Re: Digraphic letters (was: Dutch "ij")

On Friday, July 19, 2002, at 07:24 , Morgan Palaeo Associates wrote:

> Ray Brown wrote, quoting myself:
>
>>> I think it's worth remembering that the English word "letter" can only
>>> be literally translated into languages where one letter = one character.
>>
>> So letter is a synonym for grapheme or character?
>
> No, because of (1) capitalisation, whereby one letter has two
> characters and (2) punctuation, whereby non-letters have characters
> too.
Yes, of course.  I was feeling a bit tired when typed the above & not
thinking
it through.   And some letters with variant lower case forms. e.g. {a} and
{g},
have at least three different characters.  As you rightly say, punctuation
marks
are characters, as are mathematical symbols and many other things such as:
@, &, $ etc.

I'm not sure grapheme is always used the same way by all.  I seem to recall
some argument once whether lower case {i} consists of two graphemes or
not.  In Turkish I guess it must, since undotted-i occurs as a separate
letter -
but in English?

Do we count " (double quote) as one or two characters?  It is certainly two
graphemes.

>  *However* the English word "letter" does carry a connotation that
> is incompatible with the idea of being digraphic.
That I am less convinced about; especially after consulting my dictionary.
The letters of our the English version of the Roman alphabet are indeed
all monographs.   But what are we to make of, say, {æ} (a-e ligature)?  In
English it is now fairly rare and we regard as "two letters joined
together".
But when it's used in Old English we call it 'ash' and AFAIK regard it as
a
single letter.  IIRC Swedish includes it in its alphabet as a single
letter.

> Imagine that the Dutch (for example) had two seperate words, one of
> which was the regular word translated "letter" and the other of which
> denoted the thing that 'a' and 'A' are the same of, 'ij' is two of and
> ',' is not. Then, which of those two words would be a literal
> translation of the English "letter"?
Does Dutch do this?  Also does Dutch actually count {ij} as a separate
'letter'
in its alphabet so that, e.g. words beginning {ij} are listed in
dictionaries
after the words beginning {iz}?

Of the languages that actually do include digraphs as separate 'letters'
of the
alphabet, e.g. Spanish & Welsh, I am not aware of any that have two
separate words,
one denoting single character alphabet members only, and another word
denoting
each & every member of the alphabet.   Can you give me any actual examples?

> The only answer can be that it's a subjective matter and that neither
> answer is correct, which proves my point.
But does the question itself actually arise?

>> And how shall we persuade these grammarians that they have been
>> misusing the English word 'letter'?
>
> Misusing is the wrong word. I never said or implied anything about
> "misusing".
True, you did not use the word - but it seemed to me you were implying it.

>  All conlangers know that translation is not an exact
> science because words denoting complex ideas are rarely truly
> identical in two languages,
I have been well aware of that for the past half-century!  Indeed, I have
pointed this out once or twice on Conlang myself.

> and all I'm doing is pointing out that
> "letter" is an idea complex
As you now admit "letter" is a complex idea, I'm not sure what you are
trying
to say.  I was under the impression you were insisting on the definition
"one
letter" = "one character", and that, e.g. {ch} must always be two letters.

> enough for this principle to apply. I feel
> it's important not to fall victim to the illusion of objectivity.
You've lost me.  I'm not sure what point you are trying to make.

When, many years ago, I learnt that {ch}, {ll} and {ñ} were counted
separate
letters in Spanish, I wasn't aware I was falling victim to the illusion of
objectivity; I thought I was merely learning a fact about the way words
were
actually listed in Spanish.

In the meantime, until someone comes up with anything better, I shall stick
with "letter" which, as you say, is a complex idea.  Where necessary, I
shall
distinguish between: 'simple letters' (i.e. one character only), and
'composite letters' (two or more characters).

I can then say, e.g. English has only simple letters, but the Spanish
alphabet
includes three composite letters.

Even so, what do we say of the Cyrillic [bI} ?

Ray.

Replies