDDLE STAT MIDDLE

372.d. is unacceptable because of its complexes of the Relative Comp­

letive, and 374.b. because of its complexes of the Stative. 374.a.
however has complexes of the Emphatic Completive, and is therefore
acceptable. The fact that neither the Relative Completive nor the
Stative can be used in a clause with verb-focus shows that certain
distinctions are neutralised with verb-focus, just as they are with
NP focus (cf. §6.2.1.!.).
 277

It is perfectly possible for verb-focus to occur in a complex

of the Negative Completive. For example:

375. Naa Be — piil-aaki to, Be - kad d - aaki to.

not they put not yet they gird not yet
head-tie
VAP VAP

'It's not that they haven't yet put their head-ties

on yet, they haven1t put their w r a p p e r s on y e t . '

In such instances, focus is realised by means of stress, and by the

juxtaposition of contrasting clauses. There is no special negative
conjugation for verb-focus.

6.2.2.2. Incompletive Aspect

In incompletive aspect, there is no special conjugation for verb-focus,

either positive or negative: the General Incompletive or the Negative
Incompletive is used. Thus, the incompletive equivalents of 371.a.

and 374.a. are:

376.a. Be - "ngar-ay
they come
VAP
GEN INCOMPL ACT
'They'll come . 1

377.a. Naa mi "cfaan-oto, mi jjoocf-oto nii.

not I sleep I sit just
VAT VAP
GEN INCOMPL GEN INCOMPL MID
MID
'I won't go to sleep, I'll just sit down.'

In such sentences, verb-focus is realised by stress (marked by "), or

by means of the contrasting verbal roots (cfaan- vs. joocf-) .

A Relative or Progressive complex cannot occur in a positive

clause with verb-focus. Thus the following are unacceptable variants
of 376.a. and 377a.:

3 76.b. *Be - ngar-ata. c. *'E-Be - cfon-ngar-a.

they come they come
REL INCOMPL ACT PROGRESSIVE ACTIVE
 278

377.b. *Naa cfaan-otoo-mi, njoocf-otoo-mi nii.

not sleep I sit I thus
REL INCOMPL REL INCOMPL MID

c. *Naa 1e-mi-cfaan-o o , 'e-mi-joocf-oo nii.

not I sleep I sit thus
PROG MID PROGRESSIVE MIDDI£

Thus, when verb-focus occurs in Incompletive aspect, the distinction

between the Progressive, the Relative Incompletive and the General
Incompletive is neutralised. (It is not yet clear whether the Vague
Incompletive can occur with verb-focus.)

A complex of the Subjunctive (,cf. 378.) , or of the Negative Incomp­

letive (cf. 379.) is perfectly compatible with verb-focus:

378. Ndi 'un - ee koo ndi nam - ee?

it pound or it grind
VAP VAP
SUBJ PASS SUBJ PASS
'Should it be pounded, or should it be ground? *

379. Naa 'o-yerd - ataako natt-ugo asbiti, 'o-yamcfin-taako wondon.

not he agree not enter hospital he cure not is
VAP INF VAP
GEN INCOMPL MID GEN INCOMPL MID
'It's not that he refuses to go into hospital, it's that
he can't be cured.

6.2.3. Focus: Conclusions

Three other points are worth mentioning in conclusion- Firstly, NP

focus and verb-focus seem to be mutually-exclusive, ie. they cannot
co-occur in the same clause. For instance 380.a. below has verb-
focus (on h e t t - ) , and 380.b. has NP focus (on doyaaje)? but these
cannot be combined, as in 380.c .:

380.a. Faatu hett-u doyaaje.

Fatu peel yams
VAP (<Hausa)
FOCUS
'Fatu peeled the yams.'

380.b. Doyaaje Faatu hett-i.

yams Fatu peel
FOCUS VAP
'It's yams that Fatu peeled.'
 279

380.c. *Doyaaje Faatu hett-u.

Secondly/ it is difficult/ if not impossible, to have NP focus in a

relative clause. So, while 381.a. below contains a 'normal' relative
clause, and was accepted by all informants, 381.b. was found to be
only marginally acceptable, with focus on Faatu:

381.a. Raa doyaaje cfe Faatu hett-i.

see yams that Fatu peel
VAP

'Here are the yams that Fatu peeled.'

b. ?Raa doyaaje cfe Faatu hett-i.

see yams that Fatu peel
'Here are the yams that Faatu peeled.'

Thirdly, it is impossible to have verb-focus in a relative clause.

Thus, 381.c. is completely unacceptable:

381. c. *Raa doyaaje cfe Faatu hett-u.

see yams that Fatu peel

No doubt further research into the syntax and semantics of focus

constructions will help to explain these restrictions on the occur­
rence of focussed constituents. For the moment, what is important
is the fact that there is ample syntactic and morphological evidence
for the establishment of the category of focus in Fulfulde, and that
we are thus able to account for the occurrence of a relative conju­
gation in sentences with a fronted NP, and the occurrence of the
Emphatic Completive in sentences that involve verb focus plus

completive aspect.
 280

Chapter 6 : Footnotes

1. Adverbs in Fulfulde bear comparison with NP's, and this is

especially evident in the way adverbs are focussed. Whether
adverb, or adverbial phrase, is a useful syntactic (as opposed
to morphological) category in the language is not yet certain.

2. Sentence 326. is ’neutral 1 and non-verbal, with a topic. Non­

verbal sentences involving a focussed NP have yet to be invest­
igated .

3. I say 'seldom' rather than 'never', because zero object is

sometimes found, for example in recipes, where the object is
obvious. An example of zero object is in fact found in sent­
ence 61, page 72: here, however, the object is coreferential
with the preceding object, rather than with the topic.

4. The meaning 'Here he is hugging someone else's child' would be

rendered by Raa mo 'e-mo wuund-ii fcinngel m m / m u d h m .

5. I do not attempt a formal definition of 'discourse' here: I

assume that it refers to a series of sentences, in. dialogue
or monologue or amongst many peppier which treat a particular
theme. For further discusssion see, for example, Grimes (1975).

6. It has not yet been ascertained whether it is useful to d i s ­

tinguish between different sorts of focus (other than NP or
adverb focus on the one hand and verb-focus on the other).
The distinctions proposed by Dik, within the framework of
Functional Grammar, have yet to be investigated in Fulfulde.

7. This sentence raises the question of the possible differences

between focus on the verbal root and focus on the VP - if
indeed the latter is a useful syntactic category for Fulfulde.
This issue will be shelved, since it is not crucial to the
establishment of verb-focus in the language.
 281

Conclusion

In this conslusion, I shall summarise the findings of this study, and

then point the way to areas of future research.

My principle aim has been to present a systematic description of

Fulfulde verbal morphology, and thus to improve on the rather ad hoc
nature of previous descriptions. However, morphology cannot be considered
in isolation, and this study has therefore drawn on other levels of
linguistic description, particularly the phonological and the syntactic.
Phonological considerations have been used to explain such, features as
syllable structure and the occurrence of anaptyctic ij; and syntactic
considerations have been used to account for the distribution of certain
conjugations, and to establish the category of Noun Phrase, for example.
The border areas between morphology and syntax, and between morphology
and phonology have also been explored. Thus, the suffixal precedence
hierarchy and the occurrence of vowel harmony, for example, involve
morpho-phonology rather than pure phonology or pure morphology. Simi­
larly, raorpho-syntactic considerations are involved in explaining such
issues as morph-shape and morph-order within the verbal complex.

This study has provided new insights into the morpho-phonological,

morpho-syntactic and morphological principles that operate within verbal
complexes. Most importantly, it has highlighted the principle of sensitive
realisation of verbal properties, and within the description of this
sensitivity, the suffixal precedence hierarchy is the most significant
'discovery'. It has also shown that morphological patterns exist between
conjugations, despite the initially bewildering array of suffixes they display.

However, many of the non-morphological observations made in this

study are also new, or are presented in a new way, which it is hoped
will make Fulfulde more accessible to linguists seeking exemplification
of theoretical issues. Amongst these are the operation of consonant
alternation in nominal class' suffixes and in root-final position (see
§1.2.3. and §1.2.4.); the identification of different types of sentence,
and particularly the difference between sentences expressing existence
or location on the one hand, and those expressing attribution or identi-
 282

fication on the other (see §2 .1 .2 .); and the establishment of the

categories of topic and focus, and their syntactic, morphological
and phonological correlates (see §2.3., and chapter 6 ).

While this study has no doubt made a contribution to knowledge

about Fulfulde, it has at the same time exposed many gaps. I shall
therefore mention here some of the most obvious areas of possible
future investigation.

In chapter 1, certain observations were made concerning the

distribution of different types of consonants in root-final position,
and these observations were correlated to the system of consonant
alternation. A similar analysis of stem-final consonants may well
account for the distribution of the grades of nominal-class suffixes:
in other words, I believe that there may be a correlation between the
final consonant of a nominal stem and the grade of suffix with which
that stem combines. Another line of investigation that arises from
chapter 1 is the distinction, if any, between nouns and adjectives
in Fulfulde.

The non-verbal sentences discussed in chapter 2 raise many issues,

amongst which is the syntactic status of the constituent that is commented
on or predicated. Further research may well prove that a sentence such
as Nyaako 'am Sarkin Fulaani Naraayi (example 54.a., page 70 ) starts
with a topic NP rather than with a subject. The borderline between
non-verbal sentences expressing attribution, and those expressing ident­
ification also requires further investigation.

It will be remembered that various morphological and morpho-

phonological features of participles were invoked in order to dist­
inguish them from verbal complexes (see pp. 77-78) . Now, although the
distinction is valid, this is not to deny that participles and verbal
complexes are extremely similar in many respects: their morphology,
the verbal properties they realise, and in the operation of the suffixal
precedence hierarchy. At least two lines of enquiry arise with regard
to participles. Firstly, their morphological similarity with the
relative conjugations suggests a diachronic link with the latter (see
McIntosh 19 81). Secondly, the use of a participle as a 'predicating NP'
 283

in a non-verbal sentence is a common alternative to the use of a verbal

complex (in a verbal sentence); and it is as yet unclear whether there
is a semantic distinction between these two syntactic constructions
(cf. example 52.a. and footnote 10 chapter 2).

On the morpho-syntactic level, the status of the category of mood

has been thrown into doubt by the fact that there is no obvious morpho­
logical exponent of the indicative vs. non-indicative distinction,
and b y the fact that the Subjunctive conjugation apparently straddles
the two moods. Further investigation may prove that this category should
be dispensed with completely, or that a category of a different character
should be established - perhaps factivity vs. non-factivity?

Another morpho-syntactic issue concerns the status of certain

constituents inside and outside the verbal complex. The status of *e ,
cfon and of the anteriority-marker are particularly problematic when
they occur in non-verbal sentences, for there are no content words
for them to 'lean on' (as clitics), or to be appended to (as suffixes,
or, for that matter, as prefixes)? yet their word status is also
doubtful. Within the verbal complex, cfon in particular seems, in some
sense, to intrude, and this idea is corroborated b y the fact that
cfon may be replaced by other deictic adverbs such as ton and cfoo.

The discussion of topic and focus in chapter 6 is simply intended

to establish the categories and some of their correlates. However,

there is much room for fruitful research in this area. As well as the
occurrence of topic in non-verbal sentences (mentioned above), topic
and anaphoric reference, and the-use of topic-markers are two obvious
lines of investigation. As regards focus, the relationship between
sentences with. NP focus and cleft and pseudo-cleft constructions has
yet to be more fully elucidated, as has also the relationship between
NP focus constructions and relative clauses, both of which involve a
relative conjugation.

This study has been concerned with a relatively unknown dialect.

However, comparison with other works on Fulfulde reveals that the
morphological, syntactic and phonological issues in other dialects
are very similar to those raised here, though they may differ in detail.
For example, it is clear that many dialects do have the C W C and CVCC
 284

syllables that are normally unacceptable in Kaceccereere; and the

suffixal precedence hierarchy operates in some dialects in different
conjugations from those in which it operates in Kaceccereere. I believe
that the findings of this study will be useful for the study of other
dialects of Fulfulde; and it is to be hoped that one day, the further
research suggested here, and more besides, will be carried out, whether
on Kaceccereere or on another dialect or dialects.
 285

Appendix A

Summary of the Morphology of the 14

Conjugations

This appendix provides a summary of the morphology of the 14 conjuga­

tions. Although much of the discussion of the morphology has centred
around the distinction between SV and VS complexes (cf. chapter 4), a
three-way distinction will be made here. First, the morphology of
SVO complexes will be summarised (pp. 287-292), then that of VSO com­
plexes (pp. 293-296), and then that of VOS complexes (p. 297).

In considering the tables, the following points should be borne

in mind:

1. The place of the verbal root amongst the cluster of appendages

given in each table is marked by a double vertical line: in fact, it
always immediately precedes the VAP-suffix. Its form is invariable,
apart from a certain amount of phonetic assimilation (cf. example 32.b.,
p. 41).

2. To form a verbal complex of the conjugation indicated in the right

hand column of the table, any one subj ect pronoun may combine with the
VAP-suffix and any one object pronoun in the same line, semantic com­
patibility allowing. For example, in the Emphatic Completive Active
(see p. 287), any subject pronoun except mi- may combine with 0_ VAP-
suffix and -am object.(Since there are no reflexive pronouns in Ful­
fulde, a co-referential subject and object may not co-occur in a com­
plex.)

3. Only the form of a first object pronoun has been given. A second
object pronoun will always have clitic form (cf. table XXIII, p. 197),
and follow the last morph shown in the table. Thus, a second object
pronoun will follow immediately after the first object pronoun in SVO
and VSO complexes, but immediately after the subject in VOS complexes.

4. The 'basic' form of a VAP-suffix is underlined: this is the form

of the suffix that occurs in complex-final position, as discussed in
§4.2., pp. 185-192.
 286

5. Concerning the distribution of vowels in parentheses, see 14.2.2.2..

6. The pronoun be_ represents all class pronouns: c f . p. 65.

7.' An arrow indicates exceptional order. For example, -am and -~g_
precede rather than follow -no; and -no- 1intrudes 1 into the -aama
VAP-suffix for the General Completive Passive. These matters are dis­
cussed on pp. 179-180 above.

8. The significance of varies according to context:

(i) in the ANT column indicates that complexes without the
Anteriority-Marker are being considered. (It should be
remembered that certain conjugations, such as the Imperative
and the Subjunctive, never combine with anteriority.)
(ii) is also used for collocations that do not occur: for
example, coreferential subjects and objects, as mentioned
in 2 . a b o v e .

9. Further comments on the morphology of individual conjugations are

found in the footnotes on page 297.
 287

Appendix A: SVO Complexes

on

<M

•H

r4 'H

on
LTN

OJ

CQ

-p -P on on
Complexes

on
■xsi-h Iisl *h rciI oJ

cQ cQ cQ cQ tQcQcQcQ cQ cQ
SVO

OJ

I—I
'H *H
CQ

on i
—i i
—i i
—ii—i i
—i
<; <d < <

H *H •H *H

•H *rH

•H

o o
 288

0 0) 0 0 <D 0 (L) 0 0 0 0 0
cQ eQ
on I eQ cQ cQ eQ cQ I I I cQ | | cQ I cQ cQ tq
Eh PJ £h S rt rt rt rt rt
I O O O O O I I I I I i o o o o
ft
a rt rt e3 s rt rt rt rt rt rt rt
I 0 0 0 a> 0 I I I CU I 1 0 ) 10 0 0 0 0

rt a art rt rt rt rt rt rt rt
•r4 *H *H *H *H •H *H •H
i aa a s a i i i a i i a i 's '§ S S

o o o o o o o o o o O o
i a a as s i I l S I I S I S S S S S

cd rt rt rt rt rt rt rt rt
OJ I S S S S I 0 I I 0 I 0 I S S S I S
%
taq
rt \
•H
CO S 1 1 I %
ft l - I rt I I i i- - IM -

ft c—
O o o o o o o o
I i rt i rt I rtv i i i rt rt rt I I rt | i i rt

t—
(cont)

o O 0 o\
■p -p p p p p ^P ^ P
—*s^-- •H
ft rtl rt
VO -H -H rt 0 j>j rt 0 0 O O 0 rt
•H ■HI■H — '—' rt rt rt
Complexes

0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
on cQ eQ cQ cQ cQ eQ cQ cQ cQ cQ cQ eQ cQ eQ eQ cQ eQ eQ cQ cQ
rt
rt i
—1 rtrt rt rt rt rt rt rt rt rt rt rt rt rt
o o o o o o o o 1 1 o 1 1o 1o o O 1 o o
£h rt OJ
SVO

o
rt rt
ft rt rt rt rt rt rt rt rt rt rt rt rt ’ rt rt
rt •H
i— 1 0 0 0 0 0 0 1 1 0
— 1 1 0 1 —0 *-
0
—0 1 •-
0 0
—
P
O i
—1
0
•r-; ft M rt rt rt rt rt rt rt rt rt rt rt rt rt rt rt rt rt rt
ft 0 ■h •rt •rt*H •rt•rt •rt•rt•H •rt
*ri 'rt •rt *H •rt *H •H
rt T—’ 1S s S s s s 1 s s s s S a a s s s s s
CO
ft
O t o o o o o o o o o o o o o o o o o o o o
M. on
rt rt
o H
0 rt rt rt rt rt rt rt rt rt i rt rt 1 rt 1 rt rt rt i rt rt
rt hO OJ
0 rt
ft ■H
CQ *^ ‘H •rt■rH •fl•rt •rt*H *H •rt •H *I
-1 *rH •rt*H *rH
i S s s S S s 1s s I s s s s S S s s S

co
0
p ft 0 p ft 0
o •H rt o •H rt o
< ft < s ft C
i
—1
ft
S ft
ft o ftt
r
S O
o rt O
o m O
rt faD
0 0 0
cO o Ert
 289

0 0 0 0 0 0 0 0 0 0
00 cQ cQ eQ cQ cQ a £ eQ cQ & 1 eQ eQ 1 eQ eQ

'on
rt rt rt rt £

1on

'on
1on

on
0 0 0 0 O 1 1 1 1 1 1
cvj

rt £ £ £ rt £ rt £ £ £
0 0 0 0 (DO) 0 0 0 0 1 1 1 1 1 1

£ £ £ £ rt rt rt rt £ rt rt rt £ rt
'3 ’3 '3 '3 '3 ’3 ’3 1 *3 '3

Q Q Q Q Q Q 0 0 0 0
co § § a a a a a a 1 a a 1 a a

£ £ £ £ £ £ £ £ £ £
OJ a a a a a a a a a a 1 1 1 1 1 1

a a a a a a a a
£ £ £ £ £ 3 £ £ a
>>- ^ - - - - - £ I I-

EJ
no

no

no

no

no
1 1 1 1 1 1 11 111
0 0 0 0
(cont)

£ £
M£ £ 0£5 0£5
£ £ £ £
•H •rt £ £ ps
--N<
P■ -P.P pV -p CO .
Pi
< M05 M£ kS£
-
£ £
^—
£ £
.—“
£ £
^-s,1
co rt 0 001l
> £ £ £ £ N^ '— '' ' '— ''— ' 'C5>- 101 TS. £|
Complexes

0 0 e0Q e0Q 0 0 0 0 0 0
CO eQ eQ eQ eQ eQ =Q eQ eQ 1 1 1 1 1
rt CJ
£ 1
—1 rt rt rt rt rt rt rt rt rt rt rt rt
c
rt £ CM
O 0 O 0 0 O 0 0 O 0 1 1 0 1 1 0
SYO

0
rt rt
Pm rt rt rt rt rt rt rt rt rt rt rt
P
£ •H
*—
0 0 0 0 0 0 0 0 0 0 1 1 1 1 1 1
a 1
—1
0
Pm X rt rt rt rt rt rt rt rt rt rt
0 *rt
■£
cn
a '3 '3 '3 ’3 ‘3 *3 ‘3 '3 ’3 1 1 1 1 1 i

<P
0 0 0 0 0 0 0 0 0 0 0
rt CO 1 1 1 | ( i
rt £
0 H CM
0 P £ £ £ £ £ £ £ £ £ £ £ 1 £ £ 1
rt Ed CJ
0 rt
Pm
CQ •rt 'rt •h •H •H •P •1
H •H •H
T a a a a a a a a a 1 1 1 1 1 1

cn CO
nd CO -p TP cn P np
t—! £ 0 *H £ O •rH
s Pm < s PM s
rt 1
—1 rt
0 &
si 0
1
—1
p & 0 a
£ a 0 1—1
EC 0 3
£ 0 1
—1 —
'|
—c; rd
rt EO hO rt
O 0 0 O
O K K
 290

0 0 0 0 0 0 0 0 0 0 1 0 0
cn =Q eQ cQ cQ eQ =Q 1 ed eQ 1 eQ eQ eQ 1 cQ

d
d d d d d d d d d d d
o 0 1 0 1 0 0 1 0 0 10 0 10 10
d <— 1 CM
o
d d
CL, d d d d d d d d d d d
d •H 1 l CL> I <L> I C D I C L ) 1 0 1 0 I 0) I CD 0 1 0
+3 1 -
o d
0
■i“3 — 1 X
1 d d cl d d d d d

O CL,
0
1 1 a
•H
1 a
*rH
1 t 1 ’g 1 "g 1 "g 1 g 1
-P
cn CM
d
•H
ft, 1 1 a 1 a 1 a 1
CM
1^9
a 1 2
a i 9
a 1 9
a
Ch Lf\
O d d d d d d d d
d d d d $ S! £ Si
CM 0 a 1 a i a 1 a a i a 1 a 1 a 1
d
O

d
cn
0
td
•rH ■H
PM m
I I I I 'S ■g 1 a 1 ■g

EH O o o o o
o o o o o o o o o o
I I I d d i 1 d d 1 1 d d 1 t d d 1 1 d d
(cont)

o o 0) CD
-p -p -P -P
d I>s s>s O O O 0 0 0
> TS-— ■ cc5[cd od O Ol-— -'—- (D CUl— --
Complexes

CD CD CD CD 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
cn cQ cQ cQ eQ eQ cQ cQ eQ eQ eQ eQ eQ eQ eQ eQ eQ eQ

d
d rH d d d d d d d d d d
O 1 10 f O 10 0 0 1 O O 0 1 0 1 0
d d CM
SVO

o
d d -
CL, d d d d d d d d d d d d
d ’H I I 0 I 0 1 0 0 0 1 0 1 0 0 0 1 0 1 0
-p 1—1
0 I—1
0)
CP * d d d d d d d d d d d d d d d
0
1CD H
■rH
a
*rH * H
a a a a
*H »H
■a *g *g * g’g ‘g ' g ' g ’g l
<U
o 0 0 0 0 Q 0 Q 0 n 0 0 0 0000 00 00
a a a a a a a a a a a a
d
o
cn d d d d d d
d 1 d 1 d 1 d 1 d 1 d 1 d I d I d I d 1 d
0 CM
Ph
mi

thi
mi
mi
TUT
mi

mi

H t a g *g a a a a a

0
•p ed 0 -P
0 ■H d O *rH
< s Ph <
0
o >
•H •H I— I
-p 0 &
d 0
bC 0 o
d d O
b0
d' O
o d 0
o Ph PC
 291

0 CD (D CD 0 CD 0 0 0 0
00 | cQ | cQ cQ I eQ I tQ I tfl | eQ I eQ I cQ I cQ

Pi Pi Pi Pi Pi Pi pi pj Pi Pi
I O I O 0 1 0 I O I O I O I O I O I O
OJ

pi p ! pi pi pi pj Pi ■ Pi
I CD 1 (D 0 10 1 0 1 0 1 0 1 0 10 10

ri Pi Pi Pi p! pj Pi Pi pi

t 1 '0 1 1 a 1 '0 1 '0 1 *0 t 1 -a

0 0 0 0 0 0 0 Q
00 a 1 a 1 a 1 a 1 a 1 a 1 a 1 a
d d d d d d d d d
d d d d d d
CM 1 a 1 a a 1 a 1 a 1 a 1

I I I

o 0 0 0 •H O
o o 0 0 0 0 0 0 g o o
I I Pi Pi 1 pi pi 1 1 pi pi 1 i p; pi 1 a p! Pi
(cont)

0 (D
0 0 0 0 0 0 0 0
-p p •P -p p> -P +3 -p

til b b 0 0. 0 0 0 0 0 0
d d l d d 5 -h I-35
Complexes

0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
cQ cQ cQ cQ eQ tQ cQ cQ cQ cQ cQ cQ eQ eQ cQ eQ cQ cQ eQ
Pi O
pi 1
—1 Pi P! Pi Pi PI Pi PJ Pi Pi Pi
0 I o I o O I O I O I O I O I O I O I O
pi d CM
SVO

0
u P
Ph Pi Pi Pi Pi Pi Pi Pi pi Pi Pi Pi
pi ■H 10 10 0 1 0 ] 0 i 0 1 0 1 0 10 10
P t—
O
0
•i-p Ph X Pi Pi pi Pi p i p i p i p i p j p i p i p i p i p i p i Pi Pi Pi Pi
0
Pi
CQ
T— ’0 '0 a '0 t a i ^ ^ ^ ^ ‘0 ‘0 ’0 ‘a
Ch
0 O O O O O O O O O O O O O O O O O O O
Pi on - a a a
Pi d
0 H
0 p! d d
Jh bC d i d 1 d 1 d 1 d l d I d 1 d
0 Pi CM
Ph •H
CQ *rH ‘rH * H •rH •rH *H
a a a a a a ■0 ‘0 ‘0 t a a a ‘0 I

m
-p 0 -P
d a d O
Ph s Ph <c
1
—1
Pi
(
—1 a
!■
0
0
0>
O pi •H
o W p
d
H
0
—
1
0
1 p
CQ
Ph <x,
 292

0 0 0 0 0 0 0 0 0 0
eQ t cQ 1 cQ 1 eQ eQ 1 eQ 1 eQ eQ eQ eQ

H rt rt rt rt rt rt rt rt rt rt
O 0 1 0 i 0 10 0 10 10 O 0 O
rt rt CM
0
rt rt
ir t rt rt rt rt rt s s s rt rt rt
rt •H I 0 I 0 I 0 I 0 a) la ) 1 <u CD 0 0
-p T-
0 1 —1
0
•r-); (rt X rt rt rt rt
rU 0 ■H
O T- 1 a 1
-P
cn
rt
■H 0 0 0 0 o o
ph on 1 a i a 1 a 1 a a a
0 rt rt rt rt rt
rt rt rt rt I rt LT\ rt rt rt rt
rt H CM a 1 a 1 a a 1 cd a 1 a a a
0 rt
cn bL
rt rt1
Cl) *H a a a
Ph co *H rt rt rt
I I 1 a

o o
0 0 0 0
1 1 rt rt 1 1 rt rt 1 1 1 1 1 1 1
(cont)

on rt rt
a a
•HI rt on rt , o| sdI 0 0
*H *H 1'H <h rt rtl rt rt TS l ' —-■ r t | o 01 CD CDI 0 0
Complexes

0 0 0 0 0 0 0 0 CD 0 0 0 0 0
on pQ pQ r(~^ pf] p(*^ pQ r(~| eQ eQ eQ eQ eQ eQ

rt
rt 9h rt £3 rt rt rt rt rt
0 1— 1 I O I O I O I O O 1 O O 0 O
rt rt CM
SVO

0
rt rt
Ph rt rt rt a rt rt rt
4j
rt *H 1 1 1 0 ) I 0) 10) 0 (D <D
v“
O r-1
0
• X rt rt rt rt rt rt rt rt rt rt
rQ
rt
(rt
0 H
a
•H *i—l *r-j *r-|
a a a a 'I t
•h ‘P
a a
CO
<P
O 0 0 0 0 0 0 0 0
rt on a
c rt
0 H
cn rt rt rt rt rt
rt hC CVI 1 rt 1 rt 1 rt 1 rt I 1 rt l rt l rt rt rt rt
0) rt
(rt *H
CO *1—
3
a a 'g 'ft 'e! a a a a

CQ cn 0
-P rrt 0 -P TP 0
•h rt 0 -H rt 0 •H rt
S Ph < S Ph ert ^ Ph

0 a
t> 0
•H 0
-P c
0 0 pi
> rt
•H rt 0
-P " ”3 rt
rt .Q &D
-P
0
0
'rt rt
CO >
 293

Appendix A: VSO Complexes

CD 0 0 0 0 0
cH on cQ cQ eQ <Q cQ cQ
-p a3
CD Pi PI Pi Pi
0 U o ■»o I I 1 1 O O I I I I I I
CM _

Pi
§
iH X P PJ Pi Pi PJ PJ
-p
cn 0 •H •i-i
Pc Ph V* 1 1 1 1 I I a a 1 1 I 1 ■at
‘H
pH
CM CM CM
'H
O QO oo oo oo o o o o o o o
on a- a a a a a aa a i a a a a a a
Pi CM
O CM CM
tfl CdrH
?H CM ctf 03 I I II II o3 I i i ii ii
■
0
Ph
a a a
1 -I i §
i i 11 I I - i i 1 1 ii - -

PJ Pi Pi Pi
-f-* CM II It I I o o II II I I o o
O o - ”
<U
•r-j T— pj pj pi PI
II II 0 0 II II II 0 0 II
■3
cn
Complexes

CH
o
13 cd <d <d ed
pi £d . Si «3
o % CM i i e eG ti it I I tG <d II II
CQ
u Pi
0 ■H
Ph CQ a i i ii ii 11 11 11
YSO

o o o o o o
o o o o o o o o
1 Pi I Pi I Pi I Pi 1 Pi I pi I Pi 1 Pi

S3 pS pi pS pi pi pi2 pi -H *H *H
> -— ''— ' '— ■*'— ''— — * -H *H "H ’H -H »H *H »H

0 0
>
•H d
•d
-P 'd
O
<U s
Pi
O
•H ■
—I
tj t
o
Pi o
pf <—I
O CD
CD Ph
 29k
i

0 0 0 0 0 0 0 0 0
r0 0 0 0 0 0 0 0 0 0 0
cn cQ »f^ cQ cQ fQ cQ cQ cQ cQ cQ oQ cQ cQ c£l nO cQ cQ cQ cQ
1— 1

o3
i i i t it II II II I I II II
Pt CM

<— 1 X c pi Pi 0. Pi Pi me pi a
0 •Jjj *H *H *H*H
PM t i l l S i 'ii "il i i *3 *3 a a ii a a

CM

O O O O O O o o o o o o
oo a a a a a a g § a a 8 §8 i a a s & & s
CM

a CM ii it ii I II II II
bd
n
■H •H ■H
CO i i i i - - i a a i i i as i i - -
(cont)

Sh PI Pi Pi S
I I II o o I I I 1 S O I II I I O O
CM O -
p i
—1
o cd
0
Complexes

•n: 3 pi Pi Pi P! PI
-P4i P
iH •H 1 1 0 0 I I I Pi 0 I I I I I 0 0 I I
M r— — — 0 -
CO

ft
O m cd 03 a3
cd cd oJ a3 a3 cd
Pi r-j 1 tJ I 1 I aj *73 i l | | I a3 t J I I I I
o rj
CM
VSO

0 ht
P Pi
0 ■H
Ph cn r- a i i i a i i i i i i a i i i i i i

o o o o o o
o o o o o o o
i i I Pi I PI I Pi I Pi 1 Pi I Pi I Pi
O o o
o o o o o o o
P p -p -p p p p p
o3 cd 0 ° o o o o o o
> o3 03 cd -— oj o3 - cd

CO

0
t> 0 0
■H >
0 •H 3
0 p
03 CJ ■H
ft < s
I-1
ft
a
o
& CJ
O Pi
o H
-1
E i
—1
0 0
CP ft
 295

<d a> 0 a> 0) OJ 0) 0 0 0 0 0 0

PQ eQ cQ eQ eQ cQ cQ cQ cQ cQ cQ cQ eQ eQ

I f I ) I 1 I i I
OJ

g g G G G
•H *H •H *H
i s i 8 S I 8 '1 I

OQ i
2
8 9
8 9
8 § § i 9
8 2
8 2
8 i §
CM l i i i
bOl
G
•rH
I I - ■a
(cont)

CQ

G
I I I O I I O
Complexes

CM
p ■— I
o cd
0 U
•I— ; G G G
£ rH •H
i—
i i a) i I 0) 1
G Ph
rn

*H
VSO

o G cd cd
G cd cd cd
G rH CM cd l l o l l ' O i l
o G
cn 5.
u G
a) •H
IP CQ T— S i l l l i t I I I I I I

<L>
<1) CD 0 0
P P p P
—w /^*s
CD 0 0 0 O O oj 0
^ -— ^ ^ TSl ■©. O O O 0 <d a>

0 0
> CD (L) >>
•H > H *H
cn 'H td cn
cn P td cn
G O 'H G
Ph < s P„|
G i—1 0
O & t>
•H 8 •H
o P
Id a O
bCj G JZj
G H G
■r-3 ,r3
G <H P
O CD G
O PS CQ
 296

VOS Complexes

p —1
1
0 Pc QJ <L) <L) (1) 0)
0) 00 cQ cQ cQ eQ eQ

&
0

■cd 1— 1 fl rt
rt Pc O O
0 CM
CJ
<u
CQ
U bO
O CQ

P
00 § § 8 8
O
CQ
P bO
<D
Ph
CQ
CM S'S cd
a
cd
a

{50
m
Complexes

bD OO O o o
m OO O 2 9
cn a a a a a

bO cd cd cd cd cd
VOS

CQ 4 cd <ti cd a3
CM a a a a a

Em

o
p

>

<D OJ a> <u

> H t>
■h »d •H £
p td -P
0 -H O •H
<1 s
a —i
o Pc
*H 1
—1 a
p & 0
cd a 0
bC| 0 s
P 0 M
rt 1—1 1—!
o OJ QJ
o Ph crt
 297

Appendix A: Footnotes

1. In the Continuative Imperative, there is no morphological distinction

between active and middle voices (cf. £5.5.1.).
2. The subject pronoun in Imperative complexes is only evident in sur­
face structure if the complex is preceded by the negating particle, t o .
3. Although the distribution of 0 and (u) is morphologically deter­
mined (see §4*2.1.), I show-0 separately with -am and'-e, because
only 0 occurs with these vowel-initial pronouns.
4. No follows rather than precedes an -e^ or -am occurring in the same
complex. See £4.1.1.,
5. In the Exhortative, -e rather than -maa is used for formulaic greet­
ings such as 'Allah reen-e 'May God protect-you 1 1 Otherwise, and
-maa are apparently in free variation in active complexes of the Ex­
hortative, Progressive, and Subjunctive.
6. Concerning this VAP-suffix, see footnote 8 , p. 224.
7. The Anteriority-Marker 'intrudes' into the VAP-suffix of the General
Completive Passive (see 54.1.1.). It shculd also be noted that my
informants did not accept the combination of no and an object pronoun
in complexes of the General Completive Passive.
8. The Passive of the General Incompletive is not normally used: see
pp. 186 and 243.
9. Concerning the distribution of -aa and -aayi in complex-final posi­
tion, see S3.6 .1.2. (pp. 153-4), and §4,3.1. (p. 194).
10. In Relative conjugations, both SV and VS complexes occur with a sub­
ject that semantically includes first and/or■second person. For
further discussion, see §4.4.1., and footnotes 1 2 - 1 5 below.
11. In the Subjunctive, only SV complexes occur with a lsg subject ; but
with a 2sg, Ipl (incl) and 2pl subject, both SV and VS complexes may
occur. See 14.4.1..
12. In the Relative conjugations, VOS order normally occurs if (i) the
subject is lsg, (ii) the object is 2sg or 3sg (personal class), and
(iii) the complex expresses active or middle voice. However, SVO
order usually occurs if the complex contains the Anteriority Marker,
and is not of the Relative Completive Active; and VSO order usually
occurs if the complex is of the Relative Completive Passive without
the Anteriority Marker. Despite these generalisations, there is con­
siderable idiolectal variation in usage, and wherever more than one
patterning of morphs has been recorded, the alternatives are shown.
13. In the Relative Completive Passive, VSO complexes with an object pro­
noun as well as the Anteriority Marker were generally considered u n ­
acceptable, though those with a lpl (incl) or 2 pl subject were consid­
ered marginally acceptable, and have therefore been shown here.
14. For Relative Incompletive Active and Middle complexes with a lsg sub­
ject and the Anteriority Marker, my informants used SVO order.
15. It was only in a VOS complex of the Relative Completive Active that
my informants accepted the Anteriority Marker. Othersise, SVO order
was preferred if Anteriority was to be expressed in a Relative con­
jugation with lsg subject and 2 sg or 3sg (personal class) object.
 298

Appendix B

This appendix contains two fragments of recorded speech, which are

intended to convey an impression of the Kaceccereere dialect in conn­
ected speech, as opposed to the isolated sentences that have been
used as examples throughout the rest of this study.

Certain features are common to both texts: in particular, they

both employ certain 'Hausa-isms1, such as the 1discourse-fillers1,

or pause features waato (literally 'that is to s a y 1) and shi ke nan

(literally 'that's that'). Otherwise, the texts are fairly different
from each other. Text 1 is the beginning of a conversation between
Mallam Hassan Moturba and his father, Sarkin Fulani Narayi, concern­
ing the history of the Kaceccere'en. The style is fairly formal, for
Mallam Hassan has considerable experience of broadcasting, and adopt­
ed his broadcasting style for this recording. Text 2 is the recount­
ing of a recipe by Sa'adatu Rahaman Caawi, wife of Dawoobe Abayin.
Here, the style is more informal, though there are certain character­
istics that are peculiar to recipes. Such are, for example, the
frequent use of wadd- 'bring' or 'take', and the frequent use of a
zero object. Both texts display considerable repetition, but these
have not been 1edited o u t ', for the aim here is to preserve the
'naturalness 1 of the flow of speech.

The presentation of these texts is not quite the same as that

used for the examples given in the body of this study. Although a
morphological analysis of verbal forms has been given, I have not
marked absolutely every morph (for example, VAP-suffixes have not been
marked, since they are well documented elsewhere). However, I have
marked the conjugation and voice of each verbal complex, in order to
help the reader to see the use of theeonjugations in context. Below
each line of Fulfulde text is a literal translation of each word, and
below that certain syntactic or morphological pieces of information,
and then a free translation.
 299

Text 1

(HM = Hassan Moturba; SN = Sarkin Fulaani Narayi.)

HM: To, hannde kaddii, bano ko woow-aa, 'Allah wadd-ii-min

well today indeed as what usual not Allah bring us
NEG COMPL ACT GEN COMPL ACT
Well, today, God has given us the unusual opportunity to come

wuro, ngam min-nan-a habaruuji feerefeere dow no Kaceccere'en

home so that we hear histories various on how Kaceccere'en
SUBJ ACT
home, so that we can hear various aspects of the history of how

laat-or-i. Coo nii hannde, 'e-min-ngicf-i

become here thus today we want
REL COMPL ACT STAT ACT
the Kaceccere'en came into being. Thus here today we want

min-mbolw-id-a 'e Sarkin Fulaani Naarayi, ngam 'o-hokk-a-min

we speak with Sarkin Fulani Narayi so that he give us
SUBJ ACT SUBJ SCT
to speak with Sarkin Fulani Narayi, so that he can give us

habaru no Fulbe ngar-d-oy-i. To, ko min-ngicf-i

history how Fulanis come far well what we want
REL COMPL ACT REL COMPL ACT
the history of how the Fulanis came (from af a r ) . Well, what we want
1
min-'yam-a hannde, kanjum won-i, waato ..... 'En-'annd-i Fulbe
we ask today that is that is to say we know Fulani
SUBJ ACT STAT ACT
to ask today is, well, .... We know that we Fulani,
2
nii 1en-ngar-u feerefeere, wobbe Buulanko'en, wobbe kadii
thus we come variously some Bulanko'en some moreover
REL COMPL ACT
we came in various ways, some of the Bulanko'en (lineage), others

Dallaaji, wobbe Gijiya. To,

Dallaji some Gijiya
moreover of the Dalaji (lineage), others of the Gijiya. Well,

cfoo, hiddeko be - ngar-a, cfoo cfum Fulbe ngal-aa-no na,

here before they come here it (is) Fulanis live not
SUBJ ACT NEG COMPL ACT QU
here, before they came, were no Fulanis already living here,
3
koo boo a'a, Fulbe 1en-ngon-0-do cfoo?
or on the contrary no Fulanis we live here
or were we Fulanis .in fact already living here?
1
To, kanjum won-i hoore haala dow ko yewt - e t e e . To,
well that is head speech on what discuss well
REL INCOMPL PASS
Well, that is the introduction to our discusssion. Well,
 300

kanjum won-i ko min-ngicf-i .... mi-yicf-i m i - 'yam-0-maa,

that is what we want I want I ask you
REL COMPL ACT REL COMPL ACT SUBJ ACT
t h a t 's what we want ..... I want to ask y o u ,

Sarkin Fulaani Naarayi. To, bismillah ....

Sarkin Fulani Narayi well, in thename of Allah ...
Sarkin Fulani Narayi. So, please ....

SN: To, kanjum ngicf-0-cfe.a 'yam-go, ko?

so that want you ask
REL COMPL ACT INF
So, that's what you want to ask, is i t ? 1

HM: 'Ee. To, hiddeko Fulbe ngar-a cfoo, ndii leydi, beye
yes well before Fulanis come here this country who
SUBJ ACT T O P I C
Yes. Well, before the Fulani came here, who livpd in this area?
3
ngon-0-do cfoo?
live here
REL COMPL ACT

Text 2
4 5
Bete, no min-ndef-ir-ta <£um. Sey min-boh-a gawri. To
bete how we cook it must we bruise guinea-corn when
TOPIC REL INCOMPL ACT SUBJ ACT
4
As for how we cook 'bete'. We bruise guinea-corn. When we
5
min-boh — ii gawri sey min-njaar-a-ndi enjin
we bruise guinea-corn then we take it grinding-machine
GEN COMPL ACT S U B J A C T
have bruise it, then we take it to the grinding-machine

min-nam-a-ndi. To min-nam-ii-ndi, sey min-ngar-a, sey

we grind it when we grind it then we come then
SUBJ ACT GEN COMPL ACT SUBJ ACT
to grind it. When we have ground it, then we come and

min-secf-a. To min-secf-ii-ndi, shi ke nan, sey min-ngadd-a

we sift when we sift it well then we bring
SUBJ ACT GEN COMPL ACT SUBJ ACT
sift (it). When we have sifted it, well,then we take it and put

min-njonng-a.^ To min-njonng-ii, min-ngacf-a manja, min-ngacf-a

we bgil when we boil we put palm-oil we put
SUBJ ACT r GEN COMPL ACT SUBJ ACT SUBJ ACT
b
it on to boil. When we've put it on to boil, we add palm-oil,,

'inna li'o,^ min-ngacf-a maagi, min-ngacf-a borkonu^

mother soup we put maggi-cube we put pepper
SUBJ ACT 7 SUBJ ACT
we add 1locust-bean-cakes1 , we add a maggi cube, we add pepper,
 301

min-ngacf-a lamdhm, shi ke nanl.... Min-ngadd-a ceu, min-ngacf-a.

we put salt t h a t 's that we bring meat we put
SUBJ ACT SUBJ ACT SUBJ ACT
we add salt, there we are I.... (Then) we take meat and add it.

To ceu 1e dbn, min-ngadd-a min-ngacf-a. To min-ngacf-ii ceu

if meat is we bring we put when we put meat
SUBJ ACT SUBJ ACT GEN COMPL ACT
If there is any meat, we bring it and add it. When we have
9
to cfam -fat-ii, cfam-fat-ii, cfam-fat-ii, cfam-fat-ii, sey
when it boil it boil it boil it boil then
GEN COMPL ACT
added the meat, and it has (all) boiled furiously, then
10
min-'omt - a, sey lungowel sey ... d i l l - a . To min-'omt -ii,
we uncover then aroma escape when we uncover
SUBJ ACT SUBJ ACT GEN COMPL ACT
we uncover the pan, and the aroma escapes. When we have uncov-

shi ke nan, sey min-ngadd-a min-ngeer - oo-d\im bocfcfum,

well then we take we agitate it thoroughly
SUBJ ACT S U B J M I D
ered it, well,then we take it and agitate it thoroughly, and

m i n - "itt - a dumu ko min-nduf - a . Min-ngadd-a buruki

we remove bran which we throw we bring whisk
SUBJ ACT SUBJ ACT SUBJ ACT
we remove any bran, which we throw away. (Then) we take a whisk

min-ngacf-a sey min-'irt-a sey min-'irt-a sey min-'irt-a sey

we do then we stir then we stir then we stir then
SUBJ ACT SUBJ ACT SUBJ ACT SUBJ ACT
and we stir and we stir and we stir and
11
sey min-mburw-a, shi ke nan. To min-mburw-ii, tocfum-woocf-i
then we whisk that's that when we whisk if it good
SUBJ ACT GEN COMPL ACT STAT ACT
we whisk, there we are. When we have whisked (it), if it is good,

sey min-'ommb-a, sey cfum-soof-a. To cfum-soof-ii,

then we cover then it soak when it soak
SUBJ ACT SUBJ ACT GEN COMPL ACT
then we cover (it), then it should stand. When it has stood (for
12
sey min-ngadd-a kwaanooje, sey min-nokk - an-a koomoye fu.
then we take little bowls then we scoop for everyone all
SUBJ ACT SUBJ ACT
a w h i l e ) , then we take little bowls and scoop up some for every-
13
To nebbam 'e cfon, sey cfam-wacf-ee, cfum -'irt-a. To cfum-'irt-ii
if butter is then it put one stir when one stir
SUBJ PASS SUBJ ACT GEN COMPL ACT
one. If there is butter, then it should be added, and you stir.
 302

nebbam, shi ke nan. Nan-0 no min-ndef-ir-ta bete.

butter t h a t 1s it
hear how we cook bete
IMPER SG REL INCOMPL ACT
When you've stirred the butter, that's it. That's how we
c o o k ‘b e t e '.

Appendix B : Footnotes

1. When used as a copula, won-i is defective and cannot be easily

assigned to a particular conjugation. See footnote 8 , p. 116.
2. One might have expected this complex to have VS order rather than
SV order: the fact that it does not is perhaps a reflection of
the trend towards SV order that was noted in 54.4.1..

3. This -do suffix is a variant of the Anteriority-Marker. See

footnote 8 , page 116.
4. This word is borrowed from Hausa, and is often pronounced with
a low-high tone pattern, which is totally uncharacteristic of
a Fulfulde word. In Hausa, the term fatefate is also used. It
refers to a mushy soup of vegetables and flour, sometimes with
meat, as mentioned in this recipe:
5. It is difficult to give a consistent translation of sey, for with
the Subjunctive it may indicate either obligation or 'and t h e n 1.
6. The speaker has obviously forgotten to mention that water should
be added I
7. 'Inna li'o, literally 'mother of soup' is a favourite flavouring
in Northern Nigeria (as indeed are maggi cubes). 1Inna l i 1o is
m^de from the beans of the locust-bean tree. In Hausa it is
daddawa.

8. Borkonu is a generic term for 'pepper', and is borrowed from-

Hausa barkono.

9. The pronoun cfam is used without an antecedent. The choice of

pronoun is governed by the fact that the Dam class refers to
liquids, and at this stage ih the preparation of bete, it is
fairly liquid.
10. The use of sey both at the beginning of a clause and after the
subject is very frequent. It might perhaps be argued that the
subject is in fact topic in such circumstances, for it is effect­
ively separated from the predication that follows.
11. WoocT-, like certain other stative verbs, is sometimes used in the
Stative conjugation without either j_e_ or don.

12. This word is borrowed from Hausa kw a n o , with the same meaning.
13. The use of dhm as an impersonal subject meaning 'one' (French
'on', German 'man') is a caique from Hausa an. Its use in
Fulfulde is already eroding the use of the Passive.
 303

Bibliography

Ali, Saeed, and Yero Sylla (1977): 'Perceptual Transparency and

Relativisation: a Case-study in F u l a '. In Studies in
African Linguistics, Supplement 7.
Anderson, John M. (1971): The Grammar of C a s e : towards a Local-
istic T h e o r y . Cambridge University P r e s s .
Arnott, D.W. (1956): 'The Middle Voice in F u l a 1. In Bulletin of
the School of Oriental and African Studies, 28, 1.
Arnott, D.W. (1960)i The Tense System in Gombe F u l a . Ph.D. thesis,
University of London. Published in Afrika und Ubersee,
49 (1965).
Arnott, D.W. (1965): 'Sentence Intonation in the Gombe dialect of
Fu l a ' . African Language Studies, 6 .
Arnott, D.W. (1967): 'Some Reflections on the Content of Individ­
ual Classes in Fula and T i v 1. In G. Manessy (ed): La
Classification nominale dans les langues negro-afric-
aines. Paris: Colloques internationaux du Centre
National de la Recherche Scientifique.
Arnott, D.W. (1970): The Nominal and Verbal Systems of Fula.
Oxford: The Clarendon Press.
Bennett, Patrick, and Jan P. Sterk (1977): 'South-Central Niger-
Congo: a Reclassification1. In Studies in African
Linguistics, 8 , 3.
Carstairs, Andrew (1981): Constraints on Allomorphy in Inflexion.
Ph.D. thesis, University of London. Unpublished.
Chafe, Wallace L. (1976): 'Givenness, Contrastiveness, Definite­
ness, Subjects, Topics and Point of V i e w ' . In Li (ed)
1976.
Comrie, Bernard (1976): A sp e c t . Cambridge University Press.
Danes, F. (1972): 'Order of Elements and Sentence Intonation1.
In D. Bolinger (ed):Intonation. London: Penguin.
Diallo, Djibril (1979): Fuuta Jaloo and Fuuta Tooro: a Compara­
tive Study of two Pulaar Dialects. Ph.D. thesis,
University of London. Unpublished.
Dik, Simon, Sie Ing Djiang, Maria E. Hoffman, Jan R. de Jong,
Harry Stroomer, and Lourens de Vries (1980): 'On the
Typology of Focus Phenomena'. Glot, 3, 3-4. Univers­
ity of L e i d e n .

Faidherbe, Gal. L. (1075): Essai sur la langue poul, grammaire et

vocabulaire. Paris: Maisonneuve.
Fillmore, C.J. (1968): 'The Case for Case1. In E. Bach and
R. Harms (eds): Universals in Linguistics. New York:
Holt.
Fillmore, C.J. (1971): Some Problems for Case Grammar'. In
Monograph Series on Language and Linguistics, 24.
Washington D.C.: Georgetown University.
 304

Gaden, H. (1914); Le Poul ar. Tome II: Lexique Poular-Frangais.

Paris: Ernest Leroux. Reprinted Gregg, 1967.
Gaden, H. (1931): Proverbes et maximes peuls et toucouleurs. Paris:
Institut d'Ethnologie.
Givon, T. (1971): ‘Dependent Modals, Performatives, Factivity,
Bantu Subjunctives and what not'. In Studies in African
Linguistics, 2 , 1 .
Givon, T. (1973): ’Opacity and Reference in Language: an Inquiry
into the Role of Modalities'. In J. Kimball (ed):
Syntax and Semantics, vol. II.
Givon, T. (1976): 'Topic, Pronoun and Grammatical Agreement'. In
Li (ed) (1976).
Greenberg, Joseph (1966): 'Some Universals of Grammar, with partic­
ular Reference to the Order of Meaningful Elements'. In
J. Greenberg (ed): Universals of Language. Cambridge,
Mass.: MIT Press.
Greenberg, Joseph (1970): Languages of Africa. 3rd edition. The
H a g u e : Mout o n.
Grimes, Joseph E. (1975): The Thread of Discourse. The Hague:
M o uton.
Gruber, J.S. (1969): 'Topicalisation in child-language'. In
.D.A.Reibel and S.A.Shane (eds): Modern Studies in Eng­
l i s h . Englwood Cliffs, N.J.: Prentice-Hall.
Gundel, Jeanette K. (1977): Role of Topic and Comment in Linguistic
T h eory. Indiana University Linguistics Club.
Gunn, H.D. (1956): Pagan Peoples of the Central Area of Northern
Nigeria. London: International African Institute Ethno­
graphic Survey.
Hansford, Keir, John Bendor-Samuel and Ronald Stanford: rA Pro­
visional Language Map of Nigeria'. Savanna, 5, 2^
Harris, M.B. (1980): 'Noun Phrases and Verb Phrases in Romance',
Transactions of the Philological Society, 1980.
Hockett, C.F. (1954): 'Two Models of Grammatical Description'.
W o r d , 10.
Jaggar, Philip (1978): ’’And what about ...?' - Topicalisation in
H a u s a '. Studies in African Linguistics, 9, 1.
Kiparsky, P., and C. Kiparsky (1970); 'Fact'. In M.Bierwisch and
K.E.Heidolph: Progress in Linguistics. The Hague:
Mout o n .
Klingenheben, A. (1963): Die Sprache der F u l . Hamburg: J.-J.
Augustin.
Kraft, C., and A.H.M.Kirk-Greene (1973): Teach Yourself Hausa.
English Universities Press.
Labatut, R. (1973): Le Parler d'un groupe de Peuls nomades. Paris:
SELAF.
Lakoff, G. (1970): Irregularity in Syntax. Holt, Rinehart and
Winston.
 305

Li, Charles (ed) (1976): Subject and Topic. New York: Academic
Press.
Li, Charles, and Sandra Thompson (1976): 'Subject and Topic: a
new Typology of Language'. In Li (ed) (1976) .
Lyons, J. (1968): An Introduction to Theoretical Linguistics.
Cambridge University Press.
Lyons, J. (1977): Semantics. Cambrdige University Press.
Matthews, P.H. (1972): Inflectional Morphology. Cambridge Univers­
ity P r e s s .
Matthews, P.H. (1974): Morphology: an Introduction to the Theory
of Word-structure. Cambrdige University Press.
McIntosh, Mary (1980): 'Locative-based Periphrastic Verb-forms in
Fulfulde'. Unpublished Seminar Paper, School of Oriental
and African Studies, London.
McIntosh, Mary (1981): 'Fulfulde as a VSO Language1. Unpublished
Seminar Paper, School of Oriental and African Studies,
L o ndon.
Newman, Paul (1972): 1Syllable.Weight as a Phonological Variable'.
In Studies in African Linguistics, 3, 3.
Noye, Dominique (1974): Cours de Foufoulde. Maroua: Mission
Catholique, and Paris: Geuthner.
Palmer, F.R. (1971): Grammar. Penguin.
Pinkham, Jessie, and Jorge Hankamer (1975): 'Deep and Shallow
Clefts'. In Papers from the 11th Regional Meeting,
Chicago Linguistic Society.
Reichardt, C.A.L. (1876): Grammar of the Fulde Language. London:
Church Missionary Society.
Robins, R.H. (1964): General Linguistics: an Introductory Survey.
Longman.
Sapir, J. David (1971): 'West Atlantic: an Inventory of the Lang­
uages, their Noun Class Systems, and Consonant Alterna­
t i o n 1 . In Thomas A.Sebeok (ed): Current Trends in
Linguistics, vol. 7. The Hague: Mouton.
Schachter, Paul (1973): 'Focus and Relativization'. In Language,
49, 1.
Seydou, Christiane (ed) (1979): Dictionnaire General de P e u l :
Methodoloqie, Inventaire des Source existantes. Paris:
Comite scientifique international, C.N.R.S..
Skinner, Neil, and Corinne Pelletier (1979): Adamawa Fulfulde: an
Introductory Course. African Studies Program, University
of Wisconsin.
Smith, M.G. (1969): Government in Zazzau. Oxford University Press,
for International African Institute.
 306

Taylor, F.W- (1921): A First Grammar of the Adamawa Dialect of the

Fulani Language (Fulfulde). Oxford.
Taylor, F.W. (1932): A Fulani Dictionary. Oxford.
Temple, O. (1922): Notes on the Tribes, Provinces, Emirates and
States of the Northern Provinces of Nigeria. Reprinted
by Frank Cass, 1965.
UNESCO (1966): Rapport final sur la reunion d'un groupe d*experts
pour 1 1unification des alphabets des langes nationales.
UNESCO (1971): Dictionnaire elementaire Fulfulde-Frangais-English.
(Prepared by D.W.Arnott, P.-F.Lacroix, E.Mohammadou and
A.I.Sow.) Niamey: Centre regional pour la Documentation
de la Tradition Orale.
Waziri Junaidu, Alhaji (undated): Unpublished manuscript of a
Fulfulde Grammar in Fulfulde. Sokoto.
Westermann, D. (1909): Handbuch der Fulsprache: Worterbuch,
Grammatik, Ubungen und Texte. Berlin: Reimer.