Old Chinese retroflexes--a few questions.

I just got through reading a book on the historical
phonology of Mandarin, and the chart of the phonetic
inventory of old Mandarin looked like this:

p, p_h\, m, f, f_h\
t, t_h\, n
ts, ts_h\, s, s_h\
tr, tr_h\, nr
ts`, ts`_h\, s`, s`_h\
ts\, ts\_h\
k, k_h\, N, x, x_h\

My questions are:

1. Those phonemes co-articulated with [h\]---are they
breathy-voiced plosives (affricates/fricatives, too),
as in the Sanskrit /bh/, /dh/, and so on, or are they
something else entirely?
2. The plosives and nasal marked as /tr/ and so
on---what exactly are they? I assumed they were
retroflexes, until I saw the retroflex series was also
present. Did old Mandarin have an opposition between
dental/alveolar, palatoalveolar, retroflex and
/r/-ified stops?

The book explained they were similar to the English
pronunciation of the clusters /tr/ and /dr/ as [tSr\]
and [dZr\], but went on to say these clusters were
_not_ biphonemic---that they were pronounced as one
discrete sound. Were these apical palatoalveolars or
something weird like that, to contrast with the
laminal palatoalveolars, or what?

3. How on earth did the Mandarin retroflex [r\] come
about? Initially, it seemed to be the product of some
phonological change of [N], but that still doesn't
explain retroflexion of vowels, like in the words
/nàr/ and /èr/.

4. Modern Mandarin has four tones, but I'm told by
someone I know that Mandarin once had a fifth tone,
the exact phonetics of which are unknown, but has been
described by old papers as sounding something like 'an
arrow hitting wood'. Did Mandarin have a set of
'broken' tones, like in Vietnamese, or something else
like that, like a phonemic glottal stop, as in the
Danish 'stød'?

Thank you.

Reply