Porphyrogenitos' scratchpad and intermittent glossolalia

[DesEsseintes]

ajidá lavyéki ‘the old city’
ajidá e savyéki ‘the city is old’
anachidá ‘to some cities’

mbáza mvyéki ‘an old sparrow’
anopáza novyéki ‘to some old sparrows’
opáza so novyéki ‘the sparrows are old’

evyéki ‘the old one’
dumvyéki ‘of an old one’
anovyéki ‘to some old ones’

etené savyéki ‘she becomes old’
sémbya sevyéki 'he seems old’

Nice. Latin meets Bantu.

[Porphyrogenitos]

A cross between an artlang and an "oligosynthetic" engelang.

Phonology

/m n ŋ/ <m n ng>
/p t ts tʃ k/ <p t ts ch k>
/b d dz dʒ g/ <b d dz j g>
/f θ s ʃ x/ <f th s sh kh>
/v ð z ʒ ɣ/ <v dh z zh gh>
/l ɾ w j/ <l r w y>

/a e i o u ə/ <a e i o u ë>
/ai̯ au̯ eu̯ oi̯/ <ai au eu oi>

Stress is phrase-final. On all but the stressed syllable, vowels undergo the following reduction chain:

a > ə
e o > i u
oi̯ eu̯ > i u
ai̯ au̯ > e o

This is not shown in the orthography, however.

The "fun" thing about this language is that all roots are strictly (C)V, meaning there are only 270 possible phonemically-distinct roots. The goal isn't necessarily to have an oligosynthetic language, as there will be homophony, but I will be trying to get by with relatively few roots.

There is one possible coda, an archiphonemic nasal /N/ that assimilates in place to following consonant and is realized as [ŋ] before a pause. It is spelled <n>. Its role will be discussed below.

Pronouns

There are only three basic personal pronouns, one for each person.

1: la
2: mo
3: si

Each can be optionally reduplicated to specify plural number, but ordinarily number is left to context.

Morphosyntax

Phonologically speaking, there are only two real morphological operations: reduplication and suffixation of /N/. However, these each instantiate different morphological operations depending on word class.

Reduplication marks a causative on verbs. Thus, to 'die', when reduplicated, becomes toto [tuˈto] 'kill'. On nouns, as in pronouns, reduplication serves to optionally specify plural number, though it may often be more interpretable as 'a group of X' rather than a simple grammaticalized plural.

On transitive verbs, -n marks both a passive and a reflexive, semantically similar to the Spanish se reflexive. In many cases, such as with animates, its interpretation is ambiguous between the two. E.g. mo toton [mutuˈtoŋ] could mean 'you kill yourself' or 'you are killed', somewhat similar in ambiguity to the English you got yourself killed. Inanimate subjects generally preclude a reflexive interpretation, e.g. tse raran 'the rock was broken' , while imperatives generally preclude a passive interpretation e.g. toton! 'kill yourself!'

On pronouns and nouns, -n marks a genitive, possessive, or associative, e.g. lan va 'my/our tongue', which is also the name of the language, with va also being 'language'.

[Porphyrogenitos]

An amusing system of initial mutations. Ultimately, in the modern variety, only non-nasal obstruents show mutations, though /m/ and /n/ historically had mutations, so their development is shown, too.

There is the unmutated grade, a mutated grade that originated as a geminated grade (triggered by complete assimilation of some preceding segment), and a nasal grade. The modern mutation system:

Unmutated / Mutated / Nasal

Code: Select all

ɸ	p	ɸ
θ	t	θ
x	k	x
p	ɸ	m
t	θ	n
k	x	ŋ

The historical development:

Stage 1

Unmutated / Mutated / Nasal

Code: Select all

p	pp	mp
t	tt	nt
k	kk	ŋk
b	bb	mb
d	dd	nd
g	gg	ŋg
m	mm	mm
n	nn	nn

Stage 2

Unmutated / Mutated / Nasal

Code: Select all

ɸ	pp	mb
θ	tt	nd
x	kk	ŋg
p	bb	mm
t	dd	nn
k	gg	ŋŋ
m	mm	mm
n	nn	nn

Stage 3

Unmutated / Mutated / Nasal

Code: Select all

ɸ	p	b
θ	t	d
x	k	g
p	b	m
t	d	n
k	g	ŋ
m	m	m
n	n	n

Stage 4

Unmutated / Mutated / Nasal

Code: Select all

ɸ	p	β
θ	t	ð
x	k	ɣ
p	β	m
t	ð	n
k	ɣ	ŋ
m	m	m
n	n	n

Stage 5 (modern)

Unmutated / Mutated / Nasal

Code: Select all

ɸ	p	ɸ
θ	t	θ
x	k	x
p	ɸ	m
t	θ	n
k	x	ŋ
m	m	m
n	n	n

Some varieties show distinct reflexes for the mutated grade of historic /b d g/ and for the nasal grade of historic /p t k/.

[Porphyrogenitos]

A system of allocutive agreement - "a morphological feature in which the gender of an addressee is marked overtly in an utterance using fully grammaticalized markers even if the addressee is not referred to in the utterance".

This would be part of a language in which every utterance would explicitly encode the sex of the speaker, addressee, and referent (in some manner or another). This system does most of the work with the "addressee" part. I have not decided on many details of the language, but the forms are essentially arbitrary - I am envisioning it in a kind of Afroasiatic context, so I will suppose that the forms -s and -t and derived from Arabic sayyid (or another Semitic cognate) and the Afroasiatic feminine -t (perhaps reduced from a feminine version of the former).

In this language, every word in the language is explicitly marked for sex/gender with a final unstressed vowel - /ə/ for masculines, /i/ for feminines. Stress is uniformly penultimate. All verbs (at least every finite/full verb) take one of three allocutive forms that agree with the sex of the addressee, which also have an honorific function for some speaker-addressee dyads. The historic source of these forms were, in fact, honorific particles, and in previous stages of the language they did, indeed, have more of an honorific function. The language also has further systems of honorifics encoded into the pronominal and verbal agreement systems, as well as various honorific particles (all of which are also gendered).

The three allocutive forms are:

-Ø null
-s masculine
-t feminine


They are suffixed after the final unstressed vowel of the verb.

I will explain the function of each form in each possible male/female dyad.

Male speaker, male addressee

Between two male speakers, the allocutive suffixes retain a significant honorific function.

-Ø

This is the intimate form. The null form is used for address between brothers and close male relatives of roughly equal age. Depending on the situation and the circumstance of the individual relationship, it may be used in address between a father and son. It is used between close male friends, and between male acquaintances in highly casual contexts (compare a male English speaker addressing a male stranger or acquaintance as "buddy"). Moving from the -s form to the null form is often considered (whether explicitly remarked upon or not) a special moment in a male friendship.

Until the early 20th century, was used by superiors to address inferiors. Modern usage requires "equal" use of allocutive forms.

-s

The honorific form for men. Used to address respected male relatives and male strangers and coworkers. Used in professional and formal contexts, e.g. in a court of law all or a legislative proceeding males address each other with the -s allocutive form.

-t

Would normally be understood as ungrammatical. If used intentionally, may be seen as an attempt to mock or emasculate. In the gay subculture, may be used by men to address each other, often with a sarcastic tone. (Compare English speaking gay men referring to a man as "Miss So-and-so")


Male speaker, female addressee

-Ø

Does not normally appear in speech. If a male second-language learner used this form to address a woman, it would elicit a correction and an awkward or uncomfortable response (at best). If used intentionally in direct address, it is considered extremely degrading and offensive, perhaps comparable to calling a woman "cunt" or "cow" in English. It carries a desexing and dehumanizing connotation that goes above and beyond most gendered insults - even when calling a woman a whore, or catcalling a woman, men use the feminine -t form.

-s

Ungrammatical; though very rarely might be used to depict a woman as domineering or mannish"; e.g. in (fabricated) quoted speech, depicting a henpecked man as addressing his wife with -s forms.

-t

The essentially obligatory feminine form used in address by men to women, save for the uncommon exceptions outlined above.

Though having passed out of living memory long ago, historical documents show that the -t form was originally honorific in nature, and was used by men to address respected female relatives and high-status women, with the null form being used for intimates.


Female speaker, male addressee

-Ø

Not used to address adult men. May be used to address male infants or very young boys. Use for adult men would simply be perceived as ungrammatical.

-s

Normal, essentially grammatically obligatory use for females addressing males.

-t

Ungrammatical, unless a woman is attempting to insult and emasculate a man by addressing him as if he were a woman. May be used in the gay subculture by females to address close gay male friends, compare to an English-speaking fag hag addressing her friend as "girl", or, ahem, "gurl". Like this English usage, it may be considered cringey and unacceptable.


Female speaker, female addressee

-Ø

May be used by female intimates; perceived as extremely casual and unladylike; "improper" and considered prescriptively ungrammatical. May be used to emphasize "lower class", "street-smart", "badass" identity. Most women rarely or never use this form to address another woman; some women wouldn't be caught dead addressing another woman this way and would be quite offended if addressed in this manner. Would sound laughable and incongruous coming from the mouth of e.g. an elderly matron or a princess.

However, some elderly rural women recall that when growing up, it was normal for young women to address each other with the null form, and would switch to the -t form upon marriage.

-s

Normally ungrammatical and essentially unheard of; though may be used in the lesbian subculture to address "butch" or "manly" women.

-t

The normal, essentially obligatory form for a female addressing another female.

As recently in the 19th century, however, in some dialects the -t form was restricted to respected female relatives and high-status women.


Group address

The masculine form, -s, is always used in address to mixed-sex groups. -t forms are only used if the group is all female. For a male group of mixed status/intimacy, men use the -s form. On an occasion when a man is addressing a group of intimates, the null form may be used.

Undirected speech

The polite masculine form -s is used by default in newspapers, academic and government writing, TV broadcasts, signage and public announcements, etc, unless women are assumed to be the primary or sole readers/audience (e.g. a women's magazine or column).

In literary writing, the -s form is most common, but writers have extensive freedom in this matter. A writer may e.g. use the null form to give the impression of a male narrator who is close friends with the (presumed male) reader; the writer may also use the -t form to give the impression that the narrator assumes the reader to be female. Speakers have extensive metalinguistic competence in the various forms; a woman will not be offended if she picks up a book and reads a null form. Speakers may play with the forms and use them to disguise their own identity and that of their addressee; e.g. the classic conceit of a woman who writes to her male lover using the -t form to make it look like she is writing to a close female friend to anyone who happens upon the letters; or a closeted gay man who emails or speaks on the phone with his boyfriend using -t forms to make it appear that he is straight and addressing a girlfriend.

[Porphyrogenitos]

Speculating about an IE lang that underwent significant collapse/reshuffling of phonemic contrasts, inspired by the defective encoding of Mycenaean Greek.

I'm thinking:

• centum, with secondary palatalization, like Greek

• all three PIE stop series collapse into one, like Tocharian (apparently?)

• PIE /d/ and /dh/ reflected as new /l/, implying that plain voiced and voiced aspirates first merged, then /d/ > /l/, then voiced/voiceless merger

• /r/ and original PIE /l/ merge, after a RUKI-type change producing a /ʃ/

• maybe original /s/ > /h/ in many positions, like Greek, followed by /ʃ/ > /s/?

• loss of final stops

• simplification of word-initial and word-medial consonant clusters by deletion and epenthesis - syllable structure not quite reduced to (C)V(C)

• typical laryngeal outcomes, idk

• I really can never remember anything about IE verbal morphology so I don't know what happens there

PIE *gʷḗn might end up as something pleasingly irregular, like nom. s. kʷéna (or kʷénas?), acc. s. kʷénan, gen. s. nās (< kʷnās) - or perhaps vocalized to kunā́s.

Assuming both *dyḗws and *deywós have reflexes, like in Latin, they would be something like jōs (with monopthongization) and lewós.

[Porphyrogenitos]

I was revisiting an idea about a conlang with a very small, Polynesian-style phoneme inventory and Polynesian-style CV syllables, but with numerous CV bound roots (with significant homophony, often 3+ meanings paired to a (C)V unit) that combine to form all free lexical items, which uniformly take the shape CVCV (save for loans and a few exceptional items) and can form compound lexical items - though only a maximum of two, meaning a maximal word has the form CVCVCVCV.

Thus you could have things like:

ka- 'away from'
te- 'towards'
-ni 'locomote by foot'
-la 'motion of a round object'

kani 'walk away'
kala 'roll away'
teni 'walk towards'
tela 'roll towards'

The new thing I've been thinking about introducing is Rotuman-style metathesis, e.g. in a certain position kani → kain, perhaps → kɛn.

Thus in certain positions the structure of lexical items is further obscured, and in some cases kind of looks like a form of ablaut:

kɛn 'walk away'
kal 'roll away'
tejn 'walk towards'
tɛl 'roll towards'

What I'm divided on is exactly where and how the metathesis should happen. My understanding is that in Rotuman it's partially syntactic/prosodic, and partially morphological.

For this language, I kind of want to make it purely syntactic/prosodic, so that it's essentially predictable and doesn't need to be represented in the orthography. I'm thinking of a syllabic orthography. If it was represented in the orthography, I was thinking it could be represented simply by the addition of some kind of mark to the two symbols representing a metathesized CVCV segment.

But it would also be interesting if it played a morphological or morphosyntactic role, too - in fact, I assume that later in the history of the language it would be grammaticalized in this manner.

For now, assuming only a syntactic/prosodic rule, I'm thinking this is how the distribution will work:

• Metathesis effects bimoraic units, which have the form CVCV.

• Bimoraic units cannot occur across word boundaries.

• Bimoraic units are formed from left to right within a word.

• Therefore, particles (which have the form CV) are never part of a bimoraic unit and are never affected by metathesis.

• Metathesis occurs to all but the last bimoraic unit in an utterance.

• Thus, if someone utters a lone metathesized form in a non-vocative, non-quotative context, it is interpreted as an incomplete statement.

• E.g. kani 'he walks away' (a complete utterance) vs. kɛn 'he walks away, ...' - the latter utterance might prompt the listener to ask what the speaker was going to say.

• Vocative forms are always metathesized.

• Lexical items pronounced with list intonation are not metathesized.

• Speakers are capable to distinguishing between metathesized and non-metathesized form in quoted speech, but a word quoted without the context of speech (i.e. a citation form) is, as with list intonation, always unmetathesized.

[Creyeditor]

That looks interesting. Did you know that some linguists entertain(ed) for some Austronesian natlangs? I think I once came across a book that tried such an analysis for Indonesian. I will let you know if I re-find it. As for the metathesis, I think it looks good the way it is. If you want to you could add some phonological conditioning, e.g. certain vowel combinations are not allowed and metathesis is blocked in such contexts.

[Davush]

The new thing I've been thinking about introducing is Rotuman-style metathesis, e.g. in a certain position kani → kain, perhaps → kɛn.

...

What I'm divided on is exactly where and how the metathesis should happen. My understanding is that in Rotuman it's partially syntactic/prosodic, and partially morphological.

What a coincidence! I was just reading a paper on Rotuman metathesis the other day, and thought it provided good inspiration for Hakuan, although I am planning on it being a grammatical, not only prosodic. I liked your metathesis-gone-even-further with the ablaut kani ~ kɛn...perhaps something similar will happen in a Hakuan descendant if metathesis does become a prominent feature.

[Porphyrogenitos]

Thanks for the comments. I think I'll use this lang for lexember, meaning I'll have to commit to at least a tentative phoneme inventory.

I was originally thinking of going with something truly extremely small, like ~30 possible syllables, but I suppose I'll allow something slightly larger, in the range of a Polynesian language with a very attenuated inventory, like ~50 possible syllables, not counting vowel length.

I think I will give it four vowels, /i e a u/, with vowel combinations resulting from metathesis resolving in the following manner (subject, probably, to dialectal variation):

Code: Select all

  a       e       i       u
a a       ɛ <ä>   e       o
e ja <ia> e       i       ø <ö>
i ja <ia> je <ie> i       y <ü>
u wa <ua> we <ue> wi <ui> u

And for the consonant inventory...hmm. It kind of depends on how I want to handle hiatus. Hmm. Well, I suppose I will basically side-step the issue and simply inset a glottal stop before any would-be vowel-initial syllable. Although one could still say that glottal stop is non-phonemic because of the following properties:

• It appears at the beginning of stressed vowel-initial syllables (in lexical words and emphatically stressed particles)

• It does not appear at the beginning of unstressed syllables (i.e. particles), or at least only optionally

• When a vowel-initial syllable undergoes coalescence with a preceding vowel, glottal stop does not appear in the coda of the new syllable (i.e. it does not undergo metathesis, e.g. [kati] > [ket] but [kaʔi] > [keː])

In unstressed vowels, /i/ and /e/ are neutralized, with /a e i u/ being realized as [ə ɪ ɪ ʊ]. Vowel in open syllables that have undergone coalescence (like [kaʔi] > [keː] above) are always long.

Okay, for the consonants:
/m n/
/p t k (ʔ)/
/v s x/
/l/

Hm, okay, that gives 44 syllables. Good enough.

Lastly:

/n t s l/ may have the allophones [nʲ tʲ sʲ lʲ] before [ja je], sometimes with the [j] being absorbed completely. For some speakers these allophones may even be realized as [ɲ tʃ ʃ ʎ]. In this context /k x/ may also be realized as [c ç]. An alveolar in a metathesized syllable preceding a palatalized alveolar undergoes assimilation and also becomes paltalized. (As in the name of the language, hasunite [hʷoʃɲet])

/k x/ are realized as [kʷ xʷ] before rounded segments.

[x] is in free variation with [h] (and [hʷ] in rounded contexts).

[Porphyrogenitos]

Wanted to make a separate thread for this, but I don't have a name for it or really any of the morphology or morphosyntax pinned down...not even the phonology. But I am certain of a few phonological processes.

I thought about applying these to a Germanic or Slavic lang, but that's too limiting...I kind of want it to be set around Pomerania, since I want it to have German and Polish influences, maybe Russian or Nordic too, but that's a rather implausible location for a remnant pre-IE lang...oh well, it's all made up anyways. Kind of thinking the endonym would essentially be a calque of Slavic po more (Pomrania, Pomors, etc).

Not really sure about the whole consonant inventory - probably not going to go the Polish route with the three-way contrast in sibilants. Maybe just this:

/m n ɲ ŋ/ (not sure about the palatal or velar)
/p t ts tʃ k/
/b d dz dʒ/
/s ʃ x/
/v z ʒ ɣ/ (not sure about the voiced fricatives other than /v/)
/w j l r/

Vowels:

/i ɨ u/ i u ů
/e ə o/ e o o̊
/æ ɒ/ a å
/ẽ õ/ ę ǫ

What's up with the rings? Well, it has to do with the sound change that inspired this whole thing.

The Great L-Vocalization

Originally, the language had two consonants /w l/ represented w l, and five vowels /i e a o u/, represented as in IPA. /w/ alternated with /u/ in certain contexts; e.g. a prefix u- before consonants and w- before vowels.

Around the same time, /w/ became /v/, and the back rounded vowels began to centralize and unround, /u o/ → /ɨ ə/.

However, also around the same time, /l/ began to vocalize to /w/. (Obvious Polish inspiration, of course.) This happened differently in different contexts:

Coda /l/ always vocalized.
Intervocalic /l/ vocalized if adjacent to a rounded vowel; this probably varied a bit by dialect.
Onset /l/ (including in Cl clusters) vocalized if followed by a rounded vowel.

What happened is that /o/ and /u/ before coda /l/ were preserved from fronting and derounding; similarly, /a/ before coda /l/ became /ɒ/ with remaining /a/ fronting to /æ/. Coda l → w eventually coalesced into these rounded vowels completely; they were marked with a ring above.

I love the ring diacritic because it's perfectly iconic and is the obvious outcome of abbreviating the digraphs uo oo ao (or ou oo oa). I also decided that vocalized l → w would be represented orthographically as l̊ (yes, l with a ring above). Unfortunately the composed characters ů o̊ l̊ have some display issues.

Unvocalized /l/ and new /w/ eventually became phonemic due to analogical leveling and loanwords. I am thinking that the great L-vocalization took place around 1500, such that the oldest layer of Latin loanwords was affected, but newer loans from the Renaissance and scientific revolution onwards were systematically loaned in with regular /l/. E.g. label̊ /labew/ 'basin, sink' ← Latin labellum vs. albedo 'albedo' ← Latin albedo

This also means that the historic /w/ ~ /u/ alternation is now a /v/ ~ /ɨ/ alternation. It also means that historic suffixal -l now produces alternations with central and back vowels, e.g. historic apo/apol is now apo/apo̊. Oh, and modern /v/ assimilates in voicing to a preceding or following consonant, thus e.g. kw is [kf].

Nasal vowels

Nasal vowels have existed for a long time. Very early on, like probably before the year 1200, they were Vn clusters. Eventually, all short nasal vowels merged and all long nasal vowels merged, respectively, producing a single pair of nasal vowels distinguished by length alone. Later, around 1400-1600, the long nasal vowel became /õ/ ǫ and the short nasal vowel became /ẽ/ ę. (Again, directly cribbing from Polish!) Thus, there are some old, fossilized alternations between oral and nasal vowels in some paradigms.

Metaphony

Calling it metaphony, not umlaut, since it resembles the height-based metaphony in some Romance varieties. Possibly reconstructible to the proto-language, there was an alternation in which a high vowel in a following syllable caused a preceding /e o a/ to raise to /i u e/. This got a bit screwed up by the development of nasal vowels, the loss of certain final vowels, and possibly somewhat by l-vocalization. It also wasn't applied perfectly consistently in loanwords. Today, historic height metaphony is productive in a large class of native words and well-integrated loanwords.

Vowel length

I am not quite sure what to do about vowel length. As noted in the nasal vowels section, there would seem to have been distinct long and short vowels at least historically. I have no problem with having long and short vowels and having them alternate, but how do I represent them? I would prefer if the rings and the ogoneks remained the only diacritics in the language. That leaves me with vowel doubling - which is okay for the plain vowels, though it isn't really the aesthetic I want? But it's really ugly for the vowels with rings, like råål and sko̊o̊d. I mean I guess it's okay. I would also have to decide how that works with digraphs.

Otherwise, that leaves consonant doubling, like ral vs. rall or skod vs skodd. I like that. But...what if I need to double the ringed l? That would look really ugly: rel̊ vs rel̊l̊. And then there is use of a silent final e, like in English, or some combination of all of these, like English or German. I guess I just have to choose.

Other phonological notes

No /f/, at least no native /f/, like Slavic and Baltic. Consistent regressive assimilation of obstruents, and word-final devoicing of obstruents.

Morphosyntax

To be honest, I'm really not sure. I kind of thought, "Screw it, I'm not making a boring compromise between Slavic and SAE" - I recently took a seminar on Iroquoian, so why not do something crazy and give it some Iroquoian-style grammar? Maybe not all the way, but I'm thinking something with a crossover between a masculine/feminine and animate/inanimate distinction (hmm...actually rather like Polish), cislocative and translocative prefixes on verbs, and verbs having three basic aspects - habitual, stative-resultative, and factual - with tense being secondary. Also more noun incorporation than you would expect for a European language. And presumably adjectives and handled by stative verbs.

[Omzinesý]

Morphosyntax

To be honest, I'm really not sure. I kind of thought, "Screw it, I'm not making a boring compromise between Slavic and SAE" - I recently took a seminar on Iroquoian, so why not do something crazy and give it some Iroquoian-style grammar? Maybe not all the way, but I'm thinking something with a crossover between a masculine/feminine and animate/inanimate distinction (hmm...actually rather like Polish), cislocative and translocative prefixes on verbs, and verbs having three basic aspects - habitual, stative-resultative, and factual - with tense being secondary. Also more noun incorporation than you would expect for a European language. And presumably adjectives and handled by stative verbs.

I think Slavic is SAE.

stative-resultative means 'I'm seated' etc. ?
Is factual the aspect without any special aspectual specification?

If there is German influence, what about using <h> to write vowel length? It could even have some etymological reason, at least in some instances.

[eldin raigmore]

I think Slavic is SAE.

Unless I’m wrong, Slavic is a family and SAE is a sprachbund.
There’s certainly some overlap, but Slavic is genetic and SAE is geographic.
Unless I’m wrong.

I know not all SAE is Slavic; and I’ll be surprised to find out all Slavic is SAE.

[Omzinesý]

I think Slavic is SAE.

Unless I’m wrong, Slavic is a family and SAE is a sprachbund.
There’s certainly some overlap, but Slavic is genetic and SAE is geographic.
Unless I’m wrong.

I know not all SAE is Slavic; and I’ll be surprised to find out all Slavic is SAE.

I just meant that Slavic belongs to SAE.
SAE is a sprachbund. It's a matter of definition if sprachbunds should be defined as a list of shared features or a list of languages.

[eldin raigmore]

I just meant that Slavic belongs to SAE.
SAE is a sprachbund. It's a matter of definition if sprachbunds should be defined as a list of shared features or a list of languages.

Is Slavic just one language? I thought Slavic was a family of languages.

[sangi39]

I just meant that Slavic belongs to SAE.
SAE is a sprachbund. It's a matter of definition if sprachbunds should be defined as a list of shared features or a list of languages.

Is Slavic just one language? I thought Slavic was a family of languages.

Just to step in, I'm pretty sure Omzinesý isn't saying "Slavic is one language and belongs in the SAE family" but is instead saying "Slavic is a language family and belongs in the SAE sprachbund". I would have thought that was obvious, so can we drop that point and move on please.

[eldin raigmore]

https://en.wikipedia.org/wiki/Standard_Average_European
OK.
Some of the Slavic languages are core SAE members, and some are peripheral SAE members.
I thought some of those peripheral members were nonmembers.
Looks like I was wrong,

[Porphyrogenitos]

Goodness, I seem to have caused quite a stir. My understanding was that Slavic was part of SAE, but more peripherally - so I just meant "a compromise between typical SAE features and Slavic features not typical of SAE"

stative-resultative means 'I'm seated' etc. ?
Is factual the aspect without any special aspectual specification?

Basically yes, for the state-resultative. I will quote this Oneida/Mohawk/Cayuga grammar resource. In Iroquoian it's actually just called the "stative" but I find "stative-resultative" to be a clearer name:

The stative aspect is most often used to describe a state or condition. The state or condition may be one of two kinds: it may be constant and inherent, as in the first example below, or it may be the result or consequence of an action that has been performed in the past, as in the second example.

O lahnʌːyés
C hahnę́ːyeːs
M rahnénːies

'he’s tall'

O wakkáːtshi
C dewagekáhsǫː
M wakeríhsion

'I have taken it apart [and so now it's taken apart]'

I actually got mixed up on the factual. The factual is one of the modal/tense subcategories of the punctual (basically, the perfective). The other two are the future and the optative. So the factual-punctual is basically like a recent-past indicative. It can also refer to the initiation of certain events, particularly weather ("It started to rain") and emphasizes the beginning of motion verbs ("She just set off in that direction").

Personally, if I implemented a similar system of mood-tense subcategories within the punctual for this language, I would tweak the future slightly so that it had its origin in a volitive, like the English "will" - but still with some sense of intention, like the English "fixin to".

If there is German influence, what about using <h> to write vowel length? It could even have some etymological reason, at least in some instances.

Yes, that would be a good idea.

[Porphyrogenitos]

Something with mixed European, Northeast Asian, and Iroquoian influences. Very tentative phonology; this is one of my ideas where the morphosyntax came first and the sounds are an afterthought

Phoneme inventory

/m n/ m n
/pʰ tʰ tʃʰ kʰ/ p t c k
/p t tʃ k/ b d j g
/s h/ s h
/ɾ w j/ l w y

/i ɨ u/ i ü u
/e (ə) o/ e e o
/a/ a

Phonotactics and allophony

CGVC, where G can be /ɾ w j/
Vowel harmony: A word may not contain both /ɨ/ and /u/
/ɾ/ becomes [l] syllable-finally and after /t/
/ɾ n/ lost before /i j/ (restored in compounds)
Aspirated/unaspirated are neutralized after /s/ and syllable-finally
Word-final stops are unreleased
Intervocalic unaspirated stops (excluding affricate /tʃ/) are lenited to [β ð ɣ] (I may or may not make this phonemic in some manner)

Stress

Stress is root-initial
In a compound (including a noun incorporation construction) stress falls on the first element
Pretonic vowels are fully neutralized to schwa

Case marking

Core arguments (subject/agent and object/patient) are not explicitly marked; they are differentiated through word order, which is SVO

There are four case clitics and a topic clitic:

-(i)da oblique/dative
-(e)re genitive
-(o)p locative
-aker instrumental
-(i)ba topic

The instrumental is the result of the verb hak 'use' compounding with the genitive clitic. Frankly, I just now came up with this on the fly, but it seems like this would be a very productive historical process, due to the nature of relative clauses, which I'll explain in a bit. So there actually might be more 'minor' case clitics that developed this way.

Anyways, they are clitics because they attach to phrases, not words:

memere pwar 'mom's house'
mem gi dedore pwar 'mom and dad's house'

See also the relative clause construction, which uses the genitive clitic:

reyak myatere bowar
go store=GEN woman
'woman who went to the store', lit. 'woman of go to store'

Number and definiteness

There is no number marking and no articles. However, there are various strategies for encoding definiteness and quantification.

Firstly, there is the topic clitic. Topic-marked items are fronted.

bowariba ki yok ipa pwar
woman=TOP 1.exc see 3.GEN house
'As for that woman, I saw her house'

There exist demonstratives, but they are used in a narrowly spatial/pointing sense, not to keep track of referents in discourse. Constructions such as "the aforementioned..." do exist in writing and formal speech, though:

kar nebwikere bowar
AUX.PASS mention=GEN woman
'the mentioned woman', lit. 'the woman who was mentioned' or 'woman of being mentioned'

Noun incorporation plays an important role in encoding the specificity of referents. An incorporated object always has a generic sense, as opposed to an unincorporated object, which may be specific:

ki syam tu
1.exc watch bird
'I watched a/the bird'

ki tusyam
1.exc bird-watch
'I watch birds', lit. 'I birdwatch'

There is a distinction between mass and count nouns. Mass nouns used without a classifier have a generic sense, and must usually be incorporated into a verb; in certain contexts they may also have the sense 'a variety of X', cf. 'There's a beer I really like'.

However, used with the genitive clitic they have a partitive sense:

ki jan gahwere
1.exc drink coffee=GEN
'I'm drinking some coffee', lit. 'I'm drinking of coffee'

With a classifier:

ki jan duk gahwere
1.exc drink cup coffee=GEN
'I'm drinking a cup of coffee'

Incorporated:

ki gahwejan
1.exc coffee-drink
'I drink coffee [in general]'

Motion and location

There are six prefixes that encode the spatial properties of a verb:

se- cislocative
re- translocative
te- illative
ge- ellative
we- subessive
gre- superessive

(Since they occur before the primary stress, all have the reduced vowel [ə])

The semantics of all the prefixes varies somewhat arbitrarily with the verb root, but in general, the latter four are what you would expect based on their names. The cislocative and translocative are the most widely-used and have the most varied semantics. In general, the cislocative indicates motion or location near or coming towards the speaker (or another referent), and the translocative indicates motion or location at a distance, far from or going away from the speaker (or another referent).

The word translated 'go' in 'woman who went to the store', reyak, is composed of re- 'translocative' + the root yak 'go, move'.

In sentences with motion verbs, the goal does not ordinarily have to be marked with the locative clitic:

ki reyak myat
1.exc go store
'I went to the store'

The locative clitic is more typically used with verbs that do not inherently encode motion.

ki jan gahwere pwarop
1.exc drink coffee=GEN house=LOC
'I drank some coffee in the house'

The locative clitic generally may be used with the arguments of motion verbs, in which case it typically modifies the meaning, either in a significant way, or in a very subtle way, cf. "We crossed the river vs. "We crossed over the river"

Okay, I'm taking a break, and might come back to explain the following in more detail:

• Passives - two auxiliaries, one indicating volition and other indicating lack of volition

• Interaction of passives and noun incorporation

• Interaction of verbs of transfer and noun incorporation

• Relative clauses - only subjects may be relativized

• Other verbal prefixes - dualic/distributive and repetitive

• Noun affixes - diminutive, augmentative, honorific

• Zero-derivation of verbal nouns and agent/instrument nouns from verbs

• Verb roots as a closed class - new verbs derived through noun incorporation

• Pronouns - unsure how they're going to work, thinking there might just be a 1exc/1incl distinction in the first person with no number

[Porphyrogenitos]

Verbal nouns

I will briefly introduce verbal nouns before moving on to passives, since passives diachronically originated as constructions involving verbal nouns.

Any verb can be zero-derived to a verbal noun. Verbal nouns behave the same as mass nouns - if not noun-incorporated into another verb, they must usually take a classifier or be used with the partitive genitive. Although, unlike other mass nouns, the interpretation "a variety of X" is not normally available (e.g. "This isn't a good coffee.")

ki kan
1.exc run
'I run'

ki kantlik
1.exc run-like
'I like running'

ki orop nar kanere
1.exc finish round run=GEN
'I finished a round of running'

Verbs that themselves were the product of noun incorporation can also be turned into a verbal noun and noun-incorporated:

ki tusyamtlik
1.exc bird-watch-like
'I like birdwatching'

Passives

Passives diachronically originated from a construction such as "She received a hit of kicking", although the classifier has now been elided and it is just an auxiliary + the verbal noun.

bowar tanke kida
woman kick man
'The woman kicked the man'

There are two passive auxiliaries: mya, which implies lack of volition and/or some kind of tangible effect on the patient, and kar, which implies volition and/or lack of tangible effect on the patient. They roughly correspond to the English be-passive and get-passive. Mya originates from and is identical with the word mya 'eat', kar originates from and is identical with the word kar 'take, receive'. However, unlike in English, the auxiliary implying some level of volition is default and has more general semantics.

The agent may be dropped; if included, it is encoded with the oblique/dative case clitic.

kida mya tanke bowarida
man PASS.INVOL kick woman=OBL
'The man was kicked by the woman'

kida kar tanke bowarida
man PASS.VOL kick woman=OBL
'The man got kicked by the woman'