Serial Verb Constructions Research Paper

(1)       Kofi de                        pono no baae

Kofi take-PAST                       table the come-PAST

‘Kofi brought the table’ or more literally,

‘Kofi took the table and came with it’

SVCs have attracted attention from linguists for three main reasons. First, syntacticians have found them interesting because the possibility of having more than one main (i.e., nondependent, nonauxiliary) verb within a clause or clause-like unit challenges traditional assumptions that a clause contains exactly one predicate (see Foley and Olson 1985, pp. 17, 57). Second, because of their prevalence both in languages of West Africa and in certain Creole languages, they have become central to debates among Creole language researchers, being seen either as evidence for the importance of substrate (i.e., African mother-tongue) influence on the Creole, or as deriving from universal properties of human language which are manifested in the creolization process. Third, serial verbs show a tendency to be reanalyzed as other grammatical categories (complementizers and prepositions), with implications for the processes of historical change.

While different researchers use different criteria for defining serial verbs, the following set would probably be accepted by most and is based on one described by McWhorter (1997, p. 22) as a ‘relatively uncontroversial distillation of the conclusions of various scholars’:

(2)        (a) an SVC contains only one overt subject;

(b) it contains no overt markers of coordination or subordination;

(c) it expresses concomitant actions (either simultaneous or consecutive) or a single action;

(d) it falls under one intonational contour;

(e) it has tense–modality–aspect marking on none of the verbs, or on one only (usually the first or last in the series) or on all of them; and

(f ) it contains a second verb which is not obligatorily subcategorized for by the first verb.

Even the above six conditions leave it unclear where exactly to draw the line around SVCs, particularly in languages which have little morphology. In some cases of what may appear to be SVCs, we may be dealing simply with unmarked coordination of verb phrases expressing simultaneous actions.

2. Serial Verb Types

Serial constructions can be classified into different types on the basis of the functions of the verbs in the series relative to one another.

The sentences in (3) provide examples of different functional types of SVC which have been discussed in the literature.

(3)        (a) Directional complements: V₂ is go, come, or another common intransitive motion verb which functions to indicate the directionality of the action denoted by V₁.

Olu gbe          aga       wa

Olu take         chair    come

‘Olu brought a chair’ (Yoruba, West African)

(b) Other motion erb complements: similar to type

(a), but V₂ may be transitive motion verb with its own object expressing the goal of the action denoted by V₁

Kofi tow agyan no w Amma

Kofi throw-PAST arrow the pierce-PAST Amma

‘Kofi shot Amma with the arrow’

(c) Instrumental constructions: V₁ is take or a semantically similar common verb, while V₂ denotes an action performed with the aid of the object of V₁ .

Kofi teki a nefi koti a brede

Kofi take the knife cut the bread

‘Kofi cut the bread with a knife’ (Sranan, Caribbean Creole)

(d) Dative constructions: V₂ is usually give, with an object which denotes the (semantic) indirect object of V₁.

Ogyaw ne sika maa me

he-leave-PAST his money give-PAST me

‘‘He left me his money’’ (Twi, West African)

(e) Comparative constructions: a verb meaning pass or surpass is used with an object to express comparison. In this case ‘V ’ is often in fact an adjective.

Amba tranga pasa Kofi

Amba strong pass Kofi

‘Amba is stronger than Kofi’ (Sranan, Caribbean Creole)

(f ) Resultative constructions: V₂ denotes the result or consequence of an action denoted by V₁.

Kofi naki Amba kiri

Kofi hit Amba kill

‘Kofi struck Amba dead’ (or more literally,

‘Kofi hit Amba and killed her’) (Sranan, Caribbean Creole)

(g) Idiomatic constructions (lexical idioms): These are cases where the meaning of the verbs together is not derivable from the meanings of the verbs separately. They are not found in all serializing languages, but some of the West African group are particularly rich in them.

Anyi-Baule (West African):

b ‘hit’ nıa ‘look’ b …nıa ‘say, tell’

Yoruba (West African):

la ‘cut open’ ye… ‘understand’ la…ye‘explain’

3. Distribution In The World’s Languages

Serial verb constructions have been identified in many languages but it is clear that they occur in areal clusters and are not spread evenly among the languages of the world. So far serializing languages have been identified in the following areas: West Africa (Kwa and related language families), the Caribbean (Creole languages which have a historical relationship with Kwa languages), Central America (e.g., Misumalpan), Papua New Guinea (Papuan languages and Tok Pisin— Melanesian Pidgin English), South-east Asia (e.g., Chinese, Vietnamese, Thai). There are isolated reports of SVC-like constructions from elsewhere. How credible these are depends on which criteria are adopted to define SVCs.

4. Grammatical Analyses

Numerous researchers have put forward proposals for grammatical analyses of SVCs. The analyses may be classified into two types, with some degree of overlap. Semantic analyses typically seek to account for SVCs in terms of how they break down verbal concepts into more basic semantic components (take–carry–come for bring, for example). Syntactic analyses differ mainly according to whether they treat SVCs as a phrase structure phenomenon (usually under a version of X-bar theory), as a form of complementation, subordination, or secondary predication (i.e., involving more than one clause-like unit), or as a lexical phenomenon (e.g., as single but disjoint lexical items).

The following phrase structure (4) proposed for SVCs in the literature is typical of those advocated by a number of researchers:

(4)        VP → V XP VP where X is N or P

This produces a right-branching tree with a theoretically infinite number of verbs in a series. Such a structure would allow lexical relations and relations like ‘subject-of’ and ‘object-of’ to hold between verbs in the series, and captures the intuition that verbs (or verb phrases) function as complements to verbs earlier in the sequence.

Some researchers have treated SVCs as involving complementation or subordination, with clauses or clause-like units embedded within the VP and dependent on a higher verb. For example, Larson (1991, pp. 198–205) has argued in favor of analyzing serial constructions as a case of ‘secondary predication.’ A structure similar to that for Carol rubbed her finger raw in English can also, he says, account for serial combinations of the take… come and hit… kill types (a) and (f ) above).

Foley and Olson (1985) propose a distinction between ‘nuclear’ and ‘core’ serialization. In nuclear serialization the serial verbs all occur in the nucleus of the clause, in other words the verbs form a single unit which must share arguments and modifiers. The core layer of the clause consists of the nuclear layer plus the core arguments of the verb (Foley and Olson 1985, p. 34). In ‘core’ serialization two cores, each with own nucleus and arguments, are joined together. Languages may exhibit both kinds of serialization or just one. Examples (a) and (b) below (Foley and Olson 1985, p. 38) illustrate this difference; (a) is an example of core, and (b) of nuclear, serialization:

(5)         (a) fu fi fase isoe

 he sit letter write

‘He sat down and wrote a letter’

(Barai, Papuan)

(b) Fu fase fi isoe

He letter sit write

‘He sat writing a letter’

This evidence suggests that ‘serialization’ may not be a unitary phenomenon, and although a single account which explains the two or more different construction types (such as Foley and Olson’s) would be welcome, it may be that separate analyses may be needed as more different types of SVC come to light.

5. The Creolists’ Debates

SVCs are among the most salient grammatical features of many Creoles of the Caribbean area which distinguish them grammatically from their European lexifier languages. As such they have attracted attention from creolists, who, finding similar structures in West African languages which historically are connected with the Creoles in question, have claimed SVCs are evidence of substrate influence on the grammar of the Creoles concerned.

An alternative view has been offered by Bickerton (1981) and Byrne (1987), who regard the similarities between SVCs in the Creole and West African languages as coincidental. Instead, they argue, SVCs result in Creoles from universal principles of grammar, and are a consequence of the rudimentary verbal structure (V but not VP) which exist, they say, in a Creole at the earliest stage of development. They point out the existence of SVCs or serial-like structures in other pidgins and Creoles such as Tok Pisin (New Guinea Pidgin) and Hawaiian Creole English, which have no historical connections with West Africa.

McWhorter (1997) argues against Bickerton and Byrne, claiming that the Caribbean Creoles not only share SVCs with West African languages, but that their SVCs are structurally similar to each other in ways that SVCs from other areal language groupings are not. Using the nuclear core distinction proposed by Foley and Olson (see above), he argues that Papuan languages serialize at the nuclear level, while Kwa languages, Caribbean Creoles, Chinese, and other Southeast Asian languages serialize at the core level. Considering also the range of pidgins and Creoles which lack SVCs altogether, he concludes that ‘SVCs have appeared around the world in precisely the creoles which had serializing substrata’ (1997, p. 39).

SVCs have thus become a major site of contention between ‘substratists’ and ‘universalists’ in Creole studies, with both sides remaining committed to their own positions.

6. SVCs And Language Change

Particular verbs which participate in serial constructions in some cases appear to have been reanalyzed as members of another syntactic category, e.g., from verb to preposition (give → for) or complementizer (say → that). In such cases, the position of the verb usually makes it amenable to such a reanalysis (e.g., if it typically occurs immediately before its objects, it may be reanyalzed as a preposition if the semantics encourage this interpretation). Reanalyzed serial verbs lose, to varying degrees, their verbal properties and take on, again to varying degrees, the morphological characterisics of their new category. Lord (1973) describes cases of reanalysis of verbs as prepositions, comitative markers, and a subordinating conjunction in Yoruba, Ga, Ewe, and Fon. Verbs with the meaning ‘say’ are susceptible to reinterpretation as complementizers (introducing sentential clauses), while those with the meaning ‘finish’ have a tendency to be interpreted as completive aspect markers.

7. Related Constructions

Lack of a satisfactory definition of ‘serial verb construction’ makes it difficult to decide what may count as a related kind of structure. Chinese has a range of verbal structures which bear resemblances to SVCs. One of these, the class of coerbs, is widely believed to be the result of the reanalysis of serial verbs as prepositions. Others, so-called resultatives, resemble secondary predications.

The Bantu languages, though related to the serializing languages of West Africa, do not seem to have SVCs. However, some have verbal chains where the second and subsequent verb have different marking from the first. In many Bantu languages, the verb meaning say is homophonous with a complementizer. These phenomena remain to be explained by a general theory.

Bibliography:

1. Bickerton D 1981 Roots of Language. Karoma, Ann Arbor, MI
2. Byrne F 1987 Grammatical Relations in a Radical Creole: Verb Complementation in Saramaccan. Benjamins, Amsterdam
3. Foley W A, Olson M 1985 Clausehood and verb serialization. In: Nichols J A, Woodbury A C (eds.) Grammar Inside and Outside the Clause: Some Approaches to Theory from the Field. Cambridge University Press, Cambridge, UK, pp. 17–60
4. Larson R K 1991 Some issues in verb serialization. In: Lefebvre C (ed.) Serial Verbs: Grammatical Comparative and Cognitive Approaches. Benjamins, Amsterdam, pp. 184–210
5. Lefebvre C J (ed.) 1991 Serial Verbs: Grammatical, Comparative and Cognitive Approaches. Benjamins, Amsterdam
6. Lord C 1973 Serial verbs in transition. Studies in African Lingusitics 4(3): 269–96
7. McWhorter J H 1997 Towards a New Model of Creole Genesis. Peter Lang, New York
8. Sebba M 1987 The Syntax of Serial Verbs: An Investigation into Serialization in Sranan and other Languages. Benjamins, Amsterdam