Gahalese scratchpad: Noun declensions and morphophonology

[Petrichor]

Hi folks--long-time lurker and occasional poster, here and on the CBB. Lost access to my old account a while back so I made a new one. Figured I'd post a little bit about my most recent project, Gahalese, which I've been doddering on for a minute.

Nouns in Gahalese are marked for case and number. There is a singular and a plural, and a nominative and oblique...but the system is a bit defective, as the oblique singular and nominative plural have completely identical affixes (-i), and for some nouns the oblique plural is identical too. There are three declensions, named by the form of the oblique plural.

The first declension, or the -ezhi declension is the most common. As you can imagine, the affix for the oblique plural is -ezhi.

sim --> simezhi
river-NOM.SG --> river-OBL.PL

The second declension, or the -e declension, is the 2nd most common.

mono --> monoe
sweater-NOM.SG --> sweater-OBL.PL

The third declension, or the -i declension, is the least common, though it frequently surfaces in commonly used inanimates, including nearly all agricultural products.

fu --> fui
olive-NOM.PL --> olive-obl.pl

Diachronically these declensions are rooted in Gahalese's ancestor, Old Guneyic. Old Guneyic marked the plural in most cases with *-i, as in the present language, and the dative (the ancestor of the oblique) with *-e. The two coincided as /e.i/ [eji] in most forms (the ancestors of the modern -ezhi declension). However, there was another declension, the ancestor of most -e declension nouns, where the plural form was underlyingly *-/j/. This surfaced as [i.] after consonants in the nominative plural and as [j] after vowels (in the other declension it formed a hiatus). The latter feature was mostly elided in the nominative plural due to analogy, but in the dative plural it was preserved as -[ej] which eventually became -[e]. The -i conjugation comes from the relics of the ancient accusative *-y which were preserved in very common words, especially but not exclusively inanimates--as in Spanish, the dative started out used for animate direct objects.

Finally, it's worth noting two common morphophonemic alternations which occur here.

The first is velar palatalization. The phonemes /k g/ surface as [ʃ ʒ] <sh zh> before front unrounded vowels (aka basically all number and case affixes.)

pagog --> pagozhi
seaweed-NOM.SG --> seaweed-OBL.SG

mek --> meshi
heir-NOM.SG --> heir.NOM.PL

kinomik --> kinomishezhi
line-NOM.SG --> line-OBL.PL

sanzhek --> sanzheshe
polar.bear-NOM.SG --> polar.bear-OBL.PL

The one exception is the 3rd conjugation, where the oblique plural and only the oblique plural does not induce palatalization (this is due to it historically being a front rounded vowel).

oleg --> olegi
cucumber-NOM.SG --> cucumber-OBL.PL

The other morphophonemic alternation of note is high vowel lowering. In some words, word-final [i. u] in the nominative plural becomes [e o] after affixes.

idu --> idoi
chant-NOM.SG --> chant-OBL.SG

This is not true of all words ending in high vowels; one just has to know.

[Pedant]

Looking good so far! I don’t suppose there’s much in the way of prefixes or suffixes to make syntax a little easier, though? Some noun classifiers, perhaps?

[Petrichor]

Looking good so far! I don’t suppose there’s much in the way of prefixes or suffixes to make syntax a little easier, though? Some noun classifiers, perhaps?

Thank you! I'm not totally sure, but what I've been thinking is that it'll be mostly reliant on syntax, plus a fair few adpositions, including possibly one indicating direct objects. As you can probably guess, Gahalese has a kind of a French or English thing where it's gone in an analytic direction but still has some remnant inflection.

[Pedant]

Seems fair! Can’t wait to see what you do with the verbs...

[Petrichor]

Sound changes from Old Guneyic, in chronological order:

Code: Select all

Low-mid shift
a --> o/_C[-alveolar]
ø --> a

Palatalization
k g --> tʃ dʒ/_V[-back][-rounded]
kj gj --> tʃ dʒ
pj bj mj--> t d n
tj dj sj zj --> tʃ dʒ ʃ ʒ

Cluster Simplification #1
C[+stop] --> 0/_C[+nasal]

Y-unrounding
y --> i

Mid-vowel raising
e o --> i u/_#

Yod-assimilation
je ja jo ju --> ji ji jɨ jɨ

Cluster Simplification #2
mb nd ŋg --> m: n: ŋ:
mp nt ŋk --> mb nd ŋg
C1[+stop]C2[+stop] --> C1[+stop]:  ([i]i.e. pt --> p:, dg --> d:[/i])
C1[+fricative]C2[+fricative] -->C1[+fricative]:
pf bv ts dz --> f: v: s: z:
f v s z ʃ ʒ --> 0/_C[+stop]
hC[+stop] -->ʔ

Offglide Loss
V --> 0/V_

J-fortition and loss
j --> 0/C_
j --> ʒ

r-lenition
r --> w

h-shift
h --> ʔ/ V_V
h --> 0/_
v --> ɦ
0 --> ʔ/#_V,V_V (this shift was didn't fully take root in the standard dialect, and only occurred in some forms).

Lateral assimilation
kl gl --> tl dl

Affricate Lenition
tʃ dʒ --> ʃ ʒ

Loss of word-final and word-initial gemination
C: --> C/_#
C: --> iC:/#_

It's worth noting that these are the changes in the High Southern dialect, which is actually spoken natively by very few people [1]. Northern Gahalese and Southern Gahalese are each a bit different, and Far Southern and Western dialects have a bunch more differences. [2]

This results in this final phonemic inventory:
/p b t d k g ʔ/
/f s z ʃ ʒ ɦ/
/m n ŋ/
/l w/

/i e a o u ɨ/

/ŋ/ is a pretty marginal phoneme, occurring only in geminates and in cases where there were historically geminates. It has a tendency to get repaired to [m] for some speakers in the latter circumstances.

[1] High Southern is an odd dialect. Historically Northern Gahalese was by far the most prestigious dialect, but after Gahal lost its independence to the Ussar invaders about 200 years ago, Northern Gahalese lost many speakers due to forced relocations into the steppe and language loss to Ussaria. The south, with less Ussaria influence, has retained Gahalese as the language of trade, commerce, and education, while in the north, although there are still many Gahalese speakers, the language is marginalized. The result of this odd situation is that the closest thing to a standard presently is High Southern, which is essentially Southerners trying to speak with an older Northern accent and grammar. There are quite a few Southernisms however; notably High Southern has preserved intervocalic geminates, which have mostly been lost in the North. To see a map, head here.
[2] Some highlights in each dialect include:
-Northern: Loss of intervocalic geminates (discussed above), using [ɨ] for word-initial geminate repair before back vowels
-Southern: Loss of syllable-final /w/, /e/ merges with /a/ in closed syllables before a sonorant, /a/ is often realized as [æ]
-Far Southern: Preservation of historic fricative-fricative clusters, /e/ merges with /a/ in closed syllables before a sonorant, /a/ is often realized as [æ], all onsetless vowels have a glottal stop inserted, the raising of /e o/ to /i u/ remains productive
-Western: Historic /r/ is retained, with /l/ becoming [w] for some speakers; preservation of all historic geminates; a vowel shift where /ɨ u o a/ become /ɯ o a æ/.

[Petrichor]

Realized I got hit up by Janko so I figured I'd add the numerals, with a little historical discussion.

The Old Guneyic numerals were:
1. saka
2. kyl
3. vaz
4. øh
5. gjø
6. berank
7. onø
8. gaza
9. ruha
10. renta
20. lyrenta
21. lyrentavisaka (all one word, but transparently derived from lyrenta vi saka "twenty and one")
30. vazrenta
40. øhrenta
50. gjørenta
60. berenta
70. onørenta
77. onørentavionø
100. migjuv

The contemporary Gahalese numerals are mostly straightforward descendants of the Guneyic numerals, with a few wrinkles. The main deviation is the loss of historic *onø, due to its resemblance to *odø "grave." It was replaced by *gjøvikyl "five and two." The word *gjø itself was lost, replaced by the word for hand (a direct calque from Ferulan and other Pengic languages, which uses the word for hand as well.) Note that *onø is only preserved in the word for 70, it has been replaced in other contexts. By contrast, historic *gjø has been retained in all other contexts. The modern numerals are:
0. lak
1. saku
2. kil
3. hoz
4. a
5. swap
6. bewang
7. zhahikil
8. goza
9. wuʔu
10. wenda
20. liwenda
21. liwendahisaku
30. hozwenda
40. awenda
50. zhawenda
60. bewenda
70. onawenda
77. onawendahizhahikil
100. mizhuh

Gahalese also has borrowed their word for zero from Walese req, which didn't exist in old Gahalese.

Ordinal Numbers are fairly simple; ordinals are marked with the prefix ka-. The only exception is the the ordinal form of one, which is kas.

Distributive numerals also exist; they're marked regularly by reduplication.