Dear Jim,

> It just so happens that I have M. Hattori's Dignaaga on Perception,
> which I have not yet studied even thought it's been sitting on my
> bookshelf since 1980.

Yes, I know the feeling. Still it's good to get books sometimes even if you
read them years later whjern the need arises. Still, Dignaaga is very heavy
going (but not as bad as Dharmakirti) and quite different in style to
Pali-based literature. I mentioned it in my short list because it gives a
good overview from Dignaaga's Sautrantika viewpoint of the perceptual
theories current in his time.

> The 'perception' here is a translation of Skt. 'pratyak.sa'. When you
> refer to theories of perception, do you take 'perception' as
> 'pratyak.sa' or 'sa.mj~naa'?

This is the problem with translating -- you use one English word for
something that it fits and then find you need it for something else :)
However, I don't think I was intentionally using "perception" specifically
for any Indic term. I normally translation pratyak.sa as "direct
perception" and "ideation" for sa.mj~naa, though in general terms one might
say that sa.mj~naa also involves some elements of the perceptual process,
depending on what model one adopts. I think of its function as a kind of
link between perception and conception -- percepts go in and concepts come
out. Again, as I understand the role of sa.mj~aa, it does not directly
cognize vastus (which vij~naana does) but only the aakaaras (sensa) modified
by vedanaa.

In other words, visual vij~naana cognizes a red blob (the vastu "out there")
which is them present in vij~naana as an aakaara, which is also "coloured"
by a vedanaa reaction -- so it is no longer reallty a pure sensum.
Sa.mj~naa then extracts / constructs (often called nimitta-udgraha) the
salient feature -- the nimitta -- from that aakaara (sensum) and labels it
as "red" (= inference). Thus, my contention that nimittas cannot be
external to personal cognitive processes is because sa.mjn~aa does not
directly access the vastu itself, the existence of which can only be
inferred by us according to many Indian schools of Buddhism. All this seems
perfectly coherent and reasonable to me but others claiming greater powers
of understanding may disagree.

Your observation about the verbal root of nimitta and its relationship to
anumaana is well-made. If a nimitta is not directly perceived or cognized as
it would seem, then it must involve anumaana. This again is suggested by
the ambivalence of the term "sa.mjn~aa" -- which can indicate the process
and the product. As a product, sa.mj~naa can be translated by "concept" and
"designation / name", which belong to the realm of anumaana. [Just in case
a certain somebody starts to accuse me again of being a speculative
ideologue, the above is quite commonplace in any Indian Buddhist text
dealing with perception and conception -- I just don't have the time to dig
out all the references from unindexed works that may run to hundreds of
pages.]

However, you should note that often later Buddhist scholars in India,
particularly the praamaanikas, often do not use traditional Buddhist sets of
terminology such as sa.mj~naa as one of the five skandhas, because they were
engaged in debate with non-Buddhist who did not necessarily accept terms and
concepts specific to Buddhism. In other words, they were developing and
using a pan-Indian philosophical terminology which occasionally is difficult
to correlate to specifically Buddhist terminology and concepts. Another
problem is that alternative explanatory systems are used that are
semi-independent of each other and are thus difficult for us to correlate.

Best wishes,
Stephen Hodge