Chris rote:

>The trouble is that a large proportion of schwas - maybe a majority -
>correspond to full vowels in one or more derivatives. So, while it might be
>reasonable to respell "button" as *"butten" (unless you think that
>*"buttonian" and "buttonic" are latent derivatives), the penultimate vowels
>in "phonology" and "generative" should have some recognisable connection
>with the corresponding vowels in "phonological" and "generation".

But where do you draw the line?  Take adjectives formed from surnames, such
as Newtonian, Spencerian, Jespersenian, etc.  Presumably the base names do
not come ready-equipped with just the right underlying vowel phoneme to
enable the derivation of the correct adjective ([o], [e] etc.)?  We know
such proper names as Hamilton long before we ever come across the concept of
a Hamiltonian.  The only way to explain this is surely that the base names
have a final schwa, and that when we first come to use the derived adj we
form a new irregular derivative, the phonemic form of which we base on the
spelling, according to well-defined conventions.  I can tell you for certain
that before Jespersenian, i had _never_ come across an adj derived from a
surname ending in -en.  Yet when i first read the word Jespersenian i
immediately constructed the new morpheme, with a /ij/ phoneme.  In fact,
there are even cases where different people form different morphemes.  The
adj from the mathematician Abel was pronounced /@'bijlij@n/ by some
lecturers, whilst others insisted on /A:[log in to unmask]  The process of
derivation was clearly _far_ from automatic!

[This might be a good topic for a linguistics graduate to investigate, if
they hadnt all been so brainwashed by the Chomskyan paradigm that they would
reject the idea of underlying schwas out of hand.]

>I think
>it was Mike who pointed out here some time ago that some orthographic
>systems are necessarily difficult. English and French are among them, which
>isn't to say that neither couldn't in theory be improved.

Surely by saying they are difficult you are implying that the spellings
grate against the natural phonemicization by native speakers.  Otherwise
where does the difficulty come from?  And why is this difficulty necessary?
If an orthography is to be an efficient tool for writing down the native
tongue, surely such difficulties can only be to satisfy the etymologists and
classicists, who like to see the ancient forms reflected in every modern

>I agree that the aim of any spelling reform should be to aid L1 learners.
>To that end, one should probably start by discover what features of the
>current orthography are actually difficult for learners. My guess is that
>spelling-reform betes noirs like -IGHT and PH are much less of a problem
>than usually assumed and that mistakes like *ORTHOGRIPHY are a larger one,
>even though there's no good case for losing the A in -GRAPHY.

I think this shows once again that you havent grasped the full meaning of
what i am saying.  Surely if misspelling of schwas such as in 'orthogriphy'
is commonplace, this is strong evidence in favor of an underlying schwa in
the phonemic representation of such words!  If so, then there is surely an
overwhelming case for losing the 'a' and replacing it with a dedicated schwa
symbol, in order to a) conform to the native phonemicization and b)
eliminate a source of misspellings and writing difficulties!

Kordiale, James Chandler
[log in to unmask] - IALs index

"...Reinecke observes that transformational generative grammar ... has been
'immensely influential in the field of Creole studies as in all linguistic
theory'.  What he does not say is that it has been for many Creolists an
almost constant love-hate relationship, resulting in eventual divorce." -
Peter Muhlhausler, Pidgin and Creole Linguistics

Get your FREE download of MSN Explorer at