> >The writing of final _r_ follows the pronunciation of
> >Tolkien's English: Rómen, where the _r_ is pronounced,
> >where it isn't, óre. That's a problem for those who
> >pronounce every _r_: How should they write?
>They should write according to Tolkien's usage. Remember
>that this mode is largely orthographic rather than
>phonemic. The orthography is set by Tolkien's example.
>Whether or not the person transcribing a text into tengwar
>pronounces final or preconsonantal 'r' or treats it as
>silent is immaterial; he is not transcribing according to
>his spoken dialect, but rather according to what Tolkien's
I think the problem is that we have two ways of understanding
Tolkien's usage. We can understand it either phonetically: r-letters
depend on pronunciation, or 'typographically': r-letters depend on the
What arguments do we have to favour one of the two? Your argument that
the observed mode is "largely orthographic" doesn't help much. Of
course, the spelling is orientated by the traditional orthography, but
it's well known that some features follow pronunciation nonetheless,
e.g. the choice between súle and anto for <th> or between quesse and
silme nuquerna for <c>. The choice between rómen and óre for <r> can
also be explained as a reflection of two different pronunciations.
This allows a consistent explanation of the distribution of óre and
rómen, while the 'typographical' explanation is more arbitrary (why
should the r-letter depend on the following letter?).
There's one difference between the letters <th, c> and <r>: All major
dialects of English have two different pronunciations of <th, c>,
while only some have two different pronunciations of <r>. Tolkien had
two different pronunciations of <r>, but I daresay that most Tolkien
researchers haven't (assuming that most come from the U.S.), so the
neglect of that feature might be an unconscious impact of the
researchers' own dialects of English.
>Simple sibilants, such as that represented by IPA symbol
>[s], cannot be nasalised. The 'c' in the ultimae of
>'convince', 'difference' = [s]. A diacritical mark that has
>the purpose of nasalising a consonant should not be placed
>over a letter representing a consonant tat cannot be
No obstruents can be nasalized unless you consider e.g. an [m] to be a
nasalized [b]; what you're referring to is prenasalization.
Prenasalisation of fricatives requires a more complex articulatory
movement than prenasalization of stops. However, there are names like
_Nsama, Nseleni, Nshawu, Nsuze_ that suggest prenasalized fricatives,
so I wouldn't deny their existence.
In this question it's me who goes for the 'typographical' explanation:
I woulnd't use the bar over silme (nuquerna) because it isn't a
j. 'mach' wust