(Conlangs) Q&A Thread - Quick questions go here

A forum for all topics related to constructed languages
User avatar
Vlürch
greek
greek
Posts: 452
Joined: 09 Mar 2016 21:19
Location: Finland
Contact:

Re: (Conlangs) Q&A Thread - Quick questions go here

Post by Vlürch »

Kind of an "abstract" question, but...

The more I look into Nostratic and Borean and whatnot, the more I realise that regardless of whether or not the proposed macrofamilies were ever actually real or not, at least in Eurasia there seem to be some vague and elusive-to-the-point-of-probably-being-unquantifiable-but-still-noticeable universal sound symbolism and/or tendencies for words that sound certain ways to end up meaning certain things... you know, the bouba/kiki effect but not just limited to things being boubaish and/or kikish.

So, the question is: how do you break free from that? I mean, if you want to make a language that's "un-Borean" while still having a very Standard Average Borean phonology (meaning no implosives or clicks or linguolabials or whatever), how do you go about that? This isn't just about one language, but basically I'm asking "how do I make conlangs that feel right to me phonaesthetically, morphologically and lexically but don't feel shoehorned with Boreanness?"

Like, sure, it's arguable that since said conlangs are generally set in places where Borean languages are spoken, it's to be expected that they have a high degree of Boreanness, but it gets to a point where nothing feels truly a priori if that makes sense. I do of course also want to intentionally make conlangs that have high degrees of Boreanness, but it'd be nice to once in a while dabble in something that feels un-Borean even if its phonology and paradigms and whatever are still boringly Borean.

In case it's not obvious, yeah, I'm having a conlanging crisis over the realisation that every conlang I make falls squarely within Borean expectations.
Tanni
greek
greek
Posts: 600
Joined: 12 Aug 2010 02:05

Re: (Conlangs) Q&A Thread - Quick questions go here

Post by Tanni »

Vlürch wrote: 13 Apr 2021 12:05 Kind of an "abstract" question, but...

The more I look into Nostratic and Borean and whatnot, the more I realise that regardless of whether or not the proposed macrofamilies were ever actually real or not, at least in Eurasia there seem to be some vague and elusive-to-the-point-of-probably-being-unquantifiable-but-still-noticeable universal sound symbolism and/or tendencies for words that sound certain ways to end up meaning certain things... you know, the bouba/kiki effect but not just limited to things being boubaish and/or kikish.
Is there a list of that symbolism or tendencies? Can you provide it?
My neurochemistry has fucked my impulse control, now I'm diagnosed OOD = oppositional opinion disorder, one of the most deadly diseases in totalitarian states, but can be cured in the free world.
User avatar
Creyeditor
MVP
MVP
Posts: 5091
Joined: 14 Aug 2012 19:32

Re: (Conlangs) Q&A Thread - Quick questions go here

Post by Creyeditor »

I recently saw a paper on this. https://www.mdpi.com/1099-4300/12/4/844/htm
Creyeditor
"Thoughts are free."
Produce, Analyze, Manipulate
1 :deu: 2 :eng: 3 :idn: 4 :fra: 4 :esp:
:con: Ook & Omlűt & Nautli languages & Sperenjas
[<3] Papuan languages, Morphophonology, Lexical Semantics [<3]
Salmoneus
MVP
MVP
Posts: 3030
Joined: 19 Sep 2011 19:37

Re: (Conlangs) Q&A Thread - Quick questions go here

Post by Salmoneus »

I think the key thing to remember here is that there isn't actually a pervasive sound symbolism throughout 'borean' languages that makes words for certain things sound a certain way (beyond a little boubakiki and some semantically-limited onomatopoeia). However, if you stare at any random information long enough, you will eventually imagine that you see a 'vague, elusive to the point of being unquantifiable' pattern.

The second thing to consider is that 'Borean' is a concept so absurdly vast that it's tantamount to synonymous with 'language' - so far as I'm aware, only Khoisan languages haven't (yet) been frequently included in Borean, or as close relative to Borean (which (outlying member vs nearest relative) is of course just a matter of terminological difference). Unless you've extensively studied Khoisan families and are certain that the same symbolisms you see aren't also present there, then perhaps you're really seeing a 'pattern' present in human language in general, rather than just in Borean.

Finally, the third thing is: haven't you literally answered your own question?

You think you've discovered a system of sound-symbolism that is distinctively, recognisable Borean. And then you ask, how could you make a language that isn't distinctively, recognisably Borean on account of including this system. Surely then the answer is: just don't include the system you've 'discovered' in your conlang, and then it won't be distinctively 'Borean'. What am I missing here?
User avatar
Creyeditor
MVP
MVP
Posts: 5091
Joined: 14 Aug 2012 19:32

Re: (Conlangs) Q&A Thread - Quick questions go here

Post by Creyeditor »

I just wanted to mention that I think the paper that I linked might be fringe science but it is still interesting for conlangers.
Creyeditor
"Thoughts are free."
Produce, Analyze, Manipulate
1 :deu: 2 :eng: 3 :idn: 4 :fra: 4 :esp:
:con: Ook & Omlűt & Nautli languages & Sperenjas
[<3] Papuan languages, Morphophonology, Lexical Semantics [<3]
User avatar
Vlürch
greek
greek
Posts: 452
Joined: 09 Mar 2016 21:19
Location: Finland
Contact:

Re: (Conlangs) Q&A Thread - Quick questions go here

Post by Vlürch »

Tanni wrote: 13 Apr 2021 12:41Is there a list of that symbolism or tendencies? Can you provide it?
Not that I know of, but the reason things like Nostratic and even Borean can be reconstructed is that correspondences exist. Searching through the database of reconstructions of Nostratic or even Borean on Starostin's site just shows again and again that even in languages that are unrelated (unless Nostratic or even Borean is true) there are correspondences in things like whether words have velar/uvular plosives or some kind of sibilants, etc. In a lot of cases the meanings do lead to it seeming like a stretch, but in just as many cases the meanings are (at least nearly) identical.
Creyeditor wrote: 13 Apr 2021 13:16I recently saw a paper on this. https://www.mdpi.com/1099-4300/12/4/844/htm
Interesting... honestly I only understood like a tenth of that, it's so technical, but it seems like it concludes that universal sound symbolism is true at least to some degree?
Salmoneus wrote: 13 Apr 2021 14:39I think the key thing to remember here is that there isn't actually a pervasive sound symbolism throughout 'borean' languages that makes words for certain things sound a certain way (beyond a little boubakiki and some semantically-limited onomatopoeia). However, if you stare at any random information long enough, you will eventually imagine that you see a 'vague, elusive to the point of being unquantifiable' pattern.
I disagree. There are enough similarities between words even in Uralic and Sino-Tibetan or at least Sinitic (which are clearly not related (or if they are, it goes back to Borean, and while I do think it's likely that all (or at least most) langauges are ultimately related, I don't believe they can be lumped together as neatly as in Borean)) to discard as just coincidence. I don't know which is more likely, that there's universal (or at least Borean) sound symbolism leading to those similarities or if it's because of ancient relationship/contact or something else, and there's no reason it can't be both, but those kinds of things haven't been explained. I mean, yeah, you can say "it's pure coincidence" about everything but that still won't explain why the coincidences happen.
Salmoneus wrote: 13 Apr 2021 14:39The second thing to consider is that 'Borean' is a concept so absurdly vast that it's tantamount to synonymous with 'language' - so far as I'm aware, only Khoisan languages haven't (yet) been frequently included in Borean, or as close relative to Borean (which (outlying member vs nearest relative) is of course just a matter of terminological difference). Unless you've extensively studied Khoisan families and are certain that the same symbolisms you see aren't also present there, then perhaps you're really seeing a 'pattern' present in human language in general, rather than just in Borean.
Well, in general I don't really know the first thing about the indigenous languages of Africa, Papua New Guinea or the Americas. I'm just going off reconstructions of Nostratic and Borean, and comparisons. The reason I stuck with "Borean" is because I can't assume it also applies to languages not included in it since I don't know if it does. I mean, I know the words for "butterfly" do fit the kind of vague pattern for butterfly words in at least some Papuan languages as well, but I'm not going to assume anything else about them based on just that... and the few African, Papuan and American languages whose Swadesh lists I've looked at for comparisons seemed more different and so do their grammars.

Just going by Starostin's site, the correspondences between the "Borean" and "Macro-Khoisan" aren't nearly as clear as within Borean AFAICT. It's probably also indicative of that being the case since neither Starostin's arguement has been that Khoisan languages would be Borean and the Starostins are generally considered ultimate hyperlumpers to a degree of crackpottery, so...
Salmoneus wrote: 13 Apr 2021 14:39Finally, the third thing is: haven't you literally answered your own question?
No, because:
Salmoneus wrote: 13 Apr 2021 14:39Surely then the answer is: just don't include the system you've 'discovered' in your conlang, and then it won't be distinctively 'Borean'. What am I missing here?
I've tried that by trying to make things distinctively non-Uralic, non-Altaic and non-Indo-European (which I guess could be condensed as "non-Mitian"), but it often turns out that the things still apply in some other languages that are included at least in Borean. So, they're still indicative of "Boreanness" even if to a lesser degree. And more than that, doing the un-Borean thing in one case might lead to more Boreanness in other cases (even if we're just talking about noun cases lol), but especially more generally... which would be fine if the Boreanness was already there, but if it's not, it'll feel like it's not fitting and clash with the phonaesthetics of the language.

And just to be clear, I know it's a stupid question and a stupid problem to have. Sometimes I have these kinds of stupid problems because I feel like I'm pretty much making the same conlang over and over again.
Salmoneus
MVP
MVP
Posts: 3030
Joined: 19 Sep 2011 19:37

Re: (Conlangs) Q&A Thread - Quick questions go here

Post by Salmoneus »

Vlürch wrote: 13 Apr 2021 15:45
Tanni wrote: 13 Apr 2021 12:41Is there a list of that symbolism or tendencies? Can you provide it?
Not that I know of, but the reason things like Nostratic and even Borean can be reconstructed is that correspondences exist.
The reason they CAN'T be constructed is that correspondences DO exist.
The reason crackpots keep trying (and failing) to construct them is that random coincidences DO exist, and some people refuse to accept this.
Searching through the database of reconstructions of Nostratic or even Borean on Starostin's site just shows again and again that even in languages that are unrelated (unless Nostratic or even Borean is true) there are correspondences in things like whether words have velar/uvular plosives or some kind of sibilants, etc. In a lot of cases the meanings do lead to it seeming like a stretch, but in just as many cases the meanings are (at least nearly) identical.
So, when browsing through a database of coincidences specifically chosen as the most compelling similarities a crackpot could find, the most you can say is that still only half the time can you even say of two words with similar meanings that they have something as wildly vague as "having some sort of sibilant" in common? But that the other half the time, even in this small sample of the most compelling, cherry-picked comparisons possible - many of them not between actual words, but between "reconstructions" that have been created specifically for the purpose of looking similar to 'reconstructions' from other families, you STILL can't find a coincidence that compelling without an implausible stretch in semantics?

That's not good evidence!
Salmoneus wrote: 13 Apr 2021 14:39I think the key thing to remember here is that there isn't actually a pervasive sound symbolism throughout 'borean' languages that makes words for certain things sound a certain way (beyond a little boubakiki and some semantically-limited onomatopoeia). However, if you stare at any random information long enough, you will eventually imagine that you see a 'vague, elusive to the point of being unquantifiable' pattern.
I disagree. There are enough similarities between words even in Uralic and Sino-Tibetan or at least Sinitic (which are clearly not related (or if they are, it goes back to Borean, and while I do think it's likely that all (or at least most) langauges are ultimately related, I don't believe they can be lumped together as neatly as in Borean)) to discard as just coincidence. I don't know which is more likely, that there's universal (or at least Borean) sound symbolism leading to those similarities or if it's because of ancient relationship/contact or something else, and there's no reason it can't be both, but those kinds of things haven't been explained. I mean, yeah, you can say "it's pure coincidence" about everything but that still won't explain why the coincidences happen.
I'm sorry, but this is just conspiracy theorising.

There's an easy explanation for why coincidences happen: coincidences. You don't get to say "ahh, but why do coincidences happen!?" - the answer in in the question, it's a coincidence. It's like asking a physicist: "sure, the law of gravity makes objects attracted to one another... but what makes the universe obey the law?" - it's a category mistake. There is no 'why' - the concept of "physical law" is inseparable from being obeyed, there is no such thing as a physical law that isn't obeyed, by definition. Likewise, by definition, there is no reason why a coincidence has happened, it's just a coincidence. If there were a reason, it wouldn't be a coincidence!

Specifically, there are thousands of languages, each with tens or hundreds of thousands of words. If you have a large enough list of these words, you are 100% guaranteed to be able to spot a large number of surprising coincidences, even if the languages are all completely unrelated. It's impossible to create a large database of words WITHOUT finding these coincidences, even if you generate them by computer. It's only significant when the number of 'coincidences' is significantly greater than would be expected by random chance - i.e. when they're not coincidences. Have you done a statistical study of this sort that demonstrates that these universal sound-meaning correspondances aren't just what you'd expect through random chance? Because if so, well done, somebody should hand you your Nobel prize shortly!

Leaving aside that fact that, again, you're apparently deducing this from a 'database' of reconstructions designed specifically to look related, constructed on the basis of the words that a crackpot already thought looked most related [with, fwiw, little respect for accuracy in reporting either sound or meaning, and in some cases the reconstructions are contrary to everything historical linguistics believes to be true].
Salmoneus wrote: 13 Apr 2021 14:39The second thing to consider is that 'Borean' is a concept so absurdly vast that it's tantamount to synonymous with 'language' - so far as I'm aware, only Khoisan languages haven't (yet) been frequently included in Borean, or as close relative to Borean (which (outlying member vs nearest relative) is of course just a matter of terminological difference). Unless you've extensively studied Khoisan families and are certain that the same symbolisms you see aren't also present there, then perhaps you're really seeing a 'pattern' present in human language in general, rather than just in Borean.
Well, in general I don't really know the first thing about the indigenous languages of Africa, Papua New Guinea or the Americas. I'm just going off reconstructions of Nostratic and Borean, and comparisons. The reason I stuck with "Borean" is because I can't assume it also applies to languages not included in it since I don't know if it does. I mean, I know the words for "butterfly" do fit the kind of vague pattern for butterfly words in at least some Papuan languages as well, but I'm not going to assume anything else about them based on just that... and the few African, Papuan and American languages whose Swadesh lists I've looked at for comparisons seemed more different and so do their grammars.
Ah, a concrete example of a universal sound-meaning comparison! Butterfly!

So, let's see: butterfly; lepidopteran; Falter; Schmetterling; Molkendieb; dearbadan-de;pavilhao; wijfvouter; fjaril; skoenlapper; flutur; mariposa; guagóg; pumarina; palometa; somerfugl; drugys; gloyn byw; taurenis; tykki Duw.

...I'm not seeing a really compelling universal phonetic pattern here! And that's just from within western Europe, between languages that are all related (some of them very closely related). When you add in other 'Borean' words like wu dip, lyve, taqalukisaq, erveehej, ḥăṭrăṗəj, k'aalógii, doolé, tsenblale and nabi, I just don't see what this universal pattern is that's so compelling!

[interestingly, there's quite a strong coincidence linking Kartvelian, Semitic, and certain IE branches (Italic, Germanic, and Iranian), with reduplicated forms of 'par' or 'pal', which can be semantically explained in either Kartvelian or PIE. Given the geographic and cultural closeness of PS, PIE and PK, I guess it's possible that there's a genuine wanderwort there, most likely out of Kartvelian. But, notably, it's not in Austroasiatic outside of Semitic (i.e. outside of the branch that would have been in contact with Kartvelian), so a deeper connexion can be dismissed. It's also possible that this is a pure coincidence, particularly since the reduplication may be onomatopoeic (it's a small creature that flutters and is popular with children, so...). It's also worth noting that we do see some definite coincidences here: "Falter" and "flutur" aren't related, and neither is related to English "flutterby", though a reconstructionist would surely assume that they were, if they weren't from such a well-documented family!]
Salmoneus wrote: 13 Apr 2021 14:39Surely then the answer is: just don't include the system you've 'discovered' in your conlang, and then it won't be distinctively 'Borean'. What am I missing here?
I've tried that by trying to make things distinctively non-Uralic, non-Altaic and non-Indo-European (which I guess could be condensed as "non-Mitian"), but it often turns out that the things still apply in some other languages that are included at least in Borean. So, they're still indicative of "Boreanness" even if to a lesser degree.
But again, if you're able to identify Boreanness because of some features, just don't have those features. The same way that you've previously just not had Mitian features.
And more than that, doing the un-Borean thing in one case might lead to more Boreanness in other cases (even if we're just talking about noun cases lol), but especially more generally...
In a situation where both doing X and NOT doing X make something look more Borean, then by definition X is not related to Boreanness. This is a paradigm case of unfalsifiable, conspiratorial thinking!
which would be fine if the Boreanness was already there, but if it's not, it'll feel like it's not fitting and clash with the phonaesthetics of the language.
Well if it's "not there", just keep it not there and don't add it.

Although I literally cannot possible imagine what distinctively Borean phonaesthetic features you feel occur naturally in Mandarin, Inuktitun, Georgian, Berber, Finnish, Navajo, Spanish and Mongolian, that would so obviously and unbearably clash with the completely different phonaesthetics of your conlang. What is your conlang like, that you feel you could group those mentioned languages, on phonaesthetic grounds, into one group, with your conlang outside that group!?
User avatar
eldin raigmore
korean
korean
Posts: 6352
Joined: 14 Aug 2010 19:38
Location: SouthEast Michigan

Re: (Conlangs) Q&A Thread - Quick questions go here

Post by eldin raigmore »

Creyeditor wrote: 13 Apr 2021 14:53 I just wanted to mention that I think the paper that I linked might be fringe science but it is still interesting for conlangers.
I don’t think it is fringe science.
They found that about 18% of the 40 words they looked at in a “genetically” and “areally” diverse sample of more than 100 languages seem to have significant correlations between sound and meaning that can’t be explained by loan words nor “family” relations.
They said the data they analyzed is not enough to support any more specific or detailed hypothesis about sound-symbolism.
Their paper is, as far as they know, the first rigorous quantitative investigation of whether sound-symbolism is statistically significant.
None of that is fringe.
I was able to follow almost all of the statistics and maths and computation. The exception was the stuff about “metric multidimensional scaling” (“MDS”). That’s new to me, and I’m not certain the use they put it to is really appropriate.

....

For someone to jump on this paper and say it proves all the various things that have been said about sound symbolism and deep-time protolanguages and such would be fringe science. They never made any such claims.

What they did claim I think they pretty much proved.

The paper requires careful reading in my opinion.

There are several hypotheses that they explicitly say they don’t have the data to confirm or refute.

....

To the degree that they come close to a more detailed popular sound-symbolism hypothesis, there are two.
One is the finding that the word for “breast” tends to be shaped something like “muma”.
The other is that the words “I thou we(incl) name” tend to have shapes reminiscent of each other.

....

If around 18% +- (some fudge factor) of a lexicon is at least weakly guided by sound-symbolism, that still leaves about 82% +- (wiggle room) to be essentially almost arbitrary, in some sense. That is, it’ll probably be guided by the internal morphology of the language, and/or borrowing from neighboring languages, and/or pretty much any other process for innovating or changing vocabulary that you can think of besides sound-symbolism.

That 18%-ish fraction is “statistically significant” in the sense that it shouldn’t be ignored or dismissed; it won’t dominate any naturalistic or realistic conlang’s vocabulary.
User avatar
Creyeditor
MVP
MVP
Posts: 5091
Joined: 14 Aug 2012 19:32

Re: (Conlangs) Q&A Thread - Quick questions go here

Post by Creyeditor »

Vlürch wrote: 13 Apr 2021 15:45
Creyeditor wrote: 13 Apr 2021 13:16I recently saw a paper on this. https://www.mdpi.com/1099-4300/12/4/844/htm
Interesting... honestly I only understood like a tenth of that, it's so technical, but it seems like it concludes that universal sound symbolism is true at least to some degree?
What Eldin said basically. They found something, but they are not jumping to cobclusions about specific hypotheses on sound symbolism.
Creyeditor
"Thoughts are free."
Produce, Analyze, Manipulate
1 :deu: 2 :eng: 3 :idn: 4 :fra: 4 :esp:
:con: Ook & Omlűt & Nautli languages & Sperenjas
[<3] Papuan languages, Morphophonology, Lexical Semantics [<3]
User avatar
Vlürch
greek
greek
Posts: 452
Joined: 09 Mar 2016 21:19
Location: Finland
Contact:

Re: (Conlangs) Q&A Thread - Quick questions go here

Post by Vlürch »

Salmoneus wrote: 13 Apr 2021 18:03That's not good evidence!
Not everything has to be about evidence, at least I'm still (mostly) talking about this because of conlanging reasons. I don't care whether Borean is real (but like I already said, I don't think it is; I think at least most of the languages are probably related to each other, but not in such a clean-cut way as the Borean model) or even Nostratic (which I also doubt, again it seems too tidy), especially when conlanging, because for me conlanging is about fun. I just want to make more different kinds of conlangs that aren't the same thing over and over again (even though I enjoy doing that too), and it's not easy. There's a reason why stereotypically everyone's first conlang is a Romlang, and so on...
Salmoneus wrote: 13 Apr 2021 18:03So, let's see: butterfly; lepidopteran; Falter; Schmetterling; Molkendieb; dearbadan-de;pavilhao; wijfvouter; fjaril; skoenlapper; flutur; mariposa; guagóg; pumarina; palometa; somerfugl; drugys; gloyn byw; taurenis; tykki Duw.
Yeah, but for example:
Finnish perhonen
Persian پروانه (parvâne)

Māori pepeke
Indonesian kupu-kupu
(archaic) Japanese teputepu

etc.

The Finnish term is a relatively recent coinage and the Persian term had an unrelated meaning until relatively recently, yet they've become similar; it's a coincidence, but it fits the voiceless bilabial plosive and nasal pattern that's common in some butterfly terms even though in the Finnish term the -nen part is literally a productive suffix.

The Japanese term turned out that way because it's a Middle Chinese loanword made to fit Japanese phonology (and has since undergone further sound changes to become less similar). However, the Indonesian term is also a reduplicated two-syllable term with /p/ in the second syllable. The Māori term also has /p/ and /k/, and also /e/... and so on, looking at the butterfly words on Wiktionary, there are more that have some kind of similarity to the words in unrelated languages than ones that don't. It's vague, often coincidental, but...

I'd also argue that of the ones you listed, mariposa, pumarina and palometa have vague similarities that I underlined.

So, with butterfly words, it seems it's not just a Borean vs non-Borean thing, which is great because that makes it a cool thing to have similar words for butterfly even in "un-Borean-esque" conlangs.
Salmoneus wrote: 13 Apr 2021 18:03[interestingly, there's quite a strong coincidence linking Kartvelian, Semitic, and certain IE branches (Italic, Germanic, and Iranian), with reduplicated forms of 'par' or 'pal', which can be semantically explained in either Kartvelian or PIE. Given the geographic and cultural closeness of PS, PIE and PK, I guess it's possible that there's a genuine wanderwort there, most likely out of Kartvelian. But, notably, it's not in Austroasiatic outside of Semitic (i.e. outside of the branch that would have been in contact with Kartvelian), so a deeper connexion can be dismissed. It's also possible that this is a pure coincidence, particularly since the reduplication may be onomatopoeic (it's a small creature that flutters and is popular with children, so...). It's also worth noting that we do see some definite coincidences here: "Falter" and "flutur" aren't related, and neither is related to English "flutterby", though a reconstructionist would surely assume that they were, if they weren't from such a well-documented family!]
That's what I'm talking about, and I don't think it's truly a coincidence that they BECOME similar. Their origins are in some cases demonstrably unrelated, but they still converge to become more similar to each other. To me that's too much to be a coincidence and is indicative of something bigger. I know that's unscientific crackpottery and whatever since it's unfalsifiable and relies on some kind of borderline mystical stuff, but I'm not even trying to figure out the explanation. I'm talking about this from the perspective of conlanging, and what I believe doesn't make any difference.
Salmoneus wrote: 13 Apr 2021 18:03But again, if you're able to identify Boreanness because of some features, just don't have those features. The same way that you've previously just not had Mitian features.
I was saying that's what I've tried to by not having Mitian features, but it turns out they're still Borean features.
Salmoneus wrote: 13 Apr 2021 18:03In a situation where both doing X and NOT doing X make something look more Borean, then by definition X is not related to Boreanness. This is a paradigm case of unfalsifiable, conspiratorial thinking!
Conlanging isn't scientific. Bringing points of "unfalsifiability" and "conspiratorial thinking" into a desire to make less "typical" words in conlangs doesn't really make sense to me.
Salmoneus wrote: 13 Apr 2021 18:03Well if it's "not there", just keep it not there and don't add it.
In theory it's that simple, but in practice... every sound symbolism that I try to add increases Boreanness.
Salmoneus wrote: 13 Apr 2021 18:03Although I literally cannot possible imagine what distinctively Borean phonaesthetic features you feel occur naturally in Mandarin, Inuktitun, Georgian, Berber, Finnish, Navajo, Spanish and Mongolian, that would so obviously and unbearably clash with the completely different phonaesthetics of your conlang. What is your conlang like, that you feel you could group those mentioned languages, on phonaesthetic grounds, into one group, with your conlang outside that group!?
Well, anything that feels "right" for me to do in a conlang would almost certainly turn out to be a thing in one of those languages or at least some other languages that are lumped into Borean. For example, with plurals. Finnish has /t/, Spanish has /s/, Mongolian has /t/ and /s/ and Eskimo-Aleut languages have /t/ as well. Turkic languages have /lar/ and Japanese has /ra/, so liquid plurals feel "Borean". Same with /n/ because of Arabic and Sumerian plurals including it. So, I might do plurals with /b/ because that feels weird but still kinda fitting... aaaand turns out Georgian has /b/ in plurals, so maybe /w/ except that's still close to /b/ and /k/ which is "Borean" because of Hungarian, Basque, etc. Every plural that "feels right" at all would be "Borean" to some degree.

It's possible this is caused only by my (near-)total lack of knowledge about non-"Borean" languages and it's just that more common sounds "feel more right" because they're more common, but it's harder to find information about the details of the indigenous languages of Africa, Papua New Guinea, the Americas and Australia, except for the big ones but I don't know how representative those are of the smaller ones.

Well, thanks for the long reply anyway, it's helpful in that it's making me wonder if it really might be a problem I'm just making for myself because I assume things are related to Boreanness even if they're not. I don't think that's the case because of what little I've found about "un-Borean" types of equivalents, but maybe it's not Boreanness per se but something else (it's not confined to Nostraticity either, though, it's wider than that) or even just the commonness of sounds or whatever... in that case, I'm an actual idiot.
eldin raigmore wrote: 13 Apr 2021 18:48...
That makes sense, interesting! I wonder if a larger-scale study will ever be done, because that'd be really cool to see.
eldin raigmore wrote: 13 Apr 2021 18:48If around 18% +- (some fudge factor) of a lexicon is at least weakly guided by sound-symbolism, that still leaves about 82% +- (wiggle room) to be essentially almost arbitrary, in some sense. That is, it’ll probably be guided by the internal morphology of the language, and/or borrowing from neighboring languages, and/or pretty much any other process for innovating or changing vocabulary that you can think of besides sound-symbolism.
This is something that I've always struggled with and still struggle with (even with a posteriori conlangs), the arbitrariness thing. Words in natlangs often sound fitting for their meanings, even if there isn't any obvious sound symbolism.🤔
Salmoneus
MVP
MVP
Posts: 3030
Joined: 19 Sep 2011 19:37

Re: (Conlangs) Q&A Thread - Quick questions go here

Post by Salmoneus »

I'm sorry, but you're not describing an issue with languages, you're describing a mental issue. I'm not saying that as an insult, but just so that we're clear that the only way to address the 'problem' you're having is with some form of reflection, self-help or therapy. There is no actual 'solution' in the external world.

So:
Well, anything that feels "right" for me to do in a conlang would almost certainly turn out to be a thing in one of those languages or at least some other languages that are lumped into Borean. For example, with plurals. Finnish has /t/, Spanish has /s/, Mongolian has /t/ and /s/ and Eskimo-Aleut languages have /t/ as well. Turkic languages have /lar/ and Japanese has /ra/, so liquid plurals feel "Borean". Same with /n/ because of Arabic and Sumerian plurals including it. So, I might do plurals with /b/ because that feels weird but still kinda fitting... aaaand turns out Georgian has /b/ in plurals, so maybe /w/ except that's still close to /b/ and /k/ which is "Borean" because of Hungarian, Basque, etc. Every plural that "feels right" at all would be "Borean" to some degree.
As we've mentioned, almost all languages are or have been theorised to be 'Borean'. That's thousands and thousands of languages. There are only a few dozen commonly-occuring morphemes. Every possible morpheme occurs in a plural marker in at least one of those thousands upon thousands of 'Borean' languages. If Boreanism is defined in such a broad way, then literally every possible plural marker will 'feel Borean' to you. You have set yourself a task (finding a plural marker that doesn't feel Borean) that has literally no possible satisfactory conclusion.

[you've already ruled out /l/, /r/, /t/, /s/, /n/, /b/, /k/ and /w/. Presumably the -ides plural in Greek rules out /d/ as well/, and the Semitic -im plural rules out /m/. You've mentioned /a/, but of course you have to rule out /i/ as well, which should probably extend to /j/. And /p/, of course, is the plural marker in Yukaghir. You're basically just left with /g/, and I'd bet at a thousand-to-one that if you look you'll find a bunch of Borean languages with a /g/ plural marker too. Similarly, /a/, /e/ and /i/ are out right away, but I don't doubt that you can find Borean plurals with /u/ and /o/ as well. There is literally no possible plural marker that would not meet your criteria for unacceptability.]

This is not rational thinking. This is a textbook psychological/behavioural issue: repeating a task (in this case word-coining) again and again because it never quite 'feels right' and you need to keep trying until you get it 'right'... but at the same time defining 'right' in such a way that you can never, ever, make it 'feel right' (both because your criteria for wrongness are so all-encompassing, and because they are so subjective and dependent upon mood and impulse that nothing could ever escape them - and that what seems 'right' one day will seem 'wrong' the next as your mood changes). This is an endless loop, which there is no way out of - except either working on combating the desire itself, or on establishing more objective and specific guidelines for when the task is completed. [I don't know - not knowing you personally and not being a psychologist - whether this falls within the technical definition of a compulsive behaviour, but the underlying mental pattern is clearly the same that underlies compulsive behaviours. Analogous behaviours are "touching something alternately with each hand to 'balance' the sensation, but it never feels quite balanced so you can't stop" and "continually re-checking that you've locked the door because although you keep doing it none of the checks are sufficient to make you 100% confident that you didn't make a mistake in checking it and you're not just remembering yesterday's check so you have to do it again to make yourself feel sure"].

[what do I mean by an objective guideline? A specific example would be something like: make a list of ten Borean languages; discard a possible morpheme if it contains a phoneme that occurs in the equivalent marker in 6 or more of those languages; or discard a possible morpheme if it contains more than two phonemes that occur in the equivalent marker in at least 2 of those languages. Now, you don't necessarily need to have SUCH objective and specific rules, but that's an example of the sort of approach that turns an unsolvable problem ('everything feels vaguely Borean!') into a solvable problem ('some things are objectively too Borean for my needs'). Generally, the more you struggle with this sort of problem, the stricter and more objective it's helpful to be with the criteria you impose.]

Of course, this is a problem most conlangers have to some extent, myself included. But when you have a psychological problem like this, it's important to look for solutions in your own thinking, rather than in the external world - phrasing this as a problem related specifically to languages (rather than a problem with your own thinking) only makes it harder to break out of the compulsion.



This doesn't even get into the issue of why "not feeling Borean" - a property that almost no existing human languages have, and that has no objective value - is so important to you anyway.
Vlürch wrote: 14 Apr 2021 02:21
Salmoneus wrote: 13 Apr 2021 18:03That's not good evidence!
Not everything has to be about evidence... Conlanging isn't scientific. Bringing points of "unfalsifiability" and "conspiratorial thinking" into a desire to make less "typical" words in conlangs doesn't really make sense to me.
Rationality cannot evaluate your desire. If someone is desperate to, for instance, climb the fourth-tallest mountain on earth, it doesn't make much sense to call this desire irrational. "Reason is, and ought only to be, the slave of the passions, and can never pretend to any other office but to serve and obey them... it is not contrary to reason to prefer the destruction of the whole world to the scratching of my finger".

However, it does make sense to talk about the rationality of the means by which people seek to fulfill their desires. Let's say there's a large object on the other side of a partially-open door, and you want it. You wanting it is, in itself, neither rational nor irrational. But if you keep grabbing it and pulling it toward you, and it keeps bumping up against the back of the door and shutting the door close, then it IS legitimate to say that your METHOD, your MEANS, of fulfilling that want is irrational: it doesn't work. Similarly, the man who wants to climb the fourth-tallest mountain: if he goes about this by climbing each mountain he sees in the hope that it might happen to be the fourth-tallest, or if he asks a psychic to tell him which mountain Hillary's ghost says is the fourth-tallest, then of course we can say he's being irrational: the method he has chosen will not allow him (except by complete random luck) to fulfill the desire that he has - and because he is probably smart enough to realise this, he is likely to remain unhappy (even if he happens to climb the right mountain by chance, he won't feel confident that he has done so). So it's entirely legitimate to say to this man: first, let's find a way of determining which mountain actually is the fourth-tallest; and then, let's research the sort of equipment you'll need to find a mountain. It's not that his impulse needs to conform to reason (though, since most desires are only instrumental, it's probably fair to ask him: why do you want to do this?; what do you hope to accomplish?; what's your ultimate goal here?), but if he actually wants to do what he says he wants to do, then his methods need to be rational, or else he probably won't succeed. [by definition: if there is a way to do something that's just as effective as the rational way, then it IS a rational way...]

In this case, your desire to make a conlang is beyond the dictates of reason, granted. To a large extent, we can treat the desire to make a maximally un-Borean languages as likewise a-rational (though I do think it's probably wise to ask yourself: why am I fixating on this one arbitary thing? Is this really my objective, or do I have an underlying motivation here?). But it's entirely legitimate to point out that the way you are going about trying to do what you say you want to do is irrational - your definition of 'un-Borean' is incoherent, and your criteria for judging if you're doing the right thing are impossible to satisfy. If you continue to pursue this goal by this method, you will fail, and you will continue to recognise that you are failing.

And this matters because if people want to do a thing, but adopt means and methods that ensure they cannot do that thing, or cannot satisfy themselves that they have done it even if they do it, then this will be a continual source of unhappiness for them.


I just want to make more different kinds of conlangs that aren't the same thing over and over again... and it's not easy
Here we come back to my point about the underlying motivations behind the quest for unboreanness. It's good that you're already conscious of this! Because again, there is a mismatch here between your desire (to not make the same thing again and again) and your means (to instead make a conlang that isn't Borean).

Why is this a mismatch? Because 'Borean' essentially covers the whole of human language; and certainly, your definition of Borean, which includes both Borean languages and any language with features that remind you of Borean languages, covers absolutely all human languages. But even in the narrow sense, the languages you're talking about are not simply 'the same thing again and again', but are instead bizarrely, unendingly, astonishingly varied and individual and distinctive. I mean, I'm constantly struck how incredibly different - phonologically, syntactically - Irish is from English (Irish seems to do everything differently!), and those are two Indo-European languages that have grown up adjacent to one another and under one another's extensive influence. If you compare Irish to English to Inupiaq to Vietnamese to Hebrew to Navajo to Iaai to Circassian... the diversity is VAST! All-encompassing! Limitless! The idea that you have to be un-Borean in order to not be doing the same thing again and again is.... well, it is only rational if you have an incredibly, multifariously broad sense of 'the same' in mind. And if you really want something that doesn't look at all anything like any of those languages... well then you'r barking up the wrong tree worrying about this phoneme or that in your plural affix, you need to instead be thinking about much more alien things altogether, like a three-dimensional non-linear language that lacks the concept of words, or something like that....
The Japanese term turned out that way because it's a Middle Chinese loanword made to fit Japanese phonology (and has since undergone further sound changes to become less similar). However, the Indonesian term is also a reduplicated two-syllable term with /p/ in the second syllable. The Māori term also has /p/ and /k/, and also /e/... and so on, looking at the butterfly words on Wiktionary, there are more that have some kind of similarity to the words in unrelated languages than ones that don't. It's vague, often coincidental, but...
But there is no set of words that would NOT have these similarities! The word in X will have similar phonemes to the word in Y, and the word in Y will have similar phonemes to the word in Z, and so on. You're adopting a literally insane methodology in detecting 'similarity'.

In this case, you're happy to say that a word fits the "pattern" for butterfly words if it contains /p/, or /m/, or /n/, or /k/, or /u/, or /e/, or /l/, or /r/, or has four syllables or is formed by reduplication. You do see how this is a 'pattern' that everything fits into, right? And why not /a/, and /i/? (Pumarina, palometa and mariposa all have /a/, and two of them have /i/! Pattern!). And wait, the /u/ of 'pumarina' is also found in 'kupu-kupu'! So /u/ is part of the butterfly pattern too, right?
That's what I'm talking about, and I don't think it's truly a coincidence that they BECOME similar. Their origins are in some cases demonstrably unrelated, but they still converge to become more similar to each other.
But the point is: they aren't similar! I intentionally chose a list of words that look nothing alike!

However, because you've decided that they must be similar, you can find similarities between them. Because it's possible to find similarities in literally any list of words.

And to bring this back to your project: just as you were able to find a 'pattern' in a list of random words that I intentionally chose because I couldn't see any pattern in them, so to if you form a list of Borean words and words in your conlang, you will continue to find a pattern - no matter what words you choose for your conlang. And so, if you want your conlang to not share these 'Borean patterns', then you will always be unsatisfied. Because you have set yourself a criterion that it is literally impossible to fulfill.
To me that's too much to be a coincidence and is indicative of something bigger. I know that's unscientific crackpottery
And yet you don't recognise that your belief being, as you call it, 'crackpottery', means you shouldn't hold that belief.

Why shouldn't you hold that belief? Because if you hold irrational beliefs, it becomes impossible to fulfill your goals. You cannot climb the fourth-highest mountain on earth if you believe that that mountain is located on the back of an immense flying turtle that appears over the north pole every seven years and is only accessible by taking peyote and chanting the lost words of an ancient Sumerian hymn. [there's a remote chance that with enough peyote you may convince yourself you've climbed it (c.f. the periodic arrival of an auxlanger who believes he's created the perfect simple language that everyone can easily learn and that has no ambiguities), but this is very unlikely, and even that illusion of success generally fades once morning comes].

In conlanging, you can set yourself whatever objective you want, rationality be damned. Go ahead! But if you decide to pursue that objective through irrational means (often stemming from irrational beliefs), then you will never succeed in your objective, and you will be disappointed. And if you adopt irrational criteria for your success, then success will not even be conceptually possible. If you set yourself an impossible task, and do not recognise that it is impossible, then you will encounter unending disappointment.
User avatar
Vlürch
greek
greek
Posts: 452
Joined: 09 Mar 2016 21:19
Location: Finland
Contact:

Re: (Conlangs) Q&A Thread - Quick questions go here

Post by Vlürch »

Salmoneus wrote: 14 Apr 2021 12:56I'm sorry, but you're not describing an issue with languages, you're describing a mental issue. I'm not saying that as an insult, but just so that we're clear that the only way to address the 'problem' you're having is with some form of reflection, self-help or therapy. There is no actual 'solution' in the external world.
Am I the only one who needs therapy or does everyone who wants to try making a proper "exotic" conlang need therapy? You're making the problem out to be bigger than it is: this isn't causing me any mental anguish or anything, just mild annoyance, because again, my point is that I'd like to at least try to make at least one conlang that isn't "Borean-esque".

Think of it like wanting to make a non-SAE conlang but with a wider net, as a learning experience. If it's not actually possible, alright, but I for one don't think eg. Khoisan languages or Papuan languages "feel Borean", and they're not considered Borean even by the ultimate hyperlumpers, so clearly there is a possibility for languages to be "un-Borean". Maybe it's not possible for me to make a conlang like that (at the moment?), possibly because I don't know enough about those "un-Borean" natlangs, possibly because my "Borean detector" runs too deep, or some other reason. I don't know, but why are you trying to make this an argument?
Salmoneus wrote: 14 Apr 2021 12:56If Boreanism is defined in such a broad way, then literally every possible plural marker will 'feel Borean' to you. You have set yourself a task (finding a plural marker that doesn't feel Borean) that has literally no possible satisfactory conclusion.
And I admitted that that's possible, and if you know for a fact that that's the case because Boreanness really is that vast and extends even to non-"Borean" languages, alright. I'll take your word for it because I know you know a lot more about non-"Borean" languages than me.
Salmoneus wrote: 14 Apr 2021 12:56I don't know - not knowing you personally and not being a psychologist - whether this falls within the technical definition of a compulsive behaviour, but the underlying mental pattern is clearly the same that underlies compulsive behaviours.
Sure, I do have some mild compulsive problems (they've gotten significantly lesser with time), but this literally has no relation to that because it's not about that kind of a thing. It's about wanting to make a conlang that's different from what I usually do, without it being unnaturalistic. If I had infinite time and energy, of course I could just spend several hours every day on looking for information about "non-Borean" languages and reading about them, but with my lack of education, reading linguistic papers sometimes makes things even more confusing if they're about things I don't yet understand. That's why it took years to wrap my head around ergativity (and it had to be explained with examples by two people at the same time for it to truly click [:$] ), and sometimes I'm still confused by it, so the complex noun class systems in Bantu languages and click consonants in Khoisan languages and whatnot are beyond me; I mean, of course I could make a conlang with noun classes and some kind of simple click consonants that I can pronounce, but it wouldn't "feel Bantu" or "feel Khoisan" with just that.

So, yes, it's possible that the "Boreanness" isn't a bad thing if it's not actually Boreanness or if Boreanness is so vast in phonology and morphology that it's meaningless. In that case, well, that's something you just taught me and now I won't have the problem anymore because it's not an issue anymore.
Salmoneus wrote: 14 Apr 2021 12:56This doesn't even get into the issue of why "not feeling Borean" - a property that almost no existing human languages have, and that has no objective value - is so important to you anyway.
In 99% of the conlangs I've started to work on, it hasn't mattered at all. In fact, it's been a good thing because due to the vastness of Boreanness, it's at least slightly connected to naturalism. Wanting to do the opposite in at least one conlang, without it becoming unnaturalistic, doesn't sound like a serious mental health problem or anything to me, but maybe it is...
Salmoneus wrote: 14 Apr 2021 12:56And this matters because if people want to do a thing, but adopt means and methods that ensure they cannot do that thing, or cannot satisfy themselves that they have done it even if they do it, then this will be a continual source of unhappiness for them.
It's not causing me unhappiness.
Salmoneus wrote: 14 Apr 2021 12:56covers absolutely all human languages
Well, that's probably a fair point, but the plurals were just one example. It's the overall vibe of the language that matters much more, like a /t/ plural in an otherwise "Khoisan-feeling" language wouldn't increase its Boreanness to any meaningful degree. You're probably right, though.
Salmoneus wrote: 14 Apr 2021 12:56In this case, you're happy to say that a word fits the "pattern" for butterfly words if it contains /p/, or /m/, or /n/, or /k/, or /u/, or /e/, or /l/, or /r/, or has four syllables or is formed by reduplication. You do see how this is a 'pattern' that everything fits into, right? And why not /a/, and /i/? (Pumarina, palometa and mariposa all have /a/, and two of them have /i/! Pattern!). And wait, the /u/ of 'pumarina' is also found in 'kupu-kupu'! So /u/ is part of the butterfly pattern too, right?
...I'm done trying to make sense if you can't tell what I was trying to say just because you latch onto how I said it.

Are you seriously going to argue that eg. perhonen and parvâne have no similarities? Both begin with a /pVr/ syllable where the vowel is a front unrounded vowel, followed by an open syllable with a back rounded vowel, followed by a syllable with an onset /n/ followed by a front unrounded vowel. Since vowels are more unstable than consonants and even in actual cognates in related languages they can vary a lot, practically it can be said that the only difference is that /h/ and /v/ aren't similar, and that the Finnish has a final /n/. Obviously these two words aren't related to each other, but those similarities are pretty clear and came to be relatively recently. To me that seems indicative of some kind of deep-running sound symbolism or whatever that governs how languages do things.
Salmoneus wrote: 14 Apr 2021 12:56Because you have set yourself a criterion that it is literally impossible to fulfill.
Maybe, alright, whatever.
Salmoneus wrote: 14 Apr 2021 12:56And yet you don't recognise that your belief being, as you call it, 'crackpottery', means you shouldn't hold that belief.
Fucking hell, I thought you've seen enough of my posts to know I self-deprecate as a joke. I don't actually think it's crackpottery, and there's no reason to think it is because sound symbolism is a thing that exists and is generally acknowledged to exist.
Davush
greek
greek
Posts: 670
Joined: 10 Jan 2015 14:10

Re: (Conlangs) Q&A Thread - Quick questions go here

Post by Davush »

Probably getting off topic, but my take:

1) Common grammatical/morphological categories are probably more likely to be represented by the overall more-common phonemes cross-linguistically (not just in "Borean") simply because those categories mean their representative phonemes will appear a lot, and intuitively they would tend towards those that require less articulatory effort, hence the overrepresentation of /t s n m l/ etc. in plurals/whatever.

2) Sound symbolism of the form "big/small", "dangerous/pleasant" is very basic and can be associated with sounds found in nature to an extent (I don't know much in depth about sound symbolism, but i this is how I assume it fundamentally works).

3) Beyond that, i.e., believing languages (and therefore people) display an inherent (but arbitrary and not-well defined) sound-association for very specific words like 'butterfly' implies some sort of deep (but unconscious) linguistic bio-program determinism, where all words have an underlying set of sounds associated with them and will surface at varying rates...
User avatar
sangi39
moderator
moderator
Posts: 3024
Joined: 12 Aug 2010 01:53
Location: North Yorkshire, UK

Re: (Conlangs) Q&A Thread - Quick questions go here

Post by sangi39 »

So, it seems that things have started getting heated and personal within the responses to the most recent "quick question". I suggest that everyone step back from it now before it goes further down hill, because I don't think it's going to lead to any sort of progressive discussion the way it's going right now.
You can tell the same lie a thousand times,
But it never gets any more true,
So close your eyes once more and once more believe
That they all still believe in you.
Just one time.
User avatar
Omzinesý
mongolian
mongolian
Posts: 4079
Joined: 27 Aug 2010 08:17
Location: nowhere [naʊhɪɚ]

Re: (Conlangs) Q&A Thread - Quick questions go here

Post by Omzinesý »

eldin raigmore wrote: 12 Apr 2021 22:18 @Omzinesy:
Does having ejective consonant phonemes predict having aspirated consonant phonemes? It the correlation strong?
If yes, should it be explained just so that ejectives are such atypical and hard-to-pronounce sounds that other distinctive features are first utilized? Or is there some deeper connection to do with articulatory phonetics?
You could look up on UPSID or PHOIBLE how many languages have both, or either one alone, or neither, and calculate the chi-squared statistic yourself and look up its significance in a statistics handbook (CRC or something).
That will tell you whether or not the correlation you’re asking about is truly significant.
If not, there’s nothing to explain.
If so, I have no idea what the explanation is.
Doesn't seem to be an easy task.
The data frames aren't in a very handy format.
My meta-thread: viewtopic.php?f=6&t=5760
User avatar
DesEsseintes
mongolian
mongolian
Posts: 4331
Joined: 31 Mar 2013 13:16

Re: (Conlangs) Q&A Thread - Quick questions go here

Post by DesEsseintes »

Omzinesý wrote: 12 Apr 2021 20:42 This is a natlang question but it seems this section is more active.

Does having ejective consonant phonemes predict having aspirated consonant phonemes? It the correlation strong?
If yes, should it be explained just so that ejectives are such atypical and hard-to-pronounce sounds that other distinctive features are first utilized? Or is there some deeper connection to do with articulatory phonetics?

WALS didn't have data on aspiration. It isn't probably weird enough.
I know this isn’t what you are asking for, but here are some thoughts.

I know of a relatively large number of languages that only contrast ejectives with one other series of stops.

The Sahaptian languages, Sahaptin and Nez Perce, contrast a series of plain stops with ejectives. In Nez Perce at least, plain stops are unaspirated but have lightly aspirated allophones when they occur before other stops in clusters.

The South Wakashan languages, Nuuchahnulth and Makah, are similar and allophonic aspiration of plain stops also occurs in coda positions. Makah has also innovated voiced stops, but there is no direct aspiration contrast.

Of the Salish languages, at least Bella Coola contrasts ejectives and aspirated stops, but no plain stops. I don’t know the details, but there is certainly not a phonemic aspiration contrast.

The Tsimshianic and Chimakuan languages have similar systems.

Although I know next to nothing about South American languages, a quick SAPhon search yields some 15 languages that have plain and ejective stops, but no aspirated stops, listed. Some of these also have voiced stops. Edit: SAPhon lists a total of 26 languages with ejective /t’/; since 15 of those do not have aspirated stops, that is surely statistically significant, if accurate?

That’s quite a lot of languages, which suggests to me that an aspiration contrast is certainly not a prerequisite for ejectives.
Edit: Another area worth looking at is the Caucasus. Adyghe seems to have a three-way contrast: voiced-voiceless-ejective
User avatar
Dormouse559
moderator
moderator
Posts: 2945
Joined: 10 Nov 2012 20:52
Location: California

Re: (Conlangs) Q&A Thread - Quick questions go here

Post by Dormouse559 »

Omzinesý wrote: 15 Apr 2021 12:31
eldin raigmore wrote: 12 Apr 2021 22:18 @Omzinesy:
Does having ejective consonant phonemes predict having aspirated consonant phonemes? It the correlation strong?
If yes, should it be explained just so that ejectives are such atypical and hard-to-pronounce sounds that other distinctive features are first utilized? Or is there some deeper connection to do with articulatory phonetics?
You could look up on UPSID or PHOIBLE how many languages have both, or either one alone, or neither, and calculate the chi-squared statistic yourself and look up its significance in a statistics handbook (CRC or something).
That will tell you whether or not the correlation you’re asking about is truly significant.
If not, there’s nothing to explain.
If so, I have no idea what the explanation is.
Doesn't seem to be an easy task.
The data frames aren't in a very handy format.
This PHOIBLE client makes searching a bit easier. Unfortunately, you can't directly filter for aspiration, but +spread_glottis selects aspirated consonants, /h/, and breathy-voice segments. Here are some search terms:
  • both ejectives and aspiration: any +ejective any +spread_glottis and
  • only ejectives: any +ejective no +spread_glottis and
  • only aspiration: no +ejective any +spread_glottis and
  • neither: no +ejective no +spread_glottis and
User avatar
WeepingElf
greek
greek
Posts: 531
Joined: 23 Feb 2016 18:42
Location: Braunschweig, Germany
Contact:

Re: (Conlangs) Q&A Thread - Quick questions go here

Post by WeepingElf »

AFAIK, the Georgian plain (i.e., neither voiced nor ejective) stops are phonetically aspirated, at least in some contexts. And in the Georgian alphabet, whose order is clearly based on that of Greek, the letters used to write them take the places of the Greek aspirated stop letters, while the letters for the ejectives take the places of the Greek plain voiceless stop letters.
... brought to you by the Weeping Elf
My conlang pages
Salmoneus
MVP
MVP
Posts: 3030
Joined: 19 Sep 2011 19:37

Re: (Conlangs) Q&A Thread - Quick questions go here

Post by Salmoneus »

WeepingElf wrote: 15 Apr 2021 18:46 AFAIK, the Georgian plain (i.e., neither voiced nor ejective) stops are phonetically aspirated, at least in some contexts. And in the Georgian alphabet, whose order is clearly based on that of Greek, the letters used to write them take the places of the Greek aspirated stop letters, while the letters for the ejectives take the places of the Greek plain voiceless stop letters.
Conversely, the Georgian ejectives are apparently just unaspirated voiceless stops in many positions.
Nortaneous
greek
greek
Posts: 674
Joined: 14 Aug 2010 13:28

Re: (Conlangs) Q&A Thread - Quick questions go here

Post by Nortaneous »

Dormouse559 wrote: 15 Apr 2021 18:43 This PHOIBLE client makes searching a bit easier. Unfortunately, you can't directly filter for aspiration, but +spread_glottis selects aspirated consonants, /h/, and breathy-voice segments.
+spread_glottis;+consonantal;-periodic_glottal_source will select only aspirated consonants (and breathy-voiced segments entered into PHOIBLE incorrectly)
Post Reply