cc-'---'---"'=-==~-----

A TJ'Uljf!'ib1, ('liLLk llowcr') b the ;;cribc:o' way or marking the (!IHI or the main text
am! tlw b1~ginning ol' th<! col<>phon. The pn~scnt logo i,; an artistic irnprcssion by
Shubhani Sarkar bac;cd on ,;11ch a ;;cribal l!ouri,;h ,;ccn on a Nepalese 111a1111script.

C'

'\ r ..,. )t \l ,. , { '/.J

C\

,_1

I '. J

;I\ . ; .

1 '. l

()

' . ' ' .

'

'! .

PusPIKA
.
Tracing Ancient India, through Texts and 'I'raditions
Contributimrn to Current H. cscarch in In<lology

Volurnc 1
f>ItOCEEDlNCS OF TIIE
FIHST INTEllNJ\TJONAL INDOLOGY G!li\DUA'l'E i{ESEJ\HCll SYMl'OSIU!Vl
(SEPTEMBEH

2009,

OXFO!lD)

Edited by

Nina Mirnig
Pc~tcr-DAnicl SYi6.nt6
Michael Willianrn

n(,,j
Oxbow Books
Oxford & Philadelphia

l'ublif:hed in the fJnitcd Kingdr!ln in '.Wl:l l>y

OXBOW llOOl<:S
10 !Iythe Bridge Stn?ct, Oxford OX I 2EW
and in tire 1J nited State;: hy
OXBOW BOOI<:S
!J()f) Darby !toad, llavert.own, l'A l!HJfl:l
ri~

Oxbow Book;;

am]

Lhr; individwd authon:, 20 I :l

Pap<Jrhack EdiUon: !SUN !J7fl- l-fl 11217-:lx!'i-!'i

Digital Edition: ISBN !l7i-S-l-7x2!l7-!H2-2

A CIP record for this book is available from the British Library

All rights reserved. No part of this hook may br? reproduced or transmitted in any form
or by any means, electronic or mechanical including photocopying, recording or by an~ .
.rn f01m;iv1on
'.
'

1 syskrn, wit
]1out fH?rm1;:c;1on

1 rorn t,IH? P' il>l"1,,IH'r
scrn-age
an< l re t,ncva
,, ' in wnt.rng.

For a complete fo:t of Oxbow titles, plca;:c c<rntact:

UNITED KINGDOM

Oxbow Boob:
Tclephoue (Olflfi!'i) 21112 11!), Fax (Olflfi!'i) 7!J'1'J1l!J
l~rnai l: oxbow 1(l'ox bow books ,com
www.oxbowbouks.c<Jrn
UNITED STATES OF ArvrnIUCA
Oxbow Books
Tel()phonc (flOO) 7!lt-D:l!'i l, Fax (fi!O) f\!'i:l-!Jldfi
Ernai I: q ucric;/' 11ca ;:em atcaca<l<?m ic.co m
www.ca;:ernatcar:adernic.com/ ox l><>W
1

Untvareit~t

Hamburg Asien-Afrilca-lnstitut
FB"Blbllothck I 1 + 2
Edmund-Siemers-Allee 1 FIOgel Ost
D 20146 Hamburg

-,-

Contents
-------------

-----~---

---------

Preface
1

;3

ix

GIOVANNI CIOTTI

Defining the Svarn Dearing Unit in the ik.';JlLvcdr!,riga literature:

tJ nrnasking a veiled debate

Cox
Puriir.ric transformations in Cola Cidarnharnm: The Cidu:rnbaramahatmya and the Sfitasarrdl'ill/,

25

WrnTNEY

DANIELE CUNEO

Unfu:;,:;,ying the fu:;,:;,y. The distinction between rasas and blul'uas

in Dharata and Ahhinavagupta

49

I !UGO DAVID
A contribution of Vedanta to the history of l\iffrnaqrsa: Praldi~iitrnan \.; interpretation of "verbal dfoctnation" ( .fo.bdabhlL'UanrL)

77

I ms I H.AN FABKllONDEII
Married women and courtesans: Marriage and women's room for
rnannmvrn as depicted in the Kaf;/1.ri;-sarit-;r/,_(jara

105

EMMANUEL FB.ANCIS

Towards a new cditiou of the corpus of Pallava inscriptions

7

12;3

ELISA l<'B.ESC!II

Did M!ma.1psii authors formulate a theory of action?

151

Vl

l~LISJ\ GANSEH.

Trajectories of daJI(:<~ on the surface of theatrical meanings: a .contribution to tlw theory of rasa from the l'onrth chapter ol the
Abll'inu:oo,hh(in1,U:

ALASTJ\IH. GOH.NALL

Dravya as a Permanent H.dcrcnt: The Potential Sarviistiv;ida [nfiucnce on Patafijali's J>a,.'ipa,.s(l.h:n:ika

10 G1m.cELY I lmAs

Rituals in the Mahiisiiha.'JnlJffO:rrw,nla:nas1/,lra

11 NmA.JAN KAFLE

The L-i'ri,godbhu:ua Myth in Early Saiva Sources

12

22G
241

KENIClll KUH.i\NISIII

Yantras in the I3ucklhist Tantras

Literature
l:J

Yarniiritantras and Hdatcd

2GG

NINA MmNIC

Saiva Siddhiinta Sriiddba

Towards an evaluation of the socio-religious landscape envisaged
by pre-12th century sources
14 AYAKO NAKAMURA
Constituents of I3nddhahood as Presented in the 1J'U.d(J/wbli:u:m'is1/,lra and the 9th Chapter of the 1vfohiiyii:rw,.'f1/,/:nilarr1,klira: A Comparative Analysis

lG

i\NDH.EW 0LLETT

The go:u,accharula.'J in tlw Indian md.rical tradition

lG

:w:J
:n1

ANTOINI: Pi\NA'iOTI

Anll.trna,lcl., Sotcriolol-';y and Moral PsydiolOl-';Y in hidian Buddhism :JGG

i 7 I SA BELLE H.J\'l'I!.:
Pii:ramll.'rlhika or lLJHira:rn(/,rth'ika? On the 011tolo1!;i<:al status or scpa.ration accor<li11g to Ahhi11avag11pta

18

BIIIJ\NI SAH.KAH.

Thy Fierce Lotus-Fed: Danl-';cr and Bencvolc11cc iu Mediaeval Sanskrit Poems to Mahi~iismarnanlini-Dnr)i

:18
l
>

CONTENTS

vii

19 P1~;TEH-DANIEL S'!.AN'l'()
Minor Vajrayfinn t<~xts U.
A new ma11us<:ript of the Om"npaiiciisildi

20 M !CllAEL

WILLIAMS

Can we infer unestablished entities? A Madhva contribution to

the Indian theory of i uforeucc

451

Preface

We arc delighted to present the first volume iu tile new Iudological seric:-i '' Prn5pika: Tracing Ancient India Through Text and Trad.it ions". This
book contains the proceedings of the first International Indology Graduate Research Symposium (IIGRS), held at St. Hilda's College, University
of Oxford, 2009. The purpose of this conference 8eries is to draw together
early career researchers iu the field of lnclology on a regular basis in order
for them to present and di8CUS8 topics from their current research pro,iects.
Given the relatively small circle of Classical Indological study in most universities, such au international platform for fruitful exchange and cooperation already sc~emcd a dcs'idcrutmn for some time.
The IIGHS itself grew out of the Indology Graduate Seminar (IGS)
which was held at the University of Oxford. with the financial help of
the Oriental Institute, Nina Miruig established the IGS at the university
in 2008. The Seminar, originally held on a bi-weekly basis at Magdalen
college, Oxford, provided a platform for graduate students of Sanskrit and
related 8Ubjects to present and discrnss their reHearch among an audience of
their peers. The success of the IGS, in turn, led Nina Mimig and Michael
Williams to draw up plai1s for organising a similar forum on an international
level. Subsequently, they were joined in this endeavor by Virginia Greenfield,
Bihani Sarkar, ancl P6tcr-Du,nicl S:;,<into. In order to make the findings of
these early-stage researchers available to a wider arnlicncc, the decision was
made to iuitiatc a rrnblication series based on the conferences. Oxbow Books
very kindly agreed to take on the project, and the eclitor8 would here like to
express their gratitude towards Claire Litt, Publishing Director at Oxbow
for her assistance and cooperation.
The first HG HB, held over two clays in the congenial smroundiugs of
St. Hilda's Colkge, was attended by researchers from all around the world,
including India, France, rtaly, Germany, Hungary, and the UK. The symposimn was opened by Dr. James I3cuson (Oxford University), who as Lectnrer
in Sanskrit also has opened the gateways to Sanskrit learning for the organ)

ix

11
('
ll l"
'f'l /,L, [>o'i'iilJ'ilih;
for- "{/,
iscrs. The inaugural lcd11rc r1hr: Hu,r,cn
AJ ec:wns:
: ', "'.'

tl, w UJ(, was< Ie 1rvcru

, l l)y. . l>r<)I ( ' lmsLopllu
Cr.;nm1,s of / nr11c rn,rJ;n:11,.c;cr"tfJl.'i
'l'fl
l
.

f'
'1

1
)
l
'l 1so h cl< a
Minkowski (Balliol Colle:;<!, Boden Pro[essor o ,Ji.ms u1 , w Jo ' . ;,
'-'
. Irn 1mu 'l'ex t.s..di l(l il'vl"Il\ISCTlI)LS
Oll
. ,Jysterns
wor lrn lrop on ''D atmg
user l m
u.
,

t"
1>

f
fvli11lowsk1
has
the same day. Br!yond t l1es<! rmrnc( rate con1.n1rn .rons, ro
'

.
.

.-.
l
f
.

The
also given gew!rous support to the project ma my11c1.c o. Wcl.ys. second
. '-,JctlH lc'l"''()ll
(A 11 Souls
day was opened 11)y t lw I<:cynote 1cctun!. o r I). ro f' . Al ex1s
,
College, Spalding Professor of Eastc:m Hcligions and Et.hies), Ill wluch he
gave inspiring insights into the study of I<~arly Saivism. \;Ve would like to
express our utmost gratitwle to these three scholars for their snpporL aucl
participation at the varions stages of the project.
.
.
The contributors to this volume wen! all in the c;ulim stag(:s ot their
academic careers at th<~ time of the confcrcnc<!, including M. Phil. and
M.A. graduate students, PhD candidates and those already holcling postdoctoral jobs. Our criteria was simply that participants should have at least
a master's degree, ancl that tlwy should not have held a clod.orate for more
than five years. The subject matter of the contributions ranges across a
tremendous variety of Indological fields, including epigraphy, poetry, drama,
narrative literature, grammar, philosophy, epics and ritual. vVhat miitcs all.
of them, however, is what we take to he one of the key components of.
Indological research: they arc all based on detailed, philological readings of
texts written in Indic languages.
In the first contribution, G iovauni C 1OTT! discusses a long-standing
debate in the Sanskrit grarnnmtical tradition about which part of a word
bears the sua:m ("pitch modulation"), which, as he points out, echoes a
fundamental debate in modern vVesteru linguistics, namely t.he question of
tlw "Accc1 tt Bearing Uui t ".
Through a cornpa.rat.iv<! analysis of thrc<~ paralld versions of the same
legendary narrative in two South Indian pnri1r_1ic 1.<~xts, t.he Oida:mbara.rnrl.luitrn:i;a and the Sfitasa:rr1,hi/,rl., \tVhitiwy Cox, in tlw s<!cond contribution,
ckmonstra.tcs the transforrnativc relationship lwLWc!Cll t,h(!SC two Sanskrit
texts and their connect.ion to the political, social, and religious d1a11ges
taking place in the city of Cidarnharam as it rose to prominence iu the
twelfth century, CE.
Daniele CUNEO, writing on the acsthd.ic conc(!pt. of rasa, analyses the
works of Bharat.a and Abhinavagupta, arnl shows that much of t.h<~ controversy regarding the ml.Lure of rr1..m can b(! n~solvcd if we regard it. not as a
defined concept hut as a semantic fidd with nmlt,iplc mcauings in different.
contexts.
Beginning from a dose analysis of Prak~i.si:i.Lman's ,9a.bdu:n:i'f"(t,a:ya ("An
Inquiry into Verbal Knowledge"), Hugo DAVID takes up the evolution or
l

XI

tlie philosophical concept of ,<fabdabhll,vana ("verbal efficacy"), especially as

it developed iu Mlrna.rps~i schools i11 the 7th-10th centuries.
Iris lra11 FAIU<:llONlrn11 studies the cleve11th-ceutury KalhcTsar"its(/,gara
for what it might reveal regarding the status of' women in North lnclia
at the time - for in contrast to their almost total subordination to men in
normative siistric texts, this work of Sauskrit story literature foatures scvc)ral
instances wlwrn women appear to have a certain "room for mancx)uvre"
through which they arc able to hrc'1k stereotypes, go on aclventnrcs, and
achieve, at least momentarily, a killd of indcpernlcnce.
Ennnanucl FHANCIS's essay first catalogues the types of inscriptions
produced during the Pallava. period in South lnclia, and then, through a discussion of the insufficiencies and inaccuracies present within Malmlingam's
existing corpus of Pallava inscriptions - including novel re-readings of a
fow published verses - argues the need for an upclatc)d and revised corpus,
organi:;,ed along criteria of material, status, content, and language.
Elisa F1rnscrn, drawing on her study of the works of leading
Mimaqisakas, including Sahara, Kumarila and Sornesvara, explores theories
of action in lndian philosophy, and asks whether these ideas, which grew
out of Vedic exegesis, can have wider implications for the study of Indian
philosophy and philosophy in general.
Ju her chapter on Abhinavagupta's understanding of the concept of rasa
in his Abli:inavablufrati, Elisa GANSER asks about his understanding of the
role of dance in the overall aesthetic process of performance.
Alastair GOH.NALL, writing about Pataiijali's views on rl,krh: (universal
property or shape) and dravya (an individual thing or matter) as expressed
in his iVfahabhrl,,c.;ya, speculates on whether Pataiijali's views were influenced
by the Buddhist monk Dharrrmtrata.
Gergely I IIDAS's "Rituals in the MahJi,srJ,hasrapra:mardanasfitra" presents
a textual study of this Buddhist Tantric ritual text, featuring first an iutroduction detailing its sources and contcxU-;, antiquity, provenance, structurn
a1l<l contents, followed by a smvcy of the rites discussed in the text, and
finally an ediLioll aud translation of a passage describing a ritual for the
prot(~ction of the state.
, Examining clifferent versions of the Lii1godbhava myth found in early
Saiva texts, pnrii1.1ic sonrccs, awl iconographical representatious, Nirajan
KAFLE argues that this myth (detailing how Vif:)l_lU and Bralnna, install all(l
worship the first l'iri,ga of Siva) finds its earliest articulation in the ivadlwrma,foslra, and traces its later development as the l'iriga come,'"J to have
anthropomorphic and thcomorphic form.
In a study of "Yantras iu the Buddhist Tantras", Kcnichi KUH.ANISI!I
examines a range of peaceful and aggressive yantra-bar:;ccl rites outlined

xii

I>n.EFA<:E

within the l<r'?'(Wya:rrui,rilanlra, a ninth-c<~11tmy l311ddhisL TanLric L<~xt; in

addition to presenting detailed ritual proccdttres for two of tlic nine rites,
Kuranishi demonstrates how yanlras arc cmploy<~d through a critical edition
and translation of select textual passages describing their production and
deployment.
,
Through an examination of the earliest extant sources of Saiva Sniddlia
rituals (pre-12th centmy), Nina MmNIC disc1tsscs th<~ adaption of this ritual
cycle into the 8aiva ritual repertoire in the light of the tradition's movement,
towards the Brnlnnanical mainstream of housd1old<~rs in tlw early medieval
period, pointing to doctrinal awl socio-rdigious ramifications expressed in
the texts.
Ayako N J\KJ\M lJHA's contribution tackl<~s the intcrtcxtual relationship
between two important works in the Yoga.ccira school of Mah<iyana Buddhist
philosophy: the ninth chapter of the MaliJiyrinasfilrr/,lmriklira and the [Ju-.
ddhabhfrmJ.<;utra; her clos<: textual study of passages discussing the notion of
Duddhahood reveals that the latter text docs not foatnr(: a systernati:;;atioll
of the four types of gnosis (jfi(Ina) found in tlw former, leading N akamnra
to suspect it to he the earlier composition.
Andrew OLLETT's chapt(~l' on the classical Indian metrical tradition
contains an exposition of the traditional analysis of the ga:wicclw:rulas family
of metres, and a (lif.:icussion of the historical context of the texts which have
been handed clown to us.
Antoirn~ P J\NAioTI invcstigat<~s the soteriological dimensions of the Buddhist doctrine of analrnal(/,, asking about its grounding in th<: canonical
Buddhist discourses and how the ideas in q1wstion became prevalent in
classical Indian Brnldhist circks.
Isabclk~ H,ATII~; shows how the arnbignitics introduced hy the process of
en phonic combination ( sa:nrlhi) in Sanskrit kxts ca11 hav<~ very important
implications for philosophy when she examines a passage from Abhinavag1tpta's ffoarapratyabli:Uiirivfrrw:r.~in:i:
Through a reading of various ferocious (nyra) depictions of the goddes~
Durga within a rang(~ of <)arly medieval pmi11.1ic and poetic texts, 13ihaHI
SJ\H.((J\J{ argw~s that Uie early mediaeval S<ikta uwlcrstaudiu; or her forocity (ugralll) involved both a con11ectio11 with antinomian pradiccs as well
as a synthesis with a b<~nevolcnt and auspicious uatun~.
In advanc(~ of a new critical edition of the 011:r"apa:ikri.<ikd, a hrid Buddhist Tantric work describing tlw ideal charnd<)ristics of gm11s and disciples,
P0tcr-D{wicl SzANT(>'s contrilmtion presents a compld<~ diplomatic cditiou
of a previously unlmowu Sanskrit manuscript of this text, archiv<~d in Nepal.
vVorkiug from a dos<~ n~ading of a portion of the Dvaitn philosopher
.Jayatll'tha's corn1n<mtary 011 Madhva's critiqtw of Non-Dualist thought, (,]ic

xiii
Tallvodd;i;ola, Midmel W1LLIAMS explores l11<lia11 philosophers' urnlcrstandiug of the 11a1.1m~ a11d limits of i11[ercnce, in particular tlw long-standing
question o[ whdhcr we can infer uucstablislwd (apra8'iddha) entities.
The title of the present volume has a dual significa11cc. On the one
hand, 'fJ'/('iJJ'lk(/,' denotes a scribal flourish used by scribes in 11u1m1scripts
to set apart difl'cre11t parts of a text
cliaptcrs, colophons, endnotes in manuscripts. vVe thus wished to allude to the f'nct that most of the
research presented herein is text-based, ancl that many of these texts arc
only available to us in unpublished rnauuscript.s. On the other hand, given
that 'p'U.~JJ'ikci' in Sanskrit also means 'a small flower', we also hope that the
new generation of lndological scholars whose works arc pnblished in this
volume will continue to flourish and further the field in their future careers.
vVe would like to express our most sincere thanks to the Arts ancl Hnrnanities Research Council and the Oriental Institute, Oxford, who both
provided generous financial support to the project at different 8tages. vVe
arc also grateful to another contributor
who wishes to remain anonymous for making it possible to invite young scholars from as far away
as Pondichcrry. Beyond those already mentioned we would like to thank
the following: Slmbhani Sarkar for dc8igning the logos; Adrian Girstei who
has very kindly agreed to help us with hi8 cxpcrti8c in IT matters and de8igned the website of the IIGRS (still active today ancl already advertising
the fifth such event); Dr. Csalm Dezso (ELTE I3uclapest), who acted as a
senior advi8cr and provided us as8istance in type8etting; Christina Ada.ms;
the Warden and Fellows of Mert.011 College, Oxford; the staff at Oxbow with
special thanks to Clare Litt, Val Lamb, and Julie Gardiner; the staff at St.
Hilda's College with special thanks to Sarah Brett. Finally we should thank
the participants ancl contributors for their hard work and patience.

The editors

One
Defining the Svara Bearing Unit in the
sik?avedanga literature: Unmasking a
veiled debate

Giovar1ni Ciotti*
1 Defining the framework
A long-standing unspoken debate within the Sanskrit grammatical tradition concerns the definition of the nature of the Svara 1 Dearing Unit [SDU],
i.e. which part of a word bears the svarn. The very label of "SD U" docs not
in fact correspond to any specific Sanskrit term, but is a calquc -- here used
for the first ti1rn~ from the highly controversial concept of Accent Dearing Unit [ABU]. The latter bas been at the centre of Western Linguistic
speculation at least since the 19~Hls, 2 and its elaboration has substantially
*My thanks to Dr. Viw:eu:r,o Vcrgiani and Mr. Alastair Gornall for their valuable
commcuts 011 earlier drafts. All faults remain mi1w alone.
I On the definitio11
the term svara, sec :~.
2
Frorn .Jakobso11 [l!l:\l] and Trubct:r,koy [rn:rn], through Garde [lDG8], to Chomsky
and Halk [ UJG8], acceiituatiou has lH)Ctl dealt with by many scholars and by many diffcrm1t theoretical friLlll<)WOrks within the modern vVcstern Linguistic tradition. In this article
(sec:\ awl 4), the label 'Western Linguistics' refers exclusively to the post-Sound Pattern of Bn,qlish (Chomsky and Halle [1DG8]) generative strands which am dmracterized,
on Ll}() otH) lm]l(l, by the mdrical and prosodic approaches to the treatment of acn~ntna-

or

CIOVJ\NNI CIOTTI

8ontri\mtecl to the reshaping of the overall interpretation ol' tlw architecture of the phor1ological component ol' grnmmar.: 1 Through the smvcy of
various treatises, it will he shown that the Sanskrit grammarians have also
been engaged in an effort to dr~scrihe the characteristics or a sp<)cific set
of li11g11istic <~ntitics an<l their relation to tlw sua:ms, a speculation which
echoes the quest for the definition of the AI3U and which has led to the
formulation of different views.
However, before moving any further, it is essential to say a few words
about the similarity that is postulated here between Che Sl3U and the ABU.
The very nature of the language speculation is in fact intrinsically manifold: it depends on the point of vi<~w of its composer, on the tools both
intellectual arHl technical which have been used in its formnlation, ancl on
the aims which have been punmecl. 11 As a consequence, it should he asked
whether the topic that will be tn)atml here represents or docs not represent
the same problem for the Sanskrit and the Western traditions.

1.1 Sanskrit as a corrnnon ground

As with any sort of language speculation, including both the Sanskrit and
the Western linguistic traditions, the aim eonsists in formulating an interpretative and descriptive representation of its object by gathering general observations from a variety of (epi-)phenomena, i.e. by representing its
grammar. Since this can be done either through comparisons with (lifforent languages, or through the analysis of a single language, although the
Sanskrit grammatical tradition has no comparative interests, it is possible
tion and, rm the other hand, by a modular approach Lo the architectural representation
of the la11gnagc. Cf. lk<:lmm11 [lmsG], Hayes [l!HJG], and "Part Ill: l\fotrical Structure" in
H.aimy and Cairns [200~Ja).
:iit shonl<l he reuwmlH:rcd that, in Western LiHguistic terms, 'grammar' means 'U11iv1~nml Grammar', i.e. the nnivr~rsal cognitive rncdmnism which, according to the settiug
of its parmnders, can produce all tlw possible shapes by which the ard1itcctnrn of a
language can be charactcriz1~d.
1
' Suf!icc~ to say that the Sa11skrit grammatical traditio11 :mcms to havc aimed, at the
beginning of its history, to provide tlw Veda:-: with ancillary scicucc:-: (cf. 2) used for
preventing them from un<krgoilll-'; modifications, and, in a s1:co11d stage, to provide t.hr:
intellectual and political establishment with an 1111cha.11gi11g symbol of prrntig<~ ( unmn1mis
opin'io), along with a vali<l tool for the teaching of Sanskrit, (d. D(:;;hpandc [200G] and
Kiparsky [2007]). Ilowevr:r, Wcstcrn Linguistics is intcrc:-:ted in proviclinv an dcgant
in the s1:nsc that it shonlcl comply with the Ocklmrn 's razor principl(~ 1'.cpn:sm1tatio11
of the functioning of the coguitivc proecs:-:c:-: whose pmccptihlc manifestation is la11g11agc
itself and whos<: parnrnnt(:rs arc the samc for all lmman beings (U11iversa.l Grammar).

Ti!E SVAH.A BEAlllNG UNIT

to compare its methodologies ancl achievements with those of the Western

Linguistic traclition.r>
Thcrcrore, Sanskrit itself provides the common ground in which the two
traditi011s can operate: on the one lmnd describing the characteristics of
this language is the very aim of the Indian grammatical tradition and, on
the other han(l, Sanskrit represent:'! as any other language might do a
8Uitah1c field for testing the validity of the theordicnl models provided by
Western Linguistics.
Furthermore, although quite problematic in its definition, in particular
within the Western tradition, tlw object under investigation i::J basically one,
namely accentuation. As it will be shown in ;), it is possible to state that
the two traditions have gathered the same general observations while dealing with articulatory phenomena such as the modulation of the vibration of
the vocal cords. A remarkable consequence of this common unclcrntanding
can be seen in the fact that, according to both traditions, it is necessary
to split words into minor subunits and to assign to them the property of
being bearing units. In fact, jnst like some strands of the Western Linguistic
tradition have divided words into minor units, i.e. units different from morphemes, like segments, syllables, or, more generally, prosodic constituents
(sec 4), in the same way, the Sanskrit grammarians have dealt with vo,r~ias
awl ak,'ia.ras in order to understand which part of a pada ("word") bears the
svo:ras (sec 5).

1.2 Phonetics and phonology in the Sanskrit

tradition
However, what. seems to represent a common problem to which a similar
mrnwcr has been provided should actually be embedded in two remarkably
distinct. syst<~ms concerning the overall architecture of (Sanskrit) grammar.
As a. matter o[ fact, the frameworks in which Sanskrit accentuation, or Sa.11skrit sva.rns (d. :3), hav(~ becu analy~c<l arc remarkably different from those
of the \1Ve8tcrn tradition, and it wonld be rather rn.iivc to start our comparison without clarifying this point. In fact, Western Linguistics distinguishes
two separate modules concerning how the grnnnnatical architecture of a
langnag<~ deals with the sounds of the language itself, namely the phonetic
-------

"111 this sctlS<\ the diffonmcc between 'lingllistics' and 'g;rmmuar' cau be overcome.

CIOVJ\NNI C10TTI

component and the phonological component,(; whereas the Sanskrit tradition never opcrat<~s at least manifestly within such a formal distinction.
Although within vVcst<~rn Li11gnistics the debate about when~ to draw
the precise limit between phonetics and phonology is still open (cL Purnell
(2009])' it is possible to define phonetics as the study or the articulation
and perception of the sounds that human beings use in speech, whereas
phonology consists in the study of tlw sound patterns dmracteri~dng the
overall properties of contrastive sound inv<~ntori<~s, of the distrihntiou of
sounds and of their variable rcali'.l:ation in different contexts (alternations)
(cf. Dlevins [2009: :J25]). In other words, phonology studies the phenomena
concerning the sound repertoire of a language while they arc nsccl in speech
production.
It is particularly important hem to highlight a peculiar characteristic or
the phonological constituent as formulated by many strands of tlw vVcstern
Linguistic speculation: phonology may contain abstract entiti<~s. The postulation of these entities is justified by the urge to establish an interface with
the cognitive apparatus of the brain, along with the principle of descriptive
economy which hrn:i led to the forrnali'.l:ation of the phonetics/phonology dichotomy itself. According to Vaux and vVolfo [2009: l:H], "hy postulating
highly abstract formal entities such as syllables [... ], we begin to bring order
to a vast array of seemingly disparate facts that would otherwise remain
unconnected and unbelpful." 7
On the other hand, the Sanskrit grammatical tradition displays, in more
or less systematic arrangements according to each treatise, the description of
the articulation of sounds and tlw list of phenomena pertaining to them (in
the Hense of their cont<~xtnal distribution, i.e. sandh'i). Even whm1 it s<~crns
to forrnali'.l:c entities bigger than a single som1cl (e.g. the ak.~a:m, broadly au
entity made of a vowd plus a consonant), it ncv(~r op<mly st.ates that they
Gin this article the term 'sound' rnfors 1.o tlw phonetic domain, whereas the term
'segment' i.e. the rcpws<!ntational device wliern a. certain mm1hcr of tlw mticnlat.ory
features arc said to crwxist rder:s to th<! phonological domain (cf. Harris [2007: 12'1-:I L]).
The latter is prnforred to the term 'phoneme'.
7
Cormnenting on Vaux awl 'Nolfe [200)], Ckments [200!): LGGJ stat.cs: "[ ... ] not all
phonological concepts have phonetic correlates. For example, though tlic syllabic is an
essential unit in phonology (and undcrli<~s many asp<)ct.s of phmwtic and prosodic pat.terning as well), it. has no 1mivcrsally vali<l phond.ic rldinition. This fact is not. smprising
once we recognize that the :syllable is primarily a phonological c<lllst.nwt., ddiucd over
sequences of discrete phonological scgrnm11.s rntlwr than ovnr phouetic primes as such. At
thi:s level of abstraction (which incl mks most of phonology), fow cons1.rncts have direct.
I!honetic definitions. Vaux all(l wolfo rightly cmphasiz<! Uiat the ultimate justification
tr:r such co~1ccpts rlq>cnds on their success in bringing order to a vast array of seemingly
chsparatc facts. The syllabk does jnst that." Abont tl1c importance of the syllable in
connection with accentuation, sec 4.

TIIL~ 8VAB,J\ 13EAIUNG UNIT

nmy have no articulatory reality, nor arc they ::iai<l to reprc::ient any kind of
cognitive cntit.y. 8
Therefore, it should be expected that any kind of compari:son between
the spcculatiom; made by the vv(~Stern and the Sanskrit traditions will be
affected by this rn(lical difforenceY

2 Defining the sources

When I rcfor to the (Sa1rnkrit grannnatical tradition', J mean here the huge
textual production embedded in the literary label called vcdari,ga. This corpus of six ancillary sciences was initially developed in ancient India in order
to study and preserve the Vcflic scripture and knowledge. Among these
six sciences, vy{i,/;;o:ro:r.w and ,4ik:;a pertain to the speculation concerning the
Sanskrit grammar.
The first work testifying to a vyakara~w, (lit. ('discrimination") kind of
speculation have been lost, and the most ancient known text we possess is
Pai.1ini's A,<>f,adhwiyf ("The Eight Chapters", ca. Gth-11th century B.C.E.),
a sample of an already fully developed degree of grammatical speculation.
H.oughly, the whole of the subsequent vy{iko:ro:r.w litcrntnre is constituted
by commentaries on the A,<>f,(Idhyayf: those produced by the so-called JHL'r_L'irl'i/yas, i.e. the grammarians who followed more strictly Pa1.iini's teachings - in particular Katyayana (ca. :Jrd century B.C.E.) and Patafijali (ca.
2nd century D. C.E.) -- and those produced by the non-pii~iin~1;as. A definition of vyakanrr.w through vVestcrn Linguistic terms would state that it is
the science studying the morpho-phonological, morphological, ancl morphosyutactic components of the language. Included in the fJ(L'r_tin'~7Ja tradition,
although not fully in line with it (cf. :~), we find Santanava's Ph:if,sfi,tra
("Aphorisms on [the accentuation of] the llominal bases"), 10 a treatise about
the principles according to which the positions of the svaras are assigned
within the word.
8

For the implicatious of the concept of laghava ("ccouomy") in the Sauskrit traditiou,
see :L
0
This [idcl snrdy deserves to he widcuc<l. In pa.rticnlar, a comparison between prcSaussurean vVcstmu Linguistics, i.e. prn-langue/J1arolc dichotomy (e.g. the works ol' tlw
first sd1olan> c11gagcd in the rcconstrnction of Proto-Iudo-Enropcan) and the Sanskrit
grmmnatical Lraclitiou could prove to ]Jc remarkably i11teresti11g, since the former still
opcratcH in a framework iu which phonology has not yd, been delinccl as a distinct
comp01ic11t of grammar.
10
Although dating the Pli:f,s'/LJ:ra is almost impossible, a. smrniblc gncss places it at the
end of the first millennium D.C.E. (cf. Cardona [197G: 175-7]).

(: IOVJ\NNI ( J10TTl

The kik.?cl,vedciri,ga literature inclwlcs two categories o[ texts: the pn/,l:iMi11

khyas ("[belonging to] <!1H:h (Vedic] school"), from ca. 500 to 150 B.C.E.,
and the .~'ik-'j(LS ("teachings"), ca. 11th-15th ccntmy C.E. 12 In Western Linguistic terms, wherea!-l every pnil:i.fo.khya deals systematically with what we
could define as the phonetic and phonological aspects of specific :-mhsets of
the Vedic: litcrntnrc called 8arr1.h:il(l,s, the kik-?iis arc mostly short monographs
about specific phonetic awl phonological topics, which seem to pmtaiu to
Sanskrit grammar in general rather than to a. small literary corpus. Both
genres arc usually attributed to specific Vedic families or schools, most of
which remain almost completely unknown except for their narnes. 1:i
Hernaftcr, the VJJ(Lkara~rn speculation will he investigat<~d to trace a parallel between the ddinition of .'ivara and that of accent (see :~), whereas
the s'ik.?ii.vcdii:riga speculation will he used in the discu!-lsion conc<~rning the
definition of the 8BU (sec 5). In particular, the following texts will be
referred to:

(a)

IJ.gvedaprati.~iikhya:

pnit'i.Siikhya of the Jlgveclo.sarr1,ht(I,;

(b) Vc~jaso:ney'ipnlt'i.<fakhya: pnil.4akhyo. of the ,guklo.yofm"uedasaqih:iUi;

(c) /JcmnakZycT Calv:rii,dhyiiy'ikcl,: priit.Z-4lLkhya of the AthaT"1Javedasaq1,h'iUi;

(<l) Pc1,~dnfya.4ik5a: the most rnnownccl among the .{ik.~(is, traditionally not
attached to any specific Vedic school;

(c) Ycl,jiiavalkya.4'ik:wT: .{ik:wl, of the Su.klayafu:rvedasarrd1:itrl,;

(g) LomwjT.4.Zk:'irJ,: .4-ik.~a of the 8r1:mavcdasa:rr1,h:iUi. 1' 1

11

Also known as JHl'f':'i(/.d!LS (" [bdonging to] an assembly"). The two llame~s or this gcurc
clearly refer to the strict rdation bctwce)tl these (,e)xts and the) sacerdotal families, de .Ji1.clo
the Vedic schools, in charge of preserving tlic Vedic texts.
12
Any attempt to dat.<~ these texts is necessarily tent.at.iv<): for a ddail<)<l discussion
on this topic cf. Scharfo [l!J77: 127-:\4, 17fi-7], who says "The datiug or most or these
texts is 1wxt. to impossiblP" [p. L7fi], and Varma. [192D]. The corn.mnnis OJlinio that t.lw
prnli.fo/,;hyas arc older than the .~ik.~as is bascel on the fact that tlw fonrwr am compos<~d
in .mtru.s a style typical of the most ancient ucdariyn trcat.ise~s , whereas t.lw latt.<~r me
composed in verse or in prnsn.
t:ilu a later stage (cf. Paranwswara Aitltal [lD!H]), tlw kil.::'>avulo.ri,go. texts am classified
under the label of lak:'>H[W., i.e. the literary genre which comprises all tlic t.<~xts containing
infonnati()n about and instrnctions for thn Vedic recitation.
H For t.lw attrilrnti011 of these treatises to tlwir respective~ Vedic 1.<~x1.s, cf. Mishra

[l!l72].

3 Defining a pron1inence
Providing a definition ror the notion of accent has been one of the most
challenging tasks Western Linguistic scholars have been engaged in. Herc,
I will present a concise summary of the most recent ideas concerning this
notion (cf., for instance, Cairns [2009], Hayes [1995], and ldsarcli [2009]).
Nowadays, the majority of :scholars believe that accent represents a 8yntagrnatie prominence realized by various phonetic means-, i.e. a prominence
characterizing one out of a certain number of constituents proper to a specific subdivision of the spoken chain, corresponding to the word in the case
of the lexical accent. 15
Do the svaras correspoud to this notion? There is an overall agreement among the Indian treatises in recognizing three main types of svara8,
namely 'udii,tta ("raised"), anudatta ("non-'IJ,dii/;la") and svarla (literally
''sounded") Hi - the latter being defined as the combination of an 'Wll'itta
plus an o"mtd(J,lta. 17
As Par.1ini clearly states, there is only one ndatta for each pada ("word"):

6.1.158 11
"A word is without 'tJ,datta with the exception of one."
(],'fl,'l/,d(J,l[am padam ckavarjam

11

Every udaUa is then followed by a svarita, and these two arc snrroundcxl
by a variable number of armd(i/;las. Therefore, the translation of the term
svam as "accent" is rather inaccurate as it does not properly fit the phonological theory developed by Western linguistics. In this sen8C\ only the 'tUilitta can he said to be the accent - i.e. the prominence - whereas the anudlitta
simply represents the absence of accent.
On the other hand, another possible translation of armd(ilta is "unraisecl"
00 that its meaning woukl rcfor to its phonetic characteristics: in this sense
the ambiguity of the translaLiou works to om disadvantage, making any
cla0sificatory effort conccrniug the theoretical framework of the Indian traclition difiicult to define. In fact, it i0 not possible to discern whether the
1

r;Tht) ddiuition of the notion of 'word' is not treated here. For a ddinition of tlw
term 'word' in phonology, cf., for instance, N<~spor and Vogel [198G: 10!l-l4d].
1
Git slioukl he noted here that, alth011gh presented in the treatises as compulsory for
tlw correct pronmH:iation of the language, the svnras have been probably lost in the
translation from Vedic to Classical Sanskrit within au overall dmugc of the accentuation
pattern. Ad<litionally, accordi11g to sottw trnatis<~s, a fow more sva'm types should be
acldecl to this list, lrnt they arc probably variations o[ the main three, pertaining to
specific recitational styles (cf. Dcshpa11clc [ L997: tl:l8]), or particular combinations of the
basic sva:rn.s.
.
17
1\1jf,ridhyriy1, 1.2.:ll says swrruih();m.~1, svarUa(1, "svarita is the sam.r1,hii.ra ("combination") [of high pitch and low pitch]."

( ~ lOVANNl C!OT'l'l

terminology concerning th<~ svu:ras is strictly phonetic-based, or poss<~sscd of

a 'phonological flavour' ill the sense that it providc~s a pic~ce or infornmtioll
concerning the distribution or the~ accent within each word. 18
Therefore, a suitable translation [or S'll(l,rn is "pitch modulation", a translation which rcnckrs its etymological meaning more closely. It should be
noticed that the term S'IHLTa is used also to indicate the~ vocalic sounds. The
consequences of this amhignity will he shown in G.
However, at present, it is possible to state that both Western Linguistics
and the Indian tradition rccogni;1,e the existence o[ the same phenomenon
i.e. the prominence labeling it either as accent or as wlrilla, respectively.
Dnt thrcmgh which representational model do these two traditions outline
the same phenomenon? In other words, how do they ddine the concept of
prominence?
From a phonetic point of vi<~w, vVcstcrn Linguistics describes the natnrc
of this prominence as "parasitic" (cf. Hayes [1995: 7]): its articulation can in
fact be realized through quite a large variety of articulatory means used for
other purposes at the same time. Typologically, two kinds of accent have
been postulated: stress accent and non-stress accent (cf. Deckman [1986]).
The former is reali;1,ed by a combination of more than one phonetic means,
usually a combination of the vibration o[ the vocal cords (whose perceptual correlate is 'pitch'), duration, intensity (whose perceptual correlate is
'loudness'), and variation of segmental quality (for instance, in English),
whereas the latter is mainly realized hy a faster vibration of the vocal conb
(for instance, in .Japancsc). 10 In this sense, the etymological meanings of
the three .'Jvara names and their comparison with musical not<~s
quite
20
common within the Sanskrit grammatical tradition is at the base of the
cornrrmnis opfrliio that according to the Indian tradition, pitch is the main
phonetic characteristic of Vc~dic accent. vV<~st<~rn Li11guistics in particular
Historical Linguistics according to some of its strands rutchcs tlw sanie
conclusion through the comparison of the .'i'Uara system with the accentual
d1aracteristics of other [ndo-Enropean languages (d. Clackson [2007: 75-8])
and throngh a particular iutcrprdation of tJw natm<~ of the sva"rif,a. In fact,
a well-known property of tmw languages, i.e. languages in which pitch is
T\i') translation of il:'>f.wlhyayt fi. l.1G8 sho11l<l thnn IH) d1a11g<~d to: "[the s1111,rn.s] of a
word an) wm-rniscd with the exception of one". Tbc probl<~lll of tlH) do11hlc iutcrprdatio11
of the turrn anu1liilta is already discussed i11 Patafijali's Mo,ho,/1ha~y11, (cf. Cardona [I !l88:
18

442]).
1

!Jin addition, the accm1t can lie enhanced by the fact that it triggnrs some phonological
phenomena. A well-known example is tlw so-called 'V<~nwr law' (cf. Clackson [2007: 7G-

(i]).
2

For instance, d. Pau,iniya,4/,;~a 12 and Narwli,,~ik"'>a 1.8.8.

llS(Xl for lc)xical oppositions (for instance, Mawlarin Chinese), is that pitch
can spread its clmracLcristic either backward, merging then with the preceding low tone alJ(l n)a.l i'.(,ing a rising Lone, or forwards, merging in this
case with the following low tone all<l producing a falling tone. Once we take
the traditional definition of 8'IH1:rif;o, into account as well as the constraint
governing its position within the word, it seems therefore plausible to ddine
the svarita as a falling pitch rnoclulation, resulting from the merging of an
ndrYlta and an anwllYtta (cf. Halle [1D97: 28G-7, fn. 9]).
On the other hand, phonology provides a different approach to the concept of promi11cncc. At present, although Hynmn's [2009] brilliant article
demonstrates that our understanding of the phonology of the accent is far
from having reached a satisfactory stagc, 21 the two main phonological typologies according to which accent is placed within a word are considered to
be the rhythmic accent and the morphological accent (cf. Hayes [1995: 31:3]). The former represents the tendency proper to all languages to arrange
its clements in rhythmic patterns, whereas the latter represents the outcome
of a complex interaction of phonemic, morphological or syntactical contrasts
and structures, commonly known as morphological accent. Vedic accent is
usually assigned to the latter typology: its position within the word would
depend in fact on the accentual properties of the morphemes composing the
word. During the clerivational process from the underlying representation
ancl through various cycles, the position of the accent is adjusted according
to the specific properties of the morphemes, i.e. for instance accentuated vs.
non-accentuated (cf. Halle ancl Vcrgnaud [1987h: 57-Gl]). 22 In keeping with
this interpretation, Pa1.1ini assigns to each suffix the property of modifying
the position of the accent within the wonl. 2:3 Unfortunately, this represents
a further complication: from the Western point of view, such operations
do not simply happen in Lhc phonological component but in the morphoph011ological onc?1 i.e. they understood either as rnlcs or as constraints
21

Prc:mmahly, Hyman's article will havo a profound influence in reshaping the common lmowlc~clgo vVestcrn Li11gnistks has developed in interpreting the phonetic corrclatcH
of acccmtuation. Since it is irnpossibk to forecast the results of this reshaping, here arc
presented the most accepted phonetic and phonological framnworks Western Linguistics
lias developed so far.
22
h1 particular it is lwrn the phonological metric module which interacts with morphology. For the implications of these model iu the overall architecture of the linguistic
representation, cf. Cairns I200!l].
21
: It should he remembered that Pii1.1i11i does uot provide rnles for attrihnti11g the
accent to the nuclcrlying representation of 11omi11al aud verbal bases.
21
Thc cldinitiou of morpho-phouology within tlw Western linguistic tradition is not
nncontrovcrnial (cf., for instance), Molmrmu [ lD%]).

-------~-------

-----------------.....,.., ....,...,,....,.,,....
.,..w~""''"""'""""""""_,...,..,.,..........,..,=---- -

...- - -----

(j[()VJ\NNI CJO'J"l'I

10

otr soll!H ls, a.rra11e.:c11wnt

and of
pertain at Uic same Lune
to tl, 1c (l ormuns
.,
word formation.
i
On tlw other hand, a different explanation for tlw position of t\ic yro!ll -,

A
img t.o t,l ns
' t,real
ncne<~ is provided hy Santmrnva s J>/ 1:tf,s11./;ra.
nccon
'JS<'
,, the
- docs not (lcpcrn l on t l ic property o 11\.11C motPlwmcscomposition of 'UllaJta
, ." . , .
posing; ~1 sp(~cific ym.<f:a., lm_t on the lcx~cal propcrt:i(~S '.;! :J~c.1u1,~la. it~di, ~'.\,~'.:~~
semantic or phond.Jc (d. Devasthali (<~d.) [19b7]).~' I }us shows thtl
same topic can he trnat<~d in (liffm-m1t ways also within Che lwlian gi:am~
matical tradition. Ncverthd<!ss, even in this frarn<~work, the positio11 _ol tl~\
'U<iatto. depends to sorrw extent on thr~ abstract properties that arc attn bu tu
to some sounds so that they govern the position of tlw accent.
.
The fact that the two Sanskrit grammarians, although from dilfercnL

pomts
o f. view,
provI< lf! general o l)scrvat10ns
rcgarc l"mg tl, w pos1t ,1()11 of the
worels, ms
t;ea( l o 1 sunp
ly cornp11mg a ilSl,i 01c wore
ls with the
acccn t w1"tl~ un
indication of the accent position, is somewhat reminiscent of the representational device of Western Linguistics which is phonology (or morphophonology). If this is admitted, it seems plansihk to inl"er that, although the
very iclea of phonology is not found in tlw Sanskrit speculation, nevert~ic
less some of the Sanskrit grammarians moved b()yond the simple descri~>)~;IOll
r l
. 1
t
1
t.
~<> At
01 t ic art1cu at1on o sounds towards a certain <lcgrcc ot a )Strac ;Hlllthe very basis of this step lies the principle or lrighava ("economy") which,
although devoid of any cognitive implication, lc<uls the grammarians to provide general rules instead of lists of <~xccptions. In other words, laglw:ua is not
the core of the (cognitive) plausibility of lmw;nagc description but rather
of its cifoctivcncss. 27 As a consequence, as I<iparnky [200D: :H] n~marks,
"[ ... ] "syntax", "morphology", awl "phonolog;y" [... ] arr~[ ... ] emergent constellations o[ rnlcs rather than prndd,crmincd comporwnts into which the
description [of the grammar} is organizc~d." Tlicrdorc, the whole spcc11latiot1
of the Sanskrit tradition is diaracU~rizr~d by a u~nsion hct.wccn the ass11rnp' 1 --

'

. 2,;Fo;. instar1cc, it~i~ ;~ott,l~~ ~;~~1t:cr suffix which dctnnni11cs tlic position of the acccllt
\mt it is tlw propmty of the word of being 1wutcr that. irnplics a part.icular posit.ion
for the accent: atluuli(i prak .fokn[cz1, \l 2.1 \I [... ] nalmi:myasyanisantasya 11 2.:\ 11 "Fr<~nt
lwrc the initial syllabi<~ (is 11,datta) prior to .fokn[.i. (2. l) [... ] Of (a worcl) ucccssnn!r,
neuter in gcn<lm (awl) not <~nrli11g in is, (tlt<) initial syllali1<) is nrlalla) (2.:1)" (fkvasthali s
translation [UlG7: GfJI). According to S;intanava, it is uot tlin sidlix to he JH)tlt.cr hitL the
lexical item, i.e. gender is not a morphological propmty.
2 G8o:rulhi plwnomcna, which ar<) not tr<)atc<l lwrc, arc encompassed wit.hill Llic salllc
theoretical hori~on.
27 ln thi;-; S<)IlSC, thn far:t that SOllletirnes 1.he grnmtnariat1S do provide lists of cxcnptiotJS
should he sc<m as morn in keeping with the idea of layfw,v!l namely n10rc cco1101nica,l
181
than providing <~xtn~mdy mrnplex <lcrivatio11al rnks. For iusta11u~, d. Jlij[.o.dh!)<Lf/l 2.2.

:u.101, etc.

TUE SVJ\H.A BEJ\BINC UNIT

11

tio11s hascd 011 articulatory evidence and the abstractions devised for the
effectiveness ol' tlw description of the language. On the other hand, the fact
that sometimes Western Lingui:-Jtics and the Indian tradition reach similar
co11ch1:-Jio11s is clue to the application of Che Ockham's raz:or principle hy the
former for the sake of the cognitive plausibility of language description.

4 Defining the Bearing Unit

Besides the definition of the nature of the accent, Western Linguistics has
also been engaged in the definition of what clement of the word bears the
accent, i.e. occupie:-J the prominent position. This task has turned out to
he remarkably challenging, in particular because it implies an interpretative aualysi:-J of the available phonetic data characterhed by a high level of
ah:-Jtraction.
From a phonetic point of view, accent may generally he defined as multisegmcntal: specific combinatious of sounds within the same syllable allow
a better perception, in fact, of its accentual characteristics. For instance,
the vibration of the vocal cords, which is the primary characteristic of the
Vedic accent, is more easily heard when the consonants following the vowel
arc sonornnts: 28 in fact, to each fundamental frequency produced by the
vibration of the vocal cords during the articulation of voiced sounds, corresponds a series of harmonic frequencies, which arc refractions of the fundamental one within the articulatory channel. The lower harmonics arc
particularly intense in the souornut flcgrnents, whereas they arc absent
since the speaker docs not even vibrate the vocal cords in the articulatiou
of unvoiced segments (cf. Gordon [2006: 85-90]).
On the contrary, the ABU has proved to be a powerful abstraction used
to explain various phonological phenomena and whose very definition still
nuder debate (cf. Cairus [2009] am! H.aimy and Cairns [2009h]) - has deeply
contributed to rcddiuiug 1.he shape of the overall archi tectnre of phonology.
The ABU is nowadays mainly defined either as a single segment 11arnely
a pho11cnie or a slot within the word structure, depending on the different
theoretical frameworks or as a syllablc. 29
28

S011ora11tt-1 am thot-:<! t-:omHIH whmm articulation it-: voiced, like vowels, nat-:alt-:, liquids,

etc. (cf. Ilall [2007: :~ltl- G]).

'.l!J A more flexible apprnach to the ddinition of the Al3U is found in Ilallc and Vcrgnand
[l!J87a]. According to them, the ABU it-: au 'clcmm1t X', i.e. an clement which can change
from la11g11agc to language: t-:cgmcntt-: (both vowels a!Hl [+mmorn11t] c01is01m11ts in particnlar pot-:it.i011t-1), morac, t-:cgmeutt-: iu the rhyme, t-:yllablet-:, or lexically clct-:ignatecl t-:egments.

( iIOVJ\NNI C10TTI

12

In the first case (cf. Hall(~ and Verg11awl [l987a1), syllahlc awl acccu~.
information hdong on orthogonal planes wh<~rc scgiw~nts arc arrayed alonf
tlw line ddincd by their intersection. The syllabic plaw~ contains the s~ -.
lahlcs which are the domains for specific phonological operations, whcrct1s.
'
. t l ic so-ca ]]<Y l md.nca
. . ] {"ect, .i.e. S(~q11c
'
'I )("C'S 0 f
the metrical
plane contams
."
.
0
segments, whose shapes dcpcrnl on the specific accentuation properties . {.
each language. ln other words, the metrical plane rcprnsrn1ts Llw locus. lot
the application of rhythmic principles either with or without morpbologicnl
constraints.
In the second case (cf. Nespor arnl Vogel [ l 98G]), tlw ABU is usually the
syllable, understood as the proso<lic constitum1t phonological suhm1i~,
of the word. Each word is in fact sai(l to possess a prosodic structure, i.e.
a phonological structure different from the morphological one, having the
form of a node-tree, where lower c:onstit1wnts are governed by higlwr ones:
segments are governed by syllables, syllables by rrwtrical fod, and so on.
Accentuation operates on the tree constitncnts i.e. the syllables along
with other phonological operations. The syllabic option for the clcfinitioll
of the ABU is supported hy a long tradition (cf. Hayes [1995: 49-50]), i~i
particular because it leads to more straightforward gcucralizations as it
allows one to account for the influence of the syllabic weight on the accent
position within wonls.:io
To summarize, the ABU can be citlwr a Hingle Hcgrncnt of a word or sornc
sort of Hecprnuce of its segments (e.g. a syllable) awl it can occupy either an
independent plane from the locus where th<~ segments am arrayed to form
.
'll
a word, i.e. the metrical plane, or the same locns, i.<~. the prosodic tree:
(loes this complex nrnlti-lincar:i2 representation [ind auy parallel within the
Sanskrit tra(lition?
-------~~---

30
: In

-~

----

case the accent is mainly characterii':cd by tlin pitch, the rnora l1m; been snid Lo
be tlw ADU si11ce, in ccrtai11 languages (e.g. Litl111anian), ar:cmiL can neat<: oppositiollS
within the same syllable:. As Hayes [ UJWi: 1l!J-GO] rmnarks, <lc!ining the ABU as thn syllable
works in particular for the s<)-call<~d stress lang11a.gcs, wlinrcas "[ ... ] pitcl1 languages [... ]
in the generative phonology tlwy can lie trcatc<l as involvini.i; tonal representation, either
in addition to or instead of a metrical rcpres<mtati,m", \mt d. Hyman l200D] for the fact
that 'pitch language' represents an impropm phonological catq1;ory.
:ii "The nnb of the issue is wlwtlwr or not tl1e syllabk is the inviolable 11 nit for l>cariug
stress (e.g. Ilaycs lDD?"> and much rccm1t works). ff it wcrn, onn wo 11 \d expect syllal>ks
to nest neatly within f<~ct; but if souw languages were to mnploy vowds as tltn strcssb<:aring uuit, then the possibility wonl<l exist that the constituents 1iccdccl for stress
might conflict with those necdc<l for syllahic facts" (Hairny arnl Cairns [200\Jli: 11j).
:i 2Dorrowiug a Sanssmcan an<l then Firthian tnnninoloi.i;y (cf. Ga.r<lc [ U)(i8]), a n111lLilinear representation of the language correspowls to the tLccessity of explaining the syntagmatic vs. paradigmatic propmtics of Eornc phollological entities.

Trrn

SvAI{,A UEAIUNG UNIT

Because of Uw absence o! a fonnali;1,ccl distinction between phonetics

a.11d phonology, the Sanskrit tradition docs not question the mono-linear approad1 to the 80tlllds phe11omcma, i.c:. it does not iuLr()(luce different plane8
ill tlw arcbi(,cd11re o[ the grammar in which thc8C phenomena. operate.
Nonetheless, the idea of Hingling out cliffcrcnt non-morphological subunits
or a. word seems to find a straightforwanl parallel in the Sanskrit tradition.
According to the common interpretation of the ~'ik._g(/,vcrfriri,ga views, in fa.ct,
var"e,ris and ak._'wnis arc seen as ~mlmnits of a pada ("worcl").:n While the [ormer a.re well-known entities corresponding to the "speech-sounds", whether
vowds or co11s011ants,:M ak,'>arns (Literally "imperishable", "indivisible") arc
imlccd more arnhigu01rn entities. Allen [195:3: 80] presents some instances
from the primary litcrntme in order to show the reasons which support an
interpretation of the term ak:jo:m as "syllable", adding that it has ah-lo been
ambiguously extended to indicate a ''simple vowel".
However, taking a closer look at the definitions provided for a/;;f:io:ra, it
appears that often this entity corresponds to a Himple vowel, either with or
without ccmsonant(s) (Hee 5).:l'i In other words, it Heems that the notion of
ak,wra depends on a property of vowels, namely their being independently
articulated sounds, contrary to the consonants that cannot be articulated
without the support of a. vowel. 3G This interpretation of the term ak,'>ara
secrrrn to be remarkably consistent with the phonetic grounding which seems
to be characteristic of most tenrrn in the hk:ja tradition. Therefore, it also
prevents the interpretation of ak,'>ara a8 "syllable">17 Unfortunately, au exhaustive survey concerning the definition of the term ak,'>arn is Htill missing:
there arc in fact contexts in which this term seems to be used in a sense close
:i:i Althongh Pxtrcmcly controversial, mal:n!s ("rnorne") could lie added to this list, bnt
they am not relevant to the present discnssion.
:M U:-mally, after a gcticml statement about the overall rnnnher of var"(IJt.s, which varies
according to tlw various treatises, the Iwliau grammarians possibly list t.lH~m accorcliug
Lo their articulatory cl1arn.ctcristics.
:i:, At b1st 1 as it will lie shown in !\ in connectio11 with the definition of tlw SBU.
Ak,~arns am in fact also classified as dtltcr lag/rn, ("light") or gwnL ("heavy"), i.e. according
to a terminology which echoes the vVestern Li11gnistic pl1011ologica.l notion of syllabic
weigltt (cf. Alica [rn:i:~: 8G-7]).
:wThis refers in partie11lar to the stops: nasals and sibilants for instance can be art.ic11lat.ecl imlcpc11dc11tly, \mt cannot he, in Sauskrit, syllable n11clci 1 i.e. they occupy a
secondary plac<~ in the sonority seal<~ (cf. Clements [2009]).
7
:i According to vVcstcrn Li11g11istics, the phoudic ddinition of a syllable concerns
mostly tlw internal di:otrilmtion ofits sonority, whereas its phonological cldi11itio11 mostly
dq><~11ds on which muncl can occupy which position in the sonority scale. This is also
the main reason for not trnnslating a/,;:rnrn, as ":oyllabic irnclcns", i.e. tlw syllabic segment
characterized by the higlwr degree of sonority.

C: IOVANNI

()1[()'[''!'1

systems
, .

to t. l ia t, oi. " s.yl)' 1])le"

' in 1mrticu1ar It mw t.al<:cs t,Iw In< 1mn g111)hi('
into account (cf. Allen [UF.i:J: 80]). .
'
. . .
o difc<rtain
trcatis<~s dcfrnc the 513U as a cornlm1,d.1011, t
None.tlicless
, ,. '
. rnts
f.,., t xtents of either a vowd or an ak.~a:m. with mw or morn consou, "l'.
crcn. c
'
.
nee o
I..
11 tl
' l 1..
s "'Cttse
'" ' ') like the ABU, also tlw 813 U corresponds to a scquc
sounds within a given wonl.
.
,
, I<'inally it should he r<~markecl that t.lw Sa.uskrit t.n~at.is<~s rdcr t.o t, 1ic
("pitch
'
.
. when <lcf'llllllb
r the
svaras
modulat10ns")
and not only t.o the '{l(lal/,a
accent bearing unit: this surdy represents a more phonetically based approach to the language, focusc<l on providing a <ldi11itio11 of the pit.ch modulation bearers rather than based on the phonological notion of prornincn.cc.
U nfortunatdy (sec 6), since the phonetic foundations for this spccnh~ttot:
arc not spelled out, a real understanding of the nature arnl aims of the
Sanskrit speculation remains quite unclear.

5 Defining the Svara Bearing Unit

It is now time to look at the various definitions of the SBU. Herc, it

should be noted that it is impossible to stat<~ which definition is the most

widespread or authoritative, since a comprchensiv<) survey of the topic
throughout the ik.~(ivcdii:ri.go, literature is still a (ksiderntmn.

5.1 Par_iiniyasu~,~a
In the following
vowel:

vers<~s,

the

Pii(l'i'fl:~1;0...;ik.?(l

seems to define t.hc SB U as a

so 'di/t'(W: " m:tl:rdhny <lbldhalo vaktra:rn (Lpa,dya. mJirnlalJ. I

'/}(l'f''(l.iiii jano.yatc le.?ii'Tfl. vibh(igo.b. [Hl'ilr:a,dhii smrta.~1. 11 !) 11
svo.ratab. kiilatab. slhanr/,t pro.yafn(L'fl:uprad(/,no.l<llJ, I
il'i var"(w.vida)1. pnihur n:ipn(rnrri tmri 'fl:ibodhatu. 11 10 11
'lldlitta.~ cdnu.drl,tla.~ ca svarila< ca svanl,.<; lraya)1. I
hrasvo dfr:qha(1. plula 'ili /Ji,lat.o r1:i.yarrui. aci. 11 I l 11
"011ce this breath, which has h<~en npraiscd awl has t.llllrnpcd at. Lile
bead, enters the mouth, it produces tlw spc(!di-sounds, whose classification is transrnitt<~d as fiv<~fold (9), according to .'i'Uo.ra ("pitch rnodnlat.iou"),
length, place of articnlation, prayalna ("primary dl'ort. "), aw I an:n[J'radii:no.
("additional effort"). Sn say those who am lcarw~d in the souuds: list.en Lo
~------

--~Jr the pccnliar

-~-----

sa.ndhi of "so 'dtr"(l.o'', cf. Ghosh [1!):18: GG].

15
this cardully (to). The restrictions [i.e. the qualifications] concerning the
vowdfl arc the three .c.;varas ("pitch modulations"), namely 'llClat;la, ar1,'1ullIUa,
and svarila, and [those] based 011 [their] length, namely short, long, and protracted ( 11) .":rn
This translation follow:-> Allen'8 [195:3: 8:3] iutcrpretation: in his opinion,
in fact, the three svo:ms ("pitch modulations") 8cem to be clmractcri:;.i;ations
of the vowels. Allen secm8 to read aci in connection to ho th svarr!,(i, ("pitch
modulations") and /.;alato ("based on [their] lcngLh").'10 On the other hand,
if one cliviclc8 verse 11 into two iudcpcrnlenL 8entcuce8, the following trnrnlation would he possible: "The svo:ms ("pitch rnoclulatious") arc three: 'IUilitta,
a:rmdilJ;ta, and svarita. Short, long, and protracted a.re rcstrictimrn [i.e. qnalificatiorrnJ concerning the vowels, ba8cd on [their] length." According to this
interpretation, the P(Ir1:inTya:~k{J(L would simply present a list containi11g the
three svarn.8 ("pitch modulation8"), without giving any 8pecific indication
about the 11ature of their bearer.
Dccause of the ambiguity of the term svara, once divided into two parts,
it8 first half could also he read as follows: "The svo:ms Cvowel8") are three:
[articulated with] wlcLtta, anndatta, and svar'ita." Although this interpretation scern8 to he the lest> plmrnible one - the definition of svlffa as "vowd"
is already given in verse 4 it doc8 not really invalidate the meaning of the
verse as given in the first translation above.

5.2 IJ,gvedapraUsakhya
By employing the notion of ak,c.;ara, the rlgvedaprrlt'i.<cJ,khya exemplifies a
different definition of the 8DU:

udcl,tlaA cil,nwliitlaA ca svarita.< ca lrayo)1, svanib,

(L;tflL'rna:ui.fra'rnbhil,/,;,~apa:is ta ncyante I\ :3.1 11
al1:!}anI:3ray(l,(1. 11 :3. 2 11

"The svo:ras ("pitch mod11lati011s") arc three: 'Udatta, a:nudlitta, and svo,r'ita. They arc articnla.tcd through 8trctching, relaxiug, and carrying acros8
(:~.l). They am ha8ed on the ak,'>ara (:3.2)."
ln this case, it sccrn8 rca0cmahlc to translate svara8 a8 "pitch modnlatious" siucc s'1Ura :L2 8pccific8 that they arc to be nndcrntoocl a8 charactmi8tic8 of the o,/;;,'}(Ira which, in turn, is elsewhere defined as follows:
9

trmrnlations arc mine.

Ac is a pralyalw,ra ("abbreviation") usc~cl in the p1i.u:ir1:ty11 tradition to indicate the
vowels.
:i All
10

(iIOVANNI C!O'l"l'I

16

a,.c;ta:n su:mii,n(iJ;:.'ja'f'(ie,y rid'i/;0)1, \\ l.1 \\

t;i,i.a:; coJvari

sa'q1,rlhya/,:.~arii'(l'/J

de svlirrJ,~1, \\ l. '.j

\\

l.]

utf;o:ni,'(l:i \ \ l. 2 \\

-
an'u,.s'uii:ro 'V',IJ o:i~j arW'ffl d/,,,;.~
arang
mn

II i..'>'>
- - II

r
r
-
- 1
imI>k 0 };-so:ras
"The first eig;ht lsmmcls Ot t 1ie vo:rrw,.c;am,o:mno,ya am s
. , .: , . ]
.
.
1
.
.
!
.
.
.
,
(
1
'>)
Th<'S('
raksaJ as
( 1.1). Then lthem arcl lour more corn mm tor y a ...~am':'> . .~ . . ,. l . .11. ,
arc the sva.'ras ("vowels") (L'.3).'11 l... J Doth an an:ll.'i'tHira awl a consornwL < c
lto he consickrcd as] an appendix to ak.'jo:m (L22)."

l as an aJ;;.'jara not,~ Ill t,1ic seus<'

clcp1ctcc
In these sutras, the S'U U is
of a
1
syllable, i.e. a sequence of sounds in a word, hut in the sm1sc of a ;uwc '

lcut art1cu
latory somH l , i.e.
cit
l wr w1'tl, l or without
depicted
as an mdepern
'
consonants attached to it.

5.3 NiiradvyasiktJa, Vajasaneyipratisiikhya, and

Yajnavalkyasika
Other treatises (lefinc the 8U U in connection with the combination of suaro.s
("vowels") and vya'il,.Jano.s ("consonants"), without an explicit use of t,hc
term (Lk.:mra. For instance, the Niirarfi:ya.ik.?ii states:

suara ucrn'i(i svaro 'Tl'tca~i .'iVlira~i svarita C'IHL ca \

'lryaizjrinany anuvartanlc yalro. l'i.?thal'i :;o.~i suaro.(1, \\ 2.5.2 \\
.~uarapnulllii.'n.a.'rrl.

tra'isva:ryam <Ldiryli.(1.

vrn,/:fjrina(J;

Tnff~i'ivad uya:i~jo:n.arri virly(l,/. S'li.lra:uru: ca, .c;vo:rn:rr1, 1rirlu.(i \ \ 2.G.:~ \ \

''A .'ivo:rn, ("pitch modulation") can be 11.cco:i(1, (''high"), n'i.d1. ("low")',or

.'ivar'ita. vVhcre there is that pitch, the co11sorn111ts imitates [it] (2.G.2). 'l he
teachers assert that the~ l:misvaryo, ("triplet of pitch-modnlat.(~d vowds") has
the .'ivara ("vowel") as [its] basis. They hdh~vc that the consonant is like a
bead, while the sva:ra ("vowel") is the st.ring; (2.s.:n."
Similarly, the Vi~jo..~aney'iprcLt'i.~rJ,ld1:ya dcscribc'.s th<~ synergy of vowels
consonants in the artic:nlation of tlw pitch modnlatiou:

HJl(l

vaT(l.asamam:rw.ya ("colkction of souwls") which is 110(. prcs<~nt,cd in the fi.qucdaprali.fokh-yn contains the list of sounds whose articulation awl prop<~rti<~s arc described
in the text. The first twelve am thn vowels, namdy: u., 11., i, 1., u, u, [', and r followed by
11
c, ai, o, and o.n (the sonnd ! is probably rcganbl as an allophone of r). Shastri [ rn:\7: 1
translates sarrwnak.~ara as "monophthong" awl saq1.dhyo./,::;;11.m as "diphthong".
11 The

-======--.

17
vyo:i~ja,narJI, sva,n;~ui sasvaram

11

1. 107

11

"The consonant shares the svo:ra ("pitch rnod11latio11 ") with the svara
("vowel")."
Likcwis<~,

the

Y(~j'i/,a,val/;;ya,.5ik{icl,

is also explicit in Htating that:

svo:m, 'llcca,{I, sva:ro nfca{I, svarn)1, svarUa cva ca, I

sva:ro1)r<ulha:narJ1, traisvaryya:rn, vyarzja:nar1i tcna so,,svara:m

II

118 JI

"A svaru ("pitch rnodnlation") can be high, low, and 8Va't"ita. The traisvcLTya ("triplet o[ pitch-modulated vowels") has the svarn, ("vowel") [as its]
bmiis. The consonant shares the svara ("pitch modulation") with it."
Although, iu other passages, both the Vii,jasancyipr'i"itis{i)diya and the
employ the term (],k;'mra in connection with the term svara
("pitch modulation"), the rcnmrk on the capacity of consonants to he part
of the S'BU is made in connection with the vowels (svarc~ia) and not the
(],k{wrns. 112 Therefore, it is possible to ntate that these three treatises provide
an interpretation of the nature of the S'BU as consisting of 'vowel plus
consonant( s) ', i.e. as a sequence of sonnds within the word.
Yr~j'navalkyasif;;~<;ii

5.4 Saunahi:ya Caturadhyayiha

As far aH the definition of the S'BU is concerned, the Saunoki:yii Cat"nriidhycT'.IJik(L docs not provide a final answer but presents two different views on the
topic: on the one hancl, a view corresponding to an ak:'>ara-only solution
and, on the other hand, a view according to which ak.~arns a.rnl consonants
hear the s'Uo/ras ("pitch modnla.tions") together, i.e. a combination of the
two 8olutio11s shown in G.2 and 5.~~- First of all, in fact, this treatise
states:

sa-m,(/,nayarne 'kijaro"m 'u,cca'ir '1ulrJ:f,tarn, nica'ir arrndaltam (Lk:,'>'iptarri

11 1.1. Hi 11 [ ... ] suaro 'fq;aram /I l.4.2 11

8'l)(i7'tarn
12
' ' Thc

Vajnso,ncyipn/,l,i,foJ,;/i:ya iu fact states that: soaro '/;;:;;u,1nm 11 l.!H) 11 snlui<iya:i:r

vyaiijmw,i(1, II 1. lOO // uttarai.~ c1ivasit11J(l, II !.101 IJ "Au a/,;911<1a is a svara ("vowel") (L!l!l),

[aud it is so] also with tlw prcccdiug cousonaHts ( l. lOO) and the following ones in pansc
(1.101)." Tlw Y11Jiiaval/,::i;a.~'il~:~<i also employs tlw term akijarn more than once, at least
in one case in (lirect c0111wctio11 with the term udaUa: 11.dlit.l!l./;;:rnrayor madhyc bhavcn
nu:as lv o:uagraha(l. 11 84. h JI "[As iu the previous examples,] if [it occms] h(~twccn two
high-pitched ak:mras, tlw o:uo"1rmha (?) should be low-pitched" (here the sense of tho term
11.vagm.ha is not at all clear).

( j !OVANNI

( 1

;JO

'\"l'l

.
.
I
. l -u t 1ow-p1't,clll'd
, j:-;
"In the same vocal rn1w;('., a l11g;lt-p1tdt<'.d n, ,;.~an.1, ls n< .a .a, ( . . ( 1 1 lu)
..
l
l'
I l 1 l ") is -;Vffl"l
o:n:udlT,fJa, arnl an <if;;.~'ipt(J, ( "cast c own rom 11g 1 ,o ow
,,f n.
l... ] A vowel is nu aJc.~a:rn. ( 1.4.2)"
'
y

nrercnt
opill1
Thercaftcr, the Sanna},;1,yn. Ca/:u:rn.dhyo.ytl,;a pr<~scnts two ( : t l ws not
ions on the role consonants may lrn.v<~ in h('.aring th~ sun.ras, lmt 1 < ( "
state which one shon1d h<~ chosen:
a..'ruanT~i.'i v11o:ii:jnniin:i \\ '.L'.). 2G \\
svaro:1Jnntfl;y <l.nyal.are:1Ja~1. \ \ '.~.'.L 27 \ \
..

"

('

1 ')(')

"Consonants have no svrinis ("pitch modulations ) .~ .. ).~

says that [consonants1 h(~ar sva.ras. ('.).'.).27)"i:i

>

A. uyata.rcy<1

. l l'
hi(rtll't.,1c.'S
Moreover, an interesting; discussion although_ not devoi< o an~ ~ -. -ilJi
is reported in the following; .<;filras, when'. the 80:11:no.l1:Fy<'i CoJnradhy<iY
agrees with the grammarian Siii1khamiti:

ldm ak.'}u:r<L.5'.IJO. svaryrmuir_w,.<;ya suaruo.te \\ :3.:L:~1 \ \

ardhmri h:m..':1'UlL8'!Ja pado dfrghn..'iyr:ty eke \\ :).:5,:32 \\
sa'f"Vam, il'i .~iiri);:ha:m:itri. \\ :3.:3.:3:3 \\
ak.'}o:nL.'>'!Jai.~ii vi<llui no, 1J'idyal.c yad dvisuo:rffJ!uiva.(1. \ \ '.).:L'.M \\
"Of an

ak.~<mi havini!; a svarya (" suo.ril.o,"), which part hears it?

3 yi1
('.3.:3.: l.,.

'-'
l
. . . r1 (' ]
" .
l lf' t1 <1na.rtcr
,J<~me say t iat it. i~ l.t, H~ ~'rs~, half .ol. a short vowd, arn.l t .w l ll'.~ ' ,. rs it.]
of a long vowel (.3 .. L32). Sankbaimtn sa.ys tlrn.t tlu~ cntn<~ [aJ;;~u.ra he~~ . h
(:3.'.3.'.3'.3), lsincc] it is not. the paU.('.l'il of au nk.711:nr, t.o hear two suaras ( plt.c
modulations") ('.3.'.3.'.M)."
k ' )
.. .
1 . ") bnt,,
I n :j .) .. J, svar"ya has h<~(m translated as svu:m. ("p1t.d1 modn at1on
according; to D<'.shpandc l Hm7: 1Jtl8-GO], it IIH'.ans siJo:t"if.a.. '1s This iutcrprcL<t. i i} i i 1
['
11 knownl
t1011 unp ws 1, rn.1, 1; us group o .<fu,/:rn.s ls to he'. com1<~dc1l t.o the we - .
debate concerning; which part ol the 8BU is to lH'. considc~rn<l high-pit.die{
ancl which part low-pitdwd in casn Llw .r.;uo:m is a sva.ri/.a.. l lowcvcr' one
should not overlook the fad. that hem th(~ t.<~rms hro..r.;va. a11< l drrqha. a.re

1:\For the intmpretation of th,~:-;n sulm:->, d. Dn:-;hpawlc \I !J\17: 1\!17-~].

,,
'1'1 IIere, .1t 18
. wor,, l 1 notuw;

l
l
,..
I
c

t rs
t mt t 1c verb snnT".1)11./.e
l1<~rc trn.11:-;latcd as it. > " '
s IHh'VS \loW,
1. rns<>1lll< ls " or cvc11, m t l tis c011t<~xt., "it is acc,~nt.<~d". T l 118
. 10
l1.t erall y means "'t
unlike Western Ling11istics, each Sanskrit notion arnl tnrm j 11 th<~ fidd o[ .4i/,;.~a tcuds
be gronn<kd 011 articulation.
. .
1
1 "lforc, if one took svar110. as a synonym of svani ("pitch ac<:cnt"), it wonld hc dilltctt \,
to find a plansihk interpretation for tlwsc s1itras.

19
mentioned i11 comwction with aJ1:,'>a:ra, i.e. Lhc SD U: these terms arc usually
employed to indicate the length of a vowel, 80 that alqmra here woulcl simply
nwan vowel. Fnrthennorc, if one agrees with Anyatnreya (sec sfi,lra :J.:J.27),
the definition of the 8130 advanced by the Sa:u:rwJ,;fyii, Calu:nl,dhyiiy'ih:ii seems
to he closer to the one 'vowd+consona11t(s)' fonncl in the treatises discussed
above (s<:c 5.:J).

5.5

LomasisUf,~a

The last treatise that will be con8idered is the Lornafi{'ik.'/i which prescnt8 a
couple of verses where the term al1:~'iara is not mentioned lmt -- umisually for
the patterns of tlw ,4'ik{>avcdririga speculation a pseudo-syllabic definition
of the SDU seems to he upheld:

naUat sva:rat'i pfirvliri,ge na panirigc kadlicana I

na vyaiijanc na rnatriiycirri katha:rp. yogo v'idh:ryate 11 :~. l 11
svarasyaiva lu pfiTvlirddhc vyarUananldharddhapasc'imc I
tayor arddluirddlw,sarriyogc svarn:rr1, kuryrid v'icak.~a:e.a(1, 11 :3.2

11

"It [i.e. the svara ("pitch modulation")] doc8 not resound in the preceding part (pfirvriri,gc) [of the vowel], nor in the following part (panirigc) [of
the vowel (?)],'Hi nor in a [Hingle] conHouant, nor in a [single] morn: how is
itH production (yoga) [then] prescribed? (:J.1)
T'hc learned should pronounce the 8'1HLrn ("pitch modulation") at the
connection of the two halvc8 of these two [sounds], [that is] the firnt half of
the 8Vara ("vowel"), [and] the last half (arddha-JH1,idm.a) of the [preceding]
half consonant (vya'il:jo:na-arddlw,). (:J.2)"
The iuterprctatiou of the expression vya:i~ja:nrirddh(/,rddlw.pa..<fcirnc iH
rnLlicr dubious: the proposed translation docs not seem to match with the
next half-verse, where we find the cxprcsHion tayor arddluirddha,,'io:rr1.yo_qe
("the co1mcctio11 of the two halves of these two [8otmds] "). Thi8, iu fact,
implies that the sva:m is found at the juncture hctwecm the two halves of
the two sounds, rather than between the quarter of a consonant and the
half of a vowcl. 17 H.cgardkss of the correct interpretation of this passage,
------------~

HiThc) tc)rlll ari,ga, is rntlwr obscure: the tentative translation proposc<l licre ("part. [of
tlw vowel]") echoes tile debate OH what suh-mtit of the SDU beam the udo.tta iu case of
a svarifa (cf. G.4).
17
Fnrthcrmorc, it would he arbitrary to interpret a:nllw, as "half [morn]" instead of
"lm]J [somtd]" since verse :l. l clearly states that, a sva:m ("pitch modulation") docs not
rest 011 a 1110ra.

(; JOVANNl

( 1

;\(

)'!"!'!

20

.
.
.
r . .. r I , C'f >lJ is the term
what matters hen! m colllH'.d1011 with the d<'. nnt1oll 01 t 1c J > ' . . ,.
l i mo< l n1at10n,

['ad, emerges as som<),t l1i11 ,_,,,. 1wrtauung,

sam,i;oqo,: the p1tc
m
'
l
1
t l . . ouc son UC .
not only to a smg;lc sonrnl, hut to a combrnat1011 o more) , lc1ll
, . . .

L
,_,., 1 1
1
. 11 ])J.(' defuuLIOll
Therefore, the omaststr.:sa seems to >C lll avour o a sy d.
.
',
..
1 Hr t . I irwnist1c
of the SBU, although not m the sense proper to t. ic vv<!s .ct Il _, h .
.
1
l
r
1 t
)'1l'l S o( (,WO
speculation of a sequc!lcc o so111H s, hut 01 a S<'.cpwncc o .wo I ' "
1

<

contig;nous soumls.
.
r

., I 11'l>nrY 0
According to two mamiscripts prescrvc<l at tlw Umvcrst ,y
<
Carnhridgc/1K svaro:m in vers<) :t2 is to be rcarl as sur/,ra:rn,, usually a syuour:~
for svarif,o.. If this variant WCI'<) the corr<!d on<), it woulcl menu that t_ns
coup l c o f, verses <lo not. re f"er to a 11 tlw t l iree svo:ras l mt. on l Y t,o t\1('
<;'/J(l'{"lfa
' t
In this case, the Lorna.ci.Z.<f'ik.7li would h<) pn)srmting a view on what pa{',
of a word hears the high-pitched part of tlH'. sua:rita that is close
t
,Sa'll:nakiyci Catv:radh:uci?}'ilJi, (sec 5.4). 111 any case, this wonl<l 11ot iuvahdat~c.

o[ consonant. is
use
'(l in order
my m t.erpre t.ation o f. t;l ie verses smce
t l ie notion
to define the 8DU.

v: l(:

6 Open conclusions
rI'l ic, r.icl i vallc
.. t.y o t v1cwpom
.
. t.s on t l tc cl<) f'uut1on
. . o.1 t l tc ,;'-'13U exp rcssed
,, ' hY
.
the different treatises h<~ars witness to 'a lively spcc11lation' on this topic
among; difforcnt grammarians. To wit, the ,So:nnal.iyi/, Cal'nn[(l/1,yii;itikii, opcnl{
quotes_ t~vo_ pr,c~1><.~~ 11amc0, i.<). Anyatarcya aIHl S~i.i1kharnitri, w~1crcas .h.ot,;,,
the Pa~im:iya,<.;tk.7u. awl the N(inulTyo,.<iik..7ri r<~spcctivdy speak ol ua'f"~w.vtd~z.,
("thos<) who arc lcarnc<l in tlw smm<lH") awl riui'f"fJrL~/. ("teachers") nsing t. ic
plurnl. 1.J
1111fort 11. -I
l 0uc l1 a varid.y of views rnma1n
II- owcvcr, t;lw reasons 1><~1imr
natcly unknown, since th~ trcatisr~s simply state what part of Lite wot<.
11 1
hc)ars tlw svaras without providing any kind of <~xplanat.ion. lu part.ic l<i ',
those treatises according; to which the 8BU consist. 8 in a s<~qucncc of sonn<h'
are the most intri1.1;11ing from a theoretical poiut of vi<~w: they iu fact ~)n~-.
1 1

snppose some sort of rnu1ti-liw~ar sp<~c11lation. Two pla11siblc motivatIO!lS

for the formulation of this la0t approach arr~ lier<! listed:
(a) according to a strict phondical1y-oricnt.c<1 r~xplanation, it is possibl'.~
~;hat th:! g;rarnrr.mrians understoo(l that th<) articulation of tlw s~o,ffrr;;'
m particular oi the 1ula:ua, W(!re mon~ audible wlicn accompaiuc<l y
':~~~d<l. (l!)2!i) :~1Hl ~rl<l. (l!J:~~):

,, . .

I he Lonwszst.l;;~a rn an cxccpt.1011 smcc virn./;:.~a'.1.0.( 1, ("tlw ]c~arnc<l )

is

, iu rn]ar.

Lil t.hc s

21
voiced cons011ants (see :3), and therefore extended this observation
to c011sonants in general;
(h) to define a sound as a/;;{'a:m implies thaL a fundamental characteristic
of the sound has been singled out, i.e. its irnlcpcndem:c. Therefore, it
is possible that some grammarians thought that it was plmrnihlc to
attribute to a somHl seen in snch a pen.;pcctivc other characteristicr:-;,
8llch as the [act of bearing the s'1mra8, whereas other grammarians
simply attributed it to the vowels.
However, the main rea80ll8 preventing us to reach any definitive, or even
plausible, conclusion concerning tlw motivations behind this speculation arc
that, on the one hand, the fik{>lLvcdii:ri,ga genre just as the vcdlLri,ga literature
in general possesses an assertive character, and very little space is given to
explanations and, on the other hand, this tradition lacks a well-developed
cornmentarial liternture. 50

Bibliography
Allen, William Sidney [195:~] Phonetics in Ancient India, London-New York:
Oxford University Press.
Deckman, Mary E. [1992] "Evidence for Speech Rhythm across Languages" in
Y. Tohkurn, E. Vatikiotis-Bateson & Y. Sagisaka (eds.) Speech Perception,
Prodv,clion and Lingu'isf'ic Structure, pp. 457-4G:3, Amsterdam: IOS Press.
Beckman, Mary E. [198Gj Stress and Non-8/;rcss Accent, Dordrccht: Foris.
Bhisc, Usha TL (ed.) [1D8G] Nli:nuffyd, .<ik:jl/,: With Uw Cormnenta:ry of lJhaff,a
,~obfu/,/,:ara, Poona: 13lmndarkar Oriental .Rc:-;carch Institute.
13lcvins, .Juliette [2009] "Phonetic Explanations for Recurrent Sound Patterns:
Diachronic or Synchronic'!" in Raimy, Eric and Charles E. Cairns (eds.)
[200!h], pp. :32G-:Hi.
Cairns, Charles E. [2009] "Phonological Representation and the Vaux-Wolfe
Proposal" in H.a.imy, Eric and Charles E. Cairns (eds.) [200!h], pp. 14G-G4.
Cardona, George [1D7G] P<L~1,in:i: A 811/ruey of Research, The Hague-Paris: J\ilouLon.
-------------------------

----

Go An exception is Uva\.a, who composed a commentary for both the J.lgvcdopniti.fokhya

am l the Vr~jasaneyipniti.fokhya. Uufortmmtcly, his work is not useful to clncidatc the

definition of the S'BU.

CIOVANNI CIO'l"l'I

22

Cardona, George [l988] J>1i~1,in-i: !Tis Wm-!.: and its 'lhulilion, Delhi
Banarsidass.

Motilal

Chomsky, Noam and Morris Halle [l!Hi8] '!'he Sound Paller"n of l~nglish, New
York: Harper & Row.
Clackson, .James [2007] Indo-E'ILropcan Dinguis/iics: An fnl:rodud:ion, Cambridge:
Cambridge University Press.
Clements, G.N. [2009] "Docs Sonority have a Phonetic Basis?" in Hairny, Eric
and Charles E. Cairns (eds.) [2009a], pp. 1GG-7G.
De Lacy, Paul (ed.) [2007] '!'he Ca:mbr"idge JJandbook of Phonoloyy, Cambridge:
Cambridge University Press.
Deshpande, Madhav M. (ed.) [1997] S1annoJi:Tya Caturiidhyayikii: A Pniti.<i1ikhya
of the .5o:unaki:ya Atharvo:veda, w'ith commcnta:f"ies CatnnidhyayTbha9ya,
Bhlirga:ua-IJhaskara- Vrtti and Par1casandhi, London: Harvard University
Press.

Deshpande, Madhav M. [200G] "The role of linguistics in Indin society and

education" in Sylvain Auroux, E.F.K. Koerner, Hans-.Josef Nicdcrehc, and
Kees Versteegh (ed. by) History of the Langwrge Science: an International
!Iandbook from the IJcginnings to the Present, Vol. 1, pp. 17:~-177, UcrlinNcw York: Walter de Grnyter.
Devasthali, Govind Vinayak [19G7] Ph:itsiitras of /lcl:ntanava, Poona: University
of Poona.
Garde, Paul [19fi8] /,'Accent Paris: Prcss<~s univcrsitaircs de France.
Ghosh, Manrnolmn (ed.) [rn:~8] Pli:u:in:i.;11a..<iik.~1i, Calcutta: Univcrsity ol' Calcutta.
(Reprint: [JD8G])
Gordon, Matthew Kelly [200fi] Syllable wciyhl: phoncl:ics, phonology, lypoloyy,
New York-London: Routledge.

Hall, T.A. [2007] "Segmental foatmcs" in De Lacy, Paul (ed.) 'l'hc Cmnhridge
Handbook of J>/wnoloyy, pp. :n 1-:M, Carnbrid;r~: Cambridg(~ University
Press.
Halle, Morris [1997] "On Stress and Accent in Indo-Enropcan" in f,anguagc 7;~,
pp. 27G-:~ 1:~.
Halle, Morris and .Jcan-H.ogcr Vcrgnand [1D87a] An !~\say on Stress, Cambridge:
MIT Press.

Tim

SvAHA HEAH.JNC UNIT

23

Halle, Morris and Jean-Roger Vcrgnaud [1987b] "Stress and the Cycle" in Linf]'U'lsUc Inqury 18, pp. il!)-84.
Harris, .John [2007] "H,cprcsentation" in De Lacy, Paul (eel.) The Ca:mbridye
Ilu:rulbook of J>lwnology, pp. 119-:n, Cambridge: Cambridge University
Press.
Hayes, Bruce [199G] lvfcf;r'ical Stress Theory, Chicago-London: T'lrn University of
Chicago Press.
Hyman, Larry M. [20m)] "How (not) to do phonological typology: the case of
pitch-accent" in La:nguagc Sciences :31: 21:~-:~8 .
.Jakobson, Roman [19:rt] "Die Dctormng und ihrc H.olle in dcr Wort- uncl Syntagrnaphonologic" in 1'rnvaux du Cercle Linguistiquc de Prague 4, pp. 1G482, Prague: Jcdnota ceskoslovenskych matcmatiktt a fysiktt.
Kanta, Surya (ed.) [1970] J_ll.;tantrarn: A I'nitis{iJchya of the Srima:ueda, Delhi:
Mcliarchand L<1chhmandas.
Kiparsky, Paul [2007] "Piir.1ini is slick, but he isn't mean" in Nagoya Studies in
Indian Culture and Buddh:isrn: Sarn,blui~~a 26, pp. 1-28.
Kiparsky, Paul [2009] "On the Architectnre of Piir.1ini's Grammar" in Huct,
Gerard, Amba Kulkarni, and Peter Scharf (eds.) Sanskrit Computational
Ling'uistics: First and Second International Syrnposfo, pp. :~:~-94, DerlinNew York: Springer.
ldsarcli, vVilliam .J. [2009] "Calculating Metrical Structure" in H.aimy, Eric and
Charles E. Cairns (eds.) [2009a], pp. 191-211.
Laver, .John
l)rcss.

[19~)!1]

Principles of Phonetics, Cambridge: Cambridge University

MacDoncll, Arthur Anthony [J9Hi] A Vcd'ic Gra:m:mar for Shulen/;s, Oxforcl:

Oxford University Press.
lVIishrn, Vidlmta [1972] ii CriUcal Study of Sanskrit Phonc/;'ics, Varanasi: The
Chowkharnba Sanskrit Series OHice.
Mohanan, K.P. [U)9G] "The Orgard/\ation of the Gramnmr" in Goldsmith, .John
(ed.) The Ilandbook of Phonologcal Theory, pp. 24-69, Blackwell: Cambridge (Mass. )-Oxford.
Ncspor, Mmina and Irene Vogel [198G] Prosodic Phonology, Dorclrccht: Foris.
Parnmcswarn Aithal, K. [1991] Vcda-T.dLk!;ia:(uI Vedic ilndllary LUcral:nre: A Descripliive Bibliography, Stuttgart: Frnrrn Steiner Verlag.

' 'J"l'I
(: IOVANNI ( )IO

24
.
.
.
'. .. . .
. I I
l r:olled'ion of
Pii\haka, Yn;alak1~orn Vyasa. (ed.) [ 188\l-\l.~] Stlc:rns<Lrfl.IJ r r1. 1.a. J, ~
,Siksh<l..'> lry Yrifiiavall.;yrt rJ.nd oUwrs, Varanasi: B.B. Das & Co.

" in
>l
l . l OtH'ral.JOllS
Purnell, Thomas [200!)1 "Phonetic Inilncncc ou l 10no 0;1ca

1
Rairny, Eric and Charles K Cairns (eds.) [200!la], pp. :~:~7-G L

v,. 111"

on Ar-

Haimy, Eric and Charles E. Cairns (eds.) [200!la] Conkrnpmary /,(,

'

llrl' 1>1css
ch:itcdnre and Uepn:sental'ions in Plwnology, Camhri1lgc: N

t 1tion in
Rairny, Eric and Charles E. Cairns [200\H>] "Architecture arnl H.cprcscn '' ' _1().
1
Phonology" in H.aimy, Eric an<l Charles K Cairns (eds.) [2009a], PP
,

,-nu; am.:

H.astog1, Indn (c<l.) [19(i7] Kii.t1;11.yaniicil,rya-pranTtarn 8u.klo:1;a.Ju~i-pra1 tsa ~ " . 1

- Tr:_
.
;
,
.
.
' .
l. - c
+rt't Ap1us<.
at l w:ua, v a10.sancyi-priitisii.khymn, Varanasi: Caukharn m ,Ja1ps '- "

.,

, .

Scharfc, Helmut [1977] Grnmrrwtfral Litcml:u.rc Vol. V, b1sc. 2, lil ' 1

(ed.) History of India.n Diterahtre, Wiesbaden: Otto Harrassowitll.

Gonda

Byvcda-pni.ll.~il./;;hya

Shastri, Mangal Deva [lD:n] The

w'ith the cornmentar!J of
Uvaf;a, Pnnjab Oriental Series 211, Lahore: Moti Lal Banarsi Das.

T rn \ >e t 11lmy, N"l

l '
"
1 co ai S1~r;cyevich [ H):}9} "Grundzuge der Phonologic lil
da Cercle L'ingn'ist.irrne de Prague 7, Prn;lw: [s.nl

Trff/llL'IL ;
'

Vanna, Siddhcshvar [1D29J Crit'ico.l 8huHes in the Phone/Jc 0/Jscruat:imi of Irid'ia.n Gro:rnmariu:ns, London: Royal Asiatic Society.
Vanna,

v:r, mdra Kumar. (C<l.) [1!)70] ~lyvula.-prr1.f.i.foli:hya: Uua.f.a.-/Jhas~1!~(1,-.~(Lfll~

1

tam, 1lw Bauaras Ilm<lu University Sanskrit Sc~ri<~s !) Varanasi: Kasi Hut(
Visvavidyiilaya Sodha-Prnkasana.

"Th<~

Vaux, Bert and Andrew Wolfe [200D]

Appendix" in Hairny, Eric and
Charles E. Cairns (c<ls.) [2009a], pp. Hll-il:L

Two
Puranic
. transformations in Cola
Cidambaram: The
Cidambaramahatmya and the

Sutasamhita
.
Whitney Cox
The goal of this essay is quite modest: I aim here to c8tablish the relationship
of priority and irnlebteclness between three parallel passages as they occur
in two texts of South Indian pura1.tic Sanskrit. However, in working out the
details of thi8 relationship, I realized that it rnii-;ed quc8ti(m8 beyond the
8trictly technical, aud that it8 interest extended beyond these two particular
texts. I hope to demonstrate that a clearer urnlcrntanding of the textual
history of th<~se two works allow8 us to obliquely capture something of the
wider histories of the institutions and societies from which they emerged and
in which they were dis:-:;cminatecl. This suggests that philology and social
history can not only prove complementary in their methods hut that each
can provide Uie other with new research questions and new angles of vision
on existing problems.
Of the two works, by far the more extensive is the 8Fi,tasar.nh'itii, (SuSa1~1),
which describes itself a8 the second of the six sarrih-itas of the 8ko:ndapurii'(l(L, in 8ix thousand granthas or thirty-two syllable v;~rse-units (1:1.19cd-20).
The Sl:iSai~1 is snhclividccl into four kho:u4as: the S'ivo:ml'ihrilmya (700 granthas), the Jillinayogo, (7:37 gro:nlhas), the M'ukt-i (637 granthas) and the
0

\NIIITNEY COX

2G
massive Yodrio:tHL'ibhava'' (1lOOO yra:~itlws; tlw to.Lal yranlha-count is thus ac~
tnally G07 tl). The published text 1s accompanwd by th<~ comrnc11tary (the
-UiJJ;ar:i;adivik<L) by Ivbidhavanm11tri11, a schoh~r-oflicial of th<~ Vija~aw~g;1ra
empire. The lower limit of the composition ol the Si!Saip is thus i11rr11shcd
by thi!'> commentator, whose .florni/, H,ACIIAYAN dated to J;JG8 1:38 1.l CE.
RAGIIAVAN also proposed that the text's npp<~r limit conld he inlcrred
from the absence of any reference to the main shrine at Cidambararn as tlw
kanalwsabh<L ('the golden assembly'). This poiut<~d, he claimed, to composition prior to the rule of cith<~r Aditya or Parfintaka Cola in the first half of
the tenth ccutury. 1 It was my dissatisfaction with this argument 011 :-;lender
negative evidence that partly prompted this investigation.
As a whole, the SfiSaq1 is given over to a didactic purpose: arguing for the
soteriological primacy of a Saiviscd version of Advaita Vedanta. Compo:-;ed
in a pedestrian ancl dull style, much of its inten~st is source-critical: it serves
as a great digest of Upanif?<-Hlic and other sources, giving a common-reader's
eye view of the rrwdieval reception of Aclvaita (among other things). The
main didactic thrust of the work is enlivened by a nmnber of m{i/ul,lrnyas
and mythical narrative asides, most of which arc centcrc<l on Saiva sites
around the Kaveri river delta, the heartland of Cobrnanclalam in what is
now central Tarnilnadn. As this serves to show, th-c SfiS<~rp is cc!rtainly a
work of the far South.
Also assigned in its chapter c:olophons to the Bko:rulapunJ,'(W, the Cidarnbaram<Lhatrnya (CidMa), as its name indicates, is itself a product of the
Tamil country, b<~ing a collc~ction of narratives centered on the great Saiva
shrine of Cidambaram. 'The CidMa consists of twrn1ty-scvcn adhy<L:1;as, each
averaging between fifty and eighty vcrs<!8. It was to provide the basis for
a Tamil n!working, the Koy'ltvnrii'(wm attributed to tlw Saiddhfintika theologian and pod Urnapaticivar1, activc~ in tlw early fomtccnLh century (his
Ca:rikarpan'irdka:rar_1,am was completed i11 Saka 12:J5 [= 1:0:3 CI;]). \rVhilc
there arc no strong grounds to uncritically adopt this attrilmtion, the Tamil
text did receive a commentary that, judging from its language, has been
said to date to the sixt<~m1th c<~ntmy or earlicr. 2 The CidMii is much shorter
and more tightly constructed than the Sf!Sair1, and is sorndirnes very soph isticatcd within Lhc co11fincs of its g<~mc cornpkx awl h<~anLil'nl forms
arc frequently rnd with, awl it contaius a relatively high 11t11nbm of v'i7mJ<1;
'l

qnarLcr verses.'

--11~-Mill1\VJ\N l~H7: 2:;7 2:l~.

71.
r:ltations of the CidM<t refer to tlw edition or Sornmmudarn I) II<SI I !Ti\ It [=Eel.]
.. , t vrJ 111 ,111nscripts A awl B (see tlw Bibliography for details of these). D11<Slll11~: WC 11 dS ,\
''
. .
.
.
.
. ,,
., ,, ,r'Jll" to he a transcnpt1011 of a smglr~ good marn1scnpt and wHs pcrh<1-JJ,>
T;\H'~; <:< Iu,1011 ,, 1 '
2;\1WN1\<:1\LJ\M 200G: 70

:i All

l)lJHAf;JIC TIU\NSFOH.MATIONS

27

The CiclMa is hc:c;t lrnown from K UL1rn's dissertation, which was my

point of entry into it. 11 In brief, KULi<!<: argued [or a stratigraphy of the
CidNlu. based on its narrative and descriptive content; most intcrc:c;tingly,
hut also most controversially, he argued for the decisive influence upon the
work's final chronological 'layer' o[ a key event in the political history of
the Cola dynasty, the contested accession of the emperor Kulottui1ga l (r.
1070 1120). Kulke argued that the rnizhil,trnya's final narrative cycle relating
to the northern (or eastern) king Siq1lmvannan/I-Iirm,iya.varrnan (his name
changes in the courne o[ the narrative) is a mythic transposition o[ the
events surrounding Kulottui1ga's rise to power, a.ncl that the same king or
his irnmcxliatc successors were responsible for the creation of the composite
maluTtmya text as a means of legitimating what Kulkc styles his "coup
cl'6tat". Much more speculatively, Kulke goes on to argue that the final
version of the text could not have post-dated the career of Kulottui1ga I's
grandson Kulottui1ga II. 5
There arc both empirical and theoretical questions that this argument
leaves unanswered, which I cannot explore here. With some qualifications,
however, I accept KULKE's central contention of the (at least partial) identification of the text's Hirai.iyavarman with the historical Cola king. vVhat
is centrally important for my argument here is that both the SuSarp and the
CidMa arc concerned with CidambararnY This importance is patent in the
case of the rruThalrnya text, while in the SuSaq1 (at least in its later sections,
for which sec below) the site is given pride of place in lists of pan-Indian
ancl Colamai,ic,la.lam Saiva holy places.
The three parallel passages which will be examined here one from the
CiclMa and one each from the SilSai~r's first and fourth /duiti(ias--can be
fo1111d in the two Appendices that accompany this essay. The brid narrative
of all three passages is essentially identical: it is a straightforward and rather
typical bhaklii parable of a sinner redeemed. 111 all three we begin with
only lmrnght out. for private circulation: other tha.11 tltc copy in the possession of David
SM ITI I of Lancaster Univernity (of which I derived a sccoud-gcncratiou xerox from David
SillJLMAN of Hebrew University, .forusalcm: lllY thanks to both of them), I know of only
a copy iu tlw Library of C011gwss in vVashington DC. I would also like to record lllY
grntitud<~ t.o Ilcrlllallll Ku1,1rn for allowing me to use his paper transcripts of the twc
MSS.
1
' KIJLirn 1970; sec also KULKI'; JD9:L
ii l<ULKE UJ70: 207 21:1 nml lmJ:l: 20G G.
GVario11sly calkd /:i/,/,avana, vyaghrn.7J11:m,, pn'(l4a:r"di:ap'll.'f'a in these texts: the term eid
o:m/mram when it occurn is 11s11ally restricted to only a part of the Sa.iva temple complcJ<
in this period, the Tamil c'ir"mm,pala.m or 'small shrine', Thronghout this essay, I use th<
name Cidarnlmrnm to refer to the temple town, now a ta.lnk lwadqnartcrs in Cuddalon
district, Tamilnadn.

WIIITNEY COX

28

111

a moral monst,m - a hrahrninicidc, cow-kill<)!' and ba11dit who, dcspit(c '1

.
.
.
I .. l , n w io
of this happcm; to r<~crnv<~ some very ?;ood advice [rom a Ha
..
. . llrld
I 111< l i:;
l
in two of the versions happens to he th<~ anti-hero's best ncn< ' ' i 0 to
almost equally dcgc11cratc, hcing, we arc told, sexually dissolute). CJ. ',.
.
.
. . t }d cit 1HJ
Ciclarnhararn arnl rnak<) up [or your crrnws, tlw protago1ust is ,o ' . 1is
. o f' {'ate or t l w wor [m1g-out
.
. t,l"tC<'S
thro11gh the vagancs
o {' t }H) }Gtntllc
' , ,, of .l 1,
.
l'
1
1
1
l
.
rl'
l
.
.
I
{'
/"f'().s(IS outsl< (,
earlier lVCS, le ta <CS t le (1( VIC(). a <:mg up t'CSH ()lJCC JVC.
. "
r l !Cl.0
of town, where he can sc<~ the temple's Little Hall at a d1sta11c<), ou
..
111
becomes an unofiicial town watchman, whil<) distributing the loot frorn 0~
. <lq)nX1a t'ions tl) 1oca l lH'a lnnans. rvont111111ng t l iis
way ('<)1 the
rest
carl icr
,
'( [' '. at. (l JC,
his life, he receives Siva's favor, arnl as a n~snlt liberation, <~I, icr ' '
moment of his death or after his elevation to uuivcrsal kingship.
,.., G lrv 1a,
It was KULKE who first pointed out the similarities bdwecll I< ,
8:1--2~3 and SU.Sam. 4:24.l 27. 7 I:fo arPllc<l
that this shows the dependence
h
..
81 ipof the miiluLtmya on the earlier text, dc)pernling in large part on the
11c
'
l

'
f'
.
1
l

'

I)

11
1
ta
to
t
pose< l c iange o , ie protagomst s name irorn tlw ,Jll.JC.11~1 s mg < :
.
.
.. " tlllg t 1JC
name of a particular lowly caste Pulkasa, Urns allegedly clomcstH,L1' , _ ,11n
8
story to the CidMa's own concerns. Had KULKE noticed that the Sth <
l an ear.wr
1 version
of the story ( 1:4.17cd :.m ) , lll
w1uc
l1 the
' procon t amcc
11is
.
t'

.
f"
.

}
ltNCd
t agoms , s name was m act Pulkasa, he would have ccrtam Y a '. , 111 <l
conclusions. It is drnnonstrablc that it is this <:arlicr Si!Sai~1 version, 0 sc
not some other some<), that provi<lcd th<) rno<lel for the CidM~i, as a lcllL
c <',
comparison will readily bear out. In th<) first App<~udix { give the evic , ,
.
.
.
.
.
l
l sornc
for tlns rclationslup of <kpernknc<) through words givcu iu bol< , anc '
t
.UH l'ica t.,10ns o 1 tl w <l'lr<)ct1on
. o 1 .rnfltH)llC<), wh<~rn it is apparent. rl'l 11ougholl'
, t1 1
the two passag<:s tlH)rn is a clos<~ accord h<)tW<~cn th<: two, in narrative ~ ~~],c
1
tcr and often to tlw point of icl<mtical phrasing. There arc, however, n_oL c
t
[' 1 1
l

itcrcilC'
porn ,s o < i~sun~ ant~ ~tt t, JC lcvd of language, an<l t,his allows an ll . hi l
1
as to th<: <lircct1011 of 11Hl<~htcdw)SS. My und<)rstarnling of this rdat,t0Jlt:l
is as follows:
.J

"

(.1 -- (1

'

~c_1
1 l '
l (1 -
.

, }
[i,';'lj(Lte)
, ll.Jarp ' .2lc< ri:f'lo"n:1, so:ruad(/, rn,'l/,(lhnlvaJ{/, ho b:q1. /.c P i,a -, the
11
pres<mt'.s a v<:ry awkwardly co11st.rnct<)<l half-verse, <)specially S~) . ~ ;,he
case o( the noun m:1i(ll1,0L110,m, ('idiocy'), as it is strctclwd ac.1oss l ill
. ' ca<)snra. 'l'l.
. us, o f' coius<:, prcs11111c~s t,hat the rca< 11111',. !01111<
..
,-- lfi,(/,
1mes
.
.
.
'
f'
K
(

do
'(fl,//,(,
tin

l c< l i(,1011 1s 111 act correct. Ci<lMii 8.7cd htlfrn'l sa:ruo, '
'n(LT/:nUiparr1, /;:o.ro.1i co,), by contrast, n:ads m11d1 rnon: su1oothly.

7 KULKJ;

l!J70: 17:1 l7f).

.
,d l>Y a
>iThc edition actually reads m:n~llw l1111,q1., i. r:. prc~snmal>ly as a vocativc. loilowc indh<T.H
. 11c nom111at1v<!.
.
.
'I'l 11s
. 1ai1s to constnw; wl nlc
t I w sl'ig I 1t c1nc1H ht
.
pronoun mt
' :wn to ".
diopPY
1
tva'lf/, ('you an~ au idiot.') would yield s<~ns<\ it. wonld mak<~ for an cxccpLlolliil Y

J>uH.Ar;JIC 'l'HANSFORMATIONS

29

S11Sarp i:.1.2:3 includes a stylistically infelicitous sapektwsarnr/,sa. (.jar1:rnantarnsaho,,'ffc~u krtapnt1,ya'ua,cfona); while this is hy no memrn uncommon in certain registers of Sanskrit, it is notably unparalleled in
the corresponding part of the CidMa..

Iu a verse where the two texts a.re otherwise almost entirely identical (SuSaq1 1:4.3:3 ~ CidMii 8. L8), instead of the vocative chri;j(/,(l, ('oh
hral11na11s') fonnd in SuSaq1 we find a weak kvaca ('somewhere') in
CidMa.. 'l'his suggests that the latter may have been scrubbed of inappropriate speaker indicators, as at this point. in the Cicll\!lii narrative
the speakers are Vya.ghrnpiicla and his father Ma.dhyandina..
Two closely related usages in the SuSarp- -sn(J,l'l;l/, in 1:4.:J2b and sn(/,narri krtv<J, in 1:4.38a (both meaning 'bathing')--diffcr in the CidM8.,
where we find respectively ffiatv<J, ('knowing'), 8.16d and sthii:na:rri krtva ('ta.king up position'), 8.22a. The CidMii readings arc not completely senseless in either case (though jr1<Ltva, lacking an object, is
very poor), but the SuSar~1 readings arc clearly superior, and furnish
evidence of orthographic confusion in the transmission of the CidMii;
this is a good example of SANDEH.SON's dictum a.bout source texts
forming "the most valuable of tcstirnonia". 9
In all cases but one where the SuSaq1 calls the site of Pulkasa's good
works vy(J,ghrapv,rn ('Tigertown'), the CiclMa has the metrically equiva.lcnt and synonymous tillavana ('tlw forest of Tilla'; cf Tamil t'illai,
the shrub e:roccaria agallocha, whence Tillai, the oldest of the attested
names of the town now called Cidarnbaram). Changing this name is
narratively necessary in the CidMii, as Vya.ghrapiida has not yet ta.ken
up ascetic residency in the fut me town's vicinity (this is narrntecl in
the next <ulh;1;<i:ya 10 ). There is however one place where the toponym
is retained (8.20c). 11
Finally, awl most significa11tly, the compound pfi:rvokl<qrrak<LTqw,
("[he worshipped Siva] in the manner pnwiously described"), found
identically in SuSaq1 1:4.:_rn and CidIVfo. 8.22, only makes sense in the
- - - ----------

------~--

vnrsc co11tai11i11g a scric~s of three anacolntlm, and wo11ld once again awkwardly rdatc to
the ya.Li bo1111dary.
!lSANimll.SON 200 l: 2.
10
Ed. D.:m [=A p. :rn (mm1hcrc<l as D:27); I3 p. :n]: t.(],d:i;rl.rn.bhya. (1f.m.bh:i;a.] A Ed;
o.ra.IHlha D) t.dayr!dlwr(J. (day<idho.ra ] AD; do:y1is1ini, Ed.) mama nimma ]J'Wmn tv idmn I
khyato.rn. (J.sfu. S<Mna.slc~~u. blmvanc:oV a.rnlri:k:ll;pa.te 11, "Siva, Ocean of Compassion, from this
day forward, let this tow11 lw known tltronghont the world by my name.").
11
I 1i11d that .fri.rn.advya.ghmpnrarfi also occms in 7.29cl.

\NJIITNEY COX

s11Saq1, wlicrc tlw 11arratiV() of Pulkasa is prcccd<)d hy all abbreviated

..;ivap11:jiiv'irlh:i. There is 11othing in tlw immediately preceding co11tcxt
of tlw CidivHi to account for this, and it, makes little sense to suppose
that the composers of the S11Sarp i11drnkd the ritual ordinauccs siw-.
ply to fill out th<) reference. This, then, snppli<~s ddinitiV<) evidence of
the <lircction of infltwncc, while suggesting that the CidMii compilers
must have had tlw t<)Xt of the S\iSa1r1 more or l<~ss as it appears in its
edition available to thcm. 12
Thus, while the compiler-authors of the CidMu took some care in revising their source-text, enough details cscap<~d their notice to make it clcar
that the din~ction of inflw~ncc could only have been from the SliSaip to
the CidMu and not the other way around, while the closeness of these two
versions of the Pulkasa story rules out their mutual influence hy another,
third version. This sort of dose accord is lacking between either of these
two texts and the third passage found in the SflSaq1's fourth kha:(uj,a (given
in Appendix 2), the passage that KULKE had proposed to be the source for
CidMu 8::3 2:J.
All of this might seem like a lot of work in the service of pedantically
establishing that a different text-place in the SuSarp served as a model
for the Cid.Mu's Pulkasa narrative. It might also seem that this conclusion
leaves unchanged the relation between the two texts, as well as the received
ideas about their relative and absolute elating. As I will now proceed to
demonstrate, this is not the case, and our nnd<)rntauding of these texts and
of their historical context can be considerably altered in tlw light of the
relationship betW()<;n these passages.
This will however rnquirc a bri(;f cxcmsus into the early history
of Cidamhararn awl its comicction with the cornpl<;x ruling society of
Colama~H,lalam in the twelfth ceutmy. Tlw sit/; had for cc11tmi(;s held au
important place in the spatial imagination of tlw Lh<;istic religious of the
Tamil country, both Saiva and Vai0vava. Both o[ th<) ca11011s of bhoJdi hymus
dating to the latter part of the first millennium Lh(~ Tcvrira:m and the NrJ,lii,yirativiyapirapantarn contain rderenc<~s to Llw town that the hynrnists
variously called Tillai or Puliylir, as well as its residents, the famous three
thousand brnlnnans of Tillai ( l'illoJ,m:1t'tHiyiro:oar). These rd'crcrn:cs to the
place audits inhahiLauLs were, howev<~r, Uw product of iti1wra11t poets, 11011c
12

Tlie pnjavidhi, incidcutally, is itsdf qnitc itLtcrcst.iug, both in it.sown right (it, gives n
very comprcss<:d v1:rsiou of a n:cognizalJ!y Mantrnm{irga nilynpuja) and for the corrn1in11ts
it elicits frmu Miid!Javarnautrin, wl10 cites vcrsr:s from tlw lJrhat/,:<Llol/,u,m., as wdl as Llic
prulilh<Llis of SornasarnlJl111 aud .Jifanasiva (I thank Dominic ( :ooDALL who, in a pcrsoual
cornrrrnnicatiou of 21 March 2010, identified the IGilottara recension quoted !ten:).

i>IJH.Ar;JIC 'l'H.J\NSFOl{MJ\'l'IONS

o[' whom themselves were resident in Cidarnlmrarn or its smT01mdings, and

for whom it was part of a far-fiuug network of other :sacred :sites.
T'hc growth of the town into a temple city of grent importance is linked
with the fortunes o[ the Cola empcrorn, although this connection is, in certain way:s, highly anomalous. Early in the dyua:sty, Pariintaka (r. 907-955)
was allege(lly responsible for what would prove to he the paradigmatic act of
royal charity to the temple, the gilding of the roof of the ccntrnl ct'!_'!_'ampalam
shrine. Thi:s event was celebrated in the twentieth patik:arn of the T'iriw'ico"iP/Hi, a hymn supposedly composed hy Pariintaka's son Km.11;ariitittiyag,
and would go on to become a topos in later eulogistic writing. This claim,
however which, as W(~ have already seen, supplied H.AGTIJ\VJ\N with the
cornerstone for his dating of the SuSm~1 - i:s problematic, as the Saiva hyrnni:st Appar/Tinmu.vnkkarncar (the earliest of the Thcrnim poets, usually
dated to the first half of the seventh century) already celebrates the gilding of the Cidamharnrn shrine. L:l By the eleventh century-the noontide of
Cola power -- royal interest in the growing temple site and the lmbitations
that had emerged around it seems certain, although there is little trace of
this interest that can be found in Ciclambarnrn itself. 111 That the imperial
kings of the dynasty were devotees of Nat,esa is widely accepted, despite
the thin evidcntiary basis of this claim. IUijaruja I (r. 985-1014) is often
said to have taken the Ciclarnbarnrn Siva as his 'i{;_tadcva[(/, or pernonal deity
on the :strength of the title s'ivapr""idasekharn ('he who is crowned by Siva's
foot') that he adopts in his inscriptions commemorating the construction
of the Brha<ffsvara temple in Taficavur (e.g. SII II:l, ln. 55); this supposition however has little to no probative value. 15 The Taiicavlir temple itself
incorporate:s a subsidiary shrine to Natda, among many others .subordinated to the central massive lii1ga shrine awl to an iconogntphic program
<:entering on the martial Iigurc of Siva as Tripur~intaka. ](; I am unaware of
any cviclcncc to :-;nggcst that the CoJa coronation ritual or n~jyii,bhi:woka was
held at Cidarnbarnm prior to the time of Kulottui1ga II (r. l1:3:3 l150). 17 Of
11
: Scc

for instance the second Jmlikam of the fifth book of the 'l"irurnnrai, the Tamil
Saiva canon: vs. 2: ci":m.pi5nampalatt' u.l ni1_Wll ta'(l'll'Uai. .. rnara.ntu, "forgetting Stliiivu,
who starnls withi11 the hall of fine gold" and vs. 8 tiiyn ccmpot1.ni1_1.nl (~Iuti mcynta
cirrn:mz111.la /,:v.lta.rw.l. .. rrw:rn:n.lv., "forgc~tting the Dancer of tlw Littk Hall covered in
li1wly worked pnrc, bright gold." (Citations taken from SUBIU\IIMANYA J\IYAll, C111w11,LAIW AND SAHMA 2007). For Appm's date, sec ~VELEB!L lDDG: GUL
11
Tlms Otrn. 20CH: 2:n: "... of the 288 inscriptions, only G elate from thu tenth century
and 7 from tlw eleventh."
0
' As was already noted in SUlllU\llMANYAlvl 1D42: Gln.
l<iScc CllAMl'Al</\LAl<SIIMI 199G: 42D-4:\0 (critically extending SCl!WINDLim 1987).
17
Scc NIL/\l<ANTA SASTH.1 1!)5G: :M8; even this claim depends on thu interpretation
or a11 ambiguous phrase in the king's '/r/,(~ykkfr'tti.

Wr11TNEY

Cox

gn~ater evidcntiary significance is the claim ol' y1rarajcwlra

(r. lOfn 1070)

.
l
.
l
L'"
t
('"
l
to have made a co11sp1cnous < onat10n to t ie ,JI Va a, _;H am 1mr.am, <'l g('trl
,

called 'frnilokyasarn. This deed is described in tlw scV<)nty-ninth verse of

that king's Kanyakumari pillar inscription ( Ep'igraphfo Ind'ica [Er] 18: 4, ~l.
40G 41.:J, fragmentary, A) and his Chariila plates (EI 25: 25, 11. 151 153,
B), both dated to his seventh rcgnal year (1069-70 CE):
dcvasyiidrisu,f(idl1:ipasya rna/1,atas lrailokyascJ,nib/1,id/1,aq1,
.rfrrwddabhro,.<;abluiruifr1,sya makutc rna'(iikya:m ciropitmn

manyc 'Uairihul(i(frrnasya .'ia.sina.s .s'f"tka(1,p1,ac1!,if,c!:ma(icr

bhari,g(irUw:n nijavaqi.fo},;rt sa blwgo:IJ(in blui:nns samii:rop'ila~1,

mahalas] B; rn11[t.11}[ui} A

11

mii(iikyam] /\; mii(iikkam B (a Tarnilism)

"dimasya . .. bharigaT'lhan] H; /\ lacunosc .m blwyavan] B; /\ la.ctmosc .mmiiropit.11.(i] A; samaropil.am B

A ruby called the Essence of the '~hple World was set [by that
king] in the crown of the great Siva, Parvati's husband, who
dances in the glorious small assembly hall. [In so doing] it seems
to me as if the blessed Sun, the founder of his own lineage,
had been mounted there to spell the ruin of the Moon, the first
member of the family of his foes, who serves as Siva's crest-jewel.

Thim, despite the claims about the paradigmatic act of royal patronage
of the gilding of the citrmnpalarn shrine, early cpigrnphic rcforenccs to Cola
royal or court interest in Cidambarnm arc quit<) lirnite(l prior to the early
twelfth century. Other than the presence of a single (albeit significant) inscription of H.ajcndra I's time (10:rn CE), 18 there is 110 royal attention given
to the area or to its temple that can he gathnrcd from epigraphy, prior to
the final years of tlw reign of Kulottnilga l. In fact, this comwction goes
some way towards providing historical justification to KtJLKE's argnrnc11t
that this king supplies tlw prototyp() for the Cidl'vla's Hirar.1yavarrna11.
The second decade of the 1100\J mark th<) sndd<)ll appear au cc of a duster
of conspic11011sly major donations. Two of these arc by Kulottni1ga's sisters
Rajarajan Kuntavaiyalv~ir and Matnriintikiyiilwir. Thcs(~ consist of notably
grand gifts: in the first record, besides yd another claim to gilding the
temple's shrine, we learn that the princ<~ss gave a gift of gold, an on1<tmcntal
mirror, ancl a 'stone that had been prcsent<~d as a wonder to Srinijendrncola
[= Kulottui1ga] by the king of Karnboja'; tlw second records a donation of
1

"ARE 118 of J 888, Sfl IV: 22:1; it is this record that provides the sole cvidcntiary basis
for Kenneth I IALL's rr~constrnction of Cidamlmrarn as an clevm1th centmy 'ccouoniic
centre' (2001:101 lOG).

J>unA~IC THANSFOIUvlATIONS

:n

land (in the name of a local Urnhrnan) for the establi8hment of a ho8pice to
foed Saiva devotees. rn Undated, but alrno8t certainly from this 8mne period,
is the bilingual verse eulogy dc8cribiug the do11ati01rn of N aralokav1ra, a lo rel
from the northern Tamil country owing allegiaucc to the Cola emperor. The
massive project of building awl donation described in this record would have
totally transformed the built cnviromnent of the tern ple. 20
Thi8 transformation in the courtly and royal interest in the Cidamlmrarn
temple secnrn to hav(~ been 8calcd by the acts of Knlott11i1ga's sou and
succcs8or Vikrnma Cola, who unprecedentedly am10unces a series of major
benefactions to the site in the official eulogistic preamble (rncykk'l'rtti) to
this records beginning in his eleventh regnal year: 21
. . . patta~1,tu
vm"nm'U,'!_'ai rrmr_l'!JC rrwrJrJavar cumant'
'U.l.l ''ir_'ai nira"itt'llc dSrinta ccmpor);'UVaiycT'!_'
ta'!__lk'alan(Tyako"!J ta'(itavarn vnriyuii
C(~rrl,JHJ'(l,ampalarT c17J tfrurruT.likaiyuri
kop'll'('(L'/Jllyirk17ta ca_larrw,m
'11,lakavalariko'(ltav (Jf,'ivifa'nk'll ncmik
k'Ulavaraiy 'ntayakk'llr_irarru)t'I/, nir_ir' l~'!Jap
JHLC'U,rrl,]HJ'{_l mcynt'll [ . . . ]
19

Kuntavaiyiilviir: Kulottui1ga year M, 1114 cm; ARE 119 of 1888, EI vol. G, p. lOG (in
corrected Tamil orthography, the donation of the king of Kambhoja [= Cambodia?] reads
(11. 10-11): .fr[ T] n~jendracoladcvarkn kiirnbo.fan~jm1, [s'ic] kii,tciyM;ak kattina kalln); Maturiintiki: year 4G, llH> CE; ARE 117 of 1888, SII IV: 222, lines 7ff. (the local beneficiary
of the princess' gc11erosity, Viicciyau Im.vi Tirnccinampalarrmt-<1iyiiu, is mentioned in ln.

10).
20

T!tc Narnlokavlrn grant has had a complicated publication history: the earliest publislwcl rcfcrcucc Lo the inscription was l!ULTZSCI I\; notice of it as ARE 120 of 1888, while
the first a.ttmnpt at publication as Sil IV: 22G resulted in a poor edition. The Sanskrit
text, had also been copied down by one of the pandits working under Mackenzie and
was snbscqnently published (with some diffornnt readings and a Tamil paraphrase) as
8011,f;h lndian Temple Inscripl:ions nos. 1271 127:\. A somewhat difforcnt Sanskrit text
was published by B.G.L. SWAMY as au appendix to SWAMY 1979 (which, additionally,
is i(kntical to SWAMY and NAN.JlJNDAN 197:~). This latter text is acconqmnicd by a very
w1reliablc translation, hasrnl probably on the SITI gloss. 'l'hc 'l'amil text was reissued
in a heavily conjectural version by Mu. IHAKAVAIYANKAH. (19:~G<lG, nos. lOG9 lO!H).
Adclitio11ally, partial translations of tlw grnnt are iucludccl in N!Li\KANTJ\ SASTIU rn:~2
awl BALJ\SUBiti\llMANYJ\N 197!):2:1-2G. Despite all of these materials, the text remains
highly problematic. I plan to publish a revised edition and philological translation of the
eulogy in the near futmc.
21
Drawn from Vikrarna's inscription in the Apatsahaydvara temple in Alai1ku\j (yr.
11, ll:lO c1;, SIT V: 408, 11. :1, G, compared with Clll'l'lltAMAI;JIYAM 198:\, 110. 24 (pp.
1 L:l-114), 11. :l0-:\8, G2-GG).

pattii:rn <T'(/,f,'it cilliraill'irika.l

attu:m fJCTTa rit'iltu:urJ:ratl:ut

tirnvaf,arm,al'iy'i'[i tmyodruHpak!1:all'll

'iryr_iana pala:unrn 't'n:il'u, co:rrw:itl 'ar"u.lit

tan,

6'ruk:utu:i 'Tl:if,o:r /;alo:rn:u,/,'uJ:u:n /,af,<dppa . ..

From the heap of pure gold that was gathered together and
poured out before (him] by tlw lonl:-; [of tlw kingdom] out of the
abundant tax revenues of his tenth regnal year, he gilded the
beautiful enclosure that smToun<ls the hall of fine gold where
his family deity pc~rform:-; the tii:e4ava dance, ancl the hall along
with the temple-gateways, such that it appcarul as if the shining
mountain range which encircles the earth had been joined with
the mountain of the sunrise [... ] this along with many other such
donations, the king was pleased to accomplish in his tenth year,
on the Sunday that marked the thirteenth lW1:i of the bright half
of the month of Cittirai, as the moon was in I-Iasta [= Sunday,
15 April, 1128 CE 22 ], so that he might make flourish the entire
world [that lay beneath] the shade of his singular royal parasol.
This very public reference of tlw king's charitable interest in the temple
is remarkable for three reasons: it marks a notable dcpartur<~ from the model
of earlier Cola m,eyk11:frtt,is, which are almost entirely martial in character;
it punctiliously dates tlw occasion of his generosity (no other earlier or later
m(~ykkT/rtti includes a precise elate); and most interestingly - tlw record of
donations that Vikrnrna claims to his cn~dit arc evidently the exact same
as those claimed by Narnlokav!rn soni<! years earlier.
Vikrama's extraordinary cormcdion with Cidamharnrn and it:-; <lcity
(whom he calls for the first tirrw ta:ukulaniiya/;;o:u, his family god) finds
surprising confirmation in the so-called Golra8a:nluJ'i appcndc~d to tlw expanded text whose core is provided by Aghornsiva's mid-twdfth c<~ntmy

Kriya:kro:rruulyolika: 2:i
ta cc11:('i'.IJ o .r;u:a{lavr5(1,b/w(i ,~'f"t'1:<1:(l ?Jw,~ivrulr:.~'i/;;0,(1,
.~'f"trnaddabh:ra8abhe.~(/,r1,arlidrk5rJ:f;o gala8 f'I/, ya(1,

I
11

srfvif;,ramadicoJesa~i svaguruJve nyu:ue,foyan I

His pupil was the t<~ach<~r Snka1.1thasiva, an <~astcnwr, who went

forth out of his desire Lo sc<~ Siva in the Dahhrasabhii [i. c. in Cid22

The calculations am KIELIIOHN's, in EI VII: S.

-"Quoted as rntmt<lr~d by GOODALL L!)!J~:xiv.

')I)

Pun.A~Ic 'I'HANSFOH.MATIONS

amharnm]. The lords of the Coias, heginuiug with Vikrama, 211

accepted him as their master.
Over the course of just a few decades, then, Cidamlmram wtw transformed from a Saiva site of some irnportm1cc and antiquity to the linchpin
of the religious culture of the Cola conrt. This sudden rise in significance can
he linked to the new political order inaugurated by Kulottui1ga's contested
accession. The amount of resources <lircctcd to the lmrgconing temple town
appc~ar to have been prodigious, while dual clainrn made in the name of Naralolrnv1ra aucl Vikrama Cola marked by a seemingly deliberate transfer
of institutional charisma from lord to king- suggest that it may have been
a site of symbolic coutestatiou, and of the working-out of a newly forged
political order.
The long history of the temple and its urban surrounding in the 1100s
and 1200s, however, is not simply that of these courtly elites. Even in the
cases where figures from the court appear in the records, they do so alongside and, at times, at the behest of local people, seemingly the leading families and individuals of the Cidarnharam rnicroregion. Unusually for Cola
period inscriptions, these groups tend disproportionately to be hrahmans
(about half of the lGOO individuals). Individual bralmmns seem to have
been among the major landholders in the environs of Cidambaram: the
temple's epigraphical corpus presents numerous instances where such figures arc the donors of parcels of privately held property to the temple and
clo~ens more cases where these brahrnans arc adventitiously mentioned, as
when their lands provide a boundary for a donation, or when they serve as
witnesses to a transaction. The impression that emerges of the Cidambaram
bralnnans is not of temple functionaries, nrnch less of a single hornogenous
group like tlw legendary Tillai Three-Thousand, but of a substantial and
powerful lauded clite. 2" The rising fortunes of these landed Brahman groups
seems more significant given their seemingly anomalous status. Most of the
named Brahman figures arc united by their um1s1ial golras or hralnnanical 'clans': Kavuciyai1, Viicciyai1, and Ulaiccarm.mu, names that appear
to be concentrated in Cidambarnrn itself, with the last ancl most anorna21

SJ\NDlrnSON bas suggested (personal cornm1micatiou) that the uikrmn1i.di could alternately be interprnted as a periphrasis for jnst the king himself ('he whose name begi 1rn
with Vikrama,' the plmal mm1her being a mark of respect, dda:rc balwvacanmn). vVhile
this sort of construction is frequently met with, it seems unlikely to me i11 this instm1ce, as
the king's rcgnal name is in fact Vikrnma (and not, say, Vikrnrniiclitya or Vikrnmasi1!1ha
or something similar). Nevertheless, I am not certain who these three or more 'lords of
the Co!as' might be.
2
"Scc Onit 20(M: 2:11, from whom I draw my figures.

vVllITNEY

Cox

lous of these being by far the most frequently e1Jco1tutered. 2 r; This unus11<1l
onomasty suggests perhaps a period of recent brahmani>mtion, when~ local
elites adopting a place within the genealogical ideal of the pan-lrnlic caste
hierarchy had yet to accommodate their collective id<mtities to translocal
norms.
I propose that it was from these local Brahman groups that th<~ cornpilern
of the CidMa emerged, as they sought to provide a textual warrn11L to
represent their newfoun<l prosp<~rity and significanc<~ in tlw kingdom. The
seemingly recent and perhaps insecure foundations of their brahrnauical
pedigree may have induced such men to craft a text like the CidM~i, in
an effort to narrate themselves into the rww political and cnltural order of
twelfth century Colamar.H.lalam. 27 I further propose that members of thiH
brahmanical society were responsible for the redaction and expansion of
the SuSat!l into itH present form. ThiH redaction I believe can be dated
to the mid-12th century at the earliest. Partially, this conclusion rests on
circumstantial evidence: the rising importance of Cidarnhararn in the lattc~r
portions of the text congrncs with its riHing significance in this period. It is

2G01m op. cit: 2:1:~n. Of these gotra names, one can recognize in the first the panIn<lic gotra Kansika. Vacciya!l perhaps suggests a role in ritnal or other public performance (::=;:; Skt. viicaka'!); tlw Tamil/;; kalvetf,uccolloJ,;amti, s.v. vwxiyarilrn.l gives 'musical
instruments' (viidiyarika.l), citing a twelfth century record from the Kannada region [=
EC, x, 1:12]. The name nlaiccara~wn found almost exclusively in Cidamlmram remains
opaque; indeed, even its orthography is uncertain: 01m op. cit. ibid. reads ulaicmnan
(I follow the transcription in 01m's prepublication manuscript; the names given in .the
published version are incorrect). I choose instead to give the name that seems to yield
some sense ('deer foot'); the two renderings arc graphically indistinguishabk in inscriptional Tamil. Taking KAHASIIIMA, SUBilAitAYALU and MATSlJI's representative (lmt by
no meaus exhaustive) name concorclanc() ( 1!378) as a gnid<), only two of the eight, lignn:s
bearing the nanw-scgment Ulaiccara~m11 (vv .11. ulafrcap.u:n, nla:icca.lan) an) found <mt.side
of Cidamlmram (the majority of 01rn.'s references to Ulaiccarar_1a11-ideutitie(l figmcs am
found in lmpuhlished inscriptions not talrnn into account in th() couccmlanc()); of the
thirty-sevrm figmes in the concorclam:e with namc-scgrrwnt Viicciya11, seven am from
Cidarnhararn. vVhile the distribution of nindy-cight figures hearing the gotrn Kavuciyatt
or one of its varia11ts (ko:u11,c-ilrnn/ka1m:u:iy11,n/l.:avi~iiy11:n, c~tc.) show wi clc~ar regional focus,
it should he noted that among Uwse ligmes there is a high irn:idct1C() of p<~rsonal nauics
comwcti11g their bearers to the Cidarnharnm tcm plc (e.g.
/,fl lairw:1;1L/;:1Lbhaf,f1Ln
110. GIii (i,
.
'
..
m:1iv1!,1}'irapatf.an no. 1HiGD, afavifarikapafflln no. 4(;fi!J; aJ11,/;;a,o.zta1i./,:ari11.f.f. 11,n 110 . 11()!J2; /.irnu:irn.1:rnpalam:nf.ayan llO. 12:\D, i,iltafppirnn, the fatli<~l' of' the si11glc lllall recorded t1t1dcr
nos. 1818 and rn:w, etc.).
27
0rw may contrast lwre Kuurn's sngg<)stion that th<) author or authors of the CidMii
might have been among the hrnhrnaus trnnspla11tc~<l by K11!0U11i1ga [ from Vci1gi to
Cidamliaram, for which them is 110 cvidenC() in the) epigraphical n)ccml ( UJG7: 20 I 20!!;
l!)!J:l: 202 2fH). KuL1rn's hypothesis has Ileen r<)p<~atc)c] dsmvhcrc as history: sec c.q.
vVITZEL rnD:l: 2G7. The present theory has at l<~ast the advantage of an actually-existing
social constit11crn:y in which the text's ant.hors may IH~ locatc~d.

PUH.J\~IC TllJ\NSFOIUvlJ\TIONS

37

in the final arlhyayas of the Mukt:ikho"u4a aml the vast bnlk of the Yajr1ava'iblw:ual1:ha'(l-(ia where the centrality of Ciclamharam is most evident; further,
these two contiguous sections of th<~ SuSaq1 evince narrative and formal
features setting them apart from the earlier sections of the text, suggesting
an already existing version. 28
It is iu the final kho:r14a that we find the third narrative of onr sinner
rcclccmccl by Na~,csa's grace (Appendix 2), the version in which KULKI;
suggested we sec the source for the CiclMa's Pnlkasa story. Herc the central
figure is called Durghata ("Difficult" or "Obstreperous"). The narrative is
much more attenuated here, and the language less sophisticated than either
of the other two versions. 29 What this version lacks in polish it makes up for
with drnt'./,pah: the foit of Durghat;a's crimes is heavily infiated, np to and
including such anti-social activities as polluting wells and tanks. :m While it
lacks on the whole the close correspondences that are so striking in passages
given in Appendix 1, it can be demonstrated that it is based on SuSaq1
1:4: 17 39, that is, on an earlier portion of the text in which it is found, yet
again suggesting the final kha'fir.la to be a more recent aclclition. :n
The eighth adhyiiya of the fl;foktiklw~i\la begins by changing the interlocntors in
its frame narration: where the rest or the kha:(i\la is a conversation between Vii;ll_lH and
Siva, it is an exchange between the Si:ita awl the Naimis;;1 r\~is, the primary speech
situation of the rest of the SuSaqt. The same adhyiiya contains passages in longer lyric
meters demonstrating palpably Dravidian formal features like front-rhyme and heavy
rhythmic syncopation; in the ninth and final adhyaya the setting for the frame narrative
shifts from the Naimi~a forest to Cidambaram itself, where the r~9iS witness Siva and
Parvati's procession nuder the Ardra nak:~atra (sec Cox 200G: G8 72). This shift in
mnltiplc registers of the text's language and rhetoric suggests that these two final chapters
of tlw third khau\la were a Cidamlmram-spccific addendmn to an already existing tc)xt,
as well as a narrative bridge to the fourth and longest of the SuSaq1's sections. For
its part, while the Yajiiavaibhava is cqnally focnssed on Ciclarnhararn among tlin many
Saiva 8itcs it uic11tio11s, tlw section as a whole is C)8pccially repetitive and 8tylistically
clumsy, even by the) forgiving standards or the rest of the SiiSmp, something already
noticed by H.J\ClIAVAN ( lD47: 211G). Prcsmnably, the earlier, shorter SuSaq1 was already
in circnlatiou, si11ce at the very least the opening kha'(l-\ln was available to the CidMa
compilers.
wln the Appendix, I have omitted a tediously catechistic roll-call of the figmes who
cornc to worship Nat.c~sa (11:21.11 20).
:ioKlJL1<1,; suggests tl1at the CiclMil.'s Pulkasa is given a clca11c<l-11p rap sheet, in order
1.o fit i11to the morn settled world of the L2th cmitmy and the greatc~r sacrnlity of Cidarnlmrnm ( LD70: 210). Ott the contrary, the literary cult me of twelfth centmy Colairmnda.lam
seems especially proll<) to this sort of violent hyperbole, considering snch key Tamii ~orks
as the Kali'ri,kalluppanirii and the Tirutilu,f,0:1p11:nli[!J1:m (See A LI 2000, MON I US 2004, Cox
200G).
:it The telltale signs of this dependence are given in boldface in Appendix 2: the reading
ma./u1.sahasiko 'dlw:rna(l, in 4:21l.2d (fon11d only in the edition's ms. gha) resembles SilSam
l: 1l. l 7<l's pnrn:;;o:dhama(i; Ga's snhasn~janmata(1. follows SuSmr1 1.4.2:k jannu'inlamsalu;,28

\!VIIITNI;Y COX

It is however the final concluding vcrnc in Appendix 2 that is most significant, as it supplies firm evidence for tlw dating of tlw composition of this
passage, and so for the compilation of the SfiSarp as we now have: it, a date
that is considerably later than the one assigned by I(AG llAVAN. Unnsually
for the Yajr1avu:ibhavo}Jui(i,(la, this final verse is written in a lyric meter (upajrit'i, mixing varr1,.fostha and u,pendravajnL). It conce~rns the two gills that
Dmglmta receives thanks to Nat,da's favor. Both of these arc significant:
the first is that our redeemed sinner became a world-ruling mnpcror, sa:mastalokadh:ipatfr babhfi:ua. This is an innovation seen only in this version of the
story, and one that makes sense only in the context of the new courtly and
royal investment in the site, really only after 1128 CE, when Vikramacola so
spectacularly linked his awl his family's imperial fortunes with the temple.
The second of the two boons granted to Durghata is the .mmrn:nrn bonwn
of rn:ukt'i. This is de~scribed in an tmusual and, for the Vediintic SuSar~1,
highly unorthodox way as parapramritrprathanu:ikalak.~a'(ULm ("characterised
by the manifestation of the highest cogni2er").a 2 Miidhavarnantrin tries,
unsuccessfully, to explain this term:
prarruIUi 'nta(d,;aratwpah:ita(i sak.<j'l paro:rr1, n'irat'i.foyarri pramMu:r
'ILpiidh'iv'ilayena yat prathano:rri tad e.ko:m cva lak.<jo:(w,rr1, yasyas
tam 'ily artha(1,

The 'cogni;';er' is the soul, the witness [to all apparent phenomena] as it is equipped with the apparatus of internal sensation,
and it is the 'highest' or ultimate manifestation o[ this cogni;';CT
that is the sole characteristic of this [liberation]. That is the
meaning.
From this we can sec that despite his rnal Saiva learning Mite llmvamantrin fails to undc~rstand the nuance~ of this compo1111d, and tries rather
unconvincingly to slot it into a conventional Vcehintic framework. H.ather
than having an upanif?<ulic antecedent, the term fHl'm,prmnritr has here been
taken over from th<: Prntyabhijfi~i idealism of the Saiva nondualists, for
whom it denotes Siva as the unlimited awl trnnsindividual conscious agent.
This is not, in fact, a usage distinctiv<~ of Utpalad<~va\; f.foa:ra1nnlyaJ1h:Uiir/);iir'ikc/,, hnt is ratlwr a term of art from Ahhiuavagupta's Vim,a:r.frn:l; whence
it was to go on to be wi<lcly used tlmrnghout later and allied literature,
from Kf?crrmriija's tirn<~ onward. The sam<~ is true, although less d<~cisivcl.Y
src.~m (itsdf part of a phrase that is notably 1mparallclcd in tlw Cidl'vhi vcrsio11, sec P
~!~,.above); ,while !Jc mal:ipnidann:rri seemingly is a parnplm1s<~ of rn,l/,/:i'fri.... pm,d(l,/.ta.van,
S11Saq1 1:4.20(1.
'j'J

-Contrast the liarml 11pa11i::mdie paralld iu CidMi-i 8.2lli, s11.u;idnnand11.lnJ,::;;11.u11.1r1,.

39

PuH.At:HC TH.ANSFOIUvlATIONS

and more anecdotally, of the next clement o[ the compound -praf;hana-, as

derivatives of the vcrh pratlw,/;c arc frequently met with in that system's
intellectual jargon.
This then gives us an absolute limit [or the final redaction of this section of the SiiSaip, sometime a[Lcr Lhe mid-eleventh century at the very
earliest; more likely several generations later, giving time not only for these
Kashrnirian works to transmit the length of the Snbcontinent,:n but also
to have become cno11gh a part of the local Sanskrit discourse that it would
have been recognized by the middle-brow audience for which a work like the
SuSaq1 would have been intended, while still retaining some of the trendy
glamour that ::;cenicd to have attached itself to these avant-garde speculative writingi->. This brings us in all likelihood into the early or the middle
of the 1100s, exactly the time that I suggested the linkage between Cidambarnrn ancl imperial power first became a prominent clement of royal
self-presentation. The creation of the CidMa ancl the expansion and redaction of the SuSarp were thus closely parallel processes, likely to have been
undertaken by members of the same bralnnanical sodalities.
We can thus see here the ways that these particular pnra~1ic works recast
or ::;imply take over the language of their sources, whether these are materials external to themselves or even out of their own textual substance, as in
the case of the internal recycling evident in the SuSar~l. In certain cases, as
here, this allows us to situate the acts of text-making through which these
works are constituted both relatively and, to some degree, absolutely in
time and in spa.cc. These sort of pnra1.1ic transformations are potentially of
interest to Indological readers of a wide spectrum of interests, as the case of
the adaptation of the Pratyabhijiiii here shows. ThiH is because these transformations provide useful evidence of reception history--::;ometirnes more
iutdlectual-historically useful than even works of learned exegesis, as we
can trace how terms and ideas p(~rcolatc more widely in literate society,
as tlw pnrii1_iic audience is definitionally a wider constituency than that of
works of sastric scholarship. 'I'lmt the SuSa1p redactors could draw upon
the stock of Prntyabhijfia jargon and presume it to be intelligible or appealing to their prcsunied target audience casts light 011 the later success
of works 011 that and alli(~d systems in the for South or indeed in Cidambarnm itself, such as the J>aratr'is'i/;;(/,/iilpo:ryadfp'iklT (a verse epitome of the
Lagh'nvrtt'i that is atLrihutcd to Abhinavagupta) or the Ma//,(/,rlham,afijarf
of Mal1esvara11arn la. :M
------~-----~--

:nsrn~ Cox 2011.

:H 011 the provenance of the ParalTi.~ikatalparyadi.]1ikri sec
on the !vfoharll1,nm,a:i/,ja:n, see Cox 200G.

SANDE!tSON

19DO: :12 :1:1;

WI llTNEY cox

40

Equally important however is the transformation of the historical world

in which these works emerged arnl were initially cliss<~miuate<l. Purii'(W,S,
as eminently puhlic texts, served an cmmciatory or articulatory function
within their cultural orders. The changing place of Ciclamlmram within the
twelfth century world of the Colas seems to hav<~ been an especially powerful
inducement to this sort of articulation, s<~en in these two texts claiming for
themselves the status of vunL'(l,a hut also critically for a novel claimant for
that genre, the T'ir"1do'(l,f,arfmrcJ,~w:m of Cekkilar (better known by its popular
title, the Pf::riyap'll:r'ii'(w:m or "Great fJ'll'f'lL'(ia"). This was a work that in its
own way sought to articulate a particular gloss on the textual hori11,011,
Sanskrit and Tamil both, in which it was locatcd.:ir; Cekki]~ir completed his
work there during the reign of Kulottui1ga II (r. 11:3~3-1150 Cl~): that at least
three such 7mraFas were underway or brought to fruition in Ciclamlmrarn
over the middle decades of the 1100s allows us to track in remarkable detail
the complex world of literary and Saiva religious culture at work there, and
gives us some indication of its stakes.

References
ALI, Daud 2000. "Violence, Gastronomy and the Meaning of War in Medieval
South India." Medieval History Journal. Vol. :~ no. 2, pp. 2Gl 90.

A 1rni:i AC ALAM, Mu. 2005. Tamil ilukkiya va:rrLW:ca: patl1l,ram n:n'{TIL'Ut'/I,, ira:u,fJirn
fHLkam. C6rgmi: Ti Parkiir (reprint).
S.R. 1979. Laler Chola temples: [(11,lo/,/:miga I /,o Il1Ucndra III (A.D. 1070 1280}. Madras: Mudgala Trust.

BALASUBHAIIMANYAM,

IL rnDG. Trwlc, Ideology, and UrbanLzalion: 8oulh India

:JOO fJC lo 1300 CB. Delhi: Oxford University Press.

CIIJ\Ml'AKALAKSllMI,

()idarnbaro:m(J,halrnaymn. Ed.: Edited by Sornasundara Dikshitar, Kadavasal,

1968; A [= KULKJ;'s CMl]: transcript o[ GOML D. No. rn:M7 (paper,

Devanagari); B [= CM2]: transcript of GOML R.7G:~2 (pap<~r, Dcvamigarl).
Cox ' ,_
Whitney
t.,ion o t' '!'". u.nptn tlw Archer (Studies 111

. '_ ; ')()()'-)
~
". '"1'11('
_, rl'..
r,rnsf'1gura
Cekkilar s Penyapura'(W:m I)." Indo-lro:n:ian .Jo,u:rnoJ. 48 (:~-11), pp. 22:~ 2G2.

200G Making a lo:nlra in Nledieval 8oulh India: the M a!tii:rllw:mai/,jart

and lhe. Tc:rf'llal ()ull'll:re of Ciila Oidmnbo:mrn. Ph.D. Dissertation: Dcp;irtrnent of South Asian Languages and Civilizations, University of Chicago.
---

:is~'.lefJ Cox

200:> ;-;~~l 2o(J7:---

l >unA~lC 'l'HANSl<'OIUv\ATlONS

41

. 2007. "From Avm,inm to Purfir_1am" in Barrett, T. H., ed., Dimcnsfons

of South Asian Religion. SOAS Working Papen; in the Study of Religions.
The School of Oriental and African Studies, pp. G-:M.
. 2011. "Saffron in the Ra.rnrn" in Yigal Bronner, Whitney Cox and
Lawrence McCrea, eds. Sonth Asian Tc:rls 'in Ilislory: Orit'ical Engagcrncnts With Sheldon Pollock. Ann Arbor, Michigan: Association of Asian
Studies, pp. 177-201.
CuPPmAMA~lYAM,

Pu. 198:3. MeykkTrLilm1. Cc11rpti: Ulakat 'farnilarfiycci Niru-

vanmn.
Epigrapli:ia Indfra. 19:~:~-. Archeological Survey of India. Delhi: Manager of Publications.

Dominic 1998. Bho,_t,tar'iimakari,lhaviracitil.kira:u.avrtU[i: Bha,t,ta Rii:rrwkau,,tha 's corn:rncnto:ry on the Kfru(wtantra. Pondichcry: Institut Frnnc;ais
de PondichCry / Ecole Frarn,;aise cl' Extreme-Orient.

GOODALL,

LLALL, Kenneth R 2001. "Merdiant8, Rulcrn, ancl Prie8ts in an Early South

Indian Sacred Centre: Cidambaram in the Age of the Colas" in Kenneth
R. Hall, ed. Strv.ctv:rc and Society in Early So'IJ,fh India. Essays in Honour
of Norbon1, I<arn,sh:irna. Delhi: Oxford University Prcs8.
lrL\KJ\VAIYJ\NKAll, Mn. 19:~G :Hi.

Pb"antokai. Maturai: Maturait Tamilc Caitka-

rnutiracalai.
fS-uarnpral;yabli:~j'iliivfrna:r.frnf of Abhinavagupta. Edited by Mukuncl Ram Shastri.

KSTS No. 22, :~:~. Bombay: Nirnaya Sagar Press, 1918, 1921.

Noboru, Y. SUBBAHJ\YJ\LU, and 'I'orn MATSUI 1978. A Concorda:ncc of the Narncs in the Cola Inscriptions. Madurai: Sarvodaya Ilakkiya
Pannai.

KAilASIITMA,

K ULKE, I-Iennanu 1970. (}ida:rnbararnahiii:mya: cine UnlC'rsuchnng dcr rcligionsgcscli:ichllichcn u:nd historischcn INnkrgr'iindc fiir die Entstch:ung der TrndUfon cincr s1J.dindischcn Tcrnpelstad/;. Wicsba.clc11: Harrassowitt-;.

. !!)!):~. "Functional Interpretation of a South Indian Miihatrnya: The

Lcgmicl of Hirnr_1yavannan and the life of the Cola king Knlottui1ga I" in
Kings and Cults. Delhi: Ivfanohar.
Anne 2004. "Love, Violence, and the Ac8thctics of Disgust: Saivas and
Jn.ins in medieval South India." Jott'f"nal of Indian Philosophy Vol. :~2, pp.
11:~ 172.

MON!US,

cox

WlIITNEY

Nll,J\KJ\NTJ\ SASTIU, K. A. un2. 8l'lul'ies in C6J.a lhslory and Adm,inislratfon.

Madras: University of Madras .
. 19GG. The Cofas. Second edition. Madras: UniV()rsity of Madras Press.

OIUt, Leslie 2004. "Temple~ Lifo at Chidambarnm in tlw Chola Period: An E~)i
graphical Stndy." In SrT Pu:;;JHii~joii: Vr. C. ll. 8rinivo,.-;o,n Com:rncnwroJ:wn

Volume. Edited by K.V. Ramesh. Delhi: Bharatiyn, Kala Prakashan.

\{,AGIIAVAN, V. 1947. "T'hc 8Fitasmrihitii." Annals of the !Jhandarkar Or-ienf!il
Research Institute. Vol. 22, nos. :~ 11, pp. 2:rn 2G:L
Alexis 1D90. "The Visuali~ation of the Deities of the Trika." In
L 'Image Divfrw. Oulte ct lvlcditalfon dans l 'Hindouisrne. Edited by Andre
Padoux. Paris: Centre Nationale de la Rccherchc Scientifique.

SANDEllSON,

___ --. 2001. "History through Textual Criticism in the Study of Saivisrn,
the Paiicariitra, and the Ducldhist Yogin1tantrns." In Les S(mrces ct le
Temps. Edited by Frarn;ois Grimal. Ponclichfay: lnstitut Frnnc,;ais de
Pornlichery /Ecole Frarn;aise d'Extrcrne-Oricnt, pp. 1-47.

ScHWINDLER, Gary .J. 1987. "Speculations on the Theme of Siva as Tripnriintaka as it Appears during the Reign of Rajaraja I in the Thanjavnr Area,
c. A.O. 1000" Ars Oriental-is. vol. 17, pp. rn:~ -178.

So'Uth Indian Inscriptions. Mysore: Archaelogical Survey of India, 198()

(reprint).
So'Uth Indian Temple Inscr'ipt'ions. Edited by T.N. Snbrnhnmnimn. Madras,
Govt. Oriental Mannscripts Library, lDG:~-.
Su1mAIIMANYA J\JYAit, V. S., .Jean-Lnc C111w1LLAIW and S.A.S. SAIUvl/\
2007. Digital Ti:'uiirarn. Pondichfay: Institu t Franc;ais de) P011dichfay / J~;colc
Franc~ais<) d'Extrforw-Oricnt.

SmmAIIMANYAM, S.ll. UH2. "The Oldest Chidarnhararn Inscriptions (Part 2)."

Jo'Urnal of Annanwlai Unive'r"';-ily. Vol. i:~, pp. GG !H.

Sfilasamhitci,. With tlw 1'iitparyad'lpi/Ji of M<i<lhavamantrin. Edited by Wasudcv

Laxrnan Shastri Panshikar. Anawhisrama Sanskrit S<~ri<~s No.' 2G. J>1111c,
l8D:~.

SWAM_Y, D.G.L. 197D. Chidambarnm and Naf,an~ja: Problem,,-; and HalionalLwlwn. Mysore: Gectha Book Honse.
SWAMY,_?G.L. and K.R. N/\N.JUNDAN rnn. "T\vo Inscriptions from the Natarn.ia
tcrnpl<~, Cidambarnrn"

fo'tl'T"Tllll
, , ,, ()j' I'r 1,c uin '/,,<; l,m"tf ('_,o 11
< en .J111H"}('('
,,
Volume, pp. :n:~ :Ms.
'

1r

PU!LA!;JIC TlU\NSFOllMATIONS

Tarn:ilk f,;o,lvef;tuccollo,/;;a,ra:ti (Glossary of 1'a:m:l Inscriptions)

ras: Santi Sacllmna Charitable Trust.

2002-200:~.

Mad-

Michael 199:L "'L\iward a History of the Brnhmins." Jonrnal of the

American Oriental Society, Vol. 113, No. 2 (Apr.-.Jnn., urn:~), pp. 2G4-2G8

\NLTZL;L,

Ka.mil 19DG. Lc:ricon of Ta:m'il Uterah1:re. Ifondbnch der Oricntalistik.

Zwci/;e Abtclnnq, huHcn; D. Bel. Leiden: Drill.

ZVELEBTL,

Appendix 1: SiiSa 1:4.17cd-39 and

pura kascin nmhap~ipf pulkasal.1 1niru\)1idhamal.1 I 17
hriihrnar.1anaq1 lrnla1ri hat,v<1 gavarp vcdavidiiq1 vanil.1 I
apahrtya clhanmp m~irgc pr~i1.1ihiq1s<ipurnl.isaram 11 18
svacd1arnlaip nirghp.10 vipras cacara prthiv1talc I
tasya mitram ahhut kascid hral11na1.10 gar.1ik<ipatil.1 11 ID
tasnmi dattarr1 dhanarp kirpcit pulkasena dvijottamal.1 I
sa pnnar hrahnmr.1as

tu\)~;o

rnatiq1 tasmai pradattaviin

11

20

bralnnar;a uvaca
subanclho mama dmbuddhe tvaya papani nirghp_1a I
lqtani sarvada mu(Jhatvaq1 ho kiq1 tc phali0yatc 11 21 ->ity cvarr1 hahudha vipral.1 pulkasaq1 pratyahha0ata I
so 'pi vipravacal.1 srutva hahu8al.1 par.1cJitottama1.1 11 22
kalcna mahata clantal.1 pulkasa~1 pmrn.>aclhamal.1 I
janmantarasahasre~u krtapuvyavasena ca 11 2:3
:J
prar.1arnya clar.HJavacl viprarr1 papraccha bralunavittamal.1 I
pulkasa uvaca
clcva vipra subanclho me maya sarvatra sarvacla 11 24
rnahaghorar.1i papani krtani mama nayaka I
kirr1 karomy aham aclyasmirr11 lokc rnuc.1110 'tinirghp.ial.1 11 25
tad vaclatibhayavist,arr1 rnanasaq1 mama sair1tatarn I

hrahmar.ia uvaca
sacllm sacllm tvayaclyoktaq1 tava vak0yc hitarp sp,lll 11 2G
pm,iyak0ctrc rnahat!rthe puni1.w vasuclluitalr~ \
rrrnnihhil,1 sicldhagawlharvair amarais ca suscvit,<~ \I 27
sr1madvyaghrapure yatra prnnrtyaty ambikapati!.1 I

PunA~Ic TH.ANSFOB.MATIONS

45

CidMa 8.3-23 co1npared

asft kascit pura. piip1 pnlkaso nama 11iiumtal,i
kn)ai.iena sut1k~1.wna kularp hatvii dvijamnauiim II :)
margc 'pi (11w.11am alqtya pra1.1ihirpsapuras0arnm I
svccchavrttis cirmp bhurnau cacarnit_.;a hi nit_.;kqml.1 11 '1
dhurtal.1 kascid abhun rnitrn1r1 dvijauma. tasya nirghp,ial.1
hrtiid arclhmp clha11aq1 prftal,i pnlkaso '0mai pradattavan 11 5
dhanalahhcna sa1r1tnz-;t;al.1 kqmr.10 'pi ca sa dvijal.1 I
cladau tasmai mati1p fo1tarp pullrnsaya snkhavahiirn 11 G
hra.11mai.ral.1
papas1lena dnrbuddhc tvaya papani nirghp.m
k~tani sarvada. mii<.lha nanntaparp lrnrot_.;i ca 11 7

ity cvarr1 bodhitas tcna duratrna sa hi pulkasal.1

daiva.d as1t tada putrn bhpfarr1 nirvir.n.1amanasa1.1

11

a.ho papani ghora1.1i lqtany eva rnaya dvija I

ki1r1 karomi prnrnfa.lho 'ham adya 11i9knmrnanasal.1 11 9
upayarp papanasaya vada tvarr1 dvijapmpgava I
cintavi9ta11 tu me cittarp naiva janiiti kiq1cana 11 10
bra.hmar.ial.l
sadlrn s~idhu tvaya P!'t_.;\.am atrn. vak0ye hitaq1 sp_m
1n11,1yak0etrott.amc clivyc purii1.1c vasudha.talc 11 11
tnlmibhis siddhagaudharvair arnarnis ca 11i0cvitc I
sr1mattillavanc yatrn prarqtyat.y ambikapatil.1 11 i2
:1b Harnat.al,1] J\11; manaval,1 Ed. tlb prm,1i')] Ed. BJ\7"'; prn.1.1<" J\"c 4c
J:d. J 3;

VfttiS SVayatp

J\ 4d ('ClJ'.Ci.ra.i~a] fi~d.J\JlC; C(l,C().re~;i<l i'\ ac: LJPC;

vrt.t.is cirnrp]

CaC(il'CV<l

nne 5a

dhmtal.1] Ed.11 ; arUa.l,1 J\ 5h dvija11rn<i. tasya.] J\Ll; dvijanm<1H aHya Ed. 6b kqmr.10
'pi ca] I:d. llJ\I"'; kn>aha11110 'pi J\"" 7b nirghp,1<1.] Ed.; 11irghri,rnl.1 J\B 7cl na.1111tii.pa1p
kan:d] J\Jl rni.1111!.apa Irn.ro~i Ed. 8c bhrsa.q1 nirvi1,11,111] Ed.BJ\7'"; nirv;i.1p1. J\ac 9a-d

J\ ornits llh a.tra.] Ed.ll; adya J\ llb va.kfiye hit.arp] I~d.ll; vak~ye ('hi) ha1p ta1p
Kp,111 J\ 1'"7"i. "'"''" lld vas11dhata.le] Ed.; vas11dh<igat.e J\11 12d ambikil.palil,1] Ed.;
il.rnbikapa.til.1 J\; ambikii.patirn II (cancelled sec. man.)

--------------~~----------

Appendix 1: SuSa 1:4.17cd-39 and

tatra hhaktiparo bhutvii srnitvii priital.1 sarniihital.1 I 28
drstvii clahhrnsahhiitll chirc pra1.1amya bhuvi dar,1<,lavat I
yoginaq1 hhoginaq1 nf1.1iiq1 dattvii sarvasvarn arjitam 11 2D
sthanasyasya hhayc jute rnk\)ar,iaq1 kuru yatnatal.1 I
suta uvaca
cvcu!l dvijottamcnoktal.1 pulkasal.1 purn\)adharnal.1 11 :m
sr1ma<lvyaghrapurarr1 pur.war11 gatal.1 srn<ldhapmal.1sarnm 1
bralnnar.10 'pi sahancna srad<lhaya nmnipui1gava1.1 11 :n
praptavan ctad atya11taq1 sra<ldhaya sthanarn uttarnam 1
pulkasal,1 sraddhayii snatva brahrnar,1a vcclavittamal,1 11 :32 - !
yogihhyas ca tathanychhyo clattva sarvasvam arjitam I
sr1maddahhrasabhar!1 11ityarr1 brahmar.1a vcdaparagal.1 11 :3:3
clurc dr\'tva narnaslqtva paflcakrosad hahir dvijal.1 I -r
uvasa suciraq1 kalam evarp krtva dine dine 11 :34
cvaq1 ciragatc kiilc pulkasal.1 pm,iyagaurnvat I
sthanasaq1rnk9ar.mvyiijiit tatraiva marn1.mq1 gatal.1 11 :35
sa punar mara1.1ad urdhvaq1 hlmktvii bhogiin anckasal.i I
srimaclvyaghrapuresasya prasiiclad amhikapatd,1 11 :3G -)aviipa pararnarp rnuktim aviglmcna dvijottarn~il.1 I
br~1mar.io 'pi tathaivasmin sthanc pratal.1 samahita1.1 11 :37
snanarp lqtva rnahadcvarn amhikapatirn Isvaram I ->pujaya~niisa purvoktaprakarm:.ia rnahcsvarnm 11 :3s ->- !
~a~yap1 hrahmaviccl1rc9tJ1al.1 pujayii paramcsvaral.1 I
snrnaddahhrasahhii11athal.1 pradadan muktirn 1svaral,1 11 :rn

PunAr~nc TltANSFottMATIONS

47

CidMa 8.3-23 cornpared (cont'd)

tatra tvarp. 8rncldhayopcta1.1 priital.1 suatva yathaviclhi I
clurc clabhrnsabha.1r1 dn;tvii prarpunya blrnvi da1_1(_lavat 11 1:3
yogi11a1p hhogi11ii1p 11f1.18.1p clattva sarvasvam arjitam I
sthiinam ctacl bhayc jiitc rakf:la nitymy1 prayatnatal.1 11 lt1
cvam ukto dvijcndrcr.m pulkasal.1 pr1tarnii11asal.1 I
sr1rnattillavanarr1 pu1.wa1r1 8raddhaya sahito yayau
clvijaumiipi samary1 term siidaras tad aviipa sal.1 I

11

15

pulkaso 'sau prahp:j~;iitrnii jfiatva saclayamiinasal.1 11 lG

yogihhya.8 ca tath[iuyebhyo clatvii sarvasvarn arjitarn I
8rirnadclabhrasahhiiq1 divyiiq1 nitya-rn-adblrntavaibhaviim 11 17
drf?tvii dlire narnaslqtya paficakrosiid hahil.1 kvacit I
uviisa pararnapritir cvaiy1 lrnrvan dine dine 11 18
iti kale lmtipaye samatitc sa pulkasal.1 I
sthiinasarprnkf:Ja\mvyiijiit tatraiva mara.1.1ar~1 gatal.1 11 19
8rirnii11 asau tatal.1 pm_ryiid blmktvii bhogiin yathcpsitiiu
srimadvyaghrapuresasya prnsadad ambikapatel.1 11 20
atrn muktiry1 parii1y1 lcbhc saccidiinandalaks;a\1iirn I
hriilnnar.10 'pi tathaivaitat sthii11ary1 priipya rnahiimatil.1 11 21
sthanarp lqtvii mahiidcvarn ambika.patirn avyayam I
pfijayiirniisa piirvoktaprakarc1.1a natesvaram 11 22
plijiiq1 lqtavatc tasmai dvijendriiya dayiiniclhil.1 I
8rimaclclahhrasahhiiuiithal.1 pradadau muktim uttamiim 11 2:~
13b \ll"itl.al.1] Ed.I\; pnl.il.1 B (ca11ccllcd sec. man.) 13c dlire dabhrasabh<i.rp] Eel.A; dlirfi.d
ahhrasabhatp ll 111a yoginatp bhogi11a1p] Ed.A; yo;i11[1.q1 yogin;!.tp ll l4d 11it.yarp prayat.llat.a.1,1] Ed.ll; nit.yapraya.t.r1a.t.ah A lGb pnl.a.manasa.l,1] Ed.JlAJic; prn,1ama.11asa.l,1 An.c lGcd1(Jab i\ omits 171> arjit.mn] Ed. arjil.am All 17c 8rn11addabhrasabha.tp] Ed.A; srlmadabhrasahha.rp ll 18c para.rna.prltir] Ed.A (cancelled sec. man.); paramapnl.al,1 ll (wit.It
bad sandhi) J!)b sarnal.1\.(~J Ed.A; su111at.lt.c ll (cancelled) 19c st.h;i11asaq1rnk~m.1avy<ij;i.t.]
Ed.lli\P"; sl.h;i.11arp sarprak'?a1.1aq1 vyitja.I. A"" 23b da.ya11idhil,1] Ed. i\; yat.h;ividhil.1 1l 23c
snmaddabhrasabha. 0 ] l;d.A; snmadahhrnHablwO 1l

WIIITNEY

48

Appendix 2: SiiSa 4:24.1-10, 21-27

pura lrnscin rnahiipiipi dnrgha\;o niima n~irnatal.1 I
.iatya sfolro rnahakrwldho mahasahasiko '<lhamal.1 11 2
asrutcna rnahamohii<l hrahrnar.1anaq1 sataq1 hatarn I
gaviirri sataq1 hatarr1 tcua ctagdharp gchasataq1 tatki 11 :~
atlm canryaq1 lqtaq1 tcna nariir,1iim aviciiratal.1 I
baliit parastriyo hlmktii balmso <lr~tigocariil.1 11 <\
viip1klipata(.liigiidijalaq1 tcmaiva dli~itam I
varniisramasarniiciiramaryii<lii term bhcditii 11 5
sal~asrajanrnatal.1 purvarp durghatcna rrnuiisvariil,i
sivayogilmrc tcna snvan.1aq1 ni~karn uttarnam 11 G
ciattarn term rnatis tasya Jau1 kalcna sohhana I
aho rr;ohcna piipiini lqtiini subahfmi ca 11 7
rnayii tc~iirp na pasyiirni vinii8asya tu kiirnr.iarn I
briihrnar.1ii vcdavidviiq1so vaclanti narakiin mama 11 8
iti vyiilrnlacittasya durgha~~asya dmiitrnanal.1 I
matipradanarp lqtaviin hriihrnar_lal,l kascid ~istikal.1 11 9
briihmar.1a uviica
sr1madvyaghrnpurarp narna sthanarn asti malfftalc I
yatra nrtyati visvatrna siva1.1 sarrisararnocakal.1 11 10
[... ]
ity cvar~1 briihmar,wnokto clurghat,al.1 pm.1yagamav~it \
8r1rnadvyaghrapmar!l gatva sracldhaya parnya saha 11 21
pradak~ir.mtrayarp lqtvii sn~itva nityam atawlrital,1 I
namo'utaq1 sivamantrm~1 tn .iapitva~\,ottarmp satam 11 22
sr1rrm<ldahhrnsabhamadhyc pranrtyantarn urnapatim 1
dr0\,vii hlnimau mabiibhaktyii dm.i<.lavat prm.ti patya ca I\ 2:)
rndrahhaktiiya vipriiya hhasmanoddhuli tciya ca I
sivajfiiinaikani0thiiya dattvii dh:inyarp dhanatp mwHi 11 211
satyaviik 8aucasa1~1panual.1 kiirnakroclhiidivarjital.1 I
vatsarii1.1ii1p trnym~1 tatra uviistitipriy<~rrn sal.1 I\ 2G
dcvadcvo mahiidcvo mahiikarm.1ikottamal.1 I
~).rasiidm:t aka.rot tasya durghatasya dunitrna11al,1 \I 2G
s1vaprasadcna sa durghatal.1 p1mal.1
samastalokadhipatir hahhfiva I
virnuktirn apy iipa mahattariim imam

parapramatrprathanaikalakf:iar.iam

11

27

2d '~sahasiko 'dliamnl1] Ed.'s ms. 11a.; ''saltasikolLamal.1 Ed.

CoX

Three
Unfuzzying the fuzzy. The distinction
between rasas and bhavas in Bharata
and Abhinavagupta

Daniele Ct1neo
Rasa as a semantic field
raso va'i sa(I,

rasarr1, hy C'UlLya:(n labdhvffnandT bfw,vat'i

Taittir""lya Upani::md 2.7

It is well known among scholars of Sanskrit studies tlrnt. what is usually
called H.asa Theory has been first outlined in Bharata's NlI.tya.~(l,slra. Obviously enough, in the course of time, many clifforcnt a.uthori-l have interpreted, developed and transformed it hy creating, so to say, their own CT.asa
Theory. However, in the history of the scholarly reception of the theory,
there has been a tendency to consider Abhinavagupta's iuterprntatiou expouuclccl iu both the Dhva:nyalokaloco:na, [Locana ad Dhv A], a connnentary on A11anclavardha11a's Dhvanyiiloka [Dhv A] and the Abh:ina:uabh(/,raff
[A 1311], a cornrncutary on 13lrn.rnta's Nrif,yoAil,sf:ra [NS] itself - aH the traushiHtorical embodiment of the whole Hasa Theory as such. Differently, but
with similar results, the historical development of the theory has also been
regarded as a progressive development, a sort of gradual blossoming, culminating in Ahhinavagupta's vcn-;ion, which would represent 'the right inter,fl\

DANIELE CuNJW

50

. ' o f' t l H~ t l H~ory o.r rasas. l ncH. lu, l11Ily

this
prctation' or 'the full rcafo-:at1on
.c
. '
1
is for obvious masons, Abhinavagupta's own stauc<~ too.
.
. ' In my somewhat unoriginal hut rarely explicitly maintaiucd intcrprc~.ation of t.lic issue, rrmch of the coutrovcrsy and fuzziness about. the q11cst.Wll
'what is ro,.sa?' lies in the failure to recognize the obvious, 11amcly, that th('.
idea conveyed by the term rasa is not to be se<~n as a defined concept, bu~. as
a scrrmntic ffold 2 that embraces different. mcaniu;s in diiforcnt thcorct.10.d
1
formulations elaborated by different authors.:
Similarly, some recent publications especially a pioneering article on
I3hojct by Pollock ( 1998) and a ground-breaking hook on courtly cult.me by
Daud Ali (2004) hav<~ explicitly recognized awl nrnlmlincd the hist.orical
character of the H.asa Theory, in particular by highlighting the crncial difference between I3harnta's and Abhiuavagupta's versions of the theory. To
put it briefly, the divergence between the respective theories of Bharat.a and
Abhinavagupta is to he locate(l in the different ways of conceptualizing the
distinction between bha:uas and rasas.
1 Cf. AlJh ad NS fi, prose after :n, vol. 1, p. 277: amna:yasiddhe kim apurva~n et(J,t sar!ividvikrisc 'dldgal<ir;mnitvarn \ ittlwqi svayarrigni/i,yo:malwr-/1,aliehidua.nducna kuri lfo,~aytl<L
na loka?1, \\ 1irdhvonlhuam aruhya yo.d arthatnttvaqi dh1.z1, pa.<yati snintim avedo.yantz l phalarri tad adycl'i~1. par'ikalpilanarri vivek11soJHin1LJHLr1Lm/JllT1LUJLin \\ cilro:q1, n-ira,lamJmna:m eva'.
rnanye prwrne.uasiddliau prat/1,amavato.ra:rn \ sanmaryalrih/1,e so.ti seh1/mndlu1.p1uapr1Lt'l,~f,11,ru'.i,
na vfarnayc!ya \\ tasmat satam atrn na, du~ilani mato.ni I.any cua ht ,forlhitrini \ pw"uapmf.i,~thapito:yojan<Isn rmllaprati,~tfuiph(J,lo,m amananti \ \ My trnuslation nms as follows: "There
should be no wonder in following what is already known in the disclosurn of knowledge
as established by tradition. [This] being so, docs not the world spoil [everything] by
quarreling about sdf-nvidcnt arguments of great valo11r? (No!] Ascending ever higlwr,
the Intellect, unawarn of fat.ig;w), lwholds Uw t,rnt.h, which is nothing bnt the fruit of the
theories eow:cived by th() ancients, along the succession 011 t.lw ladder of discrimination.
vVon(krfnl is, I bdicv(), the first crossing in tlH) asccrtailltnC!lt of the lmowablc, as it. is
completely snpportkss. Dut, as the right road is progressively takmt, no rnor<) a source of
wonder is the foundation of cities, bridges and so forth. Tb()rcfon), the ideas of tlw wise
that hand down tlw frnit of the original fonrnlation arc not rd11tcd Item, 1>111. otlly refined
along the previously established paths." Althonglt given tlw highly polyscmic 11atmc of
these four beautiful stanzas - such a rcndmitlg caimot but lH) 'my' iutcrprntation of the
passage, in the light of my getlcral nndcrstarnling of Abhilmvagupta's thought, within the
wider context of the sastric traditional disc<)tll's<\ Abhi11avagnpta's sdf-rcpwscntat.ion as
the culminating point in the ladder of discrimination is IH)yolld any shadow of don lit. For
diff1;rcut interpretations of tlw stanzas, S()() Gnoli ( ln<iK, !) 1-2) and Ka vi rnj (2oo:i, 127)
2
lt might lw cousidnrcd safe t.o rnai11tnin that. 1.IH) t.mm m..m in tile artistic dornniu
indicat<~S the 'aesthdic experience' - ill particular, ill 1.hc form of the crnotio11a] colltc!lt
of the work of art -, lmt this would just b<) a vag1w awl 1mdnrsp<)cifkd translation that
entailH no adual ddinitiou or cxplauation of the couccpt.
:i In Pollock's words ( l!)DK, 122), "Not evcryouu s<~<~ms to hav<~ he<)tl t.alkiug about the
samn thing when they talked about rasa; their focal points and emphases diaugcd over
time."

lJNFUZZYTNG TllE FUZZY

51

My present aim is, 011 the one hand, to delineate

and to improve
the coge11cy or this very useful and comparatively dear-cut theoretical
difforcncc and, on the other band, as a brief dcconstructive corollary, to
partially clmlleugc this neat and dean division by pointing at the cases
that threaten its broad-spectrum adequacy, i.e. its capacity to account for
all the details of the texts of Dharata and Ahhinavagupta. Thus, in partial
disagreement with the explicit promises of my title, the tentative unful':'.l:ying
or tlw hIY.l':Y will be followed by a re-fnzzying of tlw unfu'.l:zicd.
As it is often customary in scholarly practice, a semantic analysis of the
terms to be dealt with can prove useful to pave the way for the necessarily
interpretive appraisal of the theories in which concepts and meanings are
constructed and negotiated.
In his introduction to the translation of Ananclavardhana's Dhva:nyii,loka
along with Abhinavagupta's Locana, Ingalls states (1990, 15) "The word
rasa in its most literal sense means juice, taste, flavour." According to the
Monier-Williams, the term has the following meanings: "the sap or juice of
plants, juice of fruit, the best or finest or prime part of anything, essence,
marrow, water, liquor, drink, milk, a constituent fluid of the human body,
mercury, quicksilver, taste, flavour, pleasure, delight, the taste or character
of a work, the feeling or sentiment prevailing in it, condiment, sauce, spice,
seasoning." 11
The meanings recorded in the Monier-Williarrrn already include an aesthetic sense, expressed in very general terms as "the taste or character of
a work, the feeling or sentiment prevailing in it." However, it is reasonable
to surmise that the, so to say, 'original' meaning of the term included such
concrete elements as the sap or juice of plants ( P.flu:n,zcns(~f't is the first
meaning given by Mayrhofor [rn8G-2001, snb voce], who adds that the term
might be etymologically connected with the Latin term ros 'dew'). Then,
the concrete grew into the abntract and rnsa became the taste or the flavour
or a given fluid, or even or an elaborate beverage or food; besides that, then,
it. ended up meaning also the plea:mre acquired from savouring those hcvcragcn and foods. Hence, the idea of 'extract,', 'best par(, of' ancl 'essence'
t\ifonier-Williarns (I DGG, .rnb vocc). Cf. also fJ(ihtliugk; Roth ( 18GG-187G, snb vou~).
Apto (lDfiG, 8'1/,b vocc) givos tlw following meanings: "l. sap (of trees); 2. a liquid; 3.
water; 4. liquor; G. a draught; G. tastn (fig. also); 7. a sauce; 8. au object of taste; !). taste
or i11cli11atio11 for a thing; 10. lovo; 11. pleasme; 12. charm; 1:3. pathos; 14. (in poetic:
compositious) a sentiment; l!i. essorn:e; Hi. a constitnout fluid of the body; 17. semen
virik; 18. Mercnry; rn. a poison; 20. any mineral metallic salt; 21. juice of the sugarcane; 22. milk; 2:1. melted butter; 24. nectar; 2G. soup; 2G. a symbolical expression for
tlw munhcr 'six'; 27. green onion; 28. myrrh; 29. gold; :m. a metal in a state of fnsiou;
:n. sec ro.sritala/l,; :12. tho tongue (as the organ of taste); :1:1. (with Vai~1.1avas) disposition
of the heart or mind (the five rasr1'> arc .<fo:nti, du,iiyo., sri/;;hyo., vatso.lya and 'miidh"wrya)."
1

'

I )ANIELE

52

CuNI;O

of something. Thc~n, in a fairly plausible iuterprdation of the famous passage of 'I'a'ilUrTya Upo:n:i!jrul 2. 7, quoted in e:1:r:r:qo, rasa becomes the essence
of everything, the absolute to be acquired with difficulty and plcasmably
. ,
savoured by the seeker of the truth."
To tip my hand, at least partially, my contention is that two 'opposite
meanings of the term ra,.c;d; have been chosen hy Bharata and Abhiuavag;npta to flesh out their idea of aesthetic experience into a viable concept for
their res1)cctive aesthetic theories. On the one hand, Bharata uses the tcnn
rasa having especially in mind the idea of 'taste' or even better 'flavour', in
the sense of a complex and manifold taste, which - though having a single
identity (for instance, the taste 'sweet') - is made up of many cli{forent clements, like all the ingredients in the case of an c~lahornted kind of pastry or
pie. 7 Hence, the aesthetic experience is seen as a full-blown and full-Hedged
experience of relishing that is the result of an expert blending of the various
elements of ordinary experience. On the other hand, Abhinavagupta uses
the term rnsa having especially in rnincl the idea of 'extract' or 'essence', in
the sense that the aesthetic expcricmce is, in other words, the sublimated
counterpart of ordinary experience. The functioning of art encompasses a
sifting of reality that brings about the separation of the wheat of a purified
and intrinsically blissful emotional experience from the chaff of attachment,
egotism and practical interests.

Bharata and the 'Ancients'

rw h'i '('(/,!;(i,d

rte /;;a.foid arlhab. prava.rtat.c

Dharata, N(/,f;yu,.~a;st:m. G, prose aft.er :H

"For a very inten~sting smvey of tlw meanings of tlw word msa. in Pa.Ii texts (rnaiuly,
"aecomplishment" and "joy"), and a tcntativ(~ and not always convincing rcco11stn1ctioll
of the inilnence of these meanings on tlw acst.lwtic sm1sc of the tmrn, sec \Vanier (I D8081).
.
';The idea of rasa as a semantic field
interprded i11 two radically opposite ways
hy diffornnt thinkers was triggernd hy Dmrida's (~ssay "Plat.o's Phannacy" aJHl by his
thought-provoking int(~rprdation of the polysernic natmc of the Greek t.mm p/1.11.rmoJ.:on.
7
Howev(~r, one might arg1w, this partic11lar kind of s111t1 is not just the aritlundical sum of different entities, bnt. a new entity which is crc~at.c(l by tlin accnrnt.c aud
sophisticated blend of disparate cntiti(~s.

lJNFUZZYINC TTTE FUZZY

From many cxprcssio11s in the N(i.tyaa:stra, such as the one just quoted
'in c:wT:qo, 8 it is clear that, according to Blmrnta,n the purpose of a theatrical pi(~cc is to rcpreseut a si.tuatiou that is imbued with a rasa, which
for the time being we might call 'an emotive mood'. All the cornponc11ts of
a theatrical representation must thcrc[orc be thought and structured with
the only aim of 'producing' the rasa. According to NS G.15 there arc eight
of these emotive foci, these flavours, rasas, viy;., the erotic (,spiqcTra), the
comic (hiisya), the pathetic (kar"Ui(l,a), the furious (niwlra), the heroic (vfra),
the fearsome ( bhaylfriaka), the loathsome ( bzbhatsa) and the wondrous (adbh'ILta) .10 These are obviously correlated to ordinary human emotions, the
stl1,c/,y'ibluivas, literally 'Stable States', which arc named and enumerated in
NS G.17. 11 They are, respectively, delight (rnt'i), laughter (hiisa), sorrow
(.fo/;;a), anger (krodha), valour (uts(iha), fear (bhaya), disgust (fuq'IL]JS(i) and
8

A tentatively unmarked trnuslation might be: "Verily, without rasa, no thing can
progress." However, it is worth remarking that Abhinavagupta (ABh ad NS G, prose after
:n, vol. 1, p. 270) gives three equally possible interpretations of the passage according
to three renderings of the term artha, respectively, in relation to the experience of the
critics, the actors or the spectators. For a tra11slation of Abhinavagupta's passage, sec
M.M. Shanna (1980-81).
0
The meaningfulness and the cogency of the followi11g discoun;e dcpc11ds, at least
partially, on my view concerning the authornhip of Dharata's text. Although different
materials have been recognizably put together to form the bulk of the Nc"if,yaifostrn, it is
my opinion that it cannot be just the result of the 'blind' accumulation of the works of
several authors over a long period of time, because, in Tieken's words (l908, 172), "[t]he
NcI.tya.sltstra is a well-planned, coherent work. Its thirty-seven chapters can roughly he
divided into several distinct groups dealing with more or less closely related topics."
This view of unitary authorship goes hack, in particular, to Vatsyayan ( l99G, G) who
argues that "the work reflects a unity of purpose and that it was the product of a
single integrated vision, pm-haps also of a single author." For a sustained argumentation
in favour of the opposite view
the work being a11 almost 'i11cohere11t' smorgasbord
of disparate materials , sec Srinivasan (1980). Pollock (1998) seems to sham, at least
partially, the latter view. For some very useful and extrcrndy co11vi11ci11g i11sights on the
concepts of 'authorship' aml 'agency', both in the restricted field of South Asian studies
and in tlw broader arena of history and ltmnauitics, sec Sq11arcini (2008, 18!'i-217).
10
Ahhinavagnpta follows a probably more recent version of the text that emnnerntcs
nine nisas by a.clcling the reposeful ( .~iintn). A discussion on the place and status of this
ninth ra..rn, a much debated awl often-treated topic, lies beyond the scope of the presc:nt
article. For a trcatrlJ(~llt thereof, sec Raghavan (1DG7, 101-114 aud 197-199), Masson;
Patwarclha.n ( U)()9) Dhatt.a.carya, K. (1!l72), Gerow; Aklujlrnr (1072), Dhattacharya, S.P.
(1D7G) awl Gerow ( lD!M).
11
The tnrm bha:ua, in this context usually translated as 'emotion', derives from the root
bh:ii-, 'to lie', awl can rnfcr to such concepts as state of being, condition or clisposition of
any kind. As sumrnarizccl in Ali (2004, 18G), "[t]he general cousc11s11s of both philosophical
and aesthetic theory was that bluivas arose within the 'mind' or manas, an internal
:organ' (l;:o:m:u,a) wliosc: function was discriminatory, constructive or analytic ( sriifi.kalpa)
m relation to sm1sc faculties ('indriyas) ."

. ( J L"rri u:yo) In ord<'l" to properly gauge the import of the dis ti ncLioll
won< ler .'1 1,., , ,
'

.
_
,
, techhctwccn hhii:uas arnl rasas, it 1s necessary to he acquamted with some , '..'
" l < <" i>nlary. 0 [ dramaturgy with regard to the basic clements, so to s,iy,
ll!Cc1 V ) ,<1 I
for 'constructing' a rasa.
. - .:
As the renowned n1..'!as.,-ilrn has established: vbluivli:n:ubhava:uyabhu:o:r /,sm[l:!Jogrid rasan-i.~fHLlt-i~1,, namely, "The rasa is pr0<l~1ccd by the m,ii,on o~ tl'.c
Determinants (v'ibhiiuas), the Conscq U<~nts ( o:n:11,blw:uas) and the I ranst t,~nY
12
States ('uyahh:idiT'iblui:uo.s) '.' The :ubhr1:uas (" Dcten~1inants") arc those .fa~;;
tors that make the emotion possible, that dctcrrrnne, or even cause it.
Thus, they am hoth the subject and the object of the emotion, as well as
the whole array of stimulating 'environmental' factors that determine the
rise of the emotion, namely, the whole emotional situation. H For instance,
in a 'love f:lituation', the lover and the beloved arc the subject and the object of the emotion, while the stir:mlating factors are springtime, garlands,
splendid mansions, and so forth.1:i The anubhiivo.s ("Consequents") arc the
consequences, the effects or, one might say, the 'symptoms' of an emotion,
namely, in the case of love, both voluntary acts, such as verbal expressions
of one's feelings, sidelong glances and the like and involuntary responses,
such as perspiration, horripilation and 80 forth. H; Obviously enough, these
actf:l are the very object of representation on the part of the actors. The
vyabh-ic{ir'ibhiivas ("Transitory States") arc a whole set of thirty-thrc<~ cornplementary, or f:lccondary, 17 emotions such as anxiety, envy, shame and
12

NS

6, prose after :n, vol. 1, p. 271.

i:Jln particular, at least in Abhinavagupta's very plausible interpretation, the fnuctioll
'.Jf v'ibl.1,11;1'.a~ is to awaken the pre-given dispositior~s (va.wmas, 8ar11.8/,:aras) accmnulatcd
man mfimtc mirnbcr of former births that form, so to say, the emotional DNA of allY
lmman being.

HAecon l'mg to a terminology latnr than Blmrnta's t<)Xt (sc<), for instance, /Ju..for11.p11J.;o.
2
4. ), ,these two sorts of vih!w:uas arc called rcspcct.ivcly 11.lnmbanavihlw.vo. ("Determinallt
qua Support") m_1d wldtpanavibluwo. ("Dckrrninant <pm Stimulant"). To nsc T.S. Eliot's
famous fornmlat10n, the vibluivns can h<~ cousid()l"Cd the "objective correlativ<~s" of all
emotion.
!GS
f .
l
cc, or mstanc<~, t H) ddinitiou of .fr'fil)ll.TH.ra 811, in NS
W8-2!JD.
. .

(j

'

r>ros<' trt <'r

.. ' . .

11

.'

vol 1 P
. . '

..
l nntary rcsp<rns<~s' is fixed. TlH)y am called .wi/.t,vU.:a.11/wuas
_m 1\ >c_r 0 f tl WS<) mvo
and._l,istcd m NS .<>.2~: Stupefaction (slo.mbhll.), Perspiration (soda), IIorripilat.ion (rornanc~L), Bn:ak of Vmcc (svan1.bheda), Tremor (vq)(/.th'l/,), Chang<) of Colour ( oa,ivarUffa),
vVccpmg ( as-rn) a1Hl Fai11ti11g ( pmln1Jo.).
.
17 0
tl 1
., 1
.
.
Hi Tlw
.

n1

. . . n ~ possi n ity to apply contemporary psychological theories t.o Bliarnta's clas. ' (
'0
sihcat1on of stabl<>' (or 'r>rinnry') , l 't .
( 00D).

' ctnr ,tansi 1.ory or secondary') mnot.ions, sec ,unc
2

lJNFUZZYINC Tim FUZzW

55

inclignation. 18 [foucc, the combination of all these elements on the stage

(letermines the 'production' of rasa.
'J'lie difference between bhJivas and rasas is not 'explicitly' stated iu the
Ni'i_lya,fostra. H.athcr, Bharata treats them in two separate chaptcn;, dearly
presupposi11g and implying a distinction between the two sets of items.
The bulk of the sixth chapter of the Nii.f;yaM,stra consists in the definitions of the eight rnsas. Such definitions, in turn, consist in stating the
s/J1iiyibluiva ('Stable State') to which every rasa is rdated and the Determinants, the Consequents and the Transitory States that pertain to it. The
hulk of the seventh chapter of the Naf;yasristra, on the other hand, consists
in the definition of the eight Stahle States and the thirty-three Transitory
States. Such definitions merely consist in stating the Determinants ancl the
Consequents that pertain to them. By cross-checking the Determinants and
the Consequents of the various rasas and their respective bhrivas, one finds
out that they are often very similar and sometimes identical, the list for the
rasa.s being usually longer and more detailed. HJ
8
L Thoy

arc listed in NS G.18-21. Ou a closer inspection, tho list is extremely heterogeneous, including what we would commonsensically call 'emotions', such as envy ( as{i,ya)
and shame (vrt{hi), as well as 'mental states', such as remembrance (sm,rti) and preoccupation ( cinfli), and 'physical or physiological conditions', such as sickness ( vylidhi) and
death (m,am~ia). On the wide semantic field covered by the word bluiva, see Ali (2004,
180-188). In this respect, on the untenability of a waterproof divide between emotional
mid cognitive phenomena both in the context of Ancient India and in general tenns, see
again Cuneo (2009).
l!lSce tho translation of the two chapters in Ghosh (19G0-19G7, Vol. 1. Translation 108-117 and 122-12G). For instance, take the definition of ka:r"u:w1.rasa ("Pathetic
msa") and of its Stable State .foka ("sorrow") according to the text and the tra11slation by Ghosh (Vol. L Trnnslation, 112-U:l and 12:1; Vol. l. Text, 87 and 94):
"Now the Pathetic (l;:aru:u.a) Sentiment [i.e. rasa] arises from the Durable Psydwlogical State [i.e. Stahle State] of sorrow. It grows from Detenninants such as affliction
11nd<~r a curse, separation from dear ones, loss of wealth, death, captivity, flight accidc11ts or any other misfortune. This is to be represented on the stage by means
of Consequents such as she<kling tears, lamentation, dryuess of the mouth, change
of colour, drooping limbs, heing out of breath, loss of memory and the like. Complinwntary [sic!] Psychological States [i.e. Transitory States] councctecl with it arc iuclifforence, languor, anxidy, yearning, excitement, delusion, fainting, sadness, <k~j<ectiou,
illness, inactivity, insanity, epilepsy, foar, indolence, death, paralysis, tremor, change
of colol!r, weeping, loss of voice and the lik<! ( atha karu:(io nil:rna .fokastlui.yibhavaprab/1.11.v11.(1,. sa 1:11, 1i1LJJ11,/;;le.<o:uinipa.lc,sf.a.fan11,vi1Jrayogo:uibh11:va1i11Aavad/1.aband/1.avidravopag/ui;l1J,vya.sa'11.11.s11.1riyog11.1libliir vib/1.1iv11i(1, 81L'/n:11.p11day1Ltc. fo,sya cri,.fr11,pataparidcvano.n1:1J.k/io..<o~11.(1.a1111ivm"(l.yo.s'T'o.s/.11,g11,/,ratar1:i.fo1isosrnrtUop1idib/1,ir an:1Lbhll:u11:ir abl1:in11ya(1, pro.yoktavya(i,.
vyab/1:icriri~1.11..< casya r1:i.,vr:daglrl>n:icintants11,kyavcg11:mo/1,ar11:1nabliayavi,s1ido,d11:inyavy1idh:i.ia(la, /,0Ji'!na1lo:pasrn1irafras dlasyarna Tlli(l.as la'Tnb lw:vcpa t11:1J.vaivm-u, y1!rusvarn. b/1. erlo:1la ya(l, [N 8
G, prose after Gl]) ." "Sorrow ( .foka) is caused by Determinants such as death of the
bdov<~d, loss of wealth, experience of sorrow clue to one's murder or captivity and

d1apt<.~r _o~ th~~

Even from a snpcdicial rea<ling of the sixth awl seventh

-l., l ,:;l- ~tra it is <'nitc evident that 'tribhiivas, an'n/Jliii,vas awl vya/Jl1"1,W"'."1,/Jha\
(J, / !J /,,, "' ,
)
1
.
.
.
.
.
1 ' . ' ' 1(
N
'uas are nothing hut the thcatncal connterparts ot real-lite causes, cf ccts. ctl ..
concomitant factors of an emotion. Dy piecing together the elements ol LhI~'
jigsaw, it might he puy;y;ling to discov<~r tl~at ra.sas and /Jl'.,(/,vas :;~c!n Lo h ,
caused by the same causes and to dctennmc the sanw dtccts. l lns wonl<\
imply that the only difference between rasas arnl /JliJivas is the presence ot
the Transitory States, i .c. of the array of secondary emotions.
A further clue on the specific nature of Ta.'ias is offered hy Dharata
himself when, just after the 'f'a,,sasfitra, he gives the paradigmatic example
( dr?tii,nta) in order to illustrate how Tasas come into being. The text reads:

What is a ~possihle1 c~xarnple? So he says in this regard: As the

rasa, (taste) arises out of the combination of various conclimcnts,
herbs and other substances, in the same way the rasa (aesthetic
experience) arises out of the corning together of various States.
To illustrate: just as the rasas (tastes or juices) such as the
.~a4<.wa, ct<:. are produced by substances, such as jaggcry, etc.,
by condiments and hcrhs, in the same way the Stable States,
brought to the fore by the various other States, achieve the
state of rasa. In this connection, he says: what is the sense of tlw
term rasa? lt is replied: lrasa is so called] as it can he savoured.
How is rasa savoured? .Just as competent persons, while eating a
dish prepared with various kirnls of spices, savour the tastes a,Jl(l
achieve joy, etc., in the same way, compdlmt spectators savour
tl~c Stahk States, when they arc rnanif<~stcd by tlw crn1ctmc11ts
ot the various Stat<~s'20 and endowed with wonls, gestures ancl
psychophysical intentness (sa/,lva), 21 and [in this way] achieve
.
')')
.
Joy, etc.-the like ..It is to be represented on the stage by Consr~qw~nts ~mcl1 as shedding tears,
10111, 1oss
, o [' v01cr~,

1oosu1css of hrnhs,

1a 11JI1"'>'
lanwntat10n ' bew'1il1
' 11g. , <!11110'('
, ,, . ril' co
on the. g;rc~nnd,. cryin~, deep breathing, paralysis, insanity, death all<l the like ( .fo/,:o
1

'~i:nn;i t.~t<?ML<LV't,y1:g11"v!JJl1,1w1~n1i.~:Lo11,d~1,11,handl1,11,d11,ld,:l1,a11"11JJlw"u11,rwdihl1,ir

vibl1,11,1111Jr 1t/,f!IJ,1i,1J<L-

,c . as ya.s r<L[!IL ta 0tla7ntap11,r"1,de1nlnvr1"l'IJILl"(l,y11,sv11"rnb /1,1:1l11,sn1,s /,r1,qr 1,/.r11,/,ahl 1,11,r r1,ifHI, /,11,11,11,81LS IJ!l,f l,ll,-

,,.,,llltlakmn' di ta
I1,1Lntsu11,.~t
, '., , 1,11}a(
: 1ntonm,11,dam11hrmwran11,dtl1hir
rt///,'l/,hhrwnir 11J1hinm;11,h pro,, , 1l-iry
yof.:ta,uyaz1, [NS 7, prose afkr 10])."

20 Tl.
l
f
1

1
, ,.
, '.s, pro m l Y rdcrs t<_> the Transitory Stat<)s, even if it is possi bk that, in this case,
m!Jluwus
annl1/wv11.s
nught ' also

'
21 'l'l dIHl
t
,
lw, iudii<l<'rl

111 t,Jrn exprcss1o!l

', 11
1 wva.
w .enns
. ) ,tn1
, l ' psyc Itophys1cal

t"r
. 'word' ('UO<')
. , , ' 'ho rlY' (anga
mtcnLn<~ss'
( sn/.l:ua) rec
to the rnam kmds of
. 1,mu1,s
, 1 " or " rnaimcrs o[ rnprcsrn1tation")
c111p loyc,1[
. abhincwa
, ', ("<'t
. , ictc
by ,~;~rl)'la::tSr;~ 8 '.111 ~1 listed in NS f\.2:\. On sallua, sc<) Ganser (fortlH:omi11g;).
ic ' ,mskut t<'xt rea1ls, Fo
" dr~I
. : :a r1, t,11,I.i. a tr11, Iw, :i1uJ Ii11, Id nu"n11"1ryai/,j11,n11,n.~1i1 LI1,// /,,1ti-

lJNFlJZZYING TITE FlJZZY

57

Thcrdorc, the 'prodnction' (n'i9palt'i) of rasa iu a theatrical n~presen

tation is similar to the production of a particular taste in an elaborate
foc)(l; and the ways of appreciating them are similar too. This i8 tantamount to 8ayiug Ll1a(; rasa is the result of a well-disposed combination of
various clemcnts/iugrcdicnts. In the case of rasa, these ingredients are the
Determinants, the C011scqnents, the Transitory Statc8 and, argnably, 2:1 the
respective Stahle State.
As their definitions arc laid out in the same way as the definitions of
rasas, namely, in the terms of causal relations with specific v'ibhlivas and
an:ubhlLVas, the Stahle States and the Transitory States can also he understood in the light of the food-simile. Consequently, in their case, the
ingredients that literally constitute them arc nothing but the Determinants
and Consequents only. Therefore, along the lines of the food-simile propounded by Bharnta, the bhavas me only a less elaborate kind of food than
the rasas, i.e. a meal prepared with fewer ingreclients. 21
One further passage from Bharata's text is, in my opinion, very telling
with regard to the relation and the distinction between msas and /Jhavas:
Just as a king surrounded by numerous attendants receives this
epithet and not any other man he he ever so great, so the
vyasarriyogad rasarl'i\<ipatti(i, tathc/, nc/,nablu!vopagmnl!d rasan,i:;;patU(i,, yathiJ hi .<J'll,flridihfrir
dmvyair vyar~janair au:;;lulhibhis ca ~'li\lavfidayo rnsa tathii nandbhiivopagatli, (},pi sthiiyino bhavd 'T'(J,Satvam 1i,pn'UVl1'f/J,fti. atn!lia rasa iti ka(i, padiir'tha(i,? w:yatc - lisviidyatvi'it.
katharn risvcidyatc rnso,(1,? yathii, hi nan1ivyarWanasmr1,skrtam annarp, bhu'i'f,jc/,ria rnscin iisvadayanl;i S'ZLm,anasa(i 1mr'11,\scl, har\'iiidnris cci,dfrigacclianti tat/w: niiniibliavll:b/iinayavyafijillin
vii,g(J,'Ti,gasattvopctiin st/io:yibluivlin (/,sva:dayant'i S'Um,anasa(1, prck\saka(1, lwr:;;ridtr[LS cad/1,,iga-

cc/uinti.
'.n As the Stable State is absent from the nisas1Ztra, it is not easy to assess whether
Bharata considmcd it as an clement pre:-mnt in the production of rnsa - as, for instance,
Bhaq;a Lollatt will arg11<~ or t1ot as will he argued, for instance, by Sr! Sai1kuka.
liowovm, in my opinion, 011 tho hasis of the food-simile a.c.; well as the simile with tlw
king; and its attendants (NS 7, prose after 7), it is highly probable that it is the Stahle
Stat<) that, iu Gharnta's theory, i8 to he considered as the vcry base for rnsa, just as
rice is the snhstrntum ancl the base of any meal in Indian cuisine, although it is the
combination of rice with the other sidc-diRlics that rnakc8 it an 'actual' meal. Blmtta
Lollata awl .Sri Sai1lrnka arc two counncntatorn of the Niif,ya,fostra whose works are 1 ~~it
()Xtant lmt who arc often quoted and referred to by Abhinavagnpta, e8pccially in the long;
comn1c11tary on the rasas1itni. For a translation of the commentary on the rasas'iitrn, sec
Gnoli ( 1!lG8). On Dhat,ta Lolla,;a, sec also Prablmkara Sastry (19G5-GG) ancl Kamimura
( l98G).
21
' This might be iutnrprcted as postulating only a quantitative difference between rnsas
and bhiivas, awl not a qualitative 011e. However, as already hinted at, one might arg;ne
that the actual tast<) of a uwal changes complctdy, if the cook uses fower ingredients
ancl that this change is not merely a quantitative clmng;e, but a qualitative one.
'

. .. i l' l)sycholcwical States li.c. th'. Stable Statc~sJ only fol,_,

,
r
. . PsyD u1 a1) c . . ,
lowcd hy Detcnninants, Co11scquc11ts and Comp nn<~ll 1,d,t y. ,' .
1.
:. l ut,i<'s r (' Trnnsitory Sta\.<~sj n:cciv<~ \.hn 1wmc ol S('.Ilcl l010g,lCd, J <.l l 1 ,.

2
tirncnt li.e. nLsa] .' "
. nothmg;
. . l mt. tic
l ' ct.
i l ' u\. t ' t l 1\ has obt<tincd
'l'l , . 'f< re the n1so. is
,, ,,w
c ,J ,,1 .c , 1c ,
[
lCIC ) . , ,
l l
t Ul('llt 0
, t,t\lS of ra..so. ' insofo.r as it is accompanied. hy. the w io c .assOl
tl,1es.u..
\l t\y tl,l,('
Determinants, Consequents and Transitory States. To put
il
.l
1
11 .
t I Slll)('t-1> /,(LU '
, . a is 1nore than a. /Jllii:va is w1c er t mn a J l,(L'Ua, is a sor , o

ras, .
,
'.
.
.
[
. , tl ,.,(ore all
a complete and widc-rnng;mg mnot10m1l cxperwucc. ?.asa, is, , iuc
'
intemiificd and licightened emotion.
"'1lThc three 'dues' l have, here, gleaned from Bharata's text the pa:' ,r

1 c arn l the
lelism in the ddnnt1ons
oi:r rasas anc l bl uwas,
t l ic 1'oo< l-sum
'. \Glln., f
simile point to the interpretation propounded by thos<) commentators <~
.
.
' ,-')<> namelY that
ni.~il
the Niitya.~rl.stra. that Ah1nnavag;upta
ca11st l\C ' ancients

_ ,
is notl~ing hut the slhiiyil1h(iva intcnsiiic(l by the iril>hJl:vos, the annlJlw:vab
and the vycibli:icii:T"ibhii:uas. 27

l'. ) llll. ,' -,

(N~!

2 GTrnnslation

by Ghosh ( 19GO- U)G7, Vol. 1. Translation, 122). The Sanskrit text

7, prose after 7) reads: ua.th<i no.rcnclro IJ11.lmjannparivaro 'pi s<i cva nmrw. lab}wtc rHin)J<L .
.sumnlwn api ]1'llT"/l.~a(1, l.nth<i vil1hnvrirmblw.vnuyo.l1hic<i1ipariurt<1.(1..~t11.a,1p. b/1.n1Jo rns11.n.nrria
lnbhatc.

it is intcn)sting to note hen) that, in an isolatc<l ancl urnkvc1opcd rcnntr:c~

Bharnta already seems to mention awl hiri;11light a foaturc of the aesthetic experience tJi,t
will come to the fore in Blmt\a Niiyalm's and Abhinavagnpta's t.lH)ori()S, i.e. tlw conccP~.
;~~,generality '.H 1.mivernality: In yart.icuh~r, Bh;:rnt;~'s t<)Xt (N~ 7, pros<) ;i:tcr ~i) ~c<~c1:j
I. he rn.sm; ansc from them [t.e. from the iorty-nmc State:-;: the Stabk, the I.rnns1\.otY ,tl ,
8
the Psychophysical 01wsj wltnn tlwy arc imhtw<l with t.h<'. qnnlity of mii vcrsality ( d1h!JiL
m srmwnyagui1.ayoycna T<L8<L nL~p<tdyanl.e) ." This stray r<)mark might be rnganlcd to \JC
enough to rea<l back into Bharata's theory Bha\1:a Niiyalrn's conception of swlhu.ra(ltA:W
Hl'J;L<L, "1mivcrnalizatiou" (SC('. hdow). Ilowrwcr, in my opinion, th<~ cvidcnc<) is ddiuitdY
too :-;emit to tell.
:ncf. Lollata':-; ~;tatmrn'.nt in NS n, prose aftnr :\l, vol. L, p. 271: tr:nn sUwyy r:oii oibhu:u<i,rmbh<iviid'i/1hir upacito Tasa(i. Dm.HJin is also to IH! r<)dmrn)<l arnong tile ancicJl\.S
that maintain snch a view, as K1ivu11.rlo:r.fo 2. 28 t states "Ddiv;ht liccom<)H .spi.yu:rn (Love)
thanks to the union with a nmltiplicity of dnuwnts" ( ral.i(1..~r1i.yu:rn.l.aq 1, y(l,/,n r11.p11,/1(1./!ILlyayogena) and Krivwulnr.fo 2.28:~ states "Ifavillg rcachc<l 81H:h an PXtrcm<'. pitch, anger
hccomcs rnu,clrn (Ragn)" (u.d.hirnhun [var. kct.: ilu ar11./1.ynj fH1.rnrr1. /,:otiq1, f,:opo rnn<LraJtT1 1'.forri gatazi). As shown by Ingalls ( i!)!lO, l8), th<) idPa that rnsn is iiothiug hut a bn:-ac
emotion that has been heightened is :-;hared alsq by J\11nrnlavanlhallct in his J)hvan!]O.lof,:<L,
the I~tost inHncntial text in the history of Samkrit literary criticism. However, even alter
A~)lunavagupta's (or, hdter, Bhatt.a Niiyaka's) rnvolnt.i01mry turn in t.hn 1m<lcrntan<li11 r,
of rnsa, then) l'f)maincd authors such as Indnr}ija wh<l "cousidcrn<l a m 8 a, t.o he sin1plY
the basic emotion which has undergone 8tn:ngthening ( Prntih}\rcnilnrlLja on lJ<lbh;1.\ 1':0
26 Howcver,

Kii:uy11.lrup.kara8u:nis11<ry1.,qmlw 't.:~-'t) ."

lJ NFUZZYTNC

Tim FUZZY

59

With respect to such a version of the H.asa Theory, 28 however, Ahhinavagupta's re-interpretation will prove to he more successful, probably
---------

28

----

In his seminal arLick, Pollock ( 1998) relates this version of tlie Rasa Theory - ro,,sas

as heightm1ed hho:uas with a specific interpretation of another important issue concerning; the 'l'ILoa, i.e. its substratum, its locns. Specifically, he mgnc:c; for a very straightforward answer as regards the view of the 'aucicnts': rasa inheres in the literary character
only. Pollock's investigation, in particular, is centred around what he cmrniders as the
culmination of the :wsthutic tradition of the 'ancients', i.e. Bhoja's Spi,,qci,1aprakasa, "tlw
most comprehensive arnl snsLained literary aualysis in prcmodcrn India (Pollock l!J!J8,
117)." In I3hoja's theory, hm1ce, rasa is an extremely vivid mnotional experience enjoyed
by the character. Iu Pollock's wordR (1998, 129), the rasa, "insofar as it pre-exists in
the character", is "manifested (vyar~j'ita) by the affective components of literary communication (the fomal<) love-object, the drncription of the scene, etc.); the emotional
response of the rnadcr is outside this causal process and ignored by it." Even if Pollock
admits that rasa is "inherently a quasi intcrsubjectivc phenomenon", that the angle of
the rnadcr/spectator and that of the character arc not "mutually exclusive pcrnpcctivcs
to adopt" aud that he wishes "to leave this matter open", in my opinion, he miderestimates the importance of the spectator's role and response in I3harnta's theory and he
tends to read back Bhoja's spcculatiorrn into I3harata's text. First of all, it is important
to k<)ep in mind that the Nc/,.tywfristrn deals with, ancl outlines, a method of performance
rather than a philosophical theory of aesthetics, i.e. a practical method to be followed
by the act.ors and the stage director: the main interest and pnrposc of the practitioners could not bnt he the success of the theatrical performance, a succes8 to be gauged
by the appreciation of the audience and nothing else (Sec, for instance, the whole 27th
chapter of the Ncif,yonfostra devoted to s'iddh:i, 'success'). Moreover, the passage on the
food-simile quoted above states clearly that the rasa is so called as it is savoured by the
competent spectators. Pollock's remark (1998, 124) that the passage "docs not di:c;allow
the assumption that for Dharnta the sthayibh<ivas and the rasa.s they prodnce arc located
in the character ( thongh "tasted" lly the audience)" appears to me slightly quibllliug.
As Pollock himself admits, the taistc of a food is present both in the food (tlw clmractcr or the theatrical performance) and in the taster (the Rpectator). Finally, one might
read the verses ab011t Lhe 'ideal spectator', i.e. NS 27.Glb-G:~a: yas hi:~.lcm t'll;;.tim ayciU
.fo/,:e .fokam v,pai/,i ca 11 /,;r1uldha(1, krodhc bha:ye bh:i:tab, sa .~rc\~f,ho,(1, 1nd,;~~o.k:a(I, smrto.(1, I
cvu:rf!. bh11:uann/,;11:m:u,c yo yasmin prav'i.fon nu:ra(l, I\ sa /;alrn vrck:rnJw fi/,eyo .<r11.~w:ir chli:ir
alo,1,n/;:rt,a(l, ("Ile is <:011siclcrccl the best spectator who focls satisfaction when satisfaction
is [portrayed], sorrow when sorrow is [portrayed], anger whnn augm is [portrayed], all(l
!'car when foar iR [portrayed]. ln such a n~-crcation of the [Stable] States, the man who
can pcnctrat<) tlwm is to be knowu as 'spectator' in their respect, as lw is crnlowc<l with
tlwse foatmcs."). They dc)arly imply that the audience had to fed, at least ideally, an
emotional expmicncc identical with the one which was lwing represented. It is hence safe
1.o state that, being; tlw Naf,yo,,~<!s/,rn a mainly dramaturgical treatise for practitioners,
Llw iss11e of tlw loc11s of ra.m was not primarily addressed by Blmrata and, probably, not
<W<)ll ccmccivccl as snc:h. However, the appreciation of rasa on the part of the audience
is at least as important as the proclncticrn of the rasa, say, in the cliarnct.cr or, maybe
better, in/by the theatrical performance. Otherwise, as Ali (2004, 202) puts it within
the co11to11rs of his clisconrsc on conrtly cnltnre, "[u]ltimatcly, tlw distinction between
i.ast<) in tlw food ancl the taste in the month is false, for one could only perceive rnsa in
another if mw posscsse<l the capacity to focl/proclucc it within itself."

GO
'directly' 2 !l entails a central fcaturn of the 'connuonplacc' cXP(_'.ri
,

.
. t Oll
cncc of art, i.e. the aesthctical clistancl:.m that allows also the apprccw. ,l

, .. ll'-'(' t
1)CCc1 . " , 1 '

11

of unplea8 a.nt situation8 and crnoti01rn:

_ iL
Nevertheless, the 8ounclncs:-> of the theory of ra8a as a. heightc11ccl IJli.~1'1! IV
did not depend only on ac:-;thctic grounds, hnt ah.;o on a cultural wor1dvJC\:
2
that irnhucd the society or ancient a.ncl medieval lrnlia.Y" Such a vVdf;a:n.schmi:u:ng was founded on the assumption tlmt the royal court was the,
paradigmatic cultural institution, an institution in which a pivotal rok was
played hy the concept of rasa as the experiential peak of an imutte as well
as cultivated capacity of emotional intensity. Thi8 issue lurn been masterfully dealt with by Ali's pioneering; study (2004) .:i:i In order to flesh out th(_~
idea of n1.sa as an enhanced emotional experience, we rn~ed to hridly outline
some of his arguments here.
On the ha.sis of l1 close analysis and a crcat.i ve interpretation of a huge
quantity of literary and epigraphical data, Ali argues for the cxi8tcncc of <1
courtly culture in which particular emphasis was placed on 8ymholic constructs such as 'hcauty', 'refinement', 'enjoyment' and 'emotionality'. The
ideal 'courtier' was therefore a man of highly refined character, a ma.n cndow~d with da;k.~ririyci, 'courte8 y' or \:ourtly rdincrnent'. This paradigmatic
quality was reflected in a set of behavioural disposition:->, psychical structures and interpersonal relations, epitomized in "an eminent yet grncdnl
~arriagc (~li 2004, 159)." The cornerstone of this peculiar conduct consisted
m a playful nonchalancc,:M a sort of charming and cultivated spontaneity,
t suet
' l <Ul
-~.~.,

. m
.
,~y tl 1",
issue ,
IS ,.1rnhrnctly'
<ka1t with

1
. vcnnon

>. "
t.lm;
o[ the h<1S<
lheory
,., n1t

the comrnot t
.
1iy mdndnw
.
,., tlw, vPry
, . ll oss1
1 .y <h" t lle ' aest 1ict1cal
drntancc'
w1thrn

, , 29 0nc

tniffht
.o

S'

(1.c:iii r~ot tnggcred 1~y art) emotional cxp()ticnc<\.

" l cilst,ancc ' as propournkd m
a famons

lc \l.Y
> ll. Compare the idm
, of 'JlSY
. cl uca
art1c
I >ll ough(l.!H2).
11
:n On
. 1t\.<~nccl cmot.1ou
. lnvltlqrhts
. .
,
'
. . ,is
. , 11c1g;
. the otlicr hand ' thP, vi<'w
, o f' rn.rn.
t.ltc coHllll0
<'XI)('J'lCl}('('
of
1rt
'IS
'Xl
'l
t"
l
'
'
11 ara ,wn arn passionate ahsorption.
,
;,,
,J
' ,
' c
ir nplymri;
. t.l ta t t l te text of the Na/,yafosl:rn was corn pose< I
. 3, 1hcre1iy,
. , . Im
. not. U<'C<'sstrily
, "'.
01 ~omp11ul m ,1 comtly rml-i.e'll..
(.l <)08)
, " to
. I)() mcnl.toncd

. , "" 1The study of

Goodwin

<l,,
cscrvcs
hem, as it was the {trst
n t,tempt
tlw HllIJOl'bll<'<'
,
l
. , . to
. l'CCO)';Ill'/,C
.
.
' , , o f tlle, l icro (nayn /,;a) of Sanskrit.
<lrnma as the idc;1.1t'/.C<
pro1cct1on
, I1!l
.. lo:.11a ("tl
. . appreciator" or "conuoisscnr ")
c ~"
, . of
I> th<~
, ..,parn<lig;nntir
, ' ' , .rn
. w sm1s1ttV<\
Ot, ,J,UlS
(flt
oct.1cs
or
of
the
nar
arnh
("
1
.
,
.

1
1
.,,
<l . .
.
1 " m l,tllc rn,i.n") ol Vat.syaymm's
Kam.o..~uf.n1. W1tO1
rcprcsentc
'
m
their
tnrn
tlw
so('h1
i<l
"
l
1
1
.
.
' 1\\"
0
(20(H).
' ' , ' '
c,t
mnrt. Y cnltnrc as rcconstrnd<\d by

:i1r'or a t.reatnwut of. th<' idet of 'I l 1 f 1

,
.
. ,
,1 . .
, ( .
,
,,
l ' Y l1 ncss , or sport1vm1ess'
and of the vanotl::i
,111s
cnt
mots
for
mstance,
hrullalr
.
_
l
._
,
)
.
.
'
.
.
'
lr-r- lr:')) ()
. l
. '
' am' as 'rw.rl-. th,i.t co1tvcycd it S<'<' Ah (200 0I '
ll- ). ne nu~; it add that this ('ttltm l l ,. l 0 f 1

" ..
\ f
'anthorit ' and 'lon1sh' , . , , , ' '.
. d H Cd
nnrt. tlnl hdiavionr wns t.hn syrnllo
Y
1P ,is sndi, for the symholic construction of 'power' m1tailcd ''ntl

Gl

lJNFUZZYINC TllE FUZZY

not far from Uw idea of sprczzalura as propounded by 13aklassarre Castiglione ( M 78-1529) in fl Cortcgia:rw.
The acsthetical <lisconrnc of the Niifya.siislra as well as the ideal of
ornamentation ( alo:rrikii:ra) ,30 prominent in the hif-ltory or Sanskrit poetics
was a. central keystone of courtly culture, exactly insofar as literatme
and art were implicitly reckoned amongst, on the one hand, the eujoyrnents
that sanctioned the hierarchical superiority of the courtiers and, on the
other hand, the pracLiccs of education, refinement and beautification that
were at the disposal of the courtly institution itself for the sake of its own
reproduction as an "interpretive comrrnmity (Ali 2001, 19) ."
Moreover, Ali argues that the underlying 'rntiouality' of courtly life was
an affective dimension that permeated all relationships at court. Life at
court was therefore constantly suffused with emotionality. If the mark of
courtly life was enjoyment, a crucial badge of distinction for the courtier
was the capacity of emotionally savouring life as such, i.e. the capacity of
experiencing an iutensc emotional lifo. However, this potentially disruptive emotional intensity was harnessed by the cultivated and formali'.l;ed
demeanour that characteri:;i;ed the ideal 'courtier'. The result of the blend
between intense emotionality and controlled bearing as the characteristics
of the ruling 6lite was the ability to experience everything as in a game,
with a sort of playfully involved clctadnnent. The name of such a mode of
savouring life was indeed rasa.:w
Consequently, the significance of Bharnta's Rasa Theory exhibits its
soundness only in the light of Ali's analysis of the courtly ethos. Raso, is
therefore an extremely intense and culturally mannered emotional experim1ce. Enjoying rasa meant enjoying, in the words of Ali (2004, 20:3), a
"once-removed savouring of emotion" that reflected the particular kind of
engagement that the people o[ the court were expected to have with the
world around them. lfouce, one might say that the acstbetical soundness of
ideological cmpliasis" on rdi1wd cnjoynwnts and pleasures as the representative rna.rlrn
of "the comt's image of itself (Ali 2004, 158)."
:v; An assessment of the poetieal tradition of ancient Iudia lies for beyond the scope of
the preseut essay. With regard to the internal logic of the ideal of figuration in Sanskrit
pod.ry and poetics, sec Gerow ( 1971). ln Ali 'fl implicitly foncaltian discourse ( 2004,
20), "Llw practice of a.la.rr1.k<ira, or adornment, functioned both afl a 'technology' of selftrnnsformation aiHl au idiom of cornmtmication." Sec also Ali (2004, 1G2-182).
:IG" ['l'j
l ie assrnnpt1011
.
. court l y c1rcl cs was I;l ia,j rnsa was cxpcncnccc

l J)Y men and

.
m
women of rank not merely in art lmt in their worldly dcalingH - that the capacity of
experiencing raso. was a way of experiencing the affective world armmcl them (Ali 2004
0

1!):\) ."

'

62

.
, of the
, l hy Uharata lies 011 the very we1us10 11
.17
l
n.
"
Theory as proponnc e<

r1d
the clS~ . , , ' . . , iVithin the common emotional cxpcricllC<~ of the WO .
aesthetic <listancc ~

Abhinavagupta

"/1.napra l-1
f
/f"'J(t'f'i!Sah
ro,.'ianiJ,lrnakavlla:my
t :1.gra 1:1~0 ' 1.a:ua c, : ,

Abhinavag;upta, ABh ad NS G, prose alter J '

rasa..'i'tl.l:ro., vol. 1, P 270
n:i?l

t 's aesthetics is charn.ctcri~cd by an actual paradigm sl n .

Ahhinavagup a " .'' ,, '

.
.
. .
, l . , , of
.. l . revolutionary thcorct1ca.l novelty. the cidrnowlu gcrncllt
ernbo(l we m a ,
.
.
.
"
~ clear distinction between aesthetic ancl common-h_fo (one mw.;ht also sd_Y~
pragmatically oriented) emotions, i.e. hctw~xm c~not10mi arons<~<l hy an a.rt
work and emotions aroused hy every-day s1tuat10ns. . .
.
. .
In hricf, ae~theti: c~m.otions (ra..'Jas)_an~ esscnti<~~l~ d1~f~~ret~\~Jron~. on:~IH:t~:::~l
crnotions ( bhauas) msofar as the urnque ontologicc.\l stc:ttus, nut.her re
one might further specify that this aestlwticisation of lnnnan crnoti~ns is
1
t
l l ran
lc w l io l u1cl l )CCil ('Clll('"Lt<'<l

rcstnctc< o men am l women o . l ug

,
' ' , to
such
a 1><trt1c11lar
.
tl
11
.
.
l
l
l
t
l
1
l
<'ttff1g('ll
.
.
way o f cxpencncmg w wor < m a s1mu tancous y c e .ac ic< aw , ,.,, . , manr1<~r ' ns tlus
. l ncrarc
.
l nca
. l snpcnonty
. . .m t l H) cour tly
was tl, lf) very mar lc 01c tl icn
. 11r1"1vl'1u<'
'" ,. F11rt
'hcnnorc,

a fnrnlamcutal part of such cdncational upbringing of the courtiers was nothing hut the
appreciation of dramas in which characters of high rank cxperim1cc<l tlicmsclvcs ro.sa.
:l 7 However,

p
''

See, again, Ali (20M, passim).

..
:i>sThc expression 'paradigm shift' is an obvious rdcrnncc to Kuhn's theory of scicntilic
revolutions (Kulm 1!)70). The rnlcvancc and ackqnacy of Kuhn's theoretical frarncwork
for the historical analysis of almrihara.~<rntra was irnkpm1dm1tly a<lvocatcd hy McCrea
(200~, 1!3-2!)) with respect to Anandavanlhaua's 'rcvolutiona.ry' view of dhuo.ni 'suggestion' as the essence) of poetry (On AuanrlavanHiana's concept of dh:oani, sec t.lic st.ill vcr-Y
11sd11l Krislmarnoorthy 1Dfi8). My contention is that. it is possihh~ t.o argue for the occmrr~ncc of two pmwli;rn shifts in tlw hist<~ry of alnqi/,:a:ra.fo8f,ro.. Tlw iirst one, poiut.c<~.
0
ont m these terms by McCrea (2008), marks the change from th<~ formalist tlwory
poetics propounded hy the earlier authors (Bhiimaha, Dm.H)in, Viirnana, Udblia\.a. nucl
R11<lrata) to the <~ssentialist and functionalist theory of poetics propo1111dcd by Arniu<h:varrlhana. Tho Hccornl paradigm shift marks the diangc from a couccptiou of acst.licti.c
exp<~ricncc (raw) that r1ocs not acco1mt. for t.lw 011\.0\ogka\ rlilfon:ncc bc~twcc11 the 11niV<)rne cxpcricnccrl in ordinary reality a!Hl the nnivcrnc c.rcai.cd by, a11d r~xpcri<~nccd in, art
to a conception of aesthetic experience (-rw;u., again), i.<~. J\ hhi11avagup1.a's view (1norc
pr<~cisdy, already l3lmtta Niiyaka's, sc~c below), that docs acco1mi. for snch a diffcrct1cC
and makes it the cmcial sp<~cnlative ar;mrwut j11sti[yi11g and legitimizing t.lic i11trit1~icallY
pkasurahk, or r)vcn beatific:, natmc of the mnotious aronsr:<l by art.
ii
,
.
I
.. ,For instance, in ABh a<l NS l. l, vol. 1, p. :l, tlic cxp<~rirn1cc o[ 1.bcat.rc is sairl t.o )C
"dissimilar from the worldly cntit.i<~s, and di[forcnt from 1.heir imitation, rdlcctiou, portrayal, ~imilarit.y, s11perirnpositio11, mental apprnhcusion, [arn:y, rlrcam, i\111sion, sorccr.Y
and so lo rtl1 (la uki/;apado.rtl w.vua l'irih f;mri l.o.1l11.ri 1i},<LnlJlra.f,il1i m.IHil<:kl i !J<1.s <i.<lr"<yarnfHl.<111, !/ 1l uw
sayotprcf;:~asvapnarrwucndro.jalo.div'ila/;:.~a~w:rn) ."

lJNFlJZZYlNC TTTE FUZZY

63

nor unreal, of the clmracters am! the situations depicted, for example, in
a theatrical perfonnance110 provokes in the spectator a particular type of
mnotiorml resp011s<~, devoid of any form o[ attac11111c11t (raga) or aversion
( dve,<Ja) with respect to the emotional stirrmli. 111
The speculative pivot on which this account of aesthetic cxpcricucc turns
is the concept of srTdhJi:ra:pJ);;aro:r_w,, which one might translate as 'generali1mtion', 'univcrnali:;,ation' or even 'trnnspcrsorm]i'.1,ation'. 112 To illustrate, the
emotions represented in art arc felt by the connoisseurs as 'gcncraliy;ccl' or
'nnivmsali:;,cd', namely as deprived of any spatial or temporal qualifications,
as well as non-belonging to any specific individual, he it the artist (say, the
poet or the playwright)' the represented character (say, n.ama)' the actori:l
10
Theatre

is the paradigmatic art form iu Abhinavagnpta's aesthetics, hut this reasoning can he theoretically applied, as I am implicitly doing here, to any other art form (or,
at least, to any representational art form). In this respect, Ahhinavagupta, in ADh ad
NS G, prose after 31, rasos11tra, vol. 1, p. 284, quotes two very telling verses by Viimana
(I<avyala'1ri1ii:ros(ilra l.:L:m-:n): "Dest among the literary compositions are the ten types
of drama, as they are multifarious by virtue of the completeness of their characteristics,
as in the case~ of a painting. ( sandarbhc,on dasar1/.pakarr1, .frcya(I.. lad vicitrarri citrapatavad

1!'i,fo,rnsrJ.lrnly<it.)"
111
Interestingly enough, Abhinavagnpta docs not directly mention the terms raga and
dvc,9a in this very respect, nor the term k<irna (desire), which might encompass both.
Rather, he mentions the suppression of all the dimensions of ordinary experience that
determine the insurgence of desire, uamcly the restrictions of time and space, the limitations of causality and the reference of the emotional experience to a specific knowing
subject, i.e. to a specific individual (Sec frdra for references). As for the reason of his
reticence, I intend to explore it in a future publication.
12
As it is well known, the idea of sri,dhJira:~iTkaro.~1.a has been borrowed from DliaHa Niiyaka's ac~sthctic speculations as developed in his lost work, the Hrdayadarpa~w..
For Ahhil1avag11pta's acconnt of Dhaq;a Ni.i.yaka's thongltt, sec, in particular, Locana
ad DhvA 2.4, translated in Ingalls (1990, 219-2:\:l) and ABh ad NS G, prose after :n,
rasasulrn, passim, translated in Gnoli ( J.!)(i8). For a fairly complete, though very personal, interpretation of Uha\,(,a Na.yaka's thought, sec Pollock (2010b). Bhatt-a Nil.yaka's
thought ha:-; most likely iufhwnccd also Dlianai\jaya.'s DasoiT"llfia/,;a and Dhauika's comnwntary ou iL. hi fact, these authors seem to have very similar conceptions as to the
'general' or 'universal' naturn of tlw world created by tlw artwork (sec, for instance,
DoAa'l"ll.]Jaka 11.:\8 and Avaloka thereon) and as to the cmtological distinction between
aesthetic emotions there called krivyaro.sa - and ordinary emotions thPrc called lav./,:ikaro.sa (sec, for instance, Da.faru.paka 4.:19-40 and Ava.loka thereon). In this regard,
S<)e again Pollock (20 lOb).
t:lTlw prcsrn1cc of the actor in snch a list is obviously justified by Abhinavagupta's
<~lreacly quoted iclea that theatre is the paracligmatic art form. ~s for the (im)po:-;sibility
lor the actor to experience rasa, sec, for instance, Abh ad NS G.:l2-:l:l, vol. I, p. 28D:
"Thcrdorn, tlin rasa is not in the actor. [... ] What about the actor, then? [He is] a means
of savouring. [... ]For this very mason, he is called vessel (a.ta eva nafe na rasa{i [... ] natc
tarhi khn'! ri:svada.nopaya,(1. [... ] ata cua ca fHitmrn Ly Hcyatc)." For a thorongh study ;)f
the i:-;sne, sec Cmwo; Ganser (forthcomi11g).
1

Ci1

. tl , . 'ct tor himself. The emotion, so experienced, is Uwrcforc fdt as if

or , lC spc , ,(1 ,
.
.

t )('I'it were undifferentiated or 'trnnspersonal', clcvo1d ot any rdc~rnncc ,o l ' ,

s~nal cliarndcriimtion such as 'l am tlw 011<~ who is reeling sncl1-a11cl-such '
.'he is experiencing this or :,hat_ crno.tion' or should he
such-arH\
. l as this or that mnot10n is hemg cxpm1c11c<~d hy hm1, with whoIIl, I
sue i,
l '1
have a particular relatior18hip, which shoul<l make nw fed sud1-and-s11c i.
Emotions felt in such a position of 'unrclatcdness' ar<), so to say, ckvntcd to
a different plane of reality, removed from the onlinary world of pleasure and
pain and freed from individuation arnl limitation. The fottcrs or the various
'empirical' selves arc temporarily shattcwd: emotions shine, tmconncctecll~,
in their own generalized essence. In Gcrow's words ( 1994, 187), the point is
that "Lw1e experience, in art, not love for X, but love as such."
But how is it that such a process of 'generalization' docs take place at
all? Abhinavagupta argues that in every aesthetic experience there is a sort
of dash between cognitive stances. On the one hand, we do have the deeply
grounded foreknowleflge that what we arc experiencing is unreal, ohvionsly
and intrinsically unreal, as it has been fictionally created hy tlw artwork. On
the other hand, we have the clear, straightforward data o[ our immediate
perccption. 11 G The result of this cognitive incongruence is the generalization
of emotions. In Ahhinavagnpta's own words, the gcncrafo--:atiou occurs "on
the hasis of a complete suppression of limiting causes such as time, spa.cc,
subject and so forth - those really existent as well as those evoked hy poetry
as a result of their reciprocal opposition"'lfi ( vash1,sa[(Lr[i kilV'!JlLT[J'il;rinii:r[L

'~

----------~~-

'14 I

.fcdiu~

's own
----n Al >11111avag11pta
won ls, ABh ad NS, G, prose after :n, rasasntra, vol. 1'
P: ~78, snch an emotion is cxpcrim1ccd by nwans of "a cognition <kvoifl of obstacles,
d1Hcrcnt ~ror'.1 ~ognitions full of ob:-;tadcs sncli ;1s 'I am afrni<l, he my enemy, my friend,
someone ll1<~1Hercnt to me is afraid', as these am bo11rnl by the rise of other ideas snch
as alrn.r'.donuw; [accepting or disregarrlin; aH inditform1t to nwj, <lctcnr1irwd by plcnsurc
<~r:d .1m1~-1. .. ( "bl~~t~J 'hrirri t:hito 'yarr1, .folrnr v1.1.yosuo marlhuo,.<;Uw ua" ity1ulipmJyayel1h!JO
'.'~ik'.ws uUwllz/,! tal ia narlib 1uld/1.yanlar'Oli<L'!}ar1:1:ua.r n<LU< 1. (. l.IL'!J<r, uig /1,nalHi./ 1, 11.lr~IJ /iy o 11il1i/,;~" a( 1.Mf 1'
mru:gl1:navra/,1,t1.fJra/1,-uu:rri)"
_. 'hlt is w01y1 i:cmcrnhcring that, in the most wirlnsprca.d Indian philosophical pcrspcc~,ivc, pen:eption IS rc;ardcd as the first and foremost means or knowkrlgc, the 011c which
18

t~~s be fr!llowed and believer\. See Matilal ( l!)8!i).

L'<'<'TtlL' t Hldl!l
. t,<llll
tl, ia't t lllS
cp1stcmw

lllCOll)'Tl!CllCC IS
()I tlY.
this
. t.ion

'
' ,,.<'tl<'r"l.,
,., , "iz,1
o t cmot1011s,
wl 1mr~a:-; tl1<~ primary
causes d t'<'
~!te :-;p<'.c1h~: q'.I<~liti<~s. possc8sc<l by tfo: 'i<kal com1ois;;r~m'' thr: so.hplrq;o.. A bltinavag11pt'.1's
l 1m_ous '.:\c~1rnt1011 of the concept, a<:cor<ling to Ingalls' tra11sla.Uo11 ( Ul!JO, 70) is the Jol'.~wmg;: I ~w wr~rd .wLlqrlaya (lit., "having tlwir lH~arts with it") dm1otcs p<~rsous who arc
0
c<tpahk
of identifynw
, t ma,tt,er, as tl tr) nmTor

..
" with th<', s11 l>Jee,
o[ their heartH has been P 1ished
t",ice
. , o r poetry, and who n~spond to it. ~ympatlicticn
11".1
. l hy
. the conHtant :-;tudy
'
'u1c\ Pidc
mt
, , I,,avyu:n1lsilanoJ1h-1;aw,v11
,
.
. ' lt //J 1,f,'
1ft
, te1r own
I hearts. (Locarw arl Dhv A 1 l"
yc.~aqi
..,~wl otsnr.
rnanorrm mrc varnan-iyatom
, , I!
, . .

naui
J uiuo,rwuormata
te svahplnyasa1rivada./1hoj11.(1.
sa.hrda,yaJI''

Ahlunavairnr>b
h
( act1nlly
'
(

the scconrlary
. : . raus<'
,

of

'

,(}

lJ NViJZZYINC

ca

G.)h

TTII; FUZZY

de.~af;;dlapramiilrii,d'ir1Jirr1,

n:iyanwJ1,et{i:ndrn anyonyo,pral:ihandliabaUirl alyanta:m apo,.'>aro:(I,(;) . On the one hand, the cnjoyer of art iH cugroHscd in the
sympatlictic contemplation of the emotional focus represented by art. Yet,
on the other hand, he remains aware of the ultimate and built-in unreality
of the imaginary univen.;c created hy the artistic medium. Therefore, rasa
is ultimately an experience cogni~cd within a frame of detachment from
our immediate egotistic interests, though still within a general structure
of i11volvemeut determined by one's personal engagement with the fictional
story.
In order to mark the unbridgeable ontological divide between the status
of ordinary life, in which bhli:uas, real-lifo emotions, are folt, aud that of the
artistic univenic, in which rasas, 'aestlwtic' ernoticms, are felt, Abhinavagupta uses a binary terminological opposition: lankika vs. alaukika, "worldly"
vs. "non-worlclly", or, in equivalent terms, la'Ukika vs. lokottara, "worldly"
vs. "transccndent".'18
To conclude, rnsas are alauJG'ika and enjoyable preci8ely insofar as they
lack reference to spatial a.nd temporal coordinates a.swell as rcforence to a.ny
particular knowing subjcct. 119 In fact, the nonexistence of these characteristics - determined in turn by the special ala'llkika nature of the emotional
117

A further and similar definition of the 'ideal spectator' is given by Ahhinavagupta in

ADh ad NS 1.19-22, p.15: krt{ifiprasf{i,va:vwUopadcsaka v'iga/;aragadve~9lL rruulhyasthavrttaya~i

'TL'i'T"malahrdayatrmk'Urc saf'i tanrnay'Eblw:uanayogyatopeta;

iil1:itarnsasv1idii~i

s<i'mcijikri it'i.

'"Spectators' am those who - beneficiaries of the teachings under the pretext of there being an occasion for an object of diversion - do savour rasa, as they have become deprived
of aversion and attachment, detached in their conduct and endowed with the faculty
of identification, given that tlw mirror of their heart has b<~cn polished." For a definition of sahrdayatva in a different context, sec Pariitri1!18ikavivarm.1il,, p. 202, in Gnoli
( l98G): adhikacatna,fkiinivca cva viryak,~obluitrnii s0Jirdayal1(, 'llCyalc. For a definition of
ahrdaya,, if in a different context, sec Tautriilolm :\.240. On sahrdaya in secondary literature, SC<) Masson (1979), Ilardikar (19\H), Km1jmrni Raja (1997). However, on a dose
inspection, I think that A bhinavagupta's contention is simply that the epistemic incongmm1cc cau only have the desired effects if the connoisseur has cnltivated the sufficim 1t
artistic-c11ltmal snnsihility proper to a sahrrlaya. I-Jenee, this sensibility (saJirdayalva) is
the prcrcqnisite for the appreciation of art, not its actual cause.
17
ABh ad NS G, prosn after :l t, rn,,rns1ttni, vol. 1, p. 278.
18
' '1'he locus class'ic'llS for the discrimination hetwe<)Il laukika and alauldka in Ahhi11avagupta's works in particular in relation with the ala:uJ;;ika natmc of vib//,livas and
anubhavas, which rcpn~s<~nt the very stuff ol' the theatrical pcrforrnm1ce is Locana acl
Dhv A I. 18. For a list of all the passages of Ahhinavagnpta in this respect and a thorough analysis of the idea of alo:nldkafva - also in relation with the Kantian concept of
the autonomy of the aesthetic cxpmicnce -, sec Kulkarni ( l!J8Gb).
1
' uin Abhinavagupta's wonls, in the aesthetic appreciation the emotion becomes rle.fo,/1;rilr!,dyarudiri,y'ita, "not embraced by space, time, etc. [i.e., the knowing subject]" (ABh
ad NS G, prose after :n, rasas1itra, vol. 1, p. 28G).

6()

.
. , , , ted hy the artwork that is kirnlling the rww clctcnnincs
situation rcprcscn ,
" l l
l l ot hN.
f tl (' hc'clonic response, plcasnn~ or pam, .. w uc 1 wou c ' ,
1
an absence o
,
,
.
'l'l . l encc
.
, , mrthlc part of every emotional plw11omc11011.
IC d, >S , .
wise be an msc1 <

, )'till
. .
.l . cc>rrii)()n<'nt causes the absence o( the clcsuc to remove I c
of tlus 1lC( ornc ,
,
.
.
. t"
"l"llI'C' and consequently the absence~ of the unpulsc to c1C
or preserve I)l(,c ,, ,,

.
.
. l')()QC) '30) It is this very mtcrcoruiectc~cl diam ol absences that. ct
lUlCO ~
' '
.
' '
.
. .
.
. (. Il
(e .
i c"'t (i' c nat nr<' of consc1ousncss to slnw~ through the apprccld JO
lows tl re i) ,u ,1 ,
,
,
.
~0
. t the hlissful experience of rasa:'
o ar ,, i.e. 111 '
'

f
To sum up, hence, Ahhinavagupta's very innovative~ interpretation o
the Ilasa Theory implies that rnsas (aesthetic emotions) are, somchow,_lcsS
than bha:vas (common-life emotions), insofar as the former lack some ol the
elements that pertain to the latter, i.e. all the clements that ddenninc tl~c1
inevitably pleasurnhle-cum-painfnl nature of real-life emotional cxistcucc."
Consequently, rasas become a sort of distillation or sublimation of bhii:uas.
The term rasa is therefore understood as meaning 'sap', 'juice' or better
'essence', 'extract' or 'elixir' of blul:ua.
As already stated, the experience of rasa, so conceived, consists of a
simultaneously detached and in vol vcd emotion, as it is aroused through a
sympathetic involvement with the portrayed fictional situation, but cannot
hut he saturated with the psychical detachment detcnninccl by the awareness of the ultimate unreality of the emotional stimuli. Therefore, this special detached-cum-involved quality of the aesthetic emotion is common to
ho th the interpretations of the H.asa theory outlined so far, the 'ancicr;~
paradigm' of I3harata as well as the 'rww paradigm' of Ahliiuavagupta.'>~
However, in the former case, it is conceived to he possible~ even in every-day
li~c as the mark of the cultural arnl <~motional snp<~riority of high class individuals, whereas, in the latter case, it is rcstrictc<l to tli(~ domain of artistic
apprcciation.s:i
"Cf
All
. 1 NS,'
('--'>') ~-;;----~\ '
-'
) l clt

J .i~-.,.,, vo . l., p. 28!): o,.~manrnalc sam:ocdarwm, coano:n<in1Jh.1i11,1i.rTL

1isuarlya.le.

.
1
" For. f nr'ti wr ea
I l >orat10n
. on the lmman emotional sphcrn as composnd ]Jy different

clements or consi:t,in; of different clillwnsions (a cognitiv(: rrnn, a physical otw, ,~volitional

('.r,1<; <~n:l a conat1vc one), sec. Cnneo (200~J). For a thorollgh analysis of t.h<: llil(}cclarccl
0
centrality of the 'element o[ rlcsire' in A hhiuavaonr>t a's wst hr'( 1(. t )\(lll'rlit s<'<' C1u1c

(200B-200!)).

,_,

' ' ' ' ' , ,

,., " ,

G~ As the~ hoth account for tlw Hpcdfir: 1iat11rn of tlic emotional (;()t)(.(~Jlt or Llic ac:-itlwt1c
1111plcasant nest I1c'
,ti<'
,
. cxperwnc<'
,
and.' in 1nrtici1l
'. , ,n, 101 tl, H., poss11 H1ity ol <m.Joymg
1
cn~c~tions --, both vm~1ons ol tlw Rasa Theory can be jHdgcd cqnally appealing frOTII '
merely phenontr'noloirw
'" . "(,1ve,

1.rom tlw
. pr;rspcctivr: of t,h(: sdf-rcprcsc11t.at1
. JI
,
. .
"' <11 I> crnpcc
i.e.
f
1
o' t .te ,Erldm:t8
hwrl
" wl to 11c1,vc
.
. an artwork and clailll to have )l,i.<
1
. by tl tosc
appwcmtcrl
.
<Ln.'.test1wt1c cxpmwnc(:'.
1
; ,. t 1H., d,l
. .t1st,
.., as he has to play the rolr) of a bridge 1>ctwccll

. '' The only .exec Pf wn \>Un~,

cornrnon crnot10ns ancl aPsth(ti.
possess an i11crcdil>ly ( leve'I,,
, c e,1no t",10ns ..II r:ncn, hr: mnst

....

(i7

lJNI'lJZZYINC Tim FUZZY

A Deconstructive Corollary
Mit elem Wissen wiiclrnt clcr Zweifel
Goethe, 8pn'idw in Prosa
In spite o[ what has been argued so far, as far as Bharnta's text is concerned,
the relationship between bhcl,va.s and rasas is uot always homogeneous. In
other words, on close scrutiny, not every rasa :seems to he nothing hut an
intensified or heightened version of its respective bhli,va.
As mentioned above, according to NS G.15 and 17, the eight rnsas and
the rcHpcctivc bhlLVas arc the following: the erotic ( S("Tigli:ra) and delight
( rnti), the eomic (hiisya) and laughter ( hlLsa), the pathetic ( kar"u:(icJ) and
sorrow ( soka), the furious ( rawlrn) and anger ( krodha), the heroic ( vTra) and
valour (ulslLha), the foanmrnc (bhaylfnaka) and fear (bhaya), the loathsome
( bTbha/;sa) and disgust (fugupsa), and, the wondrous ( adbhuta) and wonder
( v'i.wnay a).

If the model of 'intensification' is evidently valid in case8 such as anger

( krodha) intensified in fury ( raudra) or valour ( ntsliha) intensified in heroism
( vi/ra), this explicative pattern docs not seem to work in the case of .piglfm
and rat'i, and even more clearly in the case of kar"U:(W and .foka.
It can be considered safe to maintain that rati means delight, joy, pleasure, even Hexual pleasure. However, the term .srrigiira iH somehow not unambiguous. According to Warder (1980-81, G29), "[e]tymologically, it would
mean 'pointed' (like an arrow) or 'peaked' (like a mountain)", as it probably
derive8 from the term .~pi,ga 'horn'. Warder continues: "T'his might refor to
heightened Hensibility or sensitivity." Therefore, it is reasonable to think, or
at least to conjecture, that tht'original mu111ces oF the meaning of the term
.~pi,grfra oscillated between tl~< variou~ unspecific m~tuifost'.ations <~f cn~otivc
arousal such as a g<~ncral tlml or excitement. Jn tlus not 11nplaus1bk mterpretati<m, it is :-;mmd to postulate, at least 011 a hypothetical ground, that
some features of a rcspo118e-oriented aesthetics ju0t like Abhimwagupta's
theory - were alrca<ly present in Dharata's fonnnlation of the Hasa Theory
aH it can he found in the NcT,[yaslL.'itra. In other words, the UHe of the very
term trigara might ~mggcst that in a somewhat lllll'efinc<l aesthetic theory
opPd capacity of irrmgiuat.ion (vru,tib/w.) which allows him to abstract from his own practical iutercHts awl desires even in ordinary life, so that he may experience beforehand
the rasa he will have to infuse) in the artwork. Sec Locana ad Dhv A l.G. 011 ]JTa.f'iblu!.
as poetic imagination, sec Srcckautiya (rn:n), Krishuamoorthy (l~M4), Golida (rnG:3,
:H8-:~48), Sen ( 19GG ), Kri8lnmmoorthy ( UJ80-81), .Jhanji (20D:I), Slmlrnan (2008). On
the broader concept of 'imagination' in tlw philosophy of both Abhiuavagupta and hi 8
]JarnmagnTn Utpaladeva, see Hati{~ (2010).

68

1
... . ( tl c Nal'iJa.fo,stra, primarily a drnmaturgica1 trcatis<~ [or pn1,ditio'. -.
ast1ato
t
..
.
r
.
.
.1,re(la;;
1
t c tltr 11 . ., 0 me kin<l of unspcc1 ie< 1 exc1tc1n<~nt, was consH c ,
1 ,.,
ers asor Ol
'
.
r,..
.
t\0
. l r<'"l>(>IlS<' a1>IffOlffiate to a love sc<:nc. l lus is tant.arnoun' '
t l lC ('!Il0 t,10Ilct '"
' ,
ly
' , tl t J3li-'r"ta's theory, a rasa could be <iualitativdy awl not 011,1
saymg ia , 111
..,, c.. '

l t
' .. tt t'ively - different from its corrcspornliug bl/,(i:ua. However, it nll!!, '
quan ,1 ,a , , .
. ..
.
. ie
i
.
<l tli't tlt('I'" i s a concrete i>ossih1hty that, already smcc the tlll '
i><' argue
..,, , , , "

,
,
of Bliarnta, the term ,spigii:ra could relate specifically to the emotion love
only."1 This would clearly c:mtraclict t~1c ar~umcntation just ~lcvelopcd ..
However, the analogous mterprdat1on of the other case referred to, t,l1c1t
of kar'U'(W and soka, rests on much more uncontroversial grournls. The weaning of the term lw:ru'~W is safely urnlerstood as 'compassionate' or 'compassion'. This fairly straightforward, and hence highly probable, nwkrstancling
of the term strongly suggests that, in I3harata'H theory itHclf, the reHponsc
to represented sorrow or grief (.foka) waH conHiclcrccl to be Horne sort of
compassionate or sympathetic attitude or feeling. This do<~s undermine the
broad-spectrum validity of the model of intensification, aH it is quite forced
to interpret the rdation between .folw, and kar'l1:(1,a aH that between a cornrnonly folt emotion ancl a strongly felt one of the same kind. KarLL'(W ancl
,Saha seem to he two qualitatively different emotions, one hdug felt in response to the 'presence' of the other one.
Moreover, as a further confirmation of this way of m1<lcrstawli11g the
term fwru~w ancl its relation to .foka, th(~re is tl~c indirect though very
tcllin? testimony of Abhinavagnpta. He reports . ancl is in a way ohligccl
to refute - the interpretation of the word f.:a:T"ll'(W as ka'f"ll'(UL, namely HS
sarla'!}ahrdayaUi 'the fact of having a compassionate heart'.
1
In my opinion, this view although directly attri1mt<~d to SrY Sai1lrnln
represents the cornrnon-sc~ns<~ unclcrstan<ling of tlin passag<~.:i:i Moreover,
G1

-.---:-------

. t' l
o [' t 1w term srn1111,r11,
,
c]caily,
tln t By
lf the
. tune. .of Dandiu
'
' it s<'<'IIIH
' , , tl l<l
" w mcarnng
1s
(' ( an, 11 ~tcn 81 hcd version of rati. s(~C the alrcacly q11otccl KaV'lllirlar,{;J, 2.28l: "l)dight
' lrn. to tlw muon
.
. a nmltiplicity
,) of drnncnts." (ra/,t.1,
1
l)CCOlllCS
~, ,, . .8rnqan

_i (L
. ovc') tlldll
with
S!
' ') , c8 pccmllY 'Ill comparison
'
'
' '2.2' 8'"
"IIflf}IJ,Tfllll'frt
.
lj(!t(}, 1"1ipab1i11'Ul?l(l'/I(}
' ' ,, ' ycrw
with
Krw,1;0,dars(l,
'"
avmg
, , p1't cl t, anger \Jcconws r11,'urlra (H;w'(~)" ( (},(lh:iru,h1;1i
'
:, ,, rcadt<'d
, , ,81\ch
. m
, <'Xt
, ,rune
paraf!i
k0 .t:~i wpo rawlratrnalxirri go,/,a(i,).
'

'
"'Sec
ADh
acl
NS
G.fi2-:~
11

1 ."

. ,
fr . l "l
.
' vo ' P .n l. Ono shonl<l rcpn~scnt the cnactuwut ol tl1,tt.
Jr
w
uc
t,
when
savonrc<l
Kumn
r
.
,
.
'1'1

}
[
,
' I
,
, ..
(
. , '
. ' rncs.
us is t w scuS<\ ol the\ <ksi1nmtio11. 111 (ac''
cornpd881011
karnna)
18
wdl-km
.
1
.
,
hPart F 'tl '.
. ., ,
iwn m t w worlrl as tlin lad o[ having a compassiomttc,
t,ators w l 10 covui:;,c the sorrow in t,l1c~ acWr
tl ,. l" . m, turnon\
' tl' ic, spec
. it 1wrtain8
, ' to
, i,111 rn't to t 1Sie logical
rcaHons [i.c ., tl, ic, wl lol c, <LII<LY
. . .. of Dd<'nnmmts
'
.
'l'r,1
]
<'<i11c'<'<\tl<'1lts ,,111<l
, ns1 ory ' tal<'8 Snd . , tl
l .
.
,
.
,, ,
' ,, , _, '" , , "
syal1hfrtrl'yah ,,y;r~,'l'Jl:J"' ,I 18 , IC dr (~s1gnat1on ot Karmp1 according to Sri Sai1knka." ( w-,
'
.
I
,
!JO IJIL8'Y(/,8'()(l Yll'fTl,(J,'(l(l~11(L ,, . " .
,
I ft/,
hm"IL~ia lo/;;c
nnsiclrllu
, ' . ,... .
' ,, ,,ar. u,~w t1:t V',IJO,J!IU l,1;sa~1,.
s1ulu,yu,hrday11, ,11,
L. S(I, CIL 1tnqa-tr ll'fl'lll'rl'T'lll'T"1 ~ /
.
.
.
., . /o1
1"1l~ta'U'!JIL]!(l(lc,fo ili ,fr't,forikuha(t).
' ," '
' Mi "1l,'fft Jimf,t!Jaf,mr1, sama:ji/;:anafll, 1,,1, " '
>'

'

lJNFUZ/';vJNCi TIIE FOl':ZY

69

in orcler to smuggle his interpretation as the right ouc,"G Abhinavngnpta is

rather clnnrnily forced to state that it is the term ncfrna used in the text of
the Nc/,fya.fo~stra that serves to indicate that the term karn,~w means nothing hut a generalized, and hence dcsirdcss and inherently blissful, version
of .foka. m
Finally, a fnrthcr and, in my opinion, rather couchrnivc clue about
both the homogmwity of nature between .cfpigc/,rn a.nd kar'll'(UL and their
heterogeneity with respect to the other rasas can be gleaned from Dharata's
text itself. In defining the eight ro.sas, the author of the treatise docs not
use one single expression, but two different formulations. The most common
one consists in defining a. rnsa as consisting in its corre8pornling bhii:ua.
vVith the exprcs8ion (consisting in', I am translating the Sanskrit ll.tmak:a;
for irrntance, rmulra is said to be krodluis/;}u/,y'ibhl/,vc/,trrw.ka ('(consisting in
the Stable State of anger") or bhayli:nalrn is said to be bhayasthcTy'ibhllVlltm,ako)r ("consisting in the Stable State of foa.r")." 8 The very same pattern
is followed for hc!,sya, vfra, bhaylinaka, bfbhatsa n,nd adblmta. However, in
the case of srrig(/,ra and karv:(w., Bharnta uses a different formulation. He
maintains that ,1rri,gii:m is rat'istluTyibhli,vaprabha:ua?1. and that kar"ll'(W. is sokasthayibhcfoaprabhava!i. 59 Therefore, they arc not said to 'consist' in their
Stable State in particular, in a heightened version of it, according to our
lllHlcrstanding of Uharnta's theory--, hut to originate or derive from it.() 0
r,r;On Ahhinavagnpta'8 lwrrnencutical attitude towards diange and innovation, sec
Cuneo (forthcoming).
7
" Sce A13h ad NS G.G2<l: "Therefore, Kanu.m is nothing but sorrow itself. It can be
defined as what is to he savoured as completely gcnerali:t,ed thanks to the foregoing
procedure [i.e. the process of 'gcncrnlim.tion']. For this very purpose, the word n<!ma has
been employed [in the passage of the Nr!fya.<iristra] ." ( tasrnat ka:r"U'(W. il'i soko.(i. sanJO..'>'(ldhrira(w.lvcna pr<igyu.klya. l!sva.dyo.marw.sya sa:rr1.fr/,cf. ta.darlham. eva nrl.rnasabdo,(1.).
8
r; Cf. NS G, prm;c aft.c~r G:l, vol. 1, p. :l2G and NS G, prose after G8, vol. 1, p. :122.
G!lCL NS G, prose after 11G, vol. l, p. 2!)8 and NS G, prose after Gl, vol. 1, p. 31:3.
GUObviously enough, the oddity of these two rasas is accounted for in Abhinavagupt.a's <:ortmHmtary within his usual paradigm of interpretation. As, on the one harnl, they
rnpn~sent the <)Xt.rcmc manifestations of the inherently plcas11rnbk-cnm-painf11l uature of
everyday lnunan !)Xpcricnce and, on the other hand, they arc more intimately intertwined
with the limited individual self of th<~ knowing subject, the emotions of delight (ra.t'i) and
sorrow (.<ioka) a.re considmed utterly diffcrnnt from the ra.so.s related to them. In Abhinavagnpta's own words takm1 from ADh ad NS G.!19-Gl, luisya.rasa.pro./;:ara~w., vol. 1, p. :mD,
"on the other hand, the sag<~ [Dlmrnta] has used the term 'origin' for the two [definitions
of .<ipi..q1ira and ko:r"11:1w], liecansc only delight awl sorrow whoso savouring [i.e. whoso
rasas], contrarily [to other rn.sas], arc forms of consciousness of a kind different [from the
Stable States] arc gmwratcd by the mere forcn of the causes grasped in a perception
delimited by its partic11lari:t,cx1 and individnali:t,ed nature [i.e. restricted to particular persons], as they consist of pleasure and pain increased up to their oxtreme limit (ra.ti.fokrw
cva. para:rr1, rit11,jjrit{1;asa:r11:uidasvrirlo:u. d/dirrLT"IL\l!LasvJdw.d11.(1.kliar1qHilverrn nissi!d/1.r!ra(iatrn/,-

70

'
"'(l.1
l t a is
. l . , , 'l ol ra'"
' ,
ln my opinion, a quite plausihlc intcrprctat1on o\ t ;is' s\n;-;i<lcrc<l 'ccccll t
r ll.. . ,. rl'lw CUS<' of .ST'Ti!ja:m an<\ J,;(1:/"ll'p,11, \mV<~ to )(, ( ,(
'(1 s(IS cnJlllO
t l l(' 10 owm,. . ' ' ' '
. .
.
.
l 'S(' t WO 'I " . h'
, .,' ith rcS\)ect to the other six 'fnso,;-;, mso\n.r as.tic. , ~ 1 rcspcctivc ..
c l
.
l' .. f'l nHl so .. a,
nst
l,nc w ,
, . ll)\y reffar<lcd as an intens111c< vern1ou 01 rn,, < .
, ~ t,, t,hcY 111 .
\ )C sm .
h
l' tl , No f 1;0 so,,s .n",
-cs
' 't, th<' contrary a\rea<ly in the mm.c t(~xt o , lC
:. ".
., <'Xllcricnc '
(,-!iu1.c
, ,
.'
. .
\"ll . t ,110 t 1011,11
. \lY
\ e considere<l as sped Ee <ln<l <tna11tat1vdy < l ct en, <'.I. , . l drnuit1tJc<l. "'
) lll tCSllOllSC to the ., prcs<mta
. \',lOll ' 01ct ,WO VCry, s1w('\lic
<1Jl<
l .tJL
r,lt
'
, '
' . , 1(i \ r '1)(1\, \C '
ie
.
.
.
. . , , . l 'trnPlC o

very co1nmon, emot101ml s1tnat1011s: a love scun,

itll<

. ,..,

scene of some kind.

Concluding rernarks
.

11 opcll

is stJ

The <lialo;1c sitnation

) tscriP
. . l'f''tf I os,,
Halhf ass, Pre l1,nnn<'
tcr111s,
.
.
.
. .
l 11
. 1 oencn1l ; .
To conclude, ll.t least for the tnnc hmng;: it rn somi< , . h ' , .(\ l>hii:!!i1:l1
to attrihnte to Blmrata a th(~ory that umceives m,sos as hrn;:itcn~ the H,t1:it
0
especially, in opposition with Abhirn.wa;npta's reintcrprctn.tl(~ll , However,
Theory that conceives ro,sas as distilkd, a1Hl h(mcc, hlissfol bho:ua::>\. .(\ vali(l
. . .
.
..
..
.
.\
. 1sh ere
. '
tlns seemmgly vmhle pattern ol mtens1hcat1on cannot )C cm ' , . cntric
for all nisws, as two ol them .frrir1ct1n
arnl lrnrnno,
am <\dinitd. Y
cc\
tli(M
.
.
.
.
f 01, '
in their heing a direct response to, and not a mer<~ stn~ngthcnlllg.
respective hhii:vas.
. , to thC
r'
l
l
l
l
10
"\lcn[!,C
vonsequent y, t iese two n1,.<;o,s seem to rqm~scnt hot l n. c
' l it1'1vt\:'
unitary nature of Bharnta's theory awl a possihk stimulns f(.)t' j\h ~ist,illct
;upta's own intenm~\.n.tion of n1,.'Jo, arnl IJh<l:un as two qllalitat1vdY < t 0 t,hC
cmotiona1 experiences, the fonm~r hdn; cxpni(~nc<~<l as n rcspon~c '
'presence' or the ((rc-)prcscntaJ,ion' of the \aU<~r.
i
t.liC
.
l h'\l
Furthmrnor<:, as a speculative daborntion, om'. mi~ht an~nc . ' ' ' 1(\if1ll
1
. \ y \ 11;
' \ l f1rcqncncy Otc l ov(: sc<m<~s an<l pathd1c
,
1 seen,,
('8 ll I I
comparat,1vc
... ,1. t,o
. . .
.
. l 1.111,c i
drnmaJ,urgy nn;ht have rnprcscnt<'.d t\. concrd.<: spnr, or the vctY ' ~' ,ct.io11
.
.
spccuh.\.t.10ns
on thcn.trc .rn \he
' , dire 0 r1zS
tl, 1e <levelopmenl. 01r t lic t \1coret1cal
.
. as Umt dahorn.tc< l .in l 'he ""\ oriS'
o a u ll-1'lcdpl rcsponsc-oncntcd
ac:-;thd1c:-;
1
o( AhhinaV<l!J}1pta. Howcvnr, given hoth the rela.Liv<~ hridn<~ss n11<l n.P .1ricS
1 l
,._
. s1~y
l.1c
c of l)l
) mrnJ,a' s N-a!:yn.'io,stra
and t.h<~ ahs<~ne<~ ol cnrly <.01n 1ncnl"
'
----~~-------------I 0f1,11.1Lltl!l
untv<miyamugn1,hngrh:1,tn.hdtt.bnlwl, evoL7m<ludc DOJ,o 'to 'nnuor m1ininn pndJ/t.ali<~Y :1 c;li~;ht

k;rtam)." This V<~ry ismu~ is alnwly trcate<l in Tn\ih ( \.!)'.H, UH-l'.):i), a\tho11;1i with ' '
\lias in favom of A\Jhirmva;npta\; interprntation.
. ,vf11l
Gl'l'l .
l
rt l , sono
,
ie mvert<~< connnns am meant to hig)1li;ht the nU.<~r\y nli<~n natnrc OL ' ic' . .1c"11d
1
1
<lr'u11't1'c
l l n< l'mu t.l icatrn with
n~8p<:ct to thn sty \ c an< \ ,si nri11 '
' "' 81tu,1, t'10ns of ('l
_, ass1ca
ot Ch1ssical Greek trn;e<ly.

71
thereon, many of these rrurnings of mine arc, at least to an extent, doomed
to remain on hypothetical terms.

Bibliography
Prirnary Literature
Abhina.vagnpta, AbhnavabhiiratT, ed. RS. Nagar, Delhi: Pnrimal Publications,
4 Vols., 1998.
Abhinavagupta, Dhva:nyc/,lokalocana, ed. Pattiibhirarrm Siistrf, Benarcs: Kashi
Sanskrit Series l:~G, l~MO.
Abhinavagnpta, Pariitrirrufikiitatlvavi1H1ra'(W, sec Gnoli ( 198G).
Abhinavagupta, Tantriiloka with corn:menlaTy by R<l:jiinaka Jayaratha, ed. M. K.
Shastri, vol. 1-12, Allalmbad-Srinagar-Dombay: Kashmir Series of Texts
and Studies XXIII, XXVIII, XXX, XXXVI, XXXV, XXIX, XLI, XLVII,
LIX, LII, LVII, LVIII, rn1s--rn:~8.
Anandavardhana, Dhva:nyctloka, ed. Pattiibhiriima Sastrf, Dena.res: Kashi Sanskrit Series l:~G, 1940.
Dharnta, Niif;yo,,<i<Li:dra, ed. RS. Nagar, Delhi: Parimal Publications, 4 Vols. 1998.
DawJin, Ka:uyiida:r.fo, eel. Pt. H.angacharya H.addi Shastri, Poona: Dhandarkar
Oriental Research Institute, 19:38.

Dhanafijaya, Da.fo:rlipaka, with the commentary Avaloka by Dhanika, and

the subcommcntary Lagli:n_f;'l/;:ii, by Dhat-j;anrsiqilm, ed. T. Venkatacharya.
Madras: Adyar Library and Research Centre, 1DG9.

Secondary literature
Ali, D. (2004) Co'IJ,rtly Cullurc and Pol'itrnl
bridge: Cambridge University Press.

Dzf'c in

Ea:rly Medieval India, Cam-

Aptc, V.S. (l!HiG) The Practical 8o:nskrit-English Dictionary, Delhi: Motilal

Banarsiclas.
Dhattacarya, K. ( 1972) "Siintarnsa ct Advaita", ,/(mrnal Asiatquc lGO: 89- lOG.

Bhattacharya, S.P. (1D7G) ,<Jii:n/;a Rasa and 'its Scope in L'ilcraturc, Calcutta:
Sanskrit College.

72

. ,,v 1 1 Ii Sa.ukt.
Snnsf,:11.L-r or.r'.rru.<.,,'.

1 ,tershnrg:
1c

) .. lt.l" r\' o Hoth R. (l8!""J!""J-187!'>)

'
"
l .
'vVi. '('J\~-;c\wltcn.
l )o 1 tnh ,, . .,
Bnch<hnckcrci dcr kaiscrlichcll J\.lm< <~lllW <1<~1'
ss
, , Aesthetic
' t
. A rt tnd ns ,in
1 11 \l"\
\<' ( \ <) \ ')) "Psychical
Dist.au cc as a l ac .or lll
'
'
1.JH 0 h 1 ' ). . .~
Principle", [JT"il:ish .!onrruiJ of f>.'i!Jl'.holoyy !""J: 87-l 17.
. of
,
\ )1)rn1snl
. ..
.
. )
\) .
J\ J11\N\)l'<'\.lVC j l
,. th
D ('>008-200\))
"Emot.1ons wit. iont <~sire.
11

l''t"t
1 ,-, " tW~ l
( .J'\ltl<'<)

'
~

f
t
lw
'
0
H.asa Theory.
Anuot;itcd Trans\at.1oll
. '
A1)lllnava."'1lIJl.a's
h

U"
onl)\tblishc<
n,nd Seventh Chn.ptcrs of Ahhi1mvngnpt.;1's !1/Jh1.na1111.l1fw.ro. ' '
PhD Thesis, 'Sapicm~a' University of H.ornc, H.omc.
_,,
1

Cuneo, D. (200D) "The Emotimml

Sph<~rc

\ .

l1;;t'ttJ

V'\ll "

in the Light of the Ah llllt '

IUvistli di 8tu.rli Oricnl.ali 80: '21-:rn.

. .

AbhitliY

Cuneo, D. (fort.hmmin;) "Smu;;lin; Novelty Dismantling; Tra.dil.lol:l nJ. 11tth

on t l lC <l cnvat.lons

va;uptn. arn l Ill

> lOJa
01r ro..~a.s " , papc'l" Iffcscntc<

World Sanskrit Conference, Ky<Jtq lst-Gth Sept.ember 2009.
,

'
, Actor.

Cuneo, D.; Ganser, E. (forthcoming) "The J\csthdic Experience of the

.
"

);ipCl

l-''.i. (l'. ort1iconun;

) ""rl'l
I n< l'mn Then,trc
l

'
. w ('.Jonccpt o[ :;u,f.lvn m
, 'tcin\lcr
( ,.anser,
presented at the XIV 'vVorld Sanskrit Confcn~ncc, Kyoto, 2nd ScP ' '
2009.

Gerow, E. ( lD!H) "Ahhirnwa;upt.a's Aesthetics as a, Speculative

.Journal of Ame:rimn Orient.al Socie/:1; 1111: 18()-208.

. n1"'
pn,ra< \ig

. ,, 10 iir1ia!
' J\l<.HJ
l 1mr, 1\ . ( 1972) "On Siint.;t
'
H.asn iu Sanskrit 1) oc t.ic.s '
Gcrow, I',.;
of the !lm,cr"irn.n Oricntnl Socfol.fJ D2: 80-87.
r '
\rttioll
111

Ghosh, M. (U)G0-1%7) N1i[.ya.fosf,ra, of !3lw.ra.l.11,, Edition and 1 ;; 'fc;zt.;

Va.rnnasi: Chowkhamha Sanskrit. S1:ri<~s, t} Vols. (Vo1. l. Text.; Vol. ,.,.
Vol. l. Trnnshtion; Vol. 2. Trnnslation) lH.cpri11t 2om].
Gnoli, R. ( l!)(i8) The !lf~.~thc!:i.c l~:J:[!(Tfoncc n,r;conliny
V;mrnnsi: Chowkarnha Sanskrit Stll!llcs 72, 2nd <~d.

ipt(h

f,o

1\!Jhin<iOaf]l
1

~l. ( 198G! fl Comm.cn/.o di 1\11!1:i:na:o11.ynpl.11. 11.llo. Panr,/,rir[L~il.:(i,

kat11f.l:u1i-m:uo.niu.a), trn<lu:i:irHw <~ \.r~sto, Roma: lst.ituto itafowo per

t .1n.s1.-

1
(Par:

G noli,

11

~{:~<liO

crl Estn:rno Oricntc.

c~'onr lc1.,

J (. 1<)("))
l)

'I'I 1.e

v1,swn
. of. Uw

. Pods, Th<~ llag1ic: !vlont.ou 1\'.z. Co

Vc1lu;
si-

Goodwin, R. (19!)8) The Playworld of So:nskril. Drnrn.11., Ddhi: Mol.ihtl

dass.

})nJ1<l.l

. 1/1<'

Hanli.kar ' A.Il

"l tc\.1c
A pprccmt.or
.
(. t<)<)
1) "'l'lle, Acs1.
or s11.l1.pl11:yo. " , ! \ n:n.a,,.ls o '.
IJhanclarlw.r Oricnl.o.l llcscarch ln.~l:il:nl.e 7!"">: WG-272.
1

UNFUZZY!NC Till; FUZZY

Ingalls, D.H.H., (ed.) (1090) The Dhvany11,loka of )fna:nda:ua:rdho:na wilh the

Loca:na of J\ bhina1H1,_1r11pla, translated by D .H.H. Ingalls, .J.M. Masson and
M. V. Patwarda11, London: Harvard Oriental Series 49 .
.Jha.11ji, It

(200:~)

"Ahhinavagupta's corn:cpt of prat'ibful,", Sandhana :3/2: G7-78.

Kamimura, K. (lD8G) "Is the word rnukhy1/, vrlli synonymous with abh:idhii according to Bhat ta Lolla\;a?", Adyar fAbn1:ry Bulletin GO: 418-4:m.
Kavirn.j, S. (200G) "The Sudden Death of Sanskrit Knowledge", Journal of Ind-inn Philosophy :n: 110-142.
Krishnamoorthy, K. (l~H1) "Observations of Sanskrit Litcrnry Critics on Poetic
Imagination", Poona Orientalist; 9: 12:~-1:32.
Krishnamoorthy, K. (lDG8) The Dhvanyiilok:a and its Orilfr8, Mysore: Kavyalaya
Publishers.
Krislmamoorthy, K. (1980-81) "The Nature of Creative Prntibha according to
Knntaka", Indologico. Taun:ncnsia 8-9: 22:3-232.
Kuhn, T.S. (1970) The Strucf;urc of Scient'ijic Rcvoluf;fon, second enlarged edition, Chicago: University of Chicago Press.
Kulkarni, V.M. (ed.) (198Ga) Some Aspects of the Rasa 111.cory, Delhi: Dhogilal
Lclwrchand Institute of lndology.
Kulkarni, V.M. (1986b) "Abhinavagupta cm the Alaukika Nature of Rasa", in
Kulkarni (1986a: 28-42).
K1mjun11i H.a.ja, K. (1!)~)7) "Sahrcla.ya: the Ideal Art-Connoisseur", in Lienhard;
Piovano (1997: 111D-42:~).
Lienhard, S.; Piova.110, I. (eds.) (lD97) lc:r ct !Jillcro.c. 8lwlies in Honou:r of
Professor Osca:r Botto, Torino: Edi:;,ioni dcll'Orno.
Masson, .J.M.; Patwardlmn, M.V. (19G9) g(Inla:rasa and Abli:inavagupta's Philosophy of Aesthetics, Poona: Blmndarkar Oriental Series 9.
Masson, .J.M. (1979) "The Sahrdaya", in Sinha (Hl79: :n:~-:nG).
Matilal, B.K. (198G) Perception. An Essay on Classica.l Indian Theories of
r<nowlcdqc, Oxford: Clarenclo11 Press.
Mayrhofor, M. ( 1!)8G-200 l) Etynwlogischcs Wiirtcrbuch des Altindoa.rischcn,
Ifoidclherg: winter.

74
.. Cin1hric\{';C1
McCrc< , L.J. ('2008) 'L'he Tcleol.oyy oJ Podics in !vlf'.tlicual l(a.shmi.r
1

'

London: Ua.rvanl lJniv<~rsity Pn~ss.

MA

. . \ Clarcndoll
Monier-Willia.ms, M. ( tnGG) A 80.ns/,;ri/.-Bnylish IY.icl.iona.r!}, Oxlm< ,

Press.

,
. . ' A. mstorPollock, s. ( t\l\)8) "Bho.ia's Sril;i\.rn,prn.kiisn a.rnl the Prohkm ol ',nsa... , /Etnd.c~
.
1A
l 'l'
l t. " 1 .: 1sch e ')/,u,d1.cn
ica.l Introdnct1on am nnota.trn rn.ns a .1on , " s1.<1. "' '
J\.s-i<J.l-i<]'IJ,CS f)'2/\: ll7-l92.

Pollock, S. ('2010<1) Epic nnd Arrr1im.cnl. in 8o:n.skri/; Lilcmr'!}

Honor o.f Hobert P. Goldnwn, New Delhi: Mm1ohar.

.
[i's
sai;s 1,1i
1
ll1.sf.o'l'!J ' '

ti cal
'l'l , Iknncucn , ,
Pollock, S. (2010h) 'What was Blmt\.a. Nnyalm Sa.ymg.
1c , ,,
81
Transformation of Indian Aesthetics", in Pollock (20 ton.: 14.~- l - )
,

.--

r
( l%G-Gb) "Lolla.ta's Theory ot

Prabhaka.rn Sastry, J.
Oriental Instit11.lc of l3u:rodu lG: lf>7-HiG.

I.>,asa" , f(>'ll'T"IL<l,l

f, t/1<''
,

.
. ind
1
RaghaV<tn, V. (1%7) The Number of Hasas, Mn.<lras: Adya.r L1hm Y '

nc-

search Cont.re.
> t''
"A I"'ivc-'I'.run<::<~<
l l , I'our- '1'uslrn< 11~kplmnt is
runnu
win
the SkY
.1e, I . ('>()l(l)
.~
n
. 'I'',
I,,,i
,, . I
. .
.
A\ l . , virnpt<t.
0
llow In~e is nmgmation Accordm; to lJtpala.dcvn and ) 11n,i '
Asiutische Studien/Etudes Asial:iqncs M/2: :H l-:~8!'>.

1.'tic<"l

J>K
t..ton m
C'_;O1<~n< 1g;c
'- . (t<)f'l"'.)
. )) "l nmgnm

A\ )1mmvn.;np

t"\.. r!\ Cll

.
. d
1
Analysis of Chrh-;tian and Slmiva Stn.rnlpoint.s'1 , .Jmirno.l oJ Ac<';!,hd cs o.n
Ari; Criliidsm, 211: 97- l07.

'
1'JCil,

itn<l

Shanna, A. (HHH) Bs:-;cius on lhc !vfol/,(il>/uimLn, Lei< lcn: l~.J. Brill.

l0l01;ic<1
Sharma, M.M. ( U380-8 l) "On an Extract. from the J\ /J/1.in11.val!lw:ni.11.", In< ..
'li1:n1"incn:;'io, 8-9: 40G-1l l1l.

H.ii.,jaHcldiill"'."'.~~;~~-

Shulman, D. (2008) "Illnmilmtion, Irnar';iuation, Creativity:

taka,, and .Jagammtha on Prn.tihlui.", .Jov.rnal of ln<lfon 1'/1:i1.osoph'!J
f'lOG.

,)(J.

l 11.,.,
Sinha, .J.P. (e<l.) (rn79) Lwlw-i/;; 8t.ernl1ach Fdici/.u.t.ion Volu:mc, Lncknow: A.k '
mmrat.iy<t Smrnkri.

.. r (
)
riali
qna.rcnu,. '. 2008 Tnulens, Tnul:ib1:rn., Hu;iJl'icns. 8f:n1li s/.odci. c. s~ 'p\\s11.ll"1.shtnto della tra.d'iz'ione ncll. 'nnliicl1:iUI. .~11.dnsfo./:im, Fir<:11'/,c: Socicta :.;
it.rice Fionmtina.

Sr
.
'I' N (.t \):n) "lnmg;irmtion in lwliml Poc\.ics 11 lndin.n If i.s
t1rri<:Ol
' eelmnt.cuya,
'
1
Qua:rf;erly t:~: !)\)-8;1.

75
Srinivaf>an, S.A. (1D80) On t;hc cmnposUfon of lhe Naf;yaAlislro,, Rcinbck: Wc11,lcr.
'l'ickcu, II. (l!JD8) "Ikview of Po6tiquc du thcf1trc lndicn: L(~ctures du Nii(.yasil.strn by Lyne Bausat-Bondcm'', Asian Folklore 8lnd'ic!i 07 /l: 171-170.
Tubb, G.A. (lDDl) "Si"intarnsa in the Mahiihhil.rnta", in Shanna, A. (1991: 17120:~).

Vatsyayan, K. (199G) Nii_tya.ef<Istrn, of IJ!wratmmm:i, Delhi: Sahitya Aka.demi.

Warder, A.K. ( 1980-81) "The Origins of the T'cchnical Smiscs of the Word llasa",
Adyar Library Bulletin 44-40: (il4-G:M.