(WORD OF THEBUDDHA). This is indicated by their
opening with the phrase evam maya s ́rutam (“Thus
have I heard”), indicating that each text is understood
to have been recited by the Buddha’s disciple ANANDA
at the First Council. Modern scholarship situates these
texts as new if anonymous compositions, the chronol-
ogy of which tracks the evolution of Mahayana and
Vajrayana, respectively. The inclusiveness of later In-
dian Buddhism regarding canonicity also means that
it is difficult to know the precise total extent of the lit-
erature. The Pali canon by tradition has been fixed
since the first century B.C.E. and the exact content is
well known, as revealed in the fifth-century C.E. com-
mentaries attributed to BUDDHAGHOSA and others.
There is no comparable clarity for the Mahayana or
Vajrayana, and even now there exists no comprehen-
sive catalogue of works for either tradition. The near-
est we have are the ancient CATALOGUES OF SCRIPTURES
of the Chinese and Tibetan translated canons, none of
which are exhaustive. This situation makes it difficult
to write with conclusive authority on many aspects of
this literature as a whole.
This situation is further complicated in that the ma-
jor portion of canonical Buddhist literature in Sanskrit
has been lost since the time of Muslim depredations
in northern India (eleventh through twelfth centuries
C.E.) and is now known only through ancient transla-
tions made into Tibetan, Chinese, and other languages.
The exceptions to this have come from two sources:
archaeological or antiquarian recovery of ancient man-
uscripts or their active preservation through copying
in Nepal. Notable among the former are numerous
manuscripts recovered from the oases of Central Asia,
the small library of about fifty texts found in Gilgit in
the 1930s, Rahula San ̇krtyayana’s photographs made
in the 1930s in Tibet of very early Sanskrit manuscripts
originally transported there in the medieval period to
assist translation work, and the recovery in the 1990s
of very early manuscripts from Afghanistan, such as
those in the collection of Martin Schøyen in Oslo
(Braarvig). Typical of the latter category from Nepal
are numerous manuscripts of nine canonical texts
called the navadharma(the nine teachings), along with
a wide range of tantric ritual texts. The bulk of Bud-
dhist Sanskrit literature known today has been pre-
served in Nepal (Mitra).
Agama collections.The agamacollections are the
functional equivalents of the nikayasof the Pali canon—
thus there were long (dlrgha), middling (madhyama),
thematic (samyukta), incremental (ekottara), and mis-
cellaneous (ksudraka) collections in Sanskrit. The
agamacollections contain Sanskrit versions of many
of the texts found in the Pali collections, and are un-
derstood to have been the s ́ravaka canon as utilized on
the Indian subcontinent by s ́ravaka lineages other than
that of the THERAVADAschool. Overall the agamascon-
tained a larger number of texts than the nikayasand
arranged them in a different sequence. Unlike other
Buddhist literature in Sanskrit that has no s ́ravaka par-
allels, this category offers enormous potential for com-
parative study to differentiate the ideas and concerns
of the s ́ravaka schools. Regrettably, the agamasdo not
survive in their entirety and are largely known through
translations of them made into Chinese (Lamotte, pp.
153 f.). Until recently the only exceptions to this were
individual sutras—for example, the MAHAPARINIR-
VANA-SUTRA(Waldschmidt) and fragments recovered
from long abandoned Buddhist sites in Central Asia—
but this has changed with the discovery in Afghanistan
in the late 1990s of an almost complete manuscript of
the Dlrghagama,probably belonging to the Mulasar-
vastivada school (Hartmann).Vinaya and abhidharma. Although there were
seven canonical ABHIDHARMAtexts in Sanskrit belong-
ing to the Sarvastivada school, these are now lost in
their original language. The Sanskrit VINAYAcollec-
tions have fared better, and two works in particular
warrant mention. The first of these is the MULASAR-
VASTIVADA-VINAYA, which has mostly survived in a sin-
gle manuscript discovered at Gilgit. This massive text
is a compilation of narratives and case law offering nu-
merous insights into the preoccupations and realia of
monastic life in medieval India (Panglung). With this
we can compare the Mahavastu,a wonderful collec-
tion of narratives and lore built around a biography of
the Buddha that describes itself as belonging to the
vinaya of the Lokottaravada branch of the MA-
HASAMGHIKA SCHOOL(Jones). This too contains inter-
esting and important parallels to material found in the
Pali canon.Mahayana.Mahayana sutras form a diverse body of
literature produced between the first century B.C.E. and
the fifth century C.E. The earliest examples are thought
to be the perfection of wisdom texts, Astasahasrika-
prajñaparamitaand Ratnagunasamcaya-gatha,in prose
and verse, respectively. These expound a critique of the
abhidharmaand the teaching of the real existence of
dharmas and promote the BODHISATTVAas the ideal
Buddhist. While many Mahayana sutras are now only
known in Tibetan and Chinese translations, we are wellSANSKRIT, BUDDHISTLITERATURE IN