More dialects and some phrases

Just some things that I have been thinking about the last few days, regarding TjKt mostly. First is that there is a dialect common along the coastal regions best known for the sound change [+syllabic, +short]->Ø/C_$, that is short vowels deleting in open syllables (of roots, ie etymologically). This preserves words like jāmah "faith" but not jimha "to believe in something" which instead would become jem(h) (I threw in a /ɪ/->[ɛ] as well :p). Of course, this leads to problems with words like jmohi "faithful", so I'll need to put in some more rules so that not only consonants are left. Maybe only word final deletion in that case, which would then result in a closed syllable. Another alternative would be lengthening word final short vowels if the previous vowel is also short. In this cases the options would end up being jnoh or jn(h)ī respectively. I'm kind of preferring the second option right now, it leads to some interesting stuff. I also need to figure out how to avoid clusters of intital CCC coming from CCVC roots, which leads to option 3, metathesis to CVCC. Then we'd have jom(h) as the word coming out. Yeah, not sure what to do yet. I think option 3 works the best overall. Hmm, maybe some sort of chain thing. Changes start from the right side of the word, which can change a open to a closed syllable and save the word. This can end up looking with case 1. When the final vowel is already long and the form is (V)CCVCV: then metathesis occurs to CVCCV: (option three). So the word jmohi turns to jnoh but jmuhā "thing which causes belief" becomes jmahā->jam(h)ā (/u/->[ə] in analogy with the short i sound change). Ujmahū "faithfully" becomes jamhū  because the metathesis eliminates the closedness of the first syllable.

Going down the whole 4th class (ignore the colors, html is hard), I think we get the following forms (bolded if different from standard): 

CāCaC

CCoC

ūCeCC

CeCC

CōCāC

aCCōC

CCīC

CaCCū

CūCaC

CaCCā

īCCōC

CaCCū

3rd:

CōCoC

īCCaC

CaCCī

CāCC

ūCCāC

CīCūC

āCCāC

CaCC

CīCoC

āCīCC

CCūCō

CCīoC

2nd:

āCuCC

CīCeC

CaCC

aCCōC

ūCCoC

CCaC

oCCīC

CCīC

aCCīCā

CaCC

CōCaC

CaCCī

CīCoC

1st:

CīCC

CūCC

eCCūC

CCīC

CCōC

CāCaC

CeCC

ūCeCC

CeCCī

CCaC

CīCaC

Somehow there are no duplicates within a class. Praise Kīkx!

I might throw in some mergers as well, like the alveolar affricates merging with the palatal-alveolar ones. I do think this one will preserve the pharyngeals, and I think the ejectives might weaken to pharyngealized consonants, with a loss of the non-stop ejectives. It also doesn't spirantize intervocalic stops, but does contrast consonant length intervocalically.

So let's compare some sentences now (though I haven't worked on semantic/grammatical/pragmatic stuff, so it's basically only applying sound changes at this point).

"People pray to Kikxo so that they are blessed"
Úīkmo nonsīnī Kīkxo mābíi gagaxātap-gātāp[wi:kŋɔ nɔnsi:ni: xi:kʃɔ ma:bjɪ ɣəɣəʃa:θəpga:θa:p]
Úīk nonsīnī Kīkx mābī gagaxātap-gātāp[wik nɔnsi:n:i ki:kʃ ma:bi: gəg̵əʃa:təpga:ta:p]

Pretty similar with that one

"The fishermen are eating dog in the lake"
Shbīmuzō fatiúja rōxub qal gfutī[sʼbi:mʊtʃo: ħəθɪwdzə ro:ʃʊb qʼəl għʊθi:]
Spīnzō fateúj rōxab qal gafthī[spi:ntʃo: ħətewdʒ ro:ʃəb ɢǝl gəħtˤi:]

That one is pretty similar too. Let's try two more.

"I have seen stars in the desert and heard wind on the island"
Yān khopihma kōkob qal thuyī ūmpa khoniyka citham qal úlunī[ha:n kʼɔfɪʔmə xo:xɔb qʼəl tʼʊhi: ju:mpə kʼɔni:kə tsɪtʼəm qʼəl u:lʊni:]
Yān khpehm kōkob qal tahyī ūmp khnek ztham qal úalnī[ha:n qpɛʔm̩ ko:kɔb ɢəl təʔhi: ju:mp qnɛxk tʃˤtˤəm ɢəl wəlni:]

"Sentient beings talk and eat"
Tiújī ūtiúj ūmpa tiúja[tɪwdzi: ju:θɪwdz u:mpə θɪwdzə]

Teújī ūteúj ūmp teúj[tɛwdʒi: ju:tɛwdʒ u:mp tɛwdʒ]

---

Outside of dialects, I've been thinking about sayings and such. Euphemism and the like as well. One root I was thinking about is NSP "to travel (in a group)". Some important words from this root are nōsup "caravan" and ansōp "to travel (in a group); to go from one point to another for trade". Now the transitive stem onsīp isn't really used formally, at least without an applicative. Well, it makes sense that this could be used as a causative "to make a group of people travel". While this could mean like "to send off a caravan" but more normally/colloquially it means "to exile a group" or even "to ethnically cleanse/force a migration".  It doesn't have to necessarily be malevolent. For example a gafto "flood" could force people to evacuate, such as in the sentence lbupī onaxsīp gafto "The people evacuated the village because of the flood" (lit. "As for the village, a flood forced it away")

Another interesting use of NSP is nōsup-nōsōp "to give a caravan to someone" (since these types of verbs always have recipients for direct objects). This, when taken literally, is a little strange. However, in actual use it means "to invest in someone's business venture (usually by supplying capital)" since rich merchants would earn money by giving up-and-coming merchants the right to use their goods and caravan in exchange for a cut of the profits. This later extended to a general meaning of "to invest (in someone)". It then later also gained the meaning of "to give an inheritance to someone" since those same goods later became the basis of many a merchant's son's inheritance.

This construction in general is known for having a large number of idiomatic meanings. A classic is sīqro-sīqri "to give give a butt to someone". This has come to mean "to kiss up to someone, to show deference to someone" since within Kikxotian culture, to bend over like this would be a sign of making yourself vulnerable to someone (specifically for their gain...they aren't a very progressive culture).  pōjop-pōjōp "to give death" is another example, here meaning "to execute" as opposed to the more generic pījūp "to kill". There's āruyt-ārāyt "to give a tongue" or rather "to claim a bounty (from someone), stemming from the tradition of cutting off the tongue of a person or animal to show that you killed them.

Ācutr-ācātr "to give a choice" is a very interesting one. At face value, this is a good thing. However in actual use this means "to threaten". Kikxotians value choices and agency. Therefore, in sentencing and other such arrangements, the criminal would often be given a choice in their punishment. In many private arrangements, "choices" are given as well even if everyone knows that only a certain option will be decided. Plenty of corrupt individuals and criminals would preface (such as in a tax shakedown or protection scheme) their demands with "I'm giving you a choice" until this became a euphemism for "to threaten". For some people, some back euphemism stuff has even happened where the original ācutr "choice" has come to mean "threat".

A few others ones. Wsuzī-wsūzū "to give a marsh" is to give someone something that is utterly worthless in order to mock them. It's like a white elephant gift but without the prestige. Tōwow-tōwōw "to give ice" is to do something to someone that has only temporary benefits before fading away. For example, a really terrible doctor might be accused of "only giving people ice" instead of actually healing them. Nōvos-nōvōs "to give a plow" is to endebt or enslave someone. 

Point is, this is a very productive but also highly context based construction. Many of these simply have to be learned, especially in the dialects where the first part is dropped. Like in any language, the meanings can't simply be learned as a derivation of common root, but instead as component of the culture. How else would we know that rōxub-rōxōb is "to invite someone on a day outing" or wxurā-wxūrū is "to be a quisling". Let alone that naxōíox-nōíōx is "to take someone in for the night" (this one is almost exclusively used in the passive, the active has a meaning like "to give thanks to a host".  Neither would be guessable from the root which means "grass").

Just a lot of musings