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

A forum for all topics related to constructed languages
User avatar
Omzinesý
runic
runic
Posts: 3545
Joined: 27 Aug 2010 08:17
Location: nowhere [naʊhɪɚ]

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

Post by Omzinesý »

Znex wrote: 08 Sep 2022 18:08 the hypothetical Temematic and some related hypothetical substrate languages, which I have a soft spot for... [:3]
What are they by the way?
My meta-thread: viewtopic.php?f=6&t=5760
User avatar
WeepingElf
sinic
sinic
Posts: 288
Joined: 23 Feb 2016 18:42
Location: Braunschweig, Germany
Contact:

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

Post by WeepingElf »

Omzinesý wrote: 15 Sep 2022 16:09
Znex wrote: 08 Sep 2022 18:08 the hypothetical Temematic and some related hypothetical substrate languages, which I have a soft spot for... [:3]
What are they by the way?
"Temematic" is a hypothetical extinct IE language somewhere in eastern Europe for which the Austrian linguist Georg Holzer claims to have found evidence in Slavic words of unknown etymology. The language would have voiced the PIE voiceless stops, and changed the PIE breathy-voiced stops into plain voiceless stops instead, among other changes. For instance, svobodi 'free' is claimed to be such a loanword, which would be ultimately from PIE *swo-pot- 'self-dominated'. Holzer lists 45 such items. Most relevant scholars remain sceptical.
... brought to you by the Weeping Elf
Nortaneous
greek
greek
Posts: 663
Joined: 14 Aug 2010 13:28

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

Post by Nortaneous »

Sequor wrote: 08 Sep 2022 17:48 There's a prefix in Tibetan (marking animal nouns among other things, IIRC) that alternates d- with g- depending on the following consonant, reflecting dissimilation (I don't know Tibetan but let's imagine d-gik but g-thom).
The Tibetan preinitials d- and g- are in entirely complementary distribution, at least for the Old Tibetan period according to Hill's phonological description. g- appears before coronals, otherwise d-. There's a similar alternation of final -s/-d, with -d appearing after /r l n/ (coronal codas that aren't /d/) and -s appearing otherwise.

IIRC the d-/g- merger is secondary and these were originally separate preinitials. I think this even shows up in words beginning with resonants (in which the prefix would take initial rather than preinitial position).
Omzinesý wrote: 09 Sep 2022 16:59
Sequor wrote: 08 Sep 2022 17:48 /b d/ ~ /m n/ could alternate, maybe as a remnant of what used to be phonemic nasalized vowels, or consonant clusters with a nasal, or nasalization harmony between a root with /m n/ and an affix with /-b -d/.
That is a good and simple idea that I didn't occur to me.
Does the alteration appear in some language?
Word-level nasalization harmony with various blocking rules appears in (to name the two I can remember) Guarani and Enggano
User avatar
LinguistCat
sinic
sinic
Posts: 288
Joined: 06 May 2017 07:48

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

Post by LinguistCat »

Omzinesý wrote: 14 Sep 2022 21:38 I continue with my question about morphophonological alternations.

If the proto-lang had prenasalized voiced stops

mb nd

Could there dissimilation that they become voiced stops if the following syllable has nasals but nasals in other environments?

ambana -> abana, ambata -> amata

Or they become nasals if there are stops in the following syllable and voiced stops in other environments?

ambara -> abara, ambata -> amata

Does one of those ideas sound plausible?
I don't see why either of these couldn't work. I personally find the first form of dissimilation more interesting.
User avatar
Znex
roman
roman
Posts: 947
Joined: 12 Aug 2013 14:05
Location: Australia

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

Post by Znex »

WeepingElf wrote: 15 Sep 2022 17:58
Omzinesý wrote: 15 Sep 2022 16:09
Znex wrote: 08 Sep 2022 18:08 the hypothetical Temematic and some related hypothetical substrate languages, which I have a soft spot for... [:3]
What are they by the way?
"Temematic" is a hypothetical extinct IE language somewhere in eastern Europe for which the Austrian linguist Georg Holzer claims to have found evidence in Slavic words of unknown etymology. The language would have voiced the PIE voiceless stops, and changed the PIE breathy-voiced stops into plain voiceless stops instead, among other changes. For instance, svobodi 'free' is claimed to be such a loanword, which would be ultimately from PIE *swo-pot- 'self-dominated'. Holzer lists 45 such items. Most relevant scholars remain sceptical.
There are similar substrate languages proposed by Garnier and Sagot in Italy and Greece, though to some level Temematic would be distinct to these. Some key differences include this Mediterranean substrate being a centum language (Temematic is satem), some word-initial voicings of the breathy-voiced series, and no complete merger of the voiceless series and voiced series because....! The Mediterranean substrate is also proposed to have a Verner-like lenition rule where voiceless stops following unstressed syllables are voiced/lenited, but voiceless stops following stressed syllables are geminated!

Other than the shared fortition of the breathy-voiced series, the two substrates also share some local metathesis in a movement from PIE closed syllables to open syllables and an accentual shift.
:eng: : [tick] | :grc: : [:|] | :chn: :isr: :wls: : [:S] | :deu: :ell: :rus: : [:x]
Conlangs: Hawntow, misc.
User avatar
WeepingElf
sinic
sinic
Posts: 288
Joined: 23 Feb 2016 18:42
Location: Braunschweig, Germany
Contact:

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

Post by WeepingElf »

Znex wrote: 17 Sep 2022 08:14 There are similar substrate languages proposed by Garnier and Sagot in Italy and Greece, though to some level Temematic would be distinct to these. Some key differences include this Mediterranean substrate being a centum language (Temematic is satem), some word-initial voicings of the breathy-voiced series, and no complete merger of the voiceless series and voiced series because....! The Mediterranean substrate is also proposed to have a Verner-like lenition rule where voiceless stops following unstressed syllables are voiced/lenited, but voiceless stops following stressed syllables are geminated!

Other than the shared fortition of the breathy-voiced series, the two substrates also share some local metathesis in a movement from PIE closed syllables to open syllables and an accentual shift.
Where can I find more about these substratum languages? This sounds very interesting, and be it as material for conlangs.
... brought to you by the Weeping Elf
User avatar
Znex
roman
roman
Posts: 947
Joined: 12 Aug 2013 14:05
Location: Australia

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

Post by Znex »

WeepingElf wrote: 17 Sep 2022 16:47Where can I find more about these substratum languages? This sounds very interesting, and be it as material for conlangs.
This is the paper in question: https://www.academia.edu/79112163/A_sha ... and_Italic
:eng: : [tick] | :grc: : [:|] | :chn: :isr: :wls: : [:S] | :deu: :ell: :rus: : [:x]
Conlangs: Hawntow, misc.
User avatar
WeepingElf
sinic
sinic
Posts: 288
Joined: 23 Feb 2016 18:42
Location: Braunschweig, Germany
Contact:

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

Post by WeepingElf »

Znex wrote: 17 Sep 2022 16:56
WeepingElf wrote: 17 Sep 2022 16:47Where can I find more about these substratum languages? This sounds very interesting, and be it as material for conlangs.
This is the paper in question: https://www.academia.edu/79112163/A_sha ... and_Italic
Thank you. First impression: not implausible, but not very compelling either. There is only so much that can be deduced from so few words. At least, they don't claim that this was the only substratum in the languages discussed. I wonder whether the North Picene language, which with forms such as rotnem vs. rotneš, IMHO has some Indo-European ring to it, could be a residue of such a substratum.
... brought to you by the Weeping Elf
User avatar
eldin raigmore
korean
korean
Posts: 6094
Joined: 14 Aug 2010 19:38
Location: SouthEast Michigan

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

Post by eldin raigmore »

“WeepingElf” wrote: "Temematic" is a hypothetical extinct IE language somewhere in eastern Europe for which the Austrian linguist Georg Holzer claims to have found evidence in Slavic words of unknown etymology. The language would have voiced the PIE voiceless stops, and changed the PIE breathy-voiced stops into plain voiceless stops instead, among other changes. For instance, svobodi 'free' is claimed to be such a loanword, which would be ultimately from PIE *swo-pot- 'self-dominated'. Holzer lists 45 such items. Most relevant scholars remain sceptical.
Theories that most relevant scholars remain skeptical about, are very good grist for the mills of conlangers!
User avatar
Ahzoh
mongolian
mongolian
Posts: 4134
Joined: 20 Oct 2013 02:57
Location: Canada

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

Post by Ahzoh »

I'm stuck in a predicament. I originally only had masculine and feminine, but I came up with a third gender, the inanimate/neuter. I don't entirely know how this new noun class could come into existence from a noun class binary, so it creates problems with how nouns in this gender concord with pronouns and suffixes.

Essentially I now have this (in order: masculine, feminine, inanimate, masculine, feminine, inanimate):
Image
Every noun has a final vowel that determines its class (u = feminine, i = masculine, a = inanimate)

But I only have possessive suffixes and subject suffixes for masculine and feminine only:
Image
These affixes are intended to be morphologically the oldest features, older than even the case system.

There are three possible ways for the two genders to split into three:
Image
Possibility A is most sensible and elegant, especially if one could say the masculine is a reanalysis of animate suffixes combined with some i-colouring affix. But it would result in the problem of needing new masculine suffixes because the possessive and subject affixes would originally be animate and inanimate.

Possibility B can work by having inanimate/neuter nouns concord with feminine possessive and subject suffixes, but there isn't anything in the morphology that could suggest the feminine gender splitting into two. It would also mean that the possessive and subject affixes were originally masculine and feminine and now new inanimate suffixes must be created.

There's Possibility C where both genders split into animate and inanimate (so 4 genders) and the feminine and masculine inanimate nouns merge into one gender, but that still leads to the same issues as possibility B.

Now, I could just take the existing affixes and make versions where the vowel is <a>: (-ma, -man, -ta/-sa, -tan/-san, -ka, -kan), but that doesn't really make sense and it would create a lot of awkward asymmetry (when would you ever refer to an inanimate object with the second person?) and I don't really want to have a pyramid of gender distinctions (first person not distinguishing gender, 2nd person only distinguishing two, third person distinguishing three). So the solution is undesirable and maybe not even feasible. Making it so the suffixes don't make distinctions based on gender is also undesirable.
Image Śād Warḫālali (Vrkhazhian) [ WIKI | CWS ]
User avatar
Znex
roman
roman
Posts: 947
Joined: 12 Aug 2013 14:05
Location: Australia

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

Post by Znex »

Ahzoh wrote: 17 Sep 2022 22:33 I'm stuck in a predicament. I originally only had masculine and feminine, but I came up with a third gender, the inanimate/neuter. I don't entirely know how this new noun class could come into existence from a noun class binary, so it creates problems with how nouns in this gender concord with pronouns and suffixes.
You might already know about the relatively new development of animacy in Slavic nominals, as a secondary category to M-F-N gender (and plurals are their own "gender" too). Although in the case of Slavic, nouns don't receive new suffixes depending on their animacy, but rather how the nominal morphology developed (ie. the various different declensions) was used and reappropriated to express a new category.

On the other hand, the feminine gender in PIE is generally agreed to be focused on the feminine suffix -(e)H2, to which is then affixed the nominal case and number endings; not that all feminine nouns end up receiving the suffix, and many even remain similar in appearance to their masculine and neuter counterparts, but the feminine suffix becomes very productive particularly as a feminine marker. In later languages, feminine nouns become more and more distinctly marked by the feminine suffix.

Just some ideas from natlangs, off the top of my head.
:eng: : [tick] | :grc: : [:|] | :chn: :isr: :wls: : [:S] | :deu: :ell: :rus: : [:x]
Conlangs: Hawntow, misc.
User avatar
WeepingElf
sinic
sinic
Posts: 288
Joined: 23 Feb 2016 18:42
Location: Braunschweig, Germany
Contact:

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

Post by WeepingElf »

eldin raigmore wrote: 17 Sep 2022 22:23
“WeepingElf” wrote: "Temematic" is a hypothetical extinct IE language somewhere in eastern Europe for which the Austrian linguist Georg Holzer claims to have found evidence in Slavic words of unknown etymology. The language would have voiced the PIE voiceless stops, and changed the PIE breathy-voiced stops into plain voiceless stops instead, among other changes. For instance, svobodi 'free' is claimed to be such a loanword, which would be ultimately from PIE *swo-pot- 'self-dominated'. Holzer lists 45 such items. Most relevant scholars remain sceptical.
Theories that most relevant scholars remain skeptical about, are very good grist for the mills of conlangers!
Right! For instance, I use a homegrown hypothesis about a lost branch of IE in Bronze Age western Europe, which most IEists would surely remain skeptical about, in my main conlang project. Also, some ideas about the structure of Early PIE (the common ancestor of Late PIE from which the known non-Anatolian IE languages descend, Anatolian and the hypothetical lost branch).
... brought to you by the Weeping Elf
User avatar
LinguoFranco
greek
greek
Posts: 537
Joined: 20 Jul 2016 17:49
Location: U.S.

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

Post by LinguoFranco »

How do you evolve stress in a conlang?

I want my conlang to have a lexical pitch accent a la Japanese or Ancient Greek. I'm thinking about having it start out with a standard stress accent, but overtime the things that distinguishes the stressed syllable from unstressed (like length) disappear overtime, leaving only pitch.

I tried to making a register tone language as a starting point for pitch accent, but I find tonal system more complicated than pitch accent to be confusing.
Salmoneus
MVP
MVP
Posts: 2674
Joined: 19 Sep 2011 19:37

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

Post by Salmoneus »

Your four main options for evolving stress are:
- stress appears in a language that didn't have it*
- stress stays where it is
- stress moves
- stress goes away.

*all languages have stress, but in some it's not very important so they may as well not have it. And in some languages stress is not tied to words, but to larger phrases.

When stress moves, it can move linearly (backward or forward), or it can jump (perhaps because a secondary stress has become primary).

Non-phonemic stress is obviously more likely to move around, because there's no reason for it not to.

Stress can sometimes be defined metrically. Changes in that definition, or changes in the metre, can result in stress shifts.

Stress is often defined by morae, rather than by syllables (that is, stress falls in the syllable containing the accented mora).

It's possible that stress can also be directly attracted by heavy syllables. Some people I think deny this and claim that all examples can be explained metrically, but I'm not sure whether that's convincing.

Analogy can be a powerful motivation for stress shifts. Exceptions (providing the seeds of analogy) can come from sound changes (eg loss of weight) or reanalysis of word boundaries (eg affixation of unstressed particles into positions that 'should' be stressed).

Stress can develop into pitch and vice versa. However, it is also common (perhaps more so?) for pitch to develop independently of stress. A language can have both stress and pitch; or, if pitch becomes phonemic, stress can become less salient.

Pitch can become phonemic in at least three ways independent of stress:
- certain consonants naturally lower or raise adjacent pitches, and if those consonants are lost, or neutralised, the pitch can remain phonemically (eg voiced consonants are often depressors of following pitch, so loss of voicing contrasts yields tonal/pitch phonemicity)
- long vowels often have a distinct pitch contour, so if length is lost pitch can become phonemic
- pitch contour is often defined by word shape, so if word shape changes in some cases (eg some final syllables are lost) pitch can remain phonemically
- I'm sure there are other ways too.
User avatar
LinguistCat
sinic
sinic
Posts: 288
Joined: 06 May 2017 07:48

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

Post by LinguistCat »

I don't remember if it was here or the ZBB, but when I was more actively working on my Japanese-based cat language's sound changes, someone mentioned that Japanese morae are pretty strongly bound. So that the C and the V of CV sequences tend to be affected as a unit for the most part by sound changes. However, part of my ideas of how things would work would be a group of foreign language speakers arriving in Japan, picking up the language with some mistakes/innovations, and use this altered variation as the base for future sound and grammar changes. So I'm not sure that the CV sequences would inherently be closely tied in that variation of the language.

Does that seem plausible? I'd also like the language to become less recognizable as a relative of Japanese before borrowing back more Standard Japanese attributes, but I only have 1,000-1500 years to work in, depending on the exact timeline I give myself.
Nortaneous
greek
greek
Posts: 663
Joined: 14 Aug 2010 13:28

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

Post by Nortaneous »

LinguoFranco wrote: 20 Sep 2022 02:57 How do you evolve stress in a conlang?

I want my conlang to have a lexical pitch accent a la Japanese or Ancient Greek. I'm thinking about having it start out with a standard stress accent, but overtime the things that distinguishes the stressed syllable from unstressed (like length) disappear overtime, leaving only pitch.

I tried to making a register tone language as a starting point for pitch accent, but I find tonal system more complicated than pitch accent to be confusing.
What do you mean by "lexical pitch accent"? How did the feature you're looking at develop in Greek? How did similar features develop in other languages that have them, like Latvian (a large number of obscure developments) or the Germanic and Celtic varieties with contrasts in pitch contours on stressed syllables? (IIRC mostly syllable-counting processes phonemicized by morphological differences and sound changes that affected the number of syllables in the stem)
Post Reply