More autoconlanging, this time: A Conlang Like No Other

[vec]

From the creator of the avrelang, the most average language ever. I want to challenge myself to create a language unlike all others I've created. A language featuring as few features that I have up to this point used as I can. Again, I'm going to be using WALS.INFO for reference but this time not going systematically through it but rather, by three criteria: 1) have I used this feature before? 2) am I interested in this feature?, and 3) does it make sense with all the other features?

The languages being considered as DO NOTs are Uscaniv, Kiassan turasta, Imutan, Lomanin, Kìn-Sang and the aforementioned Avrelang. In addition, I'm considering Proto-Ronquian but since it's still under wraps I won't discuss it.

I'm not going to do shit that I won't enjoy here. Dealing with clicks is one thing I know I won't enjoy, so it's off the list, even though I've never done it before. There's a reason I haven't. I'm sure there are other such features out there but I'm pretty open-minded overall.

So, without further ado, let's start.

—————

Feature #1 is number of consonants. Uscaniv has 13 consonants, Kiassan Turasta has 11, Kìn-Sang has 19, Avrelang has 22, Lomanin has 24 Imutan has 29. I have a bunch of sketches of languages with very small consonantal inventories, but it looks like I generally shy away from many consonantal phonemes. I'm going to go with 34 or so consonant phonemes, the bottom limit of the large category because we don't want this to get TOO crazy.

Feature #2 is vowel quality inventories. Uscaniv has 4, Kìn-Sang has 4, Kiassan Turasta has 5, Avrelang has 5, Lomanin has 6 and Imutan has 8. They are all in the average range (5-6), except Uscaniv which has a small inventory (2-4) and Imuthan which has a large inventory (7-14). I should mention that Apalin and Aircin, relatives of Uscaniv, both have 3, so it seems like small is out. This language is going to also have lots of vowel qualities, at least 9 or more.

For purposes of this project I'm going to ignore feature #3, consonant and vowel ratios.

#4 Voicing in plosives and fricatives. Now, the options here are No voicing contrast (which applies Uscaniv, Kìn-Sang and Kiassan Turasta), Voicing contrast in plosives alone (which applies to Imutan and Avrelang), Voicing contrast in both plosives and fricatives (Lomanin) and Voicing contrast in fricatives alone (None! Probably because that's super uncommon and weird!). So I'll be going with that.

As for gaps in the plosive system, I've never done both /p/ and /g/ missing but it doesn't look like this language will have any /g/ anyway. So we'll see.

In the past, I have had uvulars and I've had not uvulars, but I've never had uvular continuants only. So expect some gargling.

As for glottalized consonants, I've done ejectives, implosives, but I've never done glottalized resonants. I hope you've got you're Listerine ready because we're about to gargle a lot together.

When it comes to laterals I've done it all. Because I fucking love laterals. The only thing I haven't done with laterals is not have laterals. So I won't have laterals.

Same with the velar nasal, except there I've also tried not having it. This one I'll figure out once the entire inventory starts shaping up, but my feeling is, there will be some dorsal nasal, if only because I'm going to be needing a lot of consonants to reach my target number.

When it comes to nasal vowels, there are only two options: phonemic nasal vowel present – or not. I've done both and since I'm not the hugest fan, I'll leave them out. I'll have enough vowels on my hands as it is already.

Front rounded vowels, chapter 11. Interestingly, with me being Germanic and all, I've not really ever used front rounded vowels much. Lomanin has /y/ and Kìn-Sang has some as allophones but that's it. So I'll be doing some of these, which is good for the target vowel inventory.

Syllable structure: I've done simple, moderately complex and complex, which are the only options. I've never actually done truly simple, exclusively (C)V, so I'm going to keep that option open, but this one is going to get decided on aesthetics.

Tones: I've only done simple tone systems, mostly because complex tone systems scare the crap out of me. I think I'm going to leave tone be this time, even though I probably should have it but I'll make a final decision later.

The next four chapters are about stress, and some of them are not applicable to all languages. Basically, I've done it all. Imutan has second-syllable stress, which is already weird. So I'll see what happens with this one. Another aesthetic concern that I'm pushing till later.

The next one is Absence of common consonants. I have skipped bilabial stops or had gaps in the bilabial system but I love /m/ so I've always had that. So I'll try to get by without it and skip bilabials entirely.

Finally, for phonology, we have the flip-side Presence of uncommon consonants. I've done nearly everything here except commiting to pharyngeals. So I'm going to have some pharyngeals. Listerine, go!

—————

In the next post, I'll be putting together a phonology based on the above findings.

[Zaarin]

it looks like all languages that have no nasals also have very small consonant inventories.

A number of languages of the Pacific Northwest (known for their large inventories) have no nasals, including Ditidaht (formerly Nitinaht), Quileute, and Eyak.

[vec]

To summarize the typological specs I have set for myself:

• 34 or more consonants
• 9 or more vowels
• Plosives: no voicing, fricatives: voicing
• No /p/ or /g/
• Uvular continuants
• Glotallized resonants
• No laterals
• Front rounded vowels
• Either CV or something else
• No bilabials
• Pharyngeals

So, what to do about plosives? The above specs would give me, say, {tʰ, t · kʰ · ʔˤ} where ʔˤ refers to a pharyngeal stop, if I interpret /g/ as [k]. It's looking a bit unbalanced so I'm going to add a retroflex, something I've never fully committed to in a previous conlang.

Plosives: tʰ, t · ʈʰ, ʈ· kʰ · ʔˤ

I believe I'll be needing some affricates as well, but these will have a voicing distinction rather than aspirates:

Affricates: ts, dz · tʃ, dʒ · ʈʂ, ɗʐ

Nasals are going to be easy: Just match those up but skip the pharyngeals since, you know, there are no pharyngeal nasals (despite how awesome that sounds). But looking at n · ɳ · ŋ, I'm thinking ŋ is looking very improbable there, so I'm going to skip it.

Nasals: n · ɳ

In addition, these will have glottalized versions.

Glottalized nasals: nˤ · ɳˤ

And there will be voiced nasal fricatives, with a surprise appearance by a labio-velo-dental:

Nasal fricatives: ṽˠ · z̃ · ʐ̃

Trills! I can't have a bilabial trill, but I can have a uvular one. (I do have it in one conlang but it's a minor one): {r ɽ ʀ}. What about a voicing distinction? Yes, let's do it!

Trills: r̥, r · ɽ̥, ɽ · ʀ̥, ʀ

Fricatives will have a voicing distinction and of course an unvoiced pharyngeal fricative will make an appearance.

Fricatives: s, z · ʃ, ʒ · ʂ, ʐ · x, ɣ · ħ

As for approximants, I'm going to go boring, because something needs to be boring in this language, and go with:

Approximant: j

In the end there are 35 consonant phonemes, one more than expected.

Apparently, Google Drive Photos can't be embedded on the ZBB but you can see a chart here: https://drive.google.com/file/d/0B3QoiX ... sp=sharing

—————

Now, for vowels, I'm thinking:

i, y · u
e, ø · o
ɛ, œ · ɔ
a

That's 10 vowels qualities. In addition, there's a few diphthongs:

ai, au
eu, øy, oi

Now that I'm looking at it, I'm almost thinking I actually want to add some tones and keep to a strict CV syllable structure. Thoughts?

—————

Next up, what I and the board love the most, respectively: I figure out some morphology, while you guys help me with a romanization! All suggestions welcome!

[vec]

I've decided to keep working in a Google Doc. Anyone is free to have a look at the in-progress grammar:

LINK TO GOOGLE DOC

I'll be posting status reports here.

—————

One thing I do have to report is that I have decided to go with a five tone system (low, mid, high, rising, falling) and a strict (C)V structure. Morphology in the works!

[Nortaneous]

Its not hard to have a large consonant inventory that doesn't get crazy. Of my conlangs, Kannow has 46 consonants, Gehui has 53 (but it has clicks), and Miar has... 34, huh, I thought it had more than that.

Here is a chart:

Code: Select all

tʰ t  ʈʰ ʈ        kʰ    ʔˤ
ts dz tʂ dʐ tʃ dʒ
s  z  ʂ  ʐ  ʃ  ʒ  x ɣ   ħ
   n     ɳ
   nˤ    ɳˤ
ṽˠ z̃     ʐ̃
r̥  r  ɽ̥  ɽ        ʀ̥ ʀ
               j

Now, about that consonant inventory...

I'm not sure if there are any natlangs that treat plosives and affricates differently in that way. The velar nasal gap looks odd but it's probably fine if CV. And I'm not sure about the nasalized fricatives. Uvular trills are vastly overplayed in conlangs compared to natlangs but w/e.

Here's what I'd do given these constraints:

Code: Select all

 tʰ  t  tsʰ  ts  ʈʰ  ʈ  c  ɟ  kʰ  k
ⁿtʰ ⁿd ⁿtsʰ ⁿdz ⁿʈʰ ⁿɖ ⁿc ⁿɟ ⁿkʰ 
         s    z  ʂ   ʐ  ɕ     x   ɣ ħ ʕ h
	        n      ɳ
		  nˀ     ɳˀ
              r      ɽ     j          ʀ
              rˀ     ɽˀ    jˀ

And maybe some labiovelars, prenasalized fricatives and trills, and /v/.

Anyway:

/tʰ t ʈʰ ʈ kʰ ʔˤ/ <t d tr dr k x>
/ts dz tʂ dʐ tʃ dʒ/ <ts dz chr jr ch j>
/s z ʂ ʐ ʃ ʒ x ɣ ħ/ <s z shr zhr sh zh h gh xh>
/n ɳ/ <n nr>
/nˤ ɳˤ/ <nx nrx>
/ṽˠ z̃ ʐ̃/ <nv nz nzr>
/r̥ r ɽ̥ ɽ ʀ̥ ʀ/ <rrh rr rh rr kr gr>
/j/ <y>
/i y u/ <i ü u>
/e ø o/ <e ö o>
/ɛ œ ɔ/ <ê ë ô>
/a/ <a>

[Matrix]

Easy typing romanization:

/tʰ t ʈʰ ʈ kʰ ʔˤ/ <th t thr tr kh q>
/ts dz tʃ dʒ ʈʂ ɗʐ/ <c j ch jh chr jhr>
/n ɳ/ <n nr>
/nˤ ɳˤ/ <nq nrq>
/ṽˠ z̃ ʐ̃/ <v zn znr>
/r̥ r ɽ̥ ɽ ʀ̥ ʀ/ <rh r rrh rr qrh qr>
/s z ʃ ʒ ʂ ʐ x ɣ ħ/ <s z sh zh sr zr k g h>
/j/ <y>

/i y u/ <i iu u>
/e ø o/ <e eo o>
/ɛ œ ɔ/ <eh eoh oh>
/a/ <a>

/ai au/ <ai au>
/eu øy oi/ <eu eoiu oi>

/ɽ̥øy ʐ̃a tu ʒi dʒa ɳa kʰa ṽˠœ nˤo/
Rrheoiu znra tu zhi jha nra kha veoh nqo.


Diacritirrhea:

/tʰ t ʈʰ ʈ kʰ ʔˤ/ <ṫ t ẗ ṭ ḳ q>
/ts dz tʃ dʒ ʈʂ ɗʐ/ <c ȷ ċ j č ǰ>
/n ɳ/ <n ṇ>
/nˤ ɳˤ/ <ň ṋ>
/ṽˠ z̃ ʐ̃/ <v ƶ ẕ>
/r̥ r ɽ̥ ɽ ʀ̥ ʀ/ <ṟ r ṝ ṛ ř ṙ>
/s z ʃ ʒ ʂ ʐ x ɣ ħ/ <s z š ž ṣ ẓ k g h>
/j/ <y>

/i y u/ <i ü u>
/e ø o/ <e ö o>
/ɛ œ ɔ/ <ẹ ộ ọ>
/a/ <a>

/ai au/ <á à>
/eu øy oi/ <è ǒ ó>

/ɽ̥øy ʐ̃a tu ʒi dʒa ɳa kʰa ṽˠœ nˤo/
Ṝǒ ẕa tu ži ja ṇa ḳa vộ ňo.

[vokzhen]

I'm not sure if there are any natlangs that treat plosives and affricates differently in that way.

It's only one example, but I think a strong one: Basque has a voiced-voiceless /p t c k/ distinction in stops, but only voiceless and a laminodental-apicoalveolar-postalveolar contrast in both fricatives and affricates, so barring something weird in the minute details I'd say it's safe to say the affricates actually do pattern as fricatives. With premodern Spanish you've got /tʃ/, but at least for the majority of dialects I'd say now the pairs /tʃ ɟʝ~ʝ/ phonemically pattern like /p b~β/ etc; Buenos Aires might be an exception thanks to its "sheismo." Aragonese, from what little I've been able to find, still has /tʃ/ as its only affricate and no voiced counterpart. I thought I'd run into a couple others that patterned them strongly, like Basque, but I've been unable to find them now - all I can remember is that they were probably Tibeto-Burman, which doesn't narrow things down much.

Most of the other ones I know of where the affricates don't clearly pattern with the stops, they don't really appear to pattern with the fricatives either: things like Rukai (Formosan) that just doesn't have much symmetry /p b t d ɖ k g/ /v θ ð s/ /ts/), or the Northeast Caucasian languages where the stops are usually voiceless-voiced-ejective, fricatives voiceless-voiced, but affricates voiceless-ejective (Avar particular stands out to me, /t d t'/ /s z s:/ /ts ts' ts: ts':/).

EDIT: And if you meant specifically voiceless stops, voiced+voiceless fricatives and affricates, I'd say given the rarity of a) languages with voicing only in fricatives and b) languages that pattern affricates as fricatives, I'd say there's good reason to argue that lack of such a natlang is an accidental gap rather than something deeper.

[opipik]

Now that I'm looking at it, I'm almost thinking I actually want to add some tones and keep to a strict CV syllable structure. Thoughts?

I'm for it.

[gestaltist]

Now that I'm looking at it, I'm almost thinking I actually want to add some tones and keep to a strict CV syllable structure. Thoughts?

I'm for.

Is it a valid English sentence? Shouldn’t it say „I’m for it.“? Honest question. I am not a native speaker.

[KathTheDragon]

Native speakers will understand "I'm for" to mean "I am in favour of this specific thing under consideration".

[gestaltist]

Native speakers will understand "I'm for" to mean "I am in favour of this specific thing under consideration".

Well, I understood it. My question is: is it something native speakers say.

[KathTheDragon]

Oh, right. Probably. I know extending it to "I'm all for" is something I would do, so I'd imagine I'd also say the shorter form.

[gestaltist]

Oh, right. Probably. I know extending it to "I'm all for" is something I would do, so I'd imagine I'd also say the shorter form.

Cool, thanks. Maybe English is on its first baby step to becoming pro-drop?

[vec]

Before I get into nominal morphology, here are some updates on sounds:

I decided to ditch the unvoiced uvular trill for a /ʋ/.

There are five tones, tentative diacritics below:
Low: ă
Mid: a
High: ā
Rising: á
Falling: à

The rising and falling tones are relative to the previous tone, but the tonal level can never go above or below the tone of the low tone and the high tone. Thus, within a word, a falling tone cannot follow a high tone, and a rising tone cannot follow the high tone.

—————

Let's see what the nouns do!

How are nouns inflected? Now that I have tone, I can go with tonal change and concatenation (chapter 20), which I have never really done. Well, Kìn-sang has a few tonal changes occur but mostly due to sandhi. So in this language, I'm going to commit to tonal inflection à la Afrique.

The next chapter is about exponence of certain inflectional formatives. Exponence is when a single morpheme encodes many categories. I've done the entire spectrum of this so I'll make the call later. I've probably done less languages with full-on exponence so I might go with that.

Chapter 22 is Inflectional Synthesis of the Verb and what I gather from that is that I've never done a really inflection heavy language, really. Uscaniv encodes person, tense, aspect, mood, voice and interrogation on the verb, and Imutan encodes finiteness, mood and tense/aspect. Imutan's system is only complex because the non-finite forms all decline like nominals but those are not verbal categories and won't get counted as per the definition in the chapter. I'm going to go with 7+ categories on the verb and really run with it.

Chapter 23 is about whether verbs mark for dependents or not. In this matter I have had subject marking (Uscaniv), subject and object marking (Kiassan Turasta), and dependent marking (no marking on verb, only case on nouns) (Imutan), and no marking. I have never tried subject-object marking in a nom-acc language, nor have I tried object-only marking (which the chapter doesn't really cover), nor have I tried applicatives. So I might do some of those. (WALS pointed me in the direction of two languages Hunzib, a Caucasian language, and Datooga, a Nilotic language from Tanzania that are inspiring to me right now. The latter is a marked nominative language, which I have never done, so that's something I'm considering – might go nice with object-only marking on the verb, hinting at an ergative past).

Chapter 24 continues the head-marking theme which they call locus. This one is about possessives. I have done head-marking and dependent-marking, but never double marking, so I may go with that, although I'll have to see.

Finally, finishing off the locus discussion, it goes into the whole language typology. Is the language consistently head-marking, consistently dependent-marking or something else. Most languages I've done are dependent-marking overall, rather than head-marking, but most have some IE-like mixture of the two. Consistent double-marking is something I've definitely not dealt with before and sounds like an interesting proposition. Dependent-marking is more IE-ish so I'll be steering away from that for sure. Let's go with: mostly head-marking, but double-marking in some instances.

Chapter 26 is about whether the language has more prefixes or suffixes. While Kiassan Turasta is mostly prefixing, especially for required categories, most of my other languages are mostly suffixing. I think this language will be a somewhat balanced mix.

Next is reduplication. Again, Hunzib pops up with partial reduplication, which I have definitely done in the past, as well as full reduplication. And I have done no reduplication. I'll make a final call later depending on what seems right.

Case syncretism is the next subject. Syncretism is when two forms within a paradigm are identical. Since I'm leaning towards a marked-nominative system, this is an interesting chapter for me. Certain minority paradigms could collapse the nominative and the accusative, or mark the accusative more than the nominative. Definitely interesting to keep in mind. The following chapter deals with syncretism in person/number marking on the verb. I've never done a great deal of syncretism. As conlangers, we seem to have a tendency to want to eliminate ambiguity and strive for ideological purity. I'll try to fend those inclinations off.

Next three chapters are on number of genders. I'm so taken with Hunzib I'm going to steal it's system whole-sale. I've only ever done animate-inanimate (Əktoś Duəmeuk, a in-progress daughter of Proto-Ronquian) and masculine-feminine (Lomanin) before, so a Caucasian-style five-gender seems like what's up. I'm going to encode male-female in the system, like Hunzib does, but try to keep it a bit genderfluid around the edges. Finally, to top off the gender discussion is a chapter that seems new, because I've never read it before. It's about semantic vs. formal assignment of gender in the language. Semantic assignment is for example "all nouns pertaining to vegetation belong to class X". Formal assignment is e.g. "all nouns ending in -a are feminine". Lomanin, my only well-developed conlang assigns gender mostly by form (masc. -a or consonant, feminine vowel other than -a or -t). Əktoś Duəmeuk, having an animate-inanimate distinction uses semantic criteria. So I've technically done it all. I'll therefore go with what makes sense and in a five-category system, a mixture of formal and semantic criteria are likely to affect gender assignment.

Encoding of plurality is up. I have had plural suffixes and plural infixes, but I have not done plural prefixes, nor have I done plural by tone nor a plural clitic. Another option is: Morphological plural with no method primary which sounds somewhat appealing – different genders might use different strategies. This aspect of the language seems wide-open for exploration. The following chapter is about whether plurality is optional or required, and whether all nouns receive it or not. I've done a lot of these already so plurality continues to be an open book. Just no fucking suffixes, OK? Jeez. Then it goes into plurals in pronouns. Does the stem change? (I've had that happen in nearly every language). Are the affixes different from the nouns? (I've done both different and same). Is there a different stem for each person but then plurality is indicated with affixing? (Kiassan Turasta) Is there one stem that then receives person/number affixes? (Never done that!) So one stem, different affixes. That seems cool and may interact interestingly with the verbal system. Finally, it asks if there's an associative plural. I've never done an associative plural and perhaps I'll do one using a syntactic device of some sort.

To finish the discussion of nouns are two chapters on articles. I've never had only an indefinite article with unmarked noun phrases being either definite or indefinite. (Apparently, no language obligatorily has unmarked nouns as definite). So that's definitely interesting.

—————

That's it for nouns. Pronouns are next. Hope you guys are enjoying.

[KathTheDragon]

a falling tone cannot follow a high tone

I think you mean a low tone.

[vec]

a falling tone cannot follow a high tone

I think you mean a low tone.

Yes indeed. I've completely rewritten the description of the tones in the Google Doc.

Question – does anyone have a pdf of a grammar of an inflectional strict-CV language lying around? I'm having a tough time getting to a satisfying place with the CV restriction and I need to see some good examples.

[vokzhen]

Just ran across this dissertation of a Guarani dialect/language. Strictly (C)V.

If you can get your hands on a grammar of one of the Southern Qiang "subdialects" (the "subdialects" aren't even mutually intelligible, but the term seems pretty engrained), they're apparently maximally CCV+tone (well, with the complex vowels that seem to be common in the area), but all I've been able to find is this based on one of the Northern subdialects with a much less helpful CCVCC.

[Qwynegold]

I have two suggestions. Skip /y/ since you've had it before and stick to either /ø/ or /œ/. I don't know how likely it is to have both without /y/. You could keep /ø/ only and then add two back-unrounded vowels, or two central vowels. When it comes to plurals, I think plural tone is an interesting option. But so is different plurality marking for different noun classes, so you could have plural tone for one or a few classes.

[Nortaneous]

Lemerig has /œ ø/ without /y/. Vurës comes close. Mwotlap has /i ɪ ɛ a ɞ ɵ ʉ ɔ ʊ/ + /ɛ͡a ɔ͡ɞ ʊ͡ɵ/ -- that's /ʉ/ without /u/, and then /ɞ ɵ/

Wobzi Lavrung doesn't like coda consonants, but it does like initial consonants, so it's not helpful.

[vec]

Icelandic has /ʏ/ and /œ/ but no /y/. When /œ/ is long it's kind of like /øœː/. I think Faroese only has /y/ in loanwords. So there's plenty of precedents probably.

[Kilanie]

This is a cool project; you've inspired me to make my own attempt at it. I'm excited to see your next segment.