ASLiJM AND AFRICAN S T U D E S , 79, 2010, 1, 109-133

ON TH E CA T E G O R Y OF ASPECT IN B E N G A L I1

Anna RÁCOVÁ Institute of Oriental Studies, Slovak Academ y of Sciences, K lem ensova 19, 813 64 Bratislava, Slovakia kaorraco@ savba.sk

A lthough in som c languages the research o f the category o f aspect goes as far back as the end o f 19th century,2 for the new Indian languages this is a relatively new problem . A ttention started to be paid to aspcct only as late as in the m iddle o f 20th century, and in this area up to now relatively little research has bccn done. T his also applies for Bengali, w here w ithin investigating this verbal category several approaches have appeared. T he aim o f oitr article is to present a b rie f survey o f the history o f the research o f aspect in B engali as well as on the basis o f existing findings and the analysis o f Bengali texts to provide our ow n view o f the m eans by w hich this sem antic category is expressed in the B engali language.

Key words: category o f aspect, category o f tensc, pcrfectivity, im perfectivity. habituality, continuousness, perfectness, im perfectness

I . Introduction 1.1 In addition to the fact that the speaker locates the action or state expressed by the verb with regard to the mom ent of speech into the past, the present or the future (doing so usually with the help of gramm atical tense) and com m ents on the mode o f the course of action or state (describing it for instance as sudden, regressive or com pleted), the speaker also expresses his/her subjective view of the action in the sense that he/she views it as if from the outside3 and then can

1 This study is published w ithin the grant project V EG A 2/0153/09. 2 E.g. H ER B IG , G., 1896. 3 CO M R IE , B. Aspect. A n Introduction to the Study o f Verbal A sp ect a n d R elated Problems, p. 4. 109

view it ‘as a single w hole’,4 ‘as a photograph’,5 as a ‘com plex, w holesom e, com plete ev en t’,6 applying to it an ‘integrative’ attitude.7 On the contrary, a person can view a situation from w ithin,8 as a series of parts,9 ‘as a film ’,10 as ‘being ju st carried out,’11 hence applying with regard to it a ‘fractionative attitude’.12 It is from such a view point of the ‘internal tem poral constituency’13 that we speak about aspect. O ver the num erous years during which linguists have been dealing with the questions concerning the category o f aspect, various approaches to researching it have been applied. Those who understand aspect from a w ider perspective include into its definiton aktionsart, i.e. the mode of verbal action,14 whereas others separate the mode of verbal action from aspect;15 still others take into consideration not only the opposition perfectivity - im perfectivity, but they discuss three aspects, i.e. the perfective, imperfective and m ultiplied aspect,16 or the perfective, im perfective and general aspect into which they include the sim ple present, sim ple past and habitual past.17 Such varying approaches can stem from the author’s conception of this verbal category as well as from the fact of which particular language is analysed. W hile on the sem antic level the opposition of perfectivity and im perfectivity could be understood in various languages in the same way, the varied system o f verbal form s nevertheless leads to different linguistic m eans by which this opposition will be expressed and consequently also to various problem s which will have to bc resolved in connection with thc category of aspect. T his is evident already when com paring the research dealing with the category o f aspect in Slavic and Germ anic or Rom ance languages. 4 CO M R IE. B. ibid., p. 18. 5 SO K O LO V Á , M. Sémantika slovesa a aspektovéformy (Verbal Sem antics and A spect Form s), p. 28. 6 Encyklopédia jazykovedy (Encyclopaedia ofL ing u istics), pp. 402 404. ' K E N ISTO N , H. V erbal A spect in Spanish, p. 164. 8 C O M R IE 1B. ibid., p. 24. 9 K E N ISTO N , H., ibid., 164. 10 SO K O LO V Á , M., ibid., p. 28. 11 Encyklopédia jazykovedy (Encyclopaedia ofL in g u istics), pp. 4 0 2 -4 0 4 . 12 K EN ISTO N , H., ibid., p. 164. 13 CO M R IE , B., ibid. p. 3. 14 For instance T rudgill (from Stekauer 1995:36), M athesius (from Stekauer 1995: 36). I5IS A Č E N K 0 , A. V., Grammaticheskiy stroy russkogo yazyka vsopostavlenii s slovackim. Morfologiya (G ram m atical Structure o f Russian in C ontrast to Slovak. M orp h o lo g y ),p . 130. 16 K O PEČ N Ý . F. Slovesný vid v češtině (Verbal A spect in C zech), 1962, p. 15 et seq. 17ZO G R A F, G. A. Morfologicheskiy stroy novykh indoariyskikh jazykov (M orphological Structure o fN e w Indo-A ryan L anguages), p. 305.

110

In Slavic languages, where the verb can be included am ong perfectives or im perfectives already on the basis o f its infinitive (Slovak perf. bo d n ú ť vs. imperf. bo d a t'{stab), Rus. perf. brosit’v s. imperf. brosat’(throw), etc.), linguists differ on various questions of viewing the aspect as has been shown in a survey undertaken by Sokolová.ls This concerns for instance the question o f whether frequentatives belong into the category o f aspect or the category o f aktionsart, and if they belong to the category o f aspect, whether they should be included among im perfective verbs or w hether they form a separate aspect or sub-aspect. The acknow ledgm ent or non-acknow ledgm ent of the existence o f aspectual prefixes and so on is also discussed. In Germ anic or Rom ance languages aspect is not m anifested within the infinitive, but the aspectual opposition is reflected on the finite verbal forms: in English the sem antic distinction perfective - im perfective is reflected in the m orphological opposition of sim ple past versus progressive past, in French in the m orphological opposition passé simple versus imparfait. These languages deal with different problem s than the Slavic languages: e.g. to what extent the perfect form belongs to the category of aspect, etc. O f course, there can arise the need for a different approach in other languages with a different gramm atical structure and different system o f verbal forms, above all to the m eans with the help of which this verbal category is expressed. This also applies for Bengali, which constitutes the subject of our article. 2. History o f the R esearch of A spect in Bengali 2.1 In Indian languages the category of aspect does not traditionally rank among the num ber o f central problem s in describing their gram m ar, which is also the case of Bengali. This is also docum ented by the fact that aspect as a specific verbal category is not tackled even in the detailed description o f the m orphology of Bengali, The O rigin and D evelopm ent o f the Bengali Language by S. K. Chatterji published in 1926.19 O f course, Chatterji is fam iliar with aspect in other languages, as is evident from his statem ent on page 1050 where he discusses com pound verbs which in m odern Indo-A ryan speeches he considers as equally characteristic as aspect in Slavic languages. Linguists investigating Indian languages have m anifested interest in aspect only as late as the second half of the 20th century thanks to Soviet Indologists,

18 SOKOLOVÁ, M. Verbálny aspekt vslo ve n čin e - bilancia doterajšieho výskumu (Verbal Aspect in Slovak - A Survey of the Existing Research), p. 8. 19 CHATTERJI, S. K. The O rigin and D evelopm ent o flh e B engali L anguage. Calcutta 1926. 111

who in com piling dictionaries of new Indian languages searched for equivalents of R ussian perfective verbs. In the Hindi-Russian D ictionary2u Russian perfectives were translated with the help of Hindi com pound verbs, which started the considerations that in Hindi it is the com pound verbs with the help of which perfectivity of action is expressed.21 Sim ilarly, in the B engali-R ussian Dictionary by E. M. Bykova et al.22 Bengali com pound verbs correspond to Russian perfective verbs, e. g. rāgā ‘serditsya’ (be angry) as against rāgiyā uthā - ‘rasserditsya’ (becom e angry); eaIā - ‘idti, dvigatsya, echať (go, move, ride) as against caliya yāoyā ‘uchodit, uyezžať (depart), etc. Otherwise, only im perfective equivalents are given with sim ple verbs. A Czech Indologist D. Zbavitel in the dictionary part of his textbook on B engali23 som etim es gives a com pound verb as a perfective counterpart to the im perfective verb (khuje paoya - ‘najíť (find), as against khöjä - ‘hledať (look for), m ārā yāoyā - ‘um říť (die) as against m arā - ‘um írať (be dying)). Som etim es with the sam e Bengali form he gives both a perfective and an im perfective Czech equivalent, e.g. pheIā - ‘házeť and also ‘hodiť (throw). W ith the basic dictionary form he quite often gives a perfective Czech equivalent: khoIā - ‘otevříť (open) (in contrast to the im perfective form in the B-R Dictionary ‘otkryvát(sya)’), mākhā - ‘pom azať (coat) (in contrast to the BR Dictionary ‘mazat, natirat’), ghatā - ‘přihodit se’ (happen) (B-R 'sluchatsya, proischodiť), nām ā - ‘sestoupiť (descend) (B-R ‘snizhatsya, spuskatsya, opuskatsya’), pāItāna - ‘obrátiť (revoke) (B-R ‘perevrachivat, perevertyvať), etc. He obviously used equivalents corresponding to the text in the textbook. This differing interpretation of the dictionary form - thc verbal noun - is enabled by the fact that this form only carries the basic lexical m eaning. This is specified only in the finite form of the verb, in its com bination with auxiliary words and with the help of the context.24 The lexical m eaning plays an important role in determ ining the aspect o f the verbs which were form ed by a nom inal-verbal com bination, i.e. by com bining the substantive or the adjective with the verb.25 For exam ple, the verbs formed 20 B ESK R O V N Y Y , V. M. Khindi-nisskiy slovar (H indi-R ussian D ictionary). 21 Cf. e.g. PO Ř ÍZK A , V. On the perfective verbal aspect in Hindi , p. 35 - 37. Or HO OK , P eter Edw in, The compound verb in Hindi, 318 p. 22 BY K O V A , E. M. et al., Bengal’s ko-russkij slovar (B enga!i-R ussian D ictionary). 23 ZBA V ITE L , D. Bengálština (Bengali), 489 pp. 24 In m ore detail see B Y K O V A , E. M. The Bengali Language, p. 77. 25 This is the way to supplem ent in Bengali a relatively small num ber o f verbs which according to S. K. C hatterji are estim ated to am ount to 800 (cf. B ykova, op. cit., p. 75). M ost often it is the verb karā - ‘to d o ’ (Jijnas /question/ karci - ‘to ask ’, kcij /w o rk / karā - ‘to w ork’, guli /bullet/ karā - ‘to shoot’, phon /telephone/ karā - ‘to telephone’,

112

from adjectives with the help of the verb haoyä - ‘to b e ’ often denote a change of state: khārāp haoyä - ‘to go w rong’, pākā haoyä - ‘to ripen’, bara haoyä ‘to grow ’, chata haoyä - ‘to become sm aller’; i.e. they denote an action which is perceived as perfective. 2.2 Although in the B engali-R ussian dictionary, sim ilarly to the H indi-Russian dictionary, com pound verbs are given as equivalents o fR u ssian perfective verbs when describing the Bengali gramm ar, the Soviet authors do not search for the category of aspect in com pound verbs, but they rather search for it in temporal forms which som etim es are referred to as aspectual-tem poral ones. 2 .2 .1 According to the Russian Bengalist E. M. Bykova the derivational structure the Bengali verb “has no category of aspect”26. On the basis of com paring the system of the Bengali and Russian verb she com es to the consclusion that “ in Russian, and in all Slavonic languages in general, suffixation and prefixation play the main role in creating pairs of verbs opposed in aspect: rasskazat - rasskazyvat (narrate), chistit - vychistit, pochistit, ottchistit (clean)” . O f course, we cannot find such opposition in Bengali, but as Bykova states further on, “this does not mean that the category o f aspect is absent from the Bengali gramm atical system. It exists, but is expressed in a different way. The category of aspect (perfectness-im perfectness)27 is reflected in the tense forms of the verb” .2ii However, she does not mention all the tense forms, she only states the suffixes -it-/-t-29 (i.e. the suffix of the past habitual tense), -iI-/-l- (she considers its occurrence in the sim ple past tense only), and -i(y)a-/-e- (the suffix o f perfect tenses)30 which, as she stipulates, and haoyä - ‘to b e ’, but possible are also com binations with other verbs, e. g.: deoya ‘to give’ ^ā Iiā rā /guard/ deoya - ‘to guard", Iaph /jum p/ deoya - ‘to ju m p “ , ghanda /sm ell/ deoya - ‘to stink“ ), päoyä - ‘to get’ (kasta /suffering/ päoytJ - 't o su ffer’), para - ‘to fall’ (asukhe parā - ‘to fall ill’); pheIā - ‘to throw ’ (tliuthu /saliva/ pheIā - ‘to sp iť etc.). 26 BY K O V A , E. M. Bengali language , p. 77. 27 In literature there often occur the term s perfectness, im perfectness in the m eaning o f aspcct. W e preserve such term s when quoting authors, otherw ise we distinguish the term s perfectness and im perfectness in the case o f tenses, and perfectivity and im perfectivity to denote the aspectual opposition. 28BY K O V A , E. M., ib id ..p . 78. 29 H ere as well as in the w hole article before the slant there are suffixes occurring in the literary language Sädhu B häsä and after the slant the suffixes o f the standard language Calit Bhāsā. 30 H ere in contrast to the earlier Russian edition o f this book, obviously by m istake, the suffix o f im perfect tenses -ite- is dropped. 113

“ indicate the character of action from the point of view of aspect (perfectness im perfectness) and mode o f action (m omentary - repeated, including habituality, frequency; continuity - non-continuity)” .'11 As is evident, Bykova does not differentiate between the aspect and the mode o f verbal action. Further on she claim s that “com pound form s32 synthesize the gramm atical m eanings o f the non-finite form, the lexical meaning of the auxiliary verb and the gram m atical m eanings of its finite forms. The synthesis of all these meanings makes it possible to express the resultativity of an action in the present and past (in forms with the conjunctive ending in -i(y)a -/-e-) plus its perfectivity (under certain conditions it expresses frequentation), or the nonresultance of an action in the present and the past (in patterns in -ite form s) plus continuousness, etc.”33 E. M. Bykova com es to the conclusion that “on the basis o f the interaction of the m eanings of the stem and the auxiliary units in the structure of finite forms, the follow ing groups o f temporal meanings can be stated: 1. Perfectness, if there is a result in the present (perfect) or in the past, when the action belongs to the relatively rem ote past (pluperfect). 2. Im perfectness and continuousness in the present or in the past (present and past im perfect). 3. Habituality, frequency, reiteration (frequentative, present). 4. M om entary perfectness in the indefinite past (preterite). 5. Relegation of the action to the future w ithout the distinction of perfect - im perfect (future tense).”34 As is evident above, Bykova delim its the m orphem es by which aspect is expressed in a linite verbal form, but she perceives the whole verbal form rather synthetically and speaks about it as about a temporal form. 2.2.2 In her textbook o f the Bengali language35 another R ussian Bengalist E. A. Alekseeva does not analyse the category of aspect, but the preterite, which she labels as a past narrative tense is denoted by her as a special aspectual-tem poral form. How ever, she uses the term s perfective and im perfective participle (sovershennoe, nesovershennoe deeprichastiye {kare, karte). The perfective participle usually expresses a perfective, preceding action.'1' At the same time it can express the reason and the condition of the action expressed by the second

31 BY K O V A , E. M., ib id ..p . 86. 32 B ykova has in mind perfect and im perfect tenses. 33 BY K O V A , E. M., ibid., p. 86. 34 BY K O V A , E. M ., ibid., p. 86 - 87. 33 A L EK SE EV A , E. A. U čebnik bengaIskogo yazyka (Textbook of Bengali), Part I 1976, 270 p., P art II 1977,272 p. 36 A LEK SE EV A , E. A., ibid. I, p. 170. 114

verb.37 N evertheless, it can also express an im perfective action taking place sim ultaneously with another one.38 She denotes the finite verbal form s as tenses and with regard to the present perfect tense she says that it denotes a com pleted action if its result continues to exist when the focus is not on the action but on its result. H ow ever, the perfect tense denotes also an activity started in the past but not yet finished.39 The forms o f the future tense do not distinguish between perfectivity and im perfectivity. If the duration of the activity is to be stressed, it is expressed analytically: likte thäkba - 1 shall be writing.40 2.2.3 M ore system atic attention to aspect is devoted by I. A. Svetovidova in her study Vrem ya, vid i sposob deystviya v bengali.41 She understands aspect as expressing the com pleteness or non-com pleteness o f the action. In her opinion only the present (kari), the present com pound (karchi), the frequentative (kartām ) and the im perfect (karehiIām) unanim ously express a non-com pleted action. The other four tense forms do not indicate com pletedness or noncom pletedness of the action if the particular form is assessed outside of any context. In her opinion, while the perfect {karechi), the preterite (karIām ) and the plusquam perfect (kareehiiām ) most often express a com pleted action, they can also express a non-com pleted action. Hence, with regard to the category of aspect she considers these tenses as being neutral.42 The activity expressed by the future tense can be understood both as com pleted and non-com pleted. Further on, Svetovidova shows that in Bengali aspect can be expressed by various means, both gram m atical and lexico-gramm atical. Gram m atical m eans expressing aspect can be either m orphological (tense form s) or syntactic jo in in g the perfective participle with the verbs o f existence resulting in com plex predicate). As lexical-gramm atical m eans she identifies the perfect participle with the verbs of motion, i.e. com pound verbs which she calls ‘složnointensivnye glagoly’ (com pound-intensive verbs). She com es to the conclusion that aspectual m eanings have to be investigated within their mutual interaction with the m eanings o f the m ode of verbal action, while the mode o f verbal action characterizes the action (state) from the point o f view of its being carried out. It indicates the m om entariness or the duration o f the action, its one-tim e character or m ultiplicity, interruptedness 37 A L EK SEEV A , E. A., ibid. I, p. 171. 38 A L EK SEEV A , E. A., ibid. I, p. 170. 39 A L EK SEEV A . E. A., ibid. Vol. I. p. 216. 40 A LEK SEEV A , E. A., ibid. Vol. II, p. 29. 41 SV E T O V ID O V A , I. A. Vremya, vid i sposob deystviya v bengali (Tense, A spect, and M ode o fV e rb a l A ction in Bengali), p. 218. 42 S V E T O V ID O V A , I. A., ibid", p .2 1 9 .

or uninterruptedness, usualness, unexpectedness, intensity, etc. The mode of verbal action is expressed by the same gramm atical and syntactic m eans as aspect. T he participle ending in -iy ā with the verbs o f existence ach-, thākā and rahā in addition to the duration of action also expresses the m anner o f its duration, while with ūeh- it stresses the duration of the action at a particular m om ent in the present or in the past (...tini cup kariyä basiyä āehen - he/she is silently sitting), with thākā it expresses habituality/usualness (obychnost) o f the given lasting state (ām i särädin ekhānei base thāki - I have been sitting here for days, with rahā it expresses rem aining in a state within a particular section of time (aneksan dujane cup kare base rahila - they have been silently sitting for long) (p .2 1 9 ).43 2.2.4 T he discussion about the category of aspect in new Indo-Aryan (NIA) languages has also been included in the work M oifologicheskiy stroy novykh indoariyskikh yazykov44 by R ussian linguist G. A. Zograf. IIe points out that many English-speaking researchers who deal with N L \ languages do not differentiate the category of aspect from the category of tense. Even in the cases when the category o f tense is acknowledged, it is by far not understood unanim ously and is often interpreted very widely. He considers it indispensable to use all the accessible formal m eans to differentiate the category o f aspect from that o f mode of verbal action.45 On page 228 he introduces three aspect-tense bases of the verb: past habitual: cal-it-, sim ple past: cal-il- and future cal-ib-. Z ograf presents a schcme depicting the system of the personal verbal forms in Bengali (he presents sim ilar schemes for other NIA languages, too).

43 In the book R eference B engali G ram m ar on page 156, W. Sm ith says that āeh- ‘b e ’ with stative verbs describes the present existence o f a state and chil- ‘w as’ with stative verbs describes the past existence o f a state: suye ache - is sleeping, Iukiye chila - was hiding; thākā ‘rem a in ’ with stative verbs describes continued existence o f a state: āmi eirkāi tomār path ceye base thäkba - I will be waiting for you a long tim e; raoyä ‘rem ain ’ with stative verbs describes continued existence o f a state: āmi sandhye theke sobär ghare lukiye raiIutn. - I rem ained hiding in the bedroom since evening. 44ZO G R A F, G. A. Morfologicheskiy stroy novykh indoariyjskikh yazykov (M orphological Structure o f N ew Indo-A ryan Languages), 368 p. 45 ZO G R A F, G. A., ibid., p. 208. 116

0 0 0 1

0 -che /-c 'e/ 2 -chila /-c 'ilo/ 3 thākihē /t 'akbe/ 4 thākē 5 thākita

1 karitē A a r[te ]/

2 kariyöi /kare/

karē A ore/ karila A o r lo / karibě A orbe/

karitēehē A o rc 'e/ karitēehiia A o r c 'ilo/ karite thākihē A o rte t 'akbe/ karitē thākē

karita A o r to /

karitē thākita

kariyachě A o re c 'e/ kariyächila A o rec 'ilo/ kariyá thakibě A o re t 'akbe/ kariyä thākē A o re t 'ake/ kariyä thākita

Adapted from Zograf, p. 271. He interprets the oppositions on the horizontal axis as aspectual ones and the oppositions on the vertical axis as tem poral ones while stating that in Bengali the forms lying on the temporal axis in lines 4 and 5 can be explained as forms having the meaning of the mode of verbal action. On the basis of the system of personal verbal forms in various NIA languages Z ograf com es to the consclusion that ‘the m eans for expressing the m orphological category of aspect in NIA languages are constituted by a non­ predicative elem ent o f a finite form ’. A ccording to Z ograf the forms of the im perfective aspect represent the action in the process of its progress as related to some m om ent o f each of the time levels expressed by the other action or by other means. O f param ount importance for them is the meaning of non-com pletion, duration and sim ultaneity of action. The forms of perfective aspect show the action within its result and they entail the meanings of com pletion, resultativeness and precedence of verbal action.46 In the form s w ithout a participial com ponent the previous groups of meanings are not expressed. In Bengali these are represented by kare and karila which are usually labelled as the ‘general present’ and ‘general past’. A dded to them is also karita (Zograf labels it as the form of the conditional mood). All the three are labelled by him as form s of the ‘general aspect’ /obshchiy vid/, for which the m eanings o f non-com pletion or com pletion, sim ultaneity or precedence, duration, resultativeness and so on arc irrelevant.

46 ZO G R A l5, G. A., ibid., p. 283. 117

Z ograf points out the fact that in many cases, above all in the colum n of the forms o f ‘general aspect’, sem antic dynam ism occurs. For exam ple, the simple form o f the future tense is sem antically close to the perfective aspect, which evokes the tendency that in the position of the im perfective aspcct a secondary analytical form is used (karba - karite thäkba).47 A fter his analysis Z ograf com es to the conclusion that ‘the aspectual opposition in northern Indian languages is relevant for the m ajority of predicative forms and such non-predicative forms as participles. The other non­ predicative forms can be included into the general aspcct. In personal forms the aspectual opposition is neutralized (in favour of the general aspect) in the , 4« imperative. In conclusion, he claim s that “ in north-Indian languages, with the exception of the A ssam language, the m orphological category of aspect is represented by the opposition of three groups of forms connected with two types o f participial form ations and one type of non-participial form ation w hich are interpreted as the perfective, im perfective and ‘general’ aspect” .49 2.3 T he problem o f aspect has also been dealt with by the Czech Indologist Dusan Zbavitel in his book N on-finite Verbal Form s in B engali.'" Although in his work he deals with non-finite verbal forms, in the introductory chapter on the verbal system in Bengali he also touches upon the aspect o f verbal action. He does not analyse how the aspect is expressed in Bengali, but from several of his notes included it seems that he is convinced that it is m anifested on the finite verbal form. IIe does not indicate which m orphem e in it expresses the aspect, but rather he understands the finite verbal form in its com plexity sim ilarly to Svetovidova. He m entions ‘ten different form s’ o fth e verb: sim ple present, sim ple future, simple preterite, habitual past, present imperative, future im perative, present perfect, preterite perfect, present progressive and preterite progressive. He com es to a sim ilar conclusion as Svetovidova had earlier: that ‘these finite forms taken out o f their content are incapable of expressing exactly delim ited and m utually distinguished categories of tense, aspect and m ood’5i A bit further on he says: ‘owing to certain features of the developm ent o f the Bengali verb, the individual finite form s have lost much of their unam biguity in respect of tense, mood and aspect.’52 47 Z O G R A F, G. A., ibid., p. 284. 48 ZOGRAI-', G. A., ibid., p. 285. 49 ZO G R A F, G. A., ibid., p. 305. 50 Z B A V IT EL , D. N on-F initeV erbal Form s in B engali, 137 p. 51 Z B A V IT EL , D ., ibid., p. 11. 52Z B A V IT EL , D .,ib id .,p . 11. 118

Zbavitel also claim s that ‘syntactic constructions are needed, instead of pure m orphological means, w herever a more definite denotation of tense, aspect and mood is required’.53 He also uses the term aspect with regard to some com pound verbs (according to A. Pal he calls them verbal com pounds), but here he offers a wider understanding of aspect. He does not put into the opposition of perfectivity versus im perfectivity, but he speaks about ‘the aspect of suddenness’ ( ‘the prevailing aspect of the suddenness o f the verbal action’ (■othā)),54 or about ‘the aspect o f continuity’ (e.g. the vector yāoyā v CV baliyä gela conveys the ‘aspect of continuity’ to the first verb).55 He says for instance that ‘the first verb is the bearer of thc main action and the second only modifies its aspect’, (p. 82) Also, that ‘all com pound makers... may add more than one unam biguous aspect to the directing verb’.56 Elsewhere in the sam e connection he does not use the term ‘aspecť but he speaks about expressing ‘the idea of suddenness or unexpectedness’.57 2.4 The authors of Bengali gram m ar writing in English som etim es make as if only a passing note of aspect without specifying in any detail by what linguistic m eans it is expressed. For instance, W. L. Smith writes: ‘In Bengali aspect is as much stressed as tense; as a consequence tense usage is rather loose.’58 He then speaks about aspect in connection with com pound verbs which he calls ‘aspective com pounds’, i.e. such com pounds which are form ed with the help of perfective participle + an aspective auxiliary. How ever, as the following exam ples illustrate, the same vector can serve both as a perfective auxiliary and as an auxiliary expressing continuity. For example, the vector yäoyä as a perfective auxiliary indicates a com pleted action: pulis ese geche, pāIāo - The police have come, flee. However, on the other hand, yāoyā describes a continuous action or process. In that case it is often in the form o f the im perfect tense: se cup kare sune yäcchila. - He silently listened.59 W ithin such usage the com pound m aker (vector) yäoyä in all the cases is presented in im perfect tenses, i.e. the duration is indicated by the very usage of tense.

53 Z B A V IT E L, D., ibid., p. 11. 54Z B A V IT E L, D., ibid., p. 82. 55 Z B A V IT E L, D., ibid., p. 76. 56Z B A V IT E L, D., ib id .,p . 80. 57 Z B A V IT E L, D., ibid., p. 88. 58 SM ITH , W. L., B engali Reference G ram m ar, p. 99. 59 SM ITH , W. L., ibid., p. 150. 119

2.5 A different understanding o f aspect in New Indo-A ryan languages is provided by Colin P. M asica.60 He considers it as a m orphological category, and in contrast to the previous authors he asserts that aspect has to be differentiated from tense. He also refuses the search for aspectual functions in com pound verbs; he relegates them to the dom ain of aktionsart. In his opinion, com pound verbs belong more to the dom ain o f derivation, that is, to lexicon, than to gramm ar. H e does not agree with the opinion that vectors have an ‘aspect’m arking role, and a perfective-m arking role in particular, because ‘they are not predictably the same for all verbs, they are governed by lexical sem antics’,61 ‘they do modify the m eaning of the verb itself, however subtly and in many cases untranslatably, which pure gramm atical elem ents should not do; moreover, there is already an aspect-m arking m orphological system in place in NIA, which is consequently ignored (or confused with tense, which is thereby also m isconstrued)’ .62 M asica also refuses the Slavic model for Indo-Aryan as m isleading and unnecessary63 and neither does he consider as appropriate any attem pts at searching for analogy betw een Slavic verbal prefixes and Indo-Aryan vector verbs.64 H e com es to the conclusion that aspect has to be differentiated not only from time but also from aktionsart. A ccording to M asica aspect in NIA is ‘em bedded in the suffixal m orphology’.65 H e claim s that in NIA the basic aspectual system is created with the help of participles, the sem antics of which already entails the basic distinction into perfective/im perfective/unspecified, while in some languages further differentiation takes place. Most com m on is the splitting o f the im perfective into habitual and continuous. W ithin his description he also concentrates upon the languages in the eastern group, i.e. on Bengali as well, where for perfective m arker he considers the sufix -il-/-l-', but he has in mind only the suffix of the finite verbal form which is already traditionally denoted as simple past or preterite. He considers this form as tense-unspecified in contrast to the tense-specified ‘perfect’. The marker o fth e perfect is the suffix -iya-/-e-. As im perfective m arker (or habitual, or continuous) he denotes the sufix

60 M A SICA , 61 M A SICA , 62 M A SICA , 63 M A SICA , 64 M A SICA , 65 M A SICA ,

C. C. C. C. C. C.

P. The Indo-A iyan Languages, pp. 262-279. P., ibid., p. 327. P., ibid., p. 268. P., ibid., p. 267. P., ibid., p. 266. P., ibid., p. 269. 120

-ite-/(Ii).66 He points out that ‘the available aspectual m arkers must not be confused with the aspectual distinctions them selves. The latter are som etim es marked in different ways in conjunction with different tenses’.67 This theory o f his is clearly reflected in the finite verb paradigm o f Bengali (Colloquial): sim ple present (he denotes it as General Unspecified > Present Habitual) and future considers for aspect-unspecified; the past habitual/contrafactual has an im perfective aspectual m arker (-y-), while being tim e-unspecified; the present continuous and the past continuous have a continuous aspectual marker (H), the simple past (which he denotes as U nspecified Perfective) has a perfective aspectual m arker (-/-) and it is timeunspecified; the present perfect and the past perfect have the perfective aspectual m arker (-