Africana Linguistica

Verbal system and diathesis derivations in Seereer

Souleymane Faye, Maarten Mous

Citer ce document / Cite this document :

Faye Souleymane, Mous Maarten. Verbal system and diathesis derivations in Seereer. In: Africana Linguistica 12, 2006.
pp. 89-112;

doi : https://doi.org/10.3406/aflin.2006.963

https://www.persee.fr/doc/aflin_2033-8732_2006_num_12_1_963

Fichier pdf généré le 05/04/2018

 Verbal system and diathesis derivations in Seereer
Souleymane FAYE* & Maarten MOUS**

1. Introduction1

Seereer, an Atlantic language spoken in Senegal, has a rich system of verbal

derivations comprising about 23 different suffixes and combinations thereof (Faye
1982). The verbal derivations convey meanings such as causative, reflexive,
reciprocal, applicative, reversive, direction, repetition and attitude. Several verbal
derivations operate on verbal diathesis: there are various causatives, a reflexive,
various reciprocals, an applicative and a beneficiary, and a derivation that
indicates that the possession of a subject is affected. Diathesis in the verbal system
operates more at the semantic argument structure than at the syntactic argument
structure. The anti-causative is also expressed on the verb but at a position where
the inflectional markers are found. § 2 presents the basic structure of the verb; § 3
discusses the various derivations that operate on diathesis, the final section
concludes.

2. Verb morphology

The verbal inflection is complex. Mäkelä (1989:190-216) presents over 200

paradigms. The following categories are expressed in the verbal conjugation: (i)

*
Centre de Linguistique Appliquée Dakar, Université Cheikh Anta Diop.
**
Leiden University & Research Centre for Linguistic Typology, La Trobe University.
1. We would like to thank Maarten Kossmann, Sasha Aikhenvald and an anonymous refe-
ree for their very useful comments. We use the Seereer orthography in which Y stands for
the voiced implosive palatal and _n for the palatal nasal. The following abbreviations are
used: ANTICAUS anti-causative, APPL applicative, CAUS causative, CL (nominal) class, DEF
definite pronoun, DEM demonstrative, FOC focus, IMPFV imperfective, INSTR instrument, NEG
negation, OBJ object, OBLIG obligation PCLE particle, PART participle, PF perfective, PL plu-
ral, PREP prepositional, PROX proximal, QUES questioning particle, RECIP reciprocal, sb
somebody, SG singular, sth something, SUBJ subject, SUBORD subordinate, SUF suffix, THITH
thither, VBLZ verbalizer.
90 AFRICANA LINGUISTICA 12 (2006)

subordination, (ii) subject, (iii) number of subject, (iv) object, (v) voice, (vi) tense-
aspect, (vii) polarity, (viii) mood and (ix) focus.
The verb can have a number of suffixes, the order of which is quite rigid. The
slots named TAM and ASPECT in the column TAM can be combined. There are
additional restrictions on co-occurrence. The slots following the verb stem are
suffixes; those preceding it are independent words or pro-clitics.

PCLE SUBJ VERB DER TAM NEG ASPECT VOICE OBJ SUBJ SUBD NBER
kaa oox ’ ir(a) a e(l) am um na iyo
xan ir oog eer aa ong o
ndax in ang ee kaa in
fat and at aand
etc. etc. aful

Table 1: Structure of the verb

The following slots and their possible fillers are involved:

1. Particles: kaa verb focus, xan future, ndax questioning, fat obligation, ba
prohibitive, bo ‘until’, ndax ‘in order to’.
2. Subject: The post-particle and preverbal subject is either a full subject noun,
a strong pronoun for subject focus (in some cases followed by the particle na
and in case of the progressive/durative followed by the proximal or distant
definite marker of human classes 1 and 2 for singular and plural,
respectively), a weak pronoun which is in some paradigms realised as a
suffix in case of 1SG or 2 SG only. The semantic basis of the ‘tenses’ of the
verb involves the choice of subject marking as well as the choice of the
morphemes in the columns PCLE , TAM , ASPECT and NEG . The third person
weak pronoun has three variants, i.e. zero, a or ta/da SG/PL, according to
the conjugation.
3. Verb: The verb stem which includes derivational suffixes, some of which are
very similar in function to some of the inflectional suffixes. The initial
consonant of the verb changes from a continuant to a prenasalized stop if the
subject is plural. Initial consonants that are voiceless stops do not change,
nor do the consonants that do not participate in consonant mutation. See
McLaughlin (2000).
4. Derivation: There are many different derivations. Of the ones that we discuss
here, most appear in this slot: middle -oox, reciprocal -ir, subject indirectly
affected -iD , direct causative -in , indirect causative -noor, causative of
state -and , causative of body state -il, applicative -it , reversive -it
(reducing preceding syllable), benefactive -an. There are more derivations in
this slot that are not mentioned here. There are also derivations that appear in
other slots, such as the anti-causative in the voice slot. The markers aful
and at in slot 5 TAM have also been regarded as derivations.
S. FAYE & M. MOUS – Verbal system and diathesis derivation in Seereer 91

5. TAM :
A glottal stop for past reference; ang for conditional clauses; at for
‘no more’ and aful for ‘not yet’, oog for the imperfective past (which
intervenes with the separate durative and past marking options) and koog for
the fulfillment of an irreal condition (cf. Faye 1982 and Mäkelä 1989).2
6. Negation: -ir, -eer; and for the negative anti-causative -aand, -ee.
7. Aspect: a for perfective, aa for imperfective; k-aa for prospective.
8. Voice: anti-causative -e(l).
9. Object: -(x)am 1SG.O, -(x)ong -2SG.O, -in 3SG.O.
10. Subject: -um 1SG, -o 2SG.
11. Subordination: na.
12. Number: -iyo optional expression of plurality of subject.

There is a main distinction between perfective and imperfective paradigms. The

differences between the two are first of all in the length of the aspectual suffix -a
for perfective and -aa for imperfective. But there are also major differences in the
encoding of the subject. In the imperfective the subject is marked by a pronoun
before the verb stem, while in the perfective the subject tends to be marked by a
suffix for the persons for which it is possible (first and second singular). A group
of conjugational paradigms such as the perfective consists of the basic paradigm,
three different focus paradigms and the subordinate paradigm. The subject-focus
paradigms consist of a left dislocated subject or strong subject pronoun and a verb
form. The non-subject-focus paradigms consist of a left dislocated non-subject
phrase and a verb form. The verb-focus paradigms consist of an initial particle
kaa followed by a weak pronoun and a verb form. The perfective paradigms are
presented in table 2, the imperfective paradigms in table 3. Both tables are based
on the verb jik, ‘to buy’, and modelled after Mäkelä (1989:190-216).3

BASIC NON-SUBJ FOC SUBJ FOCUS VERB FOCUS SUBORD

1sg jikaam X jikum mi jiku kaam jik jikuuma
2sg jikaa X jiko wo jiku koo jik jikoona
3sg a jika X a jiku ten jiku kaa (ta) jik ta jikna
1pl i njika X i njiku in njiku kaa i jik i njikna
2pl nu njika X nu njiku nuun njiku kaa nu jik nu njikna
3pl a njika X a njiku den njiku kaa (da) jik da njikna

Table 2: Perfective paradigms

2. -koog is probably a frozen combination of -k thither and -oog imperfective past. The
imperfective past marker -oog is the only marker in this slot with aspectual value; there is
another slot, named ASPECT, which contains only aspectual markers.
3. In W.Faye (1980) there is mention of a simpler conjugated verb form consisting of the
verb stem followed by a subject suffix which is used for the narrative.
92 AFRICANA LINGUISTICA 12 (2006)

BASIC NON-SUBJ FOCUS SUBJ FOCUS

1sg mexe jikaa X um jikaa mi na jikaa
2sg woxe jikaa X o jikaa wo na jikaa
3sg oxe jikaa X a jikaa ten na jikaa
1pl in we njikaa X i njikaa in na njikaa
2pl nuun we njikaa X nu njikaa nuun njikaa
3pl owe njikaa X a njikaa den njikaa

VERB FOCUS SUBORD

1sg kaam jikaa um jikaa
2sg koo jikaa o jikaa
3sg kaa (ta) jikaa ta jikaa
1pl kaa i njikaa i njikaa
2pl kaa nu njikaa nu njikaa
3pl kaa (da) njikaa da njikaa

Table 3: Imperfective paradigms

The imperfective paradigms have subject marking before the verb and in the case
of the basic (non-focus) paradigm the subject is a definite pronoun. The definite
pronouns are members of the noun classes for humans; for the third person, they
consist of the nominal prefixes plus concord (xe/we ); for the first and second
person, they consist of the emphatic pronouns plus concord of the human classes
o-oxe and we.
In the perfective, but not in the imperfective paradigms, the verb form has the
vowel u instead of the aspectual vowel a when a subject or another constituent is
left dislocated for focus. The verb-focus paradigms have an aspectual vowel in the
imperfective paradigms which is however absent in the perfective paradigms.
The perfective subordinate paradigm, which is used for relative clauses,
consists of the verb form identical to that of a non-subject-focus construction,
followed by na. The imperfective subordinate paradigm has the aspectual suffix
aa but no suffix na and a weak subject pronoun before the verb.
The differences and the schemes for the perfective and imperfective
conjugation types are presented in table 4. For each of the cells in the table, we
can in principle have the different forms of the TAM column, the anti-causative
forms, the various subject and object combinations and negative forms.
The prospective (or future) paradigms are based on the imperfective paradigms
with the addition of k before the aspectual suffix -aa. In the basic, non-focussed
forms the prospective has an initial particle xan followed by the weak subject
pronoun and the verb stem.
S. FAYE & M. MOUS – Verbal system and diathesis derivation in Seereer 93

PERFECTIVE IMPERFECTIVE
basic V-a-SUBJ definite SUBJ pro V-aa
non-subject focus OBJ V-u-SUBJ Obj SUBJ V-aa
subject focus emphatic SUBJ pro V-u emphatic SUBJ pro na V-aa
verb focus kaa SUBJ V kaa SUBJ V-aa
subord V-u-SUBJ-na SUBJ V-aa

Table 4: The basic conjugations

In table 4, the subject is represented as -SUBJ for the subject suffixes, but for
subjects other than 1SG or 2 SG it stands for weak subject preverbal pronouns;
when in front of the verb SUBJ stands for weak pronouns. Subjects are represented
by emphatic pronouns for subject focus forms and by definite subject pronouns in
the basic imperfective forms. The subject forms are presented in table 5. The third
person weak pronoun is either zero, a or ta/da sg/pl, depending on the
conjugational form: zero is used for the basic negative and verb focus negative
forms; ta/da are used for subordinate forms and verb focus forms; a is used for
the rest. Plurality of subject is indicated by making the root-initial consonant
prenasalized (consonant mutation). This does not apply to voiceless stops nor to
consonants that do not undergo consonant mutation (Faye 1985); and plurality of
subject is optionally indicated by a suffix -iyo for 3pl subjects, which can
disambiguate certain verb forms, e.g. _naamee-yo, ‘they haven’t eaten’, compared
to _naamee, ‘s/he hasn’t eaten’.

EMPHATIC WEAK SUBJ SUBJ SUF DEF OBJ SUF

1sg mi um -um me(xe) -aam, -xam
2sg wo o -o wo(xe) -ong, -xong
3sg ten ta, a o(xe) -in
1pl in i in( we)
2pl nuun nu nuun(we)
3pl den da, a o(we)

Table 5 : Personal pronouns and person marking suffixes

The focus forms of the verb are related to the fact that the focussed constituent is
left dislocated; there are other strategies for focus, such as the addition of the
particle de after the focussed element, and in that case the verb form is the basic
conjugation and not the focus form; for example, kene de layeem, ‘that has been
told to me’, with the initial demonstrative pronoun kene focussed by the particle
de and the verb form in the basic conjugation of the perfective anti-causative.
Focus forms of the verb are used both for contrastive focus and for selective focus.
Questions use focus forms of the verb. The vowel u that is added to the verb in
focus forms (in the perfective) renders the verb form identical to that of a
participle. This is not the case for the non-perfective focus forms of the verb.
A participle can be formed from any verb by the addition of -u for affirmative
active, -e for affirmative anti-causative, -eer for negative active and -aand for
94 AFRICANA LINGUISTICA 12 (2006)

negative anti-causative (Mäkelä 1989:88-93). The resulting form agrees in noun

class (noun class prefix) and consonant mutation to the head noun in the same way
as the other adjectives do. The construction can be followed by a definite pronoun.
Faye (1979:105) refers to these participles as verbes qualificatifs, e.g. a jik-a
pay Pal-g-u 3 buy-PF cloth black-VBLZ-PART, ‘he bought a black cloth’, or o-
kiin o-qon-u CL-man CL-die-PART, ‘a dead man’. The examples in (1) show the
adjectival behaviour of these verb forms.4
(1) o-Box o-soxod-u ‘a wicked dog’ < soxod ‘bad’
CL-dog CL-bad-PART
o-Box o-soxod-eer ‘a dog which is not wicked’ < soxod
CL-dog CL-bad-NEG.PART
o-koor o-qaw-u ‘a hitting person’ < xaw ‘hit’
CL-person CL-hit-PART
o-koor o-qaw-e ‘person (who is) hit’ < xaw
CL-person CL-hit-ANTICAUS.PART
faap fa-mbaax-u fee ‘the good father’ < paax ‘good’
father CL-good-PART DEF
o-koor o-paax-u ‘a good man’ < paax
CL-person CL-good-PART
o-koor o-paax-eer oxe ‘the evil man’ < paax
CL-person CL-good-NEG.PART DEF
Sitoor maaf-aand ‘Sitoor, (who is) not knocked down’ < maaf ‘let fall’
Sitoor let.fall-ANTICAUS.NEG.PART

The participles allow addition of the inflection that is close to the root in the TAM
column in table 1, but not the inflection of the ASPECT column. In order to express
future in the example in (2) we have to make a relative verb form of the verb (2h).
The participle form can take full noun direct and indirect objects (2d, e), but not
an object suffix (2f); it needs a real relative clause (2g). Note that the class prefix
for pay, ‘cloth’, in (2) is zero.

(2) a. pay Paal ‘a black cloth’

cloth black
b. pay Pal-g-u ‘a cloth, a black one’
cloth black-VBLZ-PART
c. pay Pal-ig-’-u ‘a cloth that was black’
cloth black-VBLZ-PAST-PART
d. o-koor o-qaw-u o-ndeP^andoong ‘a child hitting man’
CL-person CL-hit-PART CL-child
e. o-koor o-ci’-e o-montar ‘a man given a watch to’
CL-person CL-give-ANTICAUS.PART CL-watch
f. *o-koor o-qaw-aam-u ‘a man hitting me’
CL-person CL-hit-O1SG-PART

4. The last two examples are from W.Faye (1980:51), the other from Mäkelä (1989:88).
S. FAYE & M. MOUS – Verbal system and diathesis derivation in Seereer 95

g. o-koor o-xaa xaw-axam-a ‘a man who is hitting me’

CL-person CL-DEF hit-O1SG-IMPFV
h. pay Pal-k-oog-u ‘a cloth which was becoming black’
cloth black-THITH-PAST.IMPFV-PART

The perfective verb-focus forms seem to be based on the participle, but the focus
forms do allow for subject and object marking and even aspect marking; anti-
causative verb forms have imperfective forms with a long vowel which is
presumably the imperfective suffix -aa from the ASPECT slot in the verb form.
Specifically, the subject-focus forms seem to be based on participles which follow
the emphatic (focussed) subject pronoun and the non-subject focus forms likewise
seem to be based on participles. These focus forms must have arisen through left
dislocation plus a participle, but now allow for inflectional expression. This origin
also explains the similarity in form between the subordinate (relative) verb form
and the focus verb form for the perfective.

3. The diathesis derivations

There are many different verbal derivations. Here we only discuss those that are
directly linked to the diathesis of the verb. Apart from the anti-causative formation
there is a reflexive or middle derivation, a reciprocal derivation, several
causatives, two kinds of applicatives and a derivation to indicate that the subject is
affected in his/her possession. These are derivations with varying degrees of
productivity. They can be called derivations because they are positioned before the
inflectional suffixes that are discussed in § 2 (except for the anti-causative); they
have some instances with specialized meaning (except for the anti-causative);
there are derived verbs for which there is no base, and there are verbs that do not
take the derivation in question (except for the anti-causative). The derivations
discussed below all affect the semantic argument structure of the verb but it is less
clear whether the number of syntactic argument positions is manipulated by any of
these morphological processes. There is little known about the order and possible
combinations of the various derivational suffixes. Some combinations that have
been attested are mentioned below. Also, we have not yet studied the productivity
of these derivations.

3.1. The anti-causative (formally termed ‘passive’)

The main form of the ‘passive’ or anti-causative is -el. The anti-causative takes
different shapes in the course of inflection: the ending for the first person anti-
causative is -eem ; the negative anti-causative is -aand , the perfective anti-
causative is -e ; the imperfective anti-causative is -eel . The anti-causative
derivation removes the subject from the sentence and the object becomes the
subject. When there are several objects either of them may become the subject.
The derived verb does not allow the expression of an agent, hence it is
anti-causative rather than passive. A sentence like a fad-el, ‘he had been hit’,
allows the addition of all sorts of circumstantial adjuncts: of time, place and
96 AFRICANA LINGUISTICA 12 (2006)

manner, but not of a phrase containing an agent. In fact, the derivation does not
always even suppose the presence of an agent, for example, in an experience verb
like xeexel, ‘to be hungry’, which is a lexicalized anti-causative. The derivation
is an anti-causative, taking away the semantic presence of the causer or agent of
the verb. There is no impersonal subject construction that translates as a passive in
Seereer.
(3) a. i mbind-a xa-letaar a Njaay
1PL PL:write-PF CL-letters to Njaay
‘We wrote letters to Njaay’
b. xa-leetar a mbind-e a Njaay
CL-letter 3 PL:write-ANTICAUS:PF to Njaay
‘Letters were written to Njaay’
c. _njaay a bind-e xa-leetar
Njaay 3 write-ANTICAUS:PF CL-letters
‘Njaay was written letters to’

The anti-causative deviates from the other verbal derivations in its position: it
follows the past tense marker, which is a glottal stop. On the basis of its position,
the fact that the anti-causative is the only derivation that is expressed in the
participles, the fact that there are no lexicalized anti-causatives with specialized
meaning, and that every verb can form an anti-causative, one could argue that the
anti-causative is inflectional rather than derivational. On the other hand, there are
lexicalized anti-causatives such as xeexel, ‘be hungry’,5 and nominalizations of
anti-causatives are possible and productive, which suggest again that the
anticausative is derivational. Anti-causative verbs in -el can be nominalized by
placement into class o-ole with their plural in a-ake ; some nouns denoting
masses are in class ke (W. Faye 1980:49). The nominalization is in either of the
classes, that is, a nominalization *o-_naam-el ole is not possible, next to existing
_naam-ke, ‘food’. The nouns represent abstract concepts; some examples are given
in (4). For all the forms an anti-causative verb form with el does exist.6
(4) mos ‘be nice’ o-mos-el ole ‘beauty’
faax ‘be good’ o-paax-el ole ‘goodness’
neer ‘be fat’ o-neer-el ole ‘plumpness’
bon ‘be evil’ o bon-el ole ‘evilness’

_naam ‘eat’ _naam-el ke ‘food’

yer ‘drink’ yer-el ke ‘drink’
jal ‘work’ cal-el ke ‘works’
fiD ‘sift’ mbiD-el ke ‘flour’

5. There is no verb root xeex , but there is a noun nqeex , ‘hunger’. Derivation like
xee-xoox is not possible; xeexin is , ‘cause somebody to be hungry’.
6. The first three examples are from Fal (1980), who has a long vowel in the suffix -eel
which should be short in these nominalizations; the next two are from Faye (1980:49) and
the last two from Faye (1985:18).
S. FAYE & M. MOUS – Verbal system and diathesis derivation in Seereer 97

Experience verbs such as suuY-el, ‘be greedy’, are anti-causative; its base, suuY,
‘to greed’, has as subject the meal. Interestingly, the anti-causative can be applied
to intransitive subjectless verbs such as time and weather verbs see (5) from Faye
(1982:4-5). Thus one can say deB-eem rain-ANTICAUS:1SG, ‘I have been rained
upon’, from the intransitive verb deB, ‘to rain’ (if an explicit subject is used with
this verb, it is the word for ‘God’); and xan i yeng-el FUT 1 PL night-
ANTICAUS, ‘night will fall upon us’; in order to say, ‘if you are not too late’,
o!yeng-ang-aand 2SG night-COND-ANTICAUS:NEG, one uses an agentless anti-
causative.
(5) _naal ‘be late in morning’ n_ aal-el ‘do sth late in the morning’
yaar ‘be late in the evening’ yaar-el ‘do sth late in the evening’
o-yeng ole ‘the night’ yeng-el ‘do sth late at night’

The anti-causative is used to express potentiality, reef-el, ‘be passable’, from

reef, ‘to pass’, in (6).
(6) a-wag a reef-el
CL-road 3 pass-ANTICAUS
‘it (the road) is passable, can be passed’

Anti-causative verbs are not necessarily intransitive, because non-subject-focus

forms of a anti-causative verb do exist for verbs that allow two objects; see
Mäkelä (1989). The dative argument is expressed as subject of the anti-causative
verb, as in (7) and (11), but the additional direct object can be clefted as focussed
complement, as in (8) and (9). The (dative) subject can also be focussed (10). The
full direct object is expressed after the general purpose preposition o that is used
for any non-human non-core constituent (11). It is, however, impossible to have
an object suffix in the anti-causative verb form, *lay-un-eem but lay-eem
kene, ‘I have it been told’ (7).
(7) lay-eem o-ten
tell-ANTICAUS:1SG CL-that
I have been told that.

(8) ndingil lay-eem

truth say-ANTICAUS:1SG
It is the truth that I have been told.

(9) ke-ne lay-eem

CL-DEM.PROX say-ANTICAUS:1SG
That is what I have been told.

(10) ke-ne um lay-eel

CL-DEM.PROX 1SG say-ANTICAUS:IMPFV
It is me who is being told.

(11) ci’-eem o montar

give-ANTICAUS:1SG PREP watch
I have been given a watch.
98 AFRICANA LINGUISTICA 12 (2006)

3.2. The reflexive or middle derivation

The reflexive or middle derivation in Seereer has two forms, -u and -oox, which
are interchangeable (Faye 1982:10). Some examples are presented in table 6.
Bog ‘wash the body’ Bogu ‘wash oneself’
Deet ‘look’ Deetu ‘look at oneself’
laxad ‘wash an object’ laxadoox ‘wash one’s hands’
geen-in7 ‘stop sth’ geenu ‘stand’
hirand8 ‘leave in the evening’ hirandu ‘go to work in the afternoon’
nomit ‘cause to go back, nomtu ‘turn around, go back’
to turn around’
supit ‘change, translate’ suptu ‘transform oneself’
wet ‘open sth’ wetoox ‘open’
wetand ‘remind’ wetandu9 ‘tell’
yip ‘pour, put’ yipu ‘flow over’
sip ‘plant’ sipu ‘land on earth’
dol-n-ir10 ‘marry’ dolu ‘enter marital stage
(for woman) ’
waadin11 ‘nurse, tend’ waadnu ‘take care of oneself’
fudin ‘let go down’ fudnoox ‘to go down’
yaq12 ‘spoil’ yaqu ‘be spoilt’
wa_n ‘to diminish’ wa_nu ‘be diminished’
wa_nin ‘get sth reduced’ wa_nnu ‘get something diminished’

Table 6: Table of derived middle verbs

The reflexive or middle derivation indicates that the action is subject internal and
this includes reflexives but not only reflexives. The term reflexive is inspired by
the model of French that has a reflexive pronoun se that covers the wide area of
middle meaning. The same is true for the Seereer middle.
The middle derivation can be used to derive verbs from nouns. The resultant
verbs are middle in meaning; see table 7.13 It is not easy to say that these are
derived from nouns or from verbs because noun roots are easily used as verbs.
Verbs like faDoox, ‘behave as a slave’, express that the subject puts him/herself
into a position of a slave, thus the expression faDoox a Roog, ‘worship God’,
literally means ‘put oneself into a position of a slave of God’.

7. There is no simple verb root *geen; there is, however, a noun consisting of this root,
a-keen, ‘stand still’.
8. There is no simple root *hir.
9. Compare: wetandoor, ‘remember’.
10. A man would use jeg, ‘have a wife’; dolin, ‘put a woman into marital stage’
11. There is no simple root waad, only as a noun, o-baad oxe, ‘the sick person’.
12. yaqin, ‘make happy’.
13. The first two examples are from W. Faye (1980:49), the next two from Mäkelä (1989).
S. FAYE & M. MOUS – Verbal system and diathesis derivation in Seereer 99

o tag ‘courtier, courtesan’ dag-oox ‘behave as a courtesan’

o paD ‘slave’ faD-oox ‘behave as a slave’
foto ne ‘picture, take a picture’ foto’u ‘take a picture of one self’
portole ‘picture, take a picture’ portole’u ‘take a picture of one self’
mukit ‘face, wash a face’ muktu ‘wash one’s face’
o-peengu14 ‘needle, inject’ peengu’oox ‘get oneself injected’
wecetMwecit (n/v) ‘money, change’ (n/v) wecetu ‘get change (money)’

Table 7: Table of denominal middle verbs

There are many verbs with the middle derivation for which there is no base root;
see table 8. We assume that verbs that are polysyllabic and end in u or oox are
middle derived.
*hir hiroox ‘greet in evening’
*fud fudu ‘go down’
*kar karu15 ‘grind in a specific way in which the pestle hits against the inner
sides of the mortar’
*geen geenoox ‘stand’
*woond woondoox ‘lie down’
*ju’ ju’oox ‘have dinner’
*fut futu ‘have lunch’
*dak dakoox ‘return’
*_naaY _naaYloox ‘go for a walk’
*topat topatu16 ‘deal with’

Table 8 : Table of frozen middles

The suffix can be preceded by other derivational suffixes such as the causatives
-in (12a), -and (12b) and the reversive/separative -it (12c), and it can be
followed by a number of other suffixes, specifically, the directional suffixes -ik,
‘thither’, and -iid, ‘hither’ (Faye 1982:24-26). The verb wirlu, ‘roam around,
wander’, from wiriil, ‘let turn’, has no base *wir, but the meaning of the verb
suggests the presence of the suffix -lu , ‘do something half-heartedly’, as in
_ n aaYloox , ‘go for a walk’, derived from _ n aax , ‘walk’. In combination with
reduplication this suffix means ‘to pretend to do X’, e.g. Daan-Daan-loox ,
‘pretend to sleep’ (Faye 1982). The suffix -lu is a grammaticalized combination

14. The verb is derived from the same noun o-peengu ole, ‘the needle’, which is bor-
rowed from French épingle. The middle derivation (which is uncommon) is also realised as
peengoox M peengu.
15. Not related to kar, ‘hunt’.
16. The verb is borrowed from Wolof toppatu and the final vowel is reinterpreted as the
middle suffix, since the variant toputoox exists.
100 AFRICANA LINGUISTICA 12 (2006)

of the causative of body state and the middle -u; the meaning comes close to the
‘behave as a slave’ type of meanings of the de-nominal middle derived verbs.17
18
(12) a. gooknu ‘to kneel’ cf. gookin, ‘to make someone kneel’
jaBnu ‘to take refuge, hide oneself’
waadnu ‘to take care of oneself’
wa_nnu ‘get something (e.g. price) diminished’
b. hirandu ‘go to work in the afternoon’ from hirand ‘leave in the evening’
wetandu ‘tell, to remember’
c. nomtu ‘go back’ from nomit, ‘return’
suptu ‘transform oneself’ from supit ‘change, translate’
pictu ‘to disperse’
moytu ‘to avoid’
Baltu ‘get broken’

Nominalization of middle verbs replaces the suffix with ax. This process is fully
productive.19
(13) Examples of nominalizations of middle verbs
saytoox ‘talk’ a saytax ale ‘the conversation’
sumb ‘have a bath’ a sumbax ‘bath’
xaYloox ‘hunt’ a qaClax ‘hunt’
wetandox ‘remember’ a betandax ‘souvenir’
huloox ‘dress’ a kulax ‘dressing’
naDoox ‘be steep’ a naDax ‘inclination’
xet ‘meet’ a qetax ‘meeting’

Like other languages with a derivational middle, the suffix -u covers the semantic
area of the middle that has the body as focus (Mous 2003, Mous & Qorro 2000).
In order to make an inventory of the meaning of the middle derivation we look at
the various semantic fields that are typically middle and that come from
Kemmer’s (1993) typological study of middles, adapted by Mous (2003) for
derivational middles in African languages. Considering derived, frozen and de-
nominal middle verbs, we can see that the following categories recognised by
Kemmer are represented by the middle in Seereer: body activity, body position,
some cognition verbs, and emotional middles, not so many spontaneous events
and no natural reciprocals. Middle derivation is not an obligatory operation. There
are many verbs that belong semantically in the categories above but that are not
middle derived.
Body activity: Several expressions of body activity require a middle
derivation; see (14). The verb ‘to itch’, _ n axa_ n, has the skin as subject and the
middle derived _naxa_noox is directed at the body and means ‘to scratch’. The verb

17. There is another verb that ends in lu, xawlu, ‘hunt’, but it is not divisable; it is not
related to xaw, ‘to hit’.
18. We give the form on which the middle derivation is based whenever there is such a
form.
19. Examples in (13) are from Crétois (1973:99).
S. FAYE & M. MOUS – Verbal system and diathesis derivation in Seereer 101

‘to blow one’s nose’, _ n iitu , is middle derived from the more marked active
object directed verb. The underived _noot, ‘breath, rest, persevere’, does not allow
a middle derived *_ n ootoox, for it needs an additional causative, -in, yielding
_ n oonoox , ‘take a rest’; the causative _noot-in means ‘to give somebody the
condition to take a rest, and prevent from being tired’, and the middle directs that
to one’s self. The verb haBaas, ‘to yawn’, has no middle derivation.
(14) Body activity middles
oqotu ‘to cough’ *oqot, oqotax (n)
disoox ‘to hiccup’ *dis, a tisax
xoPtoox ‘to loose (of skin)’
_naxa_noox ‘to scratch’ a _naxa_nax
gaajoox ‘to be in labour (for woman)’ cf. gaajin, ‘help deliver (for midwife)’
_niitu ‘to blow one’s nose’ from _niit, ‘to blow (a child’s) nose’

Body positions: Verbs that primarily denote a certain body position are often
middle derived (15). The equivalent without the suffix often has causative
meaning (whether causative derived or not) and seems to be more derived from a
semantic point of view, but not morphologically: geCnoox, ‘to lean’, from geCin,
‘to make somebody lean’, gooknoox, ‘kneel’, from gookin, ‘to make someone
kneel’; kogomoox, ‘to sit with knees pulled up’, and kongom is the causative
equivalent; rumoox , ‘to have the head resting on the hand (associated with
anxiety)’. Some body positions are not related to the middle derivation: sakal, ‘to
squat’, moof , ‘to sit’, xap , ‘to sit with the whole lower part of the body on the
ground’.
(15) Body position middles
geenoox ‘be upright, stop’ *geen
wondoox ‘to lie down, to sleep’ *wond
20
daxanoox ‘to lie on one’s back’
Daasoox ‘to lie with stomach down’
geCnoox ‘to lean’ *geC
gooknoox ‘to kneel’
kongomoox ‘to sit with knees pulled up’
jokandoox ‘to lean on one’s elbows’ *jokand21
rugmoox ‘to have the head resting on the hand’ *rugum

A distinct set of verbs that is middle derived consists of the words for having the
various meals in (16).
(16) Meal middles
_naktoox ‘to have breakfast’
fuutoox ‘to have lunch’
boCtoox ‘to have afternoon tea’
ju’oox ‘to have dinner’

20. Also wondoox ndaaxaan.

21. Ultimately derived from jok, ‘join something to lengthen it’.
102 AFRICANA LINGUISTICA 12 (2006)

Some mental activity verbs are middle derived: maToox, ‘advise, warn’, and this
verb is transitive, maTooxaanum, ‘I have warned him’ and the body oriented verb
muuYoox , ‘to smile’. However, yiif , ‘think’, weec , ‘forget’ and fuux , ‘be
angry’, are not middle derived.
Spontaneous events such as ‘to sprout’, fee_n, ‘to grow (of plants)’, sax, ‘to
ripen’, foor , are not associated with a middle derivation; only sutoox , ‘come
out’, is. A verb like suptoox, ‘transform oneself’, is middle derived from supit,
‘to change’, the reverse of suup , ‘to immerse’. This verb suptoox does not
denote a spontaneous event. The verb suqtoox , ‘to open, unlock, clear’
(undergoer is subject), from sux, ‘to be constructed’, with the reversive -it and
the middle -oox, is closer to a spontaneous event; although it does not necessarily
imply a spontaneous event, the actor is simply not expressed. The fact that
spontaneous events are not generally expressed with a middle is probably linked to
the fact that the anti-causative is used for situations in which there is no agent.
Natural reciprocals are not middle derived, which is not surprising given that
there are reciprocal derivations, one of which contains the middle derivation;
see 3.3.
Verbs that do not easily fall in any of the categories that have been set, but that
can be analysed as having middle meaning, are:jegoox , ‘take possession of
something’, from jeg , ‘to possess’, which is directed at the self (subject);
dakoox, ‘go for a second time’, is derived from dak, ‘to step’ (compare with the
English ‘to come back on one’s steps’ to appreciate the middle aspect in its
meaning; a verb that has been reported middle in several African languages is ‘to
hide oneself’ (Mous 2004), Beedoox in the Ñominka dialect of Seereer, derived
from Beed , ‘to stalk, to approach for attack silently’, but in the standard Sine
dialect the verb ‘to hide’ is das and not middle derived.
Middle derived verbs need not be intransitive; in example (17) there is an
object pronoun and in example (18) we have a locative complement of the verb.
The reflexive can have an object suffix as long as that is different from the subject,
thus we can have a third person subject and third person object suffix but only if
not co-referential: (19) is impossible but (20) with an object pronoun that does not
refer to the subject is possible. The transitive verb rok-oox, ‘wear sth’, is middle
derived from the verb rok, ‘to enter’.
(17) o feet um Bog-oox-in
tomorrow 1SG wash-MIDDLE-O3SG
‘Tomorrow I’ll wash/bathe with it’

(18) a. in fop i ndoon-oox na Tat ale Gajak

1PL all 1PL PL:prepare:MIDDLE PREP-CL road CL-DEF Gajak
‘we leave (prepare the road) for Gajak, (after Mäkelä 1989:127)’
b. in fop i ndoon-oox-in
1PL all 1PL PL:prepare-MIDDLE-O3
‘We put it on ourselves’
S. FAYE & M. MOUS – Verbal system and diathesis derivation in Seereer 103

(19) *um Bog-oox-am

1SG wash-MIDDLE-O1SG
‘I wash me’

(20) a rok-oox-in
3 enter-MIDDLE-O3SG
‘She wears something’

The middle verb can either have the body (part) or the person as subject; compare
(21) and (22).
(21) kaa ta yen a Bol-oox no Bay
VERB.FOC 3SG fall 3 break-MIDDLE PREP arm
‘He fell and broke his arm’

(22) o-Bay ole a Bol-oox-a

CL-arm DEF 3 break-MIDDLE-PF
‘His arm broke’

The middle derivation is productive as long as we can imagine a situation in which

the middle semantics apply; for example, the middle derivation of ‘to eat’ is
possible in a metaphorical sense as in (23). With the verb lay , ‘to speak’, the
middle means ‘to speak to oneself’.
(23) o-Baye ole a _naam-oox-a
CL-arm DEF 3 eat-MIDDLE-PF
‘The arm has eaten itself / has become thin through a desease’

Some verbs with a middle derivation are lexically connected to the causative.
There are a number of verbs in the lexicon for which there are parallel derivations,
one with a causative -in for the active member of a pair and one with the middle
for the undergoer member of the pair while the base verb does not exist. Such
pairs are, for example, geen-in , ‘stop sb’, geenu, ‘to be up-right, to stand’,
dol-in, ‘to marry a woman’ and dolu, ‘to be married (for a woman)’.
There are also several verbs for which the middle form is the most common
and usual, while the form without the middle does exist but is semantically more
marked and has causative meaning. It is as if it is derived by subtraction of the
middle suffix; for example, kongomoox , ‘to sit with knees pulled up’, where
kongom is the causative equivalent ‘to make somebody sit with the knees pulled
up’. There are some middle derived verbs that require a causative derivation in
addition to the middle while the base with the causative is more marked in
meaning (and derived) than the middle plus causative; for example, gooknoox,
‘kneel’, gookin, ‘to make someone kneel’.
The combination of middle and anti-causative is possible. In (24) the new
subject is ‘water’, the complement that can be added as an object to the ‘reflexive’
verb without a preposition; it can be referred to by means of an object pronoun. In
fact, the middle has introduced the possibility of the complement that can become
the subject of the anti-causative: Bog, ‘wash the body’, Bog-oox, ‘wash oneself’,
104 AFRICANA LINGUISTICA 12 (2006)

allows for the object ‘water’ as application of the action. In a similar way the anti-
causative can be combined with a reciprocal. The verb maaf means ‘to make fall’;
its reciprocal derivative maaf-ir is used in the meaning ‘to wrestle’, but these
reciprocal verbs can be used in the singular (or plural) with an associative
complement with fo , ‘with’, as in 3.3, and this complement can become the
subject of the anti-causative verb (25). The anti-causative verb can have an
infinitival object as in (26). We can even have a anti-causative of an intransitive
verb (with frozen causative) with an object (27).
(24) foof le a Bog-oox-el
water DEF 3 wash-MIDDLE-ANTICAUS
‘The water will be washed with’

(25) xan um maaf-ir-el

FUT 1SG ground-RECIP-ANTICAUS
‘I will be wrestled’

(26) yaar-eem o futu faak

late-ANTICAUS:PF:1SG PREP have.lunch yday
‘I had lunch late yesterday’

(27) mi yegn-e ngulook ne Ndening

1SG be.informed-ANTICAUS marriage of Ndening
‘I have been informed of the marriage of Ndening’

The participle -u is homophonous but different from the middle (or reflexive)
derivation in -u . The two can be combined in which case the first, the middle
derivation, becomes a glide (28). The example also shows that in the participle
distinctions between active, middle and anti-causative can be made.

(28) a. rew maar-u

women plait-PART
‘the women who have plaited hair’
b. o-tew o-maar-w-u
CL-woman CL-plait-MIDDLE-PART
‘a woman who got herself plaited’
c. xa-kulook xa-maar-e
CL-puppets CL-plait-ANTICAUS.PART
‘puppets that were plaited’

3.3. The reciprocal derivation(s)

The reciprocal derivation is -ir. This morpheme appears in two forms: -ir and
-oor; the latter is a combination of the middle derivation -u and the reciprocal
-ir. The combination of the middle with the applicative -it (nominal allomorph
-ir) also results in a suffix -oor.
The mutation of the initial consonant marks plurality of subject; plural subjects
require a verb that is either prenasalized or a voiceless plosive or a consonant that
S. FAYE & M. MOUS – Verbal system and diathesis derivation in Seereer 105

does not undergo consonant mutation. Since the reciprocal often takes a plural
subject the initial consonant is either from the prenasalized or the voiceless plosive
series in the examples in (29).22
(29) fad ‘hit’ mbad-ir ‘hit each other’
maaf ‘make fall’ maaf-ir ‘wrestle’
xas ‘quarrel, sulk’ nqas-ir ‘quarrel’
Deet ‘look’ Teet-ir ‘look at each other’
lay ‘speak’ lay-ir ‘tell each other’
nan ‘hear’ nan-ir ‘listen to each other’
waag ‘get on’ mbaag-ir ‘be on good terms’
jop ‘fall’ jop-ir ‘fall onto something’
fog ‘to be long, share’ mbog-ir ‘be similar’

The associated partner(s) of the reciprocal can be expressed in a prepositional

phrase with fa/fo, ‘with/and’, in which case the verb can have a singular subject,
as in (30) and (31); the preposition is needed, because one cannot say *xan um
maafirin. The reciprocal verb is not by necessity intransitive.
(30) um simn-ir fa den laamiisoor fa den
1SG greet-RECIP with 3PL exchange.questions with 3PL
‘I greet them and exchange information with them’
(31) xan um maaf-ir fo ten / xan i maaf-ir fa den
FUT 1SG let.fall-RECIP with 3SG FUT 1SG let.fall-RECIP with 3PL
‘I will wrestle with him’ / ‘we will wrestle with them’
(32) ta jop-ir fa Astu / da njop-ir fa Astu
3SG meet-RECIP with Astu 3PL PL:meet-RECIP with Astu
‘he met with Astu’ / ‘they met with Astu’

The reciprocal is not always completely symmetrical in the action of subject to

object and object to subject, as we can have reciprocals like adn-ir , ‘walk in
line’, from adin , ‘follow sb’ (Mäkelä 1989:139). The reciprocal can express a
notion of togetherness as in fog-ir, ‘to be interrelated, be member of the same
group’ (Mäkelä 1989:149). Reciprocals can be derived from intransitive verbs.
The suffix - o o r represents two homophonous derivations: one is a
combination of the middle -u followed by the reciprocal -ir (33), and the other is
the middle -u followed by the applicative (instrumental) -it (34). The applicative
suffix -it is used as an independent suffix (see 3.6) and it has the form -ir in
nominalizations. In the case of the combination with the applicative, all the
attested cases are based on a middle verb. This is however not the case for the
combination of middle and reciprocal; for a number of verbs there seems to be no
intermediate stage of a middle derived verb. The verbs that are not middle derived
but take the middle reciprocal are not active: ‘to be right’, ‘hear’; this is also why
they need to add the middle to active reciprocal -ir.

22. Examples from Faye (1982); the last two are from Mäkelä (1989).
106 AFRICANA LINGUISTICA 12 (2006)

(33) jof ‘be straight, correct’ jof-oor ‘be on good terms’

nan ‘understand’ nan-oor ‘understand each other’
liiP ‘measure’ liP-oor ‘compare’
lay-u ‘talk to oneself’ lay-oor ‘translate one into the other’
nan-gilu ‘listen’ nan-gil-oor ‘understand each other’
23
(34) wetand-oox ‘remember’ wetandoor ‘remember because of sth’
ju’-oox ‘have supper’ ju’-oor ‘have supper with sth’
yok-u ‘get up’ yok-oor ‘get up with (e.g. a head ache)’
wond-u ‘lie down, sleep’ wond-oor ‘lie down with (a sheet) ’
mukt-u ‘wash one’s face’ muktoor ‘wash with’
sutu ‘leave, get out’ sut-oor ‘leave from’
inu ‘get up’ in-oor ‘come from, arrive’

Verbs like *maafoor from maaf, ‘make fall’, are not possible because the base is
active and the middle derivation, maafoox, ‘to let oneself fall’, is not something
one does to another and thus only the reciprocal in ir, maafir, ‘wrestle’, is used.
Derivations in ir and oor exclude each other for a number of verbs, e.g. jof, ‘be
right’, *jofir, jofoox , ‘to be straight for sb’, jofoor , ‘settle one’s
differences’. The homophony with the applicative middle -oor has certainly
limited the lexical opportunity for having a reciprocal -oor . A number of verbs
with the applicative middle -oor have a meaning that is quite different from the
meaning of the reciprocal in -ir of the same verb: lay , ‘speak’, lay-oox, ‘to
speak to oneself’, lay-ir, ‘to tell each other something’, lay-oor, ‘to plea for
somebody in court’.
There is a combined suffix -and-oor which means ‘to do something
together’.

3.4. Subject indirectly affected -iD

The suffix indicates that the verb applies to something that belongs to the subject,
but not typically the body of the subject. In this respect it is different from the
middle. The suffix relates the action of the verb to the subject but only through an
object that is different from the subject but possessed by the subject. Thus doxiD
from dox, ‘to burn’, is not used to refer to burning oneself, which is expressed by
dox-oox but means ‘to be affected in one of one’s possessions by burning’, e.g.
M uus a dosiDa, ‘Mous got something (maybe his house) burnt’, or Muus a
mos(i)Da, Suleeman a mosDee, ‘The one owned by Mous is nice, the one
from Souleymane not’ (talking about two people who both bought a pair of
trousers); based on mos , ‘be nice, good’. The subject is the person but it is only
indirectly affected by the notion expressed in the verb. The suffix requires a verb
that expresses a quality that can be associated with something one owns or
somebody one is associated with. It is an indirect middle.

23. The preceding examples are from Mäkelä (1989: 52); the next two examples are from
W. Faye (1980). The last example is from Mäkelä (1989:72).
S. FAYE & M. MOUS – Verbal system and diathesis derivation in Seereer 107

(35) mo’ ‘be lost, get lost’ mo’-iD ‘have sth lost’
jat ‘be bought (the good)’ jar-iD ‘have one’s goods bought by sb’
xon ‘die, be dead’ xon-iD ‘suffer from sb’s death’
dox ‘burn’ dox-iD ‘get something burnt’

3.5. The causative derivations

3.5.1. Direct causative -in

The causative suffix -in expresses a direct control of the subject (causer) of the
action over the caused agent. For example, moof-in, ‘to help sb to sit’, from
moof , ‘sit’, and Yuf-in , ‘drive’, from Yuf , ‘run’ (Faye 1982:10-11). The
derivation is productive except when a direct causative is unlikely, e.g. *Dof-in
from Dof , ‘be crazy’, for which only the indirect causative and the causative of
state are possible (see below).
(36) _naam ‘eat’ _naam-in ‘make sb eat’ (W. Faye 1979:99)
yer ‘drink’ yer-in ‘make sb drink’
dap ‘wash’ dap-in ‘make sb wash’
_no’ ‘sew’ _no’-in ‘make sb sew’
jaw ‘cook’ jaw-in ‘make sb cook/boil’
juD ‘grill’ juD-in ‘make sb grill’

(37) a yer-n-ik-a pis-ne

3 drink-CAUS-THITH-PF horse-DEF
‘He went to let the horse drink’

(38) juD-n-aam a-areer

grill-CAUS-1SG:PF CL-groundnuts
‘I had the groundnuts grilled’

3.5.2. Indirect causative -noor

The causative suffix -noor is used for indirect causation through a third party
(order, advise, etc). The suffix is most likely an amalgamation of -in-u-it
CAUS-MIDDLE-INSTR. Note that other languages also use a combination of middle
and causative to express indirect causation; see Mous (2004) for an indirect
causative in Konso (Cushitic) which contains a middle. Thus the suffix can be
used with the same verbs as those that apply -in but with a difference in meaning:
moof-noor, ‘make someone sit’, does not suppose the direct control of moofin
above, and the same with Yuf-noor, ‘let somebody run’, while Yufin means ‘to
drive (e.g. a car)’ (Faye 1982:11). The verbs in -noor suppose a causer that is
different from the agent. The examples in (39) are from W. Faye (1979).
108 AFRICANA LINGUISTICA 12 (2006)

(39) Daan ‘sleep’ Daan-noor ‘let sleep’

maaB ‘lie’ maaB-noor ‘let lie’
fi ‘do’ fi-noor ‘let do’
yer ‘drink’ yer-noor ‘let boil’
lay ‘speak’ lay-noor ‘let speak’

3.5.3. Causative of state -and

The suffix -and is only used with verbs of state. It expresses the entry into the
state (causative of inchoative). It can be used with de-adjectival verbs of state in
-ig such as yax-ig , ‘be red’, from yeeq , ‘red’, yax-g-and, ‘to redden, to
render red’.24 Note that there are two additional homophonous suffixes: -and for
the negative anti-causative (see section 2) and -and for locative nominalizations
as in o-weg-and ole, ‘prison’, from weg, ‘close’ (Fal 1980:115). The causative
of state only expresses the state as result. The verbs of state can also take the
direct or the indirect causative but there are differences in meaning, e.g. sum, ‘be
warm’, sum-in , ‘to heat sth’. The causatives of state are typically used in
expressions that describe a state in which ‘heat’ is used metaphorically, such as
sum-and jeetaay, ‘to heat up a conversation’, sum-and kam no kiin, ‘to
irritate sb’, lit. ‘to warm sb’s heart’ (Faye 1982:11). The difference with the direct
causative is evident in the derivatives of Dof, ‘crazy’: Dof-noor, ‘make a fool of
sb’, while Dof-and, ‘turn sb in a mentally disturbed person’, denotes a resultant
state rather than the activity.
(40) Bal-ig ‘be black’ Balg-and ‘make black’
ran-ig ‘be white’ rang-and ‘make white’
wod ‘be cured, well’ wod-and ‘cure’
mos ‘be beautiful’ mos-and ‘to make sth beautiful’

3.5.4. Causative of body state -il

The limited data on this suffix suggest that it is a causative that renders transitive a
verb that denotes a state of the body or the mind. The few examples of the
derivation in (41a) are from W. Faye (1980). Crétois (1973:107) discusses
nominalizations with a suffix -il into the n e class for abstract concepts. His
examples in (41b) also refer to states of body/mind. Finally, we have some
examples from Mäkelä (1989:23) (41c), for which we only have the derived verb
form. For the verb xij , ‘be tired’, we have both the verbal derivation xijil,
‘tire’, and the nominalization nqijil na, ‘tiredness’.25

24. W. Faye (1979, 1980) has a variant -d-and in _now-dand, ‘let live’. Mäkelä (1989) has
this verb as _noowdand from o-_noow ole, ‘life’.
25. There is also mucil na, ‘the greeting’, a lexicalized derivation from muc-il, ‘save’,
and muc, ‘be saved’.
S. FAYE & M. MOUS – Verbal system and diathesis derivation in Seereer 109

(41) a. muc ‘be saved’ muc-il ‘save sb’

fiiq ‘be tight’ fiiq-il ‘annoy’
b. sonil sonil na ‘the passion’
fiiqil mbiiqil na ‘the inconvenience’
c. wasil ‘vomit’
naqadil ‘irritate’
miriil ‘make dizzy’

3.6. The applicative -it

The suffix -it is used to indicate instruments (42a), accompaniment (42b), the
source of the action (42c). The suffix has to be added as soon as manner is into the
foreground as is the case with ‘how?’-questions such as (43). These meanings fall
under the functions of the applicative and are comparable to the functions of the
applicative in Bantu languages. The Seereer applicative is different from the Bantu
applicative with respect to its behaviour to beneficient human complements. When
the suffix is applied to double object verbs, it makes the object primary and the
sole object (42d) and no longer allows for the expression of the benificient object;
compare example (45) a and b, and see example (44). There is the benefactive
suffix discussed in 3.7 that has the function of introducing the human beneficiary
object. The anti-causative of the applicative of lalit, ‘show’, has the introduced
referent as subject and it means ‘to be shown’: Maye a laltel, ‘Maye was
shown’ (Mäkelä 1989:139).26
(42) a. Deg ‘cut’ Deg-it ‘cut with (a knife)’
fad ‘hit’ fad-it ‘hit with (a stick)’
god ‘cut’ god-it ‘cut with’
b. gat ‘return’ gat-it (xaaliis) ‘return with (money)’
lay ‘talk’ lay-it ‘talk with’
c. fool ‘jump’ fool-it ‘jump from’
Yuf ‘run’ Yuf-it ‘run from’
d. Band ‘lend’ Band-it ‘lend sth’
coox ‘give’ coox-it ‘give sth’
ci’ ‘offer’ cit ‘offer sth’
lal ‘show sb’ lal-it ‘show sth’

(43) a. nam i yegn-it-k-eel

how 1PL be.informed-APPL-FUT-ANTICAUS:IMPFV
‘How are we going to be informed?’ (Mäkelä 1989:134)
b. nam ‘la bouteille’ a lay-t-eel na seereer
how ‘la bouteille’ 3 say-APPL-ANTICAUS:IMPFV PREP sereer
‘How does one say ‘la bouteille’ in Sereer?’ (Mäkelä 1989:134)

(44) lalit o-yarax a faaxa muu ref-na

show CL-politeness 3 good-PF place be-SUBORD
‘To show politeness is good everywhere’ (Mäkelä 1989:96)

26. It is the anti-causative of da lalit a Maye, ‘They showed Maye’.

110 AFRICANA LINGUISTICA 12 (2006)

(45) a. Cungaam um coox-it Pisnit nene (*a Samba)

wait-IMPFV:O1SG 1SG give-APPL task CL-DEM
‘Wait, I transfer that task ....’
b. Cungaam um coox Pisnit nene a Samba
wait-IMPFV:O1SG 1SG give task CL-DEM to Samba
‘Wait, I give that task to Samba’

The double object verbs commonly have a human object; with verbs that do not
usually have a human object, the applicative derivation -it can also introduce a
human beneficiary object, but in most cases the derivation introduces (non-
human) instruments, accompaniments and manner. For example, fa_nit means ‘to
refuse somebody’ while fa_n means ‘to refuse to do something’, jo’it ‘let the
words of somebody hear to another’ (Mäkelä 1989:127).
There are two suffixes -it: the applicative that does not have any effect on the
preceding sounds, and the reversive that affects the preceding syllable -it! (46).
The exclamation mark indicates the effect on the preceding syllable in order to
distinguish the two suffixes. There is no evidence to determine the behaviour of
the suffix in some of the verbs. The suffix -it! renders the final consonant of the
root a voiceless stop and reduces preceding long vowels to short vowels. The
reversive suffix -it! is used to indicate separation (46a) and diminutive (46b).
(46) a. uup ‘bury’ up-it ‘take out what is buried’
sux ‘cork’ suq-it ‘to uncork’
muur ‘buckle’ mur-it ‘disclose’
duf ‘plant’ dup-it ‘remove what is planted’
hul ‘cover’ hul-it ‘uncover’
lok ‘hook’ lok-it ‘unhook’
b. maak ‘grow’ mak-it ‘grow a bit’
sax ‘sprout’ saq-it ‘grow a bit’
seec ‘dry in sun’ sec-it ‘dry in sun a bit’
seB ‘cut in pieces’ seP-it ‘cut in small pieces’

3.7. The benefactive -an

The derivation indicates that the action has a beneficiary (47, 48a), a maleficiary
(48c) or a purpose (48b). In the example in (48a) the beneficiary object is in focus
and sentence-initial; the non-subject focus form of the imperfective verb is used
(cf. table 3). When the complements are in focus and in the sentential initial
position, they typically are introduced by a benefactive or applicative derivation.
(47) jal ‘work’ jal-an ‘work for sb’
xoox ‘cultivate’ xoox-an ‘cultivate for sb’
un ‘grind’ un-an ‘grind for sb’
riw ‘weave’ riw-an ‘weave for sb’
bind ‘write’ bind-an ‘write for sb’
Deet ‘look’ Deet-an ‘look for sb’
S. FAYE & M. MOUS – Verbal system and diathesis derivation in Seereer 111

jik ‘buy’ jik-an ‘buy for sb’

waaj ‘prepare’ waaj-an ‘prepare for sth’
diC ‘place’ diC-an ‘place for sb’
naq-an ‘save food for sb’

(48) a. Saaliiw um jal-an-aa

S S1.SG work-BEN-DUR
‘I work for Saliou’
b. xaaliis um jal-an-aa
money S1.SG work-BEN-DUR
‘I work for money’
c. fuux-an ‘be angry with sb’

The original object is still present in the sentence as a second object; the
beneficient object is usually expressed in an object pronoun (49, 50). In
constructions with an auxiliary, the object pronoun is on the auxiliary, but the
beneficient derivation on the main verb (51).
(49) fat o jaw-an-aam a-put
OBLIG 2SG prepare-BEN-O1SG CL-breakfast
‘Make me my breakfast!’ (Mäkelä 1989:20)

(50) a. ndax wo seq-an-axam xa-Taq axe

QUES 2SG have-BEN-PF:O1SG CL-10cfa DEF
‘Is it you who owes me 10 CFA?’ (Mäkelä 1989:75)
b. ndax o-maad oxe seq-an-axam xa-Taq axe
QUES CL-king DEF have-BEN-PF:O1SG CL-10cfa DEF
‘Is it the king who owes me 10 CFA?’

(51) ndax waag-axam-o jaw-an ruuy

QUES can-PF:O1SG-2SG prepare-BEN porridge
‘Can you make me the porridge?’ (Mäkelä 1989:76)

4. Summary and conclusion

The verbal system of Seereer has many different options to manipulate the
presentation of the state of affairs as represented by the verb and its complements.
The default meaning of a verb root is one with an active agent subject acting on a
second argument. The anti-causative inflection takes away the possibility to
mention the agent. As a consequence something else becomes the subject of the
verb. Foregrounding an object can be achieved by fronting and using the
complement focus verb forms. Once we focus and front a circumstantial
complement such as a manner phrase, this phrase needs to be introduced by an
applicative derivation. The function of the verbal derivations is to change the
argument structure so that another complement becomes linked to the verb. The
middle derivation indicates that the action of the verb is primarily in the subject.
There is another derivation that indicates that the subject is affected but indirectly
112 AFRICANA LINGUISTICA 12 (2006)

by something associated to the subject, a kind of indirect middle. The causative

distinguishes between control on the part of the subject and indirect causation.
The distinction between derivation and inflection becomes problematic when
we take the anti-causative into account. In its position in the verb, it is among the
inflections. In the expressing tense/mood/aspect position there are other markers
that have been presented as derivation in Seereer studies. These are the
morphemes -at and -aful. The anti-causative is also inflectional in the fact that
participles occur in active and anti-causative forms. Despite this inflectional
nature of the anti-causative, there are verbs with a frozen anti-causative and
nominalizations include the addition of the anti-causative morpheme. There are no
other inflectional markers that occur in proper nominalizations and some of the
other nominalizing suffixes are related to the verbal derivations.

References

Crétois, L. 1973. Dictionnaire sereer-francais: différents dialectes. Centre de

linguistique appliquée de Dakar. Coll. Les langues africaines au Sénégal (48).
Fal, A. 1980. Les nominaux en Sereer-Siin: parler de Jaxaaw. Dakar: Nouvelles
Éditions africaines.
Faye, S. 1981. Guide Pratique du “Seereer” Écrit. Dakar: CLAD, Les langues
nationales au Sénégal, n°82.
—1982. Morphologie du verbe sérère. Dakar: CLAD, Les langues nationales au
Sénégal, n° 84.
—1985. Morphologie du nom sérère: système nominal et alternance
consonantique. Dakar: Centre de linguistique appliquée de Dakar, Université
de Dakar, Les langues nationales au Sénégal, N.S. 10.
Faye, W.C. 1979. Étude morphosyntaxique du sereer singandum: parler de
Jaxaaw et de Naaxar. Thèse. Grenoble: Université de Grenoble III.
—1980) Précis grammatical de Sérère. Dakar: Centre de linguistique appliquée
de Dakar.
Kemmer, S. 1993. The Middle Voice. Amsterdam: John Benjamins.
Mäkelä, K. 1989. Aqatin seereer Guide pratique pour apprendre le seereer.
Dakar: La Mission luthérienne du Sénégal.
McLaughlin, F. 2000. ‘Consonant mutation and reduplication in Seereer-Siin’,
Phonology 17(3):333-63.
Mous, M. 2003. ‘The Middle in Cushitic Languages’. In A. Simpson (ed.),
Proceedings of theTwenty-Seventh Annual Meeting of the Berkeley Linguistics
Society, March 22-25, 2001: Special Session on Afroasiatic Languages, 75-86.
Berkeley: University of California, Linguistic Society.
— 2004. ‘The Causative in Konso’. In R. Kastenholz & A. Storch (eds), Sprache
und Wissen in Afrika. Beiträge zum 15. Afrikanistentag, Frankfurt an Main
und Mainz, 30. September-2. Oktober 2002, 213-228. Köln: Rüdiger Köppe.
Mous, M. & Qorro, M. 2000. ‘Middle voice in Iraqw’. In K. Kahigi, Y. Kihore &
M. Mous (eds), Lugha za Tanzania / Languages of Tanzania: Studies
Dedicated to the Memory of Prof. Clement Maganga, 141-59. Leiden: CNWS.