senzacaponecoda - Tumblr blog

[senzacaponecoda]

these are approximately the same thing cmm

{adpositions, conjunctions}

tense and aspect are analogous to the quantifier and determiner layers but seem kind of flipped relative to each other. Like aspect often encodes verbal number, maybe tense and determiners aren't like each other. but tense and quantification are both kind of like the layers underneath them while often being syntactically autonomous (all of the X, had been Xing) etc

{voice, possession}

{everything lexical}

{adjectives and adjective phrases, participles and relative clauses}

[senzacaponecoda]

Once again I am pissed that most of the Maya manuscripts are gone. So tantalizingly close to getting a look at a culture that developed genuinely in the absence of influence from Old World civilization. A fully independent data point on the development of literary culture—the kind that we know wrote philosophy, history, mathematics! And yet we barely have any of it. Can you imagine how much we might learn about humanity, about cultural universals and cultural particularities, about the history of Mesoamerica, from being able to read that work and compare it to what we know?

But it's gone. It's gone because the Spanish burned it and made learning the Maya script illegal. And it's gone because codices rot in the tropical heat and humidity, so unless you've got a continuous scribal tradition it's all going to disappear. We can't even read Maya cursive; we know there was Maya cursive, and it's likely what most codices were composed in, but it's almost all gone!

And of course this is worst for the remaining Maya themselves, who in addition to being deprived economically and legally by Spain and its colonial legacy, have been deprived of the dignity and respect that comes (for better or worse) from being part of a cultural tradition that produced great and widely respected works.

I just, you know, I just wish it wasn't gone!

But as I always say on this topic, a point of optimism: the jungles of southern Mexico and Guatemala are barely explored, they're dense and they almost certainly hide many things yet. Maybe, one day, some hidden cache will be discovered, a scribal workshop collapsed perfectly to protect the codices within from the damp. We can only hope.

[senzacaponecoda]

I've created a monstrosity

[senzacaponecoda]

Miwok takeaways

Twitter really fucked me up earlier.

I spent some time trying to summarize Miwok for some Semiticists I know in case it would be interesting. I would have asked first and made some kind of commitment but I was trying to distract myself from being in shock.

Spent some time badly summarizing a bit about the people and dipped into noun stuff. Now it's 1am and my back hurts from sitting in a chair poorly. Probably won't finish.

It's of interest to Semiticists and people who care about Proto-Afroasiatic because Miwok's said to have triconsonantal roots (but independently, being an American language) and has some syntactic stuff interesting for case. But it's oversold.

The takeaways for Semiticists

history of vowel harmony in the language and its region

VSO/VOS

It's mostly agglutinative actually. crazy numbers of tam and fusional subject/object marking and like 3 levels of clusivity distinctions make for a ton of forms.

Verb stems are like, base 1v2vv3-, 1v2v33-, 1v22v3-, 1v23v- (metathesis), biconsonantal verbs are extended with a glottal stop in some forms. Stress is non-lexical, and stresses the syllable of the second mora, so there's some stress shifting in those stems.

Basic colors form a class like 1v2vv2.

Stems are used according to VTAM, the four above are typically, respectively, basic present tense, non-present, iterative/habitual, nominal.

Vowels depend on the root. So it's not non-concatenative, there's just a mora infixing to the stem (and metathesis in the 4th stem).

There's a split between [subjunctive moods + the distant past tense + habitual aspect] and the rest where the former take possessives as subjects and the latter take nominative forms

Possession is very similar to Afroasiatic as an enclitic suffix thing. Can occur on different sides of the case morphemes.

Four example verbs:

tuyáaŋ- jump

wíktï- burn

hámme- bury

lúuš- win

tuyáaŋ-ïm - jump-1S - I am jumping

tuyáŋŋ-ak - jump-PERF - I jumped (1s is Ø here)

túyyaŋ-ïṭ - jump-1S.POS - I jump, I jump and jump

túyŋa-' - jump-NOM - jump (noun)

wḯktï-m

wïkḯṭṭ-ak

wḯkkïṭ-ïṭ

wḯktï-'

hámme-m

hamé'-ak

hámme'-ïṭ

hám'e-'

lúuš-um (vowel harmony, ï -> u / [u,o]..._)

lúšš-ak

lúššu'-uṭ

lúš'u-'

[senzacaponecoda]

This is either incredibly wrong or obviously correct I think. I had an epiphany that I'm not sure is accurate or not:

pitch, loudness, and probably length are just features in the same way [round] is a feature

They can be arbitrarily part of a base word, like /l/ is an arbitrary part of "lion". Half of Russian stress is like this. Japanese and Swedish are like this with tone to a degree, I think? What people think of as tonal languages (e.g. Mandarin) are like this but have a whole inventory of tonemes, not just one. Length is definitely like this for say Italian or Japanese

They can be part of the derivational morphology, like an affix. So like how stress changes can convert some verbs (e.g. re'cord) into nouns ('record) in English. Or tonally like Somali ínan vs inán (boy vs girl), nácas vs nacás (idiot, by gender), etc. A number of languages make aktionsart by lengthening a segment as a kind of reduplication

It can be morphological. Spanish distinguishes the 1S present ending o from the 3S preterite ending ó by stress for a partial example. I'm sure plenty of langs have a "mere" stress change in morphology I'm not remembering. Tonally case endings in Japanese can have their own tone that overrides neighboring tone, iirc in Nobiin there are number suffixes and cases that are wholly tonal in nature. Length and morphology, Finnish might be a good example.

They also act phonologically without semantic information, like in prosody. We don't necessarily think of vowel reduction as a prosodic thing in itself but downstream of stress; but maybe it could be thought like that. In anycase stress and tone are very sensitive to meter of course (to the point where I'm writing about them because they're not just active at this "level") and stress and length frequently cooccur. At higher levels of the prosodic heirarchy than the foot you'll see things like phrase final stress (e.g. French) or tonal things (question intonation), lengthening, and for normal features you often see things like debuccalization or nasalization at this level

It also acts like Sandhi. So like functional words often have stressed and unstressed variants as a kind of stress sandhi. Mandarin has tone sandhi. Many langs have shortening processes for avoiding clashing strophes.

And of course intonation and affectation obv.

I think it's just early academic linguists "fetishize" tone and stress as something different because a) they don't segmentalize easily, unlike [round] or whatever b) western europe's language area confines these features to relatively high in the phonological hierarchy

and this has inertia, so new generations of linguists get the old framing to look through and sludge through, which often reinforces the exoticism or whatever.

[senzacaponecoda]

Reading Hayward's 1990 overview on Aari, a South Omotic language.

Apparently Aari developed (for the speaker) breathy voice by losing h (visible in synchronic alternations and comparisons with related Dime), partly conditioned by tone (breathy voice is associated with high).

It's making me think about PIE of course because I don't know much about anything with breathy voice. But I kind of want to toy with a hypothesis that H in PIE developed with the beginning of the loss of laryngeals into bh dh gh gwh

[senzacaponecoda]

c'mon now

[senzacaponecoda]

Sometimes I get this vibe from afroasiatic types that uvulars are somehow connected naturally to ejectives and so there's some sense that k' is naturally q like

I think that's an illusion born out of a merger of k' into q in semitic and maybe q into k' in egyptian

k'ich'ee maya for example is perfectly comfortable with k' and q(χ)' existing. So are many other Maya langs. Affrication of ejective q is pretty common, though

Considering how marked pharyngeal fricatives are and that they usually come from uvulars, and that Amazigh at least shows evidence of palatalizing them, and that the pharyngeal fricatives distinguish voice when none of the other PAA fricatives distinguish voice (except ghayin's ancestor? but maybe it's secondary) I kinda think PAA had k' k g and q(χ)' q ɢ and Semitic turned qχʼ and ɢ into ħ and ʕ and then merged k' and q as marginal dorsal stops. Egyptian might have done the same.

[senzacaponecoda]

dumb idea i got reading about attempts to make sense of something in egyptian

what if the affix conjugations were originally tense while the infix reduplication was aspect

q-t' cut

ya-qt'a he cut (pres + perf)

ya-qatt'a he is cutting (pres + impf)

qat'-su he cut (past + perf, atelic)

qatt'a-su he was cutting (past + impf)

qat'-â (perf + telic)

sqim listen

ʔasqim

ʔasaqqim

saqmay

saqimmay

saqimaku

something

Some families then would have 'lost' the suffixing conjugation, because the "preterite" was available. Egyptian might have lost the prefixes for some reason or another, maybe all of sac'immaf/sacc'imaf/sac'maf existed

[senzacaponecoda]

[ǥ̇] as in German "ja" (IPA fricative [j])

[senzacaponecoda]

rate this conlang phonology

f is phi, not f.

' is the glottal stop.

the sibilants are dental, the š are retracted in some way, maybe postalveolar, maybe just retracted.

fricatives are allophonically voiced.

uvulars are allophonically affricates.

vowels can be short or long, so i guess max is really cvc or cvv (no cvvc).

[senzacaponecoda]

my understanding. correct if wrong

preterite and aorist don't really have great comparitive linguistic uses but tend to get used by linguists who specialize in Indo-European languages or middle eastern philology to describe analogous forms in related or nearby languages. That said

both are usually a kind of past perfect considered less marked or at least less phrasal than another past perfect

the aorist is the term when this is the unmarked stem of a verb, like in ancient Greek

preterite is the term when this is relatively marked especially when compared to an imperfect aspect or present tense, like Latin or Germanic. Don't remember what Latin did but Germanic mostly added do to the end of the aorist to get a preterite.

what distinguishes this form from another past perfect differs lang to lang. When there are two forms competing for space generally the main difference for verbs expressable as both is that one is used more as a relative tense, i.e. to establish a relative sequence of tenses when each occur in the same tense relative to present.

It might be that while both are fundamentally past and perfective, one might also get preferentially employed to cover marginal aspects like gnomic or modal aspects, and this leads to a foot in the door kind of eventual overcrowding.

The more phrasal form over time might end up morphologically more regular over time due to the head base being consistent while the historically less marked form picks up vowel or consonant mutations or an increasingly rare phi feature pattern. So the phrasal wins out on morphological awkwardness - why remember cezim becomes toro when you could just say kefim cez. Or like. Ancient langue had yut pir vs par-ta vs par-in-ta, now modern language has tfai which becomes tfai vs tfai which becomes tmfai.

Or just something sociolinguistic like prestige foreign language limiting an analogous form to an edge case so the less prestige lang starts to imitate their use patterns.

There are probably other roots to grammaticalizing one over the other

[senzacaponecoda]

my understanding. correct if wrong

preterite and aorist don't really have great comparitive linguistic uses but tend to get used by linguists who specialize in Indo-European languages or middle eastern philology to describe analogous forms in related or nearby languages. That said

both are usually a kind of past perfect considered less marked or at least less phrasal than another past perfect

the aorist is the term when this is the unmarked stem of a verb, like in ancient Greek

preterite is the term when this is relatively marked especially when compared to an imperfect aspect or present tense, like Latin or Germanic. Don't remember what Latin did but Germanic mostly added do to the end of the aorist to get a preterite.

what distinguishes this form from another past perfect differs lang to lang. When there are two forms competing for space generally the main difference for verbs expressable as both is that one is used more as a relative tense, i.e. to establish a relative sequence of tenses when each occur in the same tense relative to present.

It might be that while both are fundamentally past and perfective, one might also get preferentially employed to cover marginal aspects like gnomic or modal aspects, and this leads to a foot in the door kind of eventual overcrowding.

The more phrasal form over time might end up morphologically more regular over time due to the head base being consistent while the historically less marked form picks up vowel or consonant mutations or an increasingly rare phi feature pattern. So the phrasal wins out on morphological awkwardness - why remember cezim becomes toro when you could just say kefim cez. Or like. Ancient langue had yut pir vs par-ta vs par-in-ta, now modern language has tfai which becomes tfai vs tfai which becomes tmfai.

Or just something sociolinguistic like prestige foreign language limiting an analogous form to an edge case so the less prestige lang starts to imitate their use patterns.

There are probably other roots to grammaticalizing one over the other

[senzacaponecoda]

how many languages do you speak?

(i’m counting languages where you took one class for a semester if you retained any of it congrats you are a little multilingual)

(reblog for bigger sample size!)

[senzacaponecoda]

yapattu ramiku hisana mi pira, ka li qatilata lafirata nafarata mita 2a.rabina

yapattë ramicë hisanä më pirä, kä lë qatiLätä lafirtä nafartä mitä 2arbinä

yápattë rámic hísan mê pír, kâ lê qátíLäd láfirtä náfartä mítä 2arbin

yɛpɐttɪ rɔmic hisɔn mɨ piɾ | kə lɨ qɔtiʎʝ lɔɸɪɾtə nɔɸɐɾtə mitə ʔɐrβɪn

ipat zo romi isou mi pi ke li qɔtʒi lohirt nohart mit arve

zə rom izu zə pi mi pat yes zə zat loyr noar w arve mih zə qac li ani ke

zərom wizuli zəppimi paces zə zatlør noar warve mi zqacʎaɳ ke

ərromə izzuʎ əppimə face a noar ta taler warvemə misqaccaɳ ce

fac arrom izuy apim a mwarta talleer ervim misqaccan ce

fac en arru izuk abi an martal tallir irbi c'axx'ece

hac n arru izuŋ nar abi in c'aqqe d tamart tallir ibiŋ

the rru is hæshing izu in a vi, that so the mærry llir would shawq an ivi

idk

[senzacaponecoda]

hmm If there's an established topic, a "something" to be antecedant I find "She just loses hers" acceptable "He just loses his" ok *"It just loses its" bad "They just lose theirs" ok Why is 'its' acting as the whole argument there bad to me? Is that use bad to you?

[senzacaponecoda]

i can't think of a phonetic reason, unless the shortening of vowels before fortis obstruents is having an effect on the footing of this or something syntactically they're the same things, unless mysteriously they aren't. like does animacy matter? why would animacy matter

hmm If there's an established topic, a "something" to be antecedant I find "She just loses hers" acceptable "He just loses his" ok *"It just loses its" bad "They just lose theirs" ok Why is 'its' acting as the whole argument there bad to me? Is that use bad to you?