> At 23:33 +0200 2003-06-23, Philippe Verdy wrote: > > >What about the many symbols used to signal how clothes can be cleaned,
And Michael Everson responded: > A well-defined semantic set that I think deserves encoding. :-) If what you mean is: http://www.waschsymbole.de/en/index.html then some of those are *already* representable using currently encoded symbols: U+24B6 CIRCLED LATIN CAPITAL LETTER A = dry clean with all standard methods U+24C5 CIRCLED LATIN CAPITAL LETTER P = dry clean with perchloro-ethylene U+24BB CIRCLED LATIN CAPITAL LETTER F = dry clean with fluorine-solvent U+29BB CIRCLE WITH SUPERIMPOSED X = do not dry clean U+25B3 WHITE UP-POINTING TRIANGLE = bleaching allowed And "delicate" is the sequence <25CB, 0332>, a large circle with an underscore. And so on. But as you can see if you visit that page, there is more than one standard for such icons -- a European standard and a Canadian standard. And for all we know, there might be others as well. The Canadian standard also color-codes the icons, which was one of Philippe's criteria for where these kinds of things clearly go over the line of what is appropriate for encoding as characters. And the "sethood" of a collection of arbitrary icons is not sufficient criterion for the "characterhood". Just because a group of symbolphiles can investigate and come up with a collection of these things, and just because these things are *printed* on labels for clothing does not ipso facto make them characters, any more than the various symbols and logos related to food (and other) packaging. Look again at the icons listed above at that site. Clearly, as for many such symbologies which are supposed to communicate *WITHOUT* language, we have interesting little pictographic logics embedded in the symbols to convey meaning. For instance, a pictograph of a hand iron with one, two, or three dots inside, supposed to convey the degree of heat of the iron. Or washtub pictographs with digits in them to convey water temperature (in degrees Celsius), or with a pictograph of a hand inserted to indicate "hand wash only". Such collections of icons are, generically, part of an ongoing process of the reintroduction of pictographs and (true) ideographs into writing, to solve commercial and regulatory issues of globalization. Pictographs proliferate across Europe because "Europe" the commercial and regulatory entity is becoming so multilingual that it is utterly unwieldy to require warnings, labels, and other important captions (and even instructions) in language-specific writing. The alternative--to force everyone to use a dominant (or a few dominant) official languages--is not PC in Europe. Heck, it isn't even PC in the U.S., although it is almost official policy here. But the implication of this ongoing development needs to be *considered* by the character encoding committees -- not just be catered to, by "accident", as it were, by merely encoding as characters whatever nice little set of iconic symbols happens to attract our attention this week. There is a serious question here regarding what is plain text content and what is this "other stuff" -- an ongoing evolution of iconic and pictographic symbols that are intentionally, by design, disanchored from any particular language, and are instead intended to convey *concepts* directly. I think we are at serious risk of "getting it wrong" if we just keep encoding sets of icons and pictographs as characters without clear evidence of their use *like* characters embedded in what is otherwise clearly plain text context. What is obvious is that all this stuff is in rapid ferment right now. Hundreds of agencies and organizations make these things up for all kinds of purposes, and which ones catch on and last and get used with text remains to be seen, in many instances. Further, looking a little more longterm, it is unclear where this stuff is headed over the next century. Will such symbols remain disjunct and be very product- or situation-specific, while turning over rapidly as technology or products or regulatory environments change? Will such symbols evolve towards a global, standardized, iconography-without-words, existing as a kind of universal visual sign language for the communication-impaired who don't share a common language? Will major existing writing systems evolve to incorporate more and more such symbols (either individually or globally) in a kind of mass reintroduction of pictographic and ideographic principles into writing systems? I don't know the answers to these questions. But I don't think that we should, as character encoding specialists, behave as if they don't matter for what we do. I don't think it is appropriate to just take a "Gee whiz! Let's encode that cute set of symbols!" approach to every list of these things that comes along, without considering more carefully what the Unicode Standard is for and how it is going to have to interact with these kinds of symbols in the future. --Ken
