ON THE USE OF PRAKRIT DIALECTS
IN SANSKRIT DRAMAS
BY
P. L. Vaidya
It is usual in Sanskrit dramas to introduce some characters
that speak in some Prakrit dialect. All high characters, of course,
speak in Sanskrit, but even they are at times allowed the use of
Prakrits ; but lower characters, almost invariably, speak in some
Prakrit dialect. The use of a specific dialect for a character
seems to be determined on the strength of a nice rule of Višva-
nãtha, who, in his SUhifyadarpana says : -
HÍ&Íê HTW5T g HW STTÍtTH^
Sähityadarpana VI. 168.
But in practice it is hardly observed. At any rate, it is not
always possible to say that because a character speaks Saura-
senl, therefore he or she must be the native of Sürasena country.
The dialect of a character seems to be fixed by some convention,
and although in the Mrcchakatika the above rule is observed
fairly rigidly, we cannot say that it is universal. The conven-
tion regarding the use of Prakrit dialects is found stated in
works on Dramaturgy. Let us consider them chronologically.
In the Nãtyasãstra of Bharata we find the following passage
bearing on the use of Prakrit dialects in dramas
*rraT H H
3T«rer U'?řT: i
ff TT3% II
TTTir^gri%^T STT^qi I
siïfjîîPT fffr ÎT irmeli: il [ 5 ]
fin =R^TT<JTÍ ^ fèWTT TTSTW» fiprr: Il
*mnit 3 ^ï[T<ïïm'a:3TÎïreTT%ïTTJï i
O 1
îri^TT Í%!<^PTC
1 6 Annals of the Bhandarkar Oriental Research Institute
^riçririnn ^ i
^ ^r^nrr n
statut 5I3>TCHT H^ar»rra?? ^fr *I<ïï: I
3T^r**r<n ^T^sfr u [ 15 ]
mîïït 3T3TC*Tm H Í^T%jJTtnp*fÍ H »-Tí ¡I
IT SIT ^T^TÍ^TlTTfè^re^RÍH^TTH^TH; I
3rr*řmi%: 5iT^T wr ?rfir€t u
[ The above text is fixed with all the available material from
NSP edition, Benares edition and GOS edition with Abhinava-
gupta's com. The variant« are : [LI] G08 5TÍ*%íí ^OT?ťarfl *tTTÎ
T.[ůt% í(3Čra. [ L 4 ] NSP bttM for Jjíiwfr. { L 6 ] NSP 5t^mV
for 5T5Cm«tť ; NSP °*rert° for °^ť ; GOS °5TH5íT?^5n: for •5Í%Š7-
. I L 7 1 GOS **rt for ^čfi: . [ L 8 J Ben. nimft a won 3*1^:-
S^tTitoî»* ; G0Sofi?a^HW". [ L 10 ] GOS and Sähityadarpana
*üT?*r%5T; but GOS notes a v. 1. ÎPHT (sic) HT-rr 3î^a^? which
supports our reading. [Lil] GOS for *tV. f L 12 ]
GOS ÇTÍífflmWT : Sähityadarpana ?rt%oiTřlT if. [ L 14 ] GOS ÇT$îT-
for ÇTîWïÎ I L 15 ) NSP ; Ben. TT^äTSir for
^rt^rér. [L 16] GOS [ L 17 ] GOS
3mmw<TT for STSTCHlfT. [ L 19 ] Offl. in NSP and Ben. edit
On the other hand, Viàvanãtha, the author of Sähitya
( about 1380 A. D. ) says :-
tT^noTTíRT^RÍ #re?pí l> lM*¿ Il
îTÎl^TT tTTSSÍfaí T I
H ïTT^nw inřrsrcrac Il II
3T=TT^T ïTTltvfr *T<TT Tr^rFH^^IROTT^ I
«rig^i arfara vt il n
îTT^trr çafaí i
mvrcrnftírítaí ^rí%orr^T f|
^I^RITT ^I^FHÍTIÍ ÍIT^rfí ^frnffatiq; I
g^^igpíTT^T^Í^TTHÍ «il i <=i-áT 3tTT^^U^|
3TT*IR5 řWÍTÍT ^T"3TčřT ^JIWTT^ '
STT^TCT ^IT^fr «FTOTOrnfiraf U
Use oj Prakrit Dialects in Sanskrit Dramas 17
WT^uR^hsft U 11
gi<?RÍ ^ ííHírfís^TROíra; i
3?ïraTïn*ji3TToif %g- ®ť%3; I» 11
q-^ipJT HTTŤRtT ^TRírWfřfl^ ^ I
fírg^^VRT^RÍ íTTfří HîTm^(3[ II II
ÍŘ^FH tfïnmjïsïr r%rf4rfTTinf ^r i
^ímíHrHarè^re^fa #i%Ťp.frr%ii' 11 <1
îTr^Trâ H H^q- Trť^řf^ I
3TRH33ÍŤTITTTRÍ üFHíT *IT<Iir%T^: II II
H«fr 1
%^Ttan«ř íTTTHčq- ^1'řRRčRr II II
wrfjarçfa, VI. 158-169 ( Kaoe'g Ed. )
The Nãtakalaksanaratnakosa of Sãgaranandin ( about 13th
century A. D. ) says : -
ssmT: írêfjrí RÍÇ', Í^HTTT gurí^r: 1
<T§£ : SÍRÍHŤ f%**ÏTP* II
ŤTOTsft sunra: g?rfîR: 1
m^RTT^rfí II
qar ^ í^npr: 1 C.-N
srere^nr: Ti^rr: ^rcsa- ÏTÎT^TV. n
*nw HrTOíHr: i
vj&n ttt^T H
Cn O
*nr: TOTO fà*Trfr fèfa: II
^re^^TOR^^m, lilies 2148-2157
No other writer on dramaturgy bas expressed views on the
use of Prakrit dialects in the dramas. We therefore have to rely
upon these writers and see how far the extant dramas put these
views in practice. However, on carefully reading the passage from
the Nãtyaéãstra and comparing it with the same in the Sãhitya -
darparta , it becomes clear that the author of Sähityadarpana had
before him the text of the Nãtyasãstra as he bodily copies
several lines from that work. At the same time, he introduced
3 [ Annals B. O. R. I. ]
i8 Annals of the Bhandarkar Oriental Research Instituto
certain changes in the detailed application of Bharata's rules
without authority. Among such unauthorised innovations of
Viávanãtha, we point out the following line .
sTTtfmsr JTT2TTH spring; i
It will be seen that this line cannot be traced to Nãtyasãstra. In
fact, Nãtyasãstra does not mention Mahãrãstrl at all, probably
because its use was restricted to Kävya or poetio compositions
only. Prthvldhara, the commentator of the MrcchakatiJca has
remarked : -
q-sr I
So Mahãrãstrl has no place in drama
tioned by Bharata, nor by Sãgaran
than Viávanãtha.
The rules laid down by Bharata for the use of Prakrit dialects
in the dramas seem to be very simple. According to him, and
Sãgaranandin supports him, that the main Prakrit dialect to be
used in the dramas must be Saurasenl. He however allows some
latitude to writers to use other dialects as the dramatic form of
poetic composition draws its characters from different countries.
Bharata then enumerates seven principal Prakrit dialects for
use in dramas. They are : Mâgadhï, Avanti, Prãcyã, Saurasenl,
Ardhamãgadhl, Bãhllkã and Dãksinãtyã. He then gives the list
of sub-dialects, such as Sãkãrl, Ãbhlrl, Cãndãll, Säbail, Drävidl,
AudrI and VãnaukasI or Vãnecarl. No specific list of SauresenI
speakers is now given as this is a dialect that could be used by
any in the drama. Mãgadhl should be used by snake-charmers
( Narendra, a hard word and therefore misunderstood and para-
phrased ), servants and young princes in the harem. Ardha-
mãgadhl is to be used by merchants. Prãcyã is to be used by
jester or Vidüsaka. Avanti is to be used by heroines and her
friends but this dialect should not much differ from Saurasenl.
Dãksinãtyã should be used by warriors, police-officers and
gamblers; and Bâhlïkï by northerners. Let us now see how
Viávanãtha distributes the use of these over different characters.
We have already stated above that Mahãrãstrl has no place
in the dramas, and therefore
Use of Prakrit Dialects in Sanskrit Dramas 19
3TTtfTÌTSr H îTT^TF HORTET I
a line which is unauthentic and unwarranted. The term
•Ř^íotí in Bharata is unfortunately not understood by scholar
including Visvanãtha, and therefore, Bharata's original lines
TTíí TVTT 3 1
THTgsromH
are rendered by Visvanãtha as
OT5fTŤF>T TTÍíTvfr ^rr^TT I
TF^r^TöTT^
The change however matters little as it ®mits ^5" or snake-
charmers from the users of Mãgadhl dialect. The text of Bharata
and Visvanãtha is identical as regards the use of Ardhamãgadhl
and Präcyä, but when we come to the use of Avanti, we see that
Visvanãtha transforms Bharata's text
WWT *TPTT 3T^T%STT I
íTTÍ^RÍ CTSÍftf ^ II
into
^rTTïït I
This view of Visvanãtha is clearly opposed
Bharata. If by the term dhurta we are to under
gambler, then it is clearly redundant as his
identical with Bharata's text, reads dïvyat
gamblers. It appears therefore that Visvanãtha has created a
confusion by his view
í 1
as also by his view
STTtfTÌre 1 TTSTïH
I stated these view
of Sir William Jon
hypothesis, but now
from Rõpa Gosvãm
part of his Nãtakaca
HTrffa 1
H çsïht wrawr tmn
20 Annàls of the Bhandarkar Oriental Research institute
Let us now revert to the topic of dialects. On Dãksinãtyã, the
text of Bharata and Viávanãtha is almost identical, viz.,
5TT%arrrqT i
with only one variant, viz , Visvanãtha reads it for
Bharata's text. It then becomes possible to construe
with the line
gcTPTT I
giving us a rule that Avanti should be use
and police-officers. I think this wrong
of Prthvîdhara's remark
which view is clearly opposed to that o
Nãtyaáãstra and goes against the stat
characters, viz. Candanaka, who in the six
sa? ^ï^fPiïTrr 3T^*TT%&rì.
We therefore have to reject Prthvldhar
may be based upon a wrong constructio
Visvanâtha's text. Lastly, Viávanãtha r
the use of Bãhllkl
Coming to the use of sub-dialects or Vibhäsäs, viz., Sãkãrí etc.,
Bharata's text begins with
^TRT ^ *sr?5T3TT I
But these do not figure in the general plan, and hence we have to
take these Khasa people to be a wild tribe on par with Sakas etc.
Visvanãtha however reads
?TÍšrer I
for tßis part of the text. In fact, Bharata's text
corrupt from very early times which fact is respo
confusion. The order in which the sub-dialects a
Sãkãrí etc. as given in line 6 of Bharata's tex
this, Sãkãrí is used by foresters» Sakas and
Use of Prakrit Dialects in Sanskrit Dramas il
Cândãll is used by Pulkasas or outcastes ; Säbarl by wild tribes
as also VãnaukasI and Äbhlrl or Säbarl by such wild tribes as
deal in elephants, horses, sheep and goats, camels and those that
reside in wild herds ; Drãvidl also is used by foresters.
It has to be admitted that the text of Bharata is not yet
available in a critical form ; there may be lacunae in the exist-
ing text, but as far as our knowledge goes, we have to say that
Visvanätha used it rather carelessly and made unwarranted
changes in it as Rüpa Gosvämin noticed it as early as 1550 A. D.
In conclusion, we may summarize the views of Bharata on
the use of Prakrit dialects in dramas in the following way : The
principal Prakrit dialect to be used in dramas must be Saura-
senl both for prose and verse. Vidüsaka and others should speak
in Fräcyä which is hardly different from Saurasenl. High class
ladies and their servants should use Avanti. Low-class servants,
young princes or boys, snake-charmers and merchants should use
Mãgadhl or Ardhamãgadhl the basis of which is Saurasenl.
Warriors, police-officers and gamblers should use Dãksinãtyã
•
which is Saurasenl with a light touch of MahãrãstrI, Sãkãrl
Cändäll and dialects like Dhakkl should be used by low-class
people or tribesmen.
In Laksmldhara's SadbhUsacandrikU , ( 16tb century ) we find
the following : -
ra h ir Tfrá STTOTT i
M* n
Í5iõ<mír<iqT 3TN tfíjpní; i
^ miroir htí vrrg>a^rò n 33 11
3TW ir wir grn snrífcfr II 3V (I
vfRTi?rraTr%^ flunfr i
!l It
3T»^ řinr^ri^ irersm ii 35 n
ITÏÇT?«nW TÍWST ïT^nH^rà řT«TT II 3^ II
22 Annals of the Bhandarkar Oriental Research Institute
^ ^RGTTT% || II
According to Laksmldhara the above passage is a qu
from Rûpakaparibhâsâ, which in fact, is a part of Ras
sudhäkara of Singabhupâla ( 1330 A. D. ) ( III. 297-3
views expressed here and the classification of Prakrit
there belongs to the Western School of grammarians,
not agree with the views of Bharata.
Keeping in view the teaching of Bharata as detailed
let us proceed to see how far it has been put into prac
Mrcchakatika. The commentary of Prthvldhara, whic
sequently copied by Lalladlksita, says : -
5^i^^T%3>T!rr^TrrTJT^: I 3T<T55m^s »PTT:
=TTÍ%řT I ïTTëPïï-
irrTTïîT^fïH^T iri^rr ^TR^wir-ír i
«TTffr^r ^ ítT íTípnaar: ti
#TT ^ ¥tthht: II
t%t1Nt wptt far^rr^Ts i fì^raròr^s
mrânf: «i
ïrrffiTT3%2 gwvnfi ^ÎT irçfror
%<řr ^T^ŤT^r^oFr £HT ÇH <rq
irrorTraw i g^vrffrs'-re srrfrfr, ^rârsrsnís
Tf^řr i fôfro i 5T^TTgríT'fT-
#*prr^rraí qt Tmnnfms^r. i
arnjsRTwnal^ ^^ÍTvmrmasřT I ^uçrâninwTnrô
^fj^rer wr^^řT^fr i
H»4T iirmr THTH ^snrarspreaT i rrarai^rsTT rraaì
i STHUT wTm^^^iTïn^T i »tttvít nrey«iWTf*řřr i
íwq ^ ^spre<ïr i
V HTT 1 ) 9F**raT^rcT5nfn*?*r^i3ir ^ i
Use of Prakrit Dialects in Sanskrit Dramas 23
t f%|: il
3TTr»í i õ^«í i ^[T i%r%rTH'=ni T
^rra ^T7T«ranj 1 5^»i i farm ^Tfrfrnra;
^T^TTffT^rf^řčn^u^rTr TTÍ^T: I
^T?sfr ígí^T'^Í T%f5"îp: Il
1 ^g«frfirïTr%s? ÏTR?^W 1 r%?à
^HH; i Tsfr i TTwqçrrHHTírWTÍq": ^n
íF^ífr f^T5r^nqT%^~-ä?^ffrr^r«r ^rer%^r%m
According to this view of tha commentator, th
uses four out of seven principal Prakrit dialect
Avanti, Prãoyã and Mãgadhl ; and only three
Sãkãrl, Cãndãll and Dhakkl. He then classifie
according to the dialects they use, and also states
of these dialects. Among speakers of Sauras
names of Madanikã, Radanikã, Vasantaeenã,
maid, and Cärudatta's wife, which is rather str
of Bharata's view : -
*TF?*TT *TPTT I
^rnr^r^ri trefní ^ ^TrT^-^rfirTr r% sfr 11
On examination of the speeches of these characters as
ed in reliable Mes., I find that the dialect they use is Av
a few forms of Mahãrãstrl mixed with those of Saurasen
exactly how Mãrkandeya describes Avanti :-
arigli i*rçr*îhrcNrt*g 1
Vasantasenã uses this Avanti for prose as well as for verse. The
most palpably wrong statement of the commentator seems to be
sT^aHprrns^r ^rc^F5«r#r 1
This may be due to the influence of Visvanãtha, and we have
shown above that it is wrong and opposed to the view of Bharata.
The examination of the speeches of these characters shows that
they use Dãksinãtyã as directed by Bharata. This Dãksinãtyã is
very similar to SaurasenI and has a mixture of Sanskrit. It may
also show slight influence of Mahãrãstrl. This view agrees with
the statement of Candanaka
^3Í ?(%?r<ïïiT! 3T«fŤTíTTÍ%<ifr.
24 Annals of the Bhandarkar Oriental Research Institute
Prãcyã is almost identical with SaurasenI. According to
Prthvldhara, it seems to retain the intervocalic ka of the termina-
tion instead of eliding it. The Mss. of Mrcchakatika however
do not preserve this ka in the speech of Vidüsaka. There is a
character of a Sresthin or Merchant in this play, and according
to Bharata,he should speak in Ardhaitãgadhl. The commentator
puts him among those that speak SaurasenI, and Mss. agree with
that statement. So we should hold that SaurasenI, as Bharata
has laid down, is the main and dominating Prakrit dialect used
in the Mrcchakatika, and we fix up that Sütradhära, Karna-
puraka, Šodhanaka and Sresthi speak in pure SaurasenI ; Vidü-
saka uses Prãcyã which is identical with SaurasenI except in one
particular, i. e., Prãcyã retains the ka termination ; Sütradhära's
wife, Radanikã, Madanikã, Vasantasenã, her mother and maid,
and Cärudatta's wife speak Avanti with the domination of Saura-
senI with a few forms of MahãrãstrI ; and Vlraka and Canda»
naka speak Dãksinãtyã.
Among characters that use MägadhI in the Mrcchakatika, we
find the names of Samvähaka, Bhiksu, Carudatta's son, the three
servants of Sakãra, Vasantasenã and Cãrudatta, and herein the
commentator's statement is correct. Sãkãrl used by Sakãra and
Cãndãll used by the two executioners are only sub-dialects of
MägadhI, and from the linguistic point of view, they can be
hardly distinguished from MägadhI. In the Mss. of Mrcchakatika
they show a very close resemblance to MägadhI. Lastly we come
to Dhakkl. This sub-dialect is a mixed dialect sharing the
characteristics of many. The commentator says : -
The two gamblers Mãthura and Dyütakara use this dialect ; and
on examining their speech as represented by the Mss. of the play,
I see that TCTCTrar of the commentator must be* ^rorrar. If we
accept this modification, Dhakkl of the Mrcchakatika agrees
with its description given by grammarians like Mãrkandeya.
There are thus three features of Dhakkl, viz., the use of S ending,
cf Sanskrit words and expressions and retention of both and çr.
Our Mss. support these characteristics.
Use of Prakrit Dialects in Sanskrit Dramas 25
It is true that Mãrkandeya agrees with Prtlivldhara that
Vlraka and Candanaka should use Ãvantl. Mãrkandeya quotes
as from Bharata the following : -
This passage is not found in the existing editions of Nãtyaáãstra
and is in fact opposed to the view of Bharata. Similar is lhe
oase with reference to Bãhllkl as quoted by Mãrkaçdeya
because in the existing text of Nãtyaáãstra we find
Are there then two Bharatas writing on Dramaturgy ?
In conclusion, we may say that the Mrcohakatika uses only
two main Prakrit dialects, viz., SaurasenI and Mägadhl ; in the
speeches of women, there is a tinge of MahãrãstrI which tinge
gives it the name of Ãvantl. The speeches of Vlraka and Canda*
naka are theoretically in Dãksinãtyã, but practically they are
in SaurasenI ; Säkärl and Cãndâll are only sub-dialects of
Mãgadhl ; and Dhakkl is a mixed sub-dialeot, showing features
of Sanskrit, SaurasenI, Mãgadhl and Àpabhratâáa.
4 [ Anuais, B. O. K. I, )