Are Vector Verbs Eternal?

Recently it has been asserted (Butt 2003, Butt and Geuder 2003) that the vector (aka "light") verb constituents of the compound verbs found in most South Asian and Central Asian languages differ in fundamental ways from auxiliaries of tense and mood. While auxiliaries evolve over longer or shorter periods of time from full verbs in ways deemed by students of grammaticalization to be universally valid paths; vector verbs do not evolve. Rather, their semantics are claimed to be sufficiently abstract to allow them to function immediately as explicators in compound verbs without any of the semantic bleaching or phonological ablation characteristic of the diachronic trajectory from full verb to auxiliary to affix that has been posited by “gram-maticalizationists” for the development of verbal morphology. Thus, vector verbs are "form-identical" with their full verb counterparts while tense auxiliaries are not. And their meanings as vectors are not a diminished subset of their meanings as full verbs (provided the latter are properly "underspecified").

The striking phonological changes seen in the Shina vector verb y- GO which (under the influence of western Tibetan) has been recycled as a past tense auxiliary at first seem to support the "form-identical" claim. In the second part of our paper, however, we show that in fact the phonological argument cuts both ways. The vector verbs viD- LEAVE and koL- HOLD of Tamil have undergone significant phonological attrition while the vector verbs of Indic have largely escaped. On the other hand, the tensual auxiliaries hai and thaa ( BE and WAS ) of Hindi-Urdu are form-identical with their main verb counterparts while the corresponding auxiliaries (aa(h))e and (h(o))taa of Marathi are not. In Japanese the vector shimau PUT AWAY shows a high degree of phonological attrition. We contend that the tendency over time for verb forms to undergo (or not to undergo) phonological attrition in a particular language or language family is independent of whether they function as vectors or as tense auxiliaries.

In the third part of the paper we examine the issue of semantic bleaching. While much of the semantics of Marathi's main verb de-GIVE remains palpable in its functions as a vector verb, the same cannot be said of the corresponding vector ver- GIVE in Turkish. (From http://www.turkishlanguage.co.uk/auxiliaryverbs.htm):

 

(1) “The verb vermek ‘to give’ can also be added to the subjunctive verb stem and in this case it gives a sense of urgency and speed of action to the main verb. In this case vermek loses its meaning – ‘to give’.

 

koşa-ver-iniz! ‘Hurry up and run!’ içkis-in-i içe-ver-di ‘He gulped his drink down.’ “

 

Hindi-Urdu's vector de- GIVE has a broader range of usage than its Marathi counterpart. Some of these are at variance with those of Marathi (and are not predictable from the semantics of the main verb 'give'):

 

(2) … na jaane kidhar cal detii hai, jaise kisii kaam mE jii hii nahII lagataa

not know where move GIVE is as.if any task in soul Emp not engages

'No knowing where she goes off to, as if nothing can hold her interest.' (godān 263, sent 35)

 

The Japanese vector shimau PUT AWAY also exhibits a profound degree of semantic bleaching.

We conclude that any differences in the speed of phonological attrition or semantic bleaching that may seem to discriminate vector verbs from auxiliaries of tense and mood must be ascribed to the greater text frequency of the latter and not to some fundamental or categorical distinction.

 

 

Butt, Miriam. 2003. The Light Verb Jungle. In G. Aygen, C. Bowern, and C. Quinn (eds.) Harvard Working Papers in Linguistics, Volume 9: Papers from the GSAS/Dudley House Workshop on Light Verbs. Pp. 1-49.

Butt, Miriam, and Wilhelm Geuder. 2003. Light Verbs in Urdu and Grammaticalisation.  In  Regine Eckardt, Klaus v. Heusinger & Christoph Schwarze (eds): Words in Time. Diachronic Semantics from Different Points of View. Berlin: Mouton de Gruyter  Pp. 295-350.