'rHE ORIGIN AND DEVELOPMENT OF THE BENGALI LANGUAGE THE ORIGIN AND DEVELOPMENT OF !HE BENGALI LANGUAGE -BY SUNITI KUMAR CHA TTERJI M.A. (Calcutta), D.Lit. (London) Khaira Professor of Indian Linguistics and Phonetics IUld Lecturer in English and Comparative Philology in the University of Calcutta With a Foreword by SIR GEORGE ABRAHAM GRIERSON, K.C.I.E. In Two Parts Part II : Morphology, Bengali Index CALCUTTA UNIVERSITY PRESS .1926 PRINThD BY BHUjENDRALAL BANERJEE AT THE UNIVERSTTY PRESS, SENATE HOUSE, CALCUTTA, Reg, No. 36B, August, 1926-500 19762· 2 3 NOY !966 S'I'. VRRp . - STOCKTAKING·2011 ,.... ~IORPHOLOGY • • MORPHOLOGY CHAPTER I: FORMATIVE AFFIXES 394. Hoernle's discussion of the NIA. suffixes in his ' Gaudian Grammar' remains the most thorough study of the scope and. origin eof these, although some of his derivations are more or less speculative. The Bengali formative affixes have been discussed in some d~tail by Rai Bahadur Yogesh Chandra Vidyanidhi in his Bengali grammar {pub..: lished by the VSPJ., pp. 147-191), b~t the arrangement is capable of improvement, and some of the etymologies can be questioned. There is no doubt as to the source of most NIA. affixes, but there are a few which have not been satisfactorily explained. · The OIA. affixes, primary or secondary, became disguised, or simply dropped o:lt in NIA., through either assimilation or pho~etic decay. But a number of OIA. affixes have been preserved, and· occur in .NIA. as living forms, although in their present shape they are •widely removed from their OIA. originals (e.g., « -pan • < c -tvana- ,. ; « -ab, -ib, -iv • < ·c -tavya- •; « -i •· < « -ika~·ika•; etc.). The NIA. speeches preserve only the relics of some. OIA. affixes, which as a result of phonetic decay now form integ.ral A.nd indivisible parts of the words, and have; so far as the speech-feeling of the NIA. speaker is concerned, n'o separate. existence, and are not recognisable or analysable any more as an affix. Thus, for example, the OIA. primary affixes« -a,'-a, -as •, in noun and adjective formatio~, are no longer living ones in Bengali, as they' are lost in all inherited words ; but the affix « ;.ana •, to ·form verbal nouns, continues to bff in full force as 1:5fif, 'Sot 'c -an~, -ontj. ,. ; and in words 'like st~, ~t1f « camtJ., kam~ •, the· -~ « -mtJ. • is only partially representative of the original OIA. affix « -man •; but it is no longer possible to analyse the B erf!gali words as being made up of. a t·oot pluB an affix, ~lid neither can -~ .. 652 MORPHOLOGY: CHAPTER 1 « -m:j, • be distinguished from the -l{ « mtf, • in words like iSitlf « jam:j, • (jambu) and~« am:j, • (amra). The words in Bengali have becom~ to all intents and purposes root-words, or primitive words, the sense of\heir being composed forms being long lost, from the beginning of MIA. So in words like Sftf5 • rati • (riitri), 9{~ • p£ti • (pa:gkti), ~1 • siita • (MIA. sonta-, OIA ... srotas), <~StiSf « kaj4,. (karya), the OIA. affixes « -tri, -ti, -tas, -ya • are present, but they are no longer alive. These relics or fragments of dead forms are not necessary to consider, unless the subject of study were the history of the OIA. suffixes down to NIA. through MIA. Below are given in a rough alphabetical order all the living affixes,. primary: as well as secondary, which are found in Bengali, Old, Middle and ~e~, with indications o:li their origin. [1] Su:FFIXEs [I] 'l'ADB.HAVA SUFFIXES (1) <5I « -a, :j, •: masculine, feminine and neuter noun affix. 395. Representing the primary affixes of OIA., masculine « -a!) •, feminine « -a •, neuter « -am •, also • -as •, with or without « gui}a • of the root-v~wel, this affix may be said to have existed in Old Bengali and Early Middle Bengali, but was lost from Late Middle Bengali times with the loss or assimilation of the final vowel: e.g., 51if « cara • (ciira), l'ftot «sana • (sathjfia), l{'f « mal}a • (manas), ~ • sa:gkama • (sa:gkrama) bridge; but cf. Late MB. Of • de • (deha), ~ • man~,. (manas), ~t'§ • bat~ • (varta), c~r~ .. bol~ • speeck (MlA. *bolla), 51~ • cal~ • toay, style (cala), etc. ( 2) <51 « -a ,. , also written o.s « -o ,., 396. This is the pleonastic suffix, found in a number of nouns and adjectives-e.g., ~~. <::5te~1 • bhala, bhalo • good; tG'!, ·'fn:;'jj .. kala, kalo ,. blaclc; ~~ « dhala ,. white; etc. H comes from the pleonastic or pejorative ~ -kil. ,. of OIA., added to nouns in • -a • = • -a-ka • > MIA ... -aa- ,. > .. ~wa ,. : see ante, pp. 302, 346, 347, MS. The affix is thus of secondary • origin. 'l'he employment of the pleonastic « -ka •, giving, with• the THE AFFIXES '-A, -0,' '-AN'l'A,' '-AT(A),' '-ATA' 653 preceding «-a,. of the word, the MIA. forms« -aga, *-ag.a, -awjya •, is quite a note-"'orthy thing from the Second MIA. stage; and in Skt. as written in . that period, •as well as in later times, this vernacular habit is fully reflected : in the Skt. of inscriptions, this is very much in evidence . .(Cf. Damodarpm· Copper Plates of the Gupta Period, Ep. Ind., XV, pp. 130, 138.) The combination.of « -aa,. > « awa,. seems to have begun to be altered to an « &. ,. (long.[<>:]) sound in the OB. period; and it was accomplished 1:w the 14th century. (See pp. 260, 302, 346.) In MB., it is frequently found in personal names, where it is ·commonly written as IS « -o,.: e.g., ~<11 • SibO ,. (OB. * Sibawa= ts. Siva+ ka), ~1 « Rudo ,. (*Riidawa< Rudra+ Ira:), f~~ « Bibho,. (*Bibhawa < *Vibha = Vibbu +ka), PI~~ .. SidhO,. • (*Sidhawa < Siddha [ = Siddh~svara] + ka), etc. These common contrac~ions of personal names in MB. are thus on the basis of a form in « a- ,., shortened from the full names (e .. r;., i9fl:~1 : Jano,. < « Janardana •, f9f1:q1 « PithO •< « Prthvidhara •, etc.), with the pleonastic « -ka,. added .. Forms like the above are very common in MB. genealogical works (kula-grantha, . kulaji= kulapafijika) : which shows that they were much in use. There is another affix, also (occasionally) written -151: « ·a ,., indicating the sense of about to do something : e.g., ~tif-~t"if « kH:da-kiida •, beside ~t1:if1- ~ttlf1 « kado-kiido • 1·eady to burst into tears. 'rhis is a pr~mary affix f~rming verbal adjectives and is derived from IS « -o ,. <~ « -u ,., which see, infra. (3) Affixes derived from the OIA. Active Present Participle « -ant- ,. : 151:~ .. -ants •; '51'~ > "51'~1, '51"~ ('51"f'!i) .. -at:/. > -ata., -ati (-a.ti) •, '51"\!i « -a.ta •. (Cf. also ~\!i, ~\!i « -ait, -it- •, and 'OI't~ c -anta •.) .397. 151:~ « -anta,.: this is a stereotyped form, which occurs only in a few adjectives used attributively: e;g., ~~~ « jranta •>iSfJt~ « jyanta,. [f5rento] li1Jinq (jrvant-); 5'1~ « calanth moving (calant-); *vardhant-); "!_~~ « ghumanta" sleeping c,~.v ghum); ~'!~floating (1St'! ylbj:Jas .float); etc. It occurs in a feminine form '51"~« -anti,. = «-ant- +-ika,. used with sentient feminine nouns: ~ii-it < ~15~ « nacunti-at •· The feminine force is lost to Bengali. Examples: <:~iii"~ \St « pherat~ gak~ ,. return _post, also Wiif~ ~St « phir~ti Q_iik' ,. ("fii;iif y'phir); 5-a>!~ ! « bAh~tii jal~,. running water (~v'bah < y'vahjlow); ~~ « -ta •, and used as an adjective: e.g., ~t~tiif ~i{~ c~tt-9f~ « parag~-pak~e :r,- - vii::I}attia), *vaqQ.hattia); etc. But this derivation is rather problematical. Cases like i5t'<1•t~\!i « Q.akii.(i)t' • rob?m· (\Stt(~)~ c Q.aka(i)ti • robbery, '{;1~(~)\!i :> ,;m(~)it ·~Q.h~~(g)a(i)t~ > Q.ha~ga(i)ti • 1'0'H){ly > rou.;lines$ , ... --_,O"""JJ_---) ,_..,.,._ -~-... "'• \"'J ... \ ~~ ...,..£. ........ ......_ "'~""'"',-,..;"'-""'-'u.o......,u. \LV.L \'VU1\J1.1 btt infra), show that in examples like ~~~~ « uth~ti •, O::S1ff5 c bhar~ti,. etc. ~hove, we .have an extension of the \51~ « -at • affix ( < present participle c -anta •) · by . addition of tJ;re abstract W, ~ « -I, -i »: ~~f5 « uthlf.ti ,. = c utha.(n)t(s)+-i »the state qj1·ising=a rise. The. change o£ « -anta • to c -at- •, and not to c -it- •, would show that this affix is not native to Bengali, but borrowed, with cognate loan- words, from its western neighbours. . . , . It seems also that in the formation of this affix, ,there was. the influence of the Skt. nominal affix · « -ti » as well, through the presence of a large number ts. as well as 8ts.· forms in c -ti '., like « yukti (jugati), bhakti (bhakati), mati, gati ... praqati • etc. In post MB. times, • • 656 MORPHOLOGY: CHAPTER I the Perso-Arabic loan-words in « -at,., with the Indian or Persian « -I ,. added, re-inforced this affix: e.g., « wakalat,. > IS 'Si%i~~. 'SifiS!~~ « jajiya\, jajiyat-I ,. the work qf a J~~clge. (5) 'Ofif (lSi!', t!lil ) « -anf/. (-onf/., -enf/.). 399. This is a living suffix in Bengali, forming abstract verbal nouns, some of which have acquired a concrete sense. It is the OIA. « -ana ,., Examples: n] (jvalana); a cover (~il' « thakan~,. staying (11l't ,Vthak); C'if~il « dekban~,. siglzt; f:!~'\ « dhsral].f/. ,. manner(« I},. through Skt. influence) (dharaiJa); ilt5il' c niicanf/.,. dance (ift5 V"nac); ~ c paranf/. ,. fall (pagana < *pa~ana < patana); "P!t'Ofil, -Ptt~Sil « paan~, paon~,. (prii.pa!]a); -Pit'?il « paranf/. ,. (patana); c-Pit'?if ., poranf/. ,. burning c~ vpur < Vput) ; c~t'?if « pbOran~ ,. spiee$, chatter (vsphut); ~tf:!il' c Mdhan~,. (bandbana); -~1 c -ana > -~na ,. : this is simply 'Sf~ c -an~ ,. extended with the 1511 « -a ,. affix. The « -a- ,. in this affix drops when inter-consonan­ tal; whtn after a vowel, it becomes c e.,. or co • and forms a diphthong. The original•force of this c -a ,. was to denote simply a coarser or biggish aspect of the action, or to indicate a connected object, but this has been to a considerable amount modified, so that '5f~1, 0:0 « -ana, -~na ,. in many cases is not different from '01~ can~ •. Examples: ~nn < *tlf~1, ~t~ c kanna < Ud:j.na, kandana,. (see p. 360), beside ~tlfil c kadan~ ,. (kra~­ dana-) ; ~j c ku~~na. slicing > sliced vegetables (~~ vku~) j ~~i{j c ka~~n~,. spindle (~t~vka~ spin); <:~"1~1 « khel~ni,. playing> toy beside '~"!if c khelan~ ,. playing (c~o;v'khel); · '$\~i!1, ~~ c gahana, gayna ,. Jewels (grahaqa-) ; '$\t~~~ > '$ff'Sif1 « *gahana > gaona ,. singing ( '$\t~ v'iah < vgatha.ya-);~vch!k); ~1f~1 cjharfj.~a• water-fall (~~ v'jhar: see p .. 478); '[;f~1 « qhak~nii. • largislt tiel ('[;t debt, beside c~"t c denfj.,. giving (clfvde give=yida); so C"'il1 c lena,. beside C"'if clea~. taking (cil'f,Vle take gt·atis (marga!].a-); ~1l!1 < ~t~1 c ranna< rgdh.ana,. cooking (randhana-); ~~ill « sukh~na,. dry, rh'!J-lanil c~~vsukh < su~ka); etc. There are some formations from nouns as well : ~~~ « cha.na • ;ymmglitzg (if animals) (beside ~~ c cba :: = « saba- ,. ) ; 9ft~if1 c pakh~nli "' wings (beside 9ft~ pakh~ < pak~a). (5b) olt, ~~ (fif, ~f~) c -ani >-~ni, -uni (-~ni, -uni) •. Originally feminine in form (-ana+I), indicating a dimunitive or petty aspect of the action or object. All grammatical conne"tion with the feminine form is now lost, except whe;e the word indicates a woman. The change to c -uni,. is by Vowel Harmony (see p. 398). ~~fif < ~tlf~ c kaduni < kfi.dAnl ,. weeping, petty fit if u;eeping, supplication; SKK. ift~;{t < O{ll)o{T c nacun1 < nacani • danseuse ; c~~ « chen I.<« *cheani .. a cutttng instrument ( chedana, chedanika); ~ns~ < ~l'SfolT • chauni. < chaani • aw?zing, camp (chadanika); • 83 • 658 MORPHOLOGY: CHAPTER I ~~~ « ch£k~ni • small strainer (~t~ .Jcbik); MB. c91~=t c porani • burning(~~ \~'pur); ~1folt > ~~ c kurani > kuruni,. scraper (~1f .Jkur scrape); ~~~ c mauni,. < c *ma(h)ani ,. ekurnitt~g stick (*mathanikii); etc., etc. The word ~t~ < 1ft~~ c radhuni < ra.dhani ,. coo1c is now both masculine and feminine, but originally it seems to have been only feminine (c randhanika •: or a masc. =« *randhanika,. ?) ; when it means a kind of . flavouring spice, it is the feminine form undoubtedly. So MB. Cnet:t-fine atV;/f); t1f1 'c cora ,. ( caura-) ; tifjer-lilce,fm·ocious, strong (e.f!., 'tkiik~ra, kiik~ro • (kukku~a-); 11'~1 c Rami • contemptuous for c Ram:t •; &flfl c cada • big •round • JJwon, beside Stif c cad~ » moon ( cand ra-) : cf .. Stlfi lit~ « c!d-a mach~ » a kind .qf ?'mmdish fisk, pomfret; ~011 c thala • platt! beside ~ta; c thalEt » (sthala) ; ~~ ~"'1 c dhtia • (dhiima-); 1!'~, ~t~~;1 c ra:g(g)-ii • red in coloM (ragga-); ~a;1 c tel-a • oily ( tiila-) ; ca,tifl c Ion a • salty (laval}!t-) ; ~ c hata. ,. handle (basta-) ; ~t~~ c taw a ,. pot for lteatin9 or dry cooking (tapa-) ; [ dokhne] southern ( dak~il}a- > dakkhint1-) ; t'"l·C~t~1 « carl~-dhoa ,. rice-wasking ; ~t9f~- MIA. « -a, -ia • plus the preceding pleonastic affix '511 c a •. 'fhe following examples illustrate the line of development : « yata > jaa +-a > OB. *jaa, NB. <11~~1 *jawa • gone > a going; « krta > *karita (a non-Magadhi formation: cf. p. 375) >karia+a > OB. karia>~1 karii • done>iloing, as in New Bengali. So ~t~1 c rakha • kept > keeping (rak~ita +a) ; ~t'511, ~~m « khiiwa ,. (khadita +a), etc. We have the loss of the « -i- ,. already in OB. : e.g., Carya 20 « sa~ghara ,. struck; but it is the « -i- ,. forms, with or without the pleonastic« -a •, that are common in the Caryas (when witho'lt «-a •, the final « -a • is found to have assimilated with the preceding « i •, as follo;s: «-ita > -ia>-i, -i •) : e.g., in Carya l9 « uchalia,. (ucchalita+a), • « calia ,. ( cali ta +a) ; Cary a 13 « m unia ,. ( = jfiata-) ; Cary a 17 « sul'}ia ,. (=smta-), «milia,. (milita+t), gunia (=gat;~ita+ii.); beside c pahari • (praharita), « ahiiri,. (aharita-), etc.; and cases are found with the Western Apabhransa modification of the « -ital) ,. to « -iu ,. ( « -ita~} > -ido > -io > -iu,.: cf. p. 115 supra). The affix seems to have come into prominence only after the Early M B. period: in the SKK. it is non-existent. The c -i- ,. as in « -iii,. of OB. probably first became c -e-,. and then was drop;ed, following the forms in simple c -a,. like « yata+ -a> ~~~1 jawi •, « *dita+ii. > C'fiS~i dewa,. etc.; and possibly also through an attempt to distinguish this passive participle affix from the conjunctive participle in « -I(-1), -ia,. (e.g. ~t~, ~tf~, ftf~'5!1 « rikbi, rakhi, rakhia.,. having lcept, with or without « -a ,. < MIA. « rakkhia. ,. < OIA. c -rak~-ya ,. }. The OB. forms of the passive participle in « -iii. ,. show that « -ita+ -a ,. is the source of this passive participle affix of MB. and NB. In Pa11jabi, in the Old Western Hindi dialects and in Modern Hindostii.ni we find a parallel development with loss of « -i- ,. : e.g., c calita-, calitaka • > MIA. (Saura- - seni) « cali(d)a-, caii(d)aa- •, uominative . ., c::.!i(d)6, ea!i(fl)ao • > Saur. Apabhransa « caliu, caliaii •, whence Brajbhii.khii. « calyu, C'alyau •, Early Hindi or Hindostani « calya ,. (beside Panjabi « calia, cii.lea • ), but Modern Hindostani « cala •, Bundeli and Kanauji « calo •· Oriya has this affix • also : e.g., « jaf!J-a ,. known, « kar-a • done, « dekh-ii. ,. seen, etc. • THE SUFFIX '-AI' OF VERBAL & ABSTRACT NOUNS 66! It has been attempted to connect this « -a» affix in its nominal employ with ~he verbal noun affix~~. ~~1 « -iba, -iba • < « itavya+a,.; hut that is hardly jienable. . . . . In OB. and MB., there was a feminine form of this affix, with the ~ c -1,. ( < -ika.); e.g., SKK. (p. 381) iSJ~"t c jarr ,. Jewellerl qualifying~ c curr ,. bracelet (jaQ.ita + -ika). (8) 15ft~ c -ai,. (i): Verbal and Absti'act Nouns, 402. This is a Late MIA. formation, being a feminine verbal noun from the causative stem of the root: an OIA. c -ap-ika ,. (as in Skt. « jfi- , ap-ika,. information, whi_ch would be the basis of a possible NIA. « *jar;tai,.) wouid give c -avia, .a.via, -aw1, ai > '!>ft~ ai ... The causative in.c -apa > -ava,. became established in MIA., .and on the model of the noons in « -apana- > -avaz:a- • (see infra, § xxx, pnder 101101 • -ana ,. ), formation of abstract and verbal nouns in « -ika ,. from the « ~apa-> -ava- ,. base would be a matter of course. Examples: lltl5t~ « jacai ,. enquiry (ll115 vyac: OIA. *yacapika) ; ~cit c duli ,. [see p. 352]; aJso meaning the same thing); ~tG'I~ .. jhalai ,. solrlering (~ta stolen property ( sweetmeat ( < .mithii. < mit:~ta-); 9f1G'I~t~ « pii.l~tai • ea:change, . reiu1·ning • (palat~ = paryasta) ; MIA. «-avant-, -aant- • > OB. « -ayjwanta •>:MB. 'Ol't~\5 « -a1t8. • : ;;stq'it~\5, 15tqo;t~ « Q.ii.kaittj., Q.akattj. • . robbt;r < shouter (;;stqo; v Q.ak yell) ; MB. l!ift>S « *-iiii>ao ~: Verbal Noun Affix. 405. This affix is now confined to a limited number of words; and m those ease~ where it is found, it seems there is reinforcement from • similar Hindostant forms. It is connected with l!lt~ « -ai • (i) above. 'l'h~ source is apparently a verbal noun from the causative base : « -iipa •, with • an extension « *-ap.:uk-a (·iipa+-u-ka+-a) •, corresponding to •.*-apika • the source of .tta > -a •, a derivation which is inadmissible. Examples: ~1S~1 « -iiua>-oa,.: Adjectival as well as Abstract. 406. This affix is an extension of the preceding one with the adjec­ tival or pleonastic 1511 « -a ,. (see ante, § 400) =e.g., '<1~1~1 « gharoii ,. homely, domestic [gfioroa]=c gharao, *gharii.u+-a,. (this word shows extension of the affix to a noun); ~l:~t~1 « ja:roa,. jeweUed, adorned witk precious stones =« jarao, *jarau+-a,.; ~1!:51~ « Mcoa,. safety, safe state=« Mcao, *Mcau +-a ,. ; 1)!:~1~1 « caroa ,. aggressive=« carao, *carau +-a ... • • 407. This is not so much an affix as an independent word, phonetic­ ally decayed, and forming part of a disguised compound. It is from <~>n, <1'1~ c ka~h~, ka~hi ,. = c kii!?tha, *ka!?~hika •, wood, beam, stick. '~tt a lce!J ( C~t<;J V chor =l~:t<;J V char); C"ftiltil c son ana • causing to hea1· (MIA• *sul)aval).a = sraval)a); ')\~1"1. c garan~ • slope, $!~til c garana,. a rolling, to roll (MIA. *gaQ.Q.); etc.' 'l'here are old and recent formations from nouns also: ~~til, ~\!itei!1 c jutana, jutano •, also in Calcutta ~\!itt'i11, i!f.~t~1 c jutono, jutuno ,. shoe-beating (''!i1 « juta • ,~hoe); sts. C~l$11"\. c jogiin~ • aiding 1oith materials, c~t$ftil c jogana • act if supplying (yoga); a'ltQfttfare paz'd in crossing (*pilrapanika) ; nzoue.lJ paid foJ' a sight ( = darsa.ni); C'!iiG'!lTi! as in C"fSi-C\!ila'!t1il. « sej~-tolani ,. money paid for making or clearing the bed (~"'vtul raise); fil<~>tfil c nikani,. beside fil~f-l « nikuni • liJiping, mopping the JlQPr Cfil<~> vnik ~if, ~~1 « -imil., .-ima •, which Is found in Benga,i, as in other NIA. The source of this affix is an OB. -kamma > -kama > -amil. >15ft~ -am~; ka_rma-ka> -kamma-a > *-kamawa > -ama > 15ft~ -am a; karmika, karmika > -kammia > *-kami > .ami > l:!l'tfif -ami ... The strong initial stress, as well as M B. mutation, has brought about the loss of the « -a- ,. in IMB. and NB., and reduced the form to lf., ~'fie« -m~, -rna, -mi.• etc. In M B., the word tf~~1 = <~>tf~f-1~1 ~ kamina, -inya < -iniya ,. occurs, meaning a labourer, and a form 15ftf~i{T, 15ftfilf.l « amini, amini • also is • found (as in the 'Sunya-Puraqa ') : the latter word, explained as meaning a female devotee .if Dharma, may be connected with <~>t~ « kam~,. w'or1, ·service. Examples : ~~t~ « thakama ,. cheating, rogztery (~<~' « tha.k~,. cheat, 'tl!Jtfl '); 9fttf~ « pakama, -mi ,. behaving like a 1·ipe (i.e. aged) person, precocity (911<~>1 c paka• ripe); so C'Si~t~ c J-ethama,. behaving like a respected elde1· uncle = impntinence on the part if a boy or ?JOuth (C'Si~ « jetha,. father's elder b1·other) ; Cift~ c phil.c~kemi, phac~­ kama,. (~~f~~1 « phac~kiya,. chatterbox, 01m-smart fellow); f~D,G'ft~ f~~G'Jll', f~~"'tf~, f~~~f~ « phic~lallJa, -lama, ~!ami, -limi ,. acuteuess, triekery (Risl)t"' c phical~,. an acute one); c9fCSit~, c9ftiSitfil c pejoma, pejomi .; vicionsness (9ft~ ! paji ,. mischiez•ou; character > 9ft~~1 + '!iltf~ « paj ua + -ami»); c'i'ft~nl5!>tfli, -~f~ cg8artaini, -tami .. churlishness, rasltne88, brusquene.~& (*c'i'ft~t~~ « g8ara.t~,. one oehaving.like a cltud [see p. 668], with «-at~,. affix, see §397, + « -ami,.); Citl;c"lt~fil « chota-lok~-mi ,. meanness; etc., etc. 'fhe word 't;rt'l!f1 « u~ari • brass 01' bell-metal worlce1· (kansya-karika); ~~t"lf « kumar~,. potter, by Vowel Harmony ~t~tlf c kumor~ • (kumbha-kiira); 5t~t~ « camiir~ • (earma-kiira) ; •it~t~ « sitkhar'i • (sankha-kari-) ; r~~t~. f~f~f~ « bhikhiiri > bhik(h )iri • beggar (bhik~a-kari-); ~~t"lf1 c juari • (dyuta-kari-), beside ~~%.t « juari • (dyuta­ +v'vrt); MB. t~nr « kamar~,. bloelc-mtitlt, in Pali « kammara •, represents the Vedic « karmara • (which is probably connected with the Dravidian [Tamil]« karuma • smith, smelter: cf. P. T. Srinivas Iyengar, 'Life in • Ancient India, etc.' Madras, 1912, p. 15); and the Skt. >jfngly connected with ~ « hahn ,. = c vadhu •, after . the 'l'HE A:PFIXES '-ARt, -ARU, -AL(A)' . 069 metathesis of the « h ,. took place); ~1~t~. c majhar4,. m£ddle, ~~tit « m!.jhari ,. midclli1tg-sized ( c madhy~kara •, pleonastic; c *madhyAkarika ,. ) . • (19) 15lt~. ~t~ ( <5ttf~ ) « -ar4, -ari ( -ari) ,. (iii). 414 . . li't·om c -agii.ra,·-agarin (-agat·ika., -ika),.: ~t~t~ c bMrar~,. store (bhiit}~A.gara); MB. *t~t~, beside <~>t'St~ c kllrar~, kiil)~ar~,. keel qf boat (kary~Agiira); :lftl5t~ c sabhar4,. plaee-name (? sabhragara); c~~t~ c mehar4 ,. place-name (? mahi\gara) ; ~t~ti'T « bMrari ,. 8tore-1ree_per (bhary~i\garika); <~>t'Stjt c kary~iiri ,. helmsman (kiiQQ.agii.t·ika); c~it c mehari ,. palace (? mahligarika) ; etc. (20) ~t~ « -aru •. • 415. Found in a few MB. words, seems to be from c -\ra,. (i), with addition of « -u ,. (i.e. c -ii < -uka ,. ) : fit"'"lt~ c disiru ,. pilot beside flr-ttjt « disiiri ,. ( disa = dik- + ka~a-) ; ~t1l'l « p~cal4 • with t1lJists (CS$ c p~c4,. < Pers. c pee ,. ) ; 15f&t~t1l'l c bha~iyalq, ,. belonging to the rlo1lJn-count?'!J > a follc-melorly ( < 15t~ c bha~i ,. flrJ'w qf toater=Sout!t Bengal); t., cbacal4,. (vacala); frt~tll'f c d£tal(a),. with (big) teeth (danta+-ala-); f~t"'l c cpinal~,. charal!terles.y woman (MIA. chinnala-); MB. 9ft .,1 c -al-a > -Ia ,. : e.g., . rupoli > rnpuli • Bilvery; c~ti!t~ « sonali • golde?l; fi{ift~, f.lil'ft~ « ni(n)dali • 8leepine88; ~~til'!) « sutali • thin as a thread ; etc. . (22) 15!til'l .. al;t • (ii). • 4i 7. This affix, like the preceding one, denotes connexion. It is also used to form words indicative of trade or profession. It comes from OIA. « pala • keeper, guardian by the 'tlropping of « -p- > -w- • : « -pala >-wala> -al;t •. Examples : 1ft~t"' « rikhal;t •, M B. ~tr.:~l15!til'l « rakhoala • he1·dsman, (rak~a-pala) ; cr.:"l"' < ~it"' « Kesel;t < Kaslyal~ • low class Brahma'(h from Benares (Kasi-pala); C a clerk • (kuthi kothi < ko~~hika); ' ' · gay la • (gopala-) ; -OLA, -WALI>-ULl,-WARA; -I(I)' 671' ·' The « -al~, -al-a • of Bengali has within receut times beeri influenced by the Cj)gn~te.form in Hindostani, which ·is « -wala •: the latter is substi- , tuted in maqy Bengali woi·ds in « -ala •, and ordinarily ·., -wala,. is reduced to '€1.'11 « -ohi,. : e.g., <~'t?f~'S~fG'Jt, <~'t9f~'€'-'!1 « kapar-oala ( -wala), -ola ,. ; so o;tt1t'€f.'l1 .. gari-ola ,. cabb.'lf; 'Pft~t~'Sf.'lt c pahar-ola • coit8table, ·etc. 'l'h~ feminine of « -wala ,. _is «. -waii •, which, ·by . Vowel Harmony1 is .reduced to IS~ c -uli ,. in Bengali : e .. r;., <~t1tiS~ c bari-uli • = native . Bengali <~t~11!{t.-ft .. hari-aii •· 'l'hrough Bihari or Rindt influenCP., we have also '€stt~1 « -oara= -wala ,. in ~1?:~~1~1 « matoara (ma~wara),. zo£ltl, mad, elated, jJleased (e.,r;., tcith" peifum.e) = c matwara •: the Bengali froms are ~t~t"!, ~t~?.i, ){f~ii'l1 c matal~, matal,, mat~la ,. drnnlc= Hind. c matwal •. The name NIA. mali ,. gardener:· but the force of the nominative form seems to have beeu only of secondary impOI'tance in the evolution of the NIA. affix; (ii) « -iya • : « desiya > NIA. desr • n~tive; and (iii) « -ika •: . « gramika • > 'iff~ gai ,. nillage name, clan na1lle : the feminine form of this « -ika • is also ~. ~ « -'i, -i •, which is the most popular. feminine affix of NI1., f<:>r which see .below. Later in NIA. this affix was strengthened _by the « -I • of Persian. Examples: c -I > -i • w~th the sense of ' possessing, having' : 'St~ « bhitr-i • heavy;. sts. ~t~ « rag-'i •full of passion >c!t.ole1'ic; ift~ « diim-I • costl!J (dramma < Gk. drakhme); ift~ « dag-i• stained (Pers. diiCJ-); ~~ « gug-'i • man qf part~, a1·tist (cf. Skt. gu!)in); '!;t fwfit.• desi > disi » beton,r;ing to tke conntry, countty-bom or made, native (cf. Skt. desiya); ~tl?) c Rarhi ,. betong£ng to Racfha; M B. ~~ « -ui ,. : barui ,. carpenter (cf. Skt. vat·dhakin); barui ,., 08. name for gi,.Z., (k!Judrika); "l}avi). A number of sts. forms show this «-I » of Bengali, e.g., ~~o#t . . - « subacani,. a form qf nurga; <51~~) « apsari,. (apsaras); fW'if'nll « digambari .. a name qf Kali; ~il~i!) « sunayani ,. with beautij'1d eyes, etc., with are solecisms from point of view of Skt. n The Skt. feminine affixes « -ini, -ani ,. had an influence in reitricting uthe Bengali ~. ~ « -I, -1 ,. < « -ika ,. : see infra) under « -ni •. . With names of inanimate objects, the original feminine force is lost ft.­ in Benga~i. In ·Western Hin soldier, rough + -i) ; 9\fu~') « paQOt-1 ,. belonging to a Pundit > a Pundit's character or ways, scholasticism; etc. Cf. Hindi « Hil-i ,. redness, « hariyal-i ,. green colmw, g1·eenness, « rakhwal-i • work qf a lzerdsman, etc. In Late ~IA., the Persian « -I ,. affix of abstraction or fionnexion greatly reinforced thilil suffix. • • 85 674 MORPHOLOGY : CHAPTER I In some stereotyped words, « -i ,. represents the OIA. « -i ,. : e.g., "lftott, ~t9t « rani "' ( « rajiii "' : or < « rar;u;tia < *rajiiika ,. ?). • (25) ~ « -i,.: Verbal Noun. • 420. This affix, occurring in OIA., is practically lost to NB. It is found only in some epenthetic forms in dialectal Bengali, and it has practic­ ally become a dead suffix; only some S kt. words in « -i,. help to perpetu­ ate it. Examples : 'i'fti;, 'i'ft~l; c ga(i)t ,. (granthi) ; "lt"lf, 'iit~~ c ma(i)r :. a beating < 'iit~ c mari "' (*mar-i); ~t~ > }'f1(~)~ « sari > sa(i)r "' row (sar-i); MB. ~t}'f « hasa"' ( « has-i .. , which occurs as a ts. in NB.); '5ft1if, '5ft~"'l c ga(i)l "' abuse (*garh-i ? c£. Skt. garh-a); "1'1(~)"'1 c sa(i)l ,. a kind qf, paddy ~sa.li), also occurs in the ts. form in c -i,.; ete. Cf. also the OIA. sufM « -ti ,. in ~1 (~)\> « ~a(i)tJ .. , MB. ~~ c ~athi ,. (~a~-ti), }'f;:e~ « sattar ,. (sapta-ti), ~~. sf~~ beside • the full form s'ftf~ « pa(i)t, p[ti ,. row (pa~k-ti), etc. (26) ~\51'1, ~~1, <£~ « -ia (-iya) > -e ... 421. This is a compound affix, being from the OIA. « -ika ,. > MIA. « -ia ,. + the affix « -a •, as in § 400. It is used in forming adjectives with the sense of ' pertaining to ' or ' possessing' ; and frequently it is•used with a slightly disparaging sense ; as well as pleonastically. In NB., it is contracted to <£1 « -e "'• with accompanying modification through the « -i •, of connected -.:>wels. Examples : ~tf0'!~1, C~CO'f « kalia> kele,. the blade one (contemptuous) (kala ~lack); ~1ifnt~1, ~GJ[If~1 > ~''l?:l'f ~'"'!:.?:If c Mla.dia, haludia > h5lude, hOlde ,. yellow (~~ « ha.lud:1- ,. turmeric); ;qt~'5ff'1~1 > c;q~c"i « baigar;tiyii > begune "'pnrple, qf egg-fruit (vati~gana) colour; ~tf~1 > C~CO'f c jalia > jele ,. fisherman (jala net) ; ~tf.1~1, C~CO'f c haliya > hele •farmer (hala = hala plough); '{'5ffi-~1 > 0{~~ c nsgariya > nsgure ,. man qf the city ; c bene ,. merchant (*vanika- = vanik): "lt~1 > c~ii; c matia "> mete,. earthe-n fmii.t.i < ... ~ = ~ • " ·- • - .. .. ' • - mrttika + -ika +-aka) ; (in the word c~ as in c~cl> w~~ c mete phiri~gi,. a Eurasian, a ha{f-caste, we have, however, the French word c metis ,., approximated t., the Bengali word• by folk-etymology); @;:e~~ > .iS;:e,c~ THE AFFIXES '-IYA>-E,· .. JYE' 675 c uttariya >utture:. n01·them; ~Nfilm>~f~ti{, ~·~,ti{ .. dakbiniya>dokhine, dokhne'• southern ; 9Jf pUbe ,. eastern ; 9ffi"sf~m > 9!f"15~ c pMcimiylt>poscime,. western; MB. 'Ofl\Stf"lm· « abhagiya • > NB. 'Ofl 9!t'¥11:'>k~ c parag[iya < -g~ye » rural (pataka+ grama+ -ika +-aka); ~"1~~1> *t!Oti{ c Mdaniya>k[dune,. ready to weep; M B. *Co:{~~-~t~f'¥~1 « *neuy~-!lkariya ,. > N B. Ci!~-~t~N. « nei­ [kure,. one who cUngs to argument (nyiiya) > obstinate; i!T51~m > "i!t5\~1:~ « Nag~puriya > Nag~ pure • qf Nagpur; ~%."~1 > ~~. ~~ « Oriya > Uriya > Ure,. Oriya, !'!! the 0(/,ra country; MB. (SKK.) "i!t~m « natiya .~ = "il~ « nati ,. grandson (naptr-); ko(h)une • talker; iSlT'5tt"R~1 >iSit511ti! c jaganiya > jagane • mze who ,wakens; Cif'61'f.f~1, N~f-i~1 > ~i! « deaniya, diyaniya > diune » a give1·; etc. • Although one of the most common a.ffixes in Bengali, it is frequently disguised in the Standard Colloquial by'contraction and by mutation of vowels. (27) ~c!!, ~?:~ .. -ie, -iye ». 422. This NB. affix denotes 'one who is accustomed to do a thing, a habitue.' Typical words are ~~'!! « khaie • a (good) eater, o;tt~?:~ 4J. gaie ». a good singer, a prf'!fess1:onal singer, <~tfiSfC~ « bajie » a clever 'musician, o:t'W?:'l! « nacie » a good dancer, c~f<'f1:~ c khelie ,. a clever player, fi.r~ c karie,. an adept, ~l:~ > khaaa, khawaa, gayaka > gaaa, gawaa [*gathayaka > *gahaaa, *gaheaa], *vadya-ka [ =vadaka] >*vajjaa >*bajaa, etc.), to which the « -ika > -ia • and then the c -aka > -aa, -a,. affixes were added, pleona,stically, in the Early NIA. stage: c *kathaka • +-ika+-aka • > c *kaha.a'iaa • > c *kaha'iya • (Hindostanr c kahaiya •) • ~ v > <~''<.?:Sf« ka'iye • [koie] . • (28) ~~ « -it~ •. 423. The affix~~ « -it- • is derived £rom the OIA. present parti­ ciple base « -ant- • (see supra, pp. 653-4). In Bengali, it forms the present participle in an adverbial or conditional sense, with the locative affix « -e • : e.g., Clf'f!if?:~ « karite karite • ( < *karantahi) wkilt- doing; and it gives the base £or the past habitual or past conditional (and in some East Bengal dialects, £or the future) tense. See also supm p. 663, under i>lt~~, ~t~ « -ait;j,, -at;j, •. (29) ~<( « -iba ,., 424. Base £or the future, and £or a verbal noun, with an extended form ~~ « -ib-a •. From OIA. « -itavya- ,. > MIA. « -iabba- • > c *-ebba, -ibba •. See infra, under 'Verb.' (30) ~., « -il;j, ,. (i). • 425. In MIA. « -ilia •, from OIA. pleonastic and adjectival affix « -1-1~.:. ('}t~ Pii.l'}ini_. v; iii; 79).7 forming DOUDS a!!d a-djectives .irl Be~ge.!i, from nouns. It is not a living affix in NB., but ·is fairly common in M B . • C£. Hindi« chaji • handsome, accomplished (challla, chavila). • THE AFFIXES '-IL, -ILA, -U' 67'i Examples: ?Jt~ « patil~ » an earthen-pot for cooking (patta, patra); tf15G'f l kucil~ ,.. crooked Ct15 vkuc < vkufic); sts. t'SftfuG'f • aq.Q.il~ ,., strong one, ch,ie+' ~J?.Qa +.ina) ; MB. sts. '5lt~f~~ « arati-la,. jillecl ~vitk a great desire (arati + -ila); 9ftf~G'f « pakila,. ripe (pakva + -ila); etc. Cf. ts. ~!SG'f « kutil~ ,. crooked. The ts. word « sarira " is vulgarly pronounced as -t*G'f, -tf*-ilia-,., in Eastern Magadhan. (OIA. « -(i)ta+-alla,. probably gave« *-alla- > -al~ •, in Central and Western Magadh;n)• See below, under' Verb.' This« -ila »occurs also in a few nouns and adjectives in Bengali, although the adjectival and neminal use of « -ila » forms fro~ verb root~ is almost obsolete in the language. Examples : MB. ~f~ « bhukhila » knngr,z; (bubhukf)ita+-ila); ~f~G'f « sutila »sleeping (supta+ Lila); ~Stf~G'f « bha:ugila ,. brolcen (*bha:ugita + -ila); OB. (Carya ~3) ~~ ~~ « dubila dudhu »milk th,at has been nzilkerl o.ff (*dubita+ila); NB. C~ « gel a ,. last < gone (gata + ila) ; etc. The affix ~CG'f « -il-e » forming the conjunctive, i'l the old locative form of it: 15fG'fr:G'f « calile » on hav£ng walked (*calillahi). • (32) @ < ~ « ltU < -fi » (i). 427. From an OIA .• « -u,. (adjectival and nominal) +« -ka ». It implies a pettiness and lovableness of the object indicated, and is found in familiar and affectionate forms, specially of names: e.p., sts. ·~~ « du~tu,. ... the littte naughty one ( duf?ta-); sts. ~"' « dhuttu » t!te little sly one ( dhurta ) ; \%."' « garu,. water-jar witk spout (*gaQ.~uka); ijj~.: ~~ .... « laru, naru * sweetmeat (laqcjuka) ; ill"~ « nicu * down, ~n a low posttwn (nica); ~~g « ucu,. hi{Jh (ucca-), etc. The« -u • affix of Saur!serii Apa­ bhrarisa, from OIA. « -a~ .. , seems to have influenced this affix in the formative period of Bengali (see supra, p. 311). (33) *~, ~ > ~ > "5f .. *-u > -u > -o > -a* (ii). 428. This is a primary affix, and the forms with it are generally repeated in N H. the meaning is 'about to do someting,' 'on the point o£ doing.' .E.g., ':R~-':R~ « (_lubu-qubu * about to sink; ~~...:-~ ..... « uru-uru,. about to flY away, filled with a longing; fil<(-f.:l~ > filc nibO-nibO • on the point being extinguished (also f.:l~-fil~ « nibha-nibha •); . . ~l:lf1-~lf1, 1-?ftt~ « pako-pako » nearly• ripe; lft;s.lft:S ~·1:~ i!!&'i!'lt c dau-dau kore jwala • to bum fiercely, etc. The form ~~ c habu • [liobu] about-to-be, future (e.g., ~1:-~t~t~ c habu jamai • the son-in-law to be, ~'l:"~t~1 «hahn­ raja • the future lcing) seem~ to be for *c~t-~ c *ho-u •, with euphonic insertion of « b •: similarly ~t~-~t:S> ~tl: khabo­ khabO kara • to be eager to eat, etc. (Cf. ~l: spendthrift, « bigar-u »spoiler, « jhagt/.r-u • quarrelsome person, e\c. Hoernle connects it with the Skt. c -tr » + pleonastic c -ka • (op. cit., § 333)_. Hut a likelier derivation would be from an OIA. « -u-ka • (c£. J. T. Platts, Hindustani Grammar, p. 209). This affix is found in Skt., in words like « ghat-uka, kam-uka., bhav-uka » etc., 'having the meaning and construction of a present parti­ ciple' (see Whitney, Skt. Grammar,§ 1180). There was also an « -uka • rather rare, in OIA. : c£. Skt. c jagar-fika •. • Oriya has the present_ partic!p!c affix « -u ,. ( « kar-u, kha-u, ja-u ,. etc.), which seems to be the same affix as the Bengali and Hindostii.ni ones ( H oernle derives the Oriya c -u •, however, from the present participle . . c -anta •, op. cit., i3U l : this is precluded phonologically). • THE AFFIXES '~ UA,' AND '-KA' (34) @ is-s, -s .. -ua > -uo, -o •. 679 429. This is an· extension of the preceding with the c -a ,., used, like « -iya l (p. 67 4), to indicate relation or connexion. It is also added to proper names, to indicate contempt. When affixed to an ordinary noun, it turns the latter to an adjective. There is the usual mutation of the vowels, and this fact often disguises this affix in the NB. Standard Colloquial. Examples : iSi~ > iSi'~1 « jal-ua > jolo ,. watery ; ' c ghOro » domestic; ~t~~1 > C~'CS1 ., hat-ua > he~o ,. qf the nwrket; ~~1 > i!lil'CSI1 « j(w)ar-ua > joro ,. suffering from fever; MB. ~~~~ c sar-ua ~ like a reed> slender (NB. 'I'll',.~¥ « saru, saru •lean and thin< .. sar:t+-ii •); ~t~~1 > <:~il1 « dhan-ua > dheno ,. (i) (land) growin,r; rice, (ii) 1·ice-beer; C kethO ,. made qf wood; wt~~1 > Clf'Cifl • .. dan-ua > deno,. given as a .r;ift ; *$ft5!SI't~5!1 > c~~1 « *dia-rakh-ua > der!rho ,. a wooden o1· earthen lamJJ-stand ; *- C ke~ko ,. a wooden vessel; ~1~1 > C~1 « ~ak-ua > ~eko ,. bald-pate; . so Ofr:~1 « mecho,. fishmonger; ~t~..__~1>cbeto,. belonging to the road (bata < vartman) > a hack; 9Jt~~1 >C"!Cil1 « Pan.ua > Pen6 ,. co1ttemptuous or familiar form qf a name ( = Pra~1a-natha, Panna-lalq,, etc.), t~~j > C Kelo ,. (Kala-), c~c~1, c~c~1 « Remo, ~emo ,. (Ram a, Syama); ~~1 > t~'Ciii • Jadua > Jodo,. (Yadu), etc., etc.; *~t@>;:'(~1 > ~f.:#fi « *maus-ua > meso,. mother's sister's ltusband (« mausi = matrl(lvasr-,.: cf. "*f?f\;f~j > f?fC'I « *piusia > pise » fathe1·'s siste1·'s h1ts&ancl); etc., etc. (35) , '5!~ .. -ka, (-k~), -ak:t .,\ .. -ik~ ,., @ .. -uk~,. (i). 430. This ~ffix is quite a common one. in Bengali, and is employed primarily to form nouns (abstract, concrete, or onomatopoetic), generally from verb roots: e.g., <:~t~
  • remembranc~ ("il tan~ sense of tightness, ~til v ~an pull) ; ~ swinging by hooks from a ltigh post, the kook-swinging festival ( 5~ v car ride); ~"1'<1' « chal-ak~ ,. splaslt, spill of water (~01 chal onomatopoetic (oi' splashi1tg qf water) ; .so 5"1'<1' « cal-ak~ » spilling qf .nwter; ~~~ « mar-ak~ • pestilence ( « at-aktt • confinement ( cf. <5l'~ artt obstruction); « jhal-ak~ flame (, "t~" « jhil-iktt ,. sparkle; fil:fi!'f .c chil-ik~ ,. spilling m· scattering qf water; ~" « thaktt, thak¢ • rema£n, be arrested (stabh-krta), beside ' c naQ.-akk-ai, naQ.-akk-al, naQ.-akk-udal ,. wal!cinl; c -/iru .. to be >« iru-kk-ai,. being~ «-/say,. to do > « sey-gai,. doing, etc., !ts in Tamil, and « vuJ?.QU ,. to 'eat > « Ul)I)ike ,. eating, as "'in Kannada (Bloch, 'Langue Marathe,' p. 105; Caldwell, 'Comparative Grammar of the Dravidian Languages 3 ,' 1916, pp. 542-54:3). In its origin, it would thus seem that this affix is derived from « krta,. and other forms of« vkr '"· There seems also to have been ·present the influence of the Sld. affix « -a-ka ,., which was Prakritised into c -akka •. And it is not improbable that in MIA. times a « -kk-, -g-, -k-,. affix existed in North Indian Dravidian dialects (now superseded by IA.), and this affix of Dravidian was not without its influence in the es,ablish- rnent of « -akka ,. in MIA. of the Second and Third stages. • The~~' ~ « -ik~, -uk~ ,. forms seem to be due to Vowel Harmony, the c -a- ,. changing to « -i-; -u- ,. when th~ root or base form had thest> vowels, by attraction. Extensions of , \51 c -k~, -ak~ ,. in Bengali: (151)~1 "-~ka •, i.e. c -ak~ + -a,.: adjectival and pleonastic, with an implication of biggishness or eoarseness: "i~~1 .. pa~~ka ,. cracker, ,yquib, beside "!~ .. patak~ ,. a th.uno-in,r; down w£th, cracking noi.Ye; 'D"t1 .. hur~ka ,. a bar, a bolt; etc. 151~1, (151') f' diminutive >pleonastic in NB., generally names of concrete objects: )'f~~ c sar~ki,. javelin (that which moves quickl,'IJ anrl .Yt1'£kes : < srta < v sr) ; C~'Df c h(3c~ki ,. jerk >yawn, hiccnp ( cf. C~'D~t'i' c h~c-~-ra-na ,. pull) ; ~l>f' c tarak!J. ,. witlt a .Yztdden bound ( < trta), "'~t' « sarak~ ,. with a quic/IJ!jglide, '<1"\!?t~ c karak~ ,. witlt a r:rru:k?infi sound, etc.1 vllbich are mostly 86 682 MORPHOLOGY: CHAPTER I used with the indeclinable conjunctive ~f w1 .. -:tka. ,., ~ .. -:tki ,., fw~1 .. -kiya .. , ~~~ .. -kua ... . 431. This is an adjectival affix, also used pleonastically : it is also found .in the strong form in <1'1 c -ka •, with its feminine 't, ~ « -ki, -ki•, and also in its extensions formed by superposing other affixes, ~m « -iyi'i :o, ~~1 c ua • etc. Examples : ~1, C'!t"f1 « am:tka, tom:tka • qf us, qf you = onr, your; 'i1lf1 « gad:tka • Indian clttb, cudgel ( < ts. gada); 'ifl!<\1"1 « dam:tka ,. a rnslt if wind, a strong puff if wind, beside 'ifl!'t « mej:tki • < « majhua + ki ,., c~~1 c chO~:tkr •, etc. ; CW~1 < *t~ « ket,ko < *kath-ak-ua • a wooden vessel; ~r:w c put:tke < *piit-ak-iya,. a tiny ~ne (cf. ~! « piiti ,. a tiny girl, a small fisk); '"\.l: c ser~ke < *ser-ak-iya ,. weighing a seer; ~w « sat:tke .. < *~fwm « satak-iya .. plain numeration table upto 100 (from *~~ « haibe-k:l-,. it will be; ~~~ « kar-u-k:l- ,. let him do; t~~ « khale-k:l- ,. he ate; iltf~~ « nahi-k-ay ,. is not, beside Eady Assamese ilt~<~>~ c nahi-k-anta ,. they are not; and Magahi' « y'hi-k •, Maithili « chi-k ,. to be. · These last instances would show that « -k:l- •, like the « -r- • affix (iii) below, was added to the root : probably here was some confusion with« -k~· (i) ( -C'i!t~ c majhari-gocher:j., -giich:J. ,. about middlin.fJ-sized; C~ti¥1-C'i!t?:~q > -C'5tt~ c ch5ra-giicher:J. > -giich:J. • 1·ather like a lad, yo11ngish. (39) s, 'Olts c -~t:J., -ac:J. », 434. In a few words like <~~"t~ts c kanac:J. ,. belonging to the edge, edge, c~t'fts c kii9ac:j. ,. belonging to a corner, aslant, ' -r- ,. suffixes (see below, §§449, 450), which seem to have influenced it in its pleonastic usP. It represents a MIA. c *vaH- •, from t.he OIA. formations c varta, vartika, va.rtika; vrtta, vrttika, vrttika,. from c vvrt •. This affix occurs in Hindostani as« -awa~, -aha~ •, and in other forms of ~IA. also. 1) « -:J.t~ ,. < c vartta ,. : abstract or concrete nouns: 'ift9f~ c dapM~ ,. power, puissance, rule, high-handedness (darpa-varta); ~t9f~ « sapat:J. ,. being .in a fury, indulging in heroic gestu1·es, seizing boldly (sarpa-varta); ..Sl;, Jt..S~ « keot~ ,. [kreot] (kevatta < kaivarta) . . \; « -~t~ ,. from « vrtta ::, forming adjectives and nouns : c~11ftl; c dh5at~ ,. smoky, mass of' smoke (dhiima- > dhfia, dbaa + ~t~ < vrtta); c-r-•, see p. 689) ; ~t~\;j « jhapl}',ta » spray qf water, an ornmnentfor the head w~th tassel of pearls (cf. c jhapat~,. above); Co:{IS\;1 by Vowel Harmony ~\Sr;\;1 c neota ,. > [ nreoto] affectionate, attached (*nehat~a < l}eha-vaHa-, sneha­ vrtta-); ~~,\;1 « gha~~ta,. getting rubbed, 01' rubbing continuously, as opposed to '\)1 c uk~ta,. 1·ummage, eagerly search (*utka-vartae); 't_\;1, !>\;1 « kas~ta, kac~ta ,. rinse (kar~a-varta-) ; C~OlSi « khem~ ta » a kinrl qf style in da1wing and 8inging (? + varta); c~~\;1• c ne:gta ,., tc•m'ijj [nre:gto], calli!'\;! c le:ggl}',ta,. nalcerl (*oa:ggawat.ta-, naggavatta-, nagna-vrtta-); 9fr~\;j « pitsuta ,. ashen or grey in colour (pailsu-vrtta-); fDlj_\;1 c cim~ta » tongs; pincers (MIA. ? *cimma+vrtta); etc. t;'t, fG' « -ti, -~i ,., generally = the feminine form of the « .t- » affix . ( < -vattia = vartika, vrttika). In MB. and NB., the feminine force is absent, and c • ti, -ti » has become eithPr merely pleonastic, or vaguely diminutive: e.g., ' ~·~Wt « *siikhat;I > sut~ki,. dried fish (? su~ka-vrttika); ~~ •« chip~ti » wltip (k~ipra-); similarly we have this affix in words like '1_il~~ « khun~­ suti » biclce1'ing, picking up a quan·el ; etc. ~~ « -ta » and ~. Js « -ti, -ti » have given the so-called post-fixed 'Definite Articles,' or' Enclitic Definitives 'of Bengali,-« -ta•, the original strong masculine base, implying biggishness or crudeness of the object, and c -~I, -ti •, originally feminine, implying' smallness or daintiness. Thus, we have l!l' ~?:~1, ~(~)~ > « du(i)-~a > duto, -ti,.; f'!il~ > f~"l_cl;, f~~~ c tinta > tinte, -ti •, etc. The affix has gradually become a distinct definitive in l3tlngali, being added to names of persons and things and to adjectives : e.g., ~1~~1 « gach~ta » that tree, that big t1'ee, but ~1~~ « gach~ti ,. tltat tree, that nice l£ttle tree; ~1~~1 « Riim~ta » tlwt ungainly, awkward, or bad felloz,, if a Rama, but ~l~~ c Ram~ti » our good fellow Rama; so ~1"'1~1, ~ti71~ « bhala-ta, bhala-ti,.; etc. T;j, ~ alRo occur dialectally as \51, @t .. - ci; c -tiya > -te,. C+ + -i- + -a< -vrtta + -ika- +- -aka), indicating 'habit' or 'nature,' 1¥ 'resemblance': F.g., ~~?:1; < ~t~~~ « a~~~e < *a(w)isatiya ,. smelling if ?'anlc fish and jleslt (ami~a-vrttika+ aka); C' -r- » (i) below. • ( 4.:~) ~ « -~ ~~ • ( i i) . 437. Iu a few words, which may he called disguised compounds : from Skt. or OIA. « paHa •. Examples : ·cil'l~~. WIT~~ .. lengat~, langat~,. (*la:nga-paHa, li:nga-patta), see p. 347; ;;'l'!t~ « malat~ • cover qf a book, literally dust-board (;;'1'11 c mala ,. + ?ft1i .. pata ,. < .. paHa ,. ) ; a feminine > diminutive-pleonastic extension as in 'SIT~« -at~ •, also 'Sit~ « -ati .,, ~~ « -uti.•, ~~ .. -iti ., iS« -ti •. • 438. Like some of the other suffixei noted above, this is really a decayed word figuring in a number of disguised compounds. The source of it is c *mrtta, mrttika,. earth, Beng. ;;111, ort~ «mat~, mati •. Examples: C~ta'!t~ « dholat·~ • soil zvasherl down by ?'ains; ~i'fl~ « dhiuati ,. earth heaped up for an embankment (dhara+mati < mrttika); tC9IC~tJr,, C?!~~ « petoti, pet uti • earth and bits if Jute nzi:red, for plaste1·in,r; murl huts (pat~+ mati > *patawaty, *patuati); ~~t~, ~'I_~ « tu~a~I, tut;~uti .. cha:ff and earth. mi:recl (tu~a-+ mrttika); ~~iS « khariti • challc and earth (khatika chalk). • ( 44) i!? .. -r- ,. (i) > -~1 c -ra •, ~1 ... rr •, fi!?~i > ~ .. -riya > -re •. 439. Tl1e « -r- • aflh in a number of words indicates habit or profession, or some sort of connexion : e.q.,, ~t'lt?:~ < ~t'ltf~~j c basare > basariya • d1velle1· in a lorlgin.r; house or hired apartments ( < hasa lodging house,< vasa-); c ~t~?;i!? « *pMsariya > pMsure ,. one ~vho !trangtes and robs 688 MORPHOLOGY: CHAPTER I travellers (phas;j. hatteN·ope with loop); ~tPnrt~1 > C' ghesera ,. gmss-cutter (gh:,i,slj., *gbii.si grass); ti!tl5?:~ « naure < *niiwariii.,. a boat-man; MB. ~t\5~, ~t~f~~j « dbii.ur;j., dhau:riya » a nmiu!r; C~?:"l~%. « kheloii.rlj. ,. a playe1·, an expert > a smart and sha1p person (slang) ; ~~%.) « jua:ri ,. gamester, see p. 668 ; *1lt9ff~m < ~t9J:!;~ « *siipariya > sii.pu:re ,. snake-charme1·; *vaQ.ii> *vaQ.a,. By extension with« -ii., -iyii, -i < -ika ,. etc., we get the \1?1, f~~1, ~ etc . • (45) ~ .. -r- .. , ~ .. -ra .. , \!?) .. -ri ,. (ii). 440. This occurs in some place-names, and it comes from a Late OIA. and Early MIA. « vii.~a,. enclosure which is found in Sanskrit as a c guf}a,. formation from« vata < vr-ta .. , < « vvr,. to cover, to enclose. ~' f~ c -:ri, -ri ,. as an affix in Bengali would therefore represent « vatikii » compo•mded with words. Examples: \5!1~1 c iikhlj.ra ,., Hind. « akhara ,. enclosed grou1ul for physical exercise, club (ak~a-vii.ta-); ~51%. « bbagar:t ,. place outside 11illage where dead cattle M'e cast away (? « bhagna + vii~a ,. : • connexion with ~~ « har:t » bones is popularly suggested); C'>ltm~ c goa:rr ,. place-name = Gowari near Krishna§ar in Nacliya (gopa-vatika) ; c~~t~ c kb8ar:t »pig-sty (? Persian « xug ,. swine> Bengali c *.khoga, *kho, kh5 » + « vata ,. ). 'l'he Sanskritised « nii~-vata » a fleet qf boats or ship.~, a clock, met with in old Bengal inscriptions, represents an Old Bengali­ c *nawa-wiiQ.a •, which was Persianised into c na:vwarah » by the Mohammedans : « *nii.wa-waQ.a ,. may be connected with the dialectal word ilt~ above ( §449). 'rhe word « vita » may equally be the source of ~ « -ra » in place names, although a non-Aryan ongm, .Kol or Dravidian, for that affix is not unlikely (see ltpra, p. 66). • THE AFFIX ' IJ.' (III) 689 The word ~nrt~ c batoar¢ • high-way robber (bata < vartman) has nothing to do with c r •, (i) or (ii), although c -r- ,. (i) might have been ·~confused wtth it.: it comes from MB. 'tt~9f~ « bata-para •, in which form it still occurs in NB. ( < causative root« pataya- •). (46) ~ .. -r- ., ~ .. -ra •, ~ .. -ri ,. (iii). ·441. Pleonastic affix, derived from« -Q.a- •, which was widely. used in Late MIA. : e.g., .. vaccha-Q.a (vatsa), diaha-Q.a (divasa), gora-Q.-I (g-auri), neha-Q.-a (sneha), hiya-Q.-a (hrdaya), dosa-Q.-a (do~a), parakka-Q.-a (parakiya), bhumha-Q.-I (bhUmi), teva-Q.-u (tadvat), bh(r)anta-Q.-i (bhranti), nidda-Q.-t (nidrii.), desa-Q.a (desa), sandesa-Q.a (sandesa) •, as in the Apabhrarisa fragments in Hema-candra; « dukkha-Q.-a (dul}kha) •, as in the' Safujama­ maiijari '; etc. The c -Q.a- • affix seems to have come into special pro•mi­ nence in the spokert dialects of Northern jndia in the closing centuries of the lst millennium A. C. It is also very likely that it was largely used in the spoken language in the earlier stages of MIA., but the remains of Firs~, Transitional and Second MIA. do not use it lavishly ; and in Third MIA. literary remains, where the style is exclusively artificial and is modelled on the Prakrit kavyas (as in the Jaina works. like the., Bhavisatta­ Kaha '), it is not so common either. The Ap. verses in Hema-candra's grammar form specimens of genuine popular poetry, and hence the ··-Q.- • words are so much in evidence there. All NIA. speeches show numerous forms with « -Q.a- > -r· •; and some kinas of NIA., e.g., the Rajasthan! dialects, employ it more lavishly than the rest, affixing it to _passive partici pies quite freely. The Second and Third MIA. « -Q.a • rests on a First MIA. « -ta- •, which would normally represent an OIA. « -ta-,. (or a dental c -ta- .,.with or without a c r • or« f • to cerebralise it). (Cf. Pischel, §599.) In Skt., we have a number oft words with a c -ta • affix, words which are mostly late, but a few like c marka-ta • undoubtedly go back to pre-Buddhic times: e.g., c marka-ta (regarded as being Dravidian in origin : cf. Tamil « maram • tree) ; ka:gka-~a, karka-ta (see p. 368); karpa-ta mg ( < *:V[s]karp); parka-ti; karva-~:t; lrukku-ta; lak-u-ta > lae--u-da.: • ' ' . ' f ., y ' <(' $7 690 MORPHOLOGY ; CHAPTER I muk-u-ta (?); cip-i-tajlattened rice; vadhu-ti; kula-ta (explained as being from « kula + ylat ,. ) ; barba-ta ,. ; etc. This « -ta ,. is not found in Vedic. There is nothing in the non-Aryan languages, D~vidian and Kol, to warrant any influence from that side. And such a characteristic lA. suffix, so persistent in all the forms of NIA., can reasonably be expected to have an OIA. source. It seems quite possible that this « -ra < -Q.a < -ta ,. is only a form of a « -ta •, undoubtedly the same as' the passive participle affix, which was used as a secondary affix in . some noun and adjective formations in Vedic (cf. Whitney, Sanskrit Grammar, § 1176, § 1245; Macdonell, Vedic Grammar, § 209): e.g.,« eka-ta, dvi-ta, tri-ta; .muhur-ta; ava-ta weU ;·raja-ta; parva-ta •. Some sense of connex­ ion ~ith or modification of the original idea is implied by this « -ta •, which also is not absent in the NIA. « -ra ». By spontaneous cerebralisation (see supra, pp. 487 -488), « -ta- :can very well give « -ta,. in dialectal OIA. and in Early MIA. Thus we have OIA. « vibhi-ta-ka (beside « vibhidaka,.: < « ylbhid,. ?) > *vibhi-ta-ka •> MIA.« baheQ.aa- •> NIA. « bahera,.; OIA. « amr-a-ta-ka > *amra-ta-ka ,. > MIA. « am biiQ.aa- ,. > NIA. l!lt~lfl « am,ra,.; ~nd OIA .. « sr~g-a-ta-ka-, sr~g-a-ta-ka- • > NIA. « sigara •· Throughout th~ whole history of lA., « -ta> -ta > -Q.a> -ra,. would se!'lm to haie been fairly common in . the spoken language; and after phonetic decay had put an end to most of the characteristic 01 A. affixes in the Second NIA. stage, the « -Q.a ,. affix, with its strong cerebral sound, seems to have become quite a prominent\nd a popular one. The large number of names in « -Q.a •, Sanskritised to « -ta • during the Second and Third MIA. periods (see ante, p. 89), would also indicate its popularity in the Indo-Aryan speech from the early centuries of the first millennium A.C. Hoernle's derivation of this affix from a Skt. « -drsa ,. (Gaudian Grammar, § 218) is phonologically inadmissible; there is also the fact.that the earlier form of « -ra < -Q.a • was « -ta ,., • Examples of-« ·ra· ,. from Be~gali: OB. (Carya 38) « ~ava-r-i • (nava, nau); (Caryas 10, 20) « bap-u-r-a • a Kapalilca (cf. Sauraseni Ap. vappuQ.a); (Caryii 10) « pakh-u-r-i • petal (pak~a = parva); MB. (SKK.)·,~ "! ~udart • beside t~ : mudi • rin!J (mudra); MB. ~ THE AFFIXES '-1}-, (Ill), '-T-' (I) c diyari,. tamp ( < dipa); ~t assu1·ing (as 'a horse) (cumba-); ~'t~f'¥ « phaskuri,. a sm_all_boil beside c:~tJ'f._<11'1 « phoska,. a larg~ boil or sore ; ~tl{~ « ma~~ri ,. crust. on- the suiface, as on a healing sore (marman sensitive or exposed part); ~~' ~ ~~ « cha "' ?) ; 5t~~ • « cag(g)ari .. little 'basket, beside j)f~~ « cii:g(g)~ra • above; c:~\;$Jj • p~~~ra ,. box. (*pel)-~a = petta = pe~a~) ;· ~t'SIJ\!5 > NB. ayy3t~ > eottj,,. (avidhavatva), also with ~ .. -ti,. with .. -i < -ika,. ; ata. t• raizot • ( = rii.jya +tva) in the 'Crepar Xaxtrer Orthbhed.' 'L'his affix was later reinforced by the Perso-Arabie affix noted at p. 656. ( 48). \!5 .. -ttf. •, '!51 .. -ta •, ~"t .. -ti .. (ii). 443. From « patra- • as well as « piitra- •, a decayed word figuring m some disguised compounds in which it has acquired the force of an affix: e.g., ~\!51 « nii.mfl.tii. • mnltiplication tabte (nama-patra-); ~1~\!5'1 c rli~(g)fl.ta • tinsel, copper;(oil (ra:gga-patra·) ; 15tf1§ c cak~ti • some jlat or ma~llerl obJect out into a ring (cakra-patra + -ika); g~tf