Question: is there an explanation why S. Rodon "Power, Vala" gets plural Rodyn instead of expected **Redyn? Even if *Raudon is assumed, shouldn't it rather be pl. **Roedyn?
That's a really interesting question. I dug through my notes on the Sindarin plural of nouns involving the diphthong au and here is what I uncovered.
1) In Noldorin monosyllables, au became ui in the plural, such as iau pl. iui and thaun pl. thuin.
2) In Sindarin monosyllables, au became oe in the plural, such as naug pl. noeg and thaun pl. thoen (deleted example, PE17/81). In the second example, Tolkien also considered the plural thôn (another deleted example).
3) As a rule, in polysyllables, au reduces to o, as in Rodnor < † Raud-naur.
4) I found three plurals involving au in polysyllables:
a) Rodyn plural of Rodon († Raudon)
b) Noegyth plural of Nogoth († Naugoth), later plural Negyth
c) conin plural of caun
In the last example, Tolkien also wrote the plural as cōnin (PE17/102).
If I were to assemble this into a coherent set of rules, my guess is that au in a non-final syllable either resisted mutation or because ō (ǭ?), either of which reduced to short o giving seemingly irregular plurals.
In the case of Noegyth, perhaps it was influenced by the plural of the associated monosyllable noeg. Given the small number of examples, these rules are just a guess.
Note that the Sindarin plural list in the link is incomplete because I haven't entered all the data yet, and some of the linkage for plurals are not working (e.g. naug plural noeg is missing).
A parallel in WJ is Golodh, pl. Goelydh from ºñgolod- (pp. 379, 383), later plural form Gelydh (p. 181). In these plurals oe is the monophthong written ö in the table of the Angerthas (LOTR app E II), which was produced by i-affection of short o . It had throughout changed to e in 3rd age Sindarin.
Touché! Eep, Gœlydh and Nœgyth. I had temporarily forgotten about that, given that we were discussing au > oe. I suppose I'll have to be more careful about determining whether Sindarin oe is really the diphthong, since Tolkien didn't always write "œ" where it's implied. I would have known Goelydh was the monophthong due to its derivation, but thinking of naug distracted me with Noegyth, even when we have negyth to help.
I am pretty sure o > œ > e is the normal phonetic development of short o before i, not just in plurals. Compare the presumably archaic Arnoediad (Silmarillion) with the form Arnediad (Etym:NAY, WJ:71). At least, that is what Salo says: A Gateway to Sindarin p. 48 §4.104 and p. 63 §4.189.
I still think this all points to a general rule of au in the first element of polysyllables resisting mutation in plurals and then later simplifying to o.
I'd agree that o > œ > e is the regular phenomenon for i-affection of non-final-syllable o (which doesn't just happen with plurals), and that Arnoediad is a good example: we even have nœdia- in Etymologies, as you note in Eldamo. I'd agree it's safe to assume that OE in words like Arnoediad and Eboennin is [ö].
Beregond, Anders Stenström Mar 04, 2015 (21:26)
Tamas Ferencz Mar 04, 2015 (23:44)
Paul Strack Mar 05, 2015 (04:56)
1) In Noldorin monosyllables, au became ui in the plural, such as iau pl. iui and thaun pl. thuin.
2) In Sindarin monosyllables, au became oe in the plural, such as naug pl. noeg and thaun pl. thoen (deleted example, PE17/81). In the second example, Tolkien also considered the plural thôn (another deleted example).
3) As a rule, in polysyllables, au reduces to o, as in Rodnor < † Raud-naur.
4) I found three plurals involving au in polysyllables:
a) Rodyn plural of Rodon († Raudon)
b) Noegyth plural of Nogoth († Naugoth), later plural Negyth
c) conin plural of caun
In the last example, Tolkien also wrote the plural as cōnin (PE17/102).
If I were to assemble this into a coherent set of rules, my guess is that au in a non-final syllable either resisted mutation or because ō (ǭ?), either of which reduced to short o giving seemingly irregular plurals.
In the case of Noegyth, perhaps it was influenced by the plural of the associated monosyllable noeg. Given the small number of examples, these rules are just a guess.
Here are my sources:
http://eldamo.sourceforge.net/content/words/word-2608166765.html
Note that the Sindarin plural list in the link is incomplete because I haven't entered all the data yet, and some of the linkage for plurals are not working (e.g. naug plural noeg is missing).
Björn Fromén Mar 05, 2015 (16:01)
According to WJ:413 nogoth doesn't derive from naug . Rather it reflects "some such form as *nukotto/a".
Matt Dinse Mar 05, 2015 (22:42)
We also have Nogoth, pl. Noegyth in WJ:388.
Björn Fromén Mar 06, 2015 (23:32)
Matt Dinse Mar 07, 2015 (17:19)
Paul Strack Mar 07, 2015 (17:43)
I still think this all points to a general rule of au in the first element of polysyllables resisting mutation in plurals and then later simplifying to o.
Paul Strack Mar 07, 2015 (17:49)
Paul Strack Mar 07, 2015 (17:53)
Matt Dinse Mar 08, 2015 (07:04)