On September 12, 2013 Richard Wordingham wrote:
On Wed, 11 Sep 2013 12:57:34 +0200
Charlie Ruland ☘ <[email protected]> wrote:
Andreas,
linguistically speaking (i.e. following the tradition that was
started by Ferdinand de Saussure) when items are used contrastively
they must be considered different linguistic entities on what has
been called the “emic” level: phonemes, morphemes, graphemes, etc.
But this switching is not done at the character level, but at the word
level. To follow the linguistic comparison, this is code-switching. If
an utterance contains a mix of English and French, one doesn't normally
use one character for English /s/ and another for French /s/ in
broad phonetic transcription ('emic' level). One might choose different
fonts, or font effects, to mark which language was being used, but that
is a different matter. One might simply colour code the difference, but
that would be regarded as extravagant.
That’s correct, script switching is not done at the segmental, or
character, level. And I surely didn’t meant to say that a new set of
characters would need to be encoded. I for one would have been happy
with a unification of all the Brahmi based scripts of India, as had been
the attitude of the Indian government at one time; nor do I object to
the unification of, say, Japanese 禅 and simplified Chinese 禅 (U+7985;
I have used the locale tags ja and zh-Hans, hopefully you can see the
difference — cf. U+5358 単 and U+5355 单, which are not unified).
Dealing with contrasts in writing system doesn’t necessarily imply
encoding different characters, but finding a way to show these contrasts
on whichever level is appropriate. (Incidentally this is to some extent
paralleled in the realm of phonology, where the notion of segmental
phonemes is not exactly helpful when it comes to stress, intonation etc.)
As to English /s/ and French /s/: No matter if their symbolic
representations look the same or not one must always bear in mind that
the two /s/’s have quite different systemic values and thus refer to
clearly different things. I am sure you are aware of this.
At a practical level, when writing about a computer program, I find it
useful to use a proportional font for normal text and a monospaced
text for identifiers within the program, but I wouldn't dream of
claiming that the identifiers were really written in a different script
to the plain English in the discussion.
For speakers of languages other than English the inventory of characters
used may be quite different: for example, ⟨ß⟩ ⟨ơ⟩ are used in “normal”
text in German and Vietnamese respectively, whereas ⟨@⟩ ⟨#⟩ and in
Vietnamese also ⟨f⟩ ⟨j⟩ ⟨w⟩ ⟨z⟩ are usually confined to the sphere of
computers. But this is a matter of alphabet rather than script.
As /gebrochene Schrift/ and /Antiqua/ were habitually used
contrastively there is no doubt that they are different scripts in
that tradition, although they may be the same script in another
tradition. This is very much like [ɛ] and [æ] being different
linguistic entities (phonemes) in English, but not in German.
It's not so very different to using different fonts to indicate
different languages when the text is actually 'translated' for the
readers convenience, e.g. the use of an Arabic-looking font to
represent Klatchian in the Discworld novel 'Jingo'. I wouldn't suggest
that we have a different script for the Klatchian; rather, the novel is
not in plain text in these sections.
A solution that is more general that specifying fonts which may or may
not be installed on a system is to use different locale tags for
“blackletter Anne” from Berlin and “antiqua Anne” from Paris: Anne
(which I have tagged de-Latf) vs. Anne (de-Latn). If you can’t see a
difference here it is because your e-mail viewing software doesn’t use
different fonts for display. IMO the W3C should be prompted to stipulate
that distinguishable letters be used for spans of text that are
script-tagged Latn, Latf and Latg. Though this is beyond the scope of
Unicode each of the two standards depends on the other. /Tout se tient./
One final remark: Thinking about it I have the impression that the
blackletter vs. antiqua distinction once made in German very much
resembles that made between Hiragana and Katakana in Japanese. In both
cases the underlying systems of the corresponding scripts are
essentially the same; yet it seems impossible to read the other script
without further instruction and exercise; and in both languages one
script is used primarily for inherited, and the other for foreign words.
Charlie
Richard.
