Kasshian and Other Languages

[languagesofgalhaf]

New marginal phoneme /ɹ/ in Shivrashanian. Formerly, this was used solely as an allophone of /ɻ/ after coronal consonants. And there’s still a regular /ɻ/ → /ɹ/ after coronals. But now /ɹ/ is phonemically distinctive from /ɻ/ in clusters /mɹ/ /ŋɹ/ /pʰɹ/ and /kʰɹ/ where they’re derived from clusters /mn/ /ŋn/ /pʰn/ /kʰn/

[languagesofgalhaf]

For the time being I’m using ø for /ɤ/. It’s a bit of an unusual usage for that character, admittedly, as it’s practically the opposite of its value in IPA, but it looks better than other options, no ambiguity like a digraph would have and minimal awkward stacking of diacritics

Plus I’ve tweaked some of the sound changes to reduce the amount of unpredictable stem changes. I’ll probably also apply some generous leveling to further reduce stem changes in inflectional paradigms

[languagesofgalhaf]

I’m unable to settle on a good way to write /ɤ/, which is a fairly common phoneme in Shivrashanian. Previously, I had used ŏ, but unfortunately, that doesn’t work well when combining with other diacritics, namely the acute and the macron. I’d considered a digraph eo, /eo/ is a possible, albeit uncommon, sequence in Shivrashanian, so it would be ambiguous, unless I use something like eö to distinguish the sequence /eo/ from /ɤ/

[languagesofgalhaf]

So, I’m picking up work on Shivrashanian again, having been playing with various ideas for a while, particularly in terms of inflections, and a few tweaks in the phonology. In particular, I’m considering reducing the vowel inventory slightly. The current vowel inventory is /i ɪ e ɛ ɐ ə u ʊ o ɔ ʌ ɤ/, all of which, except for /ə/, can be long and/or creaky voice (/ə/ can also be creaky voice, but cannot be long, nor stressed). I’m pretty sure I’ll merge /ʌ/ and /ə/, but I’ve been debating several thoughts for /ɪ/ and /ʊ/. One idea is to merge them with, respectively, /ɛ/ and /ɔ/, but another is to merge them as /ɨ/. Perhaps I’ll let both happen in different dialects. I’m also working on a few other tweaks to the sound changes which wouldn’t affect the phonemic inventory, but would affect the frequency of certain phonemes

[languagesofgalhaf]

Probably a good part of last year’s Lexember entries are now out of date as I’ve been making a number of revisions to Shivrashanian.  The most recent change is that there is now a marginal phonemic distinction between /ɹ/ and /ɻ/.  After coronal consonants, only /ɹ/ can occur, while intervocalically, word-initially, and after most non-coronal consonants, both can occur, but /ɹ/ is rare.  There are probably no minimal pairs

[languagesofgalhaf]

Just when I think I’m getting the verb system down, I find myself rethinking the negative inflections, rethinking how the present and imperfect negatives would be derived.  I kinda like the new idea I’m playing with in that the negatives are more clearly distinct from the affirmaties, but the major downside is that some verbs end up with special stem changes in the negatives

[languagesofgalhaf]

This is a partial verb paradigm for Class I verbs in Shivrashanian.  Class I verbs are those descended from vowel-final verbs in Kasshian.  I’ve posted a more complete paradigm previously, but there’s been a few minor adjustments

There are several syncretisms here, and in none of these are all nine cells distinct

In the present/imperfect observational and inferential, the 2nd person singular, 3rd person singular animate, and 3rd person plural animate are all homophonous.  In the present/imperfect negative, the 3rd person singular animate and 3rd person plural animate are homophones.  In the preterite, the 3rd person forms are all distinct, but the 2nd person singular and plural are homophones

I am considering the possibility of an analogical restructuring, such that the 2nd person plural and the 3rd person animate plural are always syncretized.  Thus, in the present/imperfect observational and inferential, the endings -defne̜e̜ and -gefne̜e̜ would be extended to 3rd person plural animate, so that the plural has the same 2nd person/3rd person animate syncretism as the singular, and a number distinction is simultaneously restored to the 3rd person animate.  In the present/imperfect negative the 2nd person plural ending -dre̜fi̜ne̜e̜ is likewise extended to 3rd person animate, while in the preterite, the 3rd person animate plural endings -bŏ̜́ste̜ and -bŏ̜ste̜ste̜ are extended to 2nd person.  In those forms, 2nd person and 3rd person animate remain distinguished in the singular, but are syncretized in the plural, and number remains distinguished in all persons

The optative forms have a different pattern of syncretism:

In the affirmative, number is distinguished in all persons, while in the negative, the 2nd person lacks the number distinction.  However, both also have syncretisms that cuts across person and number, with the endings -́quf and -q in the affirmative and -qre̜ste̜ in the negative

The same 2nd/3rd animate plural merging could also happen here as well by analogy, which would also be motivated for the negative, so that -glef and -qre̜ste̜ would be 2nd plural and 3rd animate plural

But maybe even further, the pattern of 2nd/3rd animate syncretism that would exist in the present/imperfect affirmative could be extended to all of these forms, the question then being whether the 2nd person or 3rd animate would be kept.  This would effectively convert the person system into 1st vs non-1st, with non-1st being further subdivided into animate and inanimate.  I kinda like the idea of such a system.  I’d actually played with the idea previously, but in that case, the idea was a development similar to several European languages where a 3rd person form was adopted as a polite 2nd person pronoun (e.g., Spanish usted or Portuguese você), replacing the original 2nd person entirely.  If I go this route, it would have the same end result of no 2nd/3rd person distinction, but some of the endings would derive from the original 2nd person forms

[languagesofgalhaf]

The major argument against such a restructuring would be that the sound changes that created the loss of number distinctions is relatively recent.  For example, the present/imperfect affirmative forms previously had a syncretism between 2nd person singular and 3rd person animate plural, but the animate singular remained distinct, and only merged with the other two recently.  And while a 2nd singular/3rd animate plural syncretism is a bit odd, it would have very little chance of ambiguity, so that there’d wouldn’t be much pressure to restructure it.  So, even if it did happen after those sound changes, there would probably not have been enough time for it to spread beyond the immediately-motivated changes

[languagesofgalhaf]

But also, even before those changes that merged singular and plural, they were already similar.  The distinction between 3rd animate singular and plural in the present/imperfect observational, for example, was just vowel nasalization, so there could still have been a weaker motivation to extend the 2nd person plural to the 3rd animate plural.  And changes to inflectional paradigms don’t always have to be motivated by phonetic considerations anyways

[languagesofgalhaf]

This is a partial verb paradigm for Class I verbs in Shivrashanian.  Class I verbs are those descended from vowel-final verbs in Kasshian.  I’ve posted a more complete paradigm previously, but there’s been a few minor adjustments

There are several syncretisms here, and in none of these are all nine cells distinct

In the present/imperfect observational and inferential, the 2nd person singular, 3rd person singular animate, and 3rd person plural animate are all homophonous.  In the present/imperfect negative, the 3rd person singular animate and 3rd person plural animate are homophones.  In the preterite, the 3rd person forms are all distinct, but the 2nd person singular and plural are homophones

I am considering the possibility of an analogical restructuring, such that the 2nd person plural and the 3rd person animate plural are always syncretized.  Thus, in the present/imperfect observational and inferential, the endings -defne̜e̜ and -gefne̜e̜ would be extended to 3rd person plural animate, so that the plural has the same 2nd person/3rd person animate syncretism as the singular, and a number distinction is simultaneously restored to the 3rd person animate.  In the present/imperfect negative the 2nd person plural ending -dre̜fi̜ne̜e̜ is likewise extended to 3rd person animate, while in the preterite, the 3rd person animate plural endings -bŏ̜́ste̜ and -bŏ̜ste̜ste̜ are extended to 2nd person.  In those forms, 2nd person and 3rd person animate remain distinguished in the singular, but are syncretized in the plural, and number remains distinguished in all persons

The optative forms have a different pattern of syncretism:

In the affirmative, number is distinguished in all persons, while in the negative, the 2nd person lacks the number distinction.  However, both also have syncretisms that cuts across person and number, with the endings -́quf and -q in the affirmative and -qre̜ste̜ in the negative

The same 2nd/3rd animate plural merging could also happen here as well by analogy, which would also be motivated for the negative, so that -glef and -qre̜ste̜ would be 2nd plural and 3rd animate plural

But maybe even further, the pattern of 2nd/3rd animate syncretism that would exist in the present/imperfect affirmative could be extended to all of these forms, the question then being whether the 2nd person or 3rd animate would be kept.  This would effectively convert the person system into 1st vs non-1st, with non-1st being further subdivided into animate and inanimate.  I kinda like the idea of such a system.  I’d actually played with the idea previously, but in that case, the idea was a development similar to several European languages where a 3rd person form was adopted as a polite 2nd person pronoun (e.g., Spanish usted or Portuguese você), replacing the original 2nd person entirely.  If I go this route, it would have the same end result of no 2nd/3rd person distinction, but some of the endings would derive from the original 2nd person forms

[languagesofgalhaf]

The major argument against such a restructuring would be that the sound changes that created the loss of number distinctions is relatively recent.  For example, the present/imperfect affirmative forms previously had a syncretism between 2nd person singular and 3rd person animate plural, but the animate singular remained distinct, and only merged with the other two recently.  And while a 2nd singular/3rd animate plural syncretism is a bit odd, it would have very little chance of ambiguity, so that there’d wouldn’t be much pressure to restructure it.  So, even if it did happen after those sound changes, there would probably not have been enough time for it to spread beyond the immediately-motivated changes

[languagesofgalhaf]

This is a partial verb paradigm for Class I verbs in Shivrashanian.  Class I verbs are those descended from vowel-final verbs in Kasshian.  I’ve posted a more complete paradigm previously, but there’s been a few minor adjustments

There are several syncretisms here, and in none of these are all nine cells distinct

In the present/imperfect observational and inferential, the 2nd person singular, 3rd person singular animate, and 3rd person plural animate are all homophonous.  In the present/imperfect negative, the 3rd person singular animate and 3rd person plural animate are homophones.  In the preterite, the 3rd person forms are all distinct, but the 2nd person singular and plural are homophones

I am considering the possibility of an analogical restructuring, such that the 2nd person plural and the 3rd person animate plural are always syncretized.  Thus, in the present/imperfect observational and inferential, the endings -defne̜e̜ and -gefne̜e̜ would be extended to 3rd person plural animate, so that the plural has the same 2nd person/3rd person animate syncretism as the singular, and a number distinction is simultaneously restored to the 3rd person animate.  In the present/imperfect negative the 2nd person plural ending -dre̜fi̜ne̜e̜ is likewise extended to 3rd person animate, while in the preterite, the 3rd person animate plural endings -bŏ̜́ste̜ and -bŏ̜ste̜ste̜ are extended to 2nd person.  In those forms, 2nd person and 3rd person animate remain distinguished in the singular, but are syncretized in the plural, and number remains distinguished in all persons

The optative forms have a different pattern of syncretism:

In the affirmative, number is distinguished in all persons, while in the negative, the 2nd person lacks the number distinction.  However, both also have syncretisms that cuts across person and number, with the endings -́quf and -q in the affirmative and -qre̜ste̜ in the negative

The same 2nd/3rd animate plural merging could also happen here as well by analogy, which would also be motivated for the negative, so that -glef and -qre̜ste̜ would be 2nd plural and 3rd animate plural

But maybe even further, the pattern of 2nd/3rd animate syncretism that would exist in the present/imperfect affirmative could be extended to all of these forms, the question then being whether the 2nd person or 3rd animate would be kept.  This would effectively convert the person system into 1st vs non-1st, with non-1st being further subdivided into animate and inanimate.  I kinda like the idea of such a system.  I’d actually played with the idea previously, but in that case, the idea was a development similar to several European languages where a 3rd person form was adopted as a polite 2nd person pronoun (e.g., Spanish usted or Portuguese você), replacing the original 2nd person entirely.  If I go this route, it would have the same end result of no 2nd/3rd person distinction, but some of the endings would derive from the original 2nd person forms

[languagesofgalhaf]

Terminology changes

Evidentiality:

Direct → Observational

Indirect → Inferential

Reportative → No change

Case

Nominative → Direct

All other cases unchanged

[languagesofgalhaf]

So, with regular sound changes, I end up with an auxiliary for the perfect tense being lilelelee, which is kinda hilarious and obviously begging for irregular sound changes, especially haplology

[languagesofgalhaf]

I’m working more on the verb system of Modern Shivrashanian.  It uses a mixture of verbal inflections and auxiliaries, which incorporate voice, e.g., there are different future tense auxiliaries depending on voice, and in the actor-voice, there are different auxiliaries for intransitive vs transitive verbs

I’m also considering changing some of the terminology.  I may change the names of the evidentiality values, for example.  There are three of them, which I’m currently calling direct knowledge (or direct for short), indirect knowledge (or indirect) and reported.  Direct knowledge is for things that the speakers knows from direct experience, indirect knowledge is things not directly witnessed, but concluded based on past experience, present evidence, etc., while reported is for hearsay.  I may change those terms, but I’m not sure what to.  Part of the reason is that the case I’ve been tentatively calling “nominative” I’m considering renaming “direct”, and I’d prefer not to use the same term for a noun case and for a verb feature

There are several different participles and an infinitive-like form.  At least one auxiliary has different functions depending on which participle is used with it

The verb “to be” is highly irregular, and suppletive, using forms derived from three different sources - the Kasshian verbs taf (to be, equative*), klesta (to become) and gonsa (to pretend to be), with klesta being the source of the majority of its paradigm, with the present tense forms preserving an otherwise-lost set of perfect inflections - that is, “have/has become” came to be used for “am/are/is”

*Kasshian had two different copulas, a token copula indicating that X is a member of the set Y, and an equative copula, indicating that X = Y, this distinction has not survived into Shivrashanian or most other Neokasshi languages

[languagesofgalhaf]

Lexember Day 31

Pepši̜m [pʲʰeˈpʃɪ̃m] Money

Traditionally, this takes plural agreement, though sometimes singular agreement is found.  Singular agreement is considered “incorrect”, although it is fairly common.  The modern form is based on what was originally the oblique stem.  The expected nominative would be *pepšiŏ̜́ğam [pʰʲepʃiˈʌɣɑ̃m]

It is descended from the Kasshian pefshigamai, the plural of pishigama “copper”, also used to refer to copper coins used as a low denomination of currency.  The plural form came to also be used to refer to money in the generic sense as well, and the original singular was lost, in the currency sense

Pçu̜m [ˈpçʊ̃m] Copper

From pishigama above

[languagesofgalhaf]

Lexember Day 30

Tšišeʔ, stem tšišep-, obl. stem tši̜ši̜p-; pl. tšitšep, obl. stem tši̜tši̜p- [t͡ʃi̥ˈʃḛ], [t͡ʃi̥ʃepʰ-]. [t͡ʃɪ̥ʃɪpʰ-]; [t͡ʃi̥ˈt͡ʃepʰ], [t͡ʃɪ̥t͡ʃɪpʰ] “Wife”

Ižeʔ, etc., pl. nedžep [iˈʒḛ], [neˈd͡ʒepʰ] “Husband”

Ššeʔ, etc. pl. šitšep [ˈʃːḛ], [ʃi̥ˈt͡ʃepʰ] “Spouse” (androgyn/epicene)

All three of these are descended from the same Kasshian root, (chi-/na-/su-)choppa, “spouse”, derived from tai- (which becomes ch- before a vowel) “person associated with” + oppa “love” (in turn from the verb upa- plus the abstract noun suffix -la/-*a), originally literally “beloved”

In the nominative singular, the stem final -p becomes ʔ originally the phoneme /ʔ/, but now realized solely by creaky voice on the preceding vowel, a regular sound change in which word-final voiceless stops merged to /ʔ/.  The original stem-final stop is realized in other cases.  In some words in which this process happened, the paradigm was later levelled by extending the stop found in other cases to the nominative, thus effectively undoing that sound change

Ššeʔ [ˈʃːḛ] is sometimes found with a lengthened vowel, since ordinarily open monosyllables with monophthongs are obligatorily long (a 2-mora minimum, with diphthongs and long vowels counted as 2 morae, and codas being moraic).  There is an underlying /ʔ/, so that phonemically [ˈʃːḛ] is /ʃːeʔ/, but as it surfaces as an open monosyllable, some speakers apply the lengthening rule (arguably, this could be seen as evidence that those speakers now analyze creaky voice as phonemic, and thus see words like this as /ʃːḛ/)

[languagesofgalhaf]

Lexember Day 29

Ofkedifei; stem ofke̜-, 2nd stem uo̜ke̜- [ofkʰʲediˈfʲej], [ofkʰʲɛ], [uɔkʰʲɛ] “To love (romantically)”

This is descended from the Classical Kasshian upatri- meaning “to court”, in turn derived from upa- “to love” + -tri- “hope to”, thus literally “hope to love”.  The meaning gradually shifted from seeking a romantic relationship to being in one

[languagesofgalhaf]

Lexember Day 28

Yo̜q, pl. yăro̜q [ˈjɔqʰ], [jəˈɻɔqʰ] A term of endearment, especially, but not exclusively, romantic; archaically, “pet” in the sense of companian animal

Synchronically, the plural seems to have an infixed -ăr-, but there’s no such infix historically.  At an earlier stage, the singular also began with r-, and the alternation between the singular and plural was in the form of an infixed -y- for the singular and a prefixed ye- for the plural, a fairly common paradigm.  However, a regular sound change deleted r (at the time /ɹ/) before /j/, thus making the previously infixed -y- word-initial, and causing an apparent plural infix

It is derived from the Classical Kasshian çizonklai meaning “pet (companion animal)” from çenklai “animal” + the dimmunuitive prefix zu-

[languagesofgalhaf]

Lexember Day 27

Hó̜le̜f, pl. o̜flo̜f [ˈhɔlɛf], [ɔˈflɔf] Marital unit; two or more people who are married to each other

Polygamy is legal and not that uncommon in modern Shivrashan - both multiple wives and multiple husbands, and other combinations of genders

This is a mildly irregular verb.  The singular is inflected as:

Nominative: hó̜le̜f [ˈhɔlɛf] Ergative: hó̜llăf [ˈhɔlːəf] (expected *hó̜lle̜f, with infixed -l-) Accusative: qlo̜faa [ql̥ɔˈfɐː] (expected *ho̜le̜faa) Oblique stem: ho̜lăf- [hɔləf-]

The plural is inflected as:

Nominative: o̜flo̜f [ɔˈflɔf] Ergative: o̜fó̜llăf [ɔˈfɔlːəf] (expected *o̜flo̜li̜f with infixed -li̜-) Accusative: o̜flo̜fr [ɔˈflɔfɻ̩] (regular) Oblique stem: o̜fo̜lăf- [ɔfɔləf-]

In both the singular and the plural the same basic change is behind the mild irregularities, namely, the unstressed e̜/o̜ of the nominative becoming ă (/ə/).  The accusative is the result of several sound changes that followed that reduction.  First the loss of /ə/ creating the medial cluster -lf-, then metathesis of the ɔl sequence to lɔ, also a regular change as non-word-final coda liquids were lost, with either metathesis or an epenthetic /ɪ/ after the /l/ being used to repair the coda /l/.  The modern /h/ is derived from earlier /χ/.  Before another consonant, /χ/ was not lost, but instead became /q/.  Synchronically, this appears as a regular alternation of h and q, thus /χɔlf/ → /χlɔf/ → /qlɔf/

Likewise, in the plural, the o̜fo̜lăf- found in the ergative became o̜fo̜lf, and then o̜flo̜f, in both the nomative and accusative