Printer Friendly

The Serial Verb Phenomenon in Gojri: An Introduction.

Byline: Nadeem Haider Bukhari


This paper is a an attempt to bring the Gojri serial verbs constructions in focus and try to shed over the prevailing notion regarding the nature of the serial verbs in South Asian languages that they are one of the forms of complex predicates. I will come up with some suggestions and propose that serial verbs have their own domain as they exhibit some very unique properties. The detailed discussion of all the concerning issues is beyond the scope of this short paper, therefore a brief discussion will be presented to capture the basic properties of Gojri serial verb constructions.

Key Words: Serial verbs, Complex Predicates, Argument Sharing, Coordination,

Subordination, Conjunctive Particles, Embedding Structures

I. Introduction

Serial verb constructions are found cross-linguistically in many languages of the world. This phenomenon is more common in African and Asian languages. Similarly, most Creole languages of the Atlantic and the Pacific exhibit this construction quite frequently. Consider the following languages.

1. a. di uman kuk res sel(Krio, Johnson 2002:41) The woman cook rice sell 'The woman cooked some rice and sold it.'

b. bola se dran ta (Yoruba, Lord 1974) bola cooked meat sell 'Bola cooked some meat and sold it.'

c. kofi ts ati-e fo yao(Ewe, Collins 1993:34) kofi took stick-def hit Yao 'Kofi took the stick and hit Yao with it.'

South Asian languages are not well known for their serial verb constructions. However, Jayaseelan (2004) has reported this phenomenon for some Dravidian languages. The following examples were reported to show the SVCs.

2. a. raaman awan-ai aDiyttu-koND-aan (Tamil, Jayaseelan: 71) Raman he-acc. hit-take (Past)-3sg 'Ram hit himself.'

b. naan oru maanga poTTiccu tinnu-u (Malayalam,Jayaseelan:67) i a mango pluck eat-Past 'I plucked and ate a mango.'

Gojri is one of the Indo-Aryan languages in the region that exhibits serial verb construction. Though their use is very limited, their syntactic structure may be worthy of interest. There is a commonly prevailing notion that serial verbs are a type of complex predicates which are found in many South Asian languages, including Gojri. For example, Jayaseelan (2004:69-70) considers the following examples as the representative of serial verb constructions.

3. a. nii oru pustakam koNDU-war-uu you a book take-come-IMP 'You bring a book.'

b. naan oru kattu ezhuti-(y)iTT-uNDu I a letter write-put-be(Pres.) 'I have written a letter.'

Jayaseelan claims that aspectual meanings, which are expressed by the auxiliary in English, are expressed by SVs in these languages. However, it is a different case in the Gojri structure. In Gojri, Hindi, Urdu and Punjabi, aspectual meanings are generally generated with complex predicates, a common formation in Indo-Aryan languages. I will discuss the differences between the SVCs and complex predicates in the subsequent sections and show that serial verbs differ from complex predicates in many respects.

The study of serial verb construction became popular in the late 80s. However, the research targeted only a couple of languages including Yoruba and Khmer.1 Recently, Baker (2002) has initiated such studies with the comparative study of serial verbs in polysynthetic languages.

Serial verb constructions consist of two or more verbs that occur in a sequence, without any intervention in between them. Interestingly, Gojri is a language that can also accommodate more than two verbs in serial verb formations. Consider the following examples:

4. a. kaloo-ne seb chillii kut@rii khaayo kaloo-ERG apple-NOM peel-SVI cut-SVI eat-PF.M 'Kaloo peeled the apple, cut it (in pieces) and finally ate it.

b. kaloo-ne seb kut@rii khaayo kaloo-ERG apple-NOM cut-SVI eat-PF.M 'Kaloo cut the apple (in pieces) and ate it.

Example (4a) shows that there are more than two serial verbs involved in the example while (4b) simply shows two serial verbs in it. It is important to mention here that the non-final serial verbs in a Gojri sentence has an 'invariant' form and always take serial verb inflection that I represent as SVI in gloss.2Another crucial point that needs to be clarified here is that traditional grammarians consider these forms as 'conjunctive particles' in many South Asian languages. This is not true at least in Gojri. Like many other languages of the regions, Gojri has an identical conjunctive particle, ge, that serves the required meanings. Compare the following examples.

5. a. kaloo-ne seb kut@ri khaayo kaloo-ERG apple-NOM cut-SVI eat-PF.M 'Kaloo cut the apple (in pieces) and ate it.

b. kaloo-ne seb kut@r-ge khaayo kaloo-ERG apple-NOM cut-CNP eat-PF.M 'Having cut the apple, Kaloo ate it.

Following the definition of the SVC proposed by Collins (1997:462) that says 'a serial verb construction is a succession of verbs and their complements (if any) with one subject and one tense value that are not separated by any overt marker of coordination or subordination', I claim that SV phenomenon does exist in Gojri. The above given examples satisfy the basic criteria set for these constructions. In other words, the Gojri serial verbs show one tense value that is marked on the last verb. Moreover, they do not allow any marker of coordination or subordination in the structure. It means that Gojri is distinct from other languages such as Urdu and Punjabi for displaying such constructions.

II. Serial Verbs vs Complex Predicates

Unfortunately, there has been no agreed definition available for serial verbs constructions. The variations which are cross-linguistically seen are the basic reason for not having a well unified definition. Different arguments have been presented by linguists in support of their assumptions. For example, Bhatia (1993) considers a compound verb as a serial verb in Punjabi. Often these terms (serial verbs, compound verbs, complex predicates) seem to be used interchangeably. Before proceeding further, it will be helpful to make a distinction between the complex predicates and serial verbs in Gojri. Thereafter, I will discuss serial verbs in detail to prove their individual identity and status. See the following examples for illustration:

6. a. Serial verb construction kaloo-ne seb chillii khayo kaloo-ERG apple-NOM peel.SVI eat-PF.M 'Kaloo peeled the apple and ate it.'

b. Complex predicate Kaloo-ne seb chil diyo kaloo-ERG apple-NOM peel give-PF.M 'Kaloo peeled the apple (for someone else).'

The above examples illustrate the difference between the serial verb construction and complex predicates in Gojri. In (6a), the Gojri SVC indicates that there are two sub-events involved in the structure to describe a full event, and both the sub-event were described by two separate serial verbs which come together in a sequence. It means that two different events have their own individuality in the course of action. However, (6b) describes just one action because the second verb in the sequence is a light verb that does not express its full meaning but contributes some aspectual meanings of 'completiveness' and beneficiary meaning to the meanings of V1 in the sentence. It simply indicates that the agent performed the action for someone else. Butt (1995, 1997) and Akhtar (2000) argue that light verbs in complex predicates are the bleached forms of verbs and therefore lose some of their semantic content. However, they can be used as main verbs with their usual lexical meanings.

The only preferred position for the main verb in complex predicates is V1. In the Gojri complex predicates, it is the V2 that generally shows agreement and the main verb V1 appears either in root or infinitive form. I will discuss this in detail in the later sections. On the other hand, in the Gojri SVCs, the non-final serial verbs display invariant frozen form that does not have any concern with the tense; however like the complex predicates, it is the last verb that agrees with the highest nominative argument in gender and number.

III. Salient Features of Serial Verbs

Durie (1993) argues that all co-occurring verbs in serial constructions bear the tense and agreement morphology. On the other hand, Muysken and Veenstra (1995) have summed up the characteristics of serial verb constructions in the following way:

i. they should have only one subject

ii. they should have at most one expressed direct object

iii. there should be one specification for tense/aspect either described by the first verb, second verb or the last verb in the sequence

iv. only one possible negator

v. no intervening coordinating conjunction or subordinating conjunction

vi. no intervening pause is possible

A close analysis of Gojri serial verb constructions reveals that Gojri is not an exception in this regard. This proves that there is a general phenomenon that exists cross- linguistically in serial verb formations. However, some variations regarding the placement of negation and adverbs in structures can be marked from language to language. The following sections will show different features of Gojri serial verb constructions.

III.1 Agreement

The final verb in Gojri always agrees with the nominative case (Sharma, 1982). Serial verb constructions in Gojri follow the general pattern of the language. So far as tense/aspect is concerned, it is always the last verb in the construction that agrees with the highest nominative argument in the structure. In the case of gender, it is quite distinct in Gojri serial verb constructions that the verb preceding the final verb bears serial verb inflection and it is again the final serial verb that agrees with the argument in gender. Similarly, it is also a distinct feature of the final serial verb in the structure that it shows agreement in number as the non-final verbs have frozen forms. Consider the following examples:

7. a. g@dria-ne sentre chaaii s@Te girl.PL.F-ERG orange.PL.M-NOM lift.SVI throw PF.PL.M 'Girls picked up the oranges and threw them away.'

b. g@dre-ne k@tab chaaii s@Tii boy.SG.M-ERG book. SG.F-NOM lift.SVI throw-PF. SG.F 'The boy picked up the book and threw it away.'

The final verb in Gojri never agrees with any other case except the nominative one. Therefore the nominative case must possess the same number and gender agreement. Example (7a) shows that the final serial verb in the structure agrees with the highest nominative sentre 'oranges' that is plural in number and masculine in gender. The subject g@dria 'girls' does not agree with the final verb because of gender marking difference. Similarly, k@tab 'book', the object of (7b) is singular in number and the verb that agrees with it also exhibits the same number and gender. Moreover, the verb and the object bear feminine inflections.

III.2 Case Marking

In Gojri, the final serial verb shows its ability to case mark the subject. The nominative and ergative case marking on the subject are sensitive to this distinction in the past/perfective form of transitive verbs. In all other forms, the subject invariably bears the nominative case. For instance, the serial verbs which are transitive always require the subject to be in the ergative case. In contrast to this, intransitive verbs, which generally appear in the final position of such constructions, are only compatible with subjects bearing nominative case. The following examples illustrate this:

8. a. us-ne ka k@ppii b@dyo s/he-ERG grass.3.SG.M-NOM cut.SVI tie-PF 'S/he cut the grass and tied it up.'

b. wa ka k@ppii geii she-NOM grass.3.SG.M-NOM cut.SVI go-PF 'She cut the grass and went away.'

In (8a), when the serial construction involves the transitive verb b@d 'tie' as the final serial verb, the subject obligatorily requires the ergative case i.e. -ne. Contrary to this (8b), the final serial verb gaa 'go' is intransitive, which requires that the subject to be in the nominative case. These examples demonstrate clearly that there is a strict correlation between the class of verbs and the case marking on the subject in the perfective form. The violation of this correlation results in the ungrammaticality as shown in the following sentences:

9. a. *wo ka k@ppii b@dyo he.M-NOM grass.3.SG.M.-NOM cut tie-PF 'He cut the grass and tied it up.'

b. *wa-ne ka k@ppii geii she.F-ERG grass.3.SG.M.-NOM cut go-PF 'She cut the grass and went away.'

The above ungrammatical structures are due to the violation of the co-occurrence parameter in serial verb construction in Gojri; (9a) is not grammatical, because the nominative subject wo 'he' and the transitive serial verb b@d 'tie' are not compatible; therefore, the possibility of case marking is ruled out. Similarly, (9b) is an ill-formed serial structure, because the final intransitive serial verb jaa 'go' appears with the ergative subject wa 'she', which violates the co-occurrence restriction.

III.3 Tense/Aspect

Durie (1993) argues that all the co-occurring verbs in a serial verb construction, bear tense/aspect and agreement morphology. However, the case is slightly interesting in Gojri. It is quite important to note that the serial verb(s) preceding the final verb in Gojri serial formation always take(s) a frozen past form of the verb, and the final verb agrees with the highest nominative argument of the sentence. As mentioned above, these non-final verbs in serial structures have nothing to do with agreement. It is only the final verb that has to reflect all this. So, the final verb bears the agreement morphology. Interestingly, a complex predicate also shares a single marking of tense and aspect, as the first verb (main verb) in the structure appears in its stem form, which is different from the frozen forms of the serial verbs. See the following examples for illustration:

10. a. kiren-ne koof chaaii s@Tyo kiren-ERG cup.M-NOM lift.SVI throw-PF.M 'Kiren picked up the cup and threw it away.'

b. kaloo-ne k@tab chaaii s@Tii kaloo-ERG book.F-NOM lift.SVI throw-PF.F 'Kaloo picked up the book and threw it away.'

11. a. kaloo k@tab le geyo kaloo.M-NOM book.F-NOM take go-PF.M 'Kaloo took the book away.'

b. kiren k@tab le geii kiren.F-NOM book. F-NOM take go-PF.F 'Kiren picked up the book away.'

The above examples illustrate the point very clearly. Example (10) shows that the serial verb preceding the final verb does not play any role in tense/ aspect agreement in any such formation. Similarly, (11) exhibits the same phenomenon for the aspectual complex predicates in Gojri. However, contrary to the serial verb constructions, the main verb in complex predicates either appears in the root or infinitive form or does not bear any inflection.

However, serial constructions and complex predicates differ from each other in many other ways. One of the most important differences is with respect to their arguments. In an aspectual complex predicate, the Theme argument is generally the direct object in transitive constructions and the subject in intransitive ones. Akhtar (2000) argues that an essential property of an aspectual complex predicate is that it 'concentrates' on the new location, physical or abstract, of the theme. In contrast, in a serial construction, a new location of the object is not necessarily implied. However, sometimes ambiguity arises because the clause can either be interpreted as involving the external argument undergoing a change of place or the internal argument undergoing some change of place or state. This point is illustrated in the sentences given in (12):

12. a. wa caa piii geii she-NOM tea.3 .SG.F.NOM drink. SVI go-PF.F 'She took tea and went away'.

b. wa TuGRo pakaii geii she-NOM food.3.SG.M-NOM cook.SVI go-PF.F 'She cooked food and went away.'

c. wa TuGRo paka geii

she-NOM food.3.SG.M-NOM cook go-PF.F 'She cooked food.'

Taken as a serial construction, (12a) essentially means that the external argument has undergone a change of place. There involves two actions in the sentence: drinking and going away. Similarly, (12b) describes two actions: cooking and going away, when taken as a serial construction. If the same sentence is converted into an aspectual complex, as shown in (12c), the main verb takes the root form and the structure describes that the subject finished her job of cooking food.

IV. Argument Sharing

According to Baker (1989:516) and Collins (1993:93), it is one of the major characteristics of serial verb constructions that the serial verbs share internal as well as external arguments. This phenomenon is commonly known as Argument Sharing in syntax. Collins argues that verbs (V1 and V2) must share an internal argument in serial verb construction. He quotes the following examples from the Eve language to illustrate different patterns of argument sharing:

13. a. me a nu u (Collins 1993:34) I cooked thing ate 'I cooked something and ate it.'

b. kofi ts ati-e fo yao kofi took stick-def hit yao 'Kofi took the stick and hit Yao with it.'

Being an SVO language, it seems that V2 in serial verb construction misses an object in Eve. From the English gloss, it is clear that V2 has an internal argument that it shares with V1. Collins argues that in (13a), nu 'thing' is an obvious object that is shared by both the serial verbs. The instrument of V2 in (13b) is the object of V1, whereas the theme of V2 in (13c) is considered as the direct object of V1. Consider the following examples from Gojri:

14. a. kiren-ne looter pakaii wertyo kiren-ERG curry-NOM cook.SVI serve-PF 'Kiren cooked the curry and served it.'

b. kaloo-ne kiren-na soTii chaaii kuTTyo kaloo-ERG kiren-DAT stick pick.SVI beat-PF 'Kaloo picked up a stick and beat Kiren with it.'

c. kaloo-ne piilo m@llii maaryo kaloo-ERG ant-NOM grind.SVI kill-PF.M 'Kaloo killed the ant by grinding it.'

Unlike Eve, the Gojri example (14a) illustrates that the internal argument is overtly shared by both the serial verbs as it is not sandwiched between them. However, the instrument of V2 in (14b) is considered as the direct object of V1 which is a parallel phenomenon that exists in Eve. It refers to the fact that there exist some similarities as well as differences in SVCs across the languages.

V. Insertion of Conjunction

Johnson (2002:40) argues that in a serial verb construction, there should be no intervening conjunction either subordinating or coordinating between the verbs. It may be concluded from this argument that any other grammatical category can intervene in such formations. Finney (2004) quotes an example from Krio serial verb constructions to illustrate the fact:3

15. a tek nef kut di bred I take knife cut the bread 'I cut the bread with knife.'

The above example from Krio shows that the object in serial verb constructions can be placed in between the two serial verbs. This phenomenon is also very common in many other languages which have been studied for their serial verb constructions. Gojri is distinct from these languages for the reason that it never allows objects to intervene in such formations. Consider the following examples:

16. a. us-ne k@tab caaii p@Rii mukkaii he-ERG book-NOM pick.SVI read.SVI finish-PF 'He finished reading the book in one go.'

b. *us-ne caaii k@tab p@Rii mukkaii he-ERG pick.SVI book-NOM read.SVI finish-PF 'He finished reading the book in one go.'

c. *us-ne caaii p@Rii k@tab mukkaii he-ERG pick.SVI read.SVI book-NOM finish-PF'He finished reading the book in one go.'

Though three serial verbs are involved in the above example, the preceding object cannot intervene at any position in serial verb formation. Example (1 6a) can simply be explained that the subject of the sentence first picked up the book then he started reading it and finally left the book when he finished the whole version. Examples (1 6b) and (1 6c) are ungrammatical because an object cannot be inserted in the serial verb combinations.

VI. Are SVCs Embedding Structures? Quite interestingly, neither SVCs nor complex predicate structures permit any embedding structures. They are mono-clausal in nature and share a single tense /aspect. Gojri behaves like Korean in this regard. Lee (1992), Suh (2000) and Choi (2003) report the same phenomenon for Korean V-V formations. Choi (2003) quotes the following example for illustration:

17. a. amwudo [sakwa-lul an kkaka(*-ess)] mek-ess-ta nobody [apple-ACC NEG peel(*-PST)] eat-PST 'Nobody peeled an apple and ate it.'

b. *Chelswu-nun [amwudo sakwa-lul kkaka-ess-ta-ko] an mit-ess-ta Chelswu-TOP [nobody apple-ACC peel-PST.DC.KO]NEG believe PST-DC 'Chelswu believed that nobody peeled an apple.'

Note that NPIs amwudo 'nobody' and na, the negative marker collectively mean 'nobody' and are yoked together in one clause. If they are separated across the clausal boundaries, the sentences become ungrammatical as shown in (17b). Choi argues that the clause-boundedness of NPIs in Korean serial V-V structures confirm that the above example is not bi-clausal. Consider the following examples from Gojri:

18. a. kaloo-ne ka k@ppii b@dyo kaloo-ERG grass-NOM cut-PF tie-PF.M 'Kaloo cut the grass and tied it up.'

b. kisi-ne-vi ka ni k@ppii b@dyo nobody-ERG-EMPH grass-NOM NEG cut.SVI tie-PF.M 'Nobody cut the grass and tied it up.'

c. kisi-ne-vi ka k@ppii [ni b@dyo] nobody-ERG-EMPH grass-NOM cut.SVI [NEG tie-PF] 'After being cut, nobody tied the grass up.'

It is clear from the above examples that SVCs in Gojri also display mono-clausal nature. These constructions cannot be split in different clauses in any way. These examples also show that the negation marker has its variant scope in Gojri SVCs structures. If the negation marker precedes the first serial verb in Gojri, it extends its scope to both the serial verbs as shown in (1 8b). It means nobody either cut grass or tied it up. However, as shown in (1 8c), if the negation marker follows the second serial verb, the scope is limited to the second verb only. Choi (2003) presents Korean SVCs examples, which illustrate that either the negation marker or the adverb can only be placed before the first serial verb, else it would generate ambiguity in the meanings. Gojri has an edge over Korean in this regard, because these types of ambiguities can be easily resolved by variant positions of negation marker or adverbs. Consider the following Korean examples by Choi (2003):

19. a. Chelswu-ka sakwa-lul ppali kkaka (*ppali) mek-ess-ta Chelswu-NOM apple-ACC quickly peel (quickly) eat-PST-DC 'Chelswu quickly [peeled the apple and ate it].' or 'Chelswu quickly [peeled the apple] and ate it.'

b. Chelswu-ka sakwa-lul an kkaka (*an) mek-ess-ta Chelswu-NOM apple-ACC NEG peel (NEG) eat-PST-DC 'Chelswu [did not peel the apple and eat it].' or 'Chelswu [did eat the apple but did not peel it.'

Similarly, in Gojri the serial verbs may be separated from each other by an adverb without hampering the mono-clausality of the structure. Consider the following examples:

20. a. kaloo-ne ka k@ppii b@dyo kaloo-ERG grass-NOM cut. SVI tie-PF.M 'Kaloo cut the grass and tied it up.'

b. kaloo-ne ka tawli tawli k@ppii b@dyo kaloo-ERG grass-NOM quickly cut.SVI tie-PF.M 'Kaloo cut and tied the grass quickly.'

c. kaloo-ne ka k@ppii tawli tawli b@dyo kaloo-ERG grass-NOM cut.SVI quickly tie-PF.M 'Kaloo cut the grass and tied it quickly.'

Unlike the negation marker, if the adverb comes before the first verb, it yields two meanings. It simply means that the scope of adverb can be extended to either both the serial verbs or just the first verb only. So (20b) can be interpreted as either Kaloo may have carried out both actions quickly or he may have only cut the grass quickly and later tied it up slowly. Contrary to this, if the adverb precedes the second verb, its scope is limited to the second verb only. Interestingly, if either the negation marker or the adverb is placed after the second verb, it will result in the ungrammaticality of the sentence. Consider the examples:

21. a. *kaloo-ne ka k@ppii b@dyo tawli tawli kaloo-ERG grass-NOM cut. SVI tie-PF.M quickly 'Kaloo cut the grass and tied it quickly.'

b. *kisi-ne-vi ka k@ppii b@dyo ni@/ni nobody-ERG-EMPH grass-NOM cut.SVI tie-PF.M NEG 'Nobody cut the grass and tied it up.'

Contrary to SVCs, complex predicates in Gojri do not allow any intervention in V1V2 structure. If some other item is inserted in between them, the structure will be ungrammatical. See the following examples:

22. a. kaloo-ne ka k@p diyo kaloo-ERG grass-NOM cut give-PF.M 'Kaloo cut the grass (for someone else).'

b. kaloo-ne ka ni k@p diyo kaloo-ERG grass-NOM NEG cut give-PF.M 'Kaloo did not cut the grass.'

c. *kaloo-ne ka k@p ni@ diyo kaloo-ERG grass-NOM cut NEG give-PF.M 'Kaloo did not cut the grass.'

23. a. kaloo-ne ka tawli tawli k@p diyo kaloo-ERG grass-NOM quickly cut give-PF.M 'Kaloo cut the grass quickly.'

b. *kaloo-ne ka k@p tawli tawli diyo kaloo-ERG grass-NOM cut quickly give-PF.M 'Kaloo cut the grass quickly.'

Unlike Korean nun 'also', and Hindi/Urdu bhii 'also', Gojri does not allow emphatic marker vi/bi 'also' to intercept the serial verb formation or complex predicate structure. Any attempt in this regard will not be a valid structure in Gojri. Consider the following examples:

kaloo-ERG grass-NOM cut. SVI tie-PF.M 'Kaloo cut the grass and tied it up.'

b. kaloo-ne ka bi/vi k@ppii b@dyo kaloo-ERG grass-NOM also cut.SVI tie-PF.M 'Kaloo also cut and tied the grass.'

c. *kaloo-ne ka k@ppii vi/bi b@dyo kaloo-ERG grass-NOM cut.SVI also tie-PF.M 'Kaloo also cut the grass and tied it.'

25. a. kaloo-ne ka k@p diyo kaloo-ERG grass-NOM cut give-PF.M 'Kaloo cut the grass (for someone else).'

b. kaloo-ne ka vi/bi k@p diyo kaloo-ERG grass-NOM also cut give-PF.M 'Kaloo cut the grass also(for someone else).'

c. kaloo-ne ka k@p vi/bi diyo kaloo-ERG grass-NOM cut also give-PF.M 'Kaloo cut the grass also (for someone else).'

Examples (24) and (25) illustrate that any attempt to insert emphatic marker in either serial verb construction or complex predicate VV formation results in ill-formed structures. The emphatic marker can only precede the first verb in both the sequences. In serial verb constructions, it emphasises both the serial verbs. However, in complex predicate its scope is limited to the first verb only. It is natural because the first verb in complex predicate describes the event while the second verb adds some aspectual or semantic meanings to the first verb. So it can be concluded from the above examples that emphatic marker behaves differently in Gojri from those of Korean, Urdu or Hindi. It is also clear from the above discussion that Gojri displays a different syntactic behaviour of emphatic marker from the negation marker or adverbs.

VII. Conclusion

Summing up the whole discussion, it can be concluded that the Gojri serial verb construction has its own individuality in different complex verb formations. They are different from complex predicates as they describe two separate events. They also show other grammatical differences which make them distinct from complex predicates. Contrary to Jayaseelan (2004) and Lee (1992), it is confirmed from the Gojri data that serial verb constructions are not a type of complex predicates. Argument structure is another test that draws a line between different structures. Butt (1995) and Pandharipande (1989, 1990) point out that the distinction between serial and complex predicate constructions is not very easy to draw. However, it is clear from the discussion that serialisation and complex predicate formation belong to two different domains.


1 Yoruba is the official Language of Nigeria and mainly spoken in West African territories while Khmer is the second most widely spoken Austro-Asiatic language and is the official language of Cambodia.

2 Jayaseelan has used the term 'frozen' form for these non-final serial verbs for Malayalam.

3 Krio is a lingua franca language and is widely spoken in Sierra Leone. The vocabulary of Krio is primarily derived from English Language while its sound system, grammar and sentence structure is heavily influenced by African languages.


Akhtar, R.N. (2000). Aspectual Complex Predicates in Punjabi. Ph.D Dissertation.University of Essex, Colchester.

Baker, M.(1989). Object Sharing and projection in serial verb constructions. Linguistic Enquiry. (20): Pp. 513-553

Baker, M. (2002). On zero agreement and polysynthesis, ms. Rutgers University. Bhatia, T. K (1993). 'Punjabi: A cognitive-descriptive grammar'. Routledge. London and New York.

Butt, M. (1995). The Structure of Complex Predicates in Urdu. Stanford: Centre for the Study of Language and Information.

Butt, M. (1997). Complex Predicates in Urdu, In Alsina, A., Bresnan, J. and Sells, P. (Eds). Complex Predicates, Stanford. CSLI Publications.

Choi, S. (2003). Serial verbs and adjunction, the Ist CamLing Proceedings. Cambridge University Press.

Collins, C. (1993). Topics in Ewe Syntax. Ph.D. Dissertation. MIT.Cambridge, MA. Collins, C. 1997. Argument sharing in serial verb construction. Linguistic Inquiry (28): Pp. 461-497

Durie, M. (1997). "Grammatical Structures in Verb Serialisation." In Complex Predicates. (Ed). Alex Alsina, Joan Bresnan and Peter Sells. Pp. 289-354. Stanford: CSLI Publications.

Finney, M. (2004). "Substratal Influence on the Morphosyntactic Properties on Krio." Linguistic Discovery. 2 (2): Pp. 58-8 1.

Johnson, S. (2002). Revisiting the Structure of Serial Verb-construction. Michigan State University.

Muysken, P. & Veenstra, T. (1995). "Serial Verbs." In Muysken, P. & Smith, N. (Eds). Pidgin and Creoles: An Introduction. Philadelphia: Benjamins.

Lee, S. (1992). The Syntax and Semantics of Serial Verb construction. Ph.D.Thesis. University of Washington, Seattle.

Pandharipande, R. (1989). 'Split Properties of Compound Verbs in Marathi'. Paper presented at the Eleventh South Asian Languages Analysis Roundable, University of Wisconsin, Madison.

Pandharipande, R. (1990). 'Serial Verb Construction in Marathi'. In Zwicky, A. & Joseph, B. (Eds.) "When Verbs Collide: Papers from the 1990 Ohio State Mini-conference on Serial Verbs." Pp. 178-199.

Sharma, J. C. 1982. Gojri Grammar. Maysore: Central Institute of Indian Languages. Suh, Y. (2000). A Study of Auxiliary Construction and Verb Serialisation in Korean. PhD Thesis, University of Washington, Seattle.

Nadeem Haider Bukhari Department of English, University of AJ and K Muzaffarabad, AK
COPYRIGHT 2009 Asianet-Pakistan
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 2009 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Publication:The Journal of Humanities and Social Sciences
Article Type:Report
Date:Dec 31, 2009
Previous Article:Population Projections of Pakistan Using Traditional and Time Series Models.
Next Article:Challenges of Afghan Refugee Repatriation (2002-2005): Success or Failure?

Terms of use | Privacy policy | Copyright © 2021 Farlex, Inc. | Feedback | For webmasters