Re: My new project - comments appreciated
From: | Joe <joe@...> |
Date: | Wednesday, July 14, 2004, 9:50 |
Philip Newton wrote:
>On Wed, 14 Jul 2004 09:22:32 +0100, Joe <joe@...> wrote:
>
>
>>Incidentally, about the phonology:
>>
>>If a sound regularly changes in one language to another, the language in
>>which the sound is kept distinct gets the orthography, and the
>>'official' pronunciation. So, as an example - 'th'. Dutch, 'd', Scots
>>'th', Frisian 'd'. While it's a tie, Dutch(and Frisian) confuses the
>>sound with plain 'd', so, while that may be a dialectal feature of
>>Dutch, it will be written 'th'. And soforth.
>>
>>
>
>Sounds a bit like Interlingua again :)
>
>IIRC their rules about forms of words include something along the
>lines of "If language X regularly changes letter A into letter B [e.g.
>representing regular sound change], then this language-specific change
>shall not be used to infer a spelling; rather, the underlying spelling
>shall be used". That is, spelling idiosyncrasies belonging to a given
>language are not taken into account.
>
>
Indeed. This obviously only occurs for straight A->B spellings, though.
And I'm not sure about removing innovations to quite the extent that
Interlingua does. If we have a phonology of A, B, C, D, E, and languge Z
has E and C collapse(into C), I'm going to make sure that E, not C, is
the standard spelling. Equally, if a sound is eliminated in one
language, it will always be kept, just to keep maximum differentiation.
On the other hand, if E->F, then things get more difficult.
>Or were you going purely by majority, rather than what the
>"underlying" sound is? For example, if sound P split into Q and R in
>two of the four languages, would you prefer the "Q" and "R" spellings
>for those words in which the sound occurs, on the basis of
>"distinctiveness rules", or the "P" spelling, on the basis of the
>splitting being language idiosyncrasy?
>
>
There are various possibilities. The first is that I may just invent a
new letter, distinguising X->R from P->R(for instance, the example, that
I'm now not sure of, of 'ołd', instead of 'old'). On the other hand, the
alphabet gets a little unwieldy after a while. If that's the case, I'll
probably a)Compare with Low Saxon b)Take the maximum splitting. Unless
it merges a sound, in which case, I won't.
>Interlingua also uses cognates which keep a given spelling do
>determine whether a given morpheme would be recognisable - for
>example, English has "father" but also words such as "patriarchal",
>indicating that a morpheme along the lines of "patr-" would be
>recognise and needn't contain initial |f|.
>
>
I'll probably avoid that, unless it's absolutely neccesary.
And now, the first sententce - 'I ken him not' - 'I don't know him'. I
decided (arbitrarily) to keep both 'I' and 'Ik', the latter before
vowels, the former before consonants.