The Defeat of Buddhism in India 167we now have, must have been kept in the monasteries—and lost
when the monasteries were destroyed. Such records must have
existed; they formed the basis for the Pali texts, for instance, which
were preserved in Sri Lanka, as well as for Tibetan, Chinese and
other translations of Buddhist literature. Taranatha, the early 17th
century Tibetan chronicler of Buddhist Indian history, refers to
manuscripts he relied on that are not known today. The disap-
pearance of Buddhism from India also has meant the disappearance
of some of the most valuable Indian historiography, and a deep
Brahmanic bias in the existing records.
It is important to realise that this constitutes a large gap in the
overall history of Indian society. This history is not simply, as
Kosambi asserts, a record of the successive changes in the means
and relations of production (Kosambi 1975: 1); it is, like all history,
a record of human actions and human relationships as they change
and develop through time. These are related to changing modes of
production but cannot be identified with these. These human
actions and relationships are precisely what we know little about.
To take only one example of how this is relevant and not simply to
those concerned about Buddhism as such, we can take the question
of gender. One of the results of the domination of Brahmanic liter-
ature is the complete ignoring of the role of queens and matrilineal
monarchies in many parts of ancient India. For example, the
Satavahanas in Maharashtra and Andhra, and their Iksvaku succes-
sors in Andhra, were matrilineal. Judging from inscriptional
evidence, Satavahana queens like Nayanika and Gotami Bala-siri,
and Iksvaku queens such as Camti-siri and Bhati-Deva were power-
ful and important people, some of whom may even have been the
active rulers at times. Mirashi’s presentation of the Satavahana
inscriptions in Devanagari transcription (especially Mirashi 1981,
II: 5–20) and Dutt’s description of the role of Iksvaku royal women
as revealed in Nagarjunakonda ruins (1988: 128–31), as well as
Mahayana texts such as the Srimalasimhanada Sutrashow this.
However, such women are not mentioned at all in the Puranic
sources and so have been left out of the normal historians’ accounts
of India (e.g. Thapar 1996; Sastri 1999). Mirashi ignores his own
evidence to assert it was impossible that women should have had so
much power (Mirashi 1981, II: 4–16, 34n, 41–49).
This situation itself puts a large question mark on all the gener-
alisations we have seen about the decline of Buddhism in India.‘barbarian’ according to Brahmanic ideology. Aside from these
Brahmanic (and other orthodox social and philosophical) texts and
also the material provided from outside, Indian history up until the
time of the Turks and Mughals has to be reconstructed largely
from evidence of inscriptions, coins, and archaeological excava-
tions. It is a strange situation for a society whose elite prides itself
on its literary and intellectual skills.
A more genuine historical tradition requires both a sense of scep-
ticism and loyalty to empirical reality, and an organisational or
institutional independence from the rulers themselves. China had
one of the most developed historical traditions; however biased,
these recorded the action of rulers and sometimes their subjects
during the dynasties themselves. Though ‘history’ was not an early
‘science’ in Buddhism as it was in China, still Buddhism did foster
a historical approach: the early Pali texts show an orientation to a
fairly sober description of historical events, and the monasteries
were institutions of potentially great autonomy. In fact, Sri Lanka,
a Buddhist society, did produce two important historical chronicles
which add to our knowledge of early India. It is quite striking that
only the Buddhist chronicles of Sri Lanka give an account of India’s
greatest emperor, Asoka; the Brahmanic sources barely mention his
name. The single example of a similar chronicle in India is the
Rajatarangini of Kashmir, written in the 12th century under king
Jayasimha, which may have been influenced by Buddhism (Thapar
1979: 243–44; Kosambi 1985: 116n).
Thapar’s defends the use of Brahmanical sources with the argu-
ment that ‘every society has a concept of its past and therefore no
society can be called ahistorical’ (1979: 238). This does not come
to grips with the fact that what is preserved in India has been the
Brahmanical ‘concept of its past’ and not that of a nebulous ‘society’.
The concept of the past as revealed in Brahmanical sources is
quite different from the one revealed in Buddhist sources, not to
mention those of other Indian traditions or preliterate indigenous
groups within India. Moreover, the very different conceptions of
the past embodied in Buddhist literature have been erased in India
itself.
We can well believe that quite a lot of literature, including chroni-
cles, must have been available once. But mostly Buddhist records,
whether chronicles, popular works, sacred texts, vernacular ver-
sions of the Chinese and Tibetan translations from Sanskrit, that
166 Buddhism in India