TANTRIC BUDDHISM (INCLUDING CHINA AND JAPAN)division of the literature, therefore, and will begin by assessing what guidelines
for interpreting tantric texts have emerged thus far. First, we think it is obvious
why this literature-including the controversial Cm:ufamaharo$ana Tantra men-
tioned earlier-has caused so much difficulty for interpretation. The Anuttara
Yoga uses language in a complex way and is thereby necessarily difficult to
understand. Nevertheless, we do know that this division of Tantra instructs its
candidates in "inner yoga completely." As a clue to interpretation, this descrip-
tion opens up an internal landscape for the yogin's practice and for our recogni-
tion of where events of that practice are taking place. Failure to see this much
room for interpretation, in fact, may be behind the strong reactions of La Vallee
Poussin and others disturbed by what they read in Anuttara Yoga Tantras. We
need to be cautious, however, and not take all that occurs in this particular liter-
ature as internal process, since features of the Steps of Generation seem to
overlap with more clearly external rituals of other divisions. Then, too, we know
that the highest candidate in Anuttara Yoga Tantra receives instruction by way
of "twilight language," "evident meaning," and "standard terminology." Since
he alone proceeds to the Steps of Completion, we can expect to find these uses
of language operating wherever features of these Steps emerge, that is, wherever
the discussion turns to such practices as manipulating winds and entering light
stages. For the same reason, we can expect to find "nontwilight language,"
"hinted meaning," and "coined terminology" wherever features of the Steps
of Generation appear, that is, such practices as generating the mm:ufala and
offerings.
Some special problemsFor these expectations to bear fruit for the actual task of interpretation, we need
to know what the "six alternatives" actually are; and it would be appropriate for
us at this point to examine each of them closely. We must choose, however, to
limit the scope of close examination to the first pair of alternatives; yet it is here
that we have the most materials and here too that we can demonstrate most
clearly the character of problems of language in Buddhist Tantra. To establish
further the context for this special analysis, we should note that the alternatives
are paired according to "word," "meaning," and "both word and meaning."
While the meaning of these distinctions themselves is problematic, there are
immediate implications for translation here. Wayman is able to reject Edgerton's
translation "esoteric meaning" for sarrzdha-bhii$ii, since "his rendition attributes
a given meaning, which is the province of either neya and nlta."^17 We add that
"meaning" is the province also of the third pair based upon word and meaning.
But such considerations, while accurate for translation, should not lead us to
conclude that the first pair based upon "word" lacks meaning, since that in tum
would force the conclusion that the second pair based upon "meaning" lacks
words-which is not only linguistically absurd but untrue. Rather, the second
pair itself suggests another route to understanding. Wayman has shown that the