Taur in compounds

Hi everyone,

I was wondering how to use taur correctly in compound words. In polysyllables, au usually becomes o, but there are also a few exceptions to this. E.g., if Rhudaur is a compound of rhu- and taur, the soft mutation of taur would be daur here. The fan-invented word rostor as a compound word of ross and taur changes the au to o and in Thorondor we can also see the shift from au to o.

For example, if I would like to make a compound of rhovan and taur, would Rhovandaur or Rhovandor be the better choice.

There are a couple of instances in which au > o didn't happen:

• Gaurhoth, Gorthaur, Rauros, Tauron: here it seems that an o in another syllable inhibits this change. For this reason Elaran has revised rostor to rostaur (the latest version of Eldamo isn't imported to Parf Edhellen yet). In Thorondor I would explain the reduction by analogy: Tolkien cites the suffix as -dor with lenition and reduction already carried out so it might have become an independent word not imideatly connected to taur any more (just like I wouldn't expect ortha- to conjugate orthaun, orthaug... just because it contains an o).
• Rhudaur, Draugluin and Glaurung suggest that u might inhibit this sound change as well (Turgon is a counterexample but again -gon might be a recognized suffix).
• Bauglir, Anfauglith, Naugrim, ilaurui are hard to explain, I would treat them as words from dialects where the reduction of au was less prevalent (like for example in German ü became i in words that entered the standard from some dialects but not from others).

For a more detailed exploration of the evidence, take a look at Paul Strack's article about this sound change.

Now rhovan + taur is tricky for two reasons:

1. The primitive form would be srāban-taurē > rhauvan-naur so the question becomes how two aus interact.
2. Unlike in Rauros, Rhudaur etc. there is an intervening syllable that could block the interaction here.

Personally I would guess Rovathor in this case, Rauvathaur seems like a bit of a mouthful. It might be possible that one reduces but the other doesn't but I wouldn't speculate on that.

(And primive n + t in compounds developes to nth > th cf. jēn-kwantā > iphant/ifant, en-kwet- > ephed-, n + t > nd only happens if there was a primitive vowel between: kjeleperina-tal > Celebrindal)

Thanks a lot for the detailed explanation.

But I don't really understand how jēn-kwantā > iphant is a proof that primitive n+t > th. Isn't that rather a proof for n+k >ph? Wouldn't ndan+tul > dandol be a hint that n+t >nd can also happen without intermediate vowels? There are numerous words that prove that with nd happens when a vowel stands between primitive n and t (Thranduil, Thorondor, Nindalf), but I could not find a direct proof for n+t > th by now. Do you have a concrete example?

You are right, n + lost vowel + t becoming nd is very well attested, but I don’t think we have an true example of n + t without an intervening vowel, so I extrapolated n·t > þ from m·p > f (kw becomes p very early, so the relevant forms for Sindarin are really jēn-penta, en-ped-). For [n]dan- I would be cautious about assuming it has no final vowels:

[About Q. nancar- and S. dangar-:] Both probably < nana-; but in older formations there are examples of ndan-: as in Q nanwen-, return, nan-men, S damen. (PE17/166)

When no compounding is involved, -nt- first turns into -nþ- and then into -nh- > -nn- pretty regularly: Q. anta- vs S. anha-, Q. fanta- vs S. fanha-, mantinā > mannen. From ifant, ephat- and the nasal mutation tiw → i·thiw I assume that nasals vanish at morphem boundaries before fricatives f, þ χ turning -n·þ- to þ instead of usual -nh- > -nn-, but I would say it’s tenuous enough that -nn- is a good option as well.

Thinking about it more though perhaps Rovandor Rhovandor is the safer alternative, not as the proper historic development of srāban-taurē, but as a Sindarin compund in analogy to all those cases of words with final vowels. The general lenition of adjectives or the case of palan + tíriel > palan-díriel show that eventually the prevalent assumption in Sindarin becomes that every word once ended in a vowel that is now lost even in cases where it isn’t historically justified.

It makes sense to me to assume that this is how it was originally with the vowel endings.

What I still wonder is why you prefer Rovandor instead of Rhovandor. Doesn't rh usually remain at the beginning of a word and only become r in the second part of a compound? For example, we have Rhovanion, Rhúnedain and Rhosgobel instead of Rovanion, Rúnedain or Rosgobel.

That's just a spelling error, of course I mean Rhovandor with a voiceless rh.