Predication Strategies

This section will discuss predication as is particularly relevant to adjectives and other modifiers. Sereer heavily favors verbal predication - most adjectives are derived from verbs, and stative verbs serve to express many of the meaning commonly expressed cross-linguistically with adjectives. Additionally, Sereer exhibits a complex system of copularization; there are for distinct copulas markers, each of which have a distinct functional load.

The following is a table with a variety of verbs, both regular and stative, and some that lexically encode meanings prototypically adjectival (such as 'be a lot'):

Error creating thumbnail: Unable to save thumbnail to destination

All tense and aspect inflections are available to stative verbs. For instance, the present progressive xe V-aa can be used with any of the above. The following table represents a sampling of inflectional and derivational forms associated with verbs, and illustrates the broad range of predicate types each of these can be used with:

Error creating thumbnail: Unable to save thumbnail to destination

It is worth noting that, for special derived adjectives such as ciig/jigid/jigdu ‘be tall’, it is the adjectival derived form (jig(i)d, yax(i)g, etc.) which is most often used in typical verbal inflection paradigms (rather than the root forms ciig, ɓaal, etc).

The way in which a verb with negative derivational morphology -eer in the table above can become an adjective is exemplary for the way all derived verbs, stative and regular, can take adjectival and verbal morphology. For instance, in the first example, the verb 'be spicy' has one layer of derivational morphology, and in the second example, the verb 'be brave' has two layers, before each becomes an -u adjective:

 ñaay   -->  ñaayeer --> oɓiy oñaayeeru le                   'the not-spicy fruit'
 sad    -->  sadar   --> sadarnoox --> okoong osadarnooxu    'the scary gorilla'

Copular predication

Sereer exhibits four distinct copulas, shown in the table below. Both ref and xe mark predicational and specificational copular clauses. ref can mark both locational and nominal complements, which are stage-level and individual level-predicates respectively (SLP and ILP). xe can only marked stage-level predicates (SLP), and also functions as a verbal auxiliary. The two other copula markers, jeg and =oo, have more limited distributions; jeg is more commonly a lexical verb `have, own', but is used in existential constructions. The enclitic =oo is used in equative constructions. Each of the copularization strategies are discussed in more detail below.

Error creating thumbnail: Unable to save thumbnail to destination

The Main Copulas ref and xe

The most common copula is ref, which marks predicational and specificational copular clauses with locational and nominal complements, examples of which are given in <glr id="ref.loc"/> and <glr id="ref.nom"/>.

<gl id="ref.loc" fontsize=12> muus ne arefa took atabula le \gll muus ne a= ref -a took atabula le cat DET 3= COP -DV on table DET \trans The cat is on the table. (274) </gl>

<gl id="ref.nom" fontsize=12> Ami arefa otew \gll Ami a= ref -a otew Ami 3= COP -DV woman \trans Ami is a woman. (274) </gl>

The auxiliary xe is the both common copula for locational constructions (SLPs), as shown in <glr id="xe.loc"/>.

<gl id="xe.loc" fontsize=12> oɓox ole axe tafil andok um \gll oɓox ole a= xe tafil andok um dog DET 3= COP outside room 3POSS \trans The dog is outside his room. (102) </gl>

It is not possible to use xe as a copula with nominal complements (with ILPs), such as *Ami axe otoog. , which would means `Ami is a woman.' Constructions of this type must use ref as a copula. However, xe is also quite productive as a verbal auxiliary, discussed in Verb Phrases.

Unlike xe, ref is a regularly inflected verb `to be' that frequently acts as copula, and works in both predicational and specificational clauses. With a specificational clause, the specified role-filler follows the verb, as shown in <glr id="Specificational"/>

<gl id="Specificational" fontsize=12> oxe adooxan na USA a= ref -a Obama. \gll oxe adooxan na USA a= ref -a Obama pro leader Part USA 3s= COP -DVObama \trans The one who is president of the USA is Obama. </gl>

(Note: na here is not the relative pronoun, but the partitive preposition. See Noun Modification with -u for more.)

With predicational clauses, it can either provide neutral focus by acting as a verb <glr id="pred.non.foc"/>, or marked focus in which case the verb must be marked with the extraction marker -u, as shown in <glr id="subj.foc"/> for subject focus, and <glr id="obj.foc"/> with object focus, in which case the object appears before the copula.

<gl id="pred.non.foc" fontsize=12> oɓox ole arefa ɓaal fo tan \gll oɓox ole a= ref -a ɓaal fo tan sg-Cl 3sg is 3sg black and white \trans The dog is black and white. (128) </gl>

<gl id="subj.foc" fontsize=12> Jegaan refu faap fe \gll Jegaan ref -u faap fe Jegaan COP -EXTR father DET \trans Jegaan is the father. (225) </gl>

<gl id="obj.foc" fontsize=12> oɓox ole obaabar arefu \gll oɓox ole o- baabar a= ref -u dog DET AGR- killer 3= COP -EXTR \trans The do is a killer. (225) </gl>

Focus in copular clauses can also be achieved an emphatic surfacing of the subject pronoun:

 mi, o caajang ref-um      'I am a student.'
 wo, o caajang ref-o       'You are a student.'
 ten, o caajang aref-u     'He is a student.'
 ino, jaajang indef-u      'We are students.'
 nuuno, jaajang nundef-u   'We are students.'
 deno, jaajang andef-u     'They are students.'

The latter points to a second verbal paradigm reserved for the ref copula when it appears in predication with a noun. When it does so, the noun precedes the verb, whereas when the predicate is verbal or prepositional, it follows ref:

<gl id="Specificational" fontsize=12> arefa koƭu koƭu \gll a= ref -a koƭ -u koƭ -u 3= COP -DV far Adj far Adj \trans He is far away. (143) </gl>

This paradigm is available only to ref by virtue of it being the only verb that can predicate nominally (unless we see all instances of -u modification of nouns, (i.e., typical adjectives) as the same type of predication as this). The paradigms are:

Error creating thumbnail: Unable to save thumbnail to destination

In the following example involving a wh-question, ref patterns like an adjective modifying the generic (that is, noun-class-indeterminate) wh-pronoun wum (see the section on Questions for more on wh-pronouns):

<gl id="Specificational" fontsize=12> \gll wum ndefu we yaxig na Wh are Det red Rel \trans Which are the ones that are red? </gl>

When appearing in wh- constituent questions, ref must follow the syntactic pattern in (3), which is focused (that is, it must be an adjectival form inside a noun phrase). This makes sense for wh- constituent questions, which ask for new information regarding the identity of the referent inquired after with 'wum', and therefore these questions are focal in nature.

In the next example, it's clear that ref can take the full range of modifier morphology, including na-modification, which is another type of clefting strategy, in addition to the one described above, but this time via relativization:

<gl id="Specificational" fontsize=12> \gll fañiik fe ref na ondeb elephant DET COP REL small \trans The elephant that is small. </gl>

Existential verb jeg

As mentioned above, jeg appears on copular clauses to mark existential constructions. Existential constructions in English are identified by the expletive there, follow by a conjugated form of the verb </>be, a nominal expression, and finally, a predicate. In Sereer, the expletive `there' and `be' is expressed by use of the verb jeg, which normally means `have, own'. The use of jeg as a regular verb is shown in <glr id="jeg.have"/>.

<gl id="jeg.have" fontsize=12> Ami ajega fambe le \gll Ami a= jeg -a fambe le Ami 3= {have (*EXT)} -DV goat DET \trans `Ami has a goat.' (*There is an Ami goat.) (298a) </gl>

The sentence in <glr id="jeg.have"/>, by virtue of having a viable subject before the verb and an object following it, cannot be interpreted as an existential. It is only possible to interpret jeg as an existential when it is not preceded by a subject. This is shown in <glr id="jeg.existential.1"/>. Even these types of constructions can also be interpreted as `have', however. Although <glr id="jeg.existential.1"/> would be pragmatically odd, it is possible to coerce the meaning `Someone owns a man.' Additionally, as seen in <glr id="jeg.existential.2"/> and <glr id="jeg.existential.3"/>, most sentences that can be read with the existential can also take the `have' meaning.

<gl id="jeg.existential.1"fontsize=12> ajega okoor \gll a= jeg -a okoor 3= EXT -DV man \trans There is a man. </gl>

<gl id="jeg.existential.2"fontsize=12> ajega pis kam kud ne \gll a= jeg -a pis kam kud ne 3= EXT/have -DV horse in stable DET \trans There is a horse in the stable./He has a horse in the stable. (298a) </gl>

<gl id="jeg.existential.3"fontsize=12> ajega xaƥox xabaabar \gll a= jeg -a xaƥox xabaabar 3= EXT/have -a dog.PL killer.PL \trans There are killer dogs./He has killer dogs. (298a) </gl>

<gl id="ref.loc" fontsize=12> muus ne arefa took atabula le \gll muus ne a= ref -a took atabula le cat DET 3= COP -DV on table DET \trans The cat is on the table. (274) </gl>

Because most existential constructions can also be interpreted with possession/ownership readings, it is useful to have diagnostics that treat the two readings differently. One very useful diagnostic is the presence or absence of definiteness effects. Definiteness effects arise when a construction is unable to take as its complement a definite noun, proper name, or and strong quantification like `all' and `every', in contrast to weak quantifiers like `some', `few', and `two'. As shown in <glr id="jeg.definite.effects.1"/>-<glr id="jeg.definite.effects.4"/>, the existential reading of jeg is sensitive to definite effects.

<gl id="jeg.definite.effects.1"fontsize=12> ajega xaƥox kam mbine \gll a= jeg -a xaƥox kam mbine 3= EXT -DV dogs in house.DET \trans There are (some) dogs in the house. (286) <\gl>

<gl id="jeg.definite.effects.2"fontsize=12> ajega xaƥox xadak/xamayu/ondiik kam mbine \gll a= jeg -a xaƥox xadak/xamayu/ondiik kam mbine 3= EXT -DV dogs two/many/few in house.DET \trans There are two/many/few dogs in the house. (286) <\gl>

<gl id="jeg.definite.effects.3"fontsize=12>

• ajega xaƥox axe kam mbine\\

Intended: 'There are the dogs in the house.' (286) <\gl>

<gl id="jeg.definite.effects.4"fontsize=12>

1. ajega Stella kam mbine
2. 'There is a Stella in the house.' (286)

More acceptable: 'Someone owns/has Stella in the house.' <\gl>

In <glr id="jeg.definite.effects.1"/>, the unmarked NP `dogs' allows the existential reading, as do the weak quantifiers shown in <glr id="jeg.definite.effects.2"/>. However, the examples in <glr id="jeg.definite.effects.3"/> and <glr id="jeg.definite.effects.4"/> cannot be understood with the existential reading, but only the `have' meaning.

Equative enclitic =oo

Sometimes with pure nominal predicates, the particle =oo is used instead of a copula, and predication occurs simply by the subject and predicative noun plus the copular particle:

<gl id="ex2" fontsize=12> \gll Kayla o tew o. Kayla sg-Cl woman Pcl \trans Kayla is a woman. </gl>

The meaning is equivalent to a version of this substituting the particle for a verbal form of ref:

<gl id="ex2" fontsize=12> \gll Kayla o tew a- ref -u. Kayla sg-Cl woman 3sg is 3sg \trans Kayla is a woman. </gl>