Verb ‘say’ in Indic Languages: A study in Grammaticalization | Международен филологически форумСп. "Филологически форум" - хуманитарно списание за млади изследователи на Факултета по славянски филологии е вече в CEEOL!

Verb ‘say’ in Indic Languages: A study in Grammaticalization

Posted in: Библиотека, Езикови картини и тезауруси Started by

Verb ‘say’ in Indic Languages: A study in Grammaticalization

Hifzur Rahman Ansary

Jawaharlal Nehru University, New Delhi, India


The present paper proposes to examine and explore a number of cases of grammaticalization (also known as grammatization or grammaticization) of verb ‘say’ in Indic Languages (Bangla, Telugu, Kannada and Dakkhani Urdu). The verb ‘say‘ is semantically polysemous in nature in the sentence. The paper scrutinizes the verb ‘say’ in which the verb retains its natural and grammatical category but gives up its actual semantic meaning and shows a different new meaning in the contexts. In particular, the verb “say” has been found to be used as a complementizer, quotative markers, reason marker, embedded questions, introducers, purposive markers, onomatopoeic expressions, manner adverbs etc.

Key words: complementizer, desemanticization, grammaticalization, polysemous, quotative, semantic bleaching and semantic fading


DAKKHANI URDU and Telugu are largely spoken in the state of Andhra Pradesh, with the largest number of Dakkhani Urdu speakers residing in Hyderabad. Telugu speakers are the linguistic majority in the state while Dakkhani Urdu speakers constitute linguistic minority in the state. Among the districts of Andhra Pradesh, Kurnool and Guntur have a comparatively large percentage of Dakkhani Urdu speakers. Telugu is a Dravidian language native to India. It is an official language in Andhra Pradesh, Telengana and Yanam. The immediate source of Dakkhani Urdu has been a debatable issue with different scholars such as Masud H. (Masud 1969); Shakeel, A.S. (Shakeel 1975); Beg, Mirza Khaleel A. (Beg 1977); Mustafa, K.S. (Mustafa 2000), positing different sources like Punjabi, Haryanvi, Khariboli, Brajbhasha and Mewati. According to Schmidt (Schmidt 1981:4) Dakkhani Urdu is a descendent of the New Indo-Aryan dialects spoken in the Jamuna-Ganges Doab during the 13th-15th centuries, which were imported into the Deccan with the Muslim rulers who exercised control over most of that area during the latter half of the middle age.

Today, as a linguistic entity, Dakkhani Urdu has come a long way since its evolution in the 13th-14th centuries. It has undergone subtle changes in its structure as a result of coming in contact with different languages, and is marked by great diversity and heterogeneity. This has been pointed out in several studies such as Kachru’ (Kachru 1986), Mustafa’ (Mustafa 2000), Harbir’ (Harbir 1986), Subbarao and Harbir’s works (Subbarao, Harbir 1989).

The major aim of this paper is to explore the various processes of grammaticalization of verb ‘bol ke’ – ‘to say, speak’ in Dakkhani Urdu. Following are the objectives of the study –to scrutinize from the point of syntax, taking the verbs as instances of a lexical category to a functional category with effects of semantic bleaching and extension; to study and investigate the various usages of verb ‘bol ke’ in which the verb retains its natural and grammatical category but gives up its actual semantic meaning and shows a different new meaning in the contexts.

The first and second sections are focused on Dakkhani Urdu and Telugu languages. The third and fourth sections deal with the concept of grammaticalization and the various parameters of grammaticalization respectively. In the fifth section there is discussion of the grammaticalization of ‘bol ke’. The last section attempts to investigate the verb ‘bol ke’ and ‘ani’ as a complementizer and its diverse usages.


The idea (description and theory) of grammaticalization is already contained in the works of Bopp (Bopp 1816), Schlegel (Schlegel 1818), Humboldt (Humboldt 1825), Gabelentz (Gabelentz 1891) and Heine, Claudi & Hunnemeyer (Heine, B., Claudi, U.& Hunnemeyer. F 1991)  – it might even go back as far as Condillac (Condillac 1746), according to Lehmann (Lehmann 1982:1).

The term “grammaticalization” itself was apparently introduced by the French linguist Antoine Millet (Millet 1912). The probably most cited, “classical,” definition is the following one by Jerzy Kuryłowicz (Kuryłowicz 1975), where, probably for the first time, the idea of a continuum of grammaticality is contained:”Grammaticalization consists in the increase of the range of morpheme advancing from a lexical to a grammatical or from a less grammatical to a more grammatical status, e.g. from a derivative formant to an inflectional one” (Kuryłowicz [1965] 1975:52). “With the term “grammaticalization” Heine and Reh (Heine, Reh 1984:15) refer essentially to an evolution whereby linguistic unit loses its semantic complexity, pragmatic significance, syntactic freedom, and phonetic substance, respectively.”

There have been a number of alternative terms suggested by various scholars in the recent past for similar phenomenon, such as “syntacticization” (Givon 1979:208 ff), “semantic bleaching”, “semantic weakening” (Guimier 1985:155), “semantic fading” (Antilla 1989:199), “condensation” (Lehmann 1982:10–11), “reduction” (Langacker 1977:103–107), “subduction” (Guillaume 1964:73–86) etc., each of them characterizing a specific aspect of the phenomenon of grammaticalization. Most of the authors believe that grammaticalization is a diachronic process which is gradual in nature.


According to Heine and Kuteva (Heine, Kuteva2002) there are four interrelated mechanisms of grammaticalization namely:

  1. Extension, i.e. the rise of novel grammatical meanings when linguistic expressions are extended to new contexts (context-induced reinterpretation),
  2. Desemanticization (or “semantic bleaching”), i.e. loss (or generalization) in meaning content,
  3. Decategorialization, i.e. loss in morpho-syntactic properties characteristic of lexical or other less grammaticalized forms, and d. Erosion (or “phonetic reduction”), i.e. loss in phonetic substance.(Heine, Kuteva 2002:2)



Dakkhani Urdu and Telugu have a verb ‘bol ke’ and ‘ani”– ‘to say, speak’ which, in terms of morphology, behaves phonologically in the same way as any other consonant-ending verb of this type (such as kar- ‘to do’, paɽh – ‘to read’, likʰ – ‘to write’ and dekʰ – ‘to see’) would do. This paper investigates the various usages of the verb “bol ke” and “ani” in the selected South Asian Language –Dakkhani Urdu and Telugu. The samples have been taken from the informants of native speakers of Hyderabad as well as secondary sources such as Grammar books, M. Phil dissertation and PhD theses.

The verb “bol ke” and “ani” (say) have the following extended usages: Complementizer, Quotative, Adjectival Complementizer, Reason Marker, Purposive Marker. It is also used with Onomatopoeic Expressions, Constructions Expressing Desire, Intention and Thought. It is used with Question word Complementizer to express deliberateness too. The adjectival participial form of the verbs “bol ke” and “ani” has the extended usage of Introducing[2], Naming and Labeling (Subbarao et al.1989:91).


The finite Complementizer occurs only when the embedded clause is finite. Notice the following sentences where a complement sentence is introduced by “bol ke” and “ani” in Dakkhani Urdu and in Telugu.

Dakkhani Urdu

      (1) us ku    cale ʤa:o bol ke   bol  d̪yo

        he-DAT   go away   COMP  tell-Give-AGR

      ‘Tell him to go’.


(2)  O-ke         cole      ɟa:o       bole       bole  d̪a:o

      he-DEM     go      away      COMP    tell   give

        ‘Tell him to go’.

The above sentence can be said in the following way too:

      (3) O-ke             cole    ɟet̪e       bolo

            he-DEM       go    Go-INF   tell

The “bole” in Bangla functions as a complementizer can be shown if the same embedded sentences receive “ɟe” complementizer.[3]


     (4) [ a:jana  ki          pomm-ani    ]        ceppu

            he-DAT              go   COMP              tell

          ‘Tell him to go’.


   (5)           avaniɡe     hoɡu      ana      hajɭu

                     he-DAT  go       COMP      tell

Bangla has also another complementizer “ɟe” which is similar to “bole” complementizer. The following example is illustrative:

    (6) anweʃa               aʃ-b-e               bole         ami    ʃuni     ni

        Anwesha     come-FUT-AGR    COMP         I      hear    neg-PERF

        ‘I did not hear (any talk of) Anwesha coming’.

    (7) ami      ʃuni              ni                   ɟe         jʰilam        aʃbe

            I      hear-ARG   NEG-PERF  COMP    Jhilam   come-FUT-ARG

          ‘I have not heard  that Jhilam  will come’

  When a listener hears (6), he/she does not find that it is suggested that Anwesha will or not come. In case of (7), one would surely conclude that Jhilam would, in fact, come. It is as if /bole/ tends to put the truth-value of its proposition ‘within parentheses’, while /ɟe/ is somewhat more ‘factitive’.

Dakkhani Urdu

     (8) si:t̪a:      ɡã:v      ku       cale    ɡaji:     bol ke    mereku   ma:lum   naĩ:

           Sita       village DAT    went   away   COMP   I-DAT    known     NEG

          ‘I did not know that Sita had gone to the village’


      (9) [si:t̪a         u:ri    ki           weɭɭind̪  –ani    ]            na:ku       t̪elijad̪u

              Sita    village-DAT        went     COMP I-DAT     me       know-NEG

             ‘I did not know that Sita had gone to the village’


  (10)         nanaɡe   sit̪a       helliɡe              hoɡid̪d̪-ɭu          ant̪a      ɡot̪t̪-iral-illa

                 for me    Sita      village-DAT    had gone-She  COMP    know-did-NEG

                ‘I did not know that Sita had gone to the village’

            Dakkhani Urdu has another post-sentential complementizer /karke/which is syntactically similar to /bol ke/ complementizer. This is a conjunctive participle form of the verb /kar/ ‘to do’.[4]

Let us now consider some further data on Complementizers in Dakkhini Urdu, Bangla, Telugu and Kannada.

Dakkhini Urdu

        (11)  ra:m     kab      a:t̪a        ki      /*bol ke      mere  ku      kja:     ma:lum?

               Ram     when  comes     COMP/COMP        I-DAT        what   known


(12)             ra:m        kobe          aʃ-b-e               bole      ami     ki        ʤani?

                   Ram       when  come-FUT-AGR     COMP      I     what     known

                 ‘How do I know when Ram will come?’


        (13) ra:muɖu     jeppuɖu       ost̪a: -o:             na:ku      je:mi   t̪elusu?

                Ram             when        come-COMP    I-DAT    what   is known

‘How do I know when Ram will come?’


  (14)      ra:m     javaɡa    bart̪ʰane     ana       nanaɡe    heɡe    ɡot̪t̪u

              Ram     when   come-FUT      COMP      me     how   known

Dakkhini has another post sentential complementizer /kar ke/which is syntactically similar to /bol ke/ complementizer. This is a conjunctive participle form of the verb /kar/ ‘to do’.[5]

According to Harbir (Harbir 2002: 65) only the /kar ke/ complementizer is found in earlier Dakkhani Urdu texts. The existence of this post-sentential complementizer might have acted as a catalyst for the borrowing of the /bol ke/ complementizer with various functions from Telugu into Dakkhini Urdu. Telugu does not have a complementizer which corresponds translationally to /karke/ of Dakkhani Urdu.


The sentence such as (15) below from Dakkhani Urdu in which the verb of the embedded sentence is in the infinitival form and the complementizer is ‘bol ke’ or ‘kar ke’.

Dakkhani Urdu

(15) mereku      numaiʃ-ku           ʤa:na:   bolke  / karke       kʰa:iŝ       ai

       I-DAT   exhibition DAT       go-inf   COMP/COMP      desire       is

‘I have a desire to go to the book fair’.


(16)           nanaɡe      calana cit̪ra      hoɡu    beku  ant̪a   apekʃe     id̪e

                I-DAT            movie              go      need COMP   desire  have

However, Telugu uses an adjectival form of the verb say as a complementizer when the lexical head is overtly present in the matrix sentence. The following sentence is illustrative:


(17)  na:ku     sinema:     ki      weɭɭʰa:li         ani   / anina     ko:rika   unnad̪i

        I-DAT    movie     DAT   go-oblig      COMP/COMP    desire       is

‘I have a desire to go to watch the movie’

Dakkhani Urdu too, has as adjectival complementizer so which occurs postsententially.[6]


(18)           nanaɡe         balamma      barutʰane     ant̪a    viʃaja    ɡot̪t̪u

                     I                Balamma    comes         COMP  news     know

                 ‘I know the news that Ballama is coming/comes’.

Another postposed complementizer in Dakkhini Urdu is sarka: / sari:kha: / vaisa: ‘as though’, ‘it appears’. When this complementizer occurs the verb of the embedded sentence is always finite. The predicate of the matrix sentence is always a non factitive such as hona:, ma:lum hona:, dikhna:, lagna: ‘to appear, ‘karna: “to see to it’ etc. the following sentences are illustrative:                                                                               (Harbir 2002)

Dakkhhani Urdu

        (19) pa:ni:   paɽe       sarka:/vaisa:    ai

                water fall-perf COMP/COMP   is

        ‘It appears as though it is going to rain/ It has rained’.


(20)         id̪u       maɭe         barut̪t̪e      ant̪a      kanisut̪t̪a  id̪e

                It          rain          comes     COMP    appear      is

Telugu too has a complementizer /aTlu/ with the similar meaning which only when the embedded verb is finite.[7]


Both “bol ke” and “kar ke” in Dakkhani Urdu are used to express the reason function just like the complementizers “ani” in Telugu and “anta” in Kannada.

Dakkhini Urdu

        (21) a:ʤ   ɡarmi   paɽi  hai    bol ke/ kar ke      humloɡ    bahar         naĩ:    ɡja:

              today   hot      has-fallen  COMP/COMP    we        outside       neg   go

‘Because it was hot today, we didn’t go outside’.

In Bangla ‘bole‘ is used to express the reason function just like the complementizers ‘ani’ in Telugu and anta in Kannada. bole could be used as reason marker and that in all sentences, tai ‘therefore’ could replace it.

(22) a:pni       e-l-en               na    bole /tai ama:-r        kʰub         kʰoti     hoje        ɡe-l-o

        you   came-PAST-AGR  neg   COMP  I-GEN   this much     loss    happen   go-PAST-AGR

         ‘I had to bear great loss because you did not come’.

      (23) a:ɟ      ɡOrOm    poɽ-ecʰ-e             bole /tai    amra     baire      ɟai      ni

             today     hot       fall-PERF-AGR    COMP       we     outside    go  NEG.PERF

           ‘Because it was hot today, we didn’t go outside’.

The corresponding Telugu and Kannada sentences are as follows:


(24)  mi:ru       ra:le:d̪-ani,                  naku          ca:na      nuksa:n       ajind̪i

         you   did NEG come COMP    I-DAT      a lot of          loss          happened

   ‘I had to bear great loss because you did not come’.


(25)      ninu  baral-illa        ant̪a      nanu    bahaɭa      naʃta  horabekajit̪u

            You   did not come COMP     I        great        loss      had to bear

        ‘I had to bear great loss because you did not come’.


Here the verb “bol ke/kar ke” are grammaticalized (or desemanticized) and leave its own original meaning and feature and provide a new interpretation, that is, Purpose.

Dakkhani Urdu

      (26) jã:        pe     ka:m     karna:    bol ke / kar ke    a:ji:     ũ:

             here    LOC   work     do-inf COMP  COMP  come have

           ‘I have come here for work.’


    (27)   ekʰane         eʃe-cʰ-i               ka:ɟ       kor-b-o                 bole

                here     come-PERF-AGR   work   do-FUT-AGR        COMP(in order to)

   ‘I have come here for work.’

In Bangla, the complementizer is performing the PURPOSIVE FUNCTION. Another point to be mentioned here is that there is an alternate constructions in Bangla in which the postposition ‘ɟonnyo’(for) and infinitive marker ‘t̪e’ (to) can occur respectively.

    (28) ekʰane   ka:ɟer        ɟonnjo       eʃecʰi-l-am

             here    work-GEN   for        come-PAST-AGR

   ‘I came here to work’

    (29) ekʰane    ka:ɟ    kor-t̪e      eʃecʰi-l-am

           here       work   do-INF    come-PAST-AGR

             ‘I came here to work.’


 (30)   ikkaɖa     pani          cejed̪d̪a:m-ani             occina:

             here      work         do-hort  COMP           came

        ‘I have come  here for work.’


(31)    nanu    illiɡe      kelasa     maɖabeku-ant̪a          band̪id̪d̪ene

           I          here      work     in order to    COMP        come-PERF


In Dakkhani Urdu “bol ke/kar ke” occur as a quotative marker when the verb of the matrix sentence is bolna: ‘to speak’, kahna: ‘to tell’; etc. In Telugu too ani occurs with verbs such as an ‘say’/ cepp ‘tell’ etc.

Dakkhani Urdu

      (32) mon       a:t̪a: ũ:        bol ke  / kar ke   bola:   pan   naĩ:   a:ja:

              Mohan will come     COMP / COMP  said   but   neg came

‘Mohan said that he will come but he did not’.

In a paper, Kachru (Kachru 1978; 1979) had already noted that an important function of Bangla bole, Oriya boli, Dakkhini bolke, Marathi mha Nuun, Sinhalese kiya, Tamil enru, Kannada endu and Telugu enri was that these could be used as ‘quotative’ markers in the way iti was used in Sanskrit. One important difference between bole in Bangla and iti in Sanskrit was that the former could not occur freely with one of the verbs of perception, i.e., dekʰ ‘to see’, although it could occur with ʃun ‘to hear’, ɟan ‘to know’ or bʰab ‘to think’. That is why the following sentence is ungrammartical with dekh: (U.N. Singh 1980:192).

    (33) a:pni   baɽi  ɟaccʰen   bole         {*dekʰ-l-am}

                                                            { bʰəb-l-am}

                                                            { ʃun-l-am}

                                                            { ɟan-l-am }

         you   house go-PROG  COMP

see/think/hear/know -PAST-AGR

         ‘I thought/heard/came to know/*saw/ (quotative)

you  were going to house’

In Dakkhini ‘bol ke’ and ‘kar  ke’ occur as a quotative marker when the verb of the matrix sentence is bolna: ‘to speak’, kahna: ‘to tell’; etc. In Telugu and Kannada too ‘ani’ and ‘anta’ occur with verbs such as an ‘say’/ cepp ‘tell’  and hajɭu ‘say’ etc.

Dakkhani Urdu

      (34) ap    kalkatta        jare           bol ke / kar ke   suna

             you   kolkata     go+PROG  COMP/COMP   hear

‘I heard that you are going to Kolkata’.


(35)  mohan        wast̪a:n   -ani           ceppi       ka:nira:le:d̪u

          Mohan   will come  COMP      said         but did not come

‘Mohan said that he will come but he did not’.


(36)   mohana      bar-t̪ini      ant̪a     heɭid̪d̪a    ad̪re  avanu     baralilla

         Mohan    come-FUT  COMP    said         but        he       come-PAST-NEG

     ‘Mohan said that he will come but he did not come’.



When ‘bol ke’ and ‘kar ke’ complementizers are used with the verbs expressing desire, intention or thought it provides another instance of syntactic change which can be attributed to Telugu influence. In such constructions the verb in the matrix sentences in Dakkhani Urdu and Kannada is either verb be or a conjunct verb and the embedded verb is in the infinitival form. The following sentences are illustrative:

Dakkhani Urdu

      (37) mereku    naye  bike  kharidna     bolke  ai/ dil  bolta:

              I-DAT     new   bike  buy+INF    COMP  is/ feel like

            ‘I have a desire to buy a new bike’


  (38)  na:ku     ca:na    ɖabbu        unɖa:l            -ani      und̪i / ko:rika   und̪i

          I-DAT    a lot     of money    have-oblig COMP    is /desire          is

         ‘I have a desire to have a lot of money’


(39)   nanaɡe     t̪umba    haNa      hond̪a beku    ant̪a    apekʃe  id̪e

           I              a lot of   money    have -INF    COMP  desire   is

          ‘I have a desire to have a lot of money’


Dakkhani Urdu, Telugu and Kannada use a similar construction with bol ke/ kar ke, ani and anta to convey the meaning of doing something intentionally. It should be emphasized that it is the presence of the participial form of the verb /bol/ and /kar/ with the verb hona: ‘to need’ in Dakkhani Urdu and /ani/ with the verb ‘ka:va:li’ ‘to need’ in Telugu and /ma:ɭu/ and /hajɭu/ with the verb ‘beku anta’ ‘to need’ in Kannada convey the sense of deliberateness.

Dakkhani Urdu

      (40) rashid  hona:                             bolke      ɡila:s    t̪oɽa:

             Rashid    needed (deliberately)  COMP   glass     broke

            ‘Rashid broke the glass deliberately’


      (41)  rashid         ka:va:li   ani       ɡila:su       paɡa:la           ɡoʈʈe:ɖu

              Rashid        needed   COMP  glass        deliberately      broke

‘Rashid broke the glass deliberately’


(42)        rashid            beku                          ant̪a        ɡaʤannu    oɖed̪anu

               Rashid        needed(deliberately)  COMP          glass          broke

Further, Dakkhani Urdu shares another construction with Telugu in which the embedded verb and the matrix are IDENTICAL, the embedded verb is in the obligative mood and the complementizer is the conjunctive participle form of the verb bol ‘say’ or kar ‘do’.[8]


   (43)  kamala   paɖa:li        ani            paɖind̪i

            kamala  fell+obl C.P. of ‘say’    fell


     (44) kamalaɭu   beku ant̪a     bid̪d̪aɭu

             Kamala      intentionally    fell

It may be noted that in Telugu and Kannada, the sentences, (43) and (44) above express intentionality of the subject.


Dakkhani Urdu in most of the cases uses “bol ke/kar ke” in onomatopoeic expressions, where some of the expressions the occurrence “bol ke” results into ungrammatical sentences the “kar ke” is used instead. In Telugu and Kannada, the complementizers /ani/ and /anta/ is used in onomatopoeic expressions which is a typical Dravidian feature.

Dakkhani Urdu   

      (45) t̪ʰakk                  bolke/  karke[i][9]               ɟor se          a:va:z    hui:

              onomatopoeic   C.P of ‘say’/C.P. of ‘do’    loudly       sound                happened

‘there was a loud thud’

     (46) uski   ã:kʰo se     pa:ni:     ʈapʈap              karke / *bolke             ɡirra:   t̪ʰa:

              her   eyes from  water   onomatopoeic   C.P.of ‘do’/ C.P.of ‘say’  falling was

‘Tears were coming from her eyes incessantly’.

Bangla, in most of the cases uses “kore” in onomatopoeic expressions, where in the occurrence “bole” results into ungrammatical sentences the “kore” is used instead.[10]

It is also to be mentioned that in Bangla the occurrence of bole is not permitted in some onomatopoeic expressionism. Instead of bole Bangla allows kore. The following examples are illustrative:

(47)  ɟol      kol kol                          *bole/kore            boicʰe

         water  onomatopoeic    C.P of ‘say’/C.P. of ‘do’  flowing

         ‘Water is flowing with the noise ‘kolkol’.


(48)   avaɭa   kannu    ɡaɭinda      paʈa paʈa        ant̪a      kanni:ru    barut̪id̪d̪avu

           her         eyes        from     onomatopoeic COMP  tears     come-PROG-PAST

       ‘Tears were coming from her eyes incessantly’.

It is also to be mentioned that in Telugu the occurrence of ‘ani’ is not permitted in some onomatopoeic expressionism. Dakkhani Urdu, however, does not have any such constraint with regard to the use of ‘kar ke’.[11]


   (49)   ʤaɭa ʤaɭa         ant̪a     ʃabd̪a     maɭut̪t̪a    ni:r    hariut̪t̪it̪t̪u

             onomatopoeic  COMP  noise     making    water     flowing

‘Water is flowing while making the noise ‘ʤaɭa ʤaɭa’.


In this paper, it has been tried to show that the verb “say” has various usages in Dakkhani Urdu and Telugu. The verb “bol ke” and “ani” are very productive lexeme in the grammar of Dakkhani Urdu as well as Telugu. In the process of grammaticalization, the verb “say” widely combines with other grammatical categories (Noun & Verb)and produces a new kind of meaning while it loses its natural meaning that is, desemanticization but it maintains and retains its grammatical category. In particular, “bol ke” and “ani” have been found to be used as a complementizer, quotative markers, reason marker, embedded questions, introducer, purposive markers, onomatopoeic expressions, manner adverbs etc.



Anttila 1989: Anttila, Raimo. Historical and Comparative Linguistics. 2nd edition. [1st edition 1972, New York: Macmillan]. (Current Issues in Linguistic Theory 6), Amsterdam: Amsterdam Benjamins,1989.

Bopp 1816: Bopp, F. Über das Conjugationssystem der Sanskritsprache in Vergleichung mit jenem der griechischen, lateinischen, persischen und germanischen Sprache. Nebst Episoden aus dem Ramajana und Mahabharata in genauen metrischen Übersetzungen aus dem Original texte und einigen Abschnitten aus den Vedas. Herausgegeben und mit Vorerinnerungen begleitet von Dr. K. J. Windischmann, Frankfurt/M: Frankfurt Andreäsche Buchhandlung, 1816.

Condillac 1746: Condillac, É. B. de, Essai sur l’origine des connoissances humaines. Ouvrage où l’on réduit à un seul principe tout ce qui concerne l’entendement humain, 2 vols. Amsterdam: Amsterdam P. Mortier, 1746.

Gabelentz 1891: Gabelentz, George von der. Die Sprachwissenschaft. Ihre Aufgaben, Methoden, und bisherigen Ergebnisse. Leipzig: Leipzig Weigel, 1891.

Givon 1979: Givon, T. On Understanding Grammar. New York: New York Academic Press,1979.

Guillaume 1964: Guillaume, Gustave. Langage et science du langage. Paris: Paris Nizet; Québec Presses de I Université Laval, 1964.

Harbir 1986: Harbir, K. A. ‘Some Aspects of Dhakkhini Hindi-Urdu Syntax with special reference to Convergence’. Delhi: University of Delhi, 1986.

Harbir 2002: Harbir, K. A. “Syntactic Convergence: The Case of Dakkhini”. Delhi: PhD. Dissertation, Delhi University, 2002.

Heine, Kuteva 2002: Heine B., Kuteva, T. Word Lexicon of Grammaticalization. Cambridge: Cambridge University Press, 2002.

Heine, Claudi & Hunnemeyer 1991: Heine, B., Claudi, U. & Hunnemeyer .F. Grammaticalization: A Conceptual Framework. Chicago: University of Chicago Press, 1991.

Heine, Reh 1984: Heine, B. and Reh Mechthild. Grammaticalization and reanalysis in African languages. Hamburg: Hamburg Buske, 1984.

Kachru 1986: Kachru, Yamuna. ‘The syntax of Dhakkhini’: A study of Language variation and language change. New Delhi: In Bh. Krishnamurti,1986.

Mustafa 2000: Mustafa, K.S. ‘A Descriptive Grammar of Dhakkhini’. New Delhi: Munshiram Manoharlal Publishers, Pvt. Ltd, 2000.

Lehmann 1982: Lehmann Christian. Thoughts on Grammaticalization. Munich: Munich Lincom Europa, 1982.

Schlegel 1818: Schlegel, A. W. von, Observations sur la langue et la littérature provençales, Paris: Paris Librairie grecquelatine-allemande, 1818.

Schmidt 1981: Schmidt, R.L. ‘Dakhini Urdu: History and structure’. New Delhi: Bahri Publication, 1981.

Shakeel 1975: Shakeel, A.S. Dakkhini Urdu and Punjabi, In Pakha Sanjan Vol. vi, Punjab: Punjab University, 1975.

Traugott, Heine 1991: Traugott, E. C., Heine. B, “Approaches to Grammaticalization”, Vol- I, 22. Philadelphia: John Benjamins Publishing Company Masterdam, 1991.

Articles in collections:

Beg 1977: Beg, Mirza Khaleel A. ‘Review of Sharma, Shriram 1964. “Dakkhini Hindi Ka Udbhav aur Vikas,”. Allahabad: Hindi Sahitya Sammelan’ II, Vol. 38.No.4. p. 259, 1977.

Harbir, Subbarao 1989: Harbir, K.A. and Subbarao, K.V. ‘Noun Modification in Dhakkhini Hindi-Urdu and Telugu’. India: Indian Journal of Linguistics, 13, 1989.

Masud, 1969: Masud Husain. ‘Urdu’ in Current trends in Linguistics. Vol. 5. Ed. Thomas A. Sebeok . Paris: The Hague, p.270-280, 1969.

Subbarao, Rao and Saxena 1989: Subbarao, K. V., C. Viswanath Rao, N.Rao and A. Saxena. In: “verb Say in South Asian Languages”. Ed. A. Mukherjee. Language Change and Language Variation. Hyderabad, India: Centre of Advanced Study, Osmania University. 89-104, 1989.

Articles in periodicals:

Guimier 1985: Guimier, Claude. On the origin of the suffix -ly. In Fisiak, 1985, pp. 155-70.

Humboldt 1825: Humboldt, Wilhelm von. Ober das Entstehen der grammatikalischen Formen und ihren Eifluβ auf die Ideenent wicklung. A bhandlullgen der Königlichen Akademie der Wissenchaften zu Berlill, 1825, pp. 401 -30.

Kachru 1978: Kachru, Yamuna. The quotative in South Asian Languages. Paper read at the first ICSALL, University of Illinois, Urbana-Champaign. Later appeared in 1979. In : SALA, I/1979, pp.63-77.

Kuryłowicz 1975 : Kuryłowicz Jerzy. The evolution of grammatical categories. Esquisses
linguistiques 2:38–54. (Originally publ. in Diogenes 1965:55–71.), Munich:  Munich Fink,1975, pp. 38-54.

Langacker 1977: Langacker, Ronald W. Syntactic reanalysis. In Li, ed, 1977, pp. 57- 139.

Meillet 1912: Meillet, Antoine. L’evolution des formes grammaticales. Scientia (Rivista di Scienza) 12, no. 26, 6. Reprinted in Meillet 1958, pp. 130-48.

Singh 1980: Singh, U. N. ‘Bole- an unresolved problem in Bengali syntax’. Indian Linguistics 41/1980, pp.188-195.

[1]Grammaticalization is a historical process, a kind of change that has certain consequences for the morpho-syntactic categories of a language and thus for the grammar of the language. (Traugott and Heine , 1991:38 )

[2]bolke in Dakkhini is also used as an introducer, just like ani in Telugu and anta in Kannada.

Dakkhani Urdu

           je       icʰaira:dʰa:   bolkemeri:       frenɖ

           this   emph     is     Radha     C.P. of ‘say’    my          friend

          ‘This is my friend Radha’

[3]       (i)   amar    ɟana   nai    ɟe      ʃi:t̪a:     ɡra:m   cʰeɽe   cole    gecʰe

            ‘I did not know that Sita had gone to the village’.

      (ii)   amar     biʃʃaʃ    ɟe     ʃe      aʃbe

            ‘My belief (is) that he will come’

[4] (i) arfat    university ku       a:t̪a      bolke/karke  bola:      pan      naĩ    a:ja

      Arfat University DAT will come  COMP         said        but      neg   come

 ‘Arfat said that he would come to the University but did not come’.

[5] (i)      ra:m      a:t̪a             bol  ke  /  karke        bola:      pan      naĩ   a:ja:

             ram    will come    COMP /    COMP      said       but      neg   come

            ‘Ram said that he would come but did not come’.

[6]   (ii)  ba:lamma:    a:ra              sokʰabar         mere  kuma:lum

         Ballama           is coming   adj. news       I-DAT      known

        ‘I know the news that Ballama is coming/comes’.

[7]    (i)      wa:na   paɖin-aʈlu        und̪i

rain      fall  COMP      is

          ‘It appears as though it is going to rain/ has rained’.

[8]Dakkhani Urdu

      (i) kamala:    ɡirna:           bol ke/ kar ke         ɡiri

           Kamala    fall+inf.    C.P ‘say’/ C.P of ‘do’   fell

          ‘Kamala fell intentionally’

[9] Here ‘bolke’ / ‘karke’ and ‘ani’ are used as a conjunctive participle (C.P).

[10]     (i)    dhopaʃ                  kore/*bole                  ɟore      ʃobd̪o      holo

          onomatopoeic    C.P. of ‘do’/ C.P. of ‘say’    loudly   sound    happen-PAST

‘there was a loud thud’

      (ii)   Or      cokʰ    t̪ʰeke       ʈopʈopkore  / *bole                                     ɟol         poɽ-cʰi-l-o

               her     eyes     from    onomatopoeic  C.P. of ‘do’/ C.P. of ‘say’  water  fall-PROG-PAST-AGR

          ‘Tears were coming from her eyes incessantly’.

[11]Dakkhani Urdu

     (i) pa:ni:       khalkhal         *bolke  /     karke          ubalra

         water   onomatopoeic   C.P of ‘say’ / C.P. of ‘do’  flowing

            ‘Water is flowing with the noise ‘khalkhal’.


    (ii)  ni:ɭɭu      ʤala:   ʤala:         *ani/ 0      prava:hist̪unna:i

            water    onomatopoeic      COMP/0     is falling

     ‘Water is flowing while making the noise ‘ʤala: ʤala:’.

About the author

Hifzur Rahman Ansary is a PhD Scholar at the Jawaharlal Nehru University (India) and his interests include Phonology-Phonetics, Syntax, Historical Linguistics and Language Studies.