Classification of Verbal compounds in Marathi: Some Issues

.........................................................................................................................

आभ्यंतर (Aabhyantar)      SCONLI-12  विशेषांक         ISSN : 2348-7771

.........................................................................................................................

4. Classification of Verbal compounds in Marathi: Some Issues

Aaditya Kulkarni

M.Phil. Linguistics

1. Introduction

Marathi is an Indo-Aryan language of the southern sub-branch. It is mainly spoken in the states of Maharashtra and Goa by around 9 million people. Like many south Asian and Indo-Aryan languages, a lot of work has been done on this language. Many of the aspects of this language though, are still in need of a better explanation.

Verbal compounds in Marathi have been dealt with by almost all the grammarians of the language. Verbal compounds mainly have a V-V structure where different suffixes get attached to V1 for making these complex structures. However, not all V-V sequences can qualify as verbal compounds and various grammarians and linguists have dealt with this topic in different ways. The purpose of the present work however is not to recount the earlier works on the topic concerned but to deal with the problems which might emerge out of the most recent classification and description of these compounds.

2. Literature Review

Verbs, their conjugations and their morphology have been a topic of interest since the time Marathi grammars started coming into picture. The earliest of the grammars (like the ones by Tarkhadkar (1846 edition), Stevenson (1854 edition) to mention a few) do deal with verbal compounds with structure V-V but call them compound verbs. For them, almost all kinds of V-V sequences were compound verbs and it was only in the grammars of the later stages that the criteria for the identification of the same truly narrowed down. The understanding of this phenomenon grew steadily over the years especially with the contribution of linguists like Raeside.

Pardeshi (2001) in his seminal paper on compound verbs and their definitional criteria made distinction between some verbal compounds (with structure ‘V-un V’). His work not only takes a good review of the literature available on the topic, but also systematically distinguished 4 types of verbal compounds on the basis of their semanticities. However, the criteria applied by grammarians or linguists till then were either formal or semantic. The structure of Marathi verbs came into proper focus only with the works of Deoskar (2006) and Ozarkar (2014).

3. Aims and Objectives

The present study will try to focus on the classification of verbal compounds done by Pardeshi (2001) will consider it as a base. The present work will:

a) Try to understand the issues which emerge out of Pardeshi’s (2001) classification of verbal compounds.

b) Try to establish the need for structural criteria for the classification of verbal compounds.

4. Methodology

For the purpose of the present work, we will try to

a) Understand the phenomena of verbal compounding

b) Collect data by the methods of elicitation and casual observation

c) Provide analysis based on the data

5. Classification of Verbal Compounds and Issues

There are a variety of verbal compounds found in Marathi and Pardeshi (2001), while considering verbal compounds with structure V-un V (where ‘-un’ is a conjunctive participle) gives out as many as 4 categories. The types given by him are as follows:

a) Serial verbs ( both verbs semantic heads)

b) Compound verbs (v1 semantic head, v2 modifier)

c) Adverbial sequences (v1 modifier, v2 semantic head)

d) Idiomatic sequences (neither of the verbs a semantic head)

All of these expressions have the same pattern ‘V-un V’ and hence are homophonous. Also, all the criteria applied here for this classification are either formal or semantic and (syntactic) structure of these verbal complexes doesn’t play any role in this. Now, in the next sections, we will try to bring out some issues with the classification of these verbal compounds.

5.1. Serial vs. Compound verbs

Now, though distinctions have been made, some cases of compounds can pass off as both serial verb (SV) constructions as well as compound verb (CV) constructions.

1) kaam kar-un dakhav

work do-CP show.IMP

Complete the work and show it (OR) it/Complete the work and prove it (that you can)

Here, the sentence would mean both ‘do the work and show it (to the concerned person- SV construction)’ and ‘do the work and prove it (that you are capable of doing it- CV with V2 as modifier)’. So, it is with the help of structure (assuming that the context is unavailable) that this can be disambiguated.

Deoskar (2006, 10) proposes a structure for V-V compound where one verb is a light verb and the other functions as a main verb. The following figure illustrates the structure of that verbal compound.

This is the structure proposed by Deoskar (2006) can explain the structure of Marathi compound verbs. As the serial verbs don’t have a light verb in their verbal compound, the structure becomes the distinguishing criteria for the compound verbs.

5.2. Compound verbs vs. Adverbial sequences

Apart from the SV vs. CV distinction, adverbials posit another problem. They are somewhat a mirror image of CVs for they have V1 as modifier and V2 as main verb. Just as we can take the vector verb out of the CV without changing its truth predication (though the meaning expressed by vector is lost), we can remove the modifier v1 from adverbial sequence without altering its truth predication at the cost of the loss of the additional meaning expressed by the modifier. One difference however is that unlike the semantically weaker, grammaticalized light verb a CV has, the modifier of the adverbial isn’t grammaticalized and retains its lexical meaning.

An example of the adverbial sequences is as follows:

2) mi phaT-i-tun tsor-un paahi-l-a

I gap-OBL-through steal-CP see-PF-N

I peeped through the gap stealthily. (Pardeshi 2001)

Now, the structural representations have to be distinct for CV, SV and Adv. seq. as well. As there are cases of verbal compounds which can pass off as both, a CV and an adverbial sequence, a successful explanation of this construction would also need another distinct structural representation of the V-un V sequence:

3) kʰeks-un bolɳe

bellow-CP speak

4) kʰəɖəsa-un saŋgɳe

scold-CP tell

5) kəwʈaɭ-un d ʰərɳe

hug-CP hold

As these expressions can function as CVs or adverbial sequences, it becomes important to make a structural distinction here as well and hence a structural criterion is needed for this purpose.

5.3. Idiomatic sequences

The last of the verbal compounds discussed by Pardeshi are ‘idiomatic sequences.’ They are homophonous with the other compounds as well but they differ from them significantly as the meaning expressed by them is not compositional at all. How provide an account of these becomes another problem. Perhaps, they can be explained successfully if we could show the cognitive processes they are likely to have undergone to get an idiomatic sense.

Some examples of such expressions are as follows:

6) t uʈ-un pəɖɳe

break-CP fall

to attack/begin violently

7) t sal-un d zaɳe

walk-CP go

to attack

Though enough information is not available at this stage, with sufficient data, we can attribute the idiomaticity of such expressions which are found in various languages to the cognitive processes that take place and often result in idiomatic meanings.

5.4. Cline of Grammaticalization

Even if a large list of verbal complexes is examined, classifying these into above mentioned clear cut categories becomes very difficult if not impossible. Due to this, Pardeshi presents a cline of grammaticalization of V2 which, in his opinion makes this blurred distinction easier to present and understand:

Serial Verb Sequences > Adverbial Sequences > Vector Sequences > Aux. Sequences

This cline has a verbal complex with two semantic heads at the left extreme and the right side of the cline is occupied by verbal complexes that involve semantically less prominent V2. Inclusion of adverbial sequences in between these sequences looks problematic though. Compound verb sequences and the adverbial sequences present a sharp contrast. CVs have V1 as semantic head

6

and V2 as modifier and adverbial sequences have a prominent V2 and V1 is a mere modifier in these constructions.

8) mi phaT-i-tun tsor-un paahi-l-a

I gap-OBL-through steal-CP see-PF-N

I peeped through the gap stealthily. (Pardeshi 2001)

So, it seems that V1 is definitely semantically weaker than V2 (whether it is grammaticalized or not can be a matter of another study). Therefore, in a cline which represents grammaticalization of V2 (which in turn is a case of semantically weaker V2s); inclusion of a verbal sequence which has modifier V1 is inconsistent. If an ungrammaticalized, semantically prominent verb is considered strong verb and grammaticalized, modifier verb is considered semantically weaker, the cline would be

Serial Verbs

Strong V1-V2

> 

Adverbial Compounds weak V1 - strong V2

> 

Compound Verbs weak V1 - strong V2

> 

Aux. Sequences strong V1 – V2 (Aux)

This fluctuation of semanticities of both V1 and V2 in these constructions shows that the cline is semantically inconsistent and needs revision.

6. Summary

After examining the classification given by Pardeshi (2001) and the issues which seem to emerge out of it, we can say that:

·        All verbal compounds of structure ‘V-un V’ are homophonous

·        Same expressions can be classified under more than one categories

·        ‘Structure’ can be used as the criterion

Also, section 5.4 also comments on the cline of grammaticalization in the passing. Further work is needed especially from the perspective of syntax which will lead us to the explanation of the issues mentioned here.

7. References:

·        Deoskar, T. 2006. Marathi light verbs. Presentation at the 36th Annual Meeting of the Chicago Linguistics Society. url:http://homepages.inf.ed.ac.uk/tdeoskar/publications.html

·        Ozarkar, R. 2014. Structures Of Marathi Verbs. (unpublished doctoral dissertation) University of Mumbai.

·        Pandharipande, R. 1997. Marathi. New York: Routledge.

·        Pardeshi, P. 2001 The compound verb in Marathi: Definitional issues and criteria for identification.In ‘Kobe GengogakuRonsô (Kobe Papers in Linguistics)’3: 94–111.

·        Stevenson, Rev. J. 1854 (3rd edition). The Principles of Murathee Grammar. Bombay: Thacker & Co.

·        Tarkhadkar, D. 1846 (3rd edition). ‘Maharastra Bhasheche Vyakaran.’