Rituals of First Menstruation in Sri Lanka
Rituals of First Menstruation in Sri Lanka
Rituals of First Menstruation in Sri Lanka
JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide
range of content in a trusted digital archive. We use information technology and tools to increase productivity and
facilitate new forms of scholarship. For more information about JSTOR, please contact support@jstor.org.
Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at
https://about.jstor.org/terms
Royal Anthropological Institute of Great Britain and Ireland is collaborating with JSTOR to
digitize, preserve and extend access to Man
DEBORAH WINSLOW
Rituals of first menstruation are found in most of Sri Lanka's many ethnic groups. The
rituals are similar and appear to be variants of a common south Asian ceremony. Much recent
writing on south Asia has emphasised such common cultural and structural features as keys
to understanding local, particular cases. At least in this case what appears to be a common
ritual is not as unchanged from one group to the next as it seems: differences in the way in
which the ritual is seen or understood in the various ethnic communities are related to
varying religious notions of womanhood.
The general south Asian type of first menstruation rituals are found among
almost all Sri Lankan people, of whatever]jatiya.4
For simplicity, I focus discussion on three ethnic groups-Buddhist,
Catholic, and Muslim-who differ not only in religion but also in language,
social organisation and mode of subsistence. Their first menstruation rituals are
so alike in structure and content that the few differences there are appear minor
in comparison. In fact, a structural analysis devised by Leach (I970) for the
Sinhalese Buddhist version might easily be used for other versions too.
The problem then is what makes the rituals significant to all three groups ?5
It is a problem worth considering because from the time of Durkheim and
L'Anne'e Sociologique, social anthropologists generally have assumed that rituals
are part of a structured universe of social relations to which they in some way
relate; that is, they are collective representations. I suggest that for these rituals
the differences although few are significant; and that they relate to (or, are
brought together in) not so much differences in social structure itself as
contrasting images of women.6
Yalman's analysis of female puberty ceremonies of Buddhists in Sri Lanka
and Hindus in south India emphasised social structure. Beginning with the
notion that, 'Highly formalised collective rituals always do reflect the structure
of the collectivity' (i963: 54), Yalman suggested that the ceremonies express
a concern with caste purity in societies where caste status is inherited bilaterally.
However, when the focus is widened to include other groups in Sri Lanka, we
find that ritualisation of puberty and bilateral caste filiation do not inevitably
go together (also, cf Leach I970: 827). Some Catholics, including those
discussed below, say that caste is patrilineal; there are Hindus who consider it
to be matrilineal (McGilvray in press); and Muslims are not concerned with
caste at all. And yet all these people have the ritual of first menstruation. While
Yalman's theory may be appropriate for the cultures he considers, it is difficult
to extend it to situations where caste ideology varies.7
Leach's response to the filiation problems of Yalman's theory was a van
Gennep-style reinterpretatioi of Buddhist ritual. The familiar three stages of
separation, transition and reintegration, Leach suggested, compose an idiom
expressing the over-arching theme of 'maturation through separation' (Leach
I970: 825). For my purposes, Leach's analysis has the advantage that it is
applicable to all the ritual variants; however, by the same token, it does not
account for differences either in the variants themselves or in their social
contexts. Indeed, with Douglas's (i966) additions to van Gennep's (i909)
original formulation, it is likely that Leach's interpretation would hold not
only for all Sri Lankan ceremonies but for female puberty rituals in many
other places as well.
Leach seems particularly concerned with what Geertz-in an article on
art-has called 'craft terms' (I976: I474-5). Leach elucidates the technique
(in this case, oppositions) by which transition is facilitated and meaning
conveyed, but is less interested in the nature of the meaning itself. Along the
same lines as Geertz, La Fontaine has pointed out that describing the mechanism
by which transition rituals work is not the same as an explanation of what such
rituals are (I 977). She added that an analysis cannot fall back on the concept of
Information sources
These three descriptions of first menstruation rituals are from villages in
different areas of Sri Lanka. The areas are culturally as well as geographically
distinct, even though two of the villages (Walangama and Wellagoda) are
only about thirty miles apart. The Buddhist village, Walangama, is in a
predominantly Buddhist region, and although well out of the mountains, its
people consider it culturally part of the central Kandyan area. The Catholic
village of Wellagoda is in a Catholic-dominated section of the western coast.
The western and southern coastal areas, the 'low country', are separated by the
influences of longer exposure to colonial rule, as well as topography, from the
Kandyan 'up country'. My description of the Muslim ritual comes from
Akkaraipattu on the opposite coast, in an area where Buddhists and Catholics
are in a minority and Muslims and Hindus predominate.9
The Buddhist version is from my own observations of Walangama, a small
(population about 500) pottery caste (Badahala) community, which depends
almost entirely on pottery for income. Like the Sinhalese described by Yalman
and Leach, the people in Walangama view caste as bilaterally inherited,10
prefer cross-cousin marriage, and employ a Dravidian kinship terminology
system. First menstruation rituals are held regularly for almost all girls in
Walangama.
Most of the approximately 750 people in Wellagoda (about 50 miles north
of Colombo) are of Karava caste, whose traditional occupation is fishing, the
village's major source of income. Roman Catholicism, brought to Sri Lanka by
the Portuguese in the sixteenth century, was revived by nineteenth-century
French missionaries. French influence is still seen in the popularity of certain
saints (Wellagoda's patron is Saint Rogus, a minor French saint), and elsewhere
in Sri Lanka in the frequency of shrines modelled on that at Lourdes (Stirrat
I979). Both Tamil and Sinhalese are spoken in Wellagoda. These Catholics
attribute caste status patrilineally, the mother's caste being irrelevant to that of
the child, and employ a non-prescriptive variant of Dravidian kin terminology
(Stirrat I977b). I did not myself see a first menstruation ritual in Wellagoda;
fortunately, I interviewed a woman who only a few days before had held one
for her daughter so that finer points were still fresh in her mind. Otherwise,
my information represents several women's jointly discussed view of how the
ritual ought to be done and, I understand, is done for at least most Wellagoda
girls. Other information on Wellagoda is from the work of R. L. Stirrat (I 973;
I975; I977a; I977b), who introduced me to the village.
My account of the Muslim ritual is also based on interviews held in a
previously-studied village (McGilvray I974). Akkaraipattu (about 40 miles
south of Batticaloa) is typical of its region in that the village is divided between
Hindus and Muslims and has a nucleated farming population of over 50,000,
surrounded by miles of rice fields. Both Hindus and Muslims are organised
into ranked matrilineal descent units and prefer cross-cousin marriage; only
Hindus however recognise a caste system. The status of menstruation rituals in
Akkaraipattu today is unclear. McGilvray (in press) suggests that in its
elaborate form the ritual is no longer held. I talked separately with two elderly
women at some length and they discussed the rituals easily in great detail. Both
seemed to regard it as a ceremony that might still be held, and gave accounts
quite similar to what I saw done by Muslims in another area, near the town of
Kandy. My description is from the two interviews and from McGilvray's
fieldnotes and writings (I 9 7; 1 976; in press).
This first menstruation ritual was held quietly and few in the village knew it
had happened. Later, when a friend asked what had been done, Kanti's mother
a winnowing basket containing oil cakes (kdvum), milk rice (kiribat), and bananas (dmbul
kehel)-all routine celebrative sweets; a small pile of husked rice, a pumpkin, onions, dry red
chillies, dry fish, brown palm sugar, salt, and tea leaves-items considered basic to a well-
stocked kitchen; and a small amount of money. On either side of the mat were two clay water
pots which contained young coconut flowers and a burning oil lamp.
While the washerwoman instructed her, Chandra combed her hair with the comb and
looked at herself in the mirror; gazed at all the objects on the mat; and, finally, extinguished
the two oil lamps.
My informants had no explanation for the comb and mirror. Mirrors are
sometimes used like water to remove vas; however, its combination with a
comb was also reminiscent of village exorcism offerings to a goddess, which
often include either a real comb and mirror or 'play ones' cut of plantain stalks.
Having her see the other objects on the mat were to ensure the girl a good
future.
The two white cloths, that on the mat and the one used to cover the girl,
were cloths of the category pivuva.ta. These are particularly clean (sudu, which
also means white) and pure (pirisidu) sheets provided by washer caste people
for various rituals, especially domestic transitions such as births, first
menstruations, weddings and deaths. The washer people are meant to remove
dirt, actual and ritual. To make a ritual clean is their particular task and the
white cloths they provide are both the fact of that task and its symbol.
Appropriately, though perhaps only coincidentally, menstruation in colloquial
Sinhalese is 'sudu veneva', becoming white or clean.
Chandra then retired to an inner room emerging shortly in new clothes andjewellery, gifts
from relatives, and carrying a pile of betel leaves and a handkerchief. She first approached the
redi nanda, put the handkerchief on the floor in front of her, respectfully offered the woman
some betel leaves, and then knelt and touched her head three times to the handkerchief. The
redi nanda then placed her right hand on the girl's bent head and softly blessed her, forgiving
her for the killa of that day and of the future. Chandra then repeated her obeisance to each
older relative in turn.
This was much as she would do if she were either leaving for or returning
from a long or important journey, before a Buddhist monk (bhikkhu), or in
some households to her parents every evening before retiring. It suggested, I
think, her return to the world as a mature woman as well as the respect she
owed her elders.
The washerwoman then departed taking with her the mat and all the objects on it; the
clothing and jewellery the girl had been wearing when her menstruation started and during
her confinement; the girl's own sleeping mat; the two water pots used in the bath; a package
of oil cakes and bananas; and a further gift of twenty-five rupees, all of which were considered
hers by right (ayiti).
After the redi ndnda left, the family served oil cakes, bananas and milk rice to the relatives
and neighbours. Later, the astrologer made an offering of flowers, betel leaves, and incense to
demons attracted to the girl. He also said a protective formula (yantra) over a string which he
tied on Chandra's arm to keep away yakshas and vas. He reminded her mother not to allow
Chandra to be alone during the next three months, the duration of the killa. Nor should
Chandra nor anyone else of the house go to a god's shrine (devale) or make offerings to a god
during the same period.16 Shortly afterwards, I heard that Chandra's parents were giving
serious consideration to a marriage proposal.
personal gifts of clothing and jewellery. They are served oil cakes, bananas and arrak.
cases, the celebration can be quite elaborate, with a great deal of food, alcohol and loud-
speakered music-'like a wedding'.18
The minimal ritual for the Muslims consists of isolating the girl-to avoid the bad luck of
seeing men-in the 'third room', a non-communicating room off the kitchen where grain is
stored and traditionally women gave birth. The girl must be accompanied by a female person
or brother, and a piece of iron or a broomstick kept nearby to ward off demons (pey, the
Tamil equivalent of yaksha). She may be given doses of oil for pain; have turmeric applied to
her body for purification, cooling and to prevent future skin diseases; have her back poulticed
with hot rice for strength; and be restricted to a diet of vegetables, milk, and eggs, without
the meat and fish she would normally eat. When the flow stops, she is taken to the well and
bathed by her mother, older sisters, and/or female cross-cousins, who use a mixture of water
and turmeric.
One major similarity between these three sets of rituals is the way the physical
commencement of menstruation is used to mark a major change in female
social status. For this reason, girls go through the ceremonies as individuals, not
as members of a social set, in contrast to cultures in which social and
In sum, the three ritual variants are all concerned with structural transition,
with protecting the health of the girl, with warding off demonic forces, and
with the girl's new status as marriageable female. Indeed, the similarity is such
that some washerwomen do them for more than one ethnic group. In
Akkaraipattu, either a Hindu washerwoman or a Muslim specialist may assist
the Muslim girl; and in Kandy, I interviewed a washerwoman who helped
with the first menstruation rituals of Muslims, Catholics and Buddhists, as well
as Hindus like herself.
In one sense, there is nothing surprising about the cultural similarities. The
three groups all have south Asian origins and share in a common south Asian
cultural system. For the last i 6 5 years, they have been part of one political unit,
Sri Lanka. An analysis that produced a similar account no matter which
version was chosen for attention could readily bejustified in terms of persisting
cultural themes and intergroup contact. Such an analysis would, however,
cover up what I see as important differences.
These differences lie less in what is done in the ritual than in what people say
about what they are doing. Specifically, there are hints in the way the
informants from the three ethnic groups talked about menstruating girls that
suggest they held different ideas about the inherent nature of women, and that,
at least in part, they saw the ritual in terms of those ideas. For example, in the
Buddhist ceremony the girl was isolated partly for her own protection but
more importantly to protect male members of the society from the destructive
power of her killa. In the other two versions, however, she was isolated
primarily for her own protection. Likewise, they were all concerned with
setting an auspicious time for the bath, but while Buddhists derived the hour
from the girl herself through her horoscope and the hour of onset, Catholics
used the liturgy of their religion, and Muslims used numerology and the girl's
mother's name. These minor points suggest differing images of the girl herself.
The onset of menstruation seems an appropriate time for making statements
about what it means to be a mature woman and to have been an immature girl.
This is not necessarily the one true meaning of menstruation rituals in Sri
Lanka. But it is an aspect that cannot be derived solely from the ritual apparatus.
The relationship between the rituals and their social contexts is elucidated only
when we go beyond the mechanisms of separation, protection, transition and
so on, to look at what people think is being separated, protected and transferred.
When we do this we find that the rituals are not identical. Each relates to and
expresses a contrasting image of women, in each case strikingly similar to that
group's major female religious figure: the goddess Pattinm, the Virgin Mary
and the Prophet Mohammed's daughter Fatima.
It is not surprising that important differences between the three versions of
the first menstruation ritual derive from religious ideology. Religion has been
a crucial distinguishing feature of ethnicity in Sri Lanka since the sixth century
A.D., when aggressive Hinduism in south India radicalised Buddhists in Sri
Lanka (de Silva I977: 38). In the late nineteenth and early twentieth centuries,
the Independence struggle was often phrased by Buddhists in relation to their
legitimate right to power in Sri Lanka by virtue of being Buddhist (Malalgoda
I976: I9I, sqq.). Since Independence, communal conflict has arisen repeatedly
so that her worshippers can walk over them unharmed; and she ends droughts
with cooling water (Obeyesekere I 976: 203).
Pattini is constantly referred to as pure (pirisrdu). She was born from a
mango and so is free even of the pollution of birth. She married but managed
to remain a virgin and yet be a good wife by giving her husband wealth to pay
for the favours of other women. Therefore she is childless; but in Walangama
women said that PattinT was the mother and real possessor of all children. Each
child born is really with her until it first eats solid food; the parents have only
the child's image. Pattinl is thus both a gentle, loving wife and mother, and a
terrible, virginal power for good and ill.
Like Pattini, little Buddhist girls are said to be innocent of sexual knowledge
or desires (Obeyesekere I976: 2Io). Their purity and goodness is a power in
itself I attended curing exorcisms at which little girls were dressed in white
and seated behind the patient; at intervals they were asked to call out blessings
to help rid the patient of the possessing demons. At weddings, again in white,
little girls chorus blessings on the new couple.
When a girl begins to menstruate, she is almost without warning a danger
to all around her. First menstrual killa is said to be the worst of all, greater than
that of either birth or death. One city woman said that a house sparrow flying
above a ko.tahala gedara is liable to burst into flames from the killa. Villagers
repeatedly said that almost healed wounds open up again and other diseases
occur near a kopahala gedara. Reading the girl's horoscope, the astrologer may
warn of impending disaster and order a particular member of the family to
avoid her for many months. The villagers told stories of such warnings and of
the deaths and accidents that followed when they were not heeded. On the
other hand, the astrologer may foretell good news and stories were also told of
families receiving sudden riches and good fortune.
A mature Buddhist woman is excluded from many ritual undertakings.
Even when she is not menstruating, she cannot prepare food for deities or for
demons in exorcism rites, nor may women in Walangama make pottery for
use in devales. In many areas, she cannot participate in harvesting rice, a semi-
sacred food as well as a dietary staple. A menstruating woman is not allowed
onto a threshing floor, nor into an area where betel leaf (important in many
rituals) is being grown; she cannot go into a devale, make an offering to a deity,
nor go on a pilgrimage.
When a woman has passed through her menstruating years, she is again
considered to have powers of blessing. Seven such women, dressed in white,
are called in to receive a special almsgiving at a ceremony given in honour of
Pattinm to ensure the health of a child (kiriammalage dane, the grandmothers'
almsgiving). In Walangama, these women are said to represent the seven forms
of the goddess. After receiving food offerings from the child's family, the
women chant blessings over the child and through this, as well as by placating
the goddess, are said to cure or protect it. Similarly, a woman around
menopausal age is expected to wear white clothes and to spend increasing
amounts of time in Buddhist temples, meditating and observing Buddhist
precepts. At a wedding, the groom's family gives the bride's mother a length
of white cloth explicitly to dress her for her new, pure and (I think) asexual
danger in the way the Buddhist girl is. I inquired of my Catholic informants
about the killa in first menstruation. They were quite shocked at the suggestion
and responded vehemently that there is none because the girl is pure and
innocent. Killa, they said, has nothing to do with first menstruation; only at
death and birth and in the skin eruptions of diseases is there killa. The bath in
the menstruation ritual is to make the girl clean again physically and so to
protect her from demons and from vas, both external agents; menstrual blood
is dirty (kiata) not polluting (killa). The danger is to her, notfrom her, which
may be why Catholics allow her close male relatives (who pose no sexual
threat) to see her while Buddhists prefer that they stay away.
The women in Wellagoda themselves compared the girl to the Virgin
Mary when explaining why an image of Mary was kept with the girl during
her seclusic,n. The girl is an uncontaminated virgin; in menstruating she pays
for sins not of her doing, and while menstruation causes her suffering it does
not devalue her. She could yet reject her own sexuality and become a nun,
which the villagers would consider virtuous (Stirrat I975: 596).
The Virgin Mary embodies an image combining the perfect wife and
mother with the perfect virginal spiritual being (Stirrat I975: 6oi). Because
of marriage most women will not be able to maintain their own spiritual
perfection. But this is their doom, not their nature; it is not inherent in them
as female, but rather is a simple consequence of historical acts. The good and
pure Virgin differs sharply from the potent and dangerous virgin goddess,
Pattini; and in a manner almost parallel the Catholic image contrasts with the
Buddhist image of women.
Fatima
The figure of the Prophet Mohammed appears quite different from that of
either Buddha or Jesus. Mohammed was not an asexual just barely human
being; he was a normal man, though sinless and God's prophet (Ali n.d.: 220
sqq.). Mohammed was born of two human parents (both died in his early
childhood and are rarely referred to), and he married and raised a family.
Necessarily, then, the women involved in his life-his mother, wives and
daughters-were human and natural themselves. Mohammed's favourite wife,
Ayesha, and his daughter, Fatima, are important in Islam, but not as
supernatural and powerful virgins. Rather, they are portrayed as true examples
of how human wives, mothers and daughters should be. Numerous Hadith,
records of Mohammed's words and actions which guide all Muslims, refer to
these women when explaining proper female behaviour. Yet female religious
figures appear relatively unimportant in Islam. One of my informants in
Akkaraipattu did mention that during the girl's seclusion she should not eat
certain foods because Fatima did not. McGilvray (pers. comm.) had never
heard of such a reference to Fatima, so it may be idiosyncratic and unusual.
However, Fatima is also reported to be honoured by women in domestic
rituals among Colombo Muslims (Thawfeeq n.d.: 68-9), so she is of at least
minor importance in the belief system of some Sri Lanka Muslims.
While it is not clear whether Fatima (or Ayesha) provides a model of the
ideal woman for Akkaraipattu Muslims, it seems that if there were such an
ideal, then it would be of a woman like Fatima-normal, natural and human.
Similarly, I think that this is the ideal hope of women that was conveyed in the
traditional first menstruation ritual.25 The Muslim ritual of first menstruation,
far more than either of the other two versions, was in the past a celebration of
the fact that the girl could thereafter assume the roles of wife and mother. All
informants stressed the large numbers of relatives, particularly actual and
classificatory affines, who would come and the revelry involved. If the girl
were already married, the celebrating was particularlyjubilant for the marriage
could then be consummated and completed. If she were not married, then
marriage negotiations began during the context of the menstruation rituals
themselves. The girl's ritual status, even her individuality, was of hardly any
concern at all. The gifts given to her followed the overarching emphasis on her
domestic roles: while Buddhist and Catholic girls received trinkets for their
own adornment, Muslim girls received household objects such as glassware
and utensils.
With the onset of menstruation, the Muslim girl becomes a complete
woman and fulfilled her religion's female ideal more completely than she has
as a child. This contrasts sharply with the situation of the Catholic or the
Buddhist girl for whom menstruation means less rather than more religious
perfection. The image of the perfect woman in Islam is of complete and natural
humanness. Whether or not the Prophet's daughter Fatima is important in
conveying this image, her attributes are those accentuated in the Muslim
version of the menstruation ritual.
Conclusion
People everywhere have developed symbolic structures in terms of which persons are
perceived not baldly as such, as mere unadorned members of the human race, but as
representatives of certain distinct categories of persons, certain specific sorts of individuals
(Geertz I966: 6).
We have seen that the category 'woman' is different in the three religious
traditions. Because first menstruation rituals are about women, the specific
nature of femaleness affects the meaning of the ritual. The differences appear
as differences in emphasis. All three variants are concerned with restoring
ritual cleanliness, with providing protection from potential dangers, and with
defining and proclaiming adult female social status. But the Buddhist version
emphasises the first concern, in accord with a tradition that portrays fertile
women as potent and dangerous; the Catholic ceremony emphasises the second
concern, which fits with their belief that virginal women are innately pure and
so threatened by the world around them; while the traditional Muslim ritual
elaborates the third aspect and focuses attention on the girl's new domestic role,
just as Islam itself emphasises the domestic virtues of its major female religious
figures. In short, Sri Lankan rituals of first menstruation exist within different
social conceptions or images of what it means to be female at the age of
puberty.
NOTES
This article is based on information I collected during I973-I976 while doing fieldwork in
Sri Lanka on grants from the Center for Research in International Studies (Stanford), the
National Institute of Mental Health, and the National Science Foundation. I wish to acknowledge
with thanks research assistance from C. and D. McGilvray, S. Potuhera, S. Senanayake, R. L.
Stirrat, I. G. Wimala, and women in Walangama, Wellagoda, and Akkaraipattu. I am also
grateful to K. Kramer and W. Sperling for help in understanding Fatima; and to J. W. Burton,
D. McGilvray, M. Z. Rosaldo, R. Rosaldo, G. W. Skinner and R. L. Stirrat for helpfully critical
comments on earlier essay drafts.
1 Indian examples are described in Gough I955.
2Jatiya and varige were used (in Walangama, see text below) to distinguish ethnic groups, as
well as simply to mean 'kind' or 'sort'. By ethnic group, I mean a community which is 'largely
biologically self-perpetuating, shares fundamental cultural values, makes up a field of
communication and interaction, and has a membership which identifies itself, and is identified
by others, as constituting a category distinguishable from other categories of the same order'
(Mines I975: 408). The three groups considered in this article are separate ethnic communities
according to this definition, and religion is a major component of their distinctiveness. However,
ethnicity in Sri Lanka is sufficiently complex that it is frequently difficult to define communities
except in relation to one another. Religion and language even though important are not
unfailing indicators of ethnicity; and within each group, social organisation as well as economic
base may vary considerably. Thus not all Catholics attribute caste patrilineally nor speak
Sinhalese; Tamil-speaking Muslims on the west coast are culturally different from those on the
east coast; and so on.
3 Ryan (I950) provides an overview of ethnic and linguistic regional variation in Sri Lanka.
He characterises the north as Hindu Tamil; the eastern coast as Tamil-speaking Hindu and
Muslim; the western coast as predominantly Sinhalese Buddhist to the south with increasing
numbers of Catholics to the north; the central and southern areas as Sinhalese Buddhist. These
generalisations are rough: almost every area is mixed, and the fact that the country as a whole
is predominantly Sinhalese Buddhist is important even in areas where few Sinhalese Buddhists
are present. Also, see Note 9, below.
4 Published descriptions include: for the Vadda, Seligman & Seligman I9II: 94 and
Wijesekere I964: 99-IOO; for Indian Estate Tamils, Green I926: 33-6I and Velupillai I970:
9-I 3; for Sri Lanka Tamils, Cartman I957: I 50-I and McGilvray in press; and for Muslims,
McGilvray in press and Thawfeeq n.d.: 69. I have also collected accounts of ceremonies held for
Portuguese Burgher Catholic girls in the Trincomalee area, for Tamil-speaking Catholics and
Protestants in Kandy, and for Sinhalese-speaking Catholics inland near Kurunagala.
5 Another question, not considered here, is what is so compelling about this ritual that
seemingly any group given opportunity adopts it. An interesting example appears in the
Seligmans' classic account of Vadda tribals in Sri Lanka (Seligman & Seligman i9i i). Their
observations (I 9 I I: 94) suggest that while Vaddas without contact with Sinhalese peasants had
no female puberty rituals at all, those influenced by peasants even marginally had adopted the
entire ritual structure.
6 This distinction is similar to that Geertz draws between social system and culture in his
analysis of a funeral in Java (Geertz I957).
7 Yalman seems to equate caste purity with bodily purity. He describes the Sinhalese first
menstruation ceremony as a process whereby 'the ... [polluted] ... infertile child becomes a
pure and fertile woman', thereby, 'the honour and respectability of men is protected and
preserved through their women' (Yalman I963: 32-3). However, in my experience (also,
Gombrich I97I: I83) Sinhalese Buddhists do not consider these two notions of purity to be the
same. The dangerous contagion of menstruation, death, birth, and oozing sores is killa. This
word is not applied to the quality of castes lower than one's own that makes intermarriage and
ritual interdining unacceptable. It may be that caste status is related to contact with pollution,
but for the Sinhalese caste status does not appear to derive from innate states of relative purity
as it does in India. This is supported by the fact that in Walangama (described below in the text),
notions of social rank and social differentiation are also kept conceptually separate, the first being
called kula and the second varige or jatiya (also meaning 'ethnic group'). It is probable that
Sinhalese notions of caste differ from Hindu ones precisely along these dimensions.
8 This is not to imply that there has to be a one-to-one correlation between the ritual and the
experience to which it speaks. Our one ritual, without variation, could relate to quite different
social experiences. Something of the kind frequently happens within social groups-when one
family holds a first menstruation ritual for its daughter this means not quite the same thing to
either them or to others as when another family holds the same ritual for its daughter. The
variables involved might be economic status, the parents' political expectations, and so on. We
certainly cannot assume that the same ritual in two different social groups involves the same
experience either.
9According to the I97I Census of Sri Lanka, the overall religious composition of the country
is: 67.4% Buddhist, I7-6% Hindu, 7I% Muslim, 6&9% Roman Catholic, o-8% other
Christian, and o I % Other. In the area around Walangama, the religious composition is similar
to that for the country as a whole. However, around Wellagoda, the area is about 65% Roman
Catholic and only i6% Buddhist, with Hindus at 8% and Muslims at 9%. While in
Akkaraipattu itself, Buddhists are even fewer at 2.5%, and Catholics are only I 9%; Hindus are
more dominant at 36% and Muslims are in the majority with 5 8 9% of the population. These
figures clearly indicate the significant regional variation in religious and therefore ethnic
composition. All percentages are rounded to the nearest tenth, which accounts for discrepancies
in the totals.
Source: Sri Lanka Department of Census and Statistics, 1971 Census, Preliminary Release
No. 1.
IO People in Walangama do not always weight the two sides equally. I have heard women say
that a man contributes somewhat more blood to the child than does the woman and thereby has
more influence on the child's caste.
11 Neither the villagers nor a standard Sinhalese dictionary give any meaning for kotahala
besides the ritual itself. Alahackoon (i 893) defined it as 'short cloth' and asserted that the ritual
is about the girl changing her child's knee-length skirt cloth for the longer one of an adult. 'Ko.ta'
is a common word for 'short' and the explanation seems logical in that it makes sense of the girl
giving up her household goods.
12 Heating and cooling are repeated themes in south Asian rituals. Heat is powerful but
dangerous unless controlled through cooling (Beck I969).
'3One washerwoman explained that the killa of first menstruation was dangerous to her and
that she took precautions to protect herself and her family. She was careful to leave in the
morning without seeing any member of her family and to bathe before returning. Therefore
killa is a problem that caste status does not erase; the fear than an unhappy washerwoman would
not assume the burden willingly may be well-founded. But the washerwoman also said that she
had to go to give the bath if asked, or she would risk becoming ill herself.
14 Yalman says, '. .. The new horoscope, drawn upon the beginning of menstruation,
supersedes the birth-horoscope of women. It is, as it were, a new life that they are beginning.
Men have only one horoscope' (Yalman I963: 29). My informants contradicted this flatly. I
made inquiries from three professional horoscope-makers: a Buddhist priest, a deity shrine priest
(kapurala), and an exorcist (kaltadiyp). They all presented the same story. A person's horoscope
is derived from the disposition of stars and planets at the moment of his birth. This is
unchangeable. But by knowing this, interpretations can be made regarding the efficacy and
auspiciousness of other times by comparing the positions of the planets and stars at those times
with their positions at birth. For a woman, the moment at which she starts menstruating is
considered an especially informative point for comparison. It is said to be particularly useful for
predicting her domestic future-the outcome of her marriage, childbearing, and so on. But this
is not a new horoscope; it is simply a reading of her horoscope in relation to a fixed point in
time.
' Gombrich (I97I: I 3 8) describes how the potent vas of completing a statue of Buddha is
evaded with the use of a mirror and water.
16 As this indicates, the gods are considered to be sensitive to any sort of killa and would be
made angry by contamination. At an important devale near Kandy I was told that whenever any
girl in the village started menstruating for the first time, the shrine would be closed for three
months. This contrasts with Buddha and his shrines which are not said to be sensitive to killa.
17 There is variability in some aspects of Catholic culture in Sri Lanka. Other Catholics I
knew, who lived about a mile from Walangama, consulted astrology and attributed killa to the
girl. They also attributed caste bilaterally. It must be made clear that I am talking about
Catholics in Wellagoda only, not those elsewhere in Sri Lanka who live among Buddhists.
18 A first menstruation held while Stirrat was living in Wellagoda was for a crippled girl.
Her family held an especially elaborate party on the occasion of her maturation, with all the
music and food that would normally have been in the wedding they feared her handicap would
keep her from having (Stirrat pers. comm.).
I Post-Vatican II Catholicism has reached Wellagoda and the barring of women from
church after birth or during menstruation is recognised as wrong. However, women with
whom I talked said that they did not feel clean and so did not usually go, although other women
might. Stirrat reports that the attempt to prevent the 'churching' of women after childbirth by
insisting on their presence at the baptism of their children has resulted only in delaying baptism
(Stirrat I975: 605). There is a merging of Buddhist and French Catholic culture at this point,
so it is hard to say whether Wellagoda notions of pollution originate in one source or the other
(cf., Warner I976: 75-6).
20 McGilvray (in press) suggests the possibility that the ritual is seen as really Hindu and has
come into disfavour because of recent attempts to purify Sri Lanka Islam of Hindu influences
(see, Mines I975).
21 Because of the relatively late age (I3-I7 years) of onset of first menstruation, marriage
soon afterwards, and the prevalence of breast feeding which retards ovulation (Kippley I975),
menstruation is not a common condition for women in rural areas of Sri Lanka until just before
menopause. When a young woman menstruates the implications tend to be negative: she is not
married and should be, or she is not pregnant and should be. That menstruation is infrequent and
problematic may heighten the tensions surrounding first menstruation.
22 Both Buddhists and Catholics suggested that the girl might become obsessed with sex if
exposed to men during this time. This fits the fact that both regard sexual activity as polluting
and damaging to the girl.
23 The first menstruation ritual itself may be regarded as such an element. A few Catholics
allege that it is really a Buddhist ritual, while some Buddhists and Muslims label it Hindu.
Foreign origin would not negate the fact that while practised the ritual was part of each group's
separate culture, even if now it does not fit in with notions of ethnic discreteness.
24 Interestingly, villagers apparently see a contradiction between this idea of Mary and the
idea that she was a normal human being like themselves. Stirrat (pers. comm.) noted that when
asked if Mary menstruated they would say she did not because she was perfect and then would
add that she must have since she was a normal woman. Clearly, they see menstruation and
spiritual perfection as opposed.
25 This may be another reason the ritual is now reduced. Not only would it be embarrassing
to the girl and her family to have her menstruating status known, but also since she is not about
to marry, the ceremony as Muslims perceive it is no longer relevant.
REFERENCES