A counter-proposition to Jallings' law-words: I'd much rather have it *thain (Q cognate sanye) as "law" and *thaineb as "lawful, law-abiding", than *thain/thein as cognate of Q sanya for "lawful" and *thainas for "law". I might be narrow-minded, but It feels much better to have the law come first, and the preference to conform to it come after, not before.
In my view they don't necessarily need to come from one another at all. STAN is a verbal root, so the development would be the root receiving a standard adjectival ending ia/ya for th adjective and a nominal ie/ye for the abstract noun.
I'm still uncertain on how to approach this but I'm certain of one thing, the word can never (= even with suffixes) be thein. The root is STAN, not STEN: [stanja] > [stʰanja] > [sθanja] > [θanja] > [θanj] > [θain]
If it was STEN: [stenja] > [stʰenja] > [sθenja] > [θenja] > [θenj] > [θein] > [θain] (when final, but thein- with suffixes)
Following Salo's Gateway, I get Fiona's flow and result, following patterns from actual attested word formation, I end up with Ekin's process and result. I think we might have a problem here.
+Fiona Jallings Quoting you, "[θania] > [θenia] > [θeni] > [θein]", I-Affection before the intrusion. I think it's the opposite, otherwise it would turn every (base vowel) A into E, and if I'm not mistaken, you would have to change your: Non-Final AI > No change | Final AI > No change Final AI from EI > Î / I (polysyllable) Non-Final EI > No change
But you probably shouldnt, because: fain, pl. fain [PE17/179]
I haven't seen him around for quite a while but I think +Paul Strack could help us here. He studied these for long, if I'm not mistaken. And Eldamo also shows I-affection before intrusion in the Phonetics section (though it shows another, a seemingly different kind of I-affection (with diphthongs) after that + after intrusion, which I think should be the real one), yet more than a few of the S./N. entries skip I-affection, which actually makes more sense.
Fiona is much more knowledgeable about Sindarin than me. My focus has always been on Quenya. I think she is right, though, about i-affection before intrusion. There is an explicit example of this by Tolkien on PE17/148:
As for the "second i-affection rule" in Eldamo, if I recall correctly, it is a placeholder for proper analysis for how i-affection propagates forward through diphthongs. I abandoned the analysis to gather more evidence first because I was having a hard time figuring it out. As it says in Eldamo, the Sindarin and Quenya phonetic analysis is very much incomplete. eldamo.org - Eldamo : Sindarin : laich
Regarding ai > ei, the evidence is very muddled and inconsistent. I am not sure Tolkien ever made up his mind. It seems that in the Noldorin period at least he was vacillating between ai > ei or ei > ai.
Sindarin seems to be a bit more consistent, and seems to point to ei > ai in final syllables. There are only a few Sindarin examples of ei appearing in non-final syllables: edhelvein "Elven-Fair", feir "mortal" and mein "first", although the last example is archaic and replaced by minui, so it could predate ei > ai.
Conversely, I don't think there are any examples of ai appearing outside a final syllable. The example Ereinion "Son of Kings" is particularly interesting, since its initial element outside the compound is erain "king".
Ugh. I just took another look at the example I wrote above and it doesn't demonstrate what I claimed at all, and may in fact demonstrate the opposite. The example has no i-affection at all.
OK, all I can say is that it looks like Sindarin is as much of a contradictory mess as Noldorin. There are four possible rules:
1) a-affection (pretty well understood)
2) i-affection (before or after i-intrusion)
3) i-intrusion (before or after i-affection)
4) Final ei > ai (maybe)
If all four rules were applied consistently, the primitive iXja, eXja and aXja would all produce the same result in Sindarin, one of aiX, eiX or îX. But there are examples of each of these:
+Fiona Jallings Thanks for the link. I don't know why I've never seen that article before. It's amazing! It's going to take a while for me to absorb it.
I Lam Arth is a great archive of old articles on Sindarin scholarship, if you want to get seriously into studying Sindarin, I strongly suggest reading the articles in it.
+Paul Strack Everything works out if you remove the first I-affection in Eldamo's phonetics section, in my opinion. The second one is placed after the intrusion (an edit like: "[i] + its diphthongs" would take care of the problem).
There's also "[—j—] > [—i—]" which (in Eldamo) comes before both, and adds to the confusion. If there's a clear distinction between j and i, and if medial j did become i before the other two rules, the chances of j-intrusion happening is almost non-existent. Yet we see it, and "[—j—] > [—i—]" as well. A mess, indeed.
These are the examples I looked at. I searched for S words with ai as final syllable, and found none with explained word formation that showed i-affection first and i-intrusion later, but several N words showing the opposite pattern. I missed the fact that they are N words and not S, of origin, so my argument is not that valid, but it's still disturbing that the rule of i-affection before i-intrusion is not found in examples, except for the inconsistent erja>îr /or/ >air.
Fiona Jallings Oct 13, 2016 (09:27)
Tamas Ferencz Oct 13, 2016 (10:15)
Ekin Gören Oct 13, 2016 (14:52)
[stanja] > [stʰanja] > [sθanja] > [θanja] > [θanj] > [θain]
If it was STEN:
[stenja] > [stʰenja] > [sθenja] > [θenja] > [θenj] > [θein] > [θain] (when final, but thein- with suffixes)
Fiona Jallings Oct 13, 2016 (18:10)
[stanja] > [stʰanja] > [sθanja] > [θanja] > [θania] > [ θenia] > [θeni] > [θein] > [θain]
Hjalmar Holm Oct 13, 2016 (21:48)
Hjalmar Holm Oct 13, 2016 (21:50)
Fiona Jallings Oct 13, 2016 (21:58)
Ekin Gören Oct 14, 2016 (12:32)
Non-Final AI > No change | Final AI > No change
Final AI from EI > Î / I (polysyllable)
Non-Final EI > No change
But you probably shouldnt, because: fain, pl. fain [PE17/179]
I haven't seen him around for quite a while but I think +Paul Strack could help us here. He studied these for long, if I'm not mistaken. And Eldamo also shows I-affection before intrusion in the Phonetics section (though it shows another, a seemingly different kind of I-affection (with diphthongs) after that + after intrusion, which I think should be the real one), yet more than a few of the S./N. entries skip I-affection, which actually makes more sense.
http://eldamo.org/content/phonetic-indexes/phonetics-s.html
http://eldamo.org/content/words/word-2030028405.html
There is certainly a problem here, I think.
https://plus.google.com/photos/...
Paul Strack Oct 15, 2016 (01:43)
http://eldamo.org/content/words/word-3698039951.html
lisyā > liχı̯ā > leχı̯ > leich > laich
As for the "second i-affection rule" in Eldamo, if I recall correctly, it is a placeholder for proper analysis for how i-affection propagates forward through diphthongs. I abandoned the analysis to gather more evidence first because I was having a hard time figuring it out. As it says in Eldamo, the Sindarin and Quenya phonetic analysis is very much incomplete.
eldamo.org - Eldamo : Sindarin : laich
Paul Strack Oct 15, 2016 (02:04)
Sindarin seems to be a bit more consistent, and seems to point to ei > ai in final syllables. There are only a few Sindarin examples of ei appearing in non-final syllables: edhelvein "Elven-Fair", feir "mortal" and mein "first", although the last example is archaic and replaced by minui, so it could predate ei > ai.
Conversely, I don't think there are any examples of ai appearing outside a final syllable. The example Ereinion "Son of Kings" is particularly interesting, since its initial element outside the compound is erain "king".
Paul Strack Oct 15, 2016 (02:10)
lisyā > liχı̯ā > leχı̯ > leich > laich
I need to think about this some more.
Fiona Jallings Oct 15, 2016 (02:25)
What you demonstrated was A-Affection, by the way.
But, here's an excellent article that you need to read: http://www.tolkiendil.com/langues/english/i-lam_arth/vowel_affection_sindarin_noldorin
tolkiendil.com - Vowel Affection in Sindarin and Noldorin - Tolkiendil
Paul Strack Oct 15, 2016 (02:27)
1) a-affection (pretty well understood)
2) i-affection (before or after i-intrusion)
3) i-intrusion (before or after i-affection)
4) Final ei > ai (maybe)
If all four rules were applied consistently, the primitive iXja, eXja and aXja would all produce the same result in Sindarin, one of aiX, eiX or îX. But there are examples of each of these:
http://eldamo.org/content/words/word-3698039951.html
lisyā > liχı̯ā > leχı̯ > leich > S. laich (but elsewhere leich)
http://eldamo.org/content/words/word-3345570783.html
eryā > erı̯a > erı̯ > irı̯ > S. îr (but elsewhere air)
http://eldamo.org/content/words/word-4207068237.html
teleryā > S. teleir
So I am afraid I have only managed to add to the confusion.
eldamo.org - Eldamo : Sindarin : laich
Paul Strack Oct 15, 2016 (04:34)
Fiona Jallings Oct 15, 2016 (06:53)
Ekin Gören Oct 15, 2016 (14:22)
There's also "[—j—] > [—i—]" which (in Eldamo) comes before both, and adds to the confusion. If there's a clear distinction between j and i, and if medial j did become i before the other two rules, the chances of j-intrusion happening is almost non-existent. Yet we see it, and "[—j—] > [—i—]" as well. A mess, indeed.
Hjalmar Holm Oct 15, 2016 (15:22)
These are the examples I looked at. I searched for S words with ai as final syllable, and found none with explained word formation that showed i-affection first and i-intrusion later, but several N words showing the opposite pattern. I missed the fact that they are N words and not S, of origin, so my argument is not that valid, but it's still disturbing that the rule of i-affection before i-intrusion is not found in examples, except for the inconsistent erja>îr /or/ >air.
[kalarjā] > [kalarja] > [kalarj] > [kalair] > [kelair] > [keleir]
[madja] > [maðja] > [maðj] > [maið] > [meið]
[spanja] > [spʰanja] > [sɸanja] > [ɸanja] > [fanja] > [fanj] > [fain] > [fein]
http://eldamo.org/content/words/word-1240862291.html
eldamo.org - Eldamo : Sindarin : celair
Fiona Jallings Oct 15, 2016 (17:52)
Paul Strack Oct 15, 2016 (18:47)
https://github.com/pfstrack/eldamo/issues/5
I don't want to tackle until I've had time to do the background research.
github.com - Change order of I-affection and j-intrusion for Sindarin · Issue #5 · pfstrack/eldamo