. from vedic india to buddhist japan 1043
example, seem to simply identify “hearths” in various locations, with-
out indicating whether these are for cooking or for the performance
of homa. While it may be that these are indeed all simply domestic
hearths, one suspects that the question of their possible other use was
never raised.
Attention to material culture has broadened the scope of art histori-
cal studies in the recent past. It has long been suggested, for example,
that the introduction of tantric ritual to Japan by Kūkai was broadly
influential in stimulating technological developments such as casting,
weaving brocades, brush-making, and so on. The introduction of these
technologies in the service of ritual culture would have allowed them
to spread beyond their most obvious uses and thereby have a much
broader cultural impact.
Examining Continuities
When considering the issue of the continuity of tantric practice, the
first impulse of a scholar such as myself, whose training and focus
is Japanese Buddhism, is to think in terms of a single linear process
of development, matching up history and geography. The inherited
model is that things Buddhist began in India, spread to China, and
then arrived in Japan. This “three countries” model is inherited from
Japanese Buddhist historiography (Stone 1999b), which placed itself at
the end of a developmental sequence—though not necessarily progres-
sive, as Western historiography would commonly have it, but instead
frequently cast in the rhetoric of decay (mappō ). While the three-
countries model is itself problematic, most obviously for marginalizing
Korea, adding presumptions of progress or decay only compounds the
historiographic distortions.
Rather than any linear geo-historical model, which is informed
by issues of legitimacy resolved by reference to competing claims of
authoritative lineal descent (the rhetoric of lineage transmission), a
model of catalytic reactions might be more useful.^7 This model would
present the spread of tantric practice as neither having a single origi-
nary location nor a single developmental trajectory, but rather as mov-
ing in a variety of directions simultaneously and constellating into
(^7) The criterion here is utility for the generation of new understanding, i.e., a heu-
ristic criteria, being the only one relevant to such models, not some notion of truth
in the abstract.