> Teithant Alf:
> >The sounds of 'w' and 'y' are
> >always the same; it's only a matter of interpretation
> >whether you consider them "consonatal" or "second
> >part of a diphtong".
> But they're NOT the same. As the second element of a diphthong, they are
> semivowels, not consonants. That is why the 'y' in 'boy' is written with
> a different symbol from the 'y' in 'yet' in IPA transcription.
> "semivowel: a vowellike sound occurring in consonantal positions in
> the same syllable with a true vowel, characterized by brief duration and
> rapid change of duration from one position of articulation to another."
> (Webster's New World Dictionary)
> True, the 'y' in 'yet' and the 'w' in 'war' are also semivowels, but
> they are 'on-glides', ot 'off-glides' and therefore more consonantal, if
> you will. They are not the same sounds as the 'y' in 'boy' or the 'w' of
This criteria is rather poor. You're reducing the difference to a
position difference in the syllable. It's as if I'd say that the on-
glide 'k' sound in _cat_ is not the same as the off-glide 'k' sound in
_back_. There's no doubt they're not the same, but every reasonable
person would agree on representing both sounds with one letter, because
there can be no ambiguity.
> >I would rather expect to find second parts of
> >diphtongs represented with tehtar, not with tengwar,
> >because in these modes any tengwa is "consonantal".
> According to whom? Certainly not Tolkien! His tengwar were designed for
> use with the Eldarin languages, and in attested examples of both Quenya
> and Sindarin tehta modes he used tengwar, not tehtar, as the second
> elements of diphthongs.
In appendix E Tolkien points out that in oomatehtar modes, only the
consonants are represented by tengwar. (But even if this were my own
supposition, it could still be useful for research on Tolkien's
writings.) This includes even Quenya (and any other) semivowel tengwar
which aren't used but as second parts of diphtongs: yanta and uure.
But why do attested oomatehtar modes have two different tengwar for the
Quenya uses both uure and vilya for the sound of w, both yanta and anna
(with a dieresis under it) for the sound of (consonantal) y. A possible
reason for this doubling: to have two tengwar -uure and yanta- that
don't bear the following vowel tehta as all other tengwar do, but the
preceding, in order to write diphtongs.
This means the reason lies in the tengwar-tehtar sequence. If the
normal tehtar-tengwar sequence doesn't allow the vocalic tehtar part of
a diphtong to be put upon its consonantal/semivowel tengwar part,
additional sequence-inverting tengwar are used. Following this
supposition, I'd expect those modes not to have two different signs for
semivowels which normally put onto each other the parts of diphtongs,
e.g. oomatehtar Sindarin.
Oomatehtar Sindarin (as in the King's Letter) doesn't have twice all
semivowels. It uses anna in the short diphtongs ending on -y and yanta
in the long diphtongs ending on what is transcribed as -e. It's
surprising that initial consonantal 'y' (transcribed as i) isn't
represented by anna but by yanta, the same sign as in diphtongs ending
on -e, and not on -y. (This might indicate that the difference between
the Sindarin diphtongs ae and ai lies more in their duration than in
the e-i difference.) So concerning the semivowel y, the same sign is
used in diphtongs and elsewhere.
But what about the semivowel 'w' in Sindarin? The only attested example
I know is in the Moria West Gate Inscription: vilya for postvocalic '-
w' (is this a diphtong?). Do attested Sindarin oomatehtar samples (or
samples of comparable modes) use more than one tengwa for the semivowel
w? I'd expect that's not the case, because with the normal Sindarin
tengwar-tehtar sequence the vocalic parts of diphtongs are always put
onto their consonantal part. Indeed, this supposition might turn out
I'm sorry the post's become so long.