Conlang relay [relocated] (aka "The Cursed Relay")

[Corumayas]

In particular, I don't think we can have them settling down in the hills and THEN being nomadic invaders.

There does seem to be a contradiction there, yes. How likely is a group of horseriding nomadic... miners? (BTW, the Gezoro apparently developed ironworking; so actually we're not quite bronze-age anymore by Tsinakan's time.)

So the western nomads should be different from the Gezoro, or prior to the Gezoro. I thought that the latest group being related to the Gezoro would explain the enmity better than some hatred for a group of mountain guys on the far periphery of the empire.

The thing is, the Gezoro are supposed to arrive in the mountains a thousand years earlier than you had them. Maybe what happens is that a group of them, after they've already been in the mountains a long time, take to herding horses in the upper Aiwa and then go on to invade the lower valley... while still maintaining ties with their mountain brethren? Is that plausible?

I'm fine with Ndak groups being the hill barbarians, I think, and with my guys having always lived in the upper valley.

Cool. That seems like a good scenario.

Another idea: it's said that the Ndak simply absorbed some of their conquerors. So not all the invaders need to have descendants hanging around later. Nor do we need to know what languages they all spoke-- it could be lost in the mists of poorly recorded history.

Maybe, then, the Meshi/Miw once lived in Latsomo? The Ndak could have pushed down the river, sending the Miwans out to the east and the west.

That makes the most sense to me; or maybe the Miwans occupied the whole Aiwa valley; the ones that became civilized early on became the Ngauro, and the others (Meshi, Miu, et al.) were split apart by the arrival of the Ndak.

The Meshi could have held Buruja for a long time while losing their language in favour of Naidda (perhaps with loans, or with Meshi as a court language or something). Or, maybe they were never there. I just thought that was one way to deal with the "and then everyone sat still for two thousand years waiting for something to happen" appearance.

There is a gap between Buruja and Lasomo, where there seems to be nothing of interest; maybe the descendants of the Meshi could go there.

I said bay - there's nothing there but water. And Proto-Isles people, yes, but meh.

My understanding-- which may be wrong, we don't have any info on Proto-Isles online-- is that the Proto-Isles-speaking people came there from further east, and probably arrived after the Lotokan-speakers were already there. Legion can clarify whether that's what he had in mind or not.

Now, if we ignore Doroh for the time being...

Fair enough, it too is a later language... and Radius' description made me think it was spoken slightly west of that area, anyway.

While Lotoka has vocabulary that can't be reconciled with a northerly location, we can handwave it away by saying that vocabulary was inherited from proto-Peninsular. (Then again, I don't think proto-Peninsular is that much more southerly...)

It's a big enough difference, climate-wise: at 45 N winter is very cold and very snowy; at 30 N it's warm, not really winter at all. Also, the peninsula is close to the large Xshali area which seems like a fair analogue to South Asia.

I think the most sensible explanation is to put Lotoka in the south part of the peninsula originally, and have its speakers spread to the islands around the bay shortly before Legion's people arrive; then there could also be Lotokan speakers in the north as recent arrivals, of course.

[Radius Solis]

Installment 1 of Delta Naidda is now up.

This consists of two parts:

- phonology

and

- a partial lexicon.

At the moment the lexicon comprises only the core content words inherited from Ndak Ta. Grammatical words, loanwords, and the extended NT vocabulary will come in a future installment. Sorry it's not the whole lexicon yet, but I wanted to post as soon as I had finished a substantial chunk of the revision.

[Dewrad]

In particular, I don't think we can have them settling down in the hills and THEN being nomadic invaders.

I dunno, it's not impossibly. While Rathedan is more fertile than the quasi-desert upper Aiwa valley, it's not incredibly fertile, and by the classical period a lot of that fertility was a result of human intervention- irrigation works, terracing and so on. I don't think it's out of the question that they settled in the hills, over-populated and then did a viking-style marauding thing.

However, I do prefer the hill barbarians being Ndak. Perhaps the "land-based Gezoro vikings" were mainly occupied in raiding, which is why the settled Ndak didn't like them.

[Corumayas]

Installment 1 of Delta Naidda is now up.

Excellent. Does this mean that the old version is still valid for the southern dialect area, then?

Also, looking at the Proto-Isthmus to Faraghin diachronics again, I have a question-- what happens to dZ?

[Radius Solis]

Installment 1 of Delta Naidda is now up.

Excellent. Does this mean that the old version is still valid for the southern dialect area, then?

No, it's not valid for anything. I have always been unhappy with the version of Naidda on that page and even a month after I'd done it, I was already trying to revise it... not that that got anywhere at the time. I'll do a proper southern dialect after I finish with the delta dialect.

Also, looking at the Proto-Isthmus to Faraghin diachronics again, I have a question-- what happens to dZ?

Who knows? Maybe it merged with something. Or maybe it was lenited and then lost. Or maybe there just aren't any cognates left in Faraghin that would have a reflex. Or you can decide on it another way if you like - like I said, that wiki page is open for editing by anyone wanting to work more on Faraghin diachronics.

[zompist]

Installment 1 of Delta Naidda is now up.

Excellent. Does this mean that the old version is still valid for the southern dialect area, then?

No, it's not valid for anything. I have always been unhappy with the version of Naidda on that page and even a month after I'd done it, I was already trying to revise it... not that that got anywhere at the time. I'll do a proper southern dialect after I finish with the delta dialect.

I hope you don't change it too much, since I borrowed a bunch of Naidda words for Faralo...

[Radius Solis]

I hope you don't change it too much, since I borrowed a bunch of Naidda words for Faralo...

Nah. Around half of Delta Naidda's words are phonetically identical to the originals, and at least half of the rest differ by no more than a couple of phones. Besides which, Faralo probably did most of its borrowing from the North Coast dialect region of Naidda where the trading towns that belong to Huyfarah are. Since that dialect hasn't been described and probably never will be... we can just assume some of the forms Faralo borrowed can be found there.

[Salmoneus]

Cool... let me know if and when you want the Miwan wordlists that have been used thus far.

Yeah, I'd definitely be interested in looking at them.

I would also like the wordlists.

[Cedh]

As indicated earlier, I have vectorized the continent map and started to create a historical atlas. At this stage, many things are still sketchy, and the distribution of languages and peoples should not be taken as definite. If we decide to change the history, I'll adapt the maps accordingly. Here are the first five images I've made:



An overview map, with some regional names for reference.



A linguistic map showing the approximate locations of the most important language families around -4000 (when the Ngauro start to build their first cities).



A linguistic map from the early years of Tsinakan's reign, c. -1900.



A few years later: the Ndak Empire at its greatest extension (adapted from Radius' Ndak Empire map on the Almeopedia)



A map showing the Athalē Empire and Huyfárah, c. 500 YP.

[Zhen Lin]

Very beautifully done.

Tangentially on-topic: that font looks like Gentium. They've just released a new variant of Gentium - Gentium Basic - which comes in various weights including, at long last, bold!

[Radius Solis]

Wow! Looks very good. I like the bulk of that.
A number of comments/suggestions...

On the second map:

* The Eige Valley language group should extend southwest into the blank spot where Atax Petan is.
* The Western languages have not spread into the visible part of the map yet. Their expansion eastward was much more sudden, at around -3000. At -4000, Proto-Western was still largely confined to the western coastal region.

On the third map:

* The Gezoro may have spread further to the northwest than shown. The Tjakori probably are influential to a greater extent down their river valley than shown.
* Again, the Miw should be filling up the whole Atax Petan blank spot that's to the SW of where you currently have them. Also they should extend along the lower border of the Ndak languages all the way up to the Meshi; the Gezoro and Hitatc language areas shouldn't touch the Ndak area. Even by the year zero, it's still a big dialect continuum.

On the last map:

* The Athale empire probably extends into at least some of the Tjakori valley; in the western half around Xoron Eiel, it also probably extends further to the north. Possibly all the way to the mountains. Also, by +500, it has already lost control of Atax Petan.
* Huyfarah, too, should be bigger by this time. It probably should: 1. extend further east at least to the first peninsula there and maybe the second; 2. control some or all of the big island southeast of Agumosou; and 3. have a number of colonies along the coast to their south, all the way into peninsular territory (these colonies need not have any long-term linguistic impact, if such is not desired). It might also have expanded across the northern mountains to some of the north coastal region.

[Radius Solis]

Cool... let me know if and when you want the Miwan wordlists that have been used thus far.

Yeah, I'd definitely be interested in looking at them.

I would also like the wordlists.

Bah! I can't find the Miw wordlist I used for Pencek, now. But, I used almost the whole list in the Pencek lexicon, so most words are available there already if you want to take the time to pick them out of the rest. If you don't, I'l do it later. (The lists for the dialects Puoni and Naidda borrowed from are in exile on the dead computer, alas.)

The phoneme set is very vanilla: /p t k b d g f v s z m n l r w j a i u a: i: u:/. The syllable structure is (C)(1)V(2), where 1 = /l r w j/ and 2 = /f v s z m n l r w j/. Final syllables can end in any consonant. The Miwan dialects are all also supposed to be tonal, but I never described the tonal system because it was irrelevent for the loanwords. Feel free to make up whatever you want for that, or just ignore it, whichever. There are a handful of words that are exceptions to this phonology; they were intended to have come from another dialect, but I didn't actually describe it.

[Cedh]

Very beautifully done.

Wow! Looks very good. I like the bulk of that.

Thank you!

A number of comments/suggestions...

These are good ones. I'll add them in the next version.

Tangentially on-topic: that font looks like Gentium. They've just released a new variant of Gentium - Gentium Basic - which comes in various weights including, at long last, bold!

Nice to know this. A pity it doesn't have IPA yet, but bold and bold italic will be useful for these maps...

[Salmoneus]

Cool... let me know if and when you want the Miwan wordlists that have been used thus far.

Yeah, I'd definitely be interested in looking at them.

I would also like the wordlists.

Bah! I can't find the Miw wordlist I used for Pencek, now. But, I used almost the whole list in the Pencek lexicon, so most words are available there already if you want to take the time to pick them out of the rest. If you don't, I'l do it later. (The lists for the dialects Puoni and Naidda borrowed from are in exile on the dead computer, alas.)

Don't worry, I've done it, thanks. [And realised that the lexicon of Miw loanwords alone is longer than the lexicon of any conlang I've ever made. Possibly of all of them put together...]

[zompist]

The maps are very nice looking.

I'm confused by the later history of Huyfárah. It looks for 500 you used the same borders as my Almeopedia map, but that's for the year 130.

From Radius's chronology, it seems Huyfárah had a second golden age in the late 300s, and expanded into Buruja and further into Kasca. And the high points of Huyfárah and Athale shouldn't coincide.

(With this sort of thing, it's all too easy to let borders stay the same for centuries or even millenia. But if you look at, say, McEvedy's atlases, almost everything is in flux. Some areas are particularly volatile; Kasca, for instance, would be an area of near-constant change in this period, as the influence of Huyfárah and the local city-states ebbs and flows.)

[Corumayas]

I've got that Miw wordlist now too; thanks, Radius. How different were the Miwan lists you had from each other? And do you have a sense of how distant Miwan and Ngauro should be?


Regarding cedh's maps, yeah, they look very very nice. I have just a few comments to add to Radius' and zompist's.

-I'm not sure Isthmus should appear on the second map; proto-Isthmus is dated -3000. I'd suggest that at -4000 we have a single "Eige-Isthmus" continuum (and maybe it doesn't reach quite that far east yet).

-Along the same lines, I think the Isthmus language names on the third map are anachronistic; as I understand it, at that date proto-Ferogh and proto-Boesin are diverging, but the Faraghin and Ferajin still speak a single language.

-Possibly relevant to the last map: during the third century, Huyfárah expands further west too: to the eastern edge of the Ziphē valley, and maybe as far as the Eige valley below Akelodo. Later, the frontiers of both empires in this region probably fluctuate frequently, but I think they should probably actually meet somewhere just east of Lasomo (unless there's some fairly strong buffer state filling the gap).

[Radius Solis]

I've got that Miw wordlist now too; thanks, Radius. How different were the Miwan lists you had from each other? And do you have a sense of how distant Miwan and Ngauro should be?

The wordlists I used different mostly in their phoneme sets. The one I used for Naidda ages ago had the additional phonemes of /N q X ? e e:/, at least one diphthong I can't remember anymore, and iirc, a syllable pattern of (C)(1)V(C). The one I used for Puoni was a lot like the one for Pencek, but with (C)(1)V(1) and a uvular approximant. In all cases the 1 represents "any approximant or liquid", just as with the list for Pencek.

Ngauro: I dunno. Wing it! But the name may not be representative - it might just be a Ndak exonym for them. Good luck.

[Corumayas]

The wordlists I used different mostly in their phoneme sets. The one I used for Naidda ages ago had the additional phonemes of /N q X ? e e:/, at least one diphthong I can't remember anymore, and iirc, a syllable pattern of (C)(1)V(C). The one I used for Puoni was a lot like the one for Pencek, but with (C)(1)V(1) and a uvular approximant. In all cases the 1 represents "any approximant or liquid", just as with the list for Pencek.

Ngauro: I dunno. Wing it! But the name may not be representative - it might just be a Ndak exonym for them. Good luck.

Thanks for the info. I guess what I'd really need to do a good reconstruction is the correspondences among the three lists... but I'll see if I can "find" some cognates with the Faraghin wordlist, and maybe some Ngauro ones in the Ndak Ta lexicon too...

If anyone else is interested in this slightly dubious enterprise, feel free to join in! We can use the Frathwiki page Radius started for the Isthmus side of things; I'll probably start one for the Eige Valley languages once I have something to put there.

[Radius Solis]

Thanks for the info. I guess what I'd really need to do a good reconstruction is the correspondences among the three lists... but I'll see if I can "find" some cognates with the Faraghin wordlist, and maybe some Ngauro ones in the Ndak Ta lexicon too...

I can't give any advice for identifying Ngauro loans in NT, save that you should probably look at the cultural sphere of things more than the core vocab. At least for unidentified Gezoro loans we can look for CVCVC structure (nalaror, isesigen, ginig) while native NT words tend to have clusters, but for Ngauro we don't have any such easy clues.

The established correspondences between the three Miwan lists are pretty minimal, because I deliberately didn't borrow many of the same words for different languages. Only a few of the flora and fauna terms are shared by Puoni and Pencek's lists (e.g. elm, oak, acorn, bird nest), and they didn't vary by much if at all.

[Corumayas]

Paging through this thread just now, I found a live link to the original Puoni lexicon! It shows etymology too, so now we basically have two of the three Miwan wordlists.

Edit: it includes a bunch of compounds, yay!

[Salmoneus]

One thing we may have to address is the origin of Faraghin /r/. It seems to occur in clusters, but I see no feasible way to produce this. I suggest we need another source of them.

[Radius Solis]

One thing we may have to address is the origin of Faraghin /r/. It seems to occur in clusters, but I see no feasible way to produce this. I suggest we need another source of them.

There's plenty of possible sources of clustered r if you look for them:

-There's always loanwords... having r in clusters in the modern langauge doesn't necessitate a diachronic source.

-Compounds where the first element has final r or the second has initial r (there are some of these already clearly shown in the wordlist).

-Losing certain vowels (but not others) between r and other consonants, when unstressed; this could explain some of the other clusters too.

-The dz and zd clusters that originally created r may have themselves been clustered with other things; we don't know the syllable structure of Proto-Isthmus. Maybe it was really clusterry already and instead of gaining clusters, a lot of them have actually dropped out over the intervening millennia.

-One or more other phonemes may have become r in clusters, such as /t`/ or /d`/ or both.

-Or you could just come up with another source for clustered r.

[Salmoneus]

One thing we may have to address is the origin of Faraghin /r/. It seems to occur in clusters, but I see no feasible way to produce this. I suggest we need another source of them.

There's plenty of possible sources of clustered r if you look for them:

-There's always loanwords... having r in clusters in the modern langauge doesn't necessitate a diachronic source.

-Compounds where the first element has final r or the second has initial r (there are some of these already clearly shown in the wordlist).

-Losing certain vowels (but not others) between r and other consonants, when unstressed; this could explain some of the other clusters too.

-The dz and zd clusters that originally created r may have themselves been clustered with other things; we don't know the syllable structure of Proto-Isthmus. Maybe it was really clusterry already and instead of gaining clusters, a lot of them have actually dropped out over the intervening millennia.

-One or more other phonemes may have become r in clusters, such as /t`/ or /d`/ or both.

-Or you could just come up with another source for clustered r.

My second thought was loans, but there seemed to be loads of them, and it seemed a bit convenient writing them all off to loans (also, more work).
Compounds are a good point that I hadn't considered.
Vowel loss and other sources of r were things I was thinking of when I said 'a new source'. I thought it might be better raising the point here, though, as otherwise several of us might be doing different things.

My first thought was big clusters, but that would mean things like "tdz", and I'm queesy about mixed-voicing clusters. Possible though, I guess.

[Corumayas]

I was going to suggest the following outcomes for Proto-Isthmus /d`/:

> dZ > j > 0 / initially
> dZ > z > r / elsewhere

I don't think that helps much, though. The real dilemma seems to be a number of words that start with pr-, br-, tr-, dr-, kr-, and a couple that end in -rt, -rk.

Deletion of certain unstressed vowels between a plosive and /r/ might work. It might help if we determined what the syllable constraints and stress patterns were like at earlier stages.


In Miwan news, I've found seven correspondences between the two word lists:

Code: Select all

Western (Pencek)    Eastern (Puoni)    Gloss
furzi:n             furzin             elm tree (tree-life)
rakuri              rhamkuri:          oak tree
ti:vur              ti:vur             hickory tree (funny-tree)
tur                 tu:l               acorn
varja:              varja              to dance
vi:zwar larch tree  vi:zwal pine tree  [coniferous tree]
zafwi:ta            zafwita:           bird's nest

Observations:
-Sometimes final /r/ in Western corresponds to /l/ in Eastern, sometimes both have /r/. (Western does have final /l/ in other words, though.)
-Eastern has /r R/, both corresponding to /r/ in Western.
-Medial /mk/ in Eastern becomes /k/ in Western without (apparently) affecting the preceding vowel.
-Intervocallic /f/ is voiced to /v/ in both. (A glance over the rest of the wordlists shows that this rule does not apply to /s/, at least not in Western).
-Vowel length shows no discernable pattern, except that there seem to be more long vowels in Eastern.

Presumably Pencek and Puoni each borrowed from multiple Miwan dialects, and that's why the correspondences are not consistent.

[Cedh]

I was going to suggest the following outcomes for Proto-Isthmus /d`/:

> dZ > j > 0 / initially
> dZ > z > r / elsewhere

I don't think that helps much, though. The real dilemma seems to be a number of words that start with pr-, br-, tr-, dr-, kr-, and a couple that end in -rt, -rk.

Deletion of certain unstressed vowels between a plosive and /r/ might work. It might help if we determined what the syllable constraints and stress patterns were like at earlier stages.

I'd suggest regular stress on the second syllable in Proto-Isthmus, with (C)(j)V(C) and (C)V(j)(C) allowed, medial plosive+plosive clusters agreeing in voicing, voiced coda stops becoming voiceless before an unvoiced consonant, and all coronal clusters agreeing in retroflexion (which means that /sd`/ would be realised as [s`d`]).

We could also develop some /r/ from *ld *dl *ld` *d`l clusters.