Unicode: where is the Drachma sign?

Andreas Stötzner

I wonder why there is no officinal drachma sign present in Unicode so far. It seems really odd.

Other old weight measurement signs are there: scrupule ℈ (2108), libra ℔ (2114), ounce ℥ (2125). There is also a character for uncia semis 🝳 (1F773) meant to be a sort of ligature, and, nota bene, a character for drachma semis 🝲 (1F772) which is shown as a compound glyph. Especially in the light of the latter I find it illogical that the actual drachma is not to be found anywhere. It is a character very well attested in historic materials.

– OK, the shape of it resembles the latin ezh ʒ Ʒ (0292, 01B7). But that seems hardly to justify the omission of the dram. Ezh is a bicameral letter, it has other function and technical behavior than the drachma. The dram sign is a unicase ideographic character which is likely to demand different typographical treatment than the letters it looks so similar to, for instance its width in relation to other measurement signs or even figures.

Am I missing something important? Any thoughts?

0292 – lowercase ezh

01B7 – uppercase Ezh

F2E6 – Drachma sign, MUFI encoding (PUA!)

1F772 – half drachma

2125 – Ounce sign

1F773 – half ounce (ligature)

1F713 – cinnabar sign

André G. Isaak

Maybe I am missing something here, but what's wrong with using U+20AF?

EDIT:

Sorry, I misread. You are referring to the unit of volume not currency. The unicode standard recommends using U+0292. Yes, that would be a bicameral lowercase character, but this is no different from many other units of measure which employ latin characters as abbreviations (e.g. mm stands for millimetre, MM does not). The separate code points for other units of measure (Kelvin, Ångstrom, various CJK codepoints) are just there for compatability with pre-unicode codepages AFAIK.

Andreas Stötzner

however, it remains inconsequent. If I were to run a search for the particular dram-weight expression – I’m lost or I find myriads of alphabetical ezh’s which is of no use whatsoever. The question is about the character, not the glyphs.

Denis Moyogo Jacquerye

The dram or drachma weight unit symbol is currently unified with U+0292 ʒ.

The proposal Toward a Proposal for an Alchemy Unicode Plane from 2008 describes U+0292 as the dram symbol as well.

Using U+0292 ʒ, or possibly its uppercase U+01B7 Ʒ, is the correct approach.

Some authors might be using other characters that may look similar in some styles: ȝ, ᵹ, ꝫ, ȥ, ƺ, ʓ, ℨ or even 3.

André G. Isaak

Andreas Stötzner said:

however, it remains inconsequent. If I were to run a search for the particular dram-weight expression – I’m lost or I find myriads of alphabetical ezh’s which is of no use whatsoever. The question is about the character, not the glyphs.

I understand the point that you are making, but I see no way of solving this which does not involve a vast proliferation of code points. For example, we'd need a separate symbol for g (the gram character) and g (the alphabetic character). And I strongly suspect that if separate code points were introduced for all unit symbols that the overwhelming majority of users would continue to use the alphabetic characters anyways so nothing would really have been gained.

John Savard

André G. Isaak said:

For example, we'd need a separate symbol for g (the gram character) and g (the alphabetic character).

I don't see that at all. I would think that the character for "dram" is clearly a conventional sign, belonging to the same family as the ounce symbol, while "g" for gram is an abbreviation, so of course it is composed of letters.

Basically, this is because the dram symbol in print is usually bolder and larger than letters.

Of course, the issue may have been considered irrelevant because maybe texts written in a language in which yogh was a letter never included the dram symbol...

Igor Freiberger

The problem is that Unicode didn't use a consistent criterion along its existence.

Some characters from the Letterlike Symbols block are simply copies of one or two characters with different code points: Kelvin, Ohm, Plank, Celsius, and Fahrenheit. Under this approach, Drachma sign should have its own code point.

Denis Moyogo Jacquerye

In many cases Unicode has those as separate characters because previous encodings did and it tried to normalize them to only one character each in a few cases.

Applications that follow Unicode normalization will automatically convert Kelvin to K and Ohm to Ω. Celsius and Fahrenheit can als be normalized to °C and °F but that’s a less common normalization.

Andreas Stötzner

Igor Freiberger said:

The problem is that Unicode didn't use a consistent criterion along its existence.

Very true. And the inconsistency goes even further in this case:

🝳 (1F773) is actually nothing else than a script-style ligature of  ℥ (2125) and ß (00DF).
🝲 (1F772) is actually nothing else than a combination of ʒ (0292) and ß (00DF; the ß stands not only for the German Eszett but also for the Latin Abbreviation ſ-s for “ſemis“ – half). By Unicode’s own policy there is actually no case for neither 1F772 nor 1F773 since the expressions can be fully represented by the appropriate sequences.

Historically, I doubt that, when authors and typesetters used the dram sign (e.g. in the 16th century), they had in mind “this is the letter ezh”. – The fact remains: the character is missing.

John Savard

Andreas Stötzner said:

🝳 (1F773) is actually nothing else than a script-style ligature of  ℥ (2125) and ß (00DF).

🝲 (1F772) is actually nothing else than a combination of ʒ (0292) and ß (00DF; the ß stands not only for the German Eszett but also for the Latin Abbreviation ſ-s for “ſemis“ – half). By Unicode’s own policy there is actually no case for neither 1F772 nor 1F773 since the expressions can be fully represented by the appropriate sequences.

Again, I am confused here. This would be true, if those two symbols were used as ligatures within text. If they're used as symbols for half an ounce and half a dram instead, so that they appear in a distinctive style, then the sequences of corresponding characters won't completely represent them. That would be like saying that U+2133 doesn't belong in Unicode, since the letter M is already present at another codepoint.

Peter Constable

Andreas Stötzner said:

...

– OK, the shape of it resembles the latin ezh ʒ Ʒ (0292, 01B7). ... The dram sign is a unicase ideographic character which is likely to demand different typographical treatment than the letters it looks so similar to...

...

[italics added]

These are good reasons in favour of unifying dram with ezh. As for these reasons given for not unifying...

Ezh is a bicameral letter, it has other function and technical behavior than the drachma. The dram sign is a unicase ideographic character which is likely to demand different typographical treatment ... for instance its width in relation to other measurement signs or even figures.

There are several cases of an uppercase or lowercase character from a bicameral pair being used a symbol for some kind of unit and in which case mapping becomes inappropriate in that usage. (In fact, the Unicode Technical Committee just last week made this response to feedback regarding case mapping of Greek letters used as units.) For example, "ms" and "MS", when used as units, represent entirely different units. And just as the usage context can bring some implications for case mapping, it shouldn't be surprising if a usage context might have typographic implications, such as the need for tabular widths. 

From a Unicode perspective, there's long-standing precedent for not accepting these as sufficient reasons to dis-unify.