zompist bboard

Torco wrote:isn't church pretty much wholly consonantal ? [tʃɹtʃ]

Fanu wrote:Once again.

Kai_DaiGoji wrote:I''m not talking about syllabic consonants (...) functioning as the nucleus

I think the problem here is that when you start talking about words like [p] or [sxs], the concept of a syllable pretty much becomes useless. So the only way to answer the question without attacking its premises is to talk about syllabic consonants such as nasals and approximants, which (as has now been pointed out many times) are supposed to be excluded from the discussion.

Such words can certainly exist, but you may find that asking where the syllable nucleus is is like asking what colour the King of France's hair is.

Yeah, I think the way I phrased it was inherently contradictory. Sorry For what it's worth, this has been extremely helpful to me: /fp'/ is indeed a possible preposition. So, that's cool.

Kai_DaiGoji wrote:Yeah, I think the way I phrased it was inherently contradictory. Sorry For what it's worth, this has been extremely helpful to me: /fp'/ is indeed a possible preposition. So, that's cool.

Something can, of course, be a syntactic word without being a phonological word. The general term for this is "clitic". So you could have an adposition which behaves like a word as far as the syntax is concerned, but attaches to the nearest (phonological) word as far as e.g. syllabification is concerned. This, I believe, is how Russian v, s etc. work.

This leads to the question, how minimal can we make a syntactic word? Are there syntactic words which only surface as suprasegmental features? Certainly there are morphemes like that (even in natlangs, I gather), but I don't know of any clitics like that.

(Actually, Ndak Ta's copula has an essentially empty stem, but the stem by itself is not usually a valid word...)

Iirc, there are 'words' in some African languages that now only exist in the form of interactions with the downstep system.

Zhen Lin wrote:This leads to the question, how minimal can we make a syntactic word? Are there syntactic words which only surface as suprasegmental features? Certainly there are morphemes like that (even in natlangs, I gather), but I don't know of any clitics like that.

(Actually, Ndak Ta's copula has an essentially empty stem, but the stem by itself is not usually a valid word...)

According to some analyses, Tongan has a definite article which manifests as stress shift from the penultimate syllable to the final syllable. This applies to the last word of the NP no matter how many words long the NP is (making it syntactically like a word). However, it seems this analysis isn't universally accepted – some linguists seem to think it's actually reduplication of the final vowel. If I get time I may do some more research into it.

Regardless of what the best analysis of Tongan is, stress shift as a suprasegmental clitic would be fun for a conlang.

Salmoneus wrote:Iirc, there are 'words' in some African languages that now only exist in the form of interactions with the downstep system.

This is very interesting. Can you provide any more info?

In informal spoken English, at least around here, it's very common to reduce the preposition "to" into [tʰ] before consonants and [tʰɰ] before vowels.

Likewise, It's very common to reduce "it's/its", "that's", and sometimes "what's" to [ts], at least in my dialect.

Serbo-Croatian "s" ("with") is vowelless.

It's funny how people obviously didn't read my whole post...

TaylorS wrote:In informal spoken English, at least around here, it's very common to reduce the preposition "to" into [tʰ] before consonants and [tʰɰ] before vowels.

Mine's like, the other way around but [tʰɔ]

Dampantingaya wrote:

Torco wrote:isn't church pretty much wholly consonantal ? [tʃɹtʃ]

Fanu wrote:Once again.

Kai_DaiGoji wrote:I''m not talking about syllabic consonants (...) functioning as the nucleus

*pulls foot off mouth*

na'oolkili wrote:Likewise, It's very common to reduce "it's/its", "that's", and sometimes "what's" to [ts], at least in my dialect.

ts'all good!

TaylorS wrote:In informal spoken English, at least around here, it's very common to reduce the preposition "to" into [tʰ] before consonants and [tʰɰ] before vowels.

Heh - that is the opposite of the English that I am used to*, where to normally reduces to just** [ə(ː)] unless after another obstruent, where the typical fate of laminal*** /t/ in such an environment also applies, after /n/, where it typically additionally either geminates the /n/ or turns it to a nasal flap and also makes it laminal***, or initially, where it reduces to [t̻ʰə(ː)]. However, the reduction of the vowel in to is not obligatory for me, and even in informal speech I will often leave it unreduced as [u(ː)].

* I am not speaking of just my own dialect of English, even though this is what I specifically describe here, but rather I am saying that I am used to very similar phenomena in at least pretty much all the English dialects that I have had contact with between southeastern Wisconsin and upstate New York. Likewise, I have heard similar phenomena in other more General American-like varieties in the northern US, even though I am not as familiar with these varieties personally and thus cannot describe them in as much detail or with as much certainty.

** However, preceding morphemes in at least my own dialect that have vowel length variation depending on what follows them take short rather than long vowels, as if the /t/ were still there.

*** Vowel reduction for me does not mean that coronal lose their apicalness/laminalness as had been conditioned by the vowels in question; hence one can consider [ə(ː)] from historical /uː/ or /ʊ/ to still be distinguished from [ə(ː)] from other sources when such follows a coronal.

smnmnmuuc
"mute"

nuyamłamkis timantx tisyuttx ʔułtimnastx
"the father sang the song to his son"

musis tiʔimmllkītx taq̓lsxʷt̓aχ
"the boy felt that rope"


Dear god I love Nuxalk.