schools whose literature has been preserved divided
the Sutrapitaka further into sections called Agama or
Nikaya. Neither term is school-specific; the notion that
the THERAVADAschool used the term Nikayawhile
other schools used Agamais justified neither by Pali
nor by Sanskrit sources.
There are either four or five Agamas and Nikayas
considered canonical by the various MAINSTREAMBUD-
DHIST SCHOOLS: the Drghagama (Pali, Dghanikaya;
Collection of Long Discourses); the Madhyamagama
(Pali, Majjhimanikaya; Collection of Discourses of
Middle Length); the Samyuktagama (Pali, Samyuttanik-
aya; Connected Discourses); the Ekottar(ik)agama
(Pali, An ̇guttaranikaya; Discourses Increasing by One);
and the Ksudrakagama (Pali, Khuddhakanikaya; Col-
lection of Small Texts). Some schools do not accept a
Ksudraka section as part of the Sutrapitaka; others
classify it as a separate pitaka. The sequence of the five
(or four) sections varies, but if included, the Ksudraka
always comes last. The names refer to the ordering
principle of each section: the Drgha (long) contains
the longest discourses; the Madhyama (middle) con-
tains those of medium-length; and the Samyukta (con-
nected) contains shorter sutras connected by their
themes. The Ekottarika (Growing by one) or An ̇gut-
tara (Increasing number of items) comprise discourses
arranged in ascending order according to numbered
sets of terms, from sutras treating one term up to those
dealing with groups of ten or more. The contents of
the Ksudraka (small texts) vary significantly from ver-
sion to version: Most of the works that seem to form
its nucleus are composed in verse and apparently be-
long to the oldest strata of the canon. Some of them,
such as the DHAMMAPADA, rank among the best known
Buddhist texts.
It is not known how many versions of the Sutra-
pitaka were once transmitted by the various schools in
India. Equally unknown is the number of languages
and dialects used for this purpose. At present, only the
Pali Suttapitaka of the Theravada school is completely
preserved. Four Agamas are available in Chinese trans-
lation: the Drgha, the Madhyama, the Samyukta, with
three translations, two of them incomplete, and the
Ekottarika. These were translated from the collections
of different schools: The Drghagama probably belongs
to the DHARMAGUPTAKA, the Madhyamagama and
Samyuktagama to the (Mula)Sarvastivadins, and the
Ekottarikagama to the MAHASAMGHIKA SCHOOL.
In the early twentieth century, numerous fragments
of Sanskrit sutra manuscripts were found in Central
Asia, enabling scholars to recover at least a small part
of the Sutrapitaka of the (Mula)Sarvastivadins. Later,
fragments of the Ekottarikagama of the same school
came to light among the Gilgit finds. Recent manu-
script finds from Afghanistan and Pakistan also con-
tain many sutra fragments from the scriptures of at
least two schools, the (Mula)Sarvastivadins and prob-
ably the Mahasamghikas. Most notable among them is
a manuscript of the Drghagama of the (Mula)Sar-
vastivadins. Unlike colophons of vinaya texts, those of
single sutras or sutra collections never mention schools,
and this often renders a definite school ascription dif-
ficult. School affiliation of Agama texts may have been
less important than modern scholars tend to believe.The different versions of the Sutrapitaka are by no
means unanimous with regard to the number and type
of sutras included in each section. To give one exam-
ple: The Dghanikaya of the Theravada school contains
thirty-four texts, while the Drghagama in Chinese
translation contains only thirty. In the incompletely
preserved Drghagama of the (Mula)Sarvastivadins,
however, forty-seven texts are so far attested. Only
twenty of them have a corresponding text in the Chi-
nese Drghagama, and only twenty-four correspond to
texts in the Pali version. For eight of them, a parallel
text is found in the Majjhimanikaya of the Pali; at least
four have no parallel at all. The agreement between the
different versions of a sutra varies significantly. Ver-
sions may be close in some passages and loose in oth-
ers. Often a considerable part of a sutra consists of
formulaic passages, and the wording of these formu-
las is version specific. Further differences may be found
in the sequence of passages, in the names of places and
persons, and also in doctrine. All this indicates a com-
mon origin, followed by a long period of separate
transmissions with independent redactional changes.There are many examples of text duplicates in two
sections of the same Sutrapitaka. For example, the Sati-
patthana-sutta(Foundation of Mindfulness) of the Pali
canon is contained in both the Dgha- and the Maj-
jhimanikaya. This may be an indication of a separate
transmission for each Agama/Nikaya in earlier times,
another indication being terms like Dghabhanaka (re-
citer of the Dgha section) to refer to the respective spe-
cialist during the phase of oral transmission in the Pali
tradition. At least in the case of the Mulasarvastivadins,
many sutras are also duplicated in their Vinaya.When growth and redactional changes of the various
collections came to an end, they began to form what can
best be described as part of a canon of the respectiveAGAMA/NIKAYA