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

Cedh · Post by **Cedh** » Fri Sep 19, 2008 4:47 am

Corumayas wrote:I see you've got teak 'chief', which looks cognate to NT diàka. I've been working with the assumption that diàka is a loan from Ngauro, cognate with Faraghin čark; but I think the same root shows up in Proto-Xoronic too. Should I drop that correspondence, or is it still possible that this word was borrowed (either into Proto-Macro-Edastean or separately into all three langs) during the Ngauro's heyday?

Tlaliolz is located close enough to both Ngauro and Faraghin territory that it could easily have borrowed some words from there, independent of NT. Actually, I'd expect to see a rather significant portion of Faraghin loans in Tlaliolz as these people seem to have shared a fairly similar (and adjacent) habitat.

The real problem is Proto-Xoronic, which has *əhtanqa "chief, headman". This looks like a clear cognate to diàka, which in turn, if it derived from Proto-Talo-Edastean, would have to come from something like **(V)diaɴqa. I would also like to keep this word as a term from Ngauro originally, so think our best option is to ascribe at least the PX version to an early borrowing from those Miwan/Meshi speakers who migrated up the Aiwa around -3000. Given the currently reconstructed PEI root *trelk(a), which should be something like **trilk(a) in Proto-Miwan, this is problematic though. Maybe the Meshi aspirated initial *tr, backed *k to [q] before *a, which in turn could lead to lowering of the vowel in the main root syllable, and nasalized coda *l (a change that is present in some Habeo varieties a few millennia later), so we could get Old Meshi **thaɴqa. But even that is still a bit of a stretch if we keep assuming that the Meshi dialect of this time should have been very close to Proto-Miwan. Well... any ideas?

EDIT:
The reflex of PX *əhtanqa in Proto-Habeo (c. -1000 YP) is *a:tenqe, which might be easier to get from Miwan/Meshi. By that time though, the peoples of the Xōron have already had their first attempts at empire-building, which should postdate the loan.

dunomapuka · Post by **dunomapuka** » Fri Sep 19, 2008 8:44 am

I say don't worry about Proto-Xoronic etymologies for now. If you're willing to convolutedly derive that one from *trelka, you ought to also be willing to say it's a mere coincidental resemblance

Tlaliolz "teak" being a loan from NT "diàka" seems perfectly likely to me.

Do you have a pTE wordlist, Radius?

Cedh · Post by **Cedh** » Fri Sep 19, 2008 10:38 am

boy #12 wrote:I say don't worry about Proto-Xoronic etymologies for now. If you're willing to convolutedly derive that one from *trelka, you ought to also be willing to say it's a mere coincidental resemblance

Right...

Let's say it's pure chance then.

Radius Solis · Post by **Radius Solis** » Fri Sep 19, 2008 11:31 am

I do have a partial pTE wordlist from which I derived around half the Tl. words thus far. Sad to say, I completely forgot to avoid back-deriving the NT words that were supposed to be loans, so I did indeed have teak coming from pTE. If I borrowed it into Tl. from NT instead, the outcome should change to tiak which honestly I don't really want because it's too close to the NT (too many words like this have been dead obvious in their cognacy).

So I'll still do that, but call teak an irregular outcome. I need a few more of those anyway, because the s/c list is supposed to be presented as imperfectly reconstructed by Akana-internal linguists.

Corumayas · Post by **Corumayas** » Fri Sep 19, 2008 3:07 pm

Hmm. Since the estimated date of Proto-Talo-Edastean is -2500 YP, I think it could easily contain loans from Ngauro that would then be inherited in both Ndak Ta and Tlaliolz.

Where does Tlaliolz isaqʷi 'big man' come from, btw? Could that be the cognate of Proto-Xoronic *əhtanqa somehow?

Another point is that, given their fairly-nearbyness, Proto-Eigə-Isthmus and Proto-Macro-Edastean may ultimately be related. (Too remotely to reconstruct an ancestor, of course... they'd have to have diverged well before -6000 I think. But I can picture Akanaran linguists speculating about it, Ruhlen-style. They could be part of a macrofamily that spread with the first appearance of pastoralism and early agriculture, maybe ca -10000-8000....) So a couple of tantalizingly similar-looking roots between the two families might not be entirely inappropriate.

Edit: While I'm posting, I want to also mention that I've put some preliminary thoughts about PEI morphosyntax up.

Radius Solis · Post by **Radius Solis** » Fri Sep 19, 2008 3:29 pm

Corumayas wrote:
Where does Tlaliolz isaqʷi 'big man' come from, btw? Could that be the cognate of Proto-Xoronic *əhtanqa somehow?

I don't know the source of the Xoronic word. Isaqʷi is a modern compound in Tlaliolz: isaf "man" (NT aisab, husband) with qʷi "big" (NT pai "big").

Zhen Lin · Post by **Zhen Lin** » Sat Sep 27, 2008 9:27 am

I am pleased to say that there are now almost 200 (just one short!) roots and over 600 entries in the Proto-Peninsular lexicon. Unfortunately my internet connection is being annoying again so I can't upload it.

Cedh · Post by **Cedh** » Sun Sep 28, 2008 2:17 pm

I've just put together some basic information on a northern dialect of Aθáta, which is spoken in the Sawîya valley in the southwestern Xōron, just north of Rathedān. Phonologically it moves a bit towards the Habeo languages, and it contains a few morphological and syntactic features shared with or inspired by Kozado and Mavakhalan.

Etymological note:
Sawîya is this dialect's name for Aθ. Θafîa < Ad. Thabīa < Plains Habeo Tawɨʔɨya "white-stone (river)" < Proto-Xoronic *Ǝtoha:quhi (ahəttih). This meaning is the same as Adāta Tikhōdōzē, a town at the confluence of this river into the Eigə.

Cedh · Post by **Cedh** » Wed Oct 01, 2008 8:08 am

On noticing that Pencek doesn't have the Tsinakan text sample yet, I've tried writing the first two lines (i.e. those that are featured in the comparative translation). Here's my try - Radius, could you please check whether that is correct Pencek?

Si Sinkan, si tirk hazar, si tirk Kazcak, si nekat zanak o tharak, a pin sik raz: Hert aphenc asa phakacak nazi, aza tirkars runs, ak par henk then.

[ si ˈsiːn.kan | si ˈtiːɾk ˈhaː.ʂaɾ | si ˈtiːɾk ˈkaːʂ.ʈʂak | si ˈneː.ɡat ˈʂaː.nak ə ˈtʰaː.ɾak | a ˈpiːn sik ˈɾaːʂ ‖ ˈçeːɾt aˈˀpʰeːnʈʂ aˌsaː ˈpʰaː.ɡaːˌʈʂak ˈnaː.ʂi | aˌʂaː ˈtiːɾ.kaɾs ˈruːns | ak ˌpaɾ ˈçeːnk ˈtʰe:n ]

this Sinkan, this king great, this king Kazc-POSS.3P, this brother sun-POSS.3 and moon-POSS.3P, 3SG say-PRF this-PL word: up.to-1SG sit-SUP on-3P throne-POSS.3P father-POSS.1SG, all country foreign be-PRF hostile towards-1SG

(corrections should go here)

Zhen Lin · Post by **Zhen Lin** » Wed Oct 01, 2008 9:29 am

I've uploaded the expanded lexicon. As usual, it is a UTF-8 encoded CSV file which can be imported into spreadsheet applications.

I've also been trying to work out how to evolve the syllabic consonants. This pattern seems to give the nicest results:

(The period indicates a syllable boundary.)

V.N= > V.NN= (This only occurs if an onset consonant has dropped out.)
N=. > @N.
N=C. > @NC., if C is a stop
N=C. > N@C., if C is not a stop and no illegal onset cluster is formed
N= > @, otherwise

R=. > @R
R=C. > @RC.
R=CC. > R@CC.

.H= > .HH=
H= > @

e.g.

/kaktm/ > /kak.dm=/ > /kak.d@m/
/ulmhs/ > /u.lm=h\s/ > /ul.m@h\s/
/hlnr/ > /hl=.nr=/ > /h@l.n@r/
/hNitu/ > /h\.Ni.tu/ > /h\@.Ni.tu/

cedh audmanh wrote:I've just put together some basic information on a northern dialect of Aθáta, which is spoken in the Sawîya valley in the southwestern Xōron, just north of Rathedān.

It sounds interesting. What are the conditions for [@] vs [1]?

Also, I'm off to university tomorrow! So I'll probably be visiting this forum less, but feel free to contact me by email.

Cedh · Post by **Cedh** » Wed Oct 01, 2008 10:07 am

Zhen Lin wrote:What are the conditions for [@] vs [1]?

[1] occurs before all nasals.

[@] occurs
- adjacent to [j w]
- in syllables immediately before the main stress
- or when preceded by a stressed syllable whose nucleus is one of /i i: A A:/.

[1] occurs elsewhere.

Radius Solis · Post by **Radius Solis** » Wed Oct 01, 2008 1:06 pm

Ah, good old Pencek. I'm always slow with Tsinakan texts these days (I need to get one done for Naidda too) because after what, four or five translations now, I've gotten sick and tired of it. The old hark-and-behold stuff wears on me after a while.

Fortunately, looking through old files it seems I have a mostly-completed translation for Pencek. I think the reason I never finished it when I was still working on the language is that its grammar wasn't robust enough to handle all the necessary mechanics and I moved onto other projects before deciding how I wanted to handle those things. I am not certain the following is strictly correct, and it differs from what you came up with, but here is the translation of the first two lines, from back when I still remembered how the grammar worked.

Sinkan, si tirk hazar, si tirk ha si ras ha Kazc, si nekat ha si zana o si thar, a eapi sipen: Hert i aphan aza phakaca nazi, aza tirkars run, ak parn henk then.

So we can use yours for now, and later when I have more free time (things have been busy here lately) I'll compare our two translations to see which is preferable (probably yours), and correct any infelicities it might have.

dunomapuka · Post by **dunomapuka** » Thu Oct 02, 2008 3:27 pm

1. Nobody has commented on cedh's cool Sawîyaran thing yet. Looks good - but I'm having trouble figuring out where it fits in on the map; I cannot reconcile your description with the various Rathedān maps. Are you working off some new, better map of your own?

2. I fear this project is getting a bit... underpopulated. We should try to reel inactive members back in.

Cedh · Post by **Cedh** » Thu Oct 02, 2008 5:30 pm

boy #12 wrote:1. Nobody has commented on cedh's cool Sawîyaran thing yet. Looks good - but I'm having trouble figuring out where it fits in on the map; I cannot reconcile your description with the various Rathedān maps. Are you working off some new, better map of your own?

I haven't drawn such map yet, but: yes. Most of the maps we have so far are a bit outdated as we've changed some details of the history. For instance, the ethnolinguistic map of the Xoron Eiel lists Komeyech, which has since been moved to Dagaem, and "High Adatan", which was never seriously planned AFAIK. I'm working off a slight modification of that map, based on my upcoming House of Rikhus article. Basically, Athale retreats from most of the Xoron in the mid-5th century, keeping only (1) the Thabia valley, which is adjacent to the Rathedan and would offer a competitive trade route to Tjakori if not controlled by the empire, and (2) the Eige valley itself up to Meximo (after c. 600 the part above Todotin is lost too). All in all, the range of Mavakhalan is a bit smaller than shown; that of Athata (as the language of central Rathedan) is a bit larger. The "High Adatan" area speaks Habeo languages, and the Komeyech area is split between Mavakhalan (in the west) and Late Ndok Aisô (in the northeast).

There will be a "High Adatan" language, actually: I'm working on a new descendant of Adata that is highly influenced by the Habeo languages, and it will be spoken by the "urban elite" in the diaspora towns along the Eige, mostly Tikhodoze, Meximo, and Eieliatus.

boy #12 wrote:2. I fear this project is getting a bit... underpopulated. We should try to reel inactive members back in.

Absolutely!

Radius Solis · Post by **Radius Solis** » Thu Oct 02, 2008 10:15 pm

boy #12 wrote:2. I fear this project is getting a bit... underpopulated. We should try to reel inactive members back in.

No objections here, I'd be happy to see more people in Akana. New people too, perhaps. Some kind of bait to draw activity would likely help; I have two thoughts on that. First, some of us - like Sal and Zomp - seem to be drawn to post when the subject matter turns to more to history, politics, and technology instead of language. These areas need much more attention than they've been getting anyway, so perhaps we should set out a Project to hammer out a lot more of that in one or more specific spheres. Secondly, we could potentially attract a great deal more activity on the language front if we were to announce (in a new thread) a Cursed Relay Number Two, structured similarly to the original but working with a different language family and getting some good chains of successive linguistic generations built. It generated a great deal of participation and fun the first time, why not do it again?

Bonus thought #3: these two things could be readily combined, if care were taken for those with varying interests (history vs. language, e.g.) to accomodate each other. (And of course nobody should be required to pick a side; everything anyone wants to contribute has always been welcome before and there's no good reason to change this.) Two good possibilities: developing the Isles/Siixtaguna sphere, and developing the Western sphere.

dunomapuka · Post by **dunomapuka** » Thu Oct 02, 2008 11:23 pm

I think another relay thread is a smashing idea. Doing the Takuña/Isles sphere would be my preference, and specifically the Takuña family if we're doing the languages, but perhaps both in conjunction would make sense. This would require a couple things:
1. Some expansion of the Proto-Nú.-Tak. vocabulary - we need about 2 or 3 times what we've got.
2. Some further background understanding of these people and their proto-history - local biology, technological level, when writing is invented, etc.
3. We should impose a bit more structure on this Relay than the last, since there are already some constraints on the world as we have constructed it. Obviously. We'll figure it out.

I wouldn't be opposed to developing Western either. My sense is that Western should be a BIG family. We would have a greater degree of <i>carte blanche</i> for it, also.

Radius Solis · Post by **Radius Solis** » Thu Oct 02, 2008 11:54 pm

Hmm. I was thinking it might be better to work with the Isles family itself, as that would have more potential to draw the other original Team 1 folks back in (Rory, Avaja, Brandrinn, Kolya, Ran - and Legion of course has already been involved). But maybe none of them are around much these days anyway, I don't know. But then the Nualis-Takuna family would still be important as substrate languages, filling a role not unlike the Eige-Isthmus family does for the Edastean sphere.

Cedh · Post by **Cedh** » Fri Oct 03, 2008 6:57 am

We should definitely combine a new relay with developing further historical detail. It would be much better to focus on broadening the information available on the two millennia between -500 and +1500 before moving too far beyond that with specific languages. And we will have to describe the scenario to start from in more detail before we can really launch this project. I think it would be good to do some systematisation of what we have so far, and to set up an easy-to-follow starting point for newcomers. For example:

- create a true Main Page for the wiki and relegate all meta information to AkanaWiki:XYZ project pages.
- add a link to the #Akana IRC channel (which we should definitely use more!)
- rewrite some of the more general articles a bit (Akana, Edastean Languages, ...)
- update the most relevant maps (note to self: high priority!)
- develop and port some information that currently exists only somewhere in this thread, e.g. Corumayas' post about agriculture.
- ...

--------------

As for the language families open for a relay:

- Western would be great, as there would indeed be quite a bit of carte blanche, but we don't have a halfway detailed geography for their region yet. We've decided to add more land over there, but I'm still somewhat unsure how to distribute it in order to (a) make a maritime north-west passage possible (needed for Sal's Cunning Plan, if he's still at it), and (b) allow for plausible tectonics (the latest layout requires the western mountains stretching northwest). Maybe some fjordlands and islands would be good...

- We already have some Isles languages for the first part of the period we should be aiming at, and Legion is still working on Thokyunehota. There would be some space for further languages of that generation on the islands in the middle of the chain from Thumapahithi to Ttiruku, but I don't think there should be any new Isles languages on the mainland in classical times. Later, maybe. But:

- Nualis-Takuña languages should have occupied most of the northeastern peninsula by -500, with the only exceptions of Lotoka (Peninsular) and a few Eastern Isthmus. I agree that NuTak should be a major substrate for any Isles languages on the mainland, but I think Takuña itself would be even more likely to spawn some daughter languages. Accordingly, starting from PNuTak would be best IMO.

- Peninsular languages do not currently appear to be of much continental importance, but we could change that. PPI is one of the most detailed protolanguages that we have, and if we're focusing on the east someone really needs to elaborate Lotoka, at least to the point where Ndok Aisô is currently. (Zhen Lin, I remember you did some sound changes for Lotoka a while back...)

- If we're dealing with the east next, Doroh should be developed too. Some grammar and vocab for PEI, and some more ideas for Proto-Isthmus, would be good for this.

- When this information is available, one or two northern/northeastern descendants of Fáralo would also be cool to have. Or a colonial Fáralo-Isles creole or something...

krinnen · Post by **krinnen** » Fri Oct 03, 2008 10:40 am

Hm. What a coincidence. Earlier today I was thinking what a pity it was I could never get involved in the conlang relay (well, after the curse fell on me and my beloved computer...) and why was that.

I think a major obstacle is the lack of organization of the information available. All the active contributors have many details in mind when working on akana, which facilitates the task of mantaining coherence, but I'm afraid a newcomer would find it really hard to produce work without contradicting this or that text, whether canon or not.

Tell me, please, what do you think would be the best way for a newcomer to approach Akana?

Radius Solis · Post by **Radius Solis** » Fri Oct 03, 2008 12:34 pm

krinnen wrote:Hm. What a coincidence. Earlier today I was thinking what a pity it was I could never get involved in the conlang relay (well, after the curse fell on me and my beloved computer...) and why was that.

I think a major obstacle is the lack of organization of the information available. All the active contributors have many details in mind when working on akana, which facilitates the task of mantaining coherence, but I'm afraid a newcomer would find it really hard to produce work without contradicting this or that text, whether canon or not.

This is one reason I suggested starting the new relay in a different cultural sphere, so that there would be a lot less existing material to maintain consistency with.

Lack of organization of the material is a perennial pitfall, I know. But at this point I think most of the important stuff is on the wiki - have you browsed the modern AkanaWiki? There's tons of stuff there now, and though wikis still leave something to be desired for organization it's nevertheless a great deal better than what was available a year ago, let alone two.

Tell me, please, what do you think would be the best way for a newcomer to approach Akana?

The beginning of this thread - my Linkes and Infoe post, in particular - may be a good starting point. From there, the rest of the answer is: read the wiki!

And of course whenever you have specific questions you're completely welcome to ask, even if you've got dozens of them. Questions are good, and they spur us to address things that need work.

Cedh · Post by **Cedh** » Fri Oct 03, 2008 12:54 pm

A note on our official timeline: Everything before the "Year Zero" is canonical, but sometimes only approximately dated. Everything between 0 and 500 is canonical. Most of the dates after that, especially those regarding the downfall of Huyfárah, are currently being revised and should not be taken as definite.

I've also just started reorganizing the Wiki front page, and I've added comments on the map repository saying which maps are valid and which ones are out of date in what way.

And as Radius said, if you're unsure about something, please ask, either here or on [url=irc://irc.sorcery.net/akana]#akana (IRC)[/url].

Corumayas · Post by **Corumayas** » Sat Oct 04, 2008 1:08 am

A new relay sounds like a great idea. Any of the suggested protolangs would be good, though I think Proto-Western has the most potential for lots of very different daughters (such as would probably result from a relay).

Though I haven't been posting much, I'm reading everything that's posted here and on the wiki, and still working (albeit slowly) on the EI family. I'm also reading up some on the beginnings of agriculture and of urbanization in various parts of the world (mostly in preparation for doing Ngauro), so I'll definitely be interested in discussing that area of history some more.

But for now I only have computer access at work, so I probably won't be able to make very substantive contributions till I get a working computer again (hopefully soon!)...

Radius Solis · Post by **Radius Solis** » Sat Oct 04, 2008 2:10 am

Another thought: in the new relay, we could try to avoid some of the feature-discontinuity seen so far in Edastean (daughters often having little in common with any of their sisters, let alone their parent langauge) if we let each participant require one feature or characteristic of their language be retained in one or more descendents. Broad features, anyway - like "you must retain some manner of tone system" or "you must keep the aspiration distinction".

Radius Solis · Post by **Radius Solis** » Sat Oct 04, 2008 3:11 am

Oops, neglected to respond to this before.

cedh audmanh wrote:We should definitely combine a new relay with developing further historical detail. It would be much better to focus on broadening the information available on the two millennia between -500 and +1500 before moving too far beyond that with specific languages.

Well... why? That's when most of our information is for the Aiwa sphere, but is there any reason the new relay needs to closely match? And is there a reason its lines of descent should not extend several millennia into the future like the first one's did? I'm not against the idea of focusing on that period. But (my prior post notwithstanding) I'm somewhat allergic to placing many restrictions on what participants can do - too many checks and limits on creative freedom reduce the enjoyment level, and risk giving new folks the impression we're just trying to get free work out of them.

And we will have to describe the scenario to start from in more detail before we can really launch this project. I think it would be good to do some systematisation of what we have so far, and to set up an easy-to-follow starting point for newcomers. For example:

...

Agreed on all counts. One further element I'd add to your list: asking Zompist if he'd mind if we ported over all the remaining information from the Almeopedia, including his more recent entries (Faraghin), so that everything will all finally be in one place. And then we'd no longer have to be taking up space on his wiki, too.

As for the language families open for a relay:

- Western would be great, as there would indeed be quite a bit of carte blanche, but we don't have a halfway detailed geography for their region yet. We've decided to add more land over there, but I'm still somewhat unsure how to distribute it in order to (a) make a maritime north-west passage possible (needed for Sal's Cunning Plan, if he's still at it), and (b) allow for plausible tectonics (the latest layout requires the western mountains stretching northwest). Maybe some fjordlands and islands would be good...

- Nualis-Takuña languages should have occupied most of the northeastern peninsula by -500, with the only exceptions of Lotoka (Peninsular) and a few Eastern Isthmus. I agree that NuTak should be a major substrate for any Isles languages on the mainland, but I think Takuña itself would be even more likely to spawn some daughter languages. Accordingly, starting from PNuTak would be best IMO.

Should we narrow it down, then, to a choice between Western and NuTak? There are issues with both, but these seem to roughly balance out. And we don't have to decide for sure until we're prepared to actually start the new relay - so there's a bit of time yet to figure out which language's issues would be most readily addressed.

Issues with NuTak: there are a lot of cultural tie-ins with the Aiwa and Isles spheres, which is quite a double-edges sword - on the plus side, interconnection with existing material is good, but on the minus side, it creates a lot of work and restricts what participants can do. It certainly restricts what languages they can use as substrates; with Western they'd be free to make up their own.

Issues with Western: Either we have to finally hammer out a definitive map for the area (blech), or geographical uncertainty will leave people effectively working without one. Due to P-W's antiquity, it will take two generations to reach the classical era. And its descendents should, realistically, probably be polysynthetic - or close to it - and as any conlanger who's tried it knows, polysynthesis (fascinating though it is) can be an incredible creative headache.

Cedh · Post by **Cedh** » Sat Oct 04, 2008 3:54 am

Radius Solis wrote:
cedh audmanh wrote:We should definitely combine a new relay with developing further historical detail. It would be much better to focus on broadening the information available on the two millennia between -500 and +1500 before moving too far beyond that with specific languages.
Well... why? That's when most of our information is for the Aiwa sphere, but is there any reason the new relay needs to closely match? And is there a reason its lines of descent should not extend several millennia into the future like the first one's did? I'm not against the idea of focusing on that period. But (my prior post notwithstanding) I'm somewhat allergic to placing many restrictions on what participants can do - too many checks and limits on creative freedom reduce the enjoyment level, and risk giving new folks the impression we're just trying to get free work out of them.

What I mean is the following:

- The first relay suffered from a lack of historical setting for many of the langs, which made it difficult to combine them into a coherent scenario. To avoid this, people should be encouraged to work out a bit of history along with their language, and to view their relay contributions as part of a world rather than free-standing languages.

- Up to a certain date, roughly 500 YP, maybe some centuries later, there is not much contact between the Aiwa sphere and the rest of the continent. After that, there will be. Which will be very interesting, but requires fitting things into a scenario. IMO this is easier if we are working forward, on several strings of the same time in parallel, rather than having to backwards justify something said to happen a millennium later. (We haven't even tried fitting in most of the existing post-2000 languages precisely because we don't know enough about their background yet.)

- I'm not opposed to creating long relay chains at all, but I'd prefer working out more 1st or 2nd generation langs before tackling one of the 4th or 5th generation - i.e. we should aim for a broad tree, not primarily a high one.

(Also, I personally find it much more interesting and inspiring to work on a conlang in a scenario with rather little creative freedom, taking into account substrate influence and all that. But of course this need not be true for other people.)