The Mahiiyiina 239
teaching of the Buddha is that everything is empty of its own
inherent existence.
But Nagarjuna was quick to point out that we should not con-
clude that emptiness itself is equivalent to the view that nothing
exists; in fact those who see in emptiness some kind of 'annihila-
tionism have a faulty view of emptiness, and 'when it is wrongly
seen, emptiness destroys the dull-witted, like a snake that is
wrongly grasped or a magical spell that is wrongly cast'.^28 It is
not that nothing exists, but that nothing exists as an individual
essence possessed of its own inherent existence. In particular, to
see 'emptiness' as undermining the teaching of the Buddha is to
fail to take proper account of the basic Abhidharma distinction
between conventional truth and ultimate truth. The point is that,
for Nagarjuna, the Abhidharma account of the world in terms
of dharmas cannot be the ultimate description of the way things
are; rather it still falls within the compass of conventional truth.
The ultimate truth about the way things are is emptiness, but
conventional truth is still truth, not conventional falsehood, and
without it the Buddha's teaching is hopeless:
The buddhas' teaching of Dharma depends equally on two truths: ordin-
ary conventional truth and. truth from the point of view of the ultimate;
tliose who do not perceive the difference between these two truths do
not perceive the deep 'reality' (tattva) in the teaching of the buddhas.
Without resorting to ordinary conventions, what is ultimate cannot be
taught; without recourse to what is ultimate, nirvaJ:.la is not attained.^29
But nirval).a is not some 'Absolute Reality' existing beyond the
phenomenal conditioned world, behind the veil of conventional
truth, for again this would commit us to eternalism. Emptiness
is the ultimate truth of reality and of nirval).a-it too is empty of
its own existence, it is not an existent. It follows that nirval)a can-
not be understood as some thing, some .existent, which is other
than the conditioned round of existence, sa111sara:
There is nothing that distinguishes sarp.sara from nirva1.1a; there is noth-
ing that distinguishes nirva1.1a from sarp.sara; and the furthest limit of
nirva1.1a is also the furthest limit of sarp.sara; not even the subtlest dif-
ference between the two is found.^30