68Various umlauts in Sindarin plurals
- Jun 14, 2002Greetings,
It has been observed multiple times before that the Sindarin vowel _o_
exhibits different behaviour when subject to umlaut, most prominently in
noun plurals. Subject to umlaut, _o_ variously yields _y_ or _ö_ > _e_.
It has been argued (by, for example, Helge Fauskanger in his Ardalambion
Sindarin article (http://www.uib.no/people/hnohf/sindarin.htm) or
Ryszard Derdzinski in his summary of the Sindarin grammar) that this
depends on whether the sound is in a final or non-final syllable. This
may be true for certain stages of Tolkien's conception, but in this opus
I will argue that it is not so in the Sindarin we meet in 'The Lord of
My proposition is that at some point Tolkien's conception was, for a
time perhaps, envisaged, in the following way: _y_ is the umlauted form
of the _o_ stemming from earlier _u_, while _e_ (< _ö_) is the product
of the 'old' o-vowel.
One obvious reason is phonetic. Helge (and Ryszard) argue that _o_
becomes _ö_ in a non-final syllable, i.e. that the degree of
assimilation of _o_ to _i_ is there for reason of distance 'less' than
that in a final syllable. While that has a point, it may be argued that
the assimilation of _o_ to _ö_ is not much 'less' than of _o_ to _y_. We
can arrange the four vowels in question (_o_, _u_, _ö_, _y_) in the
following tidy way:
_i_-umlaut is essentially the movement of the vowel to the left on this
chart. It can be easily seen that _ö_ is the most natural product of
_i_-umlauted _o_. The umlaut of _o_ to _y_ may be seen as assimilation
by vowel height.
Another argument is presented by the behaviour of _u_ when subject to
_i_-umlaut (_tulus_ 'poplar' vs _tylys_ 'poplars', V: 395 s.v.TYUL).
This is probably a pointer in the direction of the supposition that back
vowels are subject to _i_-umlaut uniformly in final and non-final
syllables. On the other hand, _u_-evidence isn't really conclusive, as
_u_ is a high vowel, and so the umlauted version can't go any higher, or
closer to _i_, unlike _o_.
In LR, there is, seemingly, evidence for both interpretations. On one
hand, ther are forms like _Ered Mithrin_ (cf. ÓROT- in V:379), but they
are in apparent contrast with _yrch_ (form ÓROK-, ibid.). We might
however observe that all instances save two of _o_ > _y_, where _o_ <
_*o_, occur in monosyllables (_yrch_, _tyll_ (TOL2), _by^r_ BOR). It
might be surmised that in monosyllables _o_ indeed moved into _y_, and
the greater degree of assimilation may be caused, say, by stress.
There is a later form _eryd_ in WJ:6, as well as _gelydh_ (from _golodh_
< _*ñgolodo_). These date mostly from post-LR time, cf. for instance the
plural of _golodh_ in the Etymologies - _goeloeidh_, _geleidh_. It
appears that in the later conception, the rule 'final-syllable _o_ >
_y_' was indeed valid. This explanation is easier than the one proposed
by Helge, who tries to invent some kind of 'status constructus' to
explain Tolkien's persistence in not changing the LR form _ered_.
Other instances of the _o_ > _y_ shift in final syllables do not suggest
that the theory is invalid. They seem to appear in derivational
suffixes, the precise proto-form of which is actually unknown. These
_amon_ > _emyn_ (AM-)
_annon_ > _ennyn_ (AD-)
_ithron_ > _ithryn_, apparently from _*ist-rVn_ (IS-)
It is not inconceivable that these contain a prehistoric suffix _*-un_ >
May it also be noted that in the Etymologies Noldorin words with _o_'s
from _o_ generally have plurals with _ö_/_e_ or _öi_/_ei_, as for
example _gwador_, pl. _gwedeir_ (TOR-), _thoron_, pl. _therein_
(THORON-). Words with _u_, on the other hand, would, I suggest, have
'diphthong-stage' plurals with _ui_. This is not at all impossible, and
anyway was there in the twenties. Confer PE13:122 ('u(.... mutates to
*ui*' (u( is an u-breve)) We have however counter-examples like _dölio_,
_delio_ from DUL- and conversely _peringiul_ (obviosuly _peringuil_)
from ÑGOL-, but that siply shows that Tolkien was vacillating in this
matter, and so we do not need to invent explanations and simply must
accept that this is the usual Tolkienian inconsistency or
indecisiveness, since these forms are all, using Bill's good term,
'workshop-Sindarin'. May it then be noted that _amon_ has an early
plural form _emuin_, which points to a possible -u- in the suffix. One
might question why I posit an -u- in the suffix of _annon_ if we have
its apparent Q(u)enya cognate _andon_ from _*andondo_ with _o_. But
there is evidence for an *_u_ > _o_ shift in Quenya, as in _órku_ > Q.
It is interesting to note that _muindyr_as the pl. of _muindor_ (TOR) is
described as analogical. Since this is a compound, the failure of
_ui_-shift in the plural is not very likely to be due to analogy, and
probably this is intended to mean that _y_ is analogical to the _y_ in
_*tyr_ 'brothers', which appears there since it's a monosyllable, but
normally the plural of _muindor_ would be _*muinder_.
And of course there is ample evidence for an _u_ > _o_ shift in Sindarin
(or should we say Noldorin?).
So, to sum it up:
It is, in my opinion, not inconceivable that in Sindarin, at least at
one stage of its development (around the time of the completion and 2nd
edition of LR), the _i_-umlaut causes the shifting of _o_ to _y_ in a)
monosyllables and b) cases where this _o_ comes from earlier _u_.
Otherwise, _o_ shifts to _ö_ > _e_ in all cases, even in final
syllables. There is no need of 'updating' Noldorin words like _doron_
pl. _deren_, as Helge suggests (and why should we do any updating
And of course I apologise profusely if this has been discussed before.
Pavel Iosad pavel_iosad@...
'I am a philologist, and thus a misunderstood man'
--JRR Tolkien, _The Notion Club Papers_
- Next post in topic >>