Sharshali (no longer Carthaginian)
Posted: Tue Aug 02, 2011 12:42 am
Sharshali is born out of a revelation I've had about myself—try as I might, I can't succeed at making a continental romlang that's the least bit interesting to me. Whatever believable sound changes I can think of, given the location, have been undergone by a Romance language that already exists. So now I'm trying my hand at a North African romlang, and I'm pretending that the Southern Romance branch extended down into Africa.
Right now, I'm just looking for thoughts, comments or feelings about the premise in general, the phonology I've devised and the sound changes that led to it.
Consonants
Out-of-conworld notes: Such an extensive retroflex series may seem a little bit odd for a Romance language, but I took Sardinian /ɖɖ/ and analogized the changes to create an entire, fully phonemic series of them.
Plosives: /p b t d ts (dz) tɕ (dʑ) ʈ ɖ k ɡ/ - p b t d c z č ž ṭ ḍ k g
Nasals: /m mː n nː ɳ ɲ ɲː/ - m mm n nn ṇ ñ ññ
Fricatives: /f v β θ ð s z ʂ ʐ ɕ ʑ x ɣ ħ/ - f v ḇ ṯ ḏ s z ṣ ẓ š ž x γ ħ
Rhotics: /r/ - r
Laterals: /l ɭ ʎ/ - l ḷ ľ
Approximants: /w j/ - w y
Vowels
Monophthongs: /i ɛ a ɔ u/ (tonic), /i a u/ (atonic)
Diphthongs: not really. Maybe from new hiatuses where -d- and -g- used to separate vowels (something like saita from sagittam), but maybe I'll just monophthongize those too.
Random Notes
I'm just covering the important ones. They might be a little bit out of order; the Master List is in an ugly and unhelpful SCA format.
Late V. Latin Changes
C → 0 / _# (except for s, n)
s → z / V_V
ss → s
d, g → 0 / V_V
i, ɪ, (y, ʏ) → i
e, ɛ → e̞
o, ɔ → o̞
u, ʊ → u
i, e → j / _V
o, u → w / _V
kʷ → p
e → j / VC_#
u → 0 / VC_# (important to note that this occurs before p, t, k lenite)
Anaptyxis
sC → isC
Palatalization
mj → mɲ
nj, ŋn → ɲ
lj → ʎ
kj, kl → tʃ
#j, jj, dj, gj, gl → dʒ
tj, dj → ts, dz
t, d → ts, dz / _i
sj, skj → ʃ
sk → ʃ / _i, e
zj → ʒ
Intervocalic Plosive Chain Lenition
pp, tt, kk → p, t, k → b, d, g ( > β, ð, ɣ)
bb, dd, gg → b, d, g
Retroflex-ization
ll, rl, ld → ɭɭ
str → ʂʈ → ʂʂ
tr, tl, dr, rj→ ʈ, ʈ, ɖ, ʐ / #_
lt, tl, tr, dr, rj, rn, ln → ʈʈ, ʈʈ, ʈʈ, ɖɖ, ʐʐ, ɳɳ, ɳɳ / V_V
Nasal Assimilation Chain Shift
mp, nt, ɲtʃ, ŋk → mb, nd, ɲdʒ, ŋg → mm, nn, ɲɲ, ŋŋ (> ɲɲ)
Syncope
V → 0 / _CV[+stress] (where V1 ≠ a)
Final Devoicing
β, ð, ɣ → ɸ, θ, x~χ
r → r̥ → ʂ
l, ʎ → ɬ, ʎ̥ (maybe)
Random
rr → ʀ → ʁ → χ (in some dialects), ħ (in other dialects)
ts(j) → θ
#fr, ɸ → fx → ħ or χ
β, v → w / V[+back]_
s → z / _C[+voice]
ʂ ʐ → ʃ ʒ / _#
other retroflex → dentalveolar / _#
ʃ ʒ → ɕ ʑ
tʃ → tɕ
dz dʑ → z ʑ
θ, x, s → 0 / V[+high]_#
Hiatus Resolution
VV → V / if V=V
ai, ae, ei → e
ao, au, ou → o
iV, eV → ijjV, ejjV
uV, oV → uwwV, owwV
Random Vowel Changes
e, o → ɛ, ɔ / _[+stress]
e, o → i, u / _[-stress]
a → ɛ / C[+palatal]_[+stress]
a → ɐ / _[-stress]
Numbers (sample)
Fun fact: Sharshali mimics Semitic gender polarity in numerals 2 and up by analogizing new feminine (or rather, masculine) forms where Latin has none.
un, una
dus, duwwa
ṭes, ṭeza
paṭu, paṭa
pimbi, pimba
šeš, šeša (first š from Phoenician/Punic š-š)
seti, seta
otu, ota
noħ, nowwa
deč, deγa
~~
So, what do you guys think?
Right now, I'm just looking for thoughts, comments or feelings about the premise in general, the phonology I've devised and the sound changes that led to it.
Consonants
Out-of-conworld notes: Such an extensive retroflex series may seem a little bit odd for a Romance language, but I took Sardinian /ɖɖ/ and analogized the changes to create an entire, fully phonemic series of them.
Plosives: /p b t d ts (dz) tɕ (dʑ) ʈ ɖ k ɡ/ - p b t d c z č ž ṭ ḍ k g
Nasals: /m mː n nː ɳ ɲ ɲː/ - m mm n nn ṇ ñ ññ
Fricatives: /f v β θ ð s z ʂ ʐ ɕ ʑ x ɣ ħ/ - f v ḇ ṯ ḏ s z ṣ ẓ š ž x γ ħ
Rhotics: /r/ - r
Laterals: /l ɭ ʎ/ - l ḷ ľ
Approximants: /w j/ - w y
Vowels
Monophthongs: /i ɛ a ɔ u/ (tonic), /i a u/ (atonic)
Diphthongs: not really. Maybe from new hiatuses where -d- and -g- used to separate vowels (something like saita from sagittam), but maybe I'll just monophthongize those too.
Random Notes
- - Much as the Italian postalveolar and palatal consonants /ɲ ʃ ʎ/ are always geminated between vowels, the retroflex series is also always geminated between vowels. This is because all the retroflex consonants originate in some permutation of /l/ or /r/ plus a coronal consonant. All other consonants besides the nasals are never geminated between vowels. The length distinction on the nasals is truly phonemic; respectively, /mː nː ɲː/ come from the assimilation of earlier /mb nd ɲdʒ ŋɡ/.
- Just like in Sardinian, /k/ and /ɡ/ do not palatalize before /i e ɛ/ as in the other Romance languages. The phonemes /tʃ/ and /dʒ/ come from other sources.
- The pairs /b β/, /d ð/ and /g ɣ/ are only very marginally contrastive. The phonemes /β ð ɣ/ never occur initially, and intervocalic /b d ɡ/ comes from -bb-, -dd- and -gg-, which was rare in Latin (mostly, I'd surmise that the vast majority of instances of these come from the assimilation of the prefix ad- to the following consonant).
- I decided that it's okay to have /v/ (from V. Latin [β]) and /β/ (from V. Latin [p] between vowels) distinct because, apparently, Sardinian does it.
- Sharshali was coaxed into evolving the Semiticky/Berbery consonants /θ x ħ/ under the influence of Punic, the Berber languages and Arabic.
- I don't like retroflex consonants very much, so I'm thinking about shifting them all forward to create a dental-alveolar distinction, or just merging them completely with the dentalveolar consonants. Good/interesting idea?
I'm just covering the important ones. They might be a little bit out of order; the Master List is in an ugly and unhelpful SCA format.
Late V. Latin Changes
C → 0 / _# (except for s, n)
s → z / V_V
ss → s
d, g → 0 / V_V
i, ɪ, (y, ʏ) → i
e, ɛ → e̞
o, ɔ → o̞
u, ʊ → u
i, e → j / _V
o, u → w / _V
kʷ → p
e → j / VC_#
u → 0 / VC_# (important to note that this occurs before p, t, k lenite)
Anaptyxis
sC → isC
Palatalization
mj → mɲ
nj, ŋn → ɲ
lj → ʎ
kj, kl → tʃ
#j, jj, dj, gj, gl → dʒ
tj, dj → ts, dz
t, d → ts, dz / _i
sj, skj → ʃ
sk → ʃ / _i, e
zj → ʒ
Intervocalic Plosive Chain Lenition
pp, tt, kk → p, t, k → b, d, g ( > β, ð, ɣ)
bb, dd, gg → b, d, g
Retroflex-ization
ll, rl, ld → ɭɭ
str → ʂʈ → ʂʂ
tr, tl, dr, rj→ ʈ, ʈ, ɖ, ʐ / #_
lt, tl, tr, dr, rj, rn, ln → ʈʈ, ʈʈ, ʈʈ, ɖɖ, ʐʐ, ɳɳ, ɳɳ / V_V
Nasal Assimilation Chain Shift
mp, nt, ɲtʃ, ŋk → mb, nd, ɲdʒ, ŋg → mm, nn, ɲɲ, ŋŋ (> ɲɲ)
Syncope
V → 0 / _CV[+stress] (where V1 ≠ a)
Final Devoicing
β, ð, ɣ → ɸ, θ, x~χ
r → r̥ → ʂ
l, ʎ → ɬ, ʎ̥ (maybe)
Random
rr → ʀ → ʁ → χ (in some dialects), ħ (in other dialects)
ts(j) → θ
#fr, ɸ → fx → ħ or χ
β, v → w / V[+back]_
s → z / _C[+voice]
ʂ ʐ → ʃ ʒ / _#
other retroflex → dentalveolar / _#
ʃ ʒ → ɕ ʑ
tʃ → tɕ
dz dʑ → z ʑ
θ, x, s → 0 / V[+high]_#
Hiatus Resolution
VV → V / if V=V
ai, ae, ei → e
ao, au, ou → o
iV, eV → ijjV, ejjV
uV, oV → uwwV, owwV
Random Vowel Changes
e, o → ɛ, ɔ / _[+stress]
e, o → i, u / _[-stress]
a → ɛ / C[+palatal]_[+stress]
a → ɐ / _[-stress]
Numbers (sample)
Fun fact: Sharshali mimics Semitic gender polarity in numerals 2 and up by analogizing new feminine (or rather, masculine) forms where Latin has none.
un, una
dus, duwwa
ṭes, ṭeza
paṭu, paṭa
pimbi, pimba
šeš, šeša (first š from Phoenician/Punic š-š)
seti, seta
otu, ota
noħ, nowwa
deč, deγa
~~
So, what do you guys think?