Sunteți pe pagina 1din 17

FRITS STAAL

SQUARES AND OBLONGS IN THE VEDA

THE DHIS. N
. YA HEARTHS

Vedic geometry of the Sulba Sutras of the Yajurveda is generally


associated with the construction of the trapezoid Mahavedi and with
Agni, the bird-shaped ahavanya altar of the Agnicayana which is piled
in five layers, each of 200 kiln-fired bricks. The piling of altars from
bricks requires constructing figures equal in area to a given figure such
as a square or an oblong (or rectangle). The theorem of Pythagoras
or Baudhayana is used in several of these constructions. Historians of
mathematics Seidenberg, van der Waerden and others have shown that
this way of doing geometry, which Seidenberg (1983: 113) calls The
Philosophy of Equivalence Through Area, is Greek as well as Vedic
but is not found in Mesopotamia, China or anywhere else in the ancient
world. On how Greek and Vedic may be related, see Staal, 1999.
The Vedia Sulba Sutras also describe altars that are much simpler
than Agni and seem to predate the equivalence philosophy. The most
important among these brick-piled alters are the dhis. n. ya hearths. They
are sun-, not kiln-fired. They suggest a development or evolutionary
progression in the shapes of bricks from squares to oblongs, pointing
beyond to the triangles that are used in Agni. The first two steps in
this development are not obvious because, from a geometrical point
of view, the square is a special kind of oblong or rectangle: one, in
which two adjacent sides (and therefore all four sides) are equal in
length. Oblong, therefore, is a more general concept than square. In the
prehistoric development of altar bricks, however, squares came first.
That is not surprising because humans perceive the species or vises. a
prior to the general idea or jati from which it logically derives. That
preference for particular knowledge may characterize other animals: for
a bird knows its nest and may recognize another birds nest as similar;
but may not know that a monkeys or insects nest is also a nest. If this
is the pattern of development of (proto-) geometry, it appears likely that
altar shapes such as we find in the dhis. n. yas are not only precursors of
Vedic geometry, but must exist or be postulated as precursors of the
Greek as well.

Journal of Indian Philosophy 29: 257273, 2001.


c 2001 Kluwer Academic Publishers. Printed in the Netherlands.
258 FRITS STAAL

Other altars are earlier than the dhis. n. yas since they are made from
unbaked clay, not bricks. The R. gveda mentions altars thrice and since
it does not know bricks (the term is. .taka occurs only in Brahman. as and
the later Yajurveda), these must be made from clay. One mention, a
riddle about a girl with four tufts (catus. kaparda in R . gveda: 10.114.3;
Potdar, 1953: 73; Staal et al., 1983: 129) implies a quadrangular shape.
It is generally taken to refer to the vedi (Geldner; Renou, 1967: 163, ubi
alia) but since the girl is butter-faced (ghr. tapratka), an expression
elsewhere used of Agni who is fond of butter oblations, it is more
likely that it refers to a fire altar and not the vedi, a receptacle for
ritual implements in which no fire occurs. The altar intended must
be the ahavanya, traditionally square just as the traditional shape of
the garhapatya is circular. The hypothesis that brick-piled altars came
from the BMAC (Staal, 1999) is consistent with Russian archeologist
A. M. Mandelstahms idea that the square and circular shapes are
foreshadowed in burial sites of the earlier and more northern Andronovo-
inspired Bishkent culture of Ferghana, as is now generally accepted by
archeologists (Parpola, 1998: 125; Staal, 2001a, ubi alia). For geometry
or baking, circular bricks are more difficult than squares or oblongs
which explains that the old, clay garhapatya was circular but the new,
brick-piled garhapatya became square.
Arguments in support of the BMAC origin of the Vedic bricks include
the term for brick, is. .taka (Avestan istya), which cannot be explained
from Indo-Iranian or Indo-European and which Michael Witzel derived
in 1992 from a Central Asian, probably BMAC (Bactriana-Margiana
Archaeological Complex) substrate-language. I mentioned this in 1999
(page 120) but my hypothesis has since been strengthened considerably
by a flood of recent information on Vedic names and words of BMAC
and other Central Asian provenance. Animal names, such as us. .tra
camel or kasyapa tortoise, and terms of material culture associated
with agriculture and brick-built settlements come from there (Witzel,
1999a: 342343, note 26; 1999b: 5; Lubotsky, forthcoming). Some
of these words have a characteristic morphology: mayukha wooden
peg (fixed in the ground, like sankhu, perhaps also for the sulba cords
of geometry; cf. Witzel in Wezler, 1980: 155160), kapota pigeon,
kapala dish, bowl, and several others are trisyllabic nouns with a
long middle syllable, difficult to explain on the basis of Indo- European
morphology and characteristic of the Central Asian layer according
to Alexander Lubotsky.
Another and even stronger argument for BMAC origins of Sulba
geometry comes from the striking similarities between the dhis. n. ya
SQUARES AND OBLONGS IN THE VEDA 259

hearths and hearths, altars and platforms that have been excavated by
J.-F. Jarrige and others at Pirak, a BMAC outpost southeast of the Bolan
Pass. We shall study these similarities in greater detail but first look at
the Vedic background of the dhis. n. ya hearths themselves.
There are six dhis. n. ya hearths in the Agnis. .toma and other Soma
ceremonies. They are located within the Mahavedi in the Sadas along a
north-south line. Six priests sit to their west and perform their R. gvedic
sastra recitations facing east. These reciters are called hotraka, because
they assist the Hotr. ; and are all camasin, i.e. they are entitled to a
camasa or Soma cup. Going from north to south, Acchavaka, Nestr. ,
Potr. and Brahman. accham . sin sit north of the (east-west) pr. s. .thya-line;
1
the Hotr. on it; and the Maitravarun. a to its south. Two other structures
are closely associated with the dhis. n. ya altars. They are located on the
boundaries of the Mahavedi, half within and half without: to the north,
the agndhrya, the hearth of the Agndh or kindler of the fire; and to
the south, the marjalya, not a hearth but a pantry where Soma vessels,
cups and goblets are cleaned (Caland-Henry, 19061907: 104106;
Krick, 1979: 366; Minkowski, 1991: 71, etc.).
The agnidhrya hearth is sometimes called a dhis. n. ya, which would
turn their total number into seven so that it corresponds to the original
seven priests of the R. gveda but the correspondence is not easy (cf.
Henry, 1907; Krick, 1979: 417418; Minkowski, 1991: 113; Witzel,
1997: 267 with note 45 which describes the replacement during the
Kuru period of the original seven by the 4 4, four for each of
the four Vedas). Equally puzzling is the correspondence between
Vedic and Avestan ritual where the Zaotar has seven auxiliary priests.
The only two that correspond closely are Vedic Agndh and Avestan
Atr uuaxs, for the simple reason that there can be no fire altars unless
there is an expert who kindles and keeps fire. The correspondence
between Vedic Hotr. and Avestan Zaotar is clear in language, but their
functions are different, reflecting the two meanings of hu-: make an
oblation and recite, with only the latter meaning prevailing in clas-
sical Vedic ritual. The others differ in function considerably and it is
not even clear who should be supposed to correspond to whom.
In the R. gveda, sadas refers to a place of residence or ritual assembly.
It is constructed like a house (Renou, 1939: 492494). In the classical
ritual of the Yajurveda, the Sadas is where priest sit, recite, chant
and drink Soma. Each sastra recitation by a priest in front of his
dhis. n. ya is preceded by a stotra or stuti chant from the Samaveda
sung by three Samavedins sitting further southwest in the Sadas. The
Samavedins do not have hearths; they use the place between them to
260 FRITS STAAL

Figure 1. The dhis. n. ya hearths in the Sadas.

mark the progression of their chants and rounds with the help of sticks
(vis. .tuti) on a piece of cloth (vais. .tutam
. vasanam). The sastra reciters
are not all Rigvedins: Acchavaka, Hotr. and Maitravarun. a belong to the
R. gveda: Nestr. and Potr. , to the Yajurveda; and Brahman. accham . sin, to
the Atharvaveda. Nestr. and Potr. are rather minor priests. The Acchavaka
SQUARES AND OBLONGS IN THE VEDA 261

has come in later; he has been merely called, not chosen (hvayante
na vr. n. ate: Mylius, 1982; 1986). The chief victim of discrimination is
not even there: the Adhvaryu, main representative of the Yajurveda,
center of the classical Vedic tradition (see Renou in Staal, 1999: 124,
note 1). Though he is chief actor and in charge of almost all ritual
proceedings, the Adhvaryu has no seat in the Sadas and his movements
are severely restricted. He should at no time go beyond the Sadas to
the west before the Soma offering is completed and if he does, he has
to go round the agndhrya in the north. He is apparently kept away
from the R. g- and Samavedas.
There are a few important things the Adhvaryu must do. At the
installation of the dhis. n. yas (Caland-Henry, 19061907: 104 sq.), he
consecrates each of them with a mantra that mentions two names or
qualities, for example in the case of the agndhrya, which is the first:
vibhur asi pravahanah. The expanding you are, the carrier (Taittirya
Sam. hita: 1.3.3a; Maitrayan. i mantras for the dhis. n. yas are generally the
same unless they contain unintelligible names, like Taittirya Tutha
versus Maitrayan. i Tuthra in the mantra for the Brahman. accham . sins
dhis. n. ya). The corresponding Brahman. a passages tell all kinds of stories,
as is their due. The gist of it seems to be that priests were deprived of
the Soma drink so Gods told them to assume two names so that they
might gain it with one or might not.
The word dhis. n. ya occurs in the R. gveda in the sense of wise or
benevolent and Agni is said once (4.3.6) to grow dhis. n. yasu. It could
mean on the dhis. n. yas though it may be a variant of dhis. an. asu in
the worlds according to Renou (1964: 93). The main purpose of the
dhis. n. yas is to provide fire to sastra reciters. Sastra recitations and stotra
or stuti chants are replete with interesting features. The 29 sastras of
the Agnicayana, which include a good part of the R. gveda, consist of
mantras collected from all books except book IX, which is specifically
concerned with Soma and its preparation.2 Unless all these sastras are
earlier than that book, there seems to be some apprehension here about
Soma.
Much more work needs to be done and it is likely that mysteries
remain, but if I had to formulate a general impression at this point, it is
that the dhis. n. ya altars, although they do not occur there, are a remaining
stronghold of the R. gveda which reflect its continuing internal struggles
about Soma, its attempts to subordinate Sama- and Atharvavedins
or incorporate them in its ritual, and its external struggles with the
Yajurveda.
262 FRITS STAAL

PRINCIPLES OF SUBDIVISION OF DHIS. N


. YAS AND PIRAK HEARTHS

The similarities between the dhis. n. ya hearths and the hearths or altars
that have been excavated at Pirak are striking, but what are they exactly?
I shall begin with the shapes of the former as they are described in
the Sulba Sutras. These compositions are relatively late, later than the
Srauta Sutras to which they are attached, i.e. later than about 600 BC
(Witzel, 1989: 251; 1997: 316317). That lateness, however, does not
preclude that the information on shapes contained in both is older than
the Sutras themselves. To demonstrate this in detail would be a large
undertaking.3 The gist of any such demonstration lies in the fact that
bricks are consecrated with mantras so that there must exist, at least
in principle, a one-to-one correspondence between them. In a complex
construction like Agni there are one thousand bricks, for variations
such as bricks of half-thickness or additions such as sarkara pebbles
(Staal et al., 1983: I, 482485, 493503; 1989: 193195) must be of
later date. The thousand mantras needed to consecrate these bricks
are listed in the earliest Yajurveda Sam . hita, the Maitrayan. , i.e. about
1,000 BC. That rough date is consistent with conclusions arrived by
T.N. Proferes (forthcoming) and Michael Witzel on the language of
some of the mantras and the date of Tura Kavas. eya who, according to
Satapath. a Brahman. a: 9.5.2.15 (cf. 10.6.5.9), was the first to perform
an Agnicayana.
If we wish to know precisely which mantras are assigned to which
bricks, what are their shapes and where they have to be deposited,
neither Sam . hitas, nor Brahman. as, nor Srauta Sutras, nor all of them
combined provide sufficient detail. Even the Sulba Sutras themselves are
unclear and imprecise on these matters, and certainly incomplete, unlike
the grammatical literature which developed in the same milieu but is
characterized by completeness and exhaustiveness (Staal, 2001a). Since
the Sulba Sutras are so incomplete, Thibaut, followed by Satyaprakash,
Ram Swarup and other editors, complemented his editions and trans-
lations with fragments from other Sulba Sutras and extracts from
commentaries of much later date. To do so makes sense, because the
three early Sulba Sutras, unlike the Pratisakhyas, bear a strong family
likeness and though they do not always cover the same ground, their
expressions are identical or extremely similar when they do. This has
been noted not only by Thibaut but by Datta, Sarasvati Amma, van
Gelder, Michaels, Seidenberg and others. That family likeness is in
accordance with the fact, that precise knowledge of all the relevant
details must have been available when the Agni altar was piled for the
SQUARES AND OBLONGS IN THE VEDA 263

first time: for bricks cannot be deposited imprecisely and if they were,
nothing would fit.4
The dhis. n. ya hearths are much simpler than Agni. I shall confine
myself to their descriptions in the Sulba Sutras of Baudhayana and
Apastamba which are the earliest. The two are similar to each other but
Baudhayana (in Adhyaya II) is more explicit and I shall follow him,
mentioning Apastamba (end of Pa. tala II, Khan. d. a VII) separately.5
The agndhrya comes first and its subdivision into bricks is the
simplest: agndhryam . navadha vibhajya- [Apastamba: vyavalikhya-]
ekasyah. sthane smanam upadadhyat after dividing the agndhrya
into nine parts, he should in the place of one (brick) deposite a stone.
In Baudhayana, the dhis. n. ya of the Hotr. comes next: atha hotur
dhis. n. yam
. navadha vibhajya purvam . s tribhagan ekaikam . dvedha vibhajet
then, after dividing the dhis. n. ya of the Hot. r into nine parts, he should
divide each of the three eastern portions into two parts. Baudhayana
continues: athetaran navadha vibhajya madhyamapurvau dvau bhagau
samasyet and after dividing the others into nine parts, he should
combine the middle and eastern parts. And finally: atha marjalyam .
tredha vibhajya purvaparau bhagau pancadha vibhajet after dividing
the marjalya into three parts, he should divide the eastern and western
portions (together) into five parts.
Apastamba is content to combine all this information at one sweep:
yathasam . khyam itara vyavalikhya yathayogam upadadhyat after
dividing the others according to number, he should deposite as is
fit.
Much of Bauhayanas information is unambiguous provided it is
understood: (A) that the place of the agndhryas stone is at the center,
and (B) that parts means, in principle, equal squares and that, if
larger parts are needed, they must be oblongs that are half, double and,
for one brick in the marjalya, a triplet of squares. I shall call this rather
recondite second principle (B), the Sulba Principle of Subdivision. It
is illustrated by the Nambudiri performance of 1975 which included
the following constructions with numbers indicating the order in which
bricks are consecrated (Staal et al., 1983: I, Figures 41, 4345 on pages
586, 588589).
Before we proceed I should mention that according to Baudhayana
the dhis. n. yas are caturasrah. pariman. d. ala va either square or circular.
Apastamba expresses this option in two steps: caturasra ity ekes. am
according to some they are square and: pariman. d. ala ity ekes. am
according to some they are circular. It is likely that the circular
option is a later option, not merely because it is mentioned in the
264 FRITS STAAL

Agndhrya Hearth. Dhis. n. ya Hearth of the Hotr. .

Dhis. n. ya Hearths of Acchavaka, Nest.r, Pot.r, Marjalya


Brahman. accham. sin and Maitravarun. a.

Figure 2. The three basic shapes of the dhis. n. yas and that of the Marjalya.

second place, but because it is not obvious how the subdivisions into 9,
12, 8 and 6 that the sutras prescribe could be applied to circles where
the constructions would be extremely complex, if not impossible. There
may have been an oral tradition or else the commentators would have
had to sort it out.
The similarity between the dhis. n. ya hearths and the altars excavated
by Jarrige and his colleagues at Pirak lies in the principles underlying
their subdivision into bricks. In Pirak, there are almost always two sizes:
a square of 23 23 cm and an oblong of approximately twice that
size: 45 25 cm. Some bricks seem to be bigger but their proportions
are the same. Figure 3 shows four typical structures, the drawings made
SQUARES AND OBLONGS IN THE VEDA 265

Figure 3. Pirak altars and platforms after Jarrige, Santonie and Enault.

after Jarrige, Santonie and Enault (1979: Vol. II): the first two are fire
altars (their Figures B = 7 and 24), the latter two apparently some
kind of platforms (Figures 22 and 23), all equally significant from the
perspective of geometry.
According to the simple principle that underlies these constructions,
the Pirak Principle of Subdivision, bricks must be squares or double
squares. The Sulba Principle of Subdivision implies the Pirak Principle
of Subdivision and looks like a simple extension of it. Since the Pirak
structures, which exhibit greater variation in size, date from the second
and third periods of Pirak, i.e. from between 1500 and 800 BC, they
coincide with the period of composition of most mantras: the later
R. gveda was composed between 1500 and 1200 BC and the Mantra
period of the Yajurveda begins about two centuries later (for greater
detail, see Witzel, 1997: 262264, 266 sq.).
The fire altars of Figure 3 are different from the dhis. n. ya hearths
in one respect, not marked in the illustrations: the Pirak altars have
at the center not a brick but a cavity with hardened ashes.6 Does that
detract from the similarities? On the contrary, it strengthens them. For
266 FRITS STAAL

we have seen that the agndhrya altar on the northern boundary of


the Mahavedi, the one that belongs to the Agndh kindler of the fire
and is consecrated first, has an open space in the middle which is not
consecrated with a mantra. What is the significance of that space?
At a dramatic juncture in the ritual, the agndhrya hearth links
two of the most important introductory ceremonies of the Soma ritual
(cf. Krick, 1982: 365366, note 995; and cf. 379, note 1026) to each
other. During the first, Agnipran. ayana, Agni is carried forth from the
domestic to the bird-shaped altar that has just been completed. During
this procession, the Adhvaryu sits down with his Yajamana at the
agndhrya. He puts a spotted stone (pr. snir asman) at the center
and recites two high-sounding mantras, virtually the same as R . gveda:
10.139.2 and 5.47.3. The second is:
uks. a samudro arus. ah. suparn. ah. / purvasya yonim pitur a vivesa /
madhye divo nihitah. prsnir asma / vi cakrame rajasas patyantau //

Bull, ocean, ruddy bird, he entered the birthplace of his old ancestor.
The spotted stone is placed at the center of heaven, he guards from aside the ends
of space.

During the second procession, Agn.somapran. ayana, Agni is carried


together with Soma and deposited not on the bird-shaped altar, but
on the agndhrya from where it will be distributed onto the dhis. n. yas.
This is a family affair, the Yajamana joins the procession with his wife
and children all hidden under a cloth. The priests shout: somosmakam
brahmanan. am. raja Soma is the king of us brahmins and nine days
of preparation and introductory ceremonies are over. The Soma ritual
can begin.
We have not yet paid attention to a characteristic feature of the
agndhrya that distinguishes it from the other dhis. n. yas: it is located
half inside, half outside the Mahavedi. What does that mean?
The Mahavedi is, by definition, the area where most of the ritual
proceedings take place; but a few ritually significant places are located
outside it. Most prominent to the north of it are the utkara rubbish
heap, on which ritual remnants are discarded, and the catvala pit or
hole from which earth is dug for the preparation of the new offering
altar (on the relation between utkara and catvala, see Krick, 1982:
116). Earth from the catvala is also scattered on the dhis. n. yas to unite
the ritual with its birthplace, for the catvala is the birthplace of the
ritual (yonir vai yajnasya catvalam . yajnasya sayonitvaya: Taittirya
Sam . hita: 6.3.1.1). If that means anything, it is that the agndhrya
is closer to the birthplace of the ritual than the other dhis. n. yas. And
what is the birthplace of the ritual?
SQUARES AND OBLONGS IN THE VEDA 267

The first stotra or stuti chant that is sung in all Soma rituals is the
Bahis. pavamanastotra, Outdoor Chant for the Purified Soma. It is
sung at a location called astava, stated to be outside the Sadas. Most
scholars have argued or assumed that it must be situated within the
Mahavedi, but Mylius (1997) has shown that it is more likely that
the astava is located outside and to the north of it. One of his many
arguments is that the singers, before singing the chant, regard in silence
the catvala and a clay dish with water; and when the chant is over, the
chief chanter throws that dish on the catvala, saying: I send you to
the ocean; go uninjured to your birthplace (samudram . vah. prahin. omy
aks. ita svam 7
. yonim apigacchata) . . . Here we learn a little more about
the birthplace of the ritual.
Since it neither support nor invalidates his argument, Mylius does
not mention the famous dialogue that ensues if the chanters belong to
the Jaiminya school of Samaveda. In that school, the Udgat. r or chief
chanter takes the Yajamana (who has not left the ritual enclosure since
the first day of the ceremony) to the northern border of the Mahavedi,
makes him spread his legs, put his right foot down within the Mahavedi,
and repeat:
Dont cut me off from the celestial world! (ma svargal lokad avacchaitsh. )

Then he makes him put his left foot outside the Mahavedi, and
repeat:
Nor from this world! (masmat).

I conclude that the birthplace of the ritual is this world, not the
celestial world inside the Mahavedi in which the ritual unfolds; and
the agndhrya is, unlike the other dhis. n. yas, not merely in touch with
it, but half of it.
The last piece of the puzzle is a recent addition to the BMAC
vocabulary of the Veda: the word catvala itself. It belongs to the group
of trisyllabic words we have already met with in the first section that
have a long middle syllable, are difficult to explain on the basis of
Indo-European morphology and are characteristic of the Central Asian
layer. Catvala is related to Late Avestan cat-, (dug) well, Buddhist
Sogdian ct and Bactrian o, well (Lubotsky, forthcoming).
There is much to speculate about; e.g., that samudra ocean, which
has surprising occurrences in the Veda, may refer to the Caspian Sea.
But our logical conclusions are two: (A) the spotted stone of the
agndhrya, the early hearth which marks the important ritual transitions
of Agni and Soma and from which all fires are installed on the other
dhis. n. ya hearths, is a remainder and reminder of the ancient ancestral
268 FRITS STAAL

altars with their central cavities; and (B) the agndhrya itself, so neatly
subdivided into nine squares, is half-BMAC.

A RECONSTRUCTION OF PROTO-GEOMETRY

Archeological evidence for Vedic peoples has always been elusive,


so elusive that their existence has been doubted. We may doubt it on
condition that we ignore the Vedas and their language, the two things
best known about them. Movements of language do not require invasions
or mass migrations. There werent any, as far as we now know, but there
were contacts between peoples. Speakers of Indo-Aryan may not have
formed a distinct racial group, but the R. gveda shows that they consisted
of a variety of tribes and lineages. Throughout a period of centuries,
small numbers of such people trickled in from Central Asia across
the Pamirs, Hindukush and Karakorums (Staal, 2001b). They came in
contact with sedentary people who gradually adopted their language.
That language left no traces in archeology, but their geometry did insofar
as it was reflected in altars made of durable materials such as bricks.
We are fortunate that such evidence has been unearthed at Pirak which
is not merely Bactrian but already Vedic, in fact, R. gvedic. The map on
Early Vedic Texts and Archeology in Witzel (1989: 242) does not
mention Pirak but marks, at the spot where it is located, Bhalana (=
Bolan?), a tribe mentioned in R . gveda: 7.18.7.
Pirak was connected with the BMAC and, at a much earlier period, the
Indus Civilization. Has it not been shown that there were dhis. n. ya altars
at Kalibangan, a hypothesis widely accessible through the magnificant
illustrations of Asko Parpolas 1994 book Deciphering the Indus script?
Kalibangan is an eastern Indus settlement in Rajasthan of long duration:
approximately 23001800 BC. Parpolas illustration comes from a 1985
survey by B. K. Thapar, former Director-General of the Archeological
Survey of India. It is based in turn upon the original excavation reports
by B. B. Lal who started digging in Kalibangan in 1960. When I look
at any of these pictures, I do not see seven altars, though there might
have been. That is unsettling because, in the absence of any similiarity
in shape or pattern, seven was the only argument in favor of these
altars being dhis. n. ya altars. The drawing from Caland and Henry of
the sacrificial enclosure with six circular dhis. n. yas, which Parpola puts
alongside, does not help: for circular is not what we want. Circulatory
was a later option and the original form of the dhis. n. yas was not circular
but square. That is entirely relevant when discussing structures from
before 1800 BC.
SQUARES AND OBLONGS IN THE VEDA 269

Figure 4. A Reconstruction of Proto-Geometric Evolution.

I have not at any time been an archeologist, let alone Director-


General of the Archeological Survey of India. We must therefore pay
close attention to the words of B. K. Thapar who describes the platforms
in the southern half of the Citadel: From the remains extant it could
be inferred that some of these platforms might have been used for
religious or ritual purposes, as attested by the presence of a row of
seven fire-altars aligned beside a well on the one side, and a brick-
lined pit containing bovine bones and an antler on the other (Thapar,
1985: 55).
Starting in 1962, B. B. Lal wrote more reports on these so-called fire-
altars than I have space to cite. Luckily for us, he has summarized his
findings in Some Reflections on the Structural Remains at Kalibangan
in the Sir Mortimer Wheeler Commemoration Volume (1984: 5562)
from which I quote:
Atop one of the platforms was located a series of what have been termed as fire-
altars in the absence of a more suitable name. Although partly damaged, it would
appear that their number was seven, a number which surprisingly remained sacred
in India all through the ages and is so even to-day.
The altars were in fact lime-pits, each measuring about 75 55 cm. Within each
pit were noted ash, charcoal, and the remains of a clay stele as well as of what are
known as terracotta cakes (Lal, 1984: 57).

I hope that these matters will be taken up again by archeologists.


In the mean time, I am entitled to the opinion that the Kalibangan
evidence, even without comparing it to what we find at Pirak and apart
from being much too early, should be discarded as possible evidence
for dhis. n. ya hearths.
Returning to and combining the Vedic and Pirak evidence that has
been assembled earlier, we are in a position to reconstruct a picture
of a gentle evolution (Figure 4). It shows a development that starts
with squares and gradually introduces oblongs by halving, doubling
and tripling squares. The dates are what we would expect as we have
already seen: the period of Pirak (A, FGH) is approximately 1500
800 BC and the later R. gveda and Yajurveda Mantra periods (BCDE)
fall within those limits.
270 FRITS STAAL

Figure 5. The New Garhapatya.

The striking similarity between all these structures is that there are
only three proportions between the sides of all these areas: 1 : 1, for
the square, by definition, 1 : 3 for one brick of the marjalya which
is not a hearth or dhis. n. ya; and 1 : 2 for all the other oblongs. What
would we expect to happen next? For the proportions of oblongs to
become a little more interesting in numerical terms. That is exactly
SQUARES AND OBLONGS IN THE VEDA 271

what happened. The evidence comes from the new garhapatya of the
Agnicayana which consists of 3 7 5 oblong bricks, each with sides
3 : 7 (see Figure 5). Its bricks are no longer sun-dried like the dhis. n. ya
bricks but fired in a kiln. The bird-shaped ahavanya will continue the
story by adding triangles and a few other shapes.

NOTES
1
So that the old garhapatya, the Hotr. s dhis. n. ya and the new ahavanya are in the
same line of vision (Apastamba Srauta Sutra: 11.7.10).
2
A discovery by T.P.Mahadevan when we looked at these stutis in Washington
during the Spring of 2000.
3
Keiths tables (1914: I, xlviilxvi) and his headings of Taittirya prapat. hakas
suggest that the five layers of the altar were similar from the beginning, but such
headings are not part of the Mantra collections, the texts do not seem to be entirely
unambiguous and the matter deserves to be studied afresh. The idea that each of the
five layers has to be yoked (agnim . yunajmi) is expressed or anticipated in the mantras
of Maitrayan. i Sam
. hita: 2.12.34, corresponding to Taittir ya Sam. hita: 4.7.13.12,
and explained in the brahman. as, MS: 3.4.45, corresponding to TS, 5.4.10. That
three of the five are especially significant is also expatiated there and recurs in the
ceremonies on the first, third and fifth layer referring to bhuh/bhuvah. /svah. . Terms
such as avidvan and svayamatr. n. n. a are not mentioned in the Mantras though the
pertinent ceremonies occur in all of the Kr. s. n. a, but not in the Sukla Yajurveda (Staal,
19771978).
4
The many gaps in our knowledge must have been filled by oral traditions, not
only of priests but also of taks. aka brick-makers and carpenters whose duties are
referred to in the Sulba Sutras by the causative karayet he should cause to do, as
distinguished from the Adhvaryu priest where the standard expression is kuryat he
should do (Apastamba Sulba Sutra: II, 7). These taks. aka experts still exist in Kerala
where they are called tacchan (in Malayalam). The complete information necessary
for the piling of the altar in the proper manner and with the proper mantras must be
at least equivalent to what, for the Nambudiri tradition, is provided by Staal et al.,
1983: Vol. I. I am not claiming that these descriptions, figures and tables show how
exactly the ritual was performed 3,000 years ago; but that an at least comparable
mass of equivalent and equally detailed information must have been in existence
when the altar was constructed for the first time.
5
Apastamba on the dhis. n. yas is unintelligible without Baudhayana. Other sections
suggest that it may be earlier since Ikari (1980: 381382) has shown that Baudhayana
has a few quotations from it (I owe this information to Hayashi, forthcoming).
6
Sergent (1997: 219) characterizes the Pirak altars as foyers carres avec une cavite
centrale circulaire and adds: forme attestee au -IIIe millenaire a Shahr-i Sokhta et
des le -IVe a Mundigak (sites in eastern and central Afghanistan).
7
There are variations in the Samaveda Srauta Sutras but I have followed the
Jaiminya from which the following dialogue is also taken (Jaiminya Srauta Sutra:
11:13.1714.5). Cf. Staal et al., 1983: I, 605607 with a photograph by Asko Parpola.
272 FRITS STAAL

REFERENCES

A. Vedic Sources

Apastamba Srauta Sutra


Apastamba Sulba Sutra
R
. k-Sam . hita
Jaiminya Srauta Sutra
Taittirya Sam . hita
Baudhayana Sulba Sutra
Maitrayan. i Sam . hita
Satapat. ha Brahman. a

B. General

Caillat, Colette (ed.) (1989). Dialectes dans les litteratures indo-aryennes. Paris:
Institut de Civilisatiion Indienne.
Caland, Willem and V. Henry (19061907). Lagnis. .toma. Description complete de la
forme normale du sacrifice de Soma dans le culte vedique. Vol. III. Paris: Ernest
Leroux.
Hayashi, Takao (forthcoming). Sulbasutras, in Italian National Encyclopaedia
TRECCANI. Storia della Scienza: La Scienza in India.
Henry, V. (1907). Esquisse dune liturgie Indo-Eranienne, in Caland-Henry Vol. II:
469490.
Ikari, Yasuke (1980). A Japanese Translation of the Apastamba-sulbasutra, in Michio
Yano (ed.), Indo tenmongaku sugaku shu (Collection, in Japanese Translations, of
Astronomical and Mathematical Works in India), pp. 375488. Tokyo.
Jarrige, Jean-Francois, Marielle Santonie and Jean-Francois Enault (1979). Fouilles
de Pirak. Vol. III. Paris: Diffusions de Boccard.
Keith, Arthur Berriedale (1914). The Veda of the Black Yajus School entitled Taittirya
Sanhita. Vol. III. Cambridge, Mass.: Harvard University (Harvard Oriental Series
1819).
Krick, Hertha (1982). Das Ritual der Feuergrundung. Wien: Osterreichischen Akademie
der Wissenschaften.
Lal, B. B. (1984). Some Reflections on the Structural Remains at Kalibangan, in
B. B. Lal and S. P. Gupta, (eds.), Frontiers of the Indus Civilisation. Sir Mortimer
Wheeler Commemoration Volume. New Delhi: Archeological Survey of India.
Lubotsky, Alexander (forthcoming). Indo-Iranian Substratum.
Mair, Victor H. (ed.) (1998). The Bronze Age and Early Iron Age Peoples of Eastern
Central Asia. Washington: Institute for the Study of Man and Philadelphia: The
University of Pennsylvania Museum Publications.
Meij, Dick van der (ed.) (1997). India and Beyond. Aspects of Literature,
Meaning, Ritual and Thought: Essays in Honour of Frits Staal. London/New
York/Leiden/Amsterdam: Kegan Paul International in association with the Interna-
tional Institute for Asian Studies.
Minkowski, Christopher Z. (1991). Priesthood in Ancient India. A Study of the
Maitravaruna Priest. Vienna: Sammlung De Nobili.
Mylius, Klaus (1982). Acchavaka, acchavakya. Skizze eines vedischen Opfer-
priesteramtes, in T. N. Dharmakirti (ed.), Golden Jubilee Volume, pp. 177184.
Poona.
Mylius, Klaus (1986). Acchavakya und potra ein Vergleich, Sanskrit and World
Culture: 471474. Berlin.
Mylius, Klaus (1997). Wo lag der Astava?, in van der Meij (ed.), pp. 373383.
SQUARES AND OBLONGS IN THE VEDA 273

Parpola, Asko (1994). Deciphering the Indus Script. Cambridge: Cambridge University.
Parpola, Asko (1998). Aryan Languages, Archeological Cultures, and Sinkiang:
Where Did Proto-Iranian Come into Being, and How Did It Spread, in Mair (ed.),
pp. 114147.
Potdar, K. R. (1953). Sacrifice in the R . gveda. Bombay: Bharatiya Vidya Bhavan.
Proferes, T. N. (forthcoming). The Formation of Vedic Liturgies (Harvard dissertation
of 1999).
Renou, Louis (1939). La maison vedique, Journal asiatique 231: 481504.
Renou, Louis (1964). Etudes vediques et pan. ineennes. Vol. XIII. Paris: Editions E.
de Boccard.
Renou, Louis (1967). Etudes vediques et pan. ineennes. Vol. XVI. Paris: Editions E.
de Boccard.
Seidenberg, A. (1983). The Geometry of the Vedic Rituals, in Staal et al., Vol. II:
pp. 95126.
Sergent, Bernard (1997). Genese de lInde. Paris: Payot.
Staal, Frits (19771978). The Ignorant Brahmin of the Agnicyana, Annals of the
Bhandarkar Oriental Research Institute (Diamond Jubilee Volume) 43: 337348.
Staal, Frits in collaboration, with C. V. Somayajipad and M. Itti Ravi (1983). AGNI.
The Vedic Ritual of the Fire Altar. III. Berkeley: Asian Humanities Press.
Staal, Frits (1989, 1992). Rules Without Meaning. Rituals, Mantras and the Human
Sciences. New York etc.: Peter Lang.
Staal, Frits (1999). Greek and Vedic Geometry, Journal of Indian Philosophy 27:
105127.
Staal, Frits (2001a). Vyakaran. a and Sulba in the Light of Newtons Lesson, in
Tsuchida and Wezler (eds.).
Staal, Frits (2001b). How a Psychoactive Substance Becomes a Ritual: The Case
of Soma, Social Research.
Thapar, B. K. (1985). New Traits of the Indus Civilization at Kalibangan: An
Appraisal, in: Recent Archaeological Discoveries in India. Paris: UNESCO and
Tokyo: The Center for East Asian Cultural Studies.
Tsuchida, Ryutoro and Albrecht Wezler (eds.) (2001). Festschrift Minoru Hara.
Wezler, Albrecht (1980). Zum Verstandnis von Chandogya-Upanis. ad 5.1.12, Vortrag
auf dem. Vol. XXI. Berlin: Deutschen Orientalistentag.
Witzel, Michael (1989). Tracing the Vedic Dialects in Caillat (ed.), pp. 97265.
Witzel, Michael (1997). The Development of the Vedic Canon and its Schools:
The Social and Political Milieu, in Michael Witzel (ed.), Inside the Texts/Beyond
the Texts. New Approaches to the Study of the Vedas. Cambridge Mass.: Harvard
University (Harvard Oriental Series: Opera Minora II).
Witzel, Michael (1999a). Aryan and Non-Aryan Names in Vedic India. Data for the
Linguistic Situation c. 1900500 B.C., in Johannes Bronkhorst and Madhav M.
Deshpande (eds.), Aryan and Non-Aryan in South Asia, Evidence, Interpretation
and Ideology. Cambridge Mass.: Harvard University (Harvard Oriental Series:
Opera Minora III).
Witzel, Michael (1999b). Substrate Languages in Old Indo-Aryan (Rgvedic, Middle
and Late Vedic), Electronic Journal of Vedic Studies 5/1.

S-ar putea să vă placă și