Re: Greek Characters Duplicated as Latin (was: Sanskrit nasalized L)

From: Philippe Verdy <>
Date: Sun, 14 Aug 2011 23:57:50 +0200

2011/8/14 Richard Wordingham <>:
> On Sat, 6 Aug 2011 17:25:11 -0700
> tulasi <> wrote:
>>    - Why did Unicode Inc copies some letters/symbols from Greek-script
>>    irresponsibly and renamed as Latin-script?
>>    - Why din't it (Unicode Inc) use same Greek letters/symbols?
> U+00B5 MICRO SIGN is an ISO-8859-1 character, and was therefore
> included as U+00B5.  It normally precedes a Latin-script letter, and
> therefore it actually makes sense to treat it as a Latin-script
> character, and possibly give it a different shape in these contexts to
> the shape of the Greek letter in Greek text.
> [...]
> U+0216 OHM SIGN is similar to U+00B5 MICRO SIGN, except that it is used
> on its own.  Whether it should be merged with U+03A9 GREEK CAPITAL
> LETTER OMEGA is debatable, but that is what has been done.

My opinion is that you could retain the same argument for the OHM SIGN
and the MICRO SIGN, because they are *both* used along with Latin
letters to create SI unit abbreviations: the units in kiloohms and
megaohms are not exceptional and most often abbreviated, and in fact
much more frequently used than units in ohms alone (especially in
electronic). So it's natural that Latin fonts will want to line up the
style of the ohm unit symbol with the style of by Latin letters ('k'
and 'M') used for noting the standard multiples, and adopt a
typographic style that will be more coherent with Latin typography
than Greek typography (for example, sans-serif styles are much less
preferable to serif styles in Greek than in Latin, and variable
weights of strokes and rounded strokes are much more frequent and
prefered in Greek typography; additionally Greek typography is often
nearer from the handwritten cursive style, and so exposes a small
slant on glyphs, even in non-italic styles, just like what happens in
the cursive style of Latin).

If users want fine typography that will work best for Latin and Greek,
including in technical publications, they will want also a contrast
between the ohm sign and the Greek capital omega letter (there's also
a perception of the separation of scripts between technical SI units,
perceived all as Latin, and native Greek texts, even in monolingual
texts in Greek).

So I don't see the "dual" encoding of the Ohm sign and the Greek
capital Omega as a problem, even if there are compatibility mappings
between them. And I'm not even speaking here about the usage in
mathematic formulas, or in IPA which is not relevant here, but has
similar consideration about their perceived separation of scripts for
distinct usages (they actually don't speak the same language:
technical notations are more international and language-neutral,
whereas there remains a strong tradition for typesetting humane
languages differently).

This also reminds me of the differences that occur between languages
about how to best typeset many Latin letters or diacritics according
to the language (see the case of German capital letters, which should
avoid the heavy contrast of variable stroke weights, because these
capitals are much more frequent in German; see also the case of the
acute accent or the hacek which should look very differently in
Czech). But these cases can be handled by language-specific glyph
substitutions in OpenType fonts. This is not the case of technical
symbols like the ohm sign and the micro sign, for which there is no
relevant language code selectable in source documents: for this, all
we can do is to use separate code points

(But you could argue that an OpenType feature could also allow
selecting the technical symbols in technical documents, if there's
support in word processors for selecting such typographic feature;
this is probably overkill because technical documents will anyway
select separate fonts and styles for technical notations; in which
case the separation of codepoints is unnecessary, except in plain-text
documents if one wants to preserve the semantic distinction,
translated in glyphic distinctions when the document is rendered).

Anyway, all fonts that only provide a glyph for a single version of
these letters should probably also map the other one, even if they
don't display the expected distinction (and may be this is not needed
for the specific font design), but text renderers will also be able to
provide fallbacks from one style to the other if one codepoint in the
pair is not mapped (possibly by using another fallback font). But
given that most Greek fonts also contain Latin letters, they should
really map both characters (the Greek letter, designed specifically
for the Greek alphabetic usage in humane Greek texts, and the
technical symbol, possibly with a separate style matching with the
Latin letters also mapped the same font).

-- Philippe.
Received on Sun Aug 14 2011 - 16:59:29 CDT

This archive was generated by hypermail 2.2.0 : Sun Aug 14 2011 - 16:59:29 CDT