Classical Saarvian Scratchpad

[KathTheDragon]

This is a language I've had simmering for a while now, where the general thing I wanted to try was consonant gradation. I'm finally pleased to say I've figured out a system that looks like it gives paradigms I'm happy with. Have some pronominal paradigms.

[KathTheDragon]

And some noun paradigms.

[Kuchigakatai]

I'm guessing the ERG/LOC/LAT ending was preceded by a vowel that was syncope'd away, while the ABS/ACC/DIR plural show an older assimilation/dissimilation process? With the ABS plural having vowel epenthesis?

Like, *miʃt *miʃtu *miʃtum > *mist *mistu *mistum > miset mistu mistum,
but *miʒətiil *miʒətii *miʒətis > miʒtiil miʒii miʒtis;
*saθt *saθtu *saθtum > *sast *sastu *sastum > säset sästu sästum,
but *saðətiil *saðətii *saðətis > säðtiil säðtii säðtis.

But maybe I'm terribly wrong, and I just proved internal reconstruction based on a sample of n=10 should never be trusted. not that it needed any proving

[KathTheDragon]

But maybe I'm terribly wrong, and I just proved internal reconstruction based on a sample of n=10 should never be trusted. not that it needed any proving

I'm afraid this is so. Originally all the stems were at least disyllabic, and the final vowel has been absorbed into the endings, where it survives at all.

[KathTheDragon]

The phonology isn't terribly interesting, but that's ok. The main divergences from the IPA are /b d g/ which are [β ð ɣ], except when immediately before a voiceless stop, in which case they devoice to [ɸ θ x]. The [ɸ] produced by this merges with original /f/ [ɸ].

[Vardelm]

Love it.

I'd be interested to see the rules/sound changes that drive these forms since this is a topic I still struggle with.

[KathTheDragon]

I have half a mind to at some point derive some sisterlangs to this one and call it a reconstruction challenge.

[KathTheDragon]

More on nominal morphology.

Nouns can be divided into (I think) 7 declension classes: 6 classes of consonant stem, and two (possibly irregular, I haven't decided how much I want to regularise them) classes of vowel stem. The classes are largely unrelated to the form of the stem.



The absolutive, accusative, and ergative cases are used for the core arguments of the verb in a way that depends on the animacy of the arguments. An intransitive verb's sole argument is always absolutive, and the more animate of a transitive verb's two arguments is also absolutive. The remaining argument of the transitive verb is accusative if it's the patient, and ergative if it's the agent.

se-

DEF=

miij

child{ABS}

peŋktu

fox-PL-ACC

juaw-soo

bite{NPST}=3>3

"The child bites foxes"

se-

DEF=

miij

child{ABS}

peŋŋtiil

fox-PL-ERG

juaw-siel

bite{NPST}=3<3

"Foxes bite the child"

The three oblique cases, locative, lative, and directive (which denotes directions) can be extended by a set of directionality clitics: sää "into or towards", hup "out of or away from", mua "adjacent to", and wiel "through" (this set may expand later). Up to two clitics can be attached to a noun, and the meanings combine iconicly: ŋälhis-sää-hup "into then out of a room"

Rather than a genitive case, there's a possessive adjective with the suffix -j- ~ -s-. If the base noun is one of classes C-ä, C-e, C-i, or V-i, then the possessive adjective inflects as class C-e (e.g. loc.sg. -sie), while if the base is one of C-a, C-o, C-u, or V-u, then as class C-o (e.g. loc.sg. -suo). Note that when the suffix is added to consonant stems, there is a linking vowel ää for class C-ä, ee for class C-e, and so on.

[Kuchigakatai]

I'm afraid this is so. Originally all the stems were at least disyllabic, and the final vowel has been absorbed into the endings, where it survives at all.

The phonology isn't terribly interesting, but that's ok. The main divergences from the IPA are /b d g/ which are [β ð ɣ], except when immediately before a voiceless stop, in which case they devoice to [ɸ θ x]. The [ɸ] produced by this merges with original /f/ [ɸ].

So I basically got it all backwards. It's [ð] that devoices to [θ], not an old /θ/ [θ] that voiced to [ð].

[Ares Land]

I really like that system of morphosyntactic alignment.

I have half a mind to at some point derive some sisterlangs to this one and call it a reconstruction challenge.

That'd actually be pretty nice!

[bradrn]

The absolutive, accusative, and ergative cases are used for the core arguments of the verb in a way that depends on the animacy of the arguments. An intransitive verb's sole argument is always absolutive, and the more animate of a transitive verb's two arguments is also absolutive. The remaining argument of the transitive verb is accusative if it's the patient, and ergative if it's the agent.

se-

DEF=

miij

child{ABS}

peŋktu

fox-PL-ACC

juaw-soo

bite{NPST}=3>3

"The child bites foxes"

se-

DEF=

miij

child{ABS}

peŋŋtiil

fox-PL-ERG

juaw-siel

bite{NPST}=3<3

"Foxes bite the child"

Interesting (and realistic) system! I suggest renaming the ‘absolutive’ case though: given that it’s used for more animate arguments (=agentive ones), it looks more like a ‘nominative’ case than anything else.

[KathTheDragon]

I can't take credit for coming up with the alignment. A friend of mine on Discord shared the idea with me and I liked it.

I suggest renaming the ‘absolutive’ case though: given that it’s used for more animate arguments (=agentive ones), it looks more like a ‘nominative’ case than anything else.

A good point

[KathTheDragon]

So I basically got it all backwards.

Yep.

[KathTheDragon]

In future, whenever I cite a noun stem, I'll add a superscript vowel after the hyphen which refers back to the inflection class, e.g. wir-^ä "rock".

I haven't addressed it directly yet, but a significant number of consonant stems undergo a process of gradation, where the alternants can be described as basically voiceless and voiced. In the singular, the voiced alternant is used for the nominative, accusative, and directive cases (e.g. "room" ŋälb, ŋälbo, ŋälbom, while the voiceless alternant is used for the ergative, locative, and lative cases (ŋälhiil, ŋälhii, ŋälhis). In the plural, these associations are reversed (e.g. nominative ŋälhit, locative ŋälbtii). What I believe to be the full set of consonant gradations is shown below.



<edit>
Cluster gradations other than those shown can appear, but they show regular gradation of the second consonant, e.g. rh ~ rg.

All gradating consonant stems which end in a single consonant, as well as a number of non-gradating consonant stems that end in a single consonant, replace the last vowel of their stem (if it is short) with a long vowel or diphthong. The usual changes are ä e i a o u > iä ie ii ua uo uu, but in some cases (predominantly before invariant voiced fricatives) the vowel is simply lengthened.

Examples of nouns with alternations are maaj ~ maas-ⁱ "person", älg ~ älk-ⁱ "deer", oob ~ ob-^ä "cat", juag ~ jak-^o "the north".
</edit>

Vowel stems also show alternation, this time in their final vowel. This alternation is far more limited, however, with only the accusative and directive showing a distinct vowel from the others (and both the same vowel, it should be noted), e.g. miä "woman[ɴᴏᴍ]", maa "woman[ᴀᴄᴄ]"

[quinterbeck]

This is great! I also really love the alignment, and the stem gradations are great. I would be interested to see some more examples of the loc/lat/dir cases and the possessive adjective. That last one isn't a term I've come across before, so I'm curious how its behaviour distinguishes it from cases.

[KathTheDragon]

I happen to have a list of all the combinations of locative/lative/directive with the directionality clitics typed up already, so feast your eyes on it.

ŋälhii room-LOC "in/at the room"
ŋälhii-sää room-LOC=IN "inside the room"
ŋälhii-hup room-LOC=EX "outside the room"
ŋälhii-mua room-LOC=AD "next to the room"
ŋälhii-wiel room-LOC=TRANS "throughout the room"
ŋälhis room-LAT "to the room"
ŋälhis-sää room-LAT=IN "into the room"
ŋälhis-hup room-LAT=EX "out of the room"
ŋälhis-mua room-LAT=AD "around the room"
ŋälhis-wiel room-LAT=TRANS "through the room"
ŋälbum room-DIR "towards the room"
ŋälbum-sää room-DIR=IN "towards the room"
ŋälbum-hup room-DIR=EX "away from the room"
ŋälbum-mua room-DIR=AD "past the room"
ŋälbum-wiel room-DIR=TRANS "through the room"

As for the possessive adjective, it's an adjective that means "of X". E.g. se-hantaasuo ŋälhii DEF=room-POSS-LOC man-LOC "at the man's room"

[WeepingElf]

Nice language so far, has some Uralic flavour to it, without being a rip-off of Uralic ways. I hope that this time, it is meant to be spoken by humans, not dragons or similar nonsense

[KathTheDragon]

Nice language so far, has some Uralic flavour to it, without being a rip-off of Uralic ways.

Well I really know about Uralic is the aesthetic. I've never put in the time to dig particularly deep.

I hope that this time, it is meant to be spoken by humans, not dragons or similar nonsense

*nervous sweating*

[KathTheDragon]

I haven't done much with this language for a while _{(been playing too much Minecraft)} but I've added three more clitics that attach to the oblique cases. loo and ŋiew mark "above" and "below" respectively (erpie-loo "on top of a mountain", se-mores-ŋiew "to the bottom of the sea"). käll on the other hand marks possession. With the locative, it denotes static possession (se-miä-käll "belonging to the woman"), with the lative it denotes acquisition (se-miäs-käll "being acquired by the woman"), and with the directive it denotes the intended owner (se-maam-käll "intended for the woman").

[KathTheDragon]

After thinking about how to use adverbs (particularly oblique cases with clitics) as noun modifiers, here are two relative pronouns: kaaj ~ kaas-^a "who", kani ~ kanj-^a "which"