>Thank you for clearing that up. So it would be correct to say that
>the main difference between the two ts and the two ds is the place
>of articulation and the resulting presence or absence of
>aspiration. Then De Silva is certainly confusing on that point, not
>least because the difference between 'th' in 'thumb' and 'they' is
>that one is voiced and the other is not, which is a separate issue.
>
>Would it be right to say that double consonants are pronounced
>double (eg as in Italian)? This point is not mentioned in either
>book.

Here's a run-down on the letters you asked about:

.t is retroflex, unvoiced, unaspirated
.th retroflex, unvoiced, aspirated
.d retroflex, voiced, unaspirated
.dh retroflex, voiced, aspirated
(.n is retroflex nasal)
t is dental, unvoiced, unaspirated
th is dental, unvoiced, aspirated
d is dental, voiced, unaspirated
dh is dental, voiced, aspirated
(n is dental, nasal)


A cool thing, in case you haven't noticed, is that this cyclical
sequence is used for most of the other consonants as well, in the the
Indian order of the alphabet. For the consonants from k -> m you can
set them up like a periodic table of the elements and read by row and
column point of articulation/ voicing/ aspiration etc. India had
advanced phonetics when Europeans were getting drunk on mead and
emitting laryngeal sibilants in mud huts.

Yes, to the question about double consonants being pronounced double.
(Example: _nn_ pronounced like 'pen-knife'.)

hope this helps,

/Rett