----- Original Message -----
From: hfroelandshagen
To: phoNet@yahoogroups.com
Sent: Friday, May 24, 2002 11:24 PM
Subject: [phoNet] Pitch Accent

> I have three questions:
> May stressed syllables be indicated by low tune instead of high, with unstressed syllables having high tune and low tune respectively?
Stress is not necessarily correlated with tone, especially in languages that employ fundamental frequency (F0) to mark lexical tone (in such cases the location of stress is marked by duration an intensity). _If_ there is a correlation, however, raised F0 is virtually always a cue for prosodic strength, since it reflects increased vocal effort, other things being equal. Conversely, lower F0 (or even a super-low tone resulting from creaky-voiced phonation) marks lower prominence and is a common boundary signal (a kind of phonetic full stop).
Note, however, that "other things" are not always equal. In particular, _primary_ stress is typically marked with a "nuclear" intonation realised as a _contour_ tone rather than merely a high tone. The beginning ("head") of the contour is linked to the stressed syllable, but its "tail" is aligned with the rest of the intonational phrase. If the contour is [LH] (as in yes/no questions in English) the stressed syllable carries a low (or low rising) tone and F0 rises over the following syllable(s). Compare "Seven." [H.L] with "Seven?" [L.H] ([.] marks syllable division).
What is a special-purpose "general question" intonation in most accents of English (and often elsewhere) may be the default intonation of plain declarative sentences in some languages: familiar examples that spring to mind include the Scouse-type accents of northern British English, and Swiss German. Welsh accents of English (as well as the Welsh language) realise primary stress as a complex rising-falling intonation [LHL]. In words with penultimate stress this becomes [L.HL], where [HL] is aligned with the unstressed final syllable, which may therefore sound more prominent than the stressed penult to a non-Welsh ear!
I suspect that similar phenomena (the association of primary stress with tonal movement rather than height) account for the occasionally reported implemantion of stress as low rather than high pitch. 

> Does contrast in pitch in pitch accent languages necessarily be of the type HL vs LH? Or may H vs LH/HL be an equally possible contrast opposition? (H-high,L-low)
I'd say that both types are perfectly possible. For example, the contrast between "acute" and "circumflex" in Ancient Greek seems to have been realised as [H] versus [HL], the latter being the "marked" term of the opposition, restricted to long (bimoraic) nuclei.

> Are there any known instances of qualitative changes in vowel inventory, in which pitch accent is a decisive factor?

> Such as
> [á] > [ó] but [â] > [â]
As far as I know, pitch-accent contrasts are accompanied by clear qualitative differences in Modern Lithuanian (if Sergei Tarasovas is listening to us, he can give you some first-hand information about that). Divergent historical development is sometimes visible in Slavic. For example, Proto-Slavic "acuted" *óRC- (where R = *r or *l) > Polish RaC- in word-initial position, while "circumflexed" *ôRC- > RoC-.