Conlang Random Thread

[Nortaneous]

I've increased the restrictions on minor (unstressed) syllables in Pñæk so that in normal words the shape can only be CV and the only vowels are i, a, u, r̩, n̩. The issue is that I have a number of clitic or "particle" forms (e.g. agreement+TAM markers) which don't comply with these rules. The clitics are still reduced (e.g. they don't show phonation distinctions or the more marked clusters), but they do have Cr- onset clusters and other vowels sometimes.

I'm debating whether to change the clitic forms to comply with the minor syllable rules or not. My gut feeling is to leave them, since clitics are in some sense intermediate between affixes and words. But it's quite difficult to find descriptions online of languages with strong unstressed syllable reduction where clitics behave differently. If I look at my dead-tree books, Semelai seems to be an example, since it has some clitics containing /e/ and other vowels which are not allowed in the most reduced syllables in normal words. English strikes me as a possible example too, but it's complicated (see https://en.wikipedia.org/wiki/Stress_an ... ull_vowels).

Can anyone give descriptions of other languages which have strong unstressed syllable reduction, but where the shapes of (unstressed) clitics are more complex than the shapes of normal unstressed syllables?

no but this is fine. everything's synchronic before it's diachronic. sometimes things are synchronic and diachronic at the same time but sometimes they aren't

[Kuchigakatai]

English strikes me as a possible example too, but it's complicated (see https://en.wikipedia.org/wiki/Stress_an ... ull_vowels).

Can anyone give descriptions of other languages which have strong unstressed syllable reduction, but where the shapes of (unstressed) clitics are more complex than the shapes of normal unstressed syllables?

I think English is a good example... Notice the vowels in "they" (expected: diachronic /ðɪ/ or /ði/ with HAPPY vowel), "my" (expected /mə/ which some people use, or diachronic /mɪ mi/ which some people use), "our" (expected /ɚ/), "their" (expected /ðɚ/), "on" (expected /ən/), "to" (by which I mean /tu/ especially before a consonant, although the expected /tə/ also gets plenty of use). It seems like monosyllabic clitic subject and possessive pronouns as well as prepositions are subject to a middle approach between stressed and unstressed syllables, where they're phonetically largely unstressed but still able to retain some unreduced vowels.

Also, I think most of the complications Wikipedia mentions there can be explained... The /ɛ/ of "to document" is part of a greater pattern in English where trisyllabic verbs gain some secondary stress in the last syllable: "to separate" /ˈsɛpəɹeɪt/ vs. "a separate thing" /ˈsɛpɹət/. "Neon", "tofu", "royale", "piano" and "manatee" are recent borrowings, and borrowings in any language tend to bear strange phonotactics, including clear vowels in languages with vowel reduction (and in general, there is a tendency for Greek -on borrowings to have /ɒn/ due to the English traditional pronunciation of Ancient Greek, plus foreign -o as in "piano" is rendered /oʊ/ by convention).

"Outlaw" and also probably "grandma" seem to be interpreted as compounds (grand-ma), and "a discount" is interpreted as related to "count" with the prefix dis- but undergoing a noun-y stress shift. "Monday" and the other days of the week similarly seem to be compounds of [synchronic weird element] + "day", plus, there is plenty of foreign final -e rendered /eɪ/. The pre-tonic /aɪ/ of "idea" is not surprising at least, being common in Latinate/Greek vocabulary (bisexual, dynamic). "Unknown" uses a very interesting strategy of treating un- like some sort of compound modifier, producing a geminate /n:/, all in order to avoid homophony with "a known" (see also: unnoted, unnavigated, unnoteworthy).

This leaves us with "ambition" and "a convert" as truly weird cases. The former might be a spelling pronunciation of sorts and the latter might be influenced by the verb /kənˈvɜ:t/, but neither is a good enough excuse.

[Kuchigakatai]

Anyone looking for sources of [ɬ]? Apparently, Sassarese has [st] > [ɬ:]: Latin stella 'star' > (with gender change) Sassarese [iɬˈɬeɖɖu] 'star' (cf. Sardinian [isˈteɖɖu]).

[chris_notts]

I think English is a good example... Notice the vowels in "they" (expected: diachronic /ðɪ/ or /ði/ with HAPPY vowel), "my" (expected /mə/ which some people use, or diachronic /mɪ mi/ which some people use), "our" (expected /ɚ/), "their" (expected /ðɚ/), "on" (expected /ən/), "to" (by which I mean /tu/ especially before a consonant, although the expected /tə/ also gets plenty of use). It seems like monosyllabic clitic subject and possessive pronouns as well as prepositions are subject to a middle approach between stressed and unstressed syllables, where they're phonetically largely unstressed but still able to retain some unreduced vowels.

Yes, I think I'm going to stick with it. It's not the most marked thing in the language anyway. I feel like the core is finally starting to come together after multiple revisions of the 91 pages of the grammar I've written so far. Although I still have a long list of topics to address.

[Pabappa]

Some Siouan and Pomoan languages have affixes that mark out the means by which an action is performed. These are bound morphemes that, as far as I know, do not resemble the free forms of the objects to which they correspond. Generally, I think the affixes are for body parts and tools. Since Siouan and Pomoan are spoken quite far apart, it's possible the trait appears in many other Native American languages as well.

Poswa has something similar, but I've recently thought about extending it to metaphorical use. These are called U-verbs because historically they all ended in /ə/, /u/, or /ū/. Using body parts and tools metaphorically allows the language to be more expressive. One very useful morpheme is /-dž-/, which means "by use of money". Thus Poswa could be said to have grammaticalized the distinction of whether an action costs money, and shows word pairs such as fobi "i ate" vs fobiebi "i ate out". \

The morpheme /-mp-/ "hip" could serve as a metaphorical way to mention the use of a cellphone, and create words like bompibebi "i tagged you online". There is also /-ntš-/ "pocket" but I prefer /mp/. This could also be used to denote the use of a gun or other firearm, since Poswa lacks any words for such things, and since even the word for "shoot" is highly derived. Two words for hunting are fap- and pamb-; these could become fapamp_b- and panump_b- respectively (yes the stem changes are off-the-charts, I know).

I'm curious to see if any speakers of these Native American languages use words like that already. I suspect not, but I've been surprised before.

[Whimemsz]

Since Siouan and Pomoan are spoken quite far apart, it's possible the trait appears in many other Native American languages as well.

They do, though of course not of the same precise type everywhere, and the Pomoan and Siouan systems are pretty distinct themselves.

[Kuchigakatai]

Halkomelem also has that type of bound morphemes, and they're also often not at all like the regular nouns used for the body parts or tools in question (although sometimes they are).

[chris_notts]

I wanted to do something like the manner affixes and transitivity/control marking of Salish originally, but in the end went in a slightly different direction for Pñæk, since I decided that complex morphology is not my strength. The conlangs I've been happiest with have always been more dominated by syntax than morphology.

It's well described that in a number of serialising languages state change is often expressed by a transitive causation verb + intransitive state change verb. The transitive verb on its own may not encode the state change as part of its meaning, although it may be present as a cancellable implicature. Take Watam, a Papuan language, as an example. There is no generic verb that means "kill", only various multiverb combinations:

rug minik = hit die
rutki minik = slash die
wak minik = sever die
arig minik = shoot die
rutki yak minik = slash cut.open die
mo minik = do die

mo minik seems like the elsewhere case, but a more specific verb is preferred if possible.

I thought it would be interesting to combine extreme SVC productivity with a tendency for transitive verbs to not encode resultant state, and to limit most significant state change to intransitive roots. So e.g. there might be a verb slash or slice at, but to assert that something is cut into pieces without it being just an implication you'd have to say "slash sever", a bit like English phrasal verbs or Slavic prefixes often add telicity and boundedness. This produces a similar effect to manner affixes but in a more productive, expandable form. Of course certain combinations will become conventionalised, but the underlying productivity of the system is still there to form novel combinations.

EDIT: when I say change of state, I mean fundamental transformation of some sort like "break", "cook", "die" etc. I have an entire chapter in the grammar I've been writing describing the semantics of verbs, typical case frames etc.

[Whimemsz]

Halkomelem also has that type of bound morphemes, and they're also often not at all like the regular nouns used for the body parts or tools in question (although sometimes they are).

Are you referring to lexical suffixes? Since those are a much broader set of morphemes with a much broader set of usages, and any that are used in a particular instance to indicate manner/an instrument should really be viewed in those terms. So I wouldn't say the Halkomelem system (or that in other Salishan and Northwest Coast languages) is comparable to that in, say, Siouan, where there's a smaller, coherent class of dedicated and distinct instrumental/manner prefixes.

[chris_notts]

Another example of agreement clitics with more word-like phonology compared to affixes. Apologies that the link is in Spanish:

https://www.academia.edu/860683/Gram%C3 ... o_Tultepec

See p52. Apparently in that dialect of Otomi, subject agreement and TAM clitics:

1. Are in fixed preverbal position
2. Are not stressed
3. Unlike affixes, have their own tone (which is the only distinguishing feature between some combinations)
4. Are not subject to certain word-level phonological processes such as nasal spreading and harmony

[Kuchigakatai]

Halkomelem also has that type of bound morphemes, and they're also often not at all like the regular nouns used for the body parts or tools in question (although sometimes they are).

Are you referring to lexical suffixes? Since those are a much broader set of morphemes with a much broader set of usages, and any that are used in a particular instance to indicate manner/an instrument should really be viewed in those terms. So I wouldn't say the Halkomelem system (or that in other Salishan and Northwest Coast languages) is comparable to that in, say, Siouan, where there's a smaller, coherent class of dedicated and distinct instrumental/manner prefixes.

Yes, I meant the lexical suffixes. I felt they were relevant because they are a closed class of dedicated and distinct morphemes... There are about 100 of them, and the vast majority is phonologically reduced or different from the corresponding free nouns (-xən 'leg, foot', from sx̌ə́n̓ə 'leg, foot'; -əɬp 'tree, plant' vs. θqet 'tree', sáx̌ʷəl 'plant'). How small is the class of instrumental prefixes in a typical Siouan language?

It is true that Halkomelem's lexical suffixes are more than instrumental affixes though, since they're also used when deriving compound nouns and in verb-object compounds (as necessarily bound, incorporated roots basically).

[Knit Tie]

Stupid question for those who deal with the Bantu languages - could an unrelated language borrow the classifier prefixes as regular prefixes through prolonged contact - i.e. /ki-/ becomes a regular dimunitive prefix and so on?

Another question, hopefully less stupid this time - can noun serialisation over time evolve into a plethora of verb adfixes, i.e. English "see" becomes an evidentiality marker?

[chris_notts]

Another question, hopefully less stupid this time - can noun serialisation over time evolve into a plethora of verb adfixes, i.e. English "see" becomes an evidentiality marker?

Do you mean verb serialisation? If so then yes. This is often where evidentiality markers come from, and verbs in SVCs do sometimes evolve into affixes in exactly the same way that verbal auxiliaries do.

[chris_notts]

See the following on Ewe, although most are particles instead of affixes at least in terms of orthography:

https://www.google.com/url?sa=t&source= ... ohn1tN80BR

Asymmetrical SVCs are virtually absent in Ewe. The meanings expressed by such structures in other languages are signalled by forms that have grammaticalized or lexicalized from verbs, for example, the aspectual, modality, and directional preverbs. It has been generally assumed that SVCs are the vehicle for such development (Ansre 1966a; Heine and Reh 1983; Heine et al. 1991; Lord 1993).
However, other MVCs have also served as vehicles. In fact, as shown in Figure 2, each of the MVC types can serve as a vehicle for lexicalization and/or grammaticalization.

[Pabappa]

Ive had a couple of ideas for semantic drift through dreams and other random thoughts .... e.g. a dream a few years back where the police in my hometown were killing prostitutes, both male and female, and i realized towards the end of the dream that the male "prostitutes" were drug dealers and that people were just grouping them together. I had a dream a few nights ago where two men cornered a woman and started talking about rape, and one of them said, "if you have the prestige of holes" ... meaning the "honor" of being a rape victim. In another recent dream, "dot" was a word for a narrow connection, as in "there's just a dot of connection between those two things". So a word for dot or bump could shift in meaning to a tube or string.

In another dream, going waaay back to childhood, I dreamt the English word lave meant "something used to chase away unwanted customers by perversely advertising one's negative traits". Recently I had a vision (just before i fell asleep, so I dont call it a dream) where a boy talkd about a smller boy in a highchair "being an end" because the smaller boy couldnt feed himself or do much on his own at all. So "end" could mean someone who depends on others for needs, regardless of reason.

I might coin a new word for "vicariously happy at someone else's misdeeds" based on another dream, and explain its survival for thousands of years by saying it's named after a person or a mythological figure. the person in the dream was named Haber. A long time ago I dreamt the words /ha/ "shield" and /ka/ "sword". Once I woke up shivering and couldn't remember the number that came after one but after some thought I decided it must be /kə'mal/. I suppose I could use that somewhere. in antoher dream, herratic meant "hairy" and palp meant "muscular" .... words like these are difficult to use in conlangs because they're not English but yet have ties to English such that I cant use them in an a priori conlang. but perhaps e.g. i could take the word for "touch, feel" and shift that in meaning to "muscular".

[Whimemsz]

Halkomelem also has that type of bound morphemes, and they're also often not at all like the regular nouns used for the body parts or tools in question (although sometimes they are).

Are you referring to lexical suffixes? Since those are a much broader set of morphemes with a much broader set of usages, and any that are used in a particular instance to indicate manner/an instrument should really be viewed in those terms. So I wouldn't say the Halkomelem system (or that in other Salishan and Northwest Coast languages) is comparable to that in, say, Siouan, where there's a smaller, coherent class of dedicated and distinct instrumental/manner prefixes.

Yes, I meant the lexical suffixes. I felt they were relevant because they are a closed class of dedicated and distinct morphemes... There are about 100 of them, and the vast majority is phonologically reduced or different from the corresponding free nouns (-xən 'leg, foot', from sx̌ə́n̓ə 'leg, foot'; -əɬp 'tree, plant' vs. θqet 'tree', sáx̌ʷəl 'plant'). How small is the class of instrumental prefixes in a typical Siouan language?

Apparently "seven to nine" in any given language, according to Douglas Parks and Robert Rankin, "Siouan Languages," in the Handbook of North American Indians volume 13.

[Kuchigakatai]

Apparently "seven to nine" in any given language, according to Douglas Parks and Robert Rankin, "Siouan Languages," in the Handbook of North American Indians volume 13.

Oof. That is tiny.

[Whimemsz]

Yeah I had not remembered it being quite that small. Pomoan languages have more like 20ish or a bit fewer -- Southern Pomo, which apparently has the most, has 21. (The Pomoan instrumental prefixes are the best ones anyway. I still intend to just completely rip off Pomoan in a conlang one day.)

[chris_notts]

Yeah I had not remembered it being quite that small. Pomoan languages have more like 20ish or a bit fewer -- Southern Pomo, which apparently has the most, has 21. (The Pomoan instrumental prefixes are the best ones anyway. I still intend to just completely rip off Pomoan in a conlang one day.)

Just for comparison/interest, here a (possibly incomplete?) list for Sudest, an Oceanic language of Papua New Guinea:

ghala- = by putting weight on
gita- = by grasping between two fingers / both hands
go- = with multiple strokes or cuts
ki- = with a single cut
mwana- = by hand
ri- = with teeth, by biting
taga- = by bending the arm
thu- = with a pointer
thuwo- = with tongs
vi- = with fingers
vuri- = with foot
vwa- = with an instrument
yo- = by stepping on
yo- = with fire

There clearly are recurring items in the lists of different languages. Most seem to contain the obvious body parts e.g. by weight, by hands, by feet, with teeth/mouth, then a few very common instruments or manner of motion/contact, e.g. fire, by poking with long object, by slashing/swiping/cutting with a long object, with a blunt object.

[Pabappa]

Yeah I had not remembered it being quite that small. Pomoan languages have more like 20ish or a bit fewer -- Southern Pomo, which apparently has the most, has 21. (The Pomoan instrumental prefixes are the best ones anyway. I still intend to just completely rip off Pomoan in a conlang one day.)

I love that. There's just enough polysemy to make it interesting without making it confusing. "butt" and "knife" being the same prefix is pretty funny. I like the soft/hard distinction .... used to distinguish between lips and teeth without actually using the words for lips and teeth. And I like that it has adapted to modern technology, having a prefix for "by shooting" and a non-specific one that covers automobiles.

The polysemy of yo- in Sudest is also interesting. This is pretty much the system I'm looking at with Poswa ... although I think I've decided to go for more abstract concepts such as the paid/free distinction that arose from the word for coin.

__________________________

A nice one-off gimmick would be for someone to make a language where the phonaesthetic perceptions common to Earth are completely inverted, so e.g. the words for pillow, love, lipstick etc sound like your car engine wont start but the words for sword, kill, wound etc are all made of soft sounds like /wuži/. I would do it but I wouldnt have the drive to make a fully functioning language like that. Plus, my tastes are different than others'.

i cant get over something that happened about ten years ago where i was eating lunch in the mall food court and a beautiful woman at the next table started choking on her food. i looked at her and nobody around her seemed concerned. it took me a few seconds she was communicating in her language, which was probably Goergian.