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

[Radius Solis]

Interesting. An auxiliary-oriented verbal system... (I see the future auxilliary, none, appears to be cognate with the verb to go?

It is indeed. Fáralo turned 'go' into an auxiliary, and that's one of the ones I kept.

I don't know for Rad, but for me it is reminiscent of the V2 word order of germanic languages.

Yes, but the thing is, German has SOV underlying the V2... The family ancestor, though, has VSO. I suppose, since Fáralo provides the SOV base here, Radius isn't really in a position to answer the question.

Yeah, it's Fáralo that developed AuxSOV as the word order for the auxiliary conjugation, and I simply elected to keep it. One can't drastically change everything, after all. And besides, the timespan from Fáralo to Puoni is less than a millennium. So the guts of the language retain a lot of similarity; nearly everything is only a moderate or incremental change from the parent language, with the only really severe difference being the following sound change sequence:

1. V -> V: / _CC, _C# coupled with C -> 0 / _C, _# (coda loss with compensatory lengthening)
2. V: -> uV / [+velar], [+postalveolar]
3. V: -> iV / _

...which (aside from a few exceptions and some cleanup) is all it took to mutate Fáralo into something with an Italianate feel to it. All word-final consonants in Puoni are from loanwords.



Speaking of words, here is a raw text file of the lexicon: http://www.u.arizona.edu/~bpawlows/puoni/puonilex.txt (grammatical words not included yet)

The entries still need a little cleanup, but provided you can understand my notation and my ascii rendering of Faralo (probably not too hard) you can see all the raw info that will be on the nice pretty dictionary page once I get that done. .....whoops, I noticed only now that, it being .txt, the diacriticked characters I'd pasted in for Naidda etymologies have been borked. Oh well, I'll fix it on the dictionary page.

[Radius Solis]

Also : I'm in the opinion that new comers in this relay should be allowed to just make a daughter of the language they like, instead of keeping the team thing. That way people don't have to wait for months, and the resulting tree will be much more realistic.

I completely agree. The relay has already beautifully fulfilled my original purpose for it. At this point I think it would be appropriate to make what we've been doing lately anyway, which is "whatever we feel like doing", into the main game. And to say "welcome!" to anyone who wants to jump in wherever and whenever they like.

[kodé]

The IPA version of the Aghïyï page is : http://www.soapboxindustries.com/zbb/aghiyi-IPA.html. I've still got to upload the conjugation of the auxiliary verbs, description of syntax, vocabulary, and the translation. This will probably take a while; thus, Pyurio (and anyone else who wants to derive a daughterlang from Aghïyï) should start their work now on sound changes, since the phonology section is complete.

[Cedh]

Before I get too carried away, could someone point me to a map of the region where the Edastean family is spoken? I thought I'd seen one before, but can't find it now. I'm wanting to know which langs are nearby spatially and temporally that I can take borrowings and other contact influence from.

Here's one:


It's a map of the time of Adāta, which was spoken in Rathedān. Kasca is the land that the sample text refers to (koxað in Mavakhalan); its language is Naidda. Huyfárah is the country where Fáralo is spoken.

On the map here you can see the town of Khalanu, whose language is Mavakhalan. If I put things together correctly, it is approximately where the <n> of Rathedān is in the map linked to above.

And here's another map, from the time of Ndak Ta:


As for temporal relations, see here. From what I've gathered from the language descriptions and cultural notes, Ayāsth would be fairly nearby and ~1000 years after Mavakhalan, as well as having lots of prestige. E'át would also be from that time period, as would Yād.

[Zhen Lin]

Also, a note for 4pq1injbok, I've just noticed that some fine-tuning I made to the sound changes has created a new declension class. This necessitates some changes to the various examples.

Which one was it? The true/false consonantal stem distinction?

Actually, the nasal-initial class, which had to be separated from the rest of the consonant-initial classes.

I have uploaded the newest version of the lexicon and grammar of Mavakhalan.

[kodé]

Damn, sorry (especially to Pyurio ) that I haven't uploaded anything in the last few weeks; I don't really have much of an excuse, either. BTW, this page is now the one that's the most updated--and it has IPA!!

In other news, after I finish uploading ?Aghiyi's verb morphology, syntax, lexicon, and the translation (with script, if I can get a font maker onto my damn system), I'll start working on tri-consonantal "grand-niece" language to ?Aghiyi. Basically, instead of having unstressed vowels harmonize with the first vowel of the word, I'll just delete the bastards, creating a CVC(C)(C) pattern for most of the words. Then, instead of relegating prefixes to a small subset of words, I'll let them blow holes in the words they prefix to. I'm pretty sure I can get a tri- (or quatri- or quinqui-)consonantal lang out of this mess, whose speakers will probably be contemporaneous with those of Pyurio's daughterlang, depending on how many centuries worth of sound changes he makes to ?Aghiyi.

I'd very much support this move and'd be willing to derive more words to Arie-with-an-accent.

Have you done any of this? If so, that's great; I'd love more vocabulary for ?Aghiyi! Of course, Nuntar would have to do his sound changes before me, but if he doesn't mind, then I'm all up for adding my sound changes into the mix.

[TzirTzi]

I'm afraid I was talking about the idea of using more of the now expanded ndak ta vocab - but I've never received anything from con quesa, and I don't know if that idea actually ever got anywhere.

However, once I've finished Yet (morphology now finally finished!) and my exams (last one tomorrow) then I'll be happy to make some more compounds/blends/derivations/borrowings/coinings in Arie .

[Radius Solis]

I'm afraid I was talking about the idea of using more of the now expanded ndak ta vocab - but I've never received anything from con quesa, and I don't know if that idea actually ever got anywhere.

However, once I've finished Yet (morphology now finally finished!) and my exams (last one tomorrow) then I'll be happy to make some more compounds/blends/derivations/borrowings/coinings in Arie .

My fault.

Well, also dewrad's computer's fault, somewhat, as he lost the NT > Adata sound changes in a computer messup some time ago and there's been no great enthusiasm from him or from me for reconstructing the list. In addition the extra vocab is/was spread across multiple disorganized files that I was trying to consolidate and clean up for the purposes of letting everyone have it, but that task was interrupted by computer troubles of my own and I still haven't gotten around to finishing it yet, without even knowing if there'll ever be changes to run them through for Adata.

So in theory the project is far from out of the question, but nothing has been done for it yet. Exactly how much enthusiasm is there for expanding all your lexicons? If there's a bunch maybe we can start trying to get it together...

[TzirTzi]

I retain my original enthusiasm for it - as, by the sounds of it, do kodé and Nuntarin. I suppose the significant thing, though, is how many of the people nearer the top of the tree still have their sound changes..

[Cedh]

...my motivation and momentum for Puoni got a serious kick in the pants when the grammar I was writing on the KneeQuickie went poof into thin air. Entirely my fault, I know.

Rad, did you know that Google seems to have cached what you had written up at the time?

[4pq1injbok]

(Ack, shouldn't go so long without looking at the board. I'm bad that way.)

Yay for the recovery of the Adāta sound changes!

I approve of the proposal to rederive more words forward from Ndak Ta on a large scale -- bigger lexica are always a good thing -- but hope there'll be more to this push for bigger lexica than just so much mechanically running lexical items through sound changes.
As an example that's really the main thrust of this post, borrowings are another way to bulk up one's lexicon. Now, granted, if there are no large source lexica to borrow from, that won't really work. But I've been introducing new items in my own lexicon (via semantic shift and derivation and whatnot), and surrounding langs are welcome to borrow them, if their creators want to and decide that it sociolinguistically makes sense.

My lang has picked up a name, Ājat he-Heloun, and I've started jotting down stuff about it here -- beware, nothing there is really presentable yet. The lexicon containing the words I've done so far is here.

[Radius Solis]

I approve of the proposal to rederive more words forward from Ndak Ta on a large scale -- bigger lexica are always a good thing -- but hope there'll be more to this push for bigger lexica than just so much mechanically running lexical items through sound changes.

I definitely approve of the sentiment that there's far more to lexicons than just running a word list through a sound change set. Stopping there leaves a lexicon feeling mechanical and dull. There's semantic shifts to be concerned with, and all manner of other lexical processes that occur...

Really, I have never understood the "depleted-lexicon problem" - or rather, I understand feeling pressured for time and so not doing as much as one could, but one can always go back and rectify it. Every step of the relay each language could easily have more words than the prior step did. Puoni's currently got about 1.5x as many as Faralo had, and that's without using any of the extended Ndak Ta vocab. How did I accomplish that?

1. I borrowed words from other languages, including pre-existing ones in the area (Naidda), contemporary other members of the relay (Æðad? and A?áta; Ay?sth was set too far in the future for it), and two mini-conlangs I made up for the sole purpose of having more languages to borrow from (Miw and Tlaliolz). 2. Morphology! I applied Faralo derivational morphology to roots in that language and sound-changed the results, and typically shifted the meanings around a bit (or sometimes a lot) as well. En masse. It works. 3. Compounding: just as with derived words in the parent language, you can create new compounds in the parent language and sound change the results. This too is a rich potential source of new words. 4. Other processes like clippings, back-formations, contemporary compounding, and the creation of new derivational morphology. Really, there is no reason the artful and enterprising conlanger cannot double or triple the number of words of a parent language in its daughter.

Also - you may want to hold off on the -ta derivatives, since I recently coined the verb þazata (to succeed), from þaza (to do).

Is this -ta a reflex of Ndak Ta?

[Zhen Lin]

Possibly. It's a full verb in Adāta, meaning to become. I know it's a bit of an about-turn in meaning, but...

[Zhen Lin]

I have uploaded a new sample text and some revisions to morphology and the lexicon to http://www.geocities.jp/lowzhenlin/mavakhalan/.

[4pq1injbok]

Ah... /ð/ to /j/ in various places - is that Ayāsth influence or just a coincidence?

Well, Ājat itself is an Ayāsth borrowing, and I figure Ayāsth influence is also likely to be behind varieties where my /d/ (the regular reflex of /ð/) is [j].

Also - you may want to hold off on the -ta derivatives, since I recently coined the verb þazata (to succeed), from þaza (to do). I was thinking more of a resultative/perfect meaning, rather than an inceptive one.

Hmm, okay. Given that I still want an inceptive, is there a better source verb? Or maybe I'll just make the same -ta have either function, depending lexically on the base. Yeah, I like that.

Also, just a heads-up: I've just discovered that there are a category of irregular verbs which have two stems.

Neat. The alternation will be disappearing, though, unless any of them happen either to survive as an s-verb, or exhibit different stems as a gerund and as a deverbal noun.

[Zhen Lin]

Yes. For example, the verb fazôs- has the deverbal fazôs (of course) and the participle fassêję.

As for the inceptive, perhaps a prefixed sote- (to begin)?

[4pq1injbok]

Mm, prefixed, right. I had decided that my innovated morphology would be suffixing, and likewise that compounding would switch to head-last, but now that you mention that I can't for the life of me remember what justification I had for that. After all, (ttbomk) compound order is one of those things that tends to be conservative and doesn't just change at the drop of a hat, and there doesn't even seem to be a neighbouring lang I can claim influence from.

I hope I don't have to rework it all to prefixing, partly 'cause I'm already kind of attached to the suffixes, but also because any prefix with an odd number of syllables would cause pervasive stem alternations of a sort I'd rather not have (I guess I could analogise them away, though -- or let my stress rules see through derivation?).

[Corumayas]

[EDIT: I've revised this list, changing the order of a lot of rules, adding some things I missed, and trying to better explain some of the irregularities. I may tinker with it a little more, but I think it's close to being as good as I can get it now. Asterisks still mark irregular or uncertain changes, and I've given more examples for some of the more obscure or ambiguous ones.]

So now that this is actually happening, I looked over those sound changes again, and I found lots of little complications I didn't notice before. In particular, changes involving vowels (especially diphthongs) and stress seem to be often irregular or unpredictable.

Here's what I've got now... irregular changes, or ones whose conditions I'm not sure about, are marked with asterisks, and I've given all the examples of those in parentheses. The most common/regular development is listed first, followed by the exceptions/irregularities; I think that's easier to read, although an SCA program would want them in the reverse order. I think I've got most of the changes in the right order otherwise.

Anyway, if anyone wants to check my results or try to clear up some of the confusing parts that'd be great.

Ndak Ta/Adāta sound changes [revised!]

m=,n=,ŋ=/a/_


Stress moves to the first syllable in all words.


au/"ū/{closed syllable} (12 words)
*au/(")ā/{closed syllable} (2 words: maunde/"māthe, rauŋ/rān)
au/"ō/{open syllable} (~35 words; unstressed in mpau/"apō)
*au/"ē/{open syllable} (1 Gezoro loanword: kuthau/ku"thē)

ãu/"ū/{closed syllable} (regular by analogy with au; 1 word: mbãur/a"būr)
*ãu/(")ā/{closed syllable} (1 word: ntãug/"atāx)
*ãu/(")ūn/_# (1 word: tsilãu/"silūn; but cf. idãu/i"dō)
ãu/"ō/{open syllable} (3 words: idãu/i"dō, ãukwai/ō"pē, tsãuki/"sōki)

ai/"ā/{closed syllable} (~28 words)
*ai/"ē/{closed syllable} (2 words: baitrai/bē"rē [cf. gaibra/xāra], paiŋkwu/pēkhu [cf. aimbu/āphu, aiŋgàu/ākhō, etc.])
ai/"ē/{open syllable} (~48 words; unstressed in ~7 or 8 words)

ãi/"ā/{closed syllable} (no examples, but this would clearly be the regular development)
*ãi/"ān/_# (2 words: abãi/a"bān, & unstressed in kumãi/"kumān; but cf. the following)
ãi/"ē/{open syllable} (4 words: ãidai/"ēdē, ŋwãi/iē, lewaitsãi/lezē"sē, & unstressed in ŋwolãi/"iolē)

*(Gezoro oi seems to pattern with ai, more or less: thunoi/thu"nē, gekhoig/xe"khāg, gekhoija/xe"khāza; OTOH honkhoi/hokho)

NB: The latest change (as listed above) that affects a given word is assumed to be where the stress stays; in words with two identical diphthongs the stress falls on the second one (baitrai/bē"rē, naisŋai/nā"iē, bwaiiskaimel/zēi"kēmel). Exceptions to the expected stress patterns, according to these rules, include dauwai/"dōzē (instead of dō"zē) and ai"lau/ē"lō (instead of "ēlō). If these rules are correct, ē"lō appears to be the only word with irregular stress that doesn't stress the first syllable.


V~/Vn/_#
V~/V/_C


m,ŋ/n/_#


mb,mp,ŋkw/ph/V_V
nd,nt/th/V_V
ŋg,ŋk/kh/V_V

*rt/th/V_V (airti/āthi; but cf. tsurtor/sūtor, orte/ūte)


kw(h)/p(h)/_ (except paiŋkwu/pēkhu, where kwh/kh)
*(gw/b/#_ in Gez. gwaunkhōs/būkhōs)

bw/w/_
w/j/_

ia/ija/_ (miasko/mizāko)

j/z/#_, V_V

m,n,ŋ//_j

*j/i/C_V (mwarwobm/iarioba suggests this)


s/:/V_# (except Gez. gwaunkhōs/būkhōs)
s/:/V_o (disau/dī"ō, & Gez. reso/rēo, wesau/zē"ō)
*s/:/V_a (asa/ā, nasa/nā may imply this)
s//V_V
s/h/#_

*(Gezoro x seems to behave like s: nalaxempha/nalāepha, which may then be more evidence for s/:/V_a)


st,ts/s/_


*mm/m/V_V (ammi/ami)

VC/V:/_C
*(in orte/ūte--and that word only, AFAICT--there's also a change in vowel quality)


ŋ/j/_V

g/j/_i,_u
*g/j/i_V (Gez. digos/deiō; no counterexamples...)
g/x/_ (except some Xšali loan words; & oddly, Gez. gekhoig/xe"khāg)


*ji/i/e_ (deŋi/dei)
ji/ja/{most of the time} (gibm/iaba, giŋko/iakho, ŋimbu/iaphu, orŋi/ōia)
*ii or ji/ei/?C_# (apwi/apei)

*je,ie/ja/{rarely} (ŋende/iathe, malwe/malia; but not in at least 7 other words)

*ij/ej/_a,_o (imbwa/eia, & Gez. ijō/eiō, digos/deiō)
*ij/i/_a,_e (simwai/hi"ē, Gez. khiŋjara/khiara)

*ēe,ēī/ēi/_ (aise/ēi, bwaiiskaimel/zēi"kēmel)
*(possibly other similar changes turning vowel sequences into diphthongs?)

*VV/V/_ (maybe) or /V:/_ (maybe)

VV:,V:V,V:V:,V::/V:/_


(j/i/_ in writing)

[Radius Solis]

Incidentally, I asked dewrad about the change you listed of rt > th, and it seems that there's only one instance in the lexicon, for which he says the <h> was probably a typo. The rule should be Vrt > V:t.

[Corumayas]

I wonder if some of the other oddities aren't typos too.

Another one I just found: orte/ūte (instead of the expected ōte).

Also runnak/rūnak (rather than runak), so ammi/ami is unique after all; and I missed s/:/V_# (e.g. das/dā, mos/mō, peras/perā, wi"mes/zimē). And there's nasa/nā, which is either another example of compensatory lengthening when s was lost, or a counterexample to my idea that two short vowels don't merge into a long one (aa/ā).

A few of the changes seen in Gezoro loans are also strange. I've mostly ignored them, since the changes from Ndak Ta are all we should need, but some may be evidence for regular changes: e.g. reso/rēo, wesau/wē"ō, etc. are more instances of compensatory lengthening.


Of course, some irregularity isn't really a problem; it adds verisimilitude...

[4pq1injbok]

*s/:/?V_o (compensatory lengthening: disau/dī"ō is the only example)

I don't have the nonexamples to hand, of course, but could the environment conditioning this be the pretonic position?

edit: oops, didn't see the latest Gezoro thing; that does make _o look good, doesn't it.

[Corumayas]

Here are all the examples of s//V_V that I've found in the lexicon.

Vsa- asa/ā, aisab/ēab, maisa/mēa, nasa/nā, ausa/ōa, Kasadgad/Kāxad
Vse- aise/ēi
Vsi- dasi/dai, susi/hui
Vso- auso/ō, disau/dī"ō; Gez. reso/rēo, wesau/zē"ō
Vsu- esul/eul, saisung/hēun

Obviously in some words the vowel before /s/ is long for other reasons; but it looks like it always becomes long before /so/, and possibly also before /sa/ (based on asa/ā, nasa/nā, if two identical short vowels don't usually merge into a single long one... but I can't remember what made me think that was the case-- I think it was a particular word I saw, but I can't find whatever it was now).


BTW, if it'd be helpful, I can update my post above as the rules get more refined.

[Radius Solis]

Of course, some irregularity isn't really a problem; it adds verisimilitude...

Yeah, it does. It's very realistic to have to deal with such uncertainties as what really happened to intervocalic /s/, etc. Few natural languages, even among those with well-established and thoroughly-investigated diachronic histories, lack a few words here and there that just don't quite fit certain rules that appear sound in other cases. The fact that this can be simulated by, probably, a combination of tricky rules and scattered minor errors on the part of a conlanger is kinda amusing on one level, and on another, I'm grateful for the extra dose of realism.

I personally see nothing wrong with calling the current list good enough. If in retrospect there turn out to have been any systematic differences between this list and whatever the original one was, that can be quite realistically chalked up to linguistic variation within the speech community in its past or present.

[Corumayas]

I personally see nothing wrong with calling the current list good enough. If in retrospect there turn out to have been any systematic differences between this list and whatever the original one was, that can be quite realistically chalked up to linguistic variation within the speech community in its past or present.

I agree, but I'm also getting kinda hooked on puzzling this out.

It seems likely that most of the unpredictable stress is simply because Dewrad missed marking noninitial stress in a few words. So clearly some dialect(s) of early Adata had initial stress in all words. Also there was probably a dialect where word-final nasal vowels acquired a following /-n/ before the changes to diphthongs, so that /a~i# a~u#/ got it too and then patterned with the diphthongs in closed syllables instead of with the ones in open syllables (as is slightly more common: 3 words vs. 4). And another one, more poorly represented, where /au/ merged with /ai/ in at least some positions, becoming /a:/ instead of /u:/ in closed syllables (2 words, vs. 12 with /u:/), and maybe even /e:/ instead of /o:/ in open ones (one word, kuthau/ku"the:, but it's a Gezoro loan).

More importantly, I've decided that the change {syllabic nasal}/a/_ needs to come before diphthong monophthongization (on the basis of aitn/e:ta & taibm/te:ba, where /ai/ acts like it's in open syllables). And the diphthong changes should be listed in the following order: au, a~u, ai, a~i (on the basis of which one usually gets the stress if there are 2 diphthongs in a word-- only dauwai/"do:ze: contradicts this, out of six words total). [EDIT: I just found another, ai"lau/e:"lo:, so that's two exceptions out of seven words.]

I think I'll keep tinkering with the rules for a while, and maybe update that big post with any significant improvements that I come up with. At least it can be cleaned up a little, and I can add the few changes I missed.


EDIT: I revised the rule list, including major changes to the order of the rules. Random irregular words aside, I think the only significant uncertainties at this point are in the last section, with the various outcomes of certain vowel sequences and of /ji/, /ij/, etc. (where I think there just aren't enough examples to figure out the rules).

[krinnen]

Hey, I want to work in a a daughter language once again. Just in case, this time I'll do the work in pen & paper, so I don't have to worry about my new HDD...

So, what should I do?