zompist bboard

a congress of convoluted conworldery
It is currently Fri Apr 20, 2018 7:20 pm

All times are UTC - 6 hours [ DST ]




Post new topic Reply to topic  [ 12 posts ] 
Author Message
PostPosted: Sat Apr 14, 2018 3:58 am 
Smeric
Smeric
User avatar

Joined: Wed Oct 09, 2002 3:17 am
Posts: 1283
Location: the darkest depths of the bone-chilling night
we talk a lot about syllable constraints, but is it possible/realistic to have constraints on whole words as well?

specifically, my current lang has a general syllable structure of (C)V, but words can end with a nasal, and/or begin with an s- cluster: paka, spaka, pakan, spakan are all legal, while paska, panka are not. is it realistic to have different rules for whole words than for basic syllables? Is This Found In Nature??


Top
 Profile  
 
PostPosted: Sat Apr 14, 2018 5:09 am 
Sumerul
Sumerul
User avatar

Joined: Thu Apr 25, 2013 4:48 am
Posts: 3064
Location: Brittania
Yes. The only example I can think of off the top of my head in the 2 minutes it took me to type this is Ancient Greek, where syllables can end in more or less anything, but words cannot end in stops.


Top
 Profile  
 
PostPosted: Sat Apr 14, 2018 7:57 am 
Sumerul
Sumerul
User avatar

Joined: Wed Mar 08, 2006 5:00 pm
Posts: 4317
Location: Braunschweig, Germany
In Greek, the only possible word-final consonants are /n/, /r/ and /s/. But as Kath has observed, non-final syllables may end in just about anything.

_________________
...brought to you by the Weeping Elf
Tha cvastam émi cvastam santham amal phelsa. -- Friedrich Schiller
ESTAR-3SG:P human-OBJ only human-OBJ true-OBJ REL-LOC play-3SG:A


Top
 Profile  
 
PostPosted: Sun Apr 15, 2018 1:53 am 
Smeric
Smeric
User avatar

Joined: Wed Oct 09, 2002 3:17 am
Posts: 1283
Location: the darkest depths of the bone-chilling night
n0ice, thanks


Top
 Profile  
 
PostPosted: Sun Apr 15, 2018 1:56 am 
Smeric
Smeric
User avatar

Joined: Wed Oct 09, 2002 3:17 am
Posts: 1283
Location: the darkest depths of the bone-chilling night
btw deciding to allow sC onsets led me eventually to realize that this could (and would) in fact be my third successive conlang to include the word skūra as the word for "shark"


Top
 Profile  
 
PostPosted: Sun Apr 15, 2018 2:14 am 
Lebom
Lebom
User avatar

Joined: Mon Sep 10, 2012 2:28 pm
Posts: 229
Location: Vancouver, BC
It's not only possible, it's very common. There are lots of languages that behave this way. Check out Ponapean and Diola-Fogny, which have CC clusters only at the right edge and pretty severe restrictions on word-internal clusters. The way it works is that the edgemost segment is not parsed into a syllable, so e.g. Ponapean sitamp "stamp" is parsed as [si.tam.<p>] with <> indicating extrasyllabicity. It would be pretty easy to imagine your language parsing spakan as [<s>.pa.ka.<n>], with all the pesky extra consonants left unparsed to preserve the nice unmarked CV syllable shape.

According to one of my profs Douglas Pulleyblank, this is actually really common. Apparently English "strengths" is best analysed as [<s>.trɛŋ.<θ.s>] (ever wonder why English only allows coronals in those positions in those clusters? It's because only coronals can be extrasyllabic in English). This is also the favoured analysis for the insane sequences of consonants present in Nuxalk and Berber. One neat bit of morphophonology from Nuxalk is that reduplication targets the first CV sequence in a word, i.e. the first syllable, so the reduplicated form of C₁C₂C₃V ( = [<C₁.C₂>.C₃V]) would be C₃VC₁C₂C₃V. You could apply this to your language and make the reduplicated form of spakan be pa-spakan.

Edit: brackets.


Last edited by Buran on Sun Apr 15, 2018 2:38 pm, edited 1 time in total.

Top
 Profile  
 
PostPosted: Sun Apr 15, 2018 3:54 am 
Smeric
Smeric
User avatar

Joined: Wed Oct 09, 2002 3:17 am
Posts: 1283
Location: the darkest depths of the bone-chilling night
Buran wrote:
It's not only possible, it's very common. There are lots of languages that behave this way. Check out Ponapean and Diola-Fogny, which have CC clusters only at the right edge and pretty severe restrictions on word-internal clusters. The way it works is that the edgemost segment is not parsed into a syllable, so e.g. Ponapean sitamp "stamp" is parsed as [si.tam.<p>] with <> indicating extrasyllabicity. It would be pretty easy to imagine your language parsing spakan as [<s>.pa.ka.<n>], with all the pesky extra consonants left unparsed to preserve the nice unmarked CV syllable shape.

According to one of my profs Douglas Pulleyblank, this is actually really common. Apparently English "strengths" is best analysed as [<s>.trɛŋ.<θ.s>] (ever wonder why English only allows coronals in those positions in those clusters? It's because only coronals can be extrasyllabic in English). This is also the favoured analysis for the insane sequences of consonants present in Nuxalk and Berber. One neat bit of morphophonology from Nuxalk is that reduplication targets the first CV sequence in a word, i.e. the first syllable, so the reduplicated form of C₁C₂C₃V ( = [<C₁.C₂>.C₃V] would be C₃VC₁C₂C₃V. You could apply this to your language and make the reduplicated form of spakan be pa-spakan.


!!!


Top
 Profile  
 
PostPosted: Sun Apr 15, 2018 9:14 am 
Avisaru
Avisaru
User avatar

Joined: Fri Mar 16, 2012 5:34 pm
Posts: 685
Sanskrit can have /s/ in coda; however, such coda cannot surface word finally -- see visarga and the varius sandhi effects underlying /-s/ is involved in: mostly deletion or modification.

Additionally, Buran's notes above reminded me, extra-syllabicity has also been used to explain Sanskrit onsets that violate the sonority hierarchy.

Sanskrit allows onset clusters of two obstruent: stop + fricative and fricative + stop. The treatment of fricative + stop sequences in morphological operations, such as reduplication, is very different from the treatment of stop + fricative or fricative + resonant onsets. The following is mostly drawn from Kobayashi Onishi’s “Historical Phonology of Old Indo-Aryan Consonants” and to a much lesser degree some of my own notes.


CS- clusters reduplicate via CV-

Root: /tsar-/ “sneak” form a perfect: /ta~tsá:r-a/

Root: /kʂi-/ < /*dʱgʷʱ(e)i-/ “destroy” forms a desiderative via, in part, reduplication: /cí~kʂi-:ʂati/


Similarly, SR sequences reduplicate via SV-

Root /sna:/ “bathe” gives perfective /sa.~sn-ur/

Root /syand/ “gallop” gives perfective /si.-ʂya.d-ur/


BUT SC- clusters reduplicate via CV-

Root /stʰa:/ < /*steh2/ “stand” forms a present in /tí~ʂʈʰa-ti/; so, /tí.~<ʂ>ʈʰa.-ti/. C.f. Avestan which has /hi-ʃtənti/.

Root /sperdʰ/ “contend” reduplicates as /pa.~spr̩.dʱé/; so, /pa.~<s>pr̩.dʱé/

Root /cʰa(n)d-/ < /*sḱe(n)d/ “appear, seem, please” forms a perfect: <ca~ccʰand-a>. The pronunciation of <ccʰ> is a point of discussion, but it seems that this latter from was pronounced [tɕa~ɕtɕʰand-a]. Sanskrit regularly deletes the first in a series of three obstruents. So, [tɕa.~<ɕ>tɕʰan.d-a ].

Similarly, root /cʰi(n)d-/ < /*sKei(n)-d-/ “cut off, split” forms a desiderative in /cí.~ccʰit.-sa.ti/ [tɕi.~<ɕ>tɕʰit.-sa.-ti ] Again, c.f. Avestan <a.uua.-hi.~siδ.iia:t̚> “off+cut, split.”

It seems Sanskrit's limitation on /s/ surfacing in coda and different treatment in onset may be related. As a point of comparison, in addition to Avestan's different reduplicaton facts, it also allows -S and -{x, f, r}S codas.

_________________
linguoboy wrote:
So that's what it looks like when the master satirist is moistened by his own moutarde.


Top
 Profile  
 
PostPosted: Sun Apr 15, 2018 11:27 am 
Lebom
Lebom

Joined: Fri May 06, 2011 9:18 pm
Posts: 109
One of my conlangs, Séoru, allows sC onsets, which come from a prefix, *həs-, in the proto-lang, which reduced to simply s-. I think I will also have a reduplicative element, altho I haven't worked out the details. But I'm pretty sure that sCV will reduplicate to sCVCV, rather than CVsCV or sVsCV or sCVsCV. Altho, later formations, when some sC combinations have become single consonants (like /sk/ → [ʃ] before front vowels), might have ʃVʃV.


Top
 Profile  
 
PostPosted: Sun Apr 15, 2018 1:08 pm 
Osän
Osän
User avatar

Joined: Sun Feb 16, 2003 2:57 pm
Posts: 11371
Location: Lake Tašpa
Now that everybody has said yes, I say Im not sure. What is the stress pattern of this language? If the language has final-syllable stress, I can see that it might make sense to allow codas in word-final position but not elsewhere. Arabic does something similar ... its structure is CVCC. But if you dont have final syllable stress I dont think Arabic is a good example.

I dont think Gk is a good example because it has the opposite pattern to what you want: more restirctions at end of word than internally.

Nuxalk has extrasyllabic consonants but, as far as I know, they are not bound to the beginnings and ends of words but can occur in the middle as wel, so it isnt a good example from which to derive a CVCV(C) syllable structure either.

Pohnpei can have prenasals as the beginning of a word, so they may simply be seen as single consonants there and therefore the word ''sitamp'' does not necessarily end in a cluster. Can Pohnpei words end in a cluster that is not permissible as an onset ?

_________________
Sunàqʷa the Sea Lamprey says:
Image


Top
 Profile  
 
PostPosted: Sun Apr 15, 2018 4:02 pm 
Lebom
Lebom
User avatar

Joined: Mon Sep 10, 2012 2:28 pm
Posts: 229
Location: Vancouver, BC
Soap wrote:
Nuxalk has extrasyllabic consonants but, as far as I know, they are not bound to the beginnings and ends of words but can occur in the middle as wel, so it isnt a good example from which to derive a CVCV(C) syllable structure either.

The bit about Nuxalk was basically an anecdote from a discussion with Dr. Pulleyblank and was not intended to be rigorous. It was more about the neat morphological process than syllabification in Nuxalk.

Soap wrote:
Pohnpei can have prenasals as the beginning of a word, so they may simply be seen as single consonants there and therefore the word ''sitamp'' does not necessarily end in a cluster. Can Pohnpei words end in a cluster that is not permissible as an onset ?

Yes they can. We get [ll] in [kull] "roach" which is not permissible as an onset, and (C)VVC syllables as in [maasaas] "cleared". So we can't dismiss word-final clusters as really being single consonants, and we have to expand the generalisation about (C)VCC syllables being permitted only word-finally to say that (C)VXC is permitted only word-finally. We could say that [maasaas] should be syllabified as [maa.saas], but that begs the question: why don't we find CVVC syllables elsewhere? We could impose a constraint saying that (C)VXC syllables are only allowed word-finally, but that predicts that there will be languages in which (C)VXC syllables are only allowed word-initially or word-internally, patterns which are not attested. Or we could invoke extrasyllabicity, which is widely attested and resolves the problem neatly.

As for analysing initial NC as a single prenasalised segment, it's actually a lot messier than it looks. There are several points against such an analysis:
  • It requires adding more segments to the language's inventory
  • Word initial NC can optionally be realised as [iNC] or [uNC] e.g. [(i)nta] "to say", [(u)ŋkɔl] "to make a sennit", [(i)nsen] "will", suggesting that initial NC is underlyingly /VNC/ (c.f. English [(ə)n.tɪl] "until")
  • If there are segments such as [nt] in Ponapean and they are attested in V_# position, we should also expect to find them in VV_# position, since VVC# is permitted, but such sequences are unattested
  • It requires explaining why prenasalised consonants are attested in exactly three environments: #_V, V_V, and V_# (excludes VV_#), i.e. even more restrictions on syllable structure

Analysing initial NC as /VNC/ is a lot simpler and has better predictive power. It posits no additional consonants, requires only one process instead of many new constraints on syllable structure, and explains the optional vowels. It's also attested in at least one other language. So in conclusion, I think my original points about Ponapean being a good example for GreenBowTie to draw from stand.


Top
 Profile  
 
PostPosted: Mon Apr 16, 2018 4:19 pm 
Sanno
Sanno
User avatar

Joined: Thu Jan 15, 2004 5:00 pm
Posts: 16729
Location: One of the dark places of the world
With respect, that rather looks like special pleading!

In any case, I don't see why Pohnpeian is relevent here. Yes, it allows final nasal+stop clusters. But it also allows medial and initial nasal+stop clusters. And yes, it allows final geminate sonorants. But it also allows medial and initial geminate sonorants. Given that these therefore pattern in distribution with other phonemes and not with any other sort of cluster, your analysis seems strange and unnecessary.

Pohnpeian is almost an example for other reasons - because it allows word-final stop codas that are impossible medially. However, as these codas are also substituted across word boundaries, it's probably better to see this rule as operating at the surface level, rather than at the morphemic level.

However, Austronesian is certainly a place to look for examples. Proto-Austronesian words were almost all CVCVC, where the second C could be a prenasalised stop - but reduplication could create CVC-CVC words. I'm sure there must be Austronesian languages in which those cross-morpheme clusters are simplified. Indeed, apparently the retention of these clusters is rare outside of Taiwan and the Philippines.
Kelabit is almost an example - it allows word-final codas, but no intra-morphemic clusters; however, it does allow a few clusters between at morpheme boundaries due to affixes.

_________________
Blog: http://vacuouswastrel.wordpress.com/

But the river tripped on her by and by, lapping
as though her heart was brook: Why, why, why! Weh, O weh
I'se so silly to be flowing but I no canna stay!


Top
 Profile  
 
Display posts from previous:  Sort by  
Post new topic Reply to topic  [ 12 posts ] 

All times are UTC - 6 hours [ DST ]


Who is online

Users browsing this forum: No registered users and 2 guests


You cannot post new topics in this forum
You cannot reply to topics in this forum
You cannot edit your posts in this forum
You cannot delete your posts in this forum
You cannot post attachments in this forum

Search for:
Jump to:  
Powered by phpBB® Forum Software © phpBB Group