Sounds of Sipalh
▲
0▲ 0 ▼ 0
Phonotactics, allophony, historical notes
This public article was written by [Deactivated User], and last updated on 7 Aug 2020, 02:43.
[comments] sphallophonyphonotacticsgeminationvoweldiphthongsonoritystress
3. Nouns & Adjectives
?
?
4. Sounds of Sipalh
?
?
Five of the consonants are digraphs, which means they are written with two letters. This is the same way English uses the letters 'sh' to indicate a single sound. Since the language does not use sounds that would correspond to the letters 'r' or 'h', I have chosen to use them as diacritics. It is important to remember that they are not produced as two separate sounds. Instead, the <r> indicates that the following consonant will be retroflex, and the <lh> is in the same place as the [l] but is a voiceless lateral fricative, rather than a lateral approximant. I find that interpreting the digraphs this way gets easier with practice.
All continuants (fricatives, nasals, approximants) can be geminated phonemically. This only applies word-internally, for example /saz:ip/ 'moon' vs. /sazip/ 'to plan, prepare'. The exception to that rule is that nasal geminates can be word-initial, such as /m:ub/ 'sky' vs. /mub/ 'seaweed'. Gemination is more common among voiced sounds. It is indicated in the romanization by doubling the letter (e.g. <mmub> for /m:ub/). If the phoneme is written as a digraph in the romanization, only the second letter is doubled (e.g. <karssi> for /kaʂ:i/ 'ocean').
Historical note: at one point, the language had prenasalized stops /mb nd/, but they were infrequent. They were reanalyzed as m+b and n+d, but this conflicts with the language's phonotactics. The stops thus assimilated to the nasal, and it became a geminated nasal, which could occur either word-initially or word-internally. That conflated with the normal gemination process that led to all the other geminated consonants. This means that word-initial /m:/ is not a frequent occurrence. This process resonates with the intuition of the language; it's already assimilating voicing to the highest sonority consonant in a cluster. (See the Allophony section below.)
(My ideas for this development process apparently came from this article: http://www.jstor.org/stable/4178108?seq=1)
The monophthong vowels are /i e a u/ <i e a u>
/u/ is a slippery sound. It can represent any back rounded vowel ranging from [ɔ] to [u]. Its default value is [u], but it would be understood with free variation among any sound in that range. If not in the word's first syllable, it tends to take on the vowel height of the preceding vowel. For example, /netu/ = [netu] = [neto] = [netɔ] but the most common pronunciation would be [neto].
Diphthongs are only formed with the high front vowel: /ei ai ui/ <ei ai ui>
All diphthongs are falling: the emphasis is on the first vowel. They are treated as single vowels for the sake of syllable/word construction. Syllable structure prevents VV, so there is no ambiguity in the romanization between a diphthong and a pair of individual vowels.
Diphthongs developed from V + /j/ but were reinterpreted as single sounds that act as vowels. So now they can be followed by a consonant, e.g. /tuig/, where */tujg/ would be illegal. As a consequence, /j/ never appears as a coda consonant because it would have become a diphthong. /malaj/ > /malai/
Syllables have a (C)CV(C) form.
Sonority Hierarchy
4 vowels
3 approximants & nasals
2 fricatives
1 plosives
Onset clusters must have increasing sonority, with a minimum distance of one.
| Allowed Syllables | Sonorities | Not allowed syllables | Sonorities |
|---|---|---|---|
| /na/ | 34 | */mɖib/ | 31 |
| /suik/ | 24 | */zbap/ | 21 |
| /pseg/ | 124 | */lnu/ | 33 |
| /ʈʂi/ | 124 | */gbiʂ/ | 11 |
| /ɬnud/ | 234 | */ste/ | 11 |
| /djeib/ | 134 |
/n/ assimilates to the place of articulation of the following consonant. /ʐin/ + /kub/ > /ʐiŋkub/
Consonant clusters assimilate voicing to that of the consonant with higher sonority. This applies to onsets and also across syllables. (But not across word boundaries?)
/kapzan/ > [kabzan] but /kasban/ > [kaspan]
If they have the same sonority, tend towards voicing?
There is vowel epenthesis in clusters containing an alveolar and retroflex: insert /e/
/dʐi/ > /deʐi/ and /ɖlen/ > /ɖelen/
When two identical sounds are brought into contact:
If possible it becomes a geminate: /gis/ + /su/ > /gissu/
Otherwise the sound reduces to a single stop: /gip/ + /pu/ > /gipu/
More...?
Stress is predictable: word-final, with stress on alternating syllables, moving left (iambic feet). This applies to words, not phrases, so consecutive syllables may both be stressed.
XxX X xX xXxX X xX
Certain monosyllabic (function) words never receive stress.
✎ Edit Article ✖ Delete Article
Comments
