J Indian Philos (2011) 39:41 62 DOI 10.

1007/s10781-010-9116-6
Padmasri s Nagarasarvasva and the World of Medieval Kamasastra
Daud Ali
Published online: 8 January 2011
Springer Science+Business Media B.V. 2011
Abstract This essay focuses on a neglected and important text, the Nagarasarvasva
of Padmasri, as an index to the changing contours of kamasastra in the early second
millennium (1000 1500) CE. Focusing on a number of themes which linked Padmasri s wo
rk with contemporary treatises, the essay argues that kamasastra incorporated sever
al new conceptions of the body and related para-technologies as well as elements
of material and aesthetic culture which had become prominent in the cosmopolita
n, courtly milieu. Rather than seeing this development as an attenuation of the
earlier science as constituted by Vatsyayana s Kamasutra, it is possible to see that k
amasastra actually developed closer relations with fields of knowledge that had lon
g developed alongside it.
Keywords Kamasastra Nagarasarvasva Ratirahasya
Pan casayaka Body technologies
Ae
s Ala_mkarasastra Sa_nketa
Hava
Bhava - Nayikabheda Conjugal love
Courtly lo
onomy
Physiognomy ~ Mantra Nad:
Introducing the text
Among the earliest texts after the composition of Vatsyayana s Kamasutra in the
Gupta period is the Nagarasarvasva or Complete Townsman , by Padmasri. The
Nagarasarvasva is a remarkable and little studied text, but one that is key to
This essay is based on a collaborative project involving a translation and study
of the Nagarasarvasva currently being completed by the author and Dr Mattia Salv
ini.
D. Ali ()
Department of South Asia Studies, University of Pennsylvania, 820 Williams Hall,
36th and Spruce Streets, Philadelphia, PA 19104, USA
e-mail: daudali@sas.upenn.edu
42
D. Ali
understanding the evolution of the genre and its relationships with other intell
ectual traditions in medieval India. The Nagarasarvasva (NS) consists of approxim
ately 330 verses divided into 38 chapters on a wide variety of different topics
(see Appendix in Desmond).1 Manuscripts preserve a short commentary to the work
attributed to the seventeenth-century Nepali king Jagajjyotirmalla (1617 1633 CE)
of Bhaktapur.2 To date, it has received very little scholarly attention, and tho
ugh noticed in Richard Schmidt s Beitrage zur Indischen Erotik and Aufrecht s Catalog
us Catalogorum, it has been the subject of only a single scholarly article.3 The
date of the text is uncertain. It seems to have been well-known throughout the
subcontinent by the fourteenth century, when it is cited and incorporated into a
variety of literary contexts, including anthologies, treatises on poetics and c
ommentaries on classical kavya works.4 On the other side, the most clearly dateab
le work cited in the Nagarasarvasva is the famous Kut:t:anmata of Damodaragupta, a
text usually placed at the end of the eighth century CE in Kashmir. Beyond this
rather unhappily wide span of time, ranging between c. 800 and 1300 AD, it is no
t at the moment possible to gain a more precise knowledge of when the text was c
omposed.5
Very little is also known about Padmasrn, author of the Nagarasarvasva. While cita
tions of his work suggest that he was known as an author widely across northern
India in the second millennium CE, the surviving manuscript traditions of the te
xt have been mostly confined to Nepal, where the text gained its only commentary
.6 This, in addition to comparatively positive remarks about the women of Nepala
(along with Cnna [Tibet] and Kamarupa [Assam]) in the Nagarasarvasva s section on the
regions (desavis:ayabhaga) prompted Lienhard to suggest that Padmasr
1 Some manuscripts divide the text into 18 chapters. See Tanusukharam Sharma s int
roduction to NS, pp. 16 17.
2 More useful, however, is Tanusukharam Sharma s learned Sanskrit t:ippan:.
3 Because of slight variations in the names of its title and author in the manus
cript traditions, Schmidt (1922, pp. 47 48, 55), following Aufrecht, postulated tw
o distinct texts the Nagarasarvasva of Padmapan:d:ita and the Nagarakasarvasva of Pa

dmasrnjn ana. Not long after the final edition of Schmidt s Beitrage was published, two
printed editions of the text appeared, one edited by the celebrated Bombay-base
d scholar, Tanusukharam Sharma, in 1921, and the other from Calcutta in 1929 by
Srirajadhara Jha. When Lienhard (1979) published the only scholarly article on t
he text some 50 years later, he was easily able to clarify that the various titl
es and authorial names referred to in manuscript traditions denoted one and the
same text and writer, the Nagara(ka)sarvasva of Padmasrn(jn ana) or Padmapan:d:ita.
4 Lienhard (1979, pp. 102 103 ) noted some sixty verses from the NS cited in the Sar
n:gadharapaddhati, an anthology compiled at the Sakhambari Chauhan court in 1363 AD
. A passage from the text is also cited in Visvanatha s Sahityadarpan:a (3.101) dated
between 1300 1380. The NS is also cited in Dinakara Misra s well-known commentary on
Kalidasa s Raghuvam:sa, the Subodhin , dated in 1385 AD (Schmidt 1922, p. 47). Finally
he text is also mentioned in Raghavabhat:t:a s famous Arthadyotanika commentary on Abh
ijn anasakuntala (1.27; 2.2) composed in the last quarter of the fifteenth century.
5 This has not stopped scholars from giving it more precise dates. Krishnamachar
iar (1974, p. 891), takes a golden mean between the upper and lower limits to pl
ace the text at 1000 AD with no apparent justification. More recently Zysk (2002
, pp. 6 8) seems to have followed Krishnamachariar.
6 Two of the three manuscripts used by Tanusukharam Sharma were of Nepali origin
, the other being from Bikaner. See Sharma s Introduction to NS, pp. 16 17.
Padmasri s Nagarasarvasva and the World of Medieval Kamasastra
43
may have been active in Nepal.7 Padmasr tells us nothing of his own family backgrou
nd, but does inform the reader that he composed the Nagarasarvasva after the repe
ated requests of a Brahman friend by the name of Vasudeva, who was learned in the
arts. Though we may surmise that Vasudeva was a Hindu and a Vaisnava, Padmasr was n
either. It is clear from the opening verse, which has a praise invocation to Man j
usr, as well as other passages in the text, that Padmasr was a Buddhist.8 While some s
cholars have contended that Padmasr was a monk, Lienhard, building on his hypothesi
s of Padmasri s Nepali origin, has suggested that he may have been a non-celibate N
epali vajracarya .9 Whatever the case, Padmasr takes his place in an established trad
ition of men formally committed in some capacity to renunciate orders who also w
rote or compiled works on the most worldly of subjects. Padmasr himself, in his chapt
er on jewels, may have relied heavily on one such source as the first extant treat
ise on the subject of gems and jewels in early India, the Ratnaparks:a, was compose
d by the Buddhist Buddhabhatta.10 In any event, celibacy would hardly preclude a
monk from writing on the subject of erotics. Vatsyayana himself claims to have wr
itten the Kamasutra in a state of chastity, for the sake those pursuing a worldly l
ife (lokayatra).11 Padmasr is even more emphatic, stressing that his work was written
from compassion, in order to help mankind pursue the goals of the trivarga.12
Among later kamasastra works, the Ratirahasya of Kokkoka and Padmasri s Nagarasarvasva
have been considered the earliest, with Yasodhara s Jayama_ngala (the most well-know
n commentary on the Kamasutra) and Jyotirsvara s Pan casayaka following slightly later.
The chronology of these works is uncertain. Like the Nagarasarvasva, the date of
the Ratirahasya remains elusive, though we know
7 See Lienhard (1979, p. 99), commenting on NS 20.11. More compelling than Lienh
ard s claim that this verse shows an extraordinarily positive judgement for the wome
n of Nepal, Kamarupa and Cina (compare with the equally positive remarks about wom
en from Si_mhala NS 20.4) is his point, noted in a footnote, that the Nagarasarva
sva seems to be the first text to introduce Nepala and Cina as regions in this co
nventionalised feature of kamasastra. As Lienhard points out, the Kamasutra includes
neither Kamarupa, Nepala nor Cna, while the Ratirahasya mentions only Kamarupa. The Sm
radpika, a text which otherwise seems to borrow from the NS, on the other hand, men
tions Kamarupa and Nepala (though not Cna). A tentative conclusion would be that the
Nagarasarvasva is the first text to introduce the northeastern regions of Nepal (
Nepala) and Tibet (Cna) to the genre convention of describing the sex practices of
the regions, a point which may (or may not) suggest a Nepali origin for Padmasr.
8 NS 1. 1, a pun in the verse also allows it to be dedicated to Kamadeva, who is
otherwise identified with Man jusr in Mahayana Buddhism. Later in the text, in connecti
on with actions to be performed by the man desiring the birth of a son, Padmasr rec
ommends first the worship (puja) of Tara, showing a preference for Buddhism, but als

o provides an alternative ritual for the trthika , a decidedly Buddhist term for a no
n-Buddhist, NS 38.6 11.
9 Lienhard (1979, p. 99). For earlier interpretations, see Shastri, (1901 1905 vol
. 1, p. 11).
10 See the remarks of McHugh in this volume and also Finot (1896, pp. ix x). Thoug
h the study of jewels
formed one of the auxiliary knowledge to be mastered by the nagaraka according to
Vatsyayana (KS
1.3.15) it is notable that the Nagarasarvasva is the only kamasastra to include a ch
apter on the subject.
11 KS 7.2.57 His commentator glosses this as for the sake of the conduct of the h
ouseholder .
12 NS 1.8 110.
13 Raghavan (1942, p. 167; 1978, pp. 639 643) argued that the Jayama_ngala should b
e dated before the reign of Bhoja of Dhara (1013 1055) whose list of sixty-four supp
orting arts conformed closer to Yasodhara s commentary than to Vatsyayana s list at KS 1
3.16. This dating has generally not been accepted because of clearer external ev
idence for the dating of Yasodhara to the thirteenth century.
44
D. Ali
that it too was used by commentators beginning in the thirteenth century (includ
ing Yasodhara, who seems to cite a verse from it). 14 The relative dating of the
Ratirahasya and Nagarasarvasva is difficult, and perhaps undecidable.15 The two t
exts also mention substantially different authorities (See Appendix). The only w
riter mentioned in both works is Vatsyayana; otherwise, they seem to draw from, or
at least acknowledge, different traditions. Padmasri refers to substantially few
er authorities than Kokkoka and mentions none of the more commonly cited names w
hich appear in the tradition from as early as Vatsyayana, like Babhravya, Gon:akipu
tra and Nandikesvara. The most frequently cited authority in the Nagarasarvasva is
Mahesvara, whose work does not seem to form part of subsequent kamasastra textual
traditions. 16 Given this variance, it is tempting to see the Nagarasarvasva as p
art of a distinct tradition of knowledge on the subject. Yet the text itself, de
spite a number distinctive features, shares much in common with both the Ratirah
asya and the bulk of later kamasastra traditions. And later authors like Minanatha
(fourteenth century) and Anantakavi (fifteenth century) draw freely from both wo
rks.17
The genealogy of authorities in different texts needs, of course, to be correlat
ed with an extended analysis of what might be called the core content of the kamasas
tra tradition, particularly the enumeration and nomenclature of different types
of kissing, embracing, biting, nail-marking, and coital positions. This is too l
arge a task to be taken up in this essay, and remains a desideratum, though the
initial work done in this direction in the topical sections of Schmidt s Beitrage w
ill remain an important source for such a study. Comfort s assertion, based on his
reading of Schmidt, that kamasastra texts fall into three literary pedigrees by the
ir content one group consisting of the Kamasutra and its later verse rendition, the
Kandarpacud:aman:; another by the Pan casayaka, Ana_ngara_nga and Ratirahasya; and a fi
nal group consisting of Smaradpika and Ratiman jar has limited
14 RR 3.8 would seem to be cited in the Jayama_ngala at KS 2.1.20 but no authorial
ascription is given and both may be quoting a prior source. For other citations,
see Schmidt (1922, p. 67) and Lienhard (1960, p. 19), who maintain the thirteen
th century as a lower date for the RR. Other scholars have placed the text much
earlier. Raghavan (1943, p. 72) on the basis of his dating of the Jayama_ngala (s
ee footnote 14) and on an apparent reference to the text in Somadevasuri s tenth-ce
ntury Yasastilaka, and Upadhyaya (1981, pp. 8 10; followed by Mylius 2009, pp. 13 14
) by a reference to the word kokkoka in the same author s Ntivakyamr:ta (25.16 25.17), d
ate the text to the ninth or tenth century. References in the works of Somadevas
uri, however, are ambiguous (in the case of Raghavan) and incorrect (in the case
of Upadhyaya and Mylius).
15 Lienhard (1979, p. 100) contends that Padmasr names the RR as one of his authori
ties in NS 38.16. Here he follows the 1929 Calcutta edition, which reads pratham
amuditamasmin vks:ya siddhaikavram_ paramaratirahasyam. But Tanusukharam Sharma s edit

ion, superior in its critical apparatus, reads prathamamuditamasmin vks:ya siddhai

kavram_ paramamatirahasyam, with one manuscript having the variant ... vramaparamati
rahasyam ... It is thus not possible, then, to secure an earlier date for the RR
based on evidence in the NS.
16 Save two possible references in PS 1.2 and 4.56. The twentieth-century Tibeta
n writer Gedun Chopel composed a kamasastra based in part on Sanskrit sources, one o
f which was the chags pai i bstan bcos (ragasastra?) by one Mahesvara, who may be the
same as Padmasr s source (Hopkins 1992, p. 36).
17 It is moreover significant that the teachings of the NS borrowed by these ant
hologies are not in any way unique to Padmasr, but are shared by the RR and PS, mos
t notably verses dealing with dividing women by age (NS 16.1 ff.). For NS verses
in the SmD of Minanatha, see Sharma (1921, Appendix gha ); in the Kamasamuha, see Pat
hak (2008, pp. 254 266).
Padmasri s Nagarasarvasva and the World of Medieval Kamasastra
45
value, not being nuanced enough to account for the multiple overlappings and dis
tinctions within and between these pedigrees .18
This essay will focus on broader changes in the genre through a comparison of a
single text, the Nagarsarvasva, with other treatises produced before 1500. It wil
l examine aspects of the text shared with other treatises including the categoriza
tion of lovers, physiological understandings of sexual pleasure, the use of mant
ras and allied knowledges for special purposes, and procreative sex as well as top
ics which seem to be unique to the Nagarasarvasva, like the use of ala_mkarasastra c
oncepts, the incorporation of knowledge on gems and perfumes, and the use of sig
ns and sign language. For the sake of presentation, the discussion may be organi
sed under two more or less broad axes of concern knowledges of the body and aesthe
tics.
Sex and Knowledges of the Body
Perhaps the most salient of feature of later kamasastra texts, as Desmond (in this
volume) has shown, is the introduction of new categories and criteria for the cl
assification of lovers, particularly women (nayikabheda). The early kamasastra divisi
on of men and women into six animal types, based on the size or depth of their s
exual organs came to be supplemented by later writers in important ways. The dis
course expanded in both its formal categories and descriptive topoi. Many later
treatises augmented Vatsyayana s basic model by introducing supplementary or replace
ment schema to the original six-fold division. Most notable among these new sche
ma was the fourfold division of women into padmin ( lotus woman ), citrin: ( many talente
woman ), sa_nkhin ( conch-shell woman ) and hastin ( she-elephant ), first introduced i
ka s Ratirahasya alongside Vatsyayana s scheme, but gradually overshadowing it in late
r kamasastra.19 Though absent in the Nagarasarvasva, which retains the older enumera
tion of Vatsyayana (potential evidence of its chronological priority to the Ratira
hasya), what unites most later kamasastric classifications of lovers (including the
enumeration in the Nagarasarvasva), is an expanded set of topoi forming the crit
eria for the major classifications. So, for example, while Vatsyayana describes on
ly the genital size of the male lover known as the hare, or sasa, Padmasri tells us
, among other things, that he is a beautiful, joyful and well-spoken man with ev
en teeth, a wide face, compact fingers, thin extremities, and with copious and f
ragrant semen .20 Overall, what we see is a gradual expansion of the diagnostic
features of nayikabheda including information like vaginal shape, sexual odor, body
type, physiognomy, complexion, hair-type, dressing style, eating habits, comport
ment, voice, and even ethical inclinations.
18 See Comfort (1964, p. 46).
19 See RR 1.10 1.19. The fifteenth-century Pan casayaka (1.6 1.9; 2.10 2.16) contains bot
h the older, three-type and newer, four-type divisions of women, but adds (unlik
e the Ratirahasya) a final category of men (the stag, or mrga) to balance the ge
nders, making a symmetrical eightfold scheme, and a number of still later texts
drop the six fold scheme altogether in favor of this eightfold typology. See, fo
r example, SmD 15, 24; also the Ratiman jar 3 9, 35 38; Ratisastra 17 51 and Ratiramana 3
31; 5.2 5.19.
20 NS 14.2; Cf. RR 3.31 3.32.
46
D. Ali

The sources of these changes are difficult to determine. They do not seem to hav
e been based on entirely internal elaborations, not least because of their increas
ing incommensurability with the older divisions of kamasastra.21 A similar prolifer
ation of divisions of lovers is also found in Sanskrit (and later, Hindi) ala~mk
arasastra, but the relationship between the two traditions is, despite a general ov
erlap of concern and vocabulary, rather surprisingly, somewhat weak.22 More like
ly sources for the criteria deployed in such divisions are ayurveda and especiall
y the heterogeneous knowledges of bodily prognostication, divination and physiog
nomy that had been developing in India throughout the first millennium CE. Texts
like the Gargasa~mhita (first century AD) and Br:hatsa~mhita (sixth century AD) c
ombined astronomical knowledge with a wide range of materials relating to physio
gnomy, divination, and sumptuary and included chapters on the characteristics (laks:
an:a) of men and women in the context of choosing partners suitable for marriage
. 23 Vatsyayana and later kamasastra authors also make reference to observing the mark
s of maidens (kanyalaks:an:a) in their advice to men in choosing a virgin for a po
tential wife.24 But more than this, the types of characteristics enumerated for
the basic divisions of lovers in later kamasastra seem to build upon the kinds of c
haracteristics detailed in the Br:hatsa~mhita, including bodily proportions, soli
dity and thickness of the limbs, physical compactness, skin color and complexion
, teeth, and even voice, walking style, and disposition though they often add elem
ents which were specific to kamasastra.25 Padmasri, for example, describes bull (vr:
s:a) type men as having a thick neck and a graceful walk, reddish hands and feet,
steady eyes with long eyelashes, a turtle-like
21 While the fourfold classification introduced in the RR was ostensibly distinc
t and separate from Vatsyayana s more restrictive typology, and could thus theoretic
ally function alongside it in complementarity, the scheme, over time, tended to
interfere with the earlier categorization, revealing a profound incommensurabili
ty, as Desmond has shown. The idea, presumed in the RR, that the schemes were co
mpatible is belied by the fact that numerous characteristics are shared by both,
raising the question of how they would articulate with one another. So the odor
of vaginal secretion (termed variously suratapayas, smarambu, ratisalila, etc.)
is given for female types in both schemes, suggesting, for example, that a woman
of the doe variety (having a depth of six a~ngulas), because her vaginal secret
ions smelled of flowers, could only possibly be a padmin (whose fluid was to smell
of the lotus), while a woman of the mare variety, with her vaginal secretions h
aving the scent of flesh (palala, or perhaps sesamum), could not be included at
all within the fourfold typology.
22 The PS (5.32 5.38) and SmD (178 188) both include brief iterations of the eight n
ayikas known from ala~mkarasastra. While Vatsyayana s theriomorphic classifications of
n and women are not taken up by writers in ala~mkarasastra, the eightfold theriomor
phic scheme of later kamasastra works seems to have been used by theorists in Hindi
literature. See Rakesagupta (1967, pp. 30 31; 139 140).
23 For a summary of the contents of the Gargasa~mhita and Br:hatsa~mhita, see Ping
ree (1981, pp. 69 75).
24 Zysk (2002, pp. 14 22) notes passages from the KS (3.1.13ff) and RR (11.1ff) wh
ich discuss the choice of a potential bride, to which may be added SmD 126 135 and
PS 4.1 4.8 (which includes the laks:anas of the bridegroom). This use of body div
ination becomes increasingly important with the growing conjugal emphasis of pos
t fifteenth-century kamasastra (or ratisastra, as Zysk mantains).
25 The Br:hatsa~mhita s relationship to kamasastra, as to other nascent fields of worl
dly knowledge like silpasastra, gandhasastra and vrksayurveda is complex and seems to
partly overlap with it. Beyond the topic of kanyalaksana, it includes material re
lated to conjugal relations and recipes for sexual performance, two topics great
ly elaborated in later kamasastra literature.
Padmasr s Nagarasarvasva and the World of Medieval Kamasastra
47
belly, a soft voice, very plump limbs they are corpulent and prosperous and the doe
(mr:gin:) type woman as being thin, dark, cool like a moon ray, with a graceful wa
lk, large teeth, slow speech, very thick hair, kapha constitution, small appetite
, a well-concealed forehead, a glossy complexion and lots of fragrance during se
x. 26 This sharing of characteristics would seem to indicate that physiognomic con

ceptions of the body were widely influential in numerous fields of knowledge in

early medieval India.
In addition to new divisions of lovers, the Nagarasarvasva and Ratirahasya also i
ntroduce a conceptualization of bodily pleasure entirely absent in Vatsyayana s Kamas
utra. This doctrine, which Kokkoka attributes to the teachings of Goxuputraka and
Nandikesvara, was widely accepted by later kamasastra writers, and came to be known
as the doctrine of candrakala, because of its association of parts of the female
body (as loci of excitement) with the phases of the monthly cycle of the moon.
27 Kamadeva was thought to dwell or arise in different parts of the body in accordanc
e with increments of the moon s phases. According to the Nagarasarvasva, Kama moved
gradually up the left side of the body from the toe (padagra) (through the leg, th
igh, uterus, abdomen, breast, palm, neck, lip cheek eye, and ear) to the top of
the head (srsa) during the bright (waxing) half of the moon s cycle (suklapaksa), upon
which Kama pervaded her entire body for two days, before moving from the head ba
ck downward to the toe during the dark half (krsnapaksa) of the cycle, residing
at each locale for one phase.28The lover was to stimulate the region of the woma
n s body made especially susceptible by Kama s presence there by using his hands, mou
th, or by contact with similar parts of his own body .29 This had the effect of
exciting, pleasing and moistening a woman in the same way, according to Kokkoka, tha
t a doll (putrika) made of moonstone oozed or flowed at the touch of the moon s ray
s. 30
The crucial feature of this scheme, however, was that physical stimulation could
also be accompanied by a mantric visualization of efficacious syllables on the
relevant parts of the woman s body. The Ratirahasya briefly recommends that the lo
ver direct, with his eyes, various vowels (matra) onto the woman s body like sparks fro
m a fire. 31 The Nagarasarasva elaborates in more detail. It would seem that a di
fferent vocalic syllable, understood as a seed of Madana , resided on the part of t
he body visited by Kama during each phase of the lunar cycle. A man was to
26 NS 14.3; 14.5.
27 RR 2.5.
28 See NS 17.1 17.3. The parts of the body differ somewhat in the RR (2.1), most n
otably, while Kokkoka enumerates fifteen anatomical locales from the big toe (a_
ngus:t:ha) to tip of the head (murdhan), Padmasr provides only 14, maintaining that
on the fifteenth Kama resides in all parts of the body (sarvasarradesa). For Padmasr
an is susceptible over her entire body for two consecutive days monthly at purn:i
ma.
29 NS 18.3 18.5; RR 2.2ff.
30 RR 2.3. The term majjanti is used for majjayanti here, which Kan cinatha glosses,
in accord with other verses in the chapter, as dravayanti.
31 RR 2.4 2.6. According to Kokkoka, Nandikesvara and Gon:putraka differed in the num
ber and place of vowels to be visualized.
48
D. Ali
stimulate the relevant syllable by visualizing it on his lover s body and adding a c
andrabindu white like the autumn moon and shining like the sun . 32 Such a visualiz
ation caused sexual arousal and even complete satisfaction in a woman. 33 Padmasr
l adds further mantras that involved the visualization/activation of other sylla
bles on the vagina and clitoris that were to be used before touching the woman s b
ody or even for the purpose of deterring the advances of rival suitors. 34
The theory of the stations of kama found in the Nagarasarvasva and Ratirahasya was u
ndergirded by a physiology of pleasure entirely absent in the Kamasutra. Padmasrl m
entions an elaborate network of channels or tubular conveynaces , called nad: or nad:
ich acted as carriers (avaha) of pleasure or arousal (mada) in the female body.35 H
e initially notes some 24 of these channels, which seem to have connected the an
atomical pleasure points governed by the lunar cycle to the vagina, particularly
the clitoris (madanatapatra, lit. love s umbrella ), described as the mouth or meeting
point of all the nad:ikas yet in his ensuing discussion he notes a number of discret
e and consequential nad:s in the female sexual organs. 36 This physiology explained
, among other things, how sexual excitement in the erogenous zones of the body p
roduced effects in the female organ, most importantly, the emission of fluid. Th
e idea was to stimulate the nads through different means in order to excite the wom

an. Other texts like the Ratirahasya and Pan casayaka, while containing little on t
he madananad:s around the body, discuss in detail a number of nad:s or nad:cakra (gr
or wheels of nad:s coming
together) which were asociated with various parts of the vulvatic organs. 37 Pad
masrl too enumerates smaller nad:s located in various parts of the vagina good (sat)
and not good (asat) located on the left and right sides (respectively) of its surfac
e, unfortunate (durbhaga) and fortunate (subhaga) on the left and right slightly furth
er inside, and son-bearing (putr) and daughter-bearing (duhitr:n) on the left and righ
sides of its interior. Stimulating (sa_mcodana) these nad:s during
32 NS 17.4 17.7.
33 Even without coital contact, according to Mahesvara, NS 17.8.
34 Most notably, the three syllable (tryaksara) and stimulation (ksobhana) mantras,
as well as a visualization of a rival suitor s name within the vagina of his belov
ed, together with seed syllables and the like, in order to kill him, NS 18.1 ff.
35 NS (181.1). The texts seem to use nad: and its diminuitive nad:ka interchangeably.
ee for example NS 18.4 ff. and RR 10.6 10.7.
36 Padmasr (NS 18.4ff) lists nad:s in the eyes (2), face (2), mouth (1), toes (2), ear
s (2), thighs (2), flanks (2), lower back (1), and cranium (1), which together o
nly constitute only 15. The elaboration here may be an adumbration of the anatom
ical loci of the candrakala (17.1) which could, when totalled, constitute the num
ber 24: toe (2), leg (2), thigh (2), vagina (1), navel (1), abdomen (2), breast
(2), palm (2), neck (1), lips (2), cheek (2), eyes (2), ears (2), and head (1).
Such an interpretation is not without its problems, not least of which being tha
t the mouth (tun:d:a), flanks (parsva) and lower back (trika) are not mentioned am
ong the bodily regions governed by the lunar cycle.
37 The RR (10.6 10.8) mentions the nad:s called the madanagamanadola (resembling a pha
llus, at the centre of the vagina, giving forth the flow of the woman s passion fl
uid) and purn:acandra (filled with sexual fluid) along with an organ (nad:?) called t
he manmathacchatra (=kamatapatra?, the clitoris, a nose-like hood above the madana
gamanadola) as being all controlled by a unnamed nadcakra. The PS (5.2 ff.) mentions
three nads, each associated with different places and functions in the vagina. For
a detailed and learned discussion of these passages and others relating to the
physiology of the female procreatory organs, see Das (2003, pp. 409 428).
Padmasr s Nagarasarvasva and the World of Medieval Kamasastra
49
intercorse had various effects on different sorts of women, and could lead to va
ried emotional and physical outcomes, not least of which was the birth of male,
female, and hermaphroditic embryos.38
The physiology of nad:s and their stimulation (nad:ks:obhana) connected kamasastra t
ast world of contemporary thinking about the body. Theories of nad:s were particula
rly widespread at the turn of the second millennium in South Asia, and cut acros
s a variety of literatures with specific physiological and meditational meanings
, including yogic, tantric and medical. The history of these often overlapping t
raditions and uses is extraordinarily complex. 39 For the purpose at hand it may
suffice to note that though the kamasastric idea of the madananad:ka seems not to be
ound in texts from these traditions, it does seem to approximate those found in
yogic/tantric and pulse-diagnostic medical traditions .40 This conception of nad: a
s a channel was not purely biological in the modern sense - which is to say that whi
le the nad:s of the Nagarasarvasva, Ratirahasya and other texts had discernible phys
ical form and physiological functions (that could be manually excited), they wer
e at the same time part of a more subtle physiology governed by yogic principles a
nd susceptible to the powers of mantra recitation and seed syllable visualizatio
n.41 It would seem that the authors of kamasastra adapted an important physiologica
l idea circulating in their time to the specific needs of understanding the body
from the vantage points of sexual pleasure and reproduction.
We have already seen how in connection with the stations of Kamadeva (kamasthana) and
channels of pleasure (madananad:ka) the Nagarasarvasva and Ratirahasya recommended no
only manual stimulation but also the use of mantras and visualization. Such pra
ctices were part of a more general rise in the recourse to what may be called par
a-technologies relating to sex which were to be deployed alongside (or sometimes
in lieu of) the methods of beautification, courtship and sex as set out in the r

emainder of the tradition. The Kamasutra had included such techniques in its final
book under the category of aupanis:adika, or esoterica , that were recommended to
the man who could not fulfill his desires through the methods laid down in the e
arlier chapters of his work. These included both recipes and mantras for success
in romance (subhagakaran:a) enabling influence over others (vaskaran:a), creating
virility (vr:s:yayoga), rekindling exhausted passion (nas:t:aragapratyanayana), in
creasing penis size (vardhanayoga) and other unusual techniques (citrayoga).42 I
n the Ratirahasya and Nagarasarvasva the prominence and diversity of such para-tec
hnologies (both herbal
38 See NS 19.1ff. Padmasr (NS 19.11) mentions another opinion that these nad:s were l
ted elsewhere on the body. Compare the gender division of the womb (udara) in th
e context of producing children, but without reference to nad:s, in the Br:hatsa_mh
ita (78.24).
39 On the complex history of the rise of medicine based on diagnosis through nad:s,
see Bruns (2009, pp. 39 97).
40 Notable in the case of Padmasr s 24 nad:s meeting in the clitoris, is the similar en
meration of 24 nad:s meeting in a tortoise at the navel in pulse diagnosis. See Bruns
(2009, p. 6).
41 Padmasr suggests that the son-bearing channel (nad:ka putr) was connected to the
/lunar organisation of nad:s in the yogic body (yada suryen:a margen:a dehe vahati mar
utah: I samcodya nad:ikam putrm_ pascan mohanam acaret k NS 38.5).
42 See KS 7.1.1ff.
50
D. Ali
and mantra based) were significantly expanded, as well as the range of phenomena
to which they related. The Ratirahasya devotes two substantial chapters to such
topics with an impressively wide ambit of concern from ensuring seduction, increa
sing virility, controlling semen-flow, and inducing orgasms to removing body hai
r, freshening the breath, winning friends and improving the voice.43 The Nagarasa
rvasva too, in addition to its recommendations on seed syllable visualization, c
ontains other chapters detailing erotic recipes and setting out mantras for vari
ous ends .44 The source of such para-technologies for the kamasastra authors was cl
ear they mention either general or specific tantric, agamic and ayurvedic authorities
.45 The development of para-technologies was widespread after the Gupta period
and traversed a wide variety of genres and knowledges. Technologies relating to
courtship and sex were key themes in this literature. Mahuka s Haramekhala, a text c
ited by the Ratirahasya and discussed by McHugh (in this volume), contains a ver
y large number of recipes both facilitative and injurious which relate to sex and co
urtship, including recipes for the genitals, for couples, intoxication and contr
olling others. This knowledge, cutting across numerous fields, seems to have bee
n widespread, and kamasastra borrowed from it as much as later tantras and other fi
elds.
One sphere which formed a growing concern in the para-technologies of later kamasas
tra, was procreative sex and both the Nagarasarvasva and Ratirahasya anticipate t
his development. Beyond its discussion of child-bearing nad:s, the Nagarasarvasva co
ntains an entire chapter on the subject of obtaining a son .46 In it Padmasr draws on
tantras from both the Buddhist and Saiva traditions, recommending a combination o
f ritual and efficacious acts giving food to monks (or brahmins), worshipping Tara (
or some other god), reciting mantras, stimulating the proper nad:, and taking herba
l potions.47 The Ratirahasya, for its part, provides a variety of recipes to fac
ilitate conception, delivery, abortion, and the relief of postnatal discomfort.4
8 These sorts of concerns were largely absent in the Kamasutra and became an incre
asingly important feature of kamasastra knowledge as it evolved throughout the seco
nd millennium. It also might be the case that the increased interest in topics l
ike female vulvatic anatomy, female sexual fluids, and methods of female sexual
excitation (through the stimulation of the kamasthanas and nad:s) all noted above were c
nnected to concerns and debates about the physiological basis for procreation. R
ahul Peter Das has demonstrated that in a
43 See RR 14.1ff; 15.1ff; see also 1.23, above and beyond the mantra visualizati
on on the kamasthanas mentioned above RR. 2.4ff. See also PS 3.1ff.
44 NS 12.1ff; 38.1ff.

45 See RR 14.1 and NS 12.3; 38.16.

46 NS 38.2ff.
47 Padmasr (NS 38.16) claims to have consulted the Siddhaikavra and the very secret Ka
atantra of Sa_nkara in composing this chapter. A collection of Buddhist Vajrayana
mantras under the title Siddhaikavramahatantra has recently been published by the C
entral Institute of Higher Tibetan Studies (1998), which contains a Tara mantra (2
.35) nearly identical to that found at in NS 38.17. I thank Mattia Salvini for d
iscovering this reference. I have not been able to locate the Kamatantra of Sa_nka
ra.
48 See RR 15.60 15.83.
Padmasr s Nagarasarvasva and the World of Medieval Kamasastra
51
number of medieval medical texts the assumption seems to have been that the emis
sion of female procreative fluid (presumably during orgasm) was deemed necessary
for conception .49 The Kamasutra alludes to such an understanding, noting that th
e followers of Babhravya maintained that the embryo could not form without a woma
n achieving orgasm (lit. bhava) during sex interpreted by his commentator Yasodhara
as reference to the emission of female semen . 50 Yasodhara s onward discussion indica
tes, however, as Das has suggested, that there seems to have been confusion and
debate as to the origin and nature of this fluid among scholars, and that the de
bate was largely conducted in ayurvedic circles.
Kenneth Zysk has argued that by the end of the sixteenth century a distinctive s
cience by the name of ratisastra, primarily oriented around conjugal love and procre
ation, had emerged from the tradition of kamasastra.51 While this distinction is fa
r from clear within the tradition itself, Zysk is surely correct to draw our att
ention to the preoccupation of some treatises with conjugality and procreation.
52 The emphasis on reproductive sex seems to have been closely related to the in
corporation of material from the physiognomic, astronomical and medical knowledg
es.
Aesthetics and Representation
A distinctive feature of the Nagarasarvasva is its relationship with courtly aest
hetics and covert systems of signification. The connection between erotics more
generally and the world of the court had been constitutive from the beginning of
the extant kamasastra tradition. Vatsyayana s Kamasutra, as is well-known, famously a
essed itself to the nagaraka, whose lifestyle, associates, accomplishments and pa
stimes, as set out in its first book (on generalities , sadharan:a) bore a distinctiv
ely courtly idiom. 53 Early medieval court poetry, conversely, borrowed heavily
from conventions of courtship detailed in kamasastra. Overall, the world of the
49 See the extended discussion of Das (2003, 368ff.)
50 KS 2.1.18, and comm. I thank Laura Desmond for drawing Yasodhra s discussion to
my attention.
51 Zysk (2002, pp. 9 12).
52 Most later kamasastric texts and their commentators use ratisastra and kamasastra
erically, imprecisely and even interchangeably. The RR (1.6,8), for example, a t
ext that Zysk places in the kamasastra tradition, uses kamasastra and ratitantra to de
note one and the same field of knowledge, sees itself as a continuation of Vatsyay
anasastra but calls itself the Ratirahasya, or Secrets of Sexual Pleasure . Jayadeva s
fifteenth-century Ratiman jar (2, 58 59), another text of the kamasastra tradition, call
itself a ratisastra (58, 59) but also claims to be distillation of the essence of
kamasastra and ratisastra (2). Padmasr makes it clear that he presents a distillatio
f kamasastra (NS 30.15; 11.1; esp. 1.2) but uses the term ratisastra throughout to de
note the same knowledge tradition (15.10; 38.17). Jagajjyotirmalla s seventeenth-c
entury commentary glosses Padmasr s use of ratisastratattvajn a as kamasastrajn a. Th
tiveness of the texts Zysk takes up, the Ratisastra and Ratiramana, lies primarily
in their extensive adoption of large amounts of ayurvedic, physiognomic (samudri
kasastra) knowledge, as well as particular features like bed types for the differe
nt nayikas. These texts are no doubt remarkable in this respect, but perhaps not e
nough to warrant classification as a separate genre. We have seen in any event t
hat the nayikabheda of later kamasastra seems to draw implicitly upon physiognomic kn
owledge.
53 See Ali (2004, pp. 74 77).

52
D. Ali
nagaraka depicted in the Kamasutra fit well with the everyday practices and aesthet
ic registers of early Indian court life. The Nagarasarvasva extends this link wit
h courtly mores in two ways. First, it includes chapters on material commodities
and sign systems that formed major foci of aesthetic delectation at court these k
nowledges, briefly indicated as supporting arts of love in the Kamasutra, are openly
included within the ambit of the Nagarasarvasva. Second, Padmasr also engages subst
antially with formalized literary aesthetics in a way unparalleled in the histor
y of kamasastra. While the use of kamasastra by literary theorists was formally sancti
oned, diverse and widespread, the reverse was not necessarily so. 54
After a brief chapter describing the decoration of the body and house of the naga
raka in terms reminiscent of Vatsyayana s more extended treatment of the subject in
the opening book of the Kamasutra, Padmasr takes up two subjects not treated in any o
ther kamasastra the examination of gems (ratnapars:ka) and the preparation of perfumes
gandhadhikara). While these two subjects appeared among the sixty-four supporting
arts (a~ngavidya/kala) to be mastered by the ideal lover, according to the Kamasutra
, as knowledges they remained largely discrete from kamasastra. Yet, as McHugh has
pointed out, with the growth of cosmopolitan courtly culture in early medieval I
ndia and the informed consumption of luxury commodities that went with it, knowled
ge of gemstones and perfumery gained increasing prominence as sciences in themse
lves and were incorporated into a range of other discourses, from Puran:as to sum
ptuary and astrological manuals. 55 As we have mentioned, Padmasr likely drew heavi
ly on the tradition of ratnasastra as it had developed throughout the first millen
nium in South Asia. The relevance of gems to the domain of kama as objects which po
ssessed both decorative beauty and potentially special powers is simply assumed by
Padmasri. His treatment is self avowedly adumbrated, providing only general rema
rks on how the best of gems may cause good or bad and briefly describing their exc
ellences and flaws. 56 In the case of scents, Padmasr emphasizes that smells form t
he major elictors of sexual arousal (madanapradpaka) and recommends a lover s diligent
training in this field. He goes on to cite the as yet unknown authority of Loke
svara, providing a brief survey of a variety of perfuming techniques including oi
ntments powders and incense to be used both for one s body (hair, armpit, mouth an
d face) as well as objects and substances (water, air, and betelnut).
Also partly related to material culture is the Nagarasarvasva s elaborate treatment
of hints or sa~nketa, which extends over a number of chapters in the work. The te
rm sa~nketa refers to an indicatory sign, mark or hint, an agreed assignation, a
nd by extension, in ala~mkarasastra, an ppointment made between lovers.57 Pad54 For recommenations that the poet be familiar with kamasastra, see Kavyalam:kara-Sut
a-Vr:tti 1.3.8 1.3.9, 11. Patel (in this volume) shows how kamasastra categories wer
e put to further use in the emic conceptualizations of the Sanskrit literary can
on among commentators and critics.
55 See McHugh (in this volume).
56 NS 3.7.
57 Conventions of kavya understand sa~nketa as an engagement or tryst (sa~nketa), par
ticularly one attended by the abhisarikanayika or heroine going out [to meet her lover
]. See Amarakosa (2.6.10): kantarthin tu ya yati sa~nketam~ sabhisarika. This sense
ed in numerous scholarly contexts, i.e. Kavyaprakasa 1.3+ and Mallinatha on Raghuva~
msa 16.12.
Padmasri s Nagarasarvasva and the World of Medieval Kamasastra
53
masri conceives of it in the former sense, as a covert agreed upon sign used by a
woman (or, presumably, a man communicating with her), to indicate information a
bout a secret meeting. Padmasri enumerates several types of sa~nketas signs from ob
lique speech (vakrabhas:a) bodily gestures (a~ngamudra), bundles (pot:ali), garmen
ts (vastra), flowers (pus:pa) and betel (tambula). The range of material domains p
resented by Padmasri would seem to resemble that of the divinatory knowledges of
the Br:hatsa~mhita and its successor texts. The main difference, of course, is th
at the sa~nketas in the Nagarasarvasva are instruments of human agency, not its l
imits they were to be used by individuals to secretly communicate with one another
. This is most clear in Padmasri s presentation of a range of everyday words whose
usual denotative meanings were unconnected to their tryst-related assignations.

So fruit (phala) was to indicate a man, flower (pus:pa) a woman, pomegranate (dad
:ima) someone twice-born, jackfruit (panasa) a ks:atriya, sprout (a~nkura) quest
ioning about one s family, plantain fruit (kadalja phala) a vaisya, mango (amra) a sudr
, moon in the second digit (dvitiyendu) a king s son, cloud s shade (ghanacchaya) th
e king s son, while words like sa~nkha, mahasa~nkha, padma, mahapadma, rama, virama, pr
vara, and pratyus:a were assigned the various watches in the night, presumably f
or arranging meetings. 58 Non-verbal gestures included touching various parts of
the body (ears, head, breasts, hair) to indicate greetings and degrees of passi
on, as well as a very elaborate set of assignations for the fngers and hands rel
ating to the time and place of a meeting. If gestures with the fngers and hands
indicated availability for meeting, touching the fngers themselves and their joi
nt lines (rekha) indicated, respectively, the cardinal directions and days of th
e lunar month once again, presumably for the purpose of arranging clandestine meet
ings. 59 The colors and tears of garments (vastra) and the strings of flower gar
lands were to indicate degrees of passion. Perhaps the most interesting of the s
a~nketas are those relating to bundles (pot:ali) and betel-leaf rolls (bit:aka). T
he meanings of betel-leaf rolls varied depending on their ingredients and the di
fferent shapes into which they were folded (a betel leaf roll with no areca nut
[puga] indicated displeasure!) , while bundles (wrapped in red thread and sealed
with beeswax), differed depending on the precious substances they contained.60
At one level Padmasri s overview of sa~nketa, like other chapters of the Nagarasarva
sva, depicts a world where familiarity with sumptuous material goods was deemed
to be an integral support of sexual pleasure a presentation entirely consistent wi
th Vatsyayana s account of the lifestyle of the nagaraka. And, the more elaborated ma
terial world of the Nagarasarvasva is reminiscent of contemporary medieval encycl
opedia and sumptuary manuals like the Manasollasa, perhaps even the closely relate
d prognosticatory literatures which divided the world
58 NS 5.4 5.5, 10 11. The last of these words, daybreak (pratyusa) perhaps too openly i
ndicates the fnal watch of the evening, to which it is assigned. Some words had
connotative connections to their assigned meaning. Summoning, for example, was t
o be indicated by the word goad (a~nkusa), while an obstacle to meeting by the word
wall (prakara), NS 5.9.
59 NS 6.1ff.
60 NS 7.1ff; 9.1ff.
54
D. Ali
into vast signifying minutiae. Yet such sources bear only a superficial intersec
tion with the theory of sa~nketa as set out by Padmasri. The problem is made more
intruiging by Padmasri s reference to a discrete science of signs (sa~nketasastra), th
ough no such treatises seem to have survived.61
The conceptual and practical background of the Nagarasarasva s treatment of sa~nket
a, however, is probably neither sumptuary nor prognosticatory, but religious. In
the Buddhist context, which Padmasr was probably familiar with, similar sign syste
ms were most frequently encountered in texts called the yogintantras, dating from
the eighth century CE. Such signs were referred to by the general term choma, an
d were used by initiates in connection with arranging secret feasts and ritual g
atherings.62 They involved both gestures and speech. In the Hevajra Tantra, hand
gestures were used as signs of recognition among initiates. Secret speech, know
n in the Buddhist tantric materials as sandhyabhas:a or coded language , relied on a s
ystem of apparently arbitrary equivalences in the manner we have seen in the Naga
rasarvasva. According to the Hevajra sandlewood (malayaja) referred to a meeting a
nd small drum (d:in:d:ima) an untouchable.63 Like the treatment of sa~nketas in th
e Nagarasaravasva, the system of choma and sandhyabhas:a allowed tantric practitione
rs to indicate the details of secret meetings, and even refer to substances used
in antinomian ritual semen, human flesh, and urine, and alcohol were among the ob
jects one could indicate covertly through the use of sandhya- bhas:a. The connectio
n is perhaps even closer, as the Hevajra Tantra understands this coded language
as consisting of what it calls conventional signs or samayasa~nketa and here we have
a usage largely identical to that in the Nagarasarvasva.64 The range of gestures
and words, as well as their specific equivalences, differed in the Nagarasarvasv
a and yogintantras. Their ends, too, were obviously distinct. Though the romantic

tryst and tantric ritual both shared secrecy, the former was entirely devoid of
antinomian or transgressive elements, and its goal was sexual pleasure, not spir
itual transformation. Though one might expect tantric practices to build upon or
transform already existing norms and conventions as seems to be the case in the C
an:d:amaharos:an:a s enumeration of different coital positions (bandhas) in its des
cription of the yogin and yogin s ritual sex65 in this case the Nagarasarvasva, where
there exists no kamasastric precedent for the system of sa~nketas, the case would s
eem to be the reverse: where tantric ritual practice is adapted for conventional
, worldly ends.
Beyond what seem to be clear links with esoteric Buddhism, the Nagarasarvasva s use
of sa~nketa may also have been closely connected with established (but not part
icularly well understood) traditions of oblique signification at court.
61 NS 5.2.
62 See the Hevajra Tantra 1.7.1ff.; 2.3.55 2.3.67. See also the very useful discus
sion of Davidson (2002, pp. 262 269), who notes that choma is a late middle-Indic
word related to the Sanskrit chadman, deceit or covering .
63 Hevajra Tantra 2.3.56, 58.
64 Hevajra Tantra 2.3.55.
65 See the Can:d:amaharos:an:a Tantra 6.180ff. for a list of bandhas. Though like
ly inspired by kamasastric discourse, most of the names in the Can:d:amaharos:an:a c
annot be traced to the nomenclature of any specific kamasastra text.
Padmasri s Nagarasarvasva and the World of Medieval Kamasastra
55
Vatsyayana himself lists among the sixty four supporting arts of erotic pleasure t
he ability to speak in sign language (aks:aramus:t:ikakathana) and the understand
ing languages made to seem foreign (mlecchitavikalpa), which Yasodhara notes were
for the purposes of secret advice (gud:havastumantran:artha).66 Padmasr claims that k
owing the assignations set out in his sa_nketa chapters was of the utmost import
ance, for it was not enough simply to be proficient in the arts one also had to un
derstand the secret signs used by women to arrange trysts. As he puts it, althoug
h a man may be skilled in all the arts and have a host of virtues, urbane young
women, abounding in good qualities, will cast him away like a faded flower if he
does not understand their hints .67 Though it is difficult to imagine that the in
ditement of the particular significatory equivalences found in the Nagarasarvasva
made them very useful, Padmasri s account (and his admonishment) makes clear how s
uch codes functioned as a set of agreed assignations hidden behind normal communicat
ive convention. They were, to extend Padmasri s characterization of speech hints, in
direct (vakra) signs. 68 Beyond presupposing a certain hierarchy of knowledge, one
wonders to what extent such systems of indirection may themselves have been impl
icated in the production of desire and the excitation of pleasure.
The most intriguing link with aesthetics in the Nagarasarvasva is Padmasri s discuss
ion of the term hava understood as a captivating action or disposition exhibited by
the nayika in courtship. No other extant kamasastra uses this term as a category of
analysis, and Padmasr here clearly draws upon what was an already established tradi
tion within ala_mkarasastra characterizing the alluring attitudes and actions of fe
male lovers. Ala_mkarasastra writers, beginning with Bharata, tend to call these mo
des ala_mkaras (ornaments), though other terms like cesp (gesture, behavior) or sr_nga
cest:a (gesture/behavior from love) are also known, and enumerate as many as 28 of
them in some treatises though a stable core of about 20 seems to persist through
the literature.69 Padmasr uses hava as a general term to refer to some sixteen dispo
sitions, but also uses it in a more restricted sense, as one among the sixteen h
avas. All other ala_mkara authors use the term in this latter acceptation. Otherwi
se, the list presented in the Nagarasarvasva seems to correspond most closely to
that of the Nat:yasastra, and also to the Sahityadarpan: a.70 There are, however, a
few differences which make Padmasr unique, and perhaps indicate his reliance on oth
er as yet unknown sources. Bharata and other authors take bhava, hava and hela to b
e the first three ala_mkaras (=Padmasri s havas),
66 KS 1.3.15 comm.
67 kalakalapais ca yutam_ samastaih: gun:air asa_nketavidam_ hi kantam_ / pramlananir
malyam ivotsrjanti gunadhika nagarikas tarunyah // NS 5.1 5.2; see also 11.1 11.2. A com
on theme in later anthologies contrasts the sophisticated ways of desirable wome

n to the untutored approaches of country suitors. See, for example, Suktimuktaval 87

.6.
68 NS 5.3.
69 See Nat:yasastra 22.8ff; Rasakalika 40ff; Nat:yadarpan:a 4.28ff; Dasarupaka 2.30 ff
Sahityadarpana 3.93ff. At one point (NS 13.4) Padmasr equates hava with an action (c
esta) born of the bhava of erotic love (sr:_ngarabhavaja).
70 NS 13.3 13.4.
56
D. Ali
defining them as increments of one and the same phenomenon: bhava is a state either
in the sense of an inner feeling of love usually with some sort of physical dim
ension; hava is bhava when the latter becomes first apparent; and hela is its full
manifestation. These are often considered to be body-born (a~ngaja) ala~mkaras, to
be distinguished from the others, often called natural (svabhavaja) and/or spontaneo
us (ayatnaja) but there is no unequivocal consensus on this point among ala~mkarasastr
a authors. The crucial point is thus a continuity in the experience and manifest
ation of physical and behavioral states in the ala~mkarasastra tradition. In the Nag
arasarvasva, however, bhava is excluded from the list of sixteen havas, instead fo
rming its basis. Padmasri begins with the position that bhava can be pure, impure
or mixed, and that pure bhavas may be further subdivided into mild, intense or ve
ry intense. And it is from these pure bhavas, of varying intensities, particularl
y the bhava of erotic love (srhgara) that the behaviors (cesta) known as havas are b
orn. 71 Padmasri conceives of bhava, in other words, as a precondition for the eme
rgence of the havas. The separation of bhava from the expression of the havas would
seem to open the possibility of a strong distinction between a prior causative
physical state (bhava) and a subsequent behavioral consequent , but Padmasri is not f
orthcoming on this point. 72
Padmasri provides each hava with a short definition and follows it with an illustr
ation typically a brief description of a woman s behavior or appearance, sometimes i
n a dramatic scenario, and often followed by a brief comment. Here are Padmasri s v
erses on viks:epa and vikr:ta.
The occurrence of various changes in a woman, like her clothing becoming disheve
led, is called viks:epa by the great sages, from Kapila onwards.
For instance she tightens her coiffure half-way, wrongly places an incomplete head
mark, applies black collyrium to only one eye, ties her jewel girdle, sounding
with bells, on her chest, drapes a half-pearl necklace on her arms, and bears a
trace of betel juice at the edge of her lower lip.
When a woman, in such a state of viks:epa, holds up her thigh garment with effor
t, it steals the heart.
When doe-eyed women say certain things to lovers, even knowing they shouldn t be s
aid, it is called vikr:ta.
71 NS 13.1 13.4.
72 Indeed, Si~mhabhupala s fourteenth-century Rasarn:avasudhakara (1.322 1.323) consider
s all of these behaviors consequents (anubhavas), but nevertheless lists bhava among
st them! Abhinavagupta s commentary on the Nat:yasastra (22.6 comm.) distinguishes be
tween the use of bhava in Chap. 22 on the ala~mkaras as a purely physical state an
d bhava in chapters seven and eight as a pervasive internal state.
Padmasri s Nagarasarvasva and the World of Medieval Kamasastra
57
As in What s this thing on your neck? Darling, it s an anklet! This band used to be an
ornament for someone s foot. Such words, uttered by a doe-eyed woman who knows bett
er but ignores that they should not be said, is vikr:ta, and increases one s inter
est. 73
Such descriptions are remarkable, and bear testatment to the complex interrelati
on of kama and ala_mkarasastra. Here, rather remarkably, the Nagarasarvasva apparentl
y takes conventions generated within the field of literary theory as description
s of social reality at least as it imagines social reality to be. 74 This is to sa
y that the purpose of Padmasri s discussion of the havas is not, like that of ala_mk
arasastra, to set the conventions in the composition and appreciation of poetry and
drama, but for the real-life delectation of feminine charms on the part of the ma
le subject. We would seem to have here, in other words, an explicit example of l
iterary sensibilities shaping attitudes to courtship. This suggests several poin

ts. First it hints at the existence of a much closer relationship between ala_mk
arasastra and kamasastra than has been assumed, at least during the time of the Nagara
arvasva. Nor was this a complete anomaly, as indicated by passages from later wo
rks like the Pan casayaka and Kamasamuha. Moreover, and perhaps more importantly, thi
s indicates that not only was Padmasri clearly aware of the ala_mkarasastra traditio
n, but he was able to productively and creatively engage with its categories. He
even writes in the ala_mkarasastra style by following his definitions by exemplars (
drstanta). This engagement is further underscored by the fact that it is the sect
ion on the havas in the Nagarasarvasva that is most often cited in later literary
commentators. Raghavabhat:ta, the fifteenth-century commentator on Kalidasa s Abhijn ana
sakuntala, for example, cites Padmasri s definition of vilasa, while the Sar_ngadharapa
dhati and Sahityadarpan:a simply borrow his definitions without any remark.75
Concluding Remarks
Scholars studying the development of kamasastra, to the extent that the history of
the genre has been considered, have generally stressed two related but somewhat
antithetical trends in the evolution of the field after Vatsyayana. First, the bro
ad, humanistic and morally permissive vision of the early science, as embodied i
n the
73 visa_ms:t:hulavesamayo vikaro vividhah: striyah: j tam amananti viks:epam_ munaya
h: kapiladayah k1 tadyatha, dhammillam baddhamuktam tilakam asakalam nyastavr:ttat
h ca dhatte drstav ekatra kalan janam urasi ran:atki_nkin:m_ ratnakan cm j a_msotks:i
hara kramukarasakalamatrases:adharanta kantaviks:epam ity adyayi harati mano yatna
asah: k avacyanam_ pada- rthanam_ jn ane pi yadi bhas:an:am j nayakesu mrgaksuiar
prakrtitamktadyatha kanthe ka esa tava vallabha nupuro yarn tatpadabhusan:am ayath val
yas tadanmjity adyavacyam abhibhavya vaco mrgaksya jn ane pi tatvikr:tam utsukatarn
NS 13.21 13.24. In the latter two verses we have replaced the editor s rendition wit
h alternative readings followed by the commentator.
74 The PS also borrows from ala_mkarasastra in its use of the theory of the eight n
ayikas (see footnote 22 above).
75 Raghavabhat:t:a on Abhijn anasakuntala 1.7; 2.23 and Sahityadarpan:a 3.101. The Sa
dharapaddhati (3152 3188) incorporates the entire chapter on the havas from the Nagar
asarvasva.
58
D. Ali
Kamasutra, was thought to be eclipsed in medieval times by a science whose horizon
s were more dominated by magic and superstition, constrained by social morality,
and increasingly confined to conjugality. Second, and relatedly, the holistic v
ision of the science itself as found in Vatsyayana s text is increasingly deemed to
shrink as external knowledges and paradigms impinge upon it. While there may be
partial truth to some of these claims, the overall picture presented is problema
tic at a number of levels. The historical problem may rather be divided into two
related but distinct issues: first, the content or worldview of kamasastra, and seco
nd, the organization of knowledge relating to kama.
As to the vision and content of kamasastra, part of the problem lies with the origi
nal conceptualization of the science as embodied in the Kamasutra. The view of thi
s text as a holistic and humanistic treatment of human sexuality devoid of morali
ty is surely somewhat misplaced. Recent scholarship, (as well as Desmond s essay in
this volume), has argued that the Kamasutra operates with its own moral constrain
ts what might usefully be called an ethics of pleasure. From this perspective, Desmo
nd rightly focuses on the question of subjectivity. While the subject of kamasastra g
radually changed over time, the texts studied here retain a strong courtly basis
, even as they show an increasing interest in procreative sex. The very title of
Padmasri s work, referring to the courtier/townsman (nagara) clearly invokes a contin
uity with the worldview of the Kamasutra. This emphasis continued later as well the
Smaradpika defines the nagaraka as having six qualities: possessing urbane ornaments
, wisdom, purity, eminence, skill at singing, and being intent upon clever conve
rsation. 76 That courtly modes and paradigms continued to inform the kamasastra tra
dition is evident not only from the material culture of later kamasastra but also t
he persistent presence of ideas and categories from ala_mkarasastra in texts like t
he Nagarsarvasva, Pan casayaka, Smaradpika and Kamasamuha. For Padmasri s part, he cl
emonstrates that ala_mkarasastric notions of feminine beauty were deemed practicall

y useful for the world of social reality envisioned through his text.
Such continuities must be juxtaposed to significant conceptual changes in the fi
eld. The human body, particularly the female body, becomes a more complex and va
riable entity, traversed by intricate webs of pleasure channels, subject to the
cycles of the moon, and bearing the marks and signs of the future all of which ten
d to make the sexual body as a vast template for collective and individual inter
pretation and manipulation through manual, medicinal and mantric para-technologie
s that exceeded the procedures of traditional courtship and romance. While such a
change may be consistent with the growth of tantric, astronomical, and medical
ideas prevalent in elite and courtly contexts, a growing emphasis on conjugality
and procreation also characterizes later kamasastra, gradually in texts like the R
atirahasya and Nagarasarvasva, and more so in texts composed after 1500. These co
ncerns, largely absent in the Kamasutra, may point to the traditional anxieties of
dharmasastric injunction regarding the maintenance of social boundaries, the repr
oduction of the patriarchal household, and the regulation of sexual norms in the
76 agramyaman:d:anah: prajn ah: sucih: srma_ms ca gayanah: I narmagos:t:hpravis:t:
un:o nagaro matah: k SmD 173.
Padmasr s Nagarasarvasva and the World of Medieval Kamasastra
59
conjugal context, but it should be kept in mind that the conjugal and reproducti
ve themes of later kamasastra were neither totalizing nor ubiquitous, and existed a
longside extensive discourses in the genre which assumed the traditional independ
ent nayikas or elided social context altogether. To date we have no textured social
history of domesticity in medieval India against which we can make sense of the
se imbrications.
Finally, a word about the organization of knowledge in later kamasastra. Despite th
e frequent acknowledgment of Vatsyayana, later kamasastras rarely replicate the thema
tic structure of the Kamasutra s with its seven-part organization into sections on g
eneral topics (sadharan:a), sex (samprayoga), maidens (kanya), wives (bharya) other m
en s wives (paradarika), courtesans (vesya) and esoterica (aupanisadika). The material
is ordered in what would seem, at least from Vatsyayana s perspective, a largely inc
onsistent and haphazard fashion from treatise to treatise. Other than the Nagaras
arvasva, which contains a series of chapters on topics broadly related to themes
covered in the opening book of the Kamasutra on the lifestyle of the nagaraka, the
great majority of later treatises like the Ratirahasya, Pan casayaka, and Smaradpika
dispense with Vatsyayana s section on general topics altogether, beginning instead w
ith the enumeration of male and female types, the foundation of the book on sex in
the Kamasutra. Likewise, few later kamasastra texts contain much material on courtes
ans.77 The overwhelming thematic emphasis in these treatises seems to be on sex
itself, beginning with the discussion of types of lovers, but carrying on into t
he enumeration of styles and types of kissing, biting, scratching, embracing and
nail-marking, sexual positions, and the like, with occasional chapters on wives
, and often very substantial sections throughout on para-technologies (esoterica
). It is here that many of the new doctrines of the body discussed in this essay
are to be found.
Overall, the impression one might glean from these developments is that of an at
tenuated vision and concern within the later texts of the genre. This, however,
would be a hasty conclusion, for the topic of sex itself saw the introduction of
numerous new ideas and practices which effectively opened the genre outwards. I
nnovation was very much the result of interactions with a wide variety of indepe
ndent but allied knowledges, including perfumery, gemology, sign theory, poetics
, astronomy, physiognomy, medicine, tantra and particularly para-technologies of
various kinds which cut across them all. Many of these fields had seen tremendo
us expansion during the latter half of the first millennium CE, exemplified in s
ome cases by their full or partial disaggregation from the earlier knowledge for
mations in which they had been embedded, to become independent and autonomous sast
ras .78 A number of these had been included among the 64 supporting knowledges (
a_ngavidya) or arts (kala) recommended to the nagaraka in the Kamasutra. In the first
neration of surviving kamasastras after Vatsyayana, the apparent contraction of topica
l concern is counterbalanced by vast skein of con77 Courtesans remain a favourite theme in court poetry.

78 Notable once again is the diversity of subjects included in a relatively earl

y text like the Br:hatsa_mhita which while continuing to form part of encyclopedi
c genres, also becomes disaggregated into a host of independent sastric knowledges
.
60
D. Ali
nections around courtship and the body which lead outward to the ever more devel
oped sciences of elite life.
Appendix I: Authorities Mentioned in Seleted Kamasastra Texts Kamasutra
1)
Auddalaki Svetaketu (1.1.9; 2.1.19; 2.1.31; 5.4.31; 6.6.34)
2)
Nandin (1.1.8)
3)
Babhravya (1.1.10; 1.3.17?;1.5.33; 2.1.34 ; 2.1.41; 2.2.5; 2.2.22; 2.6.21
; 2.10.49;
3.2.3; 4.2.40; 5.4.32; 5.4.41; 5.6.47; 6.6.35; 7.2.6; 7.2.56)
5)
Dattaka (1.1.11; 6.2.55; 6.3.44)
4)
Carayan:a (1.1.12; 1.4.20; 1.5.22)
9)
Suvarn:anabha (1.1.13; 1.5.23; 2.2.23; 2.4.6; 2.5.34; 2.6.22; 2.8.7)
6)
Ghot:akamukha (1.1.14; 1.5.24; 3.1.3; 3.1.10; 3.2.17; 3.3.4; 3.4.29)
7)
Gonardiya (1.1.15; 1.5.25; 4.1.4; 4.1.21; 4.2.36; 4.2.42)
8)
Gon:ikaputra (1.1.16; 1.5.5; 1.5.34; 5.1.4; 5.4.9; 5.4.33; 5.4.42; 5.6.45
;)
10)
Kucumara (1.1.17)
Nagarasarvasva
1) Mahesvara (1.5; 1.9; 16.6; 16.15; 17.8 )
2) Vatsyayana (37.15)
3) Lokesvara on perfumes (4.2)
5) Siddhaikavira (38.16)
6) Kut:t:animata (37.16)
7) Kamatantra of Sa_nkara (38.16)
8) munindra, munayah: (multiple)
Ratirahasya
1)
Vatsyayana (1.9, 3.18, 11.18)
1)
Nandikesvara (2.5, 14.1)
2)
Gon:iputraka (2.5)
3)
Munindra/Muninatha (multiple)
4)
Karn:isuta (4.21)
5)
Muladeva (5.22)
6)
Gun:apataka (4.3, 4.7, 4.25)
7)
Nagarjuna (15.18)
8)
Sabdarn:avam (14.1)
9)
Haramekhala (14.1)
10) Ud:d:isa (14.1)
11) Hundred Yogavalis (14.1)
12) 3 vaidyakasa_mhitas (14.1)
13) various types of saivagama (14.1)
14) mun:ayah: (mulitple)
61
Padmasri s Nagarasarvasva and the World of Medieval Kamasastra
Pan casayaka
1)
Manmathatantra of Isvara (1.3)
2)
Vatsyayana (1.3)
3)
Gon:iputraka (1.3)
4)
Muladeva (1.3)
5)
Babhravyavakyamr:ta (1.3)
6)
Nandisvara (1.3)
7)
Rantideva (1.3; 3.36; 3.42)
8)
Ks:emendra (1.3)
9)
Bhojaraja (3.6)
10) Mahesa (4.56)
11) Candramauli (5.2)
12) munindra, kavndra, (multiple)
References Primary Sources

Amarakosa of Amarsi~mha with commentaries of Li~ngayasurin and Mallinatha (1971 83).

Edited by A. A. Ramanathan. 3 vols. Madras: Adyar Library.
Abhijn anasakuntala of Kalidasa, with the commentary of Raghavabhat:t:a (1913). Edited
y Naryana Balakrishna Godabole. Bombay: Nirnaya Sagar.
Br:hatsa~mhita of Varahamihira (1865). Edited by H. Kern. Bibliotech Indica. Calcu
tta: Baptist Mission Press.
Can:d:amaharos:an:a Tantra (Chapters I-VIII) (1974). Edited and translated by Chr
istopher George.
American Oriental Series, no. 56. New Haven, Connecticut: American Oriental Soci
ety. Dasarupaka of Dhana~mjaya with the Avaloka commentary of Dhanika and sub-comm
entary of
Bhat:t:anr:si~mha (1969). Edited by T. Venkatacharya. Madras: Adyar Library and
Research Centre.
Hevajra Tantra (1959). Edited and translated by D.L. Snellgrove. 2 vols. London
Oriental Series.
London: Oxford University Press.
Kamasutra (KS) of Vatsyayana (KS) (1929). Edited by Goswami Damodar Shastri. Benares
: Jaikrishnadas and Haridas Gupta.
Kavyalam:kara-Sutra-Vr:ttti of Vamana (1928). Edited and translated by Ganganatha Jha
. Poona: Oriental Book Agency.
Kavyaprakasa of Mammat:a (1983). Edited by Vamanacharya Ramabhatta Jhalakikar. Poon
a: Bandharkar Oriental Research Institute.
Nagarasarvasva (NS) of Padmasr (1921). Edited by Tanusukharam Sharma. Bombay: Manila
l Iccaram Desai.
Nagarasarvasva of Padmasr (1929). Edited with a Hindi translation by Srirajadhara Jh
a. Calcutta: Venkatesvara Book Agency.
Nat:yadarpan:a of Ramacandra and Gun:acandra (1986). Edited by by T. C. Upreti. Pa
rimal Publications, Delhi.
Nat:yasastra of Bharata with the Abhinavabharat commentary of Abhinavagupta (2003). Ed
ited by M. R. Kavi. vol. 3. Vadodara: Oriental Institute.
Pan casayaka (PS) of Kavisekhara Jyotsvara (1926). Edited by Jadavji Trikamji Acharya.
Bombay: Manilal Iccaram Desai.
Raghuva~msa of Kalidasa with the commentary of Mallinatha (2003). Edited by Haragovi
nda Sastri. Varanasi: Chaukhambha Sanskrit Samsthan.
Rasakalika of Rudrabhat:t:a (1988). Edited and translated by Kalpakam Sankaranara
yanan. Madras: Adyar Library and Research Center.
Rasarn:avasudhakara of Si~mhabhupala (1979). Edited by T. Venkatacharya. Madras: Ady
ar Library and Research Center.
62
D. Ali
Ratiman jara of Jayadeva (1976). Edited with a Hindi translation by Ramkant Dvived
i. Varanasi: Chaukhambha.
Ratirahasya (RR) of Kokkoka with the Dipika commentary of Kan cnatha (n.d.). Edited by
S. S. Ghiladia. Lahore: Punjab Book Depot.
Ratiraman:a of Siddhanagarjuna (2002). Edited and translated by Kenneth Zysk in Th
e Art of Conjugal Love: Ratisastra and Ratiraman:a. Leiden: Brill.
Ratisastra (2002). Edited and translated by Kenneth Zysk in The Art of Conjugal Lo
ve: Ratisastra and Ratiraman:a. Leiden: Brill.
Smaradapika (SmD) of Minanatha (1967). Edited by Dhundhiraja Shastri in Kamakun jalat
a: A Collection of Old and Rare Works on Kama Sastra. Varanasi: Chaukhamba Sanskrit Series.
Sahityadarpan:a of Visvanatha (Pariccheddas I-X) (1923). Ed. P.V. Kane. Bombay, Ni
rnaya Sagar. Siddhaikavaramahatantra (1998). Edited by Janardan Pandey. Varanasi:
Central Institute of Higher
Tibetan Studies.
Suktimuktavala of Bhagadatta Jalhan:a (1991). Edited by Embar Krishnamacharya. Bar
oda: Oriental Institute.
Secondary Sources
Ali, D. (2004). Courtly culture and political life in early medieval India. Camb
ridge: Cambridge University Press.
Bruns, P. (2009). Nad:avijn ana: A study of ayurvedic pulse diagnosis.

Comfort, A. (1964). Introduction. In Comfort (Ed.), The Koka Shastra: Being the
Ratirahasya of
Kokkoka and other medieval Indian writings on love. London: Allen and Unwin.
Das, R. P. (2003). The origin of the life of a human being: Conception and the f
emale according to
ancient Indian medical and sexological literature. Delhi: Motilal Banarsidass.
Davidson, R. (2002). Indian esoteric buddhism: A social history of the Tantric m
ovement. New York: Columbia University Press.
Finot, L. (1896). Les Lapidaires Indiens. Paris: Librairie Emile Bouillon.
Hopkins, J. (1992). Tibetan arts of love: Sex orgasm and spiritual healing. Itha
ca, New York: Snow Lion.
Krishnamachariar, M. (1974). History of classical Sanskrit literature. Delhi: Mo
tilal Banarsidass. Lienhard, S. (1979). Some observations concerning a Buddhist
text on erotics: The Nagarasarvasva of Padmasr. Central Asiatic Journal, 23, 96 103.
Lienhard, S. (1960). Kokkoka Ratirahaysa: Geheimnisse der Liebeskunst. Stuttgart
: Franz Decker Nachf.
Mylius, K. (2009). Das Ratirahasya des Kokkoka und der Anangaranga des Kalyan:ama
lla. Wiesbaden: Harrassowitz Verlag.
Pathak, A. S. (2008). Verses from other works in Kamasamuha. In A. S. Pathak (Ed.)
, Kama Samuha of Sra Ananta Kavi. Varanasi: Chaukhambha Sanskrit Samsthan.
Pingree, D. (1981). Jyotih:sastra: Astral and mathematical literature. A history o
f Indian literature (Vol. 6, fasc. 4). Wiesbaden: Otto Harrassowitz.
Raghavan, V. (1942). The Gun:apataka. The Indian Historical Quarterly, 18(2), 166 16
7. Raghavan, V. (1943). Date of the Ratirahasya. The Indian Historical Quarterly
, 19(1), 72 73. Raghavan, V. (1978). Bhoja s Sr:_ngara Prakasa. Madras: Purnavasu, 1978.
Schmidt, R. (1922). Beitrage zur Indischen Erotik. Berlin: Hermann Barsdorf Verla
g.
Sharma, T. (1921). Introduction and appendix. In T. Sharma (Ed.), Nagarasarvasva.
Bombay: Manilal Iccaram Desai.
Shastri, Hara Prasad (1901 1905). A report on the search for Sanskrit manuscripts.
Calcutta: Asiatic Society of Bengal.
Upadhyaya, S. C. (1981). Hindu secrets of love: Kokashastra (Rati Rahasya) of Pa
ndit Kokkoka. Bombay: Taraporevala.
Zysk, K. (2002). Introduction. In The art of conjugal love in India: The Ratisastr
a and Ratiraman:a. Leiden: Brill.