In an insightful message posted on July 5th 2003, n° 459,
Pavel Iosad argues for the existence of two kinds of phonological boundaries in
Sindarin, with consequence on the theory of mutation. His theory arose from the
apparent contradiction in a word-group between mutations on the one hand, that
suggests that it was treated as a unitary word, and vowel-apocope on the other hand,
that suggests that the words were kept separate. Pavel then introduces two distinct
phonological units: the *phonological phrase* and the *phonological word*, which
allow us to remove the apparent discrepancy by considering mutation as a
*phrase-property* and apocope as a *word-property*.
Now this theory would explain neatly several points in the evolution of
Sindarin. Pavel already showed that it allows _b_ from *_mb_ to be lenited into
_mb_ such as _i[-]mbas_ "the bread" (VT44:21, 27-8) despite _mb_ > _m(m)_
medially. I would like to discuss some other points further.
THE LOSS OF FINAL VOWELS
Commenting on the loss of final vowels, Pavel examines several possibilities and
rejects internal syncope:
> Well, we don't know a lot about the pre-history of Sindarin stress, but I
> still think the Sindarin syncope patterns could hardly produce a number of
> these lost vowels. First, Sindarin, or rather Noldorin (since most words with
> Tolkienian proto-forms come from the Etymologies), syncope tends to affect first
> vowels in two adjacent syllables (as in Noldorin *_tarasta_ > _trasta_ (V:391
> s.v. TARAS)), or composition vowels in compounds. Thus I wouldn't expect too
> many cases of this vowel disappearing to trigger analogy.
I do not think this kind of syncope can be involved at all, for it seems much
earlier to me, since this Noldorin syncope is also seen in Telerin and
Ilkorin, though not in Quenya. In Tolkien's later linguistic scenario it could be
regarded as a Common Telerin change. For instance we have from the root BARÁD
_*b'randâ_ "lofty, noble, fine" > N _brand, brann_ from ON _branda_, T _branda_;
(V:351), from TURÚM: _*turumbê_ "shield" > T _trumbe_, no N descendant (V:395).
It may be noted however that there are hints of another kind of internal
syncope in long words. The most conspicuous is the alternation _Narog_ /
_Nargothrond_, and there is also _Elbereth_ < _elmbereth_ < *_elen-barathi (X:387).
However, it is not necessarily contemporaneous with apocope.
> Anyway, I put rather more faith into the "word-final" explanation. It gives
> us however, an apparent contradiction - on the one hand, the final vowel of
> the proclitic is subject to "word-final" processes, on the other hand, the
> mutation is triggered in seeming accordance with "word-internal" rules. Now what I
> am about to suggest is that mutation in Sindarin operates on the phrasal
> level, while vowel dropping is a word property.
I agree, and would like to define more clearly the units that suffer apocope.
"Word" is not satisfying (not to mention that this notion is hard to define
linguistically, however useful it be) and not sufficient. We can observe apocope:
1) at the end of stressed polysyllables - this is well-known.
2) at the end of some prefixes : for instance see the Noldorin _amben(n)_
"uphill" vs. Q _ambapenda_ (V:348, 380), the N and S _amrûn_ "sunrise, east" vs.
Q _Ambaróne_ (V:348, 384, S:437, LR Appendix E), the N _dadbenn_ "down" <
_*datapenda_ (my reconstruction) (V:354, 380).
3) at the end of first elements in compounds. Unfortunately it is difficult
to be sure that compounds are old (i.e. formed before apocope), yet for
historical reasons some must be, so _Celbin_ and _Moerbin_, Q _Kalaquendi_ and
_Moriquendi_. True, Tolkien notes that the Sindarin words suffered analogical
alterations, but he does not imply that the syncope is not normal phonetic evolution
(XI:361-2). Noldorin had also some old compounds common with Quenya: _andeith_
"long-mark" from Old Noldorin _andatektha_, cf. Q _andatehta_ (the S
_andaith_, LR Appendix E, can be suspected of being formed on a Q model) , _amlug_
"dragon", cf. Q _angulóke_ (V:349, 370), _lhimlug_ "fish-dragon, sea-serpent", cf
Q _lingwilóke_ (V:370), _dúlinn_ "nightingale", cf. Q _lómelinde_ (V:364,
369; also in Sindarin: S:430 has _dúlin_).
It is interesting to note that there are a number of exceptions to the latter:
_Eglamar_ < _*ekla-mbar_ (XI:365), _Eglarest_ < _*ekla-rista_ (XI:365),
_Tinúviel_ < _Tindumhiell_ < _*tindômiselde_, cf. Q _tindómerel_ (V:385-393), and
the Noldorin words _annabon_ "elephant", cf. Q _andamunda_ (V:372), _haðafang_
"throng-cleaver" cognate with Q _sangahyando_ (V:389), _heledir(n)_
"kingfisher", cf Q _halatir(no)_ (V:363, 386, 394), _Maglaðûr_ "Blacksword" (V:371);
also _merilin_ "nightingale", cf. Q _morilinde_, but this is said to be adapted
I suggest the following: apocope depended on how the word was analysed, and
was regular at the end of every unit that was considered as "independent" to
some degree - the "phonological word" in Pavel's proposal. Prefixes were
considered so, and the first elements of compounds too, provided that the compound
was still regarded as one. Some were not and were consequently treated as
THE DEVELOPMENT OF CLUSTERS OF NASAL + UNVOICED STOP
There is a discepancy between the pattern of nasal mutation of unvoiced stops
on the one hand (_p_ > _ph_, _t_ > _th_, _c_ > _ch_) and the development of non-initial
_mp_, _nt_, _ñk_ on the other hand (seemingly > _m(m)_, _nn_, _ng_ medially, preserved
finally). This was already noticed: see Elfling's messages n° 11543
, and n° 12501
VT42:27 has brought new facts, pointing towards an alternative - and very
Welsh-like - evolution of medial _mp_, _nt_, _ñk_ > _mf_ (bilabial _f_), _nth_, _ñch_
and later > _mh_, _nh_, _ñh_ (long voiceless nasals). It also says that for the final
_-mp_, _-nt_, _-nc_, it is more probable that the spirant was re-stopped that that the
stop was never changed, and specifies that the change did not occur in Northern
This reminds me of the _Silmarillion_ Appendix, entry _nim_. It has:
_nim_ "white" (from earlier _nimf_, _nimp_) in _Nimbrethil_, _Nimloth_,
_Nimphelos_, _niphredil_ (from _niphred "pallor")...
The key might be something like this:
1) first, and before apocope, _mp_, _nt_, _ñk_ > _mf_, _nth_, _ñch_ everywhere,
including external sandhi (i.e. the change occurs at the phonological phrase level).
2) apocope takes place.
3) the clusters _mf_, _nth_, _ñch_, unstable, evolve differently according to their
position at the phonological word level:
- finally, the spirant is re-stopped
- initially, the spirant prevails and the nasal is lost: this accounts for
the nasal mutation pattern and the development of the Noldorin compound
_impanta_ "year-full > aged, :long-lived" > _ifant_ (V:400)
- medially, the nasal prevails and variously assimilates the spirant; Tolkien
seems to have experimented with several possibilities, one > _m(m)_, _nn_, _ng_,
the other in VT42 > _mh_, _nh_, _ñh_.
In Elfling n°12501, I suggested that this might be connected to a difference
of stress. I do not think now this hypothesis is necessary, the different
context relative to word boundaries might be enough. In addition, I had to posit
initial stress at the initial of unitary words, but this is not very consistent with
other changes such as the diphthongisation of open _ô_ > _au_ in ultimate
syllables, such as in _*chaðôd_ > _*chaðaud > _Hadhod_ "dwarf" (XI:388).
THE SOFT MUTATION OF S > H
Old Noldorin shows that intervocalic _s_ slackened to _h_, and this must
have happened in the ancestor of Sindarin too. The _h_ is retained in compounds,
as shown in the names _Calenhad_ "Green-place" (_sad_ "place) in UT:425,
VT42:19-20, _Edhelharn_ "Elfstone" (_sarn_ "stone"), _Iorhael_ "Oldwise > Frodo",
_Perhael_ "Halfwise > Samwise", _Panthael_ "Fullwise" (_sael_ "wise") all from
the King's Letter, IX:128-31; the _tengwar_ versions show that _lh_, _rh_, _th_
are to be read _l+h_, _r+h_, _t+h_ and not as the usual digraphs. We also see
intervocalic _h_ in _díheno_, _gohenam_ (VT44:28) derived from _sen-_ with
On the other hand _s_ disappears altogether in _Tinúviel_ < _*tindomiselde_
(V:385, 393), and old compound but probably no longer seen as such, as the
conservation of the compositions vowel would suggest (see above).
This would again be neatly explained by the influence of phonological
boundaries: _h_ would be retained at the initial of a phonological word, and lost
To conclude, all these facts go well with Pavel Iosad's idea. A consequence
is that "external sandhi environment" and "medial environment" cannot be fully
identified when dealing with the history of Sindarin: the contexts are similar
but not identical.
Language has both strengthened imagination and been freed by it. Who shall
say whether the free adjective has created images bizarre and beautiful, or the
adjective been freed by strange and beautiful pictures in the mind ?
J. R. R. Tolkien, A Secret Vice
[Non-text portions of this message have been removed]