Pitch Accent

[Pabappa]

Standard Chinese, one of the best-known languages with a contrastive tone system, has been argued to have a contrast between stressed and unstressed syllables, although it's disputed. San Duanmu presents arguments for in "Tone and Non-tone Languages: An Alternative to Language Typology and Parameters", 2004, section 3.2.2.1, pages 902-906. It does seem to be the case that, as in many languages, stress in Chinese is predictable (I'm not sure how completely) from the internal structure of a word/phrase.

Thai, though more monosyllabic than English, is widely described as having iambic stress. Toneless syllables are fairly restricted, and I think not universally acknowledged.
.

And that PDF also shows that Korean has an allophonic pitch contrast based just on the initial consonant of a word. This could theoretically collapse into a true tonal contrast in some future stage of Korean .... though I think in this case, the tones might have been original, and the three-way stop contrast in modern Korean is a result of the tones causing a split in the consonant series.

And it also confirms that its OK to allow questions to have rising intonation even in a tonal language, which is something Ive always felt uncomfortable about and never really addressed. e.g. in mandarin, if a question ends on a low tone, a rising contour will appear at the end and the syllable will be lengthened. if its on a high tone, the high tone will also be lengthened. so, even in a fully tonal language, syntactic modifications of those tones appear.

[Nerulent]

I'll try get back to [you] after reading through those Hyman papers.

Those Hyman papers are really interesting! It has definitely got me thinking. I was very mistaken including Seneca as an example of (what I'm calling) a pitch-based stress accent - it's high tone is neither culminative nor obligatory. But it is also automatic, i.e. it isn't specified underlyingly, only on the surface, which makes it pretty different from most tonal systems, and gives it something in common with Nubi, Swahili, Classical Latin etc. which are also unmarked underlyingly.

The example of Northern Bizkaian Basque is also a weird one, as it seems to show that stress can be divorced from accent (in that it's non-obligatory, and non-culminative if you consider the phrase the stress-bearing domain as opposed to the word), something which I hadn't considered at all. Interestingly, the relationship between Northern Bizkaian Basque and the Tokyo Japanese/Somali system parallels the relationship between English and Nubi.

It's unfortunate that he only mentions Swedish in passing, but I still think the categories of stress-accent (English), pitch-based stress accent (Nubi, Ainu, Mohawk) and pitch-accent (Swedish, Ancient Greek) are still valid, and grouping them together makes sense to me - while all of them are culminative, obligatory (one and only one marked syllable per word) and necessarily privative* and multisyllabic, they differ in whether that marking is stress or pitch, and how many possible distinctions for that syllable there are.

All we need to really rescue this categorisation is to allow languages to be part of multiple categories - so English is not only an accent language (which excludes Northern Bizkaian Basque), but also a stress language (which includes Northern Bizkaian Basque). And Nubi is not only an accent language, but also a language with a single, privative* pitch, along with Somali and Tokyo Japanese (which differ in that the tone is culminative but not obligatory) and Arapaho (in which the tone is obligatory but not culminative) and Navajo (in which the tone is neither obligatory nor culminative).

Of course he also mentions in passing that accent might be untied to any of these concepts. This may well tie in to that paper about stress in Chinese, which I haven't gotten to yet. This could certainly complicate things.

* Privative = there are syllables unspecified for tone/accent.