Can you try to reconstruct a proto-conlang?

For the discussion of language mechanics, grammar, vocabulary, trends, and other such linguistic topics, in english and other languages.

Moderators: gmalivuk, Moderators General, Prelates

User avatar
Caprice
Posts: 31
Joined: Mon Apr 20, 2015 12:23 am UTC

Can you try to reconstruct a proto-conlang?

Postby Caprice » Sat Jan 23, 2016 2:57 pm UTC

So I wrote a passage in a conlang, then added different sound changes to it to see if you could reconstruct the original conlang. Here goes (Note: The first two languages are more similar than any others):

Language 1:
Plazia Uri

Aasa
Lo hambu agelan
Suge
Opago plaziale aaten

Language 2:
Playa Uri

Attsa
Lo avvu ajelan
Suye
Opajo playale atten

Language 3:
Plizī ūhi

Ātsa
Hambulo agelan
Tuhe
Pago plizīje āten

Language 4:
Plazia Uxari

Axatsa
La xampu akilan
Tsuri
Paka ap plasia li axatin

Language 5:
Pluzi Ux

Axts
Xambul agela
Tsur
Pag op pluzil axte

Language 6:
Plzi uhgi

Ahsa
Luam ajelar
Suge
Pajo plzi liahter

Transcription:

Language 1:
plazja uɾi

aːsa
lo hambu agelan
suge
opago plazjale aːten

Language 2:
plaja uri

attsa
lo avvu ajelan
suje
opajo plajale atten

Language 3:
plizi: u:hi

a:tsa
hambulo agelã:
tuhe
pago plizi:je a:ten

Language 4:
plaʑa uɣaɾi

aɣatsa
la ɣambu agilan
tsuri
paga ab blaʑa li aɣadin

Language 5:
pluzi ux

axts
xambul agela:
tsur
pag op pluzil axte:

Language 6:
pəlzi uhəgi

ahəsa
lwam ajela:
suge
pajo pəlzi liahəte:
Last edited by Caprice on Sun Feb 07, 2016 4:19 am UTC, edited 2 times in total.
Musical Chronicler of the OTT (bestowed by yappobiscuits, NP(?)2297)

User avatar
ThirdParty
Posts: 347
Joined: Wed Sep 19, 2012 3:53 pm UTC
Location: USA

Re: Can you try to reconstruct a proto-conlang?

Postby ThirdParty » Sun Jan 24, 2016 10:18 am UTC

Fun! I'll play...

The first thing I notice is that Language 5 seems to have dropped the final letter of every word. I think that's a mutation, so I reconstruct:
Spoiler:
Language 5:
Pluzia Uxri

Axtsa
Xambulo agelan
Tsure
Pago op pluzile axten

Next, let's clean up the soft consonants: Language 2 has suffered an "mb"->"vv" mutation and a "z"->"y" mutation; Language 3 has suffered an "r"->"h" mutation; Language 6 has suffered an "r"->"g" mutation and (during or after) an "n"->"r" mutation. Language 4 has suffered a "mb"->"mp" mutation possibly related to one of its "g"s mutating into a "k". Languages 2 and 6 both coincidentally suffered a "g"->"j" mutation (I'm pretty sure it happened in that direction, despite the coincidence, since the alternative is that four languages suffered "j" to "g" mutations).

The word "hambu" is interesting. It's spelled without an initial consonant in Languages 2 and 6, with an "h" in Languages 1 and 3, and with an "x" in Languages 4 and 5. None of the languages uses both an "h" and an "x", so I conclude that those letters are interchangeable. Since the letter is sometimes dropped entirely, and since it makes more sense to drop an "h" than to drop an "x", I'm going to reconstruct with the letter "h".

Oh, and while we're there, this pronoun-looking particle "lo" is dancing around "hambu". Languages 1, 2, and 4 place it before the word; Language 6 attaches it to the start of the word; and Languages 3 and 5 attach it to the end of the word. Since it is so mobile, I'm going to treat it as a separate word; since 4/6ths of the languages place it before, I'm also going to place it before. I'm also going to separate out the "le" word that can't seem to decide what to attach to (and fix its mutation into "je" in Language 3).

So now I have:
Spoiler:
Language 1:
Plazia Uri

Aasa
Lo hambu agelan
Suge
Opago plazia le aaten

Language 2:
Plaza Uri

Attsa
Lo hambu agelan
Suze
Opago plaza le atten

Language 3:
Plizī ūri

Ātsa
Lo hambu agelan
Ture
Pago plizī le āten

Language 4:
Plazia Uhari

Ahatsa
La hambu agilan
Tsuri
Paga ap plazia li ahatin

Language 5:
Pluzia Uhri

Ahtsa
Lo hambu agelan
Tsure
Pago op pluzia le ahten

Language 6:
Plzi uhri

Ahsa
Lu ham agelan
Sure
Pago plzi li ahten

Next I'm going to attack the vowels. It's helpful that Language 3 distinguishes eight different ones, which do seem to be distinguished in various ways by all the languages. (For example, there's a "long a" represented as "aa" in Language 1, "at" in Language 2, "ā" in Language 3, "aha" in Language 4, and "ah" in Languages 5 and 6.)

Even though diacritics only appear in one language, I'm going to use them at least temporarily while I clean up the vowel shifts. I'm also going to introduce an apostrophe for the first vowel in "Pl'zī", because there's no agreement about what it should be.
Spoiler:
Language 1:
Pl'zī Ūri

Āsa
Lo hambu agelan
Suge
Opago pl'zī le āten

Language 2:
Pl'zī Ūri

Ātsa
Lo hambu agelan
Suze
Opago pl'zī le āten

Language 3:
Pl'zī ūri

Ātsa
Lo hambu agelan
Ture
Pago pl'zī le āten

Languages 4 and 5:
Pl'zī Ūri

Ātsa
Lo hambu agelan
Tsure
Pago op pl'zī le āten

Language 6:
Pl'zī ūri

Āsa
Lo ham agelan
Sure
Pago pl'zī le āten

Now there's just some small cleanups left. Language 6 lost the "bu" from its "hambu". The closely-related Languages 1 and 2 are the only ones that put an "o" in front of "pago", so I'm going to delete that. The two closely-related languages also have something other than an "r" in the word "Sure", but can't even agree about what it should be; I'm going to go with "r".

Four out of six languages agree that the second word should be capitalized.

I'm a little puzzled by the "op" particle on the fifth line. My linguistic intuition is that it's the sort of thing that would be unnatural to add but natural to drop. However, only two of the six languages have it, and they're two that are beginning to seem relatively-closely related. Also, the way it has exactly same sounds as the surrounding words would be an awfully big coincidence if it were a genuine word. I'm going to delete it.

Four out of six languages agree that the second line is "Ātsa", not "Āsa"; also, "long a" never occurs elsewhere except before a "t", and one language even represented it as "at": so I'm going to go with the majority. The first consonant of the fourth line is harder: both languages that had "Āsa" also have "Sure", but so does one of the languages that had "Ātsa", and another one of the "Ātsa" languages has "Ture". With four out of six agreeing that it's a different sound from the one in "Ātsa", I'm going to go with "Sure".
Spoiler:
Pl'zī Ūri

Ātsa
Lo hambu agelan
Sure
Pago pl'zī le āten

Now I just need to return to the orthographic question I deferred earlier when I adopted the diacritic marks. Only one language used diacritics, and diacritics tend to originate as abbreviations for longer strings (or sometimes as substitutes for non-Latin characters, but the "lo"/"le" particles and the "x"/"h" equivalency make this language feel sufficiently Latiny that we shouldn't need non-Latin characters), so I doubt that the diacritics were present in the ancestral language.

Languages 1, 4, and arguably 5 all use "ia" for "long i", and the language that used diacritics did not use any diphthongs at all, so I suspect that "long i" should be "ia".

I never decided what the vowel in the first word should be. The votes are three "a"s (unfortunately including both of the two languages I'm told are closely related), one "i", one "u", and one omission. So I'm going to go with "a".

Languages 4, 5, and 6 all use "ah" for "long a" and "uh" for "long u". That seems like an acceptable orthography. However, I'm not sure whether there should be an "a" after the "h". Languages 1 and 4 both have an "a" after "long a" (although Language 1 has it instead of the "h" rather than after the "h"), and if I'm right about "Plazia" then we know that Language 6 has a tendency to drop "a"s sometimes. The macrons in Language 3 could easily have originated as scribal abbreviations for an "a". But on the other hand, Languages 2 and 5 definitely don't have extra "a"s after "long a", and Languages 1, 2, and 5 all lack it before "long u". Also, if there was a "ua" diphthong anywhere, it'd be surprising that it never turned into "wa" but instead consistently degenerated into a single vowel. So on balance I think I'm better off not including the "a".

So here's what I've ended up with as my reconstruction:
Spoiler:
Plazia Uhri

Ahtsa
Lo hambu agelan
Sure
Pago plazia le ahten

How did I do?

User avatar
Lazar
Landed Gentry
Posts: 2151
Joined: Tue Dec 29, 2009 11:49 pm UTC
Location: Massachusetts

Re: Can you try to reconstruct a proto-conlang?

Postby Lazar » Mon Jan 25, 2016 3:41 am UTC

This is a cool concept. But it would help if you could show phonetic transcriptions – I'd like to see where the stresses are, and what the quality of r is. Based on what's here, my guess would be:

Spoiler:
Plzia Uxagi

Axatsa
Lo xampu akelan
Tuge
Pago op plzia le axaten

ThirdParty wrote:Language 4 has suffered a "mb"->"mp" mutation possibly related to one of its "g"s mutating into a "k".

I wouldn't be so sure. It's much more common to voice stops in intervocalic and postnasal position than to devoice them.

The word "hambu" is interesting. It's spelled without an initial consonant in Languages 2 and 6, with an "h" in Languages 1 and 3, and with an "x" in Languages 4 and 5. None of the languages uses both an "h" and an "x", so I conclude that those letters are interchangeable. Since the letter is sometimes dropped entirely, and since it makes more sense to drop an "h" than to drop an "x", I'm going to reconstruct with the letter "h".

Sound changes happen in stages and at varying rates – you don't have to imagine that there's only one step separating a proto-language from each daughter language. To me, this looks like an obvious chain shift, x > h > ∅. Both of these changes are very common. In contrast, fortition from [h] to [x] strikes me as implausible – I've never heard of any language doing it.

I'm a little puzzled by the "op" particle on the fifth line. My linguistic intuition is that it's the sort of thing that would be unnatural to add but natural to drop. However, only two of the six languages have it, and they're two that are beginning to seem relatively-closely related. Also, the way it has exactly same sounds as the surrounding words would be an awfully big coincidence if it were a genuine word. I'm going to delete it.

No, I think this morpheme is present in L1 and L2: it accounts for the initial vowel in "opago" and "opayo".

The first consonant of the fourth line is harder: both languages that had "Āsa" also have "Sure", but so does one of the languages that had "Ātsa", and another one of the "Ātsa" languages has "Ture". With four out of six agreeing that it's a different sound from the one in "Ātsa", I'm going to go with "Sure".

No, the initial consonant there can't be [s], because that couldn't account for the use of plosive or affricate forms in three of the five branches. Fortition of [s] is a highly unlikely thing to happen. Again, the most likely thing here would be a chain shift, t > ts > s, progressing at different rates in different branches.

The two closely-related languages also have something other than an "r" in the word "Sure", but can't even agree about what it should be; I'm going to go with "r".

I think you're looking at things too superficially: L1, L3 and L6 all agree that it's a guttural (dorsal or laryngeal) consonant, and L2's y is probably lenited from one as well. I don't see how an original rhotic would produce [g] and [h]. Original [g] seems most likely, especially if the r in L4 and L5 is guttural.

I never decided what the vowel in the first word should be. The votes are three "a"s (unfortunately including both of the two languages I'm told are closely related), one "i", one "u", and one omission. So I'm going to go with "a".

I'm having trouble imagining any vowel that could yield [a], [і] and [u] – those three are literally all over the place. My guess is that syllabic [l] is the original form, and that the daughter languages picked different epenthetic vowels to follow it.

Languages 4, 5, and 6 all use "ah" for "long a" and "uh" for "long u".

Well, I'm taking the h at face value as representing [h] – presumably lenited from [x]. Postvocalic [h] occurs in Persian and in some dialects of Spanish, for example.
Exit the vampires' castle.

User avatar
ThirdParty
Posts: 347
Joined: Wed Sep 19, 2012 3:53 pm UTC
Location: USA

Re: Can you try to reconstruct a proto-conlang?

Postby ThirdParty » Mon Jan 25, 2016 6:45 am UTC

Lazar: you make some excellent points about which directions changes are most likely to flow in. But I still don't like your "g"s and "t"s:

Lazar wrote:I think you're looking at things too superficially: L1, L3 and L6 all agree that it's a guttural (dorsal or pharyngeal) consonant. I don't see how an original rhotic would produce [g] and [h]. Original [g] seems most likely, especially if the r in L4 and L5 is guttural.
All six languages draw a distinction between the sound in "Uxari" and the sound in "Pago". To me it strains credibility that they would have all independently arrived at this distinction if it didn't exist in the source language. Here's a table of the words with sounds in the "k"/"g"/"r" family:

Code: Select all

       1   2   3   4   5   6
Uxa?i   r   r   h   r   -   g
Tu?e    g   y   h   r   r   g
Pa?o    g   j   g   g   g   j
a?elan  g   j   g   k   g   j
Languages 3 and 6 are particularly interesting. In 3, the sound in "Uxari" is glottal while the sound in "Pago" is uvular. In 6, the sound in "Uxari" is uvular while the sound in "Pago" is palatal. I've never heard of a consonant that moved forward before back vowels like "o" but backward before front vowels like "i".

If you want to spell them "Pajo"+"Uxagi" or "Pago"+"Uxahi", rather than "Pago"+"Uxari", fine. We're probably dealing with a contrast along the lines of /g/ vs. /ɢ/, and there's not really any good way to represent that in the Latin alphabet, so one orthography is as good as another. But I don't accept that they're the same sound.

Lazar wrote:
ThirdParty wrote:Language 4 has suffered a "mb"->"mp" mutation possibly related to one of its "g"s mutating into a "k".
It's much more common to voice stops in intervocalic and postnasal position than to devoice them.
...
ThirdParty wrote:The first consonant of the fourth line is harder: both languages that had "Āsa" also have "Sure", but so does one of the languages that had "Ātsa", and another one of the "Ātsa" languages has "Ture". With four out of six agreeing that it's a different sound from the one in "Ātsa", I'm going to go with "Sure".
No, the initial consonant there can't be [s], because that couldn't account for the use of plosive or affricate forms in three of the five branches. Fortition of [s] is a highly unlikely thing to happen. Again, the most likely thing here would be a chain shift, t > ts > s, progressing at different rates in different branches.
On the one hand, everything you say here is true.

On the other hand, let me get this straight. You think that the word-initial "t" in "Tuge" suffered lenition in five of six languages, the intervocalic "k" in "akelan" suffered lenition in (a different) five of six languages, but the intervocalic "t" in "axaten" suffered lenition in zero of six languages? That strikes me as unlikely.

Here's the table for the T's and S's:

Code: Select all

       1   2   3   4   5   6
?uge    s   s   t  ts  ts   s
Axa?a   -  ts  ts  ts  ts   s
axa?en  t   t   t   t   t   t

Do you see why I'm not satisfied with "Tuge"/"Axatsa"/"axaten"?

User avatar
Lazar
Landed Gentry
Posts: 2151
Joined: Tue Dec 29, 2009 11:49 pm UTC
Location: Massachusetts

Re: Can you try to reconstruct a proto-conlang?

Postby Lazar » Mon Jan 25, 2016 8:40 am UTC

ThirdParty wrote:All six languages draw a distinction between the sound in "Uxari" and the sound in "Pago". To me it strains credibility that they would have all independently arrived at this distinction if it didn't exist in the source language. Here's a table of the words with sounds in the "k"/"g"/"r" family:

Code: Select all

       1   2   3   4   5   6
Uxa?i   r   r   h   r   -   g
Tu?e    g   y   h   r   r   g
Pa?o    g   j   g   g   g   j
a?elan  g   j   g   k   g   j
Languages 3 and 6 are particularly interesting. In 3, the sound in "Uxari" is glottal while the sound in "Pago" is uvular. In 6, the sound in "Uxari" is uvular while the sound in "Pago" is palatal. I've never heard of a consonant that moved forward before back vowels like "o" but backward before front vowels like "i".

If you want to spell them "Pajo"+"Uxagi" or "Pago"+"Uxahi", rather than "Pago"+"Uxari", fine. We're probably dealing with a contrast along the lines of /g/ vs. /ɢ/, and there's not really any good way to represent that in the Latin alphabet, so one orthography is as good as another. But I don't accept that they're the same sound.

You're right, I overlooked that. (You mean velar rather than uvular, by the way.) I'm actually stumped by how to resolve this situation – your [g] vs. [ɢ] suggestion seems to make the most sense, although I'm not sure that OP had such "exotic" sounds in mind. Anyway, I'd write [ɢ] as q, which is used for that sound in the transliteration of Persian and some other languages.

But then there's still the problem of the medial consonant in "tu?e", which acts like q in L3, L4 and L6 but like g in L1 and like its own thing in L2. I guess we could say that L1 underwent the rule "[ɢ] > [r] before [і]; [ɢ] > [g] before [e]", while L2 underwent the rule "[ɢ] > [r] before [і]; [ɢ] > [y] before [e]".

On the other hand, let me get this straight. You think that the word-initial "t" in "Tuge" suffered lenition in five of six languages, the intervocalic "k" in "akelan" suffered lenition in (a different) five of six languages, but the intervocalic "t" in "axaten" suffered lenition in zero of six languages? That strikes me as unlikely.

Here's the table for the T's and S's:

Code: Select all

       1   2   3   4   5   6
?uge    s   s   t  ts  ts   s
Axa?a   -  ts  ts  ts  ts   s
axa?en  t   t   t   t   t   t

Do you see why I'm not satisfied with "Tuge"/"Axatsa"/"axaten"?

With "xampu akilan" I guess I might have to yield to the weight of all the other languages and accept that L4 underwent devoicing. But on the other hand, we then have to explain why L4 still has g in "paga". Maybe it only happens around high vowels.

With "?uge", I really don't see much alternative. Even if we discount crosslinguistic ideas about which direction things tend to go in, I think [s] and [ts] in the original form are still both ruled out. If the former, then in L3 [s] shifted all the way to [t] while [ts], an intermediate sound, remained unchanged – you'd expect those two to have merged. And if the latter, then we have to explain why ancestral [ts] underwent a split in three of the six languages. And again, although an unlikely change like fortition of [s] might occur somewhere (like how PIE *dwóH became Armenian erku), it wouldn't be arrived at independently by multiple branches of a family like this. We have precedent for [t] affricating before [u] in Japanese – which may have even been OP's inspiration.

On the other hand, we still have to explain why L2 has t>s in "suye" but ts>ts in "attsa". I suppose we could say that after the initial "[t] > [ts] before [u]" rule applied, a subsequent "[ts] > [s] before [u]" rule applied while leaving [ts] elsewhere untouched. It's inelegant, but it kinda works.

So I guess my current proposal is:

Spoiler:
Plzia Uxaqi

Axatsa
Lo xambu agelan
Tuqe
Pago op plzia le axaten
Exit the vampires' castle.

User avatar
Caprice
Posts: 31
Joined: Mon Apr 20, 2015 12:23 am UTC

Re: Can you try to reconstruct a proto-conlang?

Postby Caprice » Mon Jan 25, 2016 1:16 pm UTC

So you pointed out Lang 4's g in "paga." It's pronounced [paga], but the orthography should say "paka." I overlooked that, sorry. Transcriptions will come later when I can more easily copy and paste IPA characters.
Musical Chronicler of the OTT (bestowed by yappobiscuits, NP(?)2297)

User avatar
eSOANEM
:D
Posts: 3652
Joined: Sun Apr 12, 2009 9:39 pm UTC
Location: Grantabrycge

Re: Can you try to reconstruct a proto-conlang?

Postby eSOANEM » Mon Jan 25, 2016 3:13 pm UTC

This website may help you with that.

I'm not sure whether I can reasonably contribute to this, but I'll definitely be watching (and might possibly throw my conlang family in here once this has been more fully resolved).
my pronouns are they

Magnanimous wrote:(fuck the macrons)

User avatar
ThirdParty
Posts: 347
Joined: Wed Sep 19, 2012 3:53 pm UTC
Location: USA

Re: Can you try to reconstruct a proto-conlang?

Postby ThirdParty » Mon Jan 25, 2016 4:30 pm UTC

Caprice wrote:So you pointed out Lang 4's g in "paga." It's pronounced [paga], but the orthography should say "paka." I overlooked that, sorry.
Okay, so it originated as a "k". That more-or-less solves the g problem.

Caprice wrote:Transcriptions will come later when I can more easily copy and paste IPA characters.
Actually, I think the puzzle was more fun without them; they make it too easy.

Lazar wrote:we still have to explain why L2 has t>s in "suye" but ts>ts in "attsa".
I'm not concerned. If "t" decayed to "ts" without pushing the "ts" to "s" (which it easily could, if the "ts" was being conceptualized in speakers' minds as two consonants rather than one), then it seems to me totally natural that "ts" could decay to "s" in word-initial position but not when following a "t".

But I am still bothered by the "t" in "axaten" not morphing into "d" when the "k" in "akelan" morphed into a "g". Are you sure the second "a" in "axaten" belongs there? It seems to me more likely that Language 4 systematically inserts "a" in between "x" and a following consonant to avoid clusters which its speakers find awkward. Note that "a" is the vowel that Language 4 tends to insert places, e.g. into "Plzia". Also, we see now in the transcriptions that the "aa" in Language 1 represents a single long vowel, which makes the vote 5-to-1 against "axaten" having three syllables.

My current guess is:
Spoiler:
Plzia Uxgi

Axtsa
Lo xampu akelan
Tuge
Pako op plzia le axten

eSOANEM wrote:might possibly throw my conlang family in here once this has been more fully resolved
Yeah, it's tempting, isn't it? I'm reading up on interesting sound changes to see how hard it would be to construct one myself.

User avatar
Lazar
Landed Gentry
Posts: 2151
Joined: Tue Dec 29, 2009 11:49 pm UTC
Location: Massachusetts

Re: Can you try to reconstruct a proto-conlang?

Postby Lazar » Mon Jan 25, 2016 7:51 pm UTC

eSOANEM wrote:This website may help you with that.

Also this site, if you know X-SAMPA.

ThirdParty wrote:But I am still bothered by the "t" in "axaten" not morphing into "d" when the "k" in "akelan" morphed into a "g".

Yeah, I conceded that point in my post.

It seems to me more likely that Language 4 systematically inserts "a" in between "x" and a following consonant to avoid clusters which its speakers find awkward. Note that "a" is the vowel that Language 4 tends to insert places, e.g. into "Plzia". Also, we see now in the transcriptions that the "aa" in Language 1 represents a single long vowel, which makes the vote 5-to-1 against "axaten" having three syllables.

I was tempted to remove a in those words, but I didn't like the idea of reconstructing a form like "uxgi" with a voiceless fricative preceding a voiced stop (because one of those would have to change, at least phonetically). It looks like just the sort of form that would have resulted from the loss of a vowel, not something you'd want to reconstruct in a protolanguage. And "aa" could have collapsed from "axa".
Exit the vampires' castle.

User avatar
Caprice
Posts: 31
Joined: Mon Apr 20, 2015 12:23 am UTC

Re: Can you try to reconstruct a proto-conlang?

Postby Caprice » Mon Jan 25, 2016 8:50 pm UTC

That was fun, but here's the original conlang. I'm not that good at doing sound changes yet, but I hope to improve. Hopefully next time, I'll remember to put surrounding vowels into consideration! :P

Spoiler:
Plysiy Uhri

Ahca
Lo hampu akelan
Cure
Pako-op plysiyle ahten

pləziə uχʀ̥i

axtsa
lo xambu agelan
tsuʀe
pagoːp pləziəle axten

The indirect article particle is "Vp," taking the preceding vowel and lengthening it. The first changes in the daughter languages had to do with /x/, phonemicising voicing for the most part.
Musical Chronicler of the OTT (bestowed by yappobiscuits, NP(?)2297)

User avatar
Caprice
Posts: 31
Joined: Mon Apr 20, 2015 12:23 am UTC

Re: Can you try to reconstruct a proto-conlang?

Postby Caprice » Wed Jan 27, 2016 9:02 pm UTC

The second conlang reconstruction challenge:
L1 and L2 are closely related, and so are L4 and L5.

L1:
"Tera pos uríat?"
Kera ede paga. "PAGAD. URELI." Kera av kablepri kar. "Gaaaaaaa..."

L2:
"Po tera uri a?"
Kera paga ede. "PAGADA. URELI." Kera av ka be pi. "Gaaaaaaa..."

L3:
"Trak pos uhia?"
Kra pag ed. "PAGAR ATA. URELI." Kra kabrible av. "Gaaaaaaa..."

L4:
"Tege uga at pu?"
Kege page edi. "PAGE AT. URELI." Kege kabali kegu. "Gaaaaaaa..."

L5:
"Tha gi a po?"
Kha pha de. "PHA AT. URELI." Kha ghile av khi. "Gaaaaaaa..."

L6:
"Teha uhi a po?"
Keha paha etee. "PAHAA AD. URELI." Kehn av kablebri kerii. "Gaaaaaaa..."
Musical Chronicler of the OTT (bestowed by yappobiscuits, NP(?)2297)

User avatar
ThirdParty
Posts: 347
Joined: Wed Sep 19, 2012 3:53 pm UTC
Location: USA

Re: Can you try to reconstruct a proto-conlang?

Postby ThirdParty » Thu Jan 28, 2016 2:19 am UTC

Yay, a new one! (I tried to build a conlang family myself to post, but realized I'd screwed it up when I noticed "/w/ changes to /ɰʷ/" in one of my evolution rules. :wink: Maybe I'll get it working at some point.)

Here's a first try at reconstructing the second conlang:
Spoiler:
"Tegak ugi-at pos?"
Kega paka eter. "PAKAR-ATA. URELI." Kegan av ka-ble-pri kaxyr. "Gaaaaaaa..."
There's a number of things I'm not happy about with this reconstruction, but it'll at least give us a starting point for discussion.

User avatar
Caprice
Posts: 31
Joined: Mon Apr 20, 2015 12:23 am UTC

Re: Can you try to reconstruct a proto-conlang?

Postby Caprice » Thu Jan 28, 2016 2:26 am UTC

A note: The proto-conlang is the same as the last one, but the daughter languages aren't.
Musical Chronicler of the OTT (bestowed by yappobiscuits, NP(?)2297)

User avatar
eSOANEM
:D
Posts: 3652
Joined: Sun Apr 12, 2009 9:39 pm UTC
Location: Grantabrycge

Re: Can you try to reconstruct a proto-conlang?

Postby eSOANEM » Sat Feb 06, 2016 3:20 am UTC

Ok, this should be fun. Here's a translation of the PIE fable of the horses into my proto-conlang with the phonetic changes (but not grammatical changes which I've yet to flesh out properly) into seven of its many daughters. There's a definite tree structure going on here so expect certain subsets to bear a stronger family resemblance. Points will be awarded for figuring out what the tree structure (and forms at the branch points) is as well as the proto-forms. The order of the samples is random so there's no clues there


Spoiler:
language 1 wrote:Ǣhpūhcā, larjī ŋãə, ūnĩþi ŋǣ̃ãwlǣ̃j ǣhpʊ̄scā, trapo mõjnǣ̃j əpūlcā ʊposa, bərpo mõjnǣ̃j mpə̄ pʊ, rþi bərpo mõjnǣ̃j ŋǣ̃w ŋrĩafþi. Hǣsi ūþi “ǣhcwospo, pə̄r twǣhpə̄ jǣhū̃ab ŋǣ̃ǣ̃j ǣhpʊ̄scā pə̄.” Hǣscā ǣhpʊ̄scā “ūþili asa ūsi, pə̄rcā pʊ̄cā twǣhpə̄ həmbə̄ ŋǣ̃ǣ̃j, walsənõþi, hodʊzǣ cþi ŋãwlǣtʃe ūstʃe hi rūtʃi, rþi pʊ̄ nə ūnĩþi ŋãlǣ nə.” Ūsə ūnĩþi, rþi ləjʃi ūþi nlə̃tūj ūhtʃi.



language 2 wrote:Āgōvā, narbā al, āzij ajllnalēj aʔūsp, ðralv olnalēj eʒūlbā ɔ̄vō, berb olnalēj maʔ vā, riz berb olnalēj majō rilājviz. Gazēj āz “aʔūspā, ha ðɔ̄ʔā ācojā majlēj aʔūsp hō.” Gazējllv aʔūsp “ūzējl ɔ̄lz āzi, herb hōv ðɔ̄ʔā helba majlēj, ɔ̄ldzājðozn, hōzā hiz allnaðā āzið hā roðēj, riz hā ne āzij allnā ne.” Ūzi āzij, riz nelgli āz nendzēj eðēj.



language 3 wrote:Aqquēp, lārbva aj, wussin ānājɲāj aqquosp, trāp ōɲāj ēxop wapos, brjep ōɲāj majq bā, ris brjep ōɲāj mān rīlehafis. Jāɕwe wūs “aqquōspa, cer dōhajquu jaccōnab māɲāj aqquosp que.” Jāɕwep aqquosp “yɕɕi wās wuɕɕi, crjep quep dōhajquu jēb māɲāj, wājttusen , hodvessa cis ājnātʃ wussitʃ wē rūt, ris gā ne wussin ājnā ne.” Yɕɕi wussin, ris leqquihje wūs lenedvō īt.



language 4 wrote:Ācfōf, lārbvæ al, ōwsin ānālnʒǣl ācfosf, trāf ōnʒǣl ēclof wəpos, bref ōnʒǣl maʃ pā, ris bref ōnʒǣl mān rīlahfis. hǣsfo ōws “ācfōspa, ʃer tōɣv jǣcōnav mānʒǣl ācfosf cfo.” Hǣsfof ācfosf “ȳʃīl wōs ōwʃ, cref cfof tōɣv hājv mānʒǣl, wōltuzn, hədwōsa ʃis ālnātʃ ōwsitʃ ʒē rūs, ris cā ne ōwsin ālnā ne.” Ȳʃ ōwsin, ris lēbvī ōws lendāw īs.



language 5 wrote:Oʃīha, lārej ānə, uɣicni acnāɣnacni jehola, drājo ojnoni əguλa ɣāwoh, vāro ojnoni nōgə ʒaɣ, rih vāro ojnoni nōcno renāj. Hogoje uɣih "jecʃōlāw, gər dzojoʃe jogunav nōcnoni jehola ʃe." Hogojeha jehola" ujeli ɣāh uɣiho, gāra ʃeha dzojoʃe hīvə nōcnoni, ɣālgdzocni, hosiɣah gih āɣnagdʒe uɣizdʒe zɣī rugi, rih ɣaɣ nə uɣicni āɣnah nə." Uje uɣicni, rih līsī uɣih ləndzuj jeɣgi.



language 6 wrote:Usjeλha, tλahrrjeλ ahne, uɣeŋe aŋahŋaŋe ʒehela, trahe ojnune cuλa gahwā, vrru ojnune māga zaɣ, rrjē vrru ojnune muŋu rrenār. Gugeλe uɣjē "ʒehsjulahv, crr dzuesje ʒugunjav muŋune ʒehela zje." Gugeλha ʒehela “ujle gār uɣehu, crra zeha dzuesje garva muŋune, gahlgzuŋe, guseɣah cjē ahŋagje uɣezdje zdɣar rruge, rrjē gaɣ ne uɣeŋe ahŋah ne." Uje uɣeŋe, rrjē tλeλsjē uɣjē tλendzuj ʒeɣge.



language 7 wrote:Āvø̄b, lārva al, īzn ājlŋāl āvosp, trāb ēāl ēcleb wɔbe, breb ēāl mag pā, ri breb ēāl mān rīlehfi. Jāswø ī “āvōspa, cel tēv jāgēav mājl āvosp bø.” Jāswøb āvosp “ījil wɔ̄ īj, creb bøb tēv hējb mājl, wɔ̄ltizn, hedwø̄a ci ālnāj īsʃ wø̄ rīd, ri cā ne īzn ālnā ne.” Īj īzn, ri lēwe ī lendēj īd.



The orthography's pretty much phonemic and fairly consistent (and pretty much as per IPA, but with þ for the unvoiced th not theta). λ does vary a bit; in some languages it represents [ɬ], in others [λ]. Tildes show nasal vowels, macrons are for long vowels, whilst geminated consonants are just repeated. Rhotics are also somewhat non-standard, <rr>=[r], <r> can be pretty much any rhotic and, in some languages <ʁ> is [ʁ] and, in one case, [ɹ] (in both cases, it's contrasted with a more central <r> though).
my pronouns are they

Magnanimous wrote:(fuck the macrons)

User avatar
Carlington
Posts: 1588
Joined: Sun Mar 22, 2009 8:46 am UTC
Location: Sydney, Australia.

Re: Can you try to reconstruct a proto-conlang?

Postby Carlington » Sat Feb 06, 2016 8:27 am UTC

I've never done anything like this before, but I'm keen to have a try. For ease of formatting/editability I'm going to do it in a Google Doc, which anyone who's interested in my progress and thinking can view and comment on here if I've done this right.
Kewangji: Posdy zwei tosdy osdy oady. Bork bork bork, hoppity syphilis bork.

Eebster the Great: What specifically is moving faster than light in these examples?
doogly: Hands waving furiously.

Please use he/him/his pronouns when referring to me.

User avatar
eSOANEM
:D
Posts: 3652
Joined: Sun Apr 12, 2009 9:39 pm UTC
Location: Grantabrycge

Re: Can you try to reconstruct a proto-conlang?

Postby eSOANEM » Sat Feb 06, 2016 2:38 pm UTC

Oh I realised I forgot one other bit of info in my section about orthography at the end: <c> can represent [k] or [c] but will usually be [k] unless contrasted with <k> or <q>.

Edit: your groupings aren't looking too bad, 5 and 6 did split off early and form their own branch and 3 & 4 do both belong to the same branch. 1 is also in its correct position. 2 and 7 aren't quite right though.
my pronouns are they

Magnanimous wrote:(fuck the macrons)

User avatar
Caprice
Posts: 31
Joined: Mon Apr 20, 2015 12:23 am UTC

Re: Can you try to reconstruct a proto-conlang?

Postby Caprice » Sun Feb 07, 2016 5:55 am UTC

Here's my reconstruction of L5 and L6's proto-language:

Spoiler:
Usjiλha, tλahrrejλ ahnə, uhɣicni acnahɣnacni zjehɘla, trahjɘ ojnuni gūλa ɣahwɑh, vahrru ojnuni mɑhgə zjaɣ, rrih vahrru ojnuni mʊhcnu rrenahr. Hugɘλe uhɣih "zjehsjʊhlahʋ, gərr dzujɘsje zjuguhnjav mʊhcnuni zjehɘla zje." Hugɘλeha zjehɘla "uhjeli ɣahr uhɣihu, gahrra sjeha dzujɘsje hihrvə mʊhcnuni, ɣahlgdzucni, husiɣah gih ahɣnagdje uhɣizdje zdɣihr rruhgi, rrih ɣaɣ nə uhɣicni ahɣnah nə." Uhje uhɣicni, rrih tλiλsih uhɣih tλəndzuj zjeɣgi.


[Updated 2/7/16, changing /jih/ to /ih/]

λ is /ɬ/, c is /k/, ɘ, ʊ, and ɑ are placeholder vowels.

Also, have something I made to prepare myself for reconstruction:

Spoiler:

Code: Select all

L1 vowels:
i ī ĩ         ū  ū̃
            ʊ ʊ̄
            o    õ
  ǣ ǣ̃  ə ə̄
a ā ã

L1 consonants:
m   n     ŋ
p   t tʃ  c
b   d
f þ s ʃ   h
    z
    l j   w
    r

L2 vowels:
i   ū
e ē ō
    ɔ̄
a ā

L2 consonants:
m   n
p        c ʔ
b   dz   g
v ð z  ʒ   h
    r
    l  j

L3 vowels:
i ī y u ū
e ē   o ō
a ā

L3 consonants:
m n    ɲ
p t tʃ c q
b d      g
f s    ɕ x h
v
  r
  l    j w

L4 vowels:
i ī ȳ   u ū
e ē   ə o ō
æ ǣ
a ā

L4 consonants:
m n
p t tʃ c
b
f s ʃ  h
v z ʒ  ɣ
  l j  w
  r

L5 vowels:
i ī     u
e   ə   o ō
    a ā

L5 consonants:
  n     c
  d     g
  dz dʒ
  s  ʃ  h
v z  ʒ  ɣ
  λ
  l  j  w
  r

L6 vowels:
    u
e ē o
a ā

L6 consonants:
m n    ŋ
  t    c
  dz   g
  s    h
v z  ʒ ɣ
  l  j w
  λ
  r
  rr

L7 vowels:
i ī
e ē ø ø̄   ō
         ɔ ɔ̄
a ā

L7 consonants
  n   ŋ
p t   c
b d   g
f s   h
v z
  l j w
  r
Musical Chronicler of the OTT (bestowed by yappobiscuits, NP(?)2297)


Return to “Language/Linguistics”

Who is online

Users browsing this forum: No registered users and 6 guests