L2/10-289
                                       

Title:   How to Pick a Representative Glyph for a New Currency Symbol

Authors: Ken Whistler, Asmus Freytag

Date:    August 4, 2010

Action:  For consideration by the UTC


The adoption of the Euro in 2002 has set two precedents.
One is the idea that a currency needs to a have a logo to get respect.
The other is that governments sanction an official design.

In case of the Euro, it was initially not understood by all stakeholders
that the most natural representation of the new symbol was not like a logo,
with all details frozen, but as a character, with glyphic variations to 
harmonize with the font used to render the character.

The latter view became the accepted one, and now there exist as many 
different glyphs for the Euro as there are distinct font designs containing
it.

That, in turn, raised the question of which of these glyphs would be the 
most appropriate to use as representative glyph in the Unicode code
charts. That question was eventually solved (see more below),
but the purpose of this document is to question whether the 
solution found for the Euro is the always the most appropriate for any new 
currency symbol, no matter the circumstances.

To be able to discuss this, a bit of background is needed first. 

The primary and most common use for any currency symbol 
is to note prices for local consumption (labels and price lists in the 
local language).
 
Therefore, in our opinion any new currency symbol is foremost intended
to harmonize with type faces native to the area using the currency. In 
consequence, because of that, we are not convinced by the statement in WG2 
document N3862 that harmonizing the design of the sample glyph of a
new currency sign with those of *other* currency symbols is of great 
practical importance. Further, the analogy to the Euro isn't as 
straightforward as it is sometimes claimed.
 
Creating a harmonized design for the dollar sign, pound and Euro sign
did make some sense, because all of them are used primarily with the Latin
script. As a happy coincidence, the code charts for these characters 
show glyphs that actually occur in their area of use - to the degree 
a Times-style font is used to display prices.
 
The same may not necessarily be true of currency symbols that are not 
primarily intended for use with the Latin script. For example
U+20AA NEW SHEKEL SIGN is intended to be used in a Hebrew script context.
Therefore, it makes sense for the sample glyph to harmonize
with the font used to display the Hebrew script in Unicode.

Fonts for depicting scripts other than Latin, Greek and Cyrillic in the
code charts are not based on Times, but are chosen independently as
native fonts that are similar in spirit to Times. They are everyday 
designs with as few idiosyncratic or "novel" elements as possible and
usually selected to show good distinction between different characters.

In the case of the proposed new Indian Rupee Sign, the symbol will be used
primarily in the domestic context of India. Therefore, it will occur
most of the time in the typographical context of one of the Indian
scripts. Even though the symbol isn't meant to be strictly specific 
to any single script, its design took Devanagari as one point of departure.

Therefore, it would appear that the sample glyph in the code chart benefits 
less from harmonizing with Times, than from being related to whatever
font style the charts use for Devanagari in the code charts. Doing it 
that way would again allow the code charts to show a glyph that might 
actually occur in its intended area of use.
 
While it seems that the Indian government is cognizant of the possibility
of the new Indian Rupee Sign also being used outside of India (say with the
Latin script), such use would always be considered a minority usage, 
and would occur alongside of the use with other, non-Indian and non-Latin
scripts (Han, Kana, etc.).

As a result, recognizing the non-Latin nature of the new Rupee symbol
by choice of the sample glyph in the code charts would be the most 
appropriate way to handle this block of cross-cultural symbols.

On the contrary, sticking to an unanalyzed precept of "it must be Times" 
would be tantamount to European parochialism in an internationally 
used standard.

Because there have been competing opinions on how to apply the 
existing precedent as to what typestyle to use, an annotation at the 
head of the Currency Symbols block could make clear that symbols 
not natively used with the Latin script are not shown in the way 
they might appear in a Times type face, but instead in the way they 
might appear in a sample type face similar to the 
one used for the dominant script(s) for the intended usage area of 
the new symbol.