On 7/9/06, Peter T. Daniels <grammatim@...> wrote:

> How out of touch are you? Have you never looked at the list of Penguin
> Classics? at what Project Gutenberg has _already_ digitized?

Actually, I have looked at what Project Gutenberg has already
digitized, having scanned or processed about a hundred books for
Project Gutenberg. Project Gutenberg has transcribed roughly 18,000
books, dominated by works of the late 19th and early 20th century.
Compartively, UMI has roughly 100,000 books in their Early English
Books Online collection, which stops at 1700, very few of which are
accessable in any form without a major library at hand, and those only
in microfilmed forms. That count doesn't include the 18th century, an
extremely prolific era in English publishing.

Penguin Classics has 53 volume from 16th to 18th century English
literature. I think that's cut off, and for those volumes, I'm
possibly mistaken, but I do believe that they would be reprinted in
any new orthography that looked to be dominant within a few years.

> If the process would be automatic, then English
> spelling-to-pronunciation is _already_ fully predictable and no reform
> is needed.

That's a non-sequitor. The issue is not that well-trained people and
computers can't predict how to pronounce things usually; it's that
children have a hard time learning the system and even well-trained
people and computers sometimes make mistakes.

> We _have_ a system that reflects the complexities of English vowels --

I am under the impression that English writing simplifies the English
vowel system to fit the five vowel letters of the Latin alphabet, no
matter what dialect or century of the English language you're looking
at. Were, for example, bead and head, ever pronounced the same, or is
there any systematic rule that could seperate the two?

> If one of those dialects -- with all its own peculiar phonemic
> mergers -- was chosen to be the standard (Earl referred to the
> cot/caught merger), then all the others would be short a representation,
> and/or would have to deal with distinctions that meant nothing to them.

Again, there's no reason to pick one dialect; in language
standardization, a regularized form of the language corresponding to
no one dialect is often used. And I think there would still be a great
advantage to a more phonemic spelling, even if it meant remembering
distinctions that others make or merging distinctions that you do
make.