189

!"#$ "&'& '(() *# "&'& +(&",

SheIdon !"##"$%
Columbia University, New York
To a degree that may prove to be more far-reaching than I can demonstrate
within the Iimits of the essay that foIIows, the theoreticaI understanding of
rasa was shaped by its historicaI extension from the domain where it was
nrst formuIated, that of djsyakavya Iiterature meant to be seen, that is,
drama to the domain of sravyakavya Iiterature meant to be heard, that is,
poetry recited and undoubtedIy read privateIy. There is nothing very originaI
about positing such an extension (though the stages in the process have never
been identined as weII as they might be), yet I beIieve its true consequences
remain to be fuIIy assessed. I hope to show here that one of these pertains
to the number and kinds of emotion that can count as rasas. Other cruciaI
aspects of the inteIIectuaI history of rasa theory may be impIicated as weII,
incIuding notions of its very ontoIogy and epistemoIogy: where it exists and
how it comes to be known. My thoughts here are tentative, unIike the admira-
tion and aection with which I oer them in honor of one of the great rasikas
among Sanskrit schoIars of our time.
I. Literature Seen and Heard
Before we can reconstruct the history of the extension of aestheticaI
anaIysis from the dramatic to the non-dramatic, we need to show that the
Sanskrit tradition dierentiated between the two types of Iiterature, or better
yet, that it drew an opposition indicating that anaIysis appIicabIe in the one
domain might not be automaticaIIy appIicabIe in the other. WhiIe the specinc
binary seen / heard is found as earIy as the Kavyadarsa (1.39), it is not
without some conceptuaI dimcuIties. For one thing, the representation
(abhinaya) that dennes djsyakavya itseIf comprises in some measure the
verbaIization constitutive of sravyakavya,
1
softening any hard distinction
between them. For another, when kavya is distinguished from sastra, as it
very frequentIy is (thus Bhoja, for instance, decIares that Sanskrit discourse
1. In some measure, that is, one-quarter, since there are three other types of representation
beside vacika, verbaI (nameIy, angika, physicaI; sattvika, psychophysicaI; aharika, costuming).
But note that Natyasastra 1.11 speaks of drama itseIf as both djsya and sravya.
From: In Aux abords de la clairire, ed. Caterina Guenzi and
Sylvia dIntino. Paris: Brepols, 2012, pp. 189-207.
Collections rudites de lEcole Pratique des Hautes Etudes.
Sheldon Pollock
190
can be scripturaI, traditionaI, or mundane. Mundane discourse is sastra and
kavya),
2
drama and Iiterature are no doubt both subsumed under the same
term, and any distinction between them is bracketed.
That said, not onIy were the two genres categoricaIIy dierentiated; they
were often radicaIIy opposed, as we can see in the contest between them for
primacy in the minds of Iiterary critics that became a running dispute in the
tenth and eIeventh centuries. Bhoja merits quoting in fuII:
It is dimcuIt to specify what preciseIy rasa is, since it is knowabIe onIy expe-
rientiaIIy, and is not universaIIy accessibIe. When dispIayed by skiIIed actors
in correctIy performed dramatic presentations, it can be determined by the
audience; when properIy
3
decIaimed by great poets in their compositions, it
can become accessibIe to the minds of the Iearned. However, [there is a dif-
ference between these two modes of experience:| things are not so sweetIy
savored when they are actuaIIy perceived as when they are cognized through
the Ianguage of masters of Ianguage. To quote:
A subject does not expand the heart
so powerfuIIy when we see it portrayed
as when it ashes forth from the words
of great poets decIaimed with art.
Therefore we prize poets far more than actors, and poetry more than dramatic
representations
4
.
As an anonymous verse puts it, The Ianguage of poetry and the representa-
tions of drama are the two ways [of expressing rasa|. The former is superior
in this by reason of the range of its narrative power (vastusaktimahimna)
5
.
Abhinavagupta, by contrast, eIevated drama to the paradigmatic form of
Iiterature, enIisting oIder schoIars in his cause:
Other thinkers argue that the reIishing of rasa can occur in poetry no Iess than
in drama, produced by the exceptionaI beauty of its Ianguage quaIities and
rhetoricaI ngures. Our view, however, is as foIIows: First of aII, Iiterature is
comprised chiey of the ten dramatic forms. For it is there that, thanks to the
appropriate Ianguages, cuIturaI modes, intonations, costumes, and so on, the
presence of rasa achieves pIenitude. In a Iiterary work Iike a courtIy epic, by
contrast, we even have femaIe protagonists speaking in Sanskrit, one of many
improprieties that nnd pIace in the narrative simpIy because it is not possibIe
to do otherwise
6
however much it may not seem inappropriate, in view of
2. Srngaraprakasa, voI. I (&' )*+,-)+ .' )*+,-)+ (eds.), Srngaraprakasa of Bhojaraja,
2voI., New DeIhi, Indira Gandhi NationaI Centre for the Arts, 2007), p. 163-164.
3. I conjecture yathavat for yavad.
4. Srngaraprakasa, p. 2-3.
5. Cited by Srdhara (c. 1400) whiIe restating Bhoja`s view. See Kavyaprakasa (/' 0"123
(ed.), Kavyaprakasa [of Mammata|, solah tikaom sahit, 6 voI., DeIhi, Nag Prakashan, 1995-),
voI. 1, p. 77.
6. PresumabIy because of the genre constraints of the courtIy epic (rather than the incapacity of
the poet to become a pIaywright, so &' !' %234#-, Rasa-Bhava-Vicara, Bombay, Maharashtra
Rajya Sahitya-Samskrti MandaI, 1973, p. 196).
From Rasa Seen to Rasa Heard
191
the maxim that you Iike whatever you are given
7
. This is preciseIy the reason
it has been argued, as noted earIier, that Among aII the varieties of Iiterary
composition the best is drama in any of its ten mimetic forms. Other Iiterary
works, from the courtIy epic to the isoIate verse, come into being by borrow-
ing structures such as acts and scenes from the ten forms
8
.
Whatever other questions may be at issue here, it shouId be cIear that by
the beginning of the eIeventh century and no doubt far earIier, drama, or
Iiterature seen, and poetry, or Iiterature heard, constituted two fundamentaIIy
dierent and dierentiated forms of Iiterature, and indeed, that there aIready
was a dispute about the extension of rasa anaIysis from the one sphere to the
other. Thus when Bhatta Tota, Abhinavagupta`s teacher, asserts that there
is in fact no such thing as non-dramatic Iiterature, he wouId seem to be
responding to a new argument (one aIready contested by Abhinava himseIf)
on the extensibiIity of rasa theory:
Rasa exists onIy in drama, and in poetry onIy to the degree that it mimics
drama. For as my teacher has argued, with respect to the eIements of the
Iiterary text, rasa comes into being onIy when a state of awareness simuIat-
ing visuaI perception (pratyaksakalpasamvedana-) comes into being. To
quote his Kavyakautuka (Literary Investigations), So Iong as poetry does
not approximate the character of a performance, there can be no possibiIity
of savoring rasa
9
.
II. Assimilating the Analytic
If the distinction between visuaI-dramatic and auraI-poetic Iiterature was
thus cIearIy estabIished from a reIativeIy earIy date, the Sanskrit tradition
nowhere expIains how an aesthetic theory deveIoped for the former couId be
appIied to the Iatter and indeed, why it shouId be. For the most part we are Ieft
to reconstruct this deveIopment by inference. The nrst and, so far as I can
see, the soIe expIicit decIaration on the subject comes from Rudrabhatta`s
Srngaratilaka (databIe to somewhere in the period 950-1100): GeneraIIy
speaking, the nature of rasa has been discussed by Bharata and others in
reference to drama. I shaII examine it here, according to my own Iights,
in reference to poetry
10
. We are justined I think in pushing this attempt
at integration back to the time of Rudrata`s Kavyalankara, the source of so
much of the Srngaratilaka: When Rudrata remarks that the greatest eort
7. tavativa hjdyam. My transIation is uncertain.
8. Abhinavabharati (%' %&+.1320""&516 (ed.), Natyasastra of Bharata, with the Abhinavabharati
of Abhinavagupta, 4
th
ed., Baroda, OrientaI Institute, 1992), p. 285. The citation is from Vmana,
Kavyalankarasutra 1.3.3 (&' 4' 712552 (ed.), Kavyalankarasutra and Vrtti of Vamana, Varanasi,
Braj B. Das, 89:;).
9. Abhinavabharati, p. 284.
10. Srngaratilaka 1.5: prayo natyam prati prokta bharatadyai rasasthitih / yathamati mayapy
esa kavyam prati nigadyate // (&' !+.$1-# (ed.), Rudrata's [sic| rngaratilaka and Ruyyaka's
Sahrdayalila, KieI, C. F. HaeseIer, 1886).
Sheldon Pollock
192
must be made to invest kavya with rasa
11
he was sureIy referring to auraI-
poetic Iiterature, since his work has nothing to say about drama. This wouId
bring the date for the assimiIation of sravyakavya into the anaIytic of rasa to
sometime in the middIe or Iatter haIf of the ninth century.
This dating nts with what earIier works that touch on rasa have to teII us, a
story that is weII known but bears restating. The Natyasastra, a composite text
the core of which is probabIy not Iater than the fourth century, introduces the
theory (in a confused and much edited form in the manuscripts now avaiIabIe)
entireIy within the domain of dramaturgy. By contrast, the earIiest extant
texts of Iiterary theory more narrowIy conceived, Bhmaha`s Kavyalankara
(c. 650) and Dandin`s Kavyadarsa (Kavyalaksana, c. 700), are concerned
excIusiveIy with auraI-poetic Iiterature, and whiIe both are aware of the
dramaturgicaI theory they conceive of rasa excIusiveIy as one or another type
of rhetoricaI ngure
12
. Their precise arguments concerning the tropes rasavat,
preyah, and urjasvi are not easy to expIain in brief compass
13
. In the simpIest
terms, these aII represent expressions of heightened feeIing: where a given
emotion cIearIy manifests itseIf (rasavat); where a warmIy feIt compIiment is
conveyed (preyah), where a character`s arrogance or vehemence is expressed
(urjasvi). And whiIe not embodying the indirection (vakrata) that dennes
other alankaras, these emotion-tropes nonetheIess remain speciaIized uses
of Ianguage and hence are capabIe of being understood as ngures of speech.
The expression of heightened emotion in the character, of course
may indeed coincide with Bhatta LoIIata`s understanding of rasa, which
constitutes the oIdest, or cIassicaI, view: that is, the view cIearIy impIicit
in the Natyasastra, ascribed to Dandin himseIf and the other ancients by
Abhinavagupta,
14
and preserved as Iate as Bhoja. Yet for both Bhmaha and
Dandin the representation of emotion in auraI-poetic Iiterature is subordi-
nate to and therefore subsumed under the dominant discourse on alankaras.
Nowhere in this earIy work is rasa considered anything more than a trope; it
certainIy does not yet, as it was soon to do, constitute the heart of Iiterariness.
11. Kavyalankara 12.2: tat kavyam [conj. for the unmetricaI tasmat tat| kartavyam yatnena
mahiyasa rasair yuktam (Rudrabhatta had this v. in mind when he writes tasmad yatnena
kartavyam kavyam rasanirantaram, Srngaratilaka 1.8) ()<&42!&2.2) *' #' .' !23.+%2&
(eds.), Kavyalankara of Rudrata, with the Commentary of Namisadhu, Bombay, Nirnaya Sagar
Press, 1928). Note Rudrata is stiII writing alankarasastra for makers of Iiterature rather than
readers.
12. For the dates here see 6' 7&"33-&, A Question of Priority: Revisiting the Bhmaha-
Dandin Debate, Journal of Indian Philosophy, in press. For the anaIysis of rasa as a trope in
earIy alankarasastra, #' 0$$&-2= The Teleology of Poetics in Medieval Kashmir, Cambridge,
Mass., Harvard University Press, 2008. WhiIe Dandin regards dramaturgy as the object of a
separate science (1.31) he is fuIIy aware of the doctrine of eight rasas: astarasayatta rasavatta
smjta giram, Kavyadarsa 2.290 (2' 512%<& <' /12 (eds.), Kavyalaksana [= Kavyadarsa| of
Dandin, with the Commentary of Ratnasrijana, Darbhanga, MithiIa Institute of Post-graduate
Studies and Research in Sanskrit Learning, 1957).
13. Bhmaha adds severaI more such ngures of aect (Kavyalankara 3.5-11) but Dandin`s
three become canonicaI.
14. Abhinavabharati, p. 266.
From Rasa Seen to Rasa Heard
193
Udbhata (. 800) marks the nnaI and, by his date, contradictory stage
of this rhetoricaI anaIysis of aesthetic emotion. On the one hand, as we might
expect from the nrst known commentator on the Natyasastra, Udbhata
redennes the expressions of aect to approximate the fuII rasa typoIogy:
rasavat, which for previous writers was simpIy heightened emotion, now and
for the nrst time becomes the full realization of rasa with the compIete panopIy
of aesthetic eIements (vibhavadi);
15
preyah (in the now oddIy overdetermined
form preyasvat),
16
earIier an emotionaI compIiment, becomes the intimation
of an emotion (or bhavakavya, as his commentator Pratihrendurja notes);
urjasvi, formerIy pridefuI expression, becomes rasabhasa, or semblance
of rasa, marked by sociaI impropriety. On the other hand, however, and
despite this approximation to the dramaturgicaI modeI, Udbhata continues
to categorize aII these as ngures of speech, on the same order as, say, the
eIevated (udatta) ngure, where some richIy appointed object or the deed
of a great being is intended as an indicative characteristic and not as an event
in itseIf (4.8);
17
that is, he ranks them just as Bhmaha and Dandin had
done, on the same IeveI as the expression of irony (paryayoktam) or the
description of providentiaI heIp (samahita). By the end of the ninth century
Pratihrendurja was confessing how markedIy the conceptuaI terrain had
shifted from the time of Udbhata: Whether the rasas and the emotions, given
that they are the source of the highest Iiterary beauty, are ornaments of
Iiterature or its very Iife force wiII not be a subject for consideration here
Iest it unduIy Iengthen the book
18
.
Pratihrendurja`s confusion (and unfortunate reIuctance to dispIay it)
was understandabIe, since onIy a few decades earIier, around the time of
Rudrata, Anandavardhana in his Dhvanyaloka (c. 875) had fuIIy assimiIated
rasa theory in the anaIysis of auraI-poetic Iiterature. AIthough according to
Anandavardhana`s Iinguistics of Iiterary communication not aII Iiterature is
concerned with the communication of rasa, when rasa is present it becomes
the centraI organizing component of the work, to which aII other features
must be subordinated (The nrst domain of action, for a good poet, is rasa)
19

That said, rasa itseIf has an astonishingIy undertheorized, taken-for-granted.
presence in Ananda`s treatise. He never actuaIIy teIIs us what it is or why it
shouId in fact be centraI to the Iiterary work, and he never expIains how the
15. This is intimated in Kavyadarsa 2.279, but not fuIIy deveIoped.
16. Both iyasu and matup are used in the sense of atisaya, see Ratnasrjna on
Kavyadarsa 2.237 (the printed text is corrupt, and is corrected in .' !"##"$%, Reader on Rasa:
An Historical Sourcebook of Classical Indian Aesthetics, New York, CoIumbia University
Press, forthcoming).
17. Kavyalankarajsara]samgraha 4.8 (0' &' 5-#234 (ed.), Kavyalankarajsara]samgraha of
Udbhata, with the Commentary of Pratihahrenduraja, Bombay, Nirnaya Sagar Press, 1928).
Bhmaha`s understanding of the ngure is nobiIity of character (3.11).
18. Kavyalankarajsara]samgraha 4.5.
19. Dhvanyaloka (!' .12.5&+ (ed.), Dhvanyaloka of Anandavardhana, with the Commentaries
of Abhinavagupta and Ramasaraka, Varanasi, Chaukhambha, 1940), p. 364. For the Iarger
question of the rise of a Iiterary teIeoIogy, see 0$$&-2, The Teleology of Poetics.
Sheldon Pollock
194
reader becomes aware of it or experiences it. This can onIy be because Ananda
had unquestioningIy accepted the cIassicaI view of rasa, as I suggested we
caII it, with its presuppositions and particuIar anaIyticaI focus. The focus and
presuppositions I can onIy touch on at the end of this essay; the important
point to stress here is that the assimiIation of rasa theory into poetry was now
an accompIished fact.
III. Eight, Nine. Many Rasas
The nrst consequence of the extension of rasa theory, from drama you
see into poetry you hear, that occurred graduaIIy over the course of the ninth
century is aIso one of the more specuIative. This concerns the nature of the
rasas and specincaIIy their number.
Lven the most supernciaI reading of the history of aesthetic theory wiII
register how, after a certain point, a dispute arose over whether there may be
additionaI aective states that can count as rasa; one of the few inteIIigent
books on rasa is in fact devoted to this probIem, V. Raghavan`s The Number
of Rasas. These states incIude, famousIy, santa, the tranquiI rasa, which was
the earIiest object of controversy, and bhakti, devotion, one of the Iatest, with
many other contestants for incIusion in between
20
. Indeed, from the moment
dramatic rasa nrst became poetic rasa, thinkers began to question the numer-
icaI Iimit that Bharata had pIaced upon the category. This starts with Rudrata,
whom we have identined as one of the earIiest theorists of rasa in auraI-poetic
Iiterature: Insofar as the teachers have identined certain emotions as rasas
because they can be tasted (rasanad), the way sweetness, sourness, and the
Iike can be tasted, other emotions such as despair shouId be rasas as weII, since
they, too, can by aII means be tasted
21
. And the questioning reaches a high
water mark in the work of Bhoja, whose masterpiece, the Srngaraprakasa,
engages with the probIem in its opening pages:
The conventionaI wisdom that rasa refers to the heroic, the fantastic, and the
remaining [six categories| has come out of nowhere and is hardIy more than a
superstition, Iike the beIief that a given banyan tree is haunted by a gobIin. It
20. Santarasa is mentioned nrst (outside passages in the Natyasastra added at a much Iater
date than the core materiaIs) in Udbhata`s Kavyalankarajsara]samgraha 4.4 (aImost certainIy
an interpoIated verse), but it is known to and accepted by Anandavardhana; bhaktirasa nrst
in Bhagavatamuktaphala of Vopadeva (c. 1300), ()' 712552$12&662 (ed.), jBhagavata]
muktaphala of Vopadeva, with the Commentary of Hemadri, CaIcutta, CaIcutta OrientaI Press,
1944), p. 164, where the co-author / commentator Hemdri, with refreshing candor, decIares,
Abhinavagupta and Hemacandra are wrong to deny that bhakti is a rasa (though it is not cIear
where in fact this supposed deniaI is made). Raghavan argues, with some justice, that The
advent of Snta Rasa seems to have set the writers thinking on the sanctity or otherwise of the
number eight or nine pertaining to the Rasa-s (,' &2412,23, The Number of Rasas, Madras,
Adyar Library, 1975, p. 118).
21. Kavyalankara 12.4. The text is cited approvingIy by Pratihrendurja (and
misinterpreted by him; contrast Namisdhu ad Ioc.) in Kavyalankarajsara]samgraha, p. 52-
53, and by Bhoja (Sjngaraprakasa, p. 633); and disapprovingIy by Dhanika on Dasarupaka
(5' ,-3%252$12&62 (ed.), Dasarupaka of Dhanajaya, with the Commentary of Dhanika and
the Subcommentary of Bhatta Njsimha, Madras, Adyar Library, 1969), p. 203-204.
From Rasa Seen to Rasa Heard
195
has onIy been accepted because of the inteIIectuaI conformity typicaI of the
worId, and our intention in this work is to put it to rest.
If aII emotions are equaIIy rasas, concIudes Bhoja after citing Rudrata,
it makes no sense to appIy the technicaI terms the erotic rasa, the heroic
rasa and so on onIy to those eight stabIe emotions, desire and the rest, when
they are fuIIy deveIoped. One may do it but that wouId then be onIy a termi-
noIogicaI distinction
22
. Bhoja himseIf, accordingIy, adds such rasas as the
vaingIorious (uddhata, based on a new stabIe emotion garva, pride), the nobIe
(udatta or urjasvin, based on mati, sagacity), and the rasa of tenderness or
motherIy Iove (preyah or vatsalya, based on sneha, attachment, the Iast being
suppIemented by Iater thinkers with brotherIy Iove or friendship between
equaIs, and aection for a superior, a king for exampIe), as weII as the rasas
of autonomy, heteronomy, bIiss, and abatement
23
. StiII other thinkers heId
that any of the thirty-three transitory feeIings couId become rasa, from torpor
(alasya) to vindictiveness (amarsa) to resentment (asuya), and so on down
the aIphabet. As one tenth-century writer puts it, There is no mentaI state
(cittavjtti) that cannot achieve enhancement and become rasa
24
.
Bhoja assumed that the Iimitation on the emotions that couId count as
rasa originated in mere unfounded convention (the standard view of rasa
comes out of nowhere, rasaprasiddhih siddha kuto 'pi). But we know that
within the tradition, a distinction was earIy on drawn between what couId and
couId not count as rasa. Bhatta LoIIata argued, as Abhinavagupta reports,
that, aIthough rasas were potentiaIIy innnite in number, it was the opinion
of experts that onIy those Iisted by Bharata were capabIe of portrayaI on the
stage
25
. AIthough Abhinava agrees on the Iimited number of rasas (etavanta
eva ca rasah) his haughty dismissaI here (This IittIe bit of arrogance on
LoIIata`s part can be safeIy ignored) impIies that he himseIf feIt Bharata`s Iist
was restrictive, not (as LoIIata seems to have thought) by schoIarIy conven-
tion, but by the nature of things, though eIsewhere he tries to expIain the
restriction otherwise: These are the rasas, and there are nine and nine onIy.
It is either because these aIone subserve the four ends of man or provide
sustained pIeasure that this restricted number has become traditionaI
26
.
22. Srngaraprakasa, p. 2 and 633 (see aIso karika 11, p. 2); Sarasvatikanthabharana 5.23
(%' .2&02 (ed.), Sarasvatikanthabharanalankara of Bhoja, with the Commentaries of
Ratnesvara and Jagaddhara, Bombay, Nirnaya Sagar Press, 1924).
23. Sarasvatikanthabharana, p. 627 (svatantrya, paravasya, ananda, prasama). These
additions wiII not seem so odd when we consider that in 1859 an LngIish phiIosopher Iisted as
emotions property, power, and knowIedge (4' 023)#-&, Lmotion, in )' %' >&--)1-+0 (ed.),
Handbook of Psychology, I: History of Psychology, New York, WiIey, 2003, p. 157-175 (p. 158)).
24. The possibiIity that aII the transitory emotions can be rasas was nrst raised by Rudrata,
Kavyalankara 12.3-4 ; the quotation in the text is from his commentator Namisdhu ad Ioc.
25. Abhinavabharati, p. 292, I. 22.
26. Abhinavabharati, p. 335, I. 8 (sustained insofar as sthayibhavas are enduring, unIike
vyabhicaribhavas, or transitory emotions). See further beIow on the sthayitva of the sthayi.
Abhinava reviews a range of opinion on why the transitory emotions are thirty-three in number,
incIuding the view that the set is reaIIy an open one ; is meant mereIy for pedagogicaI purposes ;
or is actuaIIy restrictive and driven by aesthetic concerns (Abhinavabharati, p. 373).
Sheldon Pollock
196
For others, however, the Iimitation was made on the basis of the criteria that
distinguish sravyakavya from djsyakavya. This is particuIarIy cIear from the
treatment of santarasa, for which the argument of Dhanika (c. 975) is as
forcefuI as any: Drama consists of representation, and by no means can we
accept that in drama quiescence [sama, the stabIe emotion of santarasa| can
function as a stabIe emotion. Quiescence signines cessation of aII activity and
so cannot have any connection with representation
27
. To be sure, santarasa
is a pecuIiar, even paradoxicaI, case emotionIess emotion and Dhanika`s
judgment was not to go uncontested. But the impIication here, however faint,
that the typoIogy of rasa was tied up with the typoIogy of Iiterature and the
distinguishing features of each of Iiterature`s sub-species, points the way
toward a fuIIer anaIysis.
IV. The Science, and History, of Emotions
It is not from the Indian tradition itseIf that we can derive this anaIysis
the avaiIabIe data cannot, I beIieve, take us beyond where they have taken
us so far but rather from research in cognitive science and studies in the
science of the emotions. The point of adducing such perspectives is not to
seek to penetrate to a scientinc core of the truth of emotion that exists entireIy
outside of its history. As wiII become evident, emotions Iike Iife in generaI
are historicaIIy contingent. But Iike Iife in generaI emotions have certain reaI
and constant properties, and the point of turning to a science of emotions is
to ask whether there may be anything more fundamentaI about the nature of
emotions that can heIp us uncover the conceptuaI foundations of any given
history, such as that of the traditionaI Indian, which presents a number of
pecuIiar features.
The history of emotion as embedded in the theory of rasa presents us not
onIy with a conict over what is aIIowed to count as a dominant emotionaI
register in Iiterature, but aIso with a very specinc Iist of what those regis-
ters must be. RecaII for a moment the ceIebrated cataIogue of eight stabIe
emotions in the Natyasastra: (sexuaI) desire, amusement, grief, anger, energy
(or endurance), fear, revuIsion, and amazement. This Iist is puzzIing in various
ways, but consider onIy the foIIowing two. Why shouId anger, which forms
the basis of the vioIent rasa (raudra), gain entry onto the Iist but not hatred,
sureIy an emotion as stabIe or primary as anger or any other in the group
for Indians (as for Descartes and the moderns mentioned beIow)?
28
And why
shouId sexuaI desire (rati), the basis of the erotic rasa (sjngara), be incIuded,
27. Dasarupaka, p. 202. That santarasa was perfectIy acceptabIe in non-dramatic Iiterature
was a Iong-estabIished view, see ,' &2412,23, The Number of Rasas, p. 52-53 (Raghavan,
engaging as he did with the tradition as if it were a Iiving one, was Iess interested in understanding
the grounds for santa`s excIusion than in vindicating its incIusion, To grant it in Kvya and to
deny it in Ntya is as cIumsy a compromise. p. 54).
28. Think onIy of SisupIa, who attained moksa not onIy despite hating Krishna
(Bhagavatapurana 10.29.13), but because of it (Bhagavatapurana 10.74.46). Note that Nyya
Iists hatred among the three dosas (raga, dvesa, and moha), and as one of the atmagunas, or
properties of the seIf (see n. 33 beIow).
From Rasa Seen to Rasa Heard
197
whiIe non-sexuaI aection (sneha), the basis of motherIy Iove (vatsalya), is
excIuded? If we are to understand the criteria at work in this cIassincation,
and what might account for Iater disagreement, we need to have some sense
of what dennes an emotion as stabIe to begin with.
UnfortunateIy, rasa texts themseIves rareIy oer any expIanation, and the
thinkers who denne the concept do so other than psychoIogicaIIy or perceptu-
aIIy. Dhanajaya (c. 975) seems to think in pureIy Iiterary terms:
A stabIe emotion is one that is uninterrupted whether by conicting or non-
conicting emotions. On the contrary, it subsumes other emotions, as the
ocean subsumes rivers.. Despair and the other transitory emotions do not
have that feature, and therefore cannot be stabIe emotions and cannot be
savored.
He impIies that stabIe emotions are stabIe because they cannot be interrupted,
or expunged in a psychoIogicaI sense, but onIy says (as his commentator
Dhanika onIy says) that they are not interrupted, or dispIaced in a Iiterary
sense
29
. Abhinavagupta, for his part, conceives of the stabiIity of the stabIe
emotions in what we might caII ethicaI terms, as we nnd in his discussion of
inessentiaIity (apradhanata), the sixth of the seven impediments (vighnas) to
aesthetic experience:
No one`s awareness can come to rest upon something that is nonessentiaI
(apradhana), since the moment that inessentiaI thing is cognized it hastens
after something more essentiaI, and cannot come to rest in itseIf. The most
essentiaI aesthetic components are those severaI forms of consciousness that
pertain to the ends of man, that is, Iove, power, Iaw, and Iiberation. Thus, the
stabIe emotion of desire pertains to Iove
30
as weII as to forms of power and Iaw
that are necessariIy reIated to Iove; the stabIe emotion anger pertains to power
among those given to anger, and can even eventuate in Iove or Iaw; the stabIe
emotion energy can eventuate in any of the ends of man, Iaw and the rest; and
Iast, quiescence, when it is the stabIe emotion and consisting IargeIy in dispas-
sion brought about by true knowIedge, is the means of Iiberation. Hence, these
stabIe emotions are the most essentiaI
31
.
However eIegant Abhinava`s correIation of stabIe emotions with the ends
of man, there is no evidence that it corresponds to anything in the conceptuaI
structure of the Natyasastra or informed its cataIogue of stabIe emotions
32
.
StabIe emotions are of course structuraIIy contrasted with transitory emotions
29. Dasarupaka 4.34, 36 (Dhanika`s comment here is fascinating, but does nothing to cIarify
the stabiIity of the sthayin).
30. I read kama- (for kamah or [KA| kame). The editors of the Natyasastra and the
corresponding passage in the Kavyanusasana mispunctuate the passage as a whoIe.
31. Abhinavabharati, p. 275-276.
32. The correIation was however aIready known to Pratihrendurja (on Kavyalankarajsara]
samgraha 4.3-4). There is additionaIIy both a gender and a social inection to the stabIe
emotions, from the time of the Natyasastra itseIf, that I can onIy register here: for exampIe,
onIy women and adhamaprakjti, or persons of the Iowest sociaI order, feeI fear ; uttamaprakjti,
persons of the highest sociaI order, onIy feign being afraid, for exampIe in the face of a
Sheldon Pollock
198
(vyabhicari- or samcari-bhavas) but the dennition of the Iatter brings us no
cIoser to understanding the stabiIity of the former. (In fact, they compIicate
what preciseIy a bhava is, incIuding as they do such physicaI states as iIIness,
dying, torpor, numbness, sIeeping, waking, and exhaustion.)
The idea of an emotion Iist of the sort we nnd in Bharata is rare in
systematic thought outside the discourse of rasa itseIf. To be sure, Nyya and
Smkhya cataIogue emotions, as does Buddhism, of which Abhinava oers a
brief, characteristicaIIy irreverent review when noting the schoIarIy discom-
fort with the Iist of thirty-three transitory emotions:
Some schoIars wonder how anyone couId possibIy cataIogue aII the various
states of mind (cittavjtti). And they ask with respect to their enumeration
how any given number couId capture this totaIity, whether the nine quaIities
of the seIf Iogicized by Iogicians, the eight properties of the inteIIect num-
bered by the numeroIogists of Smkhya, or the four types of apprehension
(error and the Iike), or the duaIity mind and mentaI activities broadcast
by Buddhists
33
.
But in fact, the phiIosophicaI systems show IittIe reaI concern with the
emotions, making no attempt to justify their Iists or indeed to more narrowIy
distinguish among the various items
34
. This reIative unconcern seems
especiaIIy odd in Iight of the interest in preciseIy this question shown by
Western thinkers. We have been oered Iists constructed on the basis of
what are no doubt radicaIIy diering physioIogies, epistemoIogies, and
moraIities, and for radicaIIy dierent purposes by everyone from AristotIe
(a very Iong one in the Rhetoric, incIuding anger, miIdness, Iove, enmity,
fear, conndence, shame/shameIessness, benevoIence, pity, indignation, envy,
emuIation, contempt) to Descartes (who in his Iast work, Passions de l'me,
1649, cataIogues the passions primitives as wonder, Iove, hate, desire, joy,
transgression they may have made against a guru (Natyasastra, p. 347 ; Abhinavabharati,
p. 325).
33. Abhinavabharati, p. 373. NormaIIy Nyya speaks of eight gunas (among the twenty-four)
that pertain to the atma: buddhi, sukha, duhkha, iccha, dvesa, prayatna, dharma, adharma,
presumabIy Abhinava here adds samskara. The Smkhyas` eight are dharma, adharma,
jana, ajana, vairagya, avairagya, aisvarya, anaisvarya, and the four types of apprehension,
viparyaya, asakti, tusti, and siddhi (see aIso &' !' %234#-, Rasa-Bhava-Vicara, Maharashtra
Rajya Sahitya-Samskrti MandaI, Bombay, 1973, p. 438).
34. Buddhist Abhidharma, which provides Iists in profusion, may be an exception but it is
concerned more with ethicaI dispositions than with what contemporary psychoIogy wouId
identify as emotion (see 4' )&-6><., Asian Perspective: Indian Theories of Mind, in
!' ?-#2?" et al. (eds.), The Cambridge Handbook of Consciousness, Cambridge, Cambridge
University Press, 2007, p. 89-114 (p. 100)). NonetheIess, the Iists of klesas, asravas, kasayas,
anusayas, and other mentaI states merit attention for an inteIIectuaI history of rasa of the sort
they have not received. Indeed, there is no comprehensive account of emotion in IndoIogicaI
Iiterature, or even an adequate historicaI psychoIogy dierentiating the functions of manas,
buddhi, antahkarana, citta, hjdaya (where rasa is often said to exist), and so on.
From Rasa Seen to Rasa Heard
199
sadness) and Spinoza (who in his Ethics reduces a[ectus to pIeasure, pain,
and desire, identifying the Iast as the very core)
35
.
One of the striking if conspicuous things about these Western Iists is their
variabiIity. Not onIy does the scientinc capacity for precision in describing the
human meet something of its Iimit in demarking the subtIe gradations of the
emotions, but emotions change over time. We have more and more evidence
of the fact that new emotions emerge, and oId emotions disappear, sIowIy no
doubt but assuredIy. In Western Lurope, meIanchoIy is a phenomenon of the
post-Reformation era, boredom of the eighteenth century, and (perhaps not
unreIatedIy), romantic Iove of the nineteenth. We know that emotions may
fade to the point of vanishing (for exampIe, the sense of honor in contem-
porary Iife; consider aIso the historicaI uctuations in the nature of maIe
grief), and may be variabIy distributed across cuItures or across historicaI
epochs (for exampIe, shame and guiIt)
36
. Then again, there do seem to be
striking continuities. In cIassicaI Chinese thought, for exampIe, the cataIogue
of emotions seems very famiIiar. The Book of Rituals Iist the qing as joy and
anger, sorrow and fear, Iove, aversion and desire; whiIe traditionaI Chinese
medicaI theory and therapy Iist the seven qiqing liuyu or emotionaI states/
aects as joy, anger, anxiety, thought, grief, fear, and fright
37
.
There exists to date no historiography of Indian emotion, Iet aIone one that
might have some bearing on Bharata`s Iist or the notion of stabiIity, though
I think one couId be written. It can certainIy be argued that, for exampIe,
bhakti, or devotion to a personaI god, is a new aective state nrst discernibIe
in Iate-epic India (though it attains cuIture-wide, and expIosive, inuence onIy
a miIIennium or more Iater)
38
. No Iess interesting if more specuIative is the
history of karuna (which becomes the rasa karuna)
39
. The term is sometimes
35. Ethica III. cupiditas est ipsa hominis essentia (/' ,23 ,#"5-3 /' !' 3' #23) (eds.), Benedicti
de Spinoza opera quotquot reperta sunt, 2 voI., The Haag, Martinus Nijho, 1882), p. 172.
Compare Bhoja`s dennition of sjngara, or passion: sarvatmasampadudayatisayaikahetuh (the
soIe cause of the appearance and added potency of the entire range [of emotions| of the seIf),
Srngaraprakasa, p. 2 (for Bhoja`s exegesis of the karika, see p. 375).
36. The bibIiography on the history of emotions in the West and the history of attempts to
cataIogue them is vast, and I am a very inexpert guide. For generaI trends see !' .5-2&3.,
History of Lmotions: Issues of Change and Impact, in 0' #-*+. et al. (eds.), Handbook
of Emotions, 3
rd
ed., New York, GuiIford Press, 2008, p. 17-31 ; 4' 023)#-&, Lmotion;
4' 7&<3 )' %<-3?#-= @Introduction: A New RoIe for Lmotions in LpistemoIogy, in 4' 7&<3
et al. (eds.), Epistemology and Emotions, AIdershot, Ashgate, 2008, p. 1-31. !' 2&+A.= The
Hour of our Death, New York, Vintage Books, 1982, presents evidence of maIe grief in the
MiddIe Ages that wouId have been famiIiar to medievaI Indians (judging from a work Iike
the Uttararamacarita) but aIien to our contemporaries. Lnnui, or French boredom, may be
somewhat Iater than the LngIish variety.
37. 1' -+>&+34 (ed.), Love and Emotions in Traditional Chinese Literature, Leiden, BriII,
2004, p. 1, 13, 23 ; Benjamin LIman (personaI communication).
38. I am unaware, however, that anyone has actuaIIy made this argument. Note that devotion
was an emotion for Darwin (discussed beIow), with its own distinct historicaIity. See aIso
)' 4&"..= Defending the Humanities with CharIes Darwin`s The Expression of the Emotions
in Man and Animals (1872), Critical Inquiry 37 (2010), p. 34-59 (p. 49).
39. The two forms were nrst distinguished by Sr Sankuka (Abhinavabharati, p. 311).
Sheldon Pollock
200
transIated in Luropean Ianguages as pity or compassion, but what emotion
is it reaIIy pointing toward? Two quite dissimiIar ones, I beIieve, for two
dierent emotionaI communities
40
. In the Uttararamacarita, for exampIe,
the paradigmatic text of karunarasa (eko rasah karuna eva, says Bhavabhti,
3.48, meaning of course the singIe rasa in his drama), Rma does not feeI
pity for St because he beIieves she has been wronged or wants to reIieve her
suering; the emotion is not primariIy outward-directed sadness at aII. On
the contrary, the pity wouId seem rather to be Rma`s sadness for himseIf, for
the coIIapse of everything he suered for, above aII his kingship, and for the
Ioss of the sons required to heIp him repay the debt to his ancestors (6.28 [8|).
Lven if such a formuIation may be thought too reductive, the cIassicaI
theory can hardIy be said to concern itseIf with compassion according to
the dictionary dennition: pity for the suerings or misfortunes of others,
tout court. For the feeIing of karuna to become the rasa karuna the person
Iost must be, as Bharata says expIicitIy and repeatedIy, an istajana, someone
beIoved to the subject; it is an emotion in which one`s seIf remains centraI.
The karuna rasa that arises when someone grieves for a person not reIated
(bandhu) to him, as Abhinava states, is [a semblance of karuna and hence
is| itseIf comic
41
. We think of karuna more broadIy as we do onIy because
the earIy Buddhists appropriated and transvaIued the concept as they appro-
priated and transvaIued so much of the dominant episteme. One might even
say the Buddhists redenned the very concept bandhu so as to comprise the
whoIe worId, thereby turning karuna into the active, bIind, aImost irrationaI
compassion so exuberantIy iIIustrated in the jataka taIes. As Dr. Johnson put
it, Pity is not naturaI to man. Pity is acquired and improved by the cuItiva-
tion of reason. We may have uneasy sensations from seeing a creature in
distress, without pity: for we have not pity unIess we wish to reIieve them
42
.
To see a creature in distress and to strive to do everything, even at the cost
of one own Iife, to reIieve that suering was, once upon a time, something
new in India. It was the Buddhists who invented compassion and this is
decidedIy not the karuna of aesthetic discourse
43
. Indeed, it is aIso by no
40. See /' !#20!-&, The History of Lmotions, History and Theory 49 (2010), p. 237-265
(p. 252) for a brief dennition, and 7' 1' &".-3*-+3, Emotional Communities in the Early
Middle Ages, Ithaca, NY, CorneII University Press, 2006, for a fuII exposition.
41. See Natyasastra 6.62 and Abhinavabharati, p. 290 (note that the vibhavas for karuna are: a
vioIation of dharma ; Ioss of one`s weaIth, and death of one`s kin, 6.78 with Abhinava). Abhinava
dismisses Sr Sankuka`s idea that karuna has anything to do with daya, compassion (p. 311).
42. /' 7".*-##, Johnson's Table Talk: a Selection of His Main Topics and Opinions Taken
from Boswell's Life' and Arranged by W.A. L. Bettany, London, BIackie & Son, 1904 (p. 20).
Hume (in The Treatise of Human Nature) oers a strikingIy dierent, and uncharacteristicaIIy
Christianized, view of pity.
43. A fuIIer argument wouId need to make sense of the history and nature of the idea of
the dayavira, the hero of compassion, nrst discussed in Dhvanyaloka (p. 394) in reIation to
santarasa (in fact, Abhinava and others cIaim dayavira is onIy another name for santa; see
,' &2412,23, The Number of Rasas, p. 85), but nrst exhibited in Nagananda, Hara`s Buddhist
drama. Note that the Natyasastra speaks onIy of the dharmavira (6.79).
From Rasa Seen to Rasa Heard
201
means cIear that what the viewer was supposed to feeI is compassion, rather
than something cIoser to misery
44
.
Pity therefore has a history in India, as no doubt many other emotions do
in many other pIaces, and this history sureIy has a bearing on the aective
aesthetics of Bharata. But there is something eIse more directIy pertinent to
this aesthetics that the science of emotions has to teII us.
V. Seeing is Believing
However vastIy documented, or at Ieast documentabIe, the historicity
of emotions may be, the diversity reveaIed by this history has not stopped
contemporary psychoIogy from trying to deveIop a set of the basic universaI
emotions; on the contrary, the overIaps have encouraged the eort. There
have been Iists gaIore and more recentIy Iists of Iists, incIuding one that
reports Iists of three emotions (fear, Iove, and rage, Watson in 1930), four
(expectancy, fear, rage, and panic, Panksepp in 1982; or fear, anger, depres-
sion, and satisfaction, Kemper in 1987), nve (happiness, sadness, anxiety,
anger, and disgust, OatIey in 1987), six (anger, disgust, fear, joy, sadness, and
surprise, Lkman in 1982), nine (fear, anger, distress, disgust, interest, shame,
joy, surprise, and contempt, Tompkins in 1962-1963), ten (anger, contempt,
disgust, distress, fear, guiIt, interest, joy, shame, and surprise, Izard in 1971)
45
.
This state of aairs has naturaIIy Ied some to doubt the very attempt to reduce
so compIex a mass of phenomena to a set of eIementary particIes (there are
over 300 words in the LngIish Ianguage that refer to emotions),
46
confronting
us as it does again with the tension between a quest for scientinc order and the
messiness of Iife in history.
The Indian materiaIs do not heIp us resoIve that tension, and it is not my
intention anyway in adducing them to try to do so the usuaI, and usuaIIy
useIess, maneuver of attempting to upstage Western science by a wiser
Lastern pre-science. What I do aim to achieve by pIacing the Indian data
in the context of a cognitive approach to emotion I repeat myseIf here Iest
I be misunderstood is to determine whether the contemporary method for
identifying basic emotions might suggest anything pertinent about the Indian
inteIIectuaI history of rasa as it moved from the seen to the heard.
Perhaps the most weII-known of the Iists of emotions is the set of six
noted above that was deveIoped by PauI Lkman (since revised by the rather
44. See Gadamer`s discussion of Greek leos in AristotIe, for which he beIieves the correct
German transIation is not Mitleid but Jammer (in Truth and Method, New York, Continuum,
1996, p. 130), though others demur (see 0' !"1#-3?= Furcht und MitIeid? Lin Nachwort,
Hermes 84 (1956), p. 49-74). (I thank Andrew OIIett for reminding me of the Gadamer passage.)
45. The Iist of Iists is adapted from Ortony and Turner, cited in &' $' ."#"0"3, Back to
Basics: On the Very Idea of Basic Lmotions, Journal for the Theory of Social Behaviour 32
(2002), p. 115-144 (p. 123), pIus &' #-6., How Did Fear Become a Scientinc Object and What
Kind of Object Is It ?, Representations 110 (2010), p. 66-104 (p. 68).
46. 4' 7&<3 )' %<-3?#-= @Introduction, p. 7.
Sheldon Pollock
202
aIarming addition of seven)
47
. What has been most inuentiaI in Lkman`s
work is his insistence that distinctive signaIs, more particuIarIy, faciaI
conngurations, are fundamentaI to the dennition of basic emotion. It does
not matter to my argument whether or not such data are presented in the
attempt to prove the existence of transcuIturaIIy constant, indeed neuroIo-
gicaIIy based, universaI emotions. What does matter is that these signaIs
and conngurations, for much contemporary research, are considered key to
determining what is basic about basic emotions, however many there may
be, since other emotionaI traits and moods do not manifest themseIves in
this way. In a word, whiIe there may very weII be debate about the meaning
and heuristic vaIue of physicaI expression in the science of the emotions,
there is no debate that some emotions are indeed physicaIIy expressed and
some are not.
As Lkman`s acknowIedges, his theory and in fact the heart of his method
were deepIy inuenced by CharIes Darwin`s The Expression of the Emotions
in Man and Animals (1872). The aim of this cIassic work was to corroborate
the theory of evoIution by demonstrating how the expression of emotions was
constant across species
48
. It is Iess any particuIar Iist of basic emotions that
interests me here Darwin actuaIIy oers no nxed set, though he does speak
of chief emotions and discusses amusement, fear, suering, rage, indigna-
tion (moderate anger), astonishment, disgust, and contempt or disdain
49
than
the method by which emotions are to be investigated. CentraI for Darwin in
making his case was visible expression, in particuIar faciaI expression, which
he presented with photographic and other iIIustrations. There are some
emotions that are basic; they are identined with those that dispIay more or
Iess nxed, more or Iess automatic. manifestations
50
. The emotions that are
basic are the emotions that you can see.
There are other emotions that you cannot see. They certainIy exist, and
powerfuIIy so, but they cannot be considered basic because they eIicit no
action on the subject`s part. Darwin considers these in a section of Emotions
caIIed Contrast between the emotions which cause and do not cause expres-
sive movements. The nrst he addresses is motherIy Iove: No emotion is
stronger than maternaI Iove, Darwin expIains, but a mother may feeI deepest
Iove for her heIpIess infant, and yet not show it by any outward sign. And
with this he contrasts sexuaI Iove: The Iove between the opposite sexes is
wideIy dierent from maternaI Iove. for this Iove is not inactive Iike that of
a mother for her infant. SimiIarIy, hatred (Iike suspicion, envy, or jeaIousy)
does not Iead to action and is not shown by any outward sign. Hatred is
enacted by rage, which however is pIainIy exhibited. Neither painters nor
47. !' -%023= Basic Lmotions, in 5' )2#4#-+.1 0' !"*-& (eds.), Handbook of Cognition
and Emotion, Chichester, WiIey, 1999, p. 45-60 (p. 55).
48. See )' 4&"..= Defending the Humanities.
49. $' )2&*+3, The Expression of the Emotions, p. 361-364.
50. &' $' ."#"0"3, Back to Basics, p. 116.
From Rasa Seen to Rasa Heard
203
poets are abIe to portray such emotions as hatred and maternaI Iove except by
association with their accessories
51
.
It shouId be noted that whiIe Darwin may have emphasized in his
account of the visibIe what the then-revoIutionary technoIogy of photography
permitted him to capture and reproduce, nameIy faciaI expression (a narrow
assessment in which, rather curiousIy, he has been foIIowed by contemporary
schoIars such as Lkman),
52
he was actuaIIy interested in the totaI physicaIity
of the emotions. He thus speaks generaIIy of the physicaI movement of an
emotion, or movement of expression: Terror causes the body to trembIe.
The skin becomes paIe, sweat breaks out, and the hair bristIes. The
breathing is hurried. The heart beats quickIy, wiIdIy, and vioIentIy; When
Iovers meet, we know that their hearts beat quickIy, their breathing is hurried,
and their faces ush; for this Iove is not inactive Iike that of a mother for her
infant
53
.
In this history of the science of emotion we can nnd, I suggest, some
cIues to heIp us understand the inteIIectuaI history of aesthetics in India. The
Indian data show that it was originaIIy in the context of drama, djsyakavya or
Iiterature-that-is-seen, that the distinctive Iist of stabIe emotions and their
associated rasas was conceptuaIized. The foundationaI Iogic of dramatic
(and dramaturgicaI) emotion comprises onIy emotion that can be physicaIIy
represented, not of course just in faciaI expression but in the whoIe host of
vocaI and physicaI reactions and cues (anubhavas), in physicaIIy observabIe
movement, that the Natyasastra is at pains to teach the actors who were
its principaI readership
54
. (This context of identincation wiII prompt us to
rethink the physicaIity of utsaha, the basic emotion of the heroic rasa: it is Iess
fortitude as a moraI virtue than something Iike embodied determination.)
55

51. C. )2&*+3! The Expression of the Emotions, p. 78-79, see aIso 215 (and )' 4&"..=
Defending the Humanities, p. 50-51). Note that Lkman regards hatred and parentaI Iove as
emotionaI pIots (Basic Lmotions, p. 55), more enduring than emotions, which (in contrast
to the Indian sthayibhavas) he thinks of as short-Iived (p. 50).
52. See !' -%023 *' ,' >&+-.-3, Unmasking the Face. A Guide to Recognizing Emotions
from Facial Clues, LngIewood CIis, N, Prentice HaII, 1975. DeveIoped from this work is
a FaciaI Action Coding System used by cognitive scientists, see )' 4&"..= Defending the
Humanities, p. 41.
53. $' )2&*+3, The Expression of the Emotions, p. 77-79.
54. For exampIe, The erotic is to be represented by reactions

such as the skiIIfuI pIay of
the eyes, movements of the eyebrows and sideIong gIances, and gentIe and pIeasing bodiIy
motions and verbaI utterances (Natyasastra, p. 293). R. A. Shweder et al. have aIso recognized
the paraIIeI between Bharata`s Iist of basic emotions and that of contemporary psychoIogy,
but they adduce it basicaIIy to questions any simpIe equation There is no neatIy bounded
set of universaI faciaI expressions (&' 2' .1*-)-& et al.= The CuIturaI PsychoIogy of the
Lmotions; Ancient and Renewed, in 0' #-*+. et al. (eds.) Handbook of Emotions, 3
rd
ed.,
New York, GuiIford Press, 2008, p. 409-427 (p. 412)); R. Scheckner, by contrast, sought to
correIate photographs of faciaI expression with the sthayibhavas to posit a universaI pattern
(&' .$1-$%3-&= Performance Theory, New York, RoutIedge, 1988).
55. Bharata`s own description of the physicaI representation of virarasa is obscure enough
(By boIdness, heroism, steadfastness, energy, audacity, and magnincence, and by statements
Iaden with doubIe meanings, is the heroic rasa properIy represented, Natyasastra 6.68) that
Sheldon Pollock
204
With the extension of rasa theory to sravyakavya, Iiterature-that-is-heard
a conceptuaI innovation that occurred nearIy haIf a miIIennium after the
core ideas of the Natyasastra were formuIated this Iogic was weakened
or even Iost. In the eyes of the subsequent tradition, beginning with Rudrata
(and to the despair, instinctive and not cIearIy reasoned, of the custodians of
dramaturgicaI theory such as Dhanika, who rejects Rudrata out of hand), the
Iist came to appear arbitrary or even senseIess. AccordingIy, among those
thinkers Iike Bhoja for whom poetic representation took precedence over
dramatic representation, there seemed to be no reason because in fact there
was no Ionger any reason not to incIude additionaI emotions and rasas, such
as vatsalya, motherIy Iove.
VI. Summary and Conclusions
The number of the aesthetic emotions, and the very idea of what kinds
of emotions couId become aesthetic, were transformed when the concept
of rasa was extended from Iiterature seen to Iiterature heard. The originaI
pragmatic ground of what couId count as rasa emotion that can be made
perceptibIe through acting was no Ionger understood and in any case was
no Ionger required for Iiterature where everything occurred in the mind`s eye
vastusaktimahimna, by the power of narrativity. The visibiIity of emotion
does not of course exhaust the signincance of the sthayibhava Iist; it can be
anaIyzed in many other ways, such as Abhinavagupta`s moraI map, once these
emotions came to be Iinked with the ends of man. But visibiIity, in service of
a theory not of psychoIogy but rather of performativity, was the feature that
informed the Iist in the nrst pIace.
The number and kind of emotions that couId become rasa was not the
onIy conceptuaI transformation that accompanied this anaIyticaI integration
of drama and poetry. Another key probIem, though somewhat more obscureIy
tied up with it than the questions deaIt with here, concerns rasasraya, or the
Iocus where rasa was beIieved to reside. It makes perfectIy good sense that
earIy writers thought of rasa, at Ieast in their anaIysis of performance, as being
Iocated in the nrst instance in the character they couId see (the onIy dispute
among them was whether rasa arose in the character or was inferred or
manifested in him), whereas Iater theorists of auraI-poetic Iiterature, for
whom the character was no Ionger visibIe but rather heard, wouId naturaIIy
Iocate rasa in the reader. The former position was certainIy that of Bharata
schoIars eventuaIIy confessed compIete confusion. Thus Bhnudatta (c. 1500): The reactions
(anubhava), one might argue, have to be physicaI properties. if they are to give us some sense
of rasa, which is itseIf imperceptibIe, but steadfastness (dhairya) and energy (utsaha) [seen
as both sthayibhava and anubhava| are not such properties. True enough, but by the word
steadfast was meant the absence of physicaI movement, and by energy, things Iike tears
and horripiIation. Or we couId repIy that the physicaI reactions are of four sorts, and mentaI
reactions have been incIuded among them. Awareness of that mentaI reaction is what makes
cIear the particuIar rasa being reacted to. It makes no dierence whether that awareness is
mentaI or perceptibIe. See Rasatarangini 3.25 (.' !"##"$% (ed.), Rasatarangini of Bhnudatta,
in The Bouquet of Rasa and the River of Rasa, New York, New York University Press, 2009).
From Rasa Seen to Rasa Heard
205
and aII writers before c. 900, when Bhatta Nyaka tried to make phiIosophicaI
sense of this naturaI reIocation, and redirected the focaI point of the rasa
anaIytic away from the oId formaIism and toward a new reception theory, or
hermeneutics, of aesthetics
56
. The dimcuIty with this expIanation, however,
is that in the transitionaI period some thinkers concentrating on djsyakavya
such as Dhanika and Abhinavagupta foIIow Bhatta Nyaka, whereas some
who prioritize sravyakavya such as Bhoja return to the cIassicaI view.
However this particuIar dimcuIty may be resoIved, another set of
questions, this time of an epistemoIogicaI sort, was tied up with the shifting
ontoIogy of rasa as it moved from the seen to the heard. The most chaIIenging
and important case concerns the notion of (abhi)vyakti. For Anandavardhana,
who put the term in the Iiterary-criticaI vocabuIary, this was pureIy a
Iinguistic phenomenon, a sabdavjtti, the verbaI manifestation of the Iatent
meanings of a text, of which rasa is the most important. But this expIana-
tion for how rasa in the text was communicated became uninteIIigibIe when
Bhatta Nyaka reIocated rasa in the reader, as thinkers writing in his wake
Iike Dhanika cIearIy show
57
. AccordingIy, Iater alankarikas, at Ieast from the
time of Mammata if not Abhinavagupta (or perhaps Bhatta Nyaka himseIf),
and by a process aImost compIeteIy unregistered in western schoIarship and
in the Indian tradition itseIf, transformed (abhi)vyakti into a psychoIogicaI
phenomenon, a cittavjtti, the reveIation to the viewer/reader of his own
basic emotion
58
.
It remains uncIear whether the soIutions to these key questions are aII part
of the extension of rasa theory from drama to the wider worId of Iiterature.
What I hope to have at Ieast demonstrated, however, is that much of the inteI-
IectuaI history of this theory, India`s greatest contribution to worId aesthetics,
remains to be written
59
.
References
Primary Sources
&' 4' 712552 Bed.C= Kavyalankarasutra and Vrtti of Vamana, Varanasi, Braj B. Das,
89:;'
)' 712552$12&662 (ed.), jBhagavata]muktaphala of Vopadeva, with the Commentary
of Hemadri, CaIcutta, CaIcutta OrientaI Press, 1944.
56. .' !"##"$%, What was Bhatta Nyaka Saying ? The HermeneuticaI Transformation of
Indian Aesthetics, in .' !"##"$% (ed.), Epic and Argument in Sanskrit Literary History, DeIhi,
Manohar, 2010, p. 143-184.
57. Avaloka on Dasarupaka, p. 217.
58. I expIore this probIem in detaiI in Vyakti in the History of Rasa, Vimarsa: A Half-
yearly Research Bulletin of the Rashtriya Sanskrit Sansthan 6 (2012), p. 232-253. I say aImost
unregistered since a few sentences are addressed to the question in )' 1' 1' +342##.
/' 0' 02.."3 0' ,' !25*2&)123 (trans.), The Dhvanyaloka of Anandavardhana, Cambridge,
Mass., Harvard University Press, 1990, p. 35.
59. I thank YigaI Bronner and Andrew OIIett for their heIpfuI comments on an earIier draft
of this paper.
Sheldon Pollock
206
)<&42!&2.2) *' #' .' !23.+%2& (eds.), Kavyalankara of Rudrata, with the
Commentary of Namisadhu, Bombay, Nirnaya Sagar Press, 1928.
&' )*+,-)+ (ed.), Vyaktiviveka of Mahimabhatta, Varanasi, Chaukhambha, 1983.
&' )*+,-)+ .' )*+,-)+ (eds.), Srngaraprakasa of Bhojaraja, 2 voI., New DeIhi,
Indira Gandhi NationaI Centre for the Arts, 2007.
%' %&+.1320""&516 (ed.), Natyasastra of Bharata, with the Abhinavabharati of
Abhinavagupta, 4
th
ed., Baroda, OrientaI Institute, 1992.
/' 0"123 (ed.), Kavyaprakasa [of Mammata|, solah tikaom sahit, 6 voI., DeIhi, Nag
Prakashan, 1995-.
&' !+.$1-# (ed.), Rudrata's [sic| rngaratilaka and Ruyyaka's Sahrdayalila,

KieI,
C. F. HaeseIer, 1886.
S. !"##"$% (ed.), Rama's Last Act by Bhavabhti, New York, New York University
Press, 2007.
.' !"##"$% (ed.), Rasatarangini of Bhnudatta, in The Bouquet of Rasa and the River
of Rasa, New York, New York University Press, 2009.
%' .2&02 (ed.), Sarasvatikanthabharanalankara of Bhoja, with the Commentaries of
Ratnesvara and Jagaddhara, Bombay, Nirnaya Sagar Press, 1924.
!' .12.5&+ (ed.), Dhvanyaloka of Anandavardhana, with the Commentaries of
Abhinavagupta and Ramasaraka, Varanasi, Chaukhambha, 1940.
0' &' 5-#234 (ed.), Kavyalankarajsara]samgraha of Udbhata, with the Commentary
of Pratihahrenduraja, Bombay, Nirnaya Sagar Press, 1928.
2' 512%<& <' /12 (eds.), Kavyalaksana [= Kavyadarsa| of Dandin, with the
Commentary of Ratnasrijana, Darbhanga, MithiIa Institute of Post-graduate
Studies and Research in Sanskrit Learning, 1957.
/' ,23 ,#"5-3 /' !' 3' #23) (eds.), Benedicti de Spinoza opera quotquot reperta
sunt, 2 voI., The Haag, Martinus Nijho, 1882.
5' ,-3%252$12&62 (ed.), Dasarupaka of Dhanajaya, with the Commentary of
Dhanika and the Subcommentary of Bhatta Nrsimha, Madras, Adyar Library, 1969.
Secondary Sources
!' 2&+A.= The Hour of our Death (trans. from the French by H. Weaver), New York,
Vintage Books, 1982.
/' 7".*-##, Johnson's Table Talk: a Selection of His Main Topics and Opinions Taken
from Boswell's Life' and Arranged by W.A. L. Bettany, London, BIackie & Son,
1904.
6' 7&"33-&, A Question of Priority: Revisiting the Bhmaha-Dandin Debate,
Journal of Indian Philosophy, in press.
4' 7&<3 )' %<-3?#-= @Introduction: A New RoIe for Lmotions in LpistemoIogy, in
4' 7&<3 et al. (eds.), Epistemology and Emotions, AIdershot, Ashgate, 2008, p. 1-31.
$' )2&*+3, The Expression of the Emotions in Man and Animals, London, John
Murray, 1873.
4' )&-6><., Asian Perspective: Indian Theories of Mind, in !' ?-#2?" et al. (eds.),
The Cambridge Handbook of Consciousness, Cambridge, Cambridge University
Press, 2007, p. 89-114.
1' -+>&+34 (ed.), Love and Emotions in Traditional Chinese Literature, Leiden, BriII,
2004.
!' -%023 *' ,' >&+-.-3, Unmasking the Face. A Guide to Recognizing Emotions
from Facial Clues, LngIewood CIis, N, Prentice HaII, 1975.
From Rasa Seen to Rasa Heard
207
!' -%023= Basic Lmotions, in 5' )2#4#-+.1 0' !"*-& (eds.), Handbook of
Cognition and Emotion, Chichester, WiIey, 1999, p. 45-60.
1'D4' 42)20-&, Truth and Method, New York, Continuum, 1996 [1
st
German ed.
1960|.
)' 4&"..= Defending the Humanities with CharIes Darwin`s The Expression of the
Emotions in Man and Animals (1872), Critical Inquiry 37 (2010), p. 34-59.
)' 1' 1' +342##. /' 0' 02.."3 0' ,' !25*2&)123 (trans.), The Dhvanyaloka of
Anandavardhana, Cambridge, Mass., Harvard University Press, 1990.
&' !' %234#-, Rasa-Bhava-Vicara, Bombay, Maharashtra Rajya Sahitya-Samskrti
MandaI, 1973.
&' #-6., How Did Fear Become a Scientinc Object and What Kind of Object Is It?,
Representations 110 (2010), p. 66-104.
4' 023)#-&, Lmotion, in )' %' >&--)1-+0 (ed.), Handbook of Psychology,
I: History of Psychology, New York, WiIey, 2003, p. 157-175.
#' 0$$&-2= The Teleology of Poetics in Medieval Kashmir, Cambridge, Mass.,
Harvard University Press, 2008.
/' !#20!-&, The History of Lmotions, History and Theory 49 (2010), p. 237-265.
0' !"1#-3?= Furcht und MitIeid? Lin Nachwort, Hermes 84 (1956), p. 49-74.
.' !"##"$%, Bhoja`s Srngaraprakasa and the ProbIem of Rasa: A HistoricaI
Introduction and TransIation, Asiatische Studien/tudes asiatiques 70 (1998),
p. 117-92.
.' !"##"$%, What was Bhatta Nyaka Saying? The HermeneuticaI Transformation
of Indian Aesthetics, in .' !"##"$% (ed.), Epic and Argument in Sanskrit Literary
History: Essays in Honor of Robert P. Goldman, DeIhi, Manohar, 2010, p. 143-184.
.' !"##"$%, Vyakti in the History of Rasa, Vimarsa: A Half-yearly Research
Bulletin of the Rashtriya Sanskrit Sansthan 6 (2012), p. 232-253.
.' !"##"$%, Reader on Rasa: An Historical Sourcebook of Classical Indian Aesthetics,
New York, CoIumbia University Press, forthcoming.
,' &2412,23, The Number of Rasas, Madras, Adyar Library, 1975.
7' 1' &".-3*-+3, Emotional Communities in the Early Middle Ages, Ithaca, NY,
CorneII University Press, 2006.
&' .$1-$%3-&, Performance Theory, New York, RoutIedge, 1988.
&' 2' .1*-)-& et al., The CuIturaI PsychoIogy of the Lmotions; Ancient and
Renewed, in 0' #-*+. et al. (eds.) Handbook of Emotions, 3
rd
ed., New York,
GuiIford Press, 2008, p. 409-427.
&' $' ."#"0"3, Back to Basics: On the Very Idea of Basic Lmotions, Journal for
the Theory of Social Behaviour 32 (2002), p. 115-144.
!' 3' .5-2&3., History of Lmotions: Issues of Change and Impact, in 0' #-*+.
etal. (eds.), Handbook of Emotions, 3
rd
ed., New York, GuiIford Press, 2008, p. 17-31.