RE: [CSS3 Text] Tibetan Emphasis marks

Here are the basic rules for the Tibetan emphasis mark.

1. one mark per syllable;
2. mark is positioned at the center of each syllable;
3. mark fits close to the text;
4. all marks in a line may or may not align to each other vertically.

As such one might say that the emphasis could be applied as a CSS markup and the User Agent would implement the rules according to the manner in which they would choose to do this.

I would submit that this feature, if accepted for consideration should not be normative. 
1. The use of the Tibetan Emphasis marks are very limited in the scope in which they are used. Implementing support for this would most likely not be a high priority for any company. 
2. The tight coupling to the cluster handling of Tibetan this would require special low level code for centering the mark on the syllable. Not trivial code and not something the UA would necessarily be able to handle unless they also implement the Tibetan shaping engine.

It would be easy enough to add support for this to CSS3 'text-emphasis' property by adding something the following values (not intuitive names and prone to typographic errors):

nyizla - uses the Tibetan mark U+0F35 as the emphasis mark
sgorrtags - uses the Tibetan mark U+0F37 as the emphasis mark

The use of these marks for other than Tibetan text would need to be undefined to preclude people doing "creative" things with this property.

Doing the above would make it possible to style Tibetan emphasis marks. However, it does not guarantee that support for this property would be added in UAs.


-----Original Message-----
From: Christopher Fynn [] 
Sent: Monday, July 09, 2007 12:37 AM
Cc: Paul Nelson (ATC);
Subject: Re: [CSS3 Text] Tibetan Emphasis marks

Paul Nelson (ATC) wrote:

> I have copies of the images (attached)

> I do not see that the Tibetan Emphasis marks truly fit as an extra layer above/below text as do Chinese or Japanese for the following reasons.

Hi Paul - The semantic of these marks is emphasis - which is sometimes displayed in other ways (e.g using a larger point size or different colored font) - so IMO markup is a better representation than using explicit characters.

The existing characters are also problematic in that they have to be applied to a single base (letter or ligature) when they should really be applied to the whole run of characters in a Tibetan syllable.

Unicode are of course never going to change the combining class and other properties of these characters which they have defined. Personally I think the existing characters should be deprecated in favour of emphasis markup

With the existing characters people will of course be tempted to choose different bases within the same 'syllabe'  to achieve an approximation of the kind of "typographic interaction" you mention below.

> 1. There should be typographic interaction between the marks and the syllable cluster on which they are positioned.

While such typographic interaction is ideal it is not essential to the semantic.
Modern Tibetan language publications from China do in fact place these marks at a single level (often substituting the CJK sesame dots for the real Tibetan glyph shapes.)

2. The purpose of proposing CSS (as pointed out by Chris' mail) is to have a more immediate solution instead of working with getting this functionality into OS level.

I think you might be reading more into what I wrote than I intended. While I did point out the problems with implementing these in smart fonts on the basis of the existing Unicode characters - the main thing I wanted to point out was that these marks are simply an <em>emphasis</em> -  and best encoded that way. CSS comes in when you want to specify how you want that emphasis displayed.

The ideal placement of the Tibetan emphasis marks is a rendering / application problem which I wouldn't expect to see solved any time soon.
IMO lack of precise typographic placement of these marks is much less of an issue in most situations than conveying the semantic correctly.

> While I can sympathize with Chris wanting to get Emphasis marks handled correctly (e.g. not as part of normal text stream), I do recommend that these *not* be introduced into CSS as this is really an unwanted hack and not the right long term solution.

If you have any suggestions as to how Tibetan emphasis could be better represented in HTML , XML,  or in plain text,  I'm interested to hear them.

best regards

- Chris

> Paul

Received on Monday, 9 July 2007 07:03:28 UTC