Talk:Writing system

From Wikipedia, the free encyclopedia

This article is part of the "Writing systems" set of articles nominated for Version 0.7. Discuss this nomination, or see the set nominations page for more details.

	This article falls within the scope of WikiProject Writing systems, a WikiProject interested in improving the encyclopaedic coverage and content of articles relating to writing systems on Wikipedia. If you would like to help out, you are welcome to drop by the project page and/or leave a query at the project’s talk page.
B	This article has been rated as B-Class on the Project’s quality scale.
Top	This article has been rated as top-importance on the Project’s importance scale.

Contents 1 History of writing systems 1.1 Cuneiform not nessarily oldest 2 Types of writing systems 3 Examples of writing systems 3.1 Runic systems 4 Terminology 4.1 Orthography 5 Syllabaries vs. abugidas 6 Grapheme vs Character 7 Map 8 Graphemics 9 Proto-Elamite and OES 9.1 The Proto-Elamite Script 9.2 The Old European script 10 Question on external link 11 SignWriting 12 New article to merge? 13 Invention of writing - Rongorongo 14 General Properties

[edit] History of writing systems

[edit] Cuneiform not nessarily oldest

While it was long belived that Cuneiform antedated Egyptian Heiroglyphics, relativly recent (late 90's) finds have pushed back the date of Heiroglyphs. The BBC has an article entitled Were Egyptians the first scribes? about the discoveries. The 'History of writing systems' section should be changed, but my prose is often terrible (and my spelling even worse). -- Levi Aho 2004-12-21 13:57:52 (UTC)

[edit] Types of writing systems

Hiragana is somewhere between being a syllabary and being an abugida, isn't it? For instance, to write my nickname ("Juuitchan") in hiragana, using the hiragana set as a syllabary will not work: you must use it as an abugida. --User:Juuitchan

I suppose so. The main difference, as I see it, is that the characters of syllabary are not similar if the sounds they represent are similar, while the characters of an abugida are similar if the sounds they represent are similar. In this sense, hiragana has components of both - "ka ki ku ke ko" do not resemble one another, but "ka ga" do resemble one another, as do "kya kyu kyo". Also, some characters (especially small tsu and small ya yu yo) don't form an entire syllable but help to make up a syllable. -- pne 09:27, 20 Apr 2004 (UTC)

---

I really believe that the "classification" of featural writing systems is bogus. It is not a structural description of the elements of the writing system; it is a description of how the glyph shapes have a particular kind of internal relationship. Hangul is an alphabet, where each unit refers to a sound. Evertype 20:03, 2004 May 31 (UTC)

Agreed. I think "featural" is a valid label, but this is orthogonal to the classification as alphabet/syllabary/etc. I'd consider Hangul an alphabet, too. -- pne 08:47, 1 Jun 2004 (UTC)

I also agree, and as such I think I'm going to remove the classification. thefamouseccles 02:49, 8 Nov 2005 (UTC)

One problem: you now claim that Hangul is partially logographic. English is more logographic than Hangul! (Pair, pear, pare are logographic distinctions; Korean doesn't do this.)

This discussion from a year and a half ago was debating a very different presentation, which did misrepresent things. Yes, Hangul is an alphabet, but there are different types of alphabet. Hangul is as distinct from Latin as Latin is from Hebrew or Nagari, and deserves recognition as such. It, along with Tengwar and a few other fictional scripts, and maybe SignWriting (if SignWriting is phonemic), form a fourth structural type of alphabet. And of course it's structural: there are elements for aspiration, place, manner, vowel harmony, etc. The fact that is may not be learned that way doesn't make it a regular alphabet any more than the fact that Amharic isn't learned as an abugida makes it a syllabary with "a particular kind of internal relationship". I'll think about restoring a featural category, though maybe worded a bit differently. kwami 05:58, 8 November 2005 (UTC)

If not "logographic", then I'm not sure what the term would be for the stylised depiction of the actual shape of the vocal tract during articulation. Glossostomatographic, maybe. Nonetheless, I still don't think the fact that Hangul's alphabetic system is featural in many of its principles justifies separating it from the "Alphabets" category. (As the page stands now, the classification is fourfold, [logographic-syllabic-alphabetic-featural], which I think is overemphasising the difference between Hangul and other alphabetic scripts.) I would argue that just because the alphabet was designed to indicate features doesn't mean it can necessarily be used that way in an open-ended fashion; the top half of ㄹ cannot be used as a separate symbol, any more than we can use the leg of p as a separate symbol. As well, the basic symbol ㄴ (for instance) doesn't just represent the feature [coronal], but the constellation of features that make up the phoneme [n]. Finally, one cannot just learn the symbols for each feature and be ready to write Korean straightaway; the realisations of each feature are often different for different points of articulation (compare the pair ㅁ ㅂ m/p with the pair ㄴ ㄷ n/t). I'm not arguing with the fact that Hangul is featural - just with the fact that "featural" is being presented on the page as being separate from, not a special case of, the true alphabet. thefamouseccles 02:46, 15 Nov 2005 (UTC)

Yeah, I've struggled with some of these very issues. I think Amharic is a close parallel in a lot of ways. It's learned and used as a syllabary. It's not an open-ended system. You can't use the consonant or vowel pieces on their own as separate symbols. And you can't just learn the symbols for consonants and vowels and be ready to use the system; the symbols for the vowels are different for different consonants - all the same objections you've raised for Hangul being separated from the alphabets. But despite this, Amharic is separated from the syllabaries as the prototypical abugida, an alphabet.

At one time, we had Hangul listed as a fourth type of alphabet (Greek-type, abjad, abugida, Hangul). Maybe we could return to that?

As for what word to use for Hangul's "pictographic" origins, I don't know. "Logographic" certainly isn't it. Anyway, it may be best to leave that out - the idea that Hangul represents the shapes of the vocal tract could well be a later rationalization or just a mnemonic. Ledyard believes the basic consonant shapes come from Phagspa, and this has received good press; and the vowels always were thought to be abstract. kwami 09:41, 15 November 2005 (UTC)

Hangul is clearly not logographic since the shapes of the radicals do not bear meaning, they are phonemes.

[edit] Examples of writing systems

[edit] Runic systems

How come "Runic Futhark" and "Anglo-Saxon Futhorc" both gets mentioned? The Old English variant is a bit different, I agree, but so is Slovene and Hungarian and German, you don't see them listed with special entries?? --Gabbe 00:07 Jan 10, 2003 (UTC)

I belive the Anglo-Saxon runes are are considered disctinct from the Germantic runes due to the addition of a number of letters (seven in all, if I remember correctly). Wether this justifies listing as a seperate writing system is debateable, but the Anglo-Saxon runes are a not simply specific usage of Germantic runes. -- Levi Aho 2004-12-21 13:32:27 (UTC)

[edit] Terminology

[edit] Orthography

What is the meaningful distinction between this article and "orthography"? Most of the latter article could be transplanted here. --Ryguasu 00:15 27 Jun 2003 (UTC)

I always understood "orthography" to refer to systems of spelling. So French and English use the Latin alphabet with different orthographies. Joe Cetina 03:54, 5 Mar 2004 (UTC)

---

The new definition of orthography given in the text is wrong. Cbdorsett 20:40, 3 Jun 2005 (UTC)

Would you be able to help by outlining in what way the proposed definition may be perceived as deficient? I'm sure it could probably be phrased more clearly.

When composing it I relied not only on the sense in which I understand the term, but definitions offered such as that found in the American Heritage Dictionary at Answers.com, which lists:

1. The art or study of correct spelling according to established usage.
2. The aspect of language study concerned with letters and their sequences in words.
3. A method of representing a language or the sounds of language by written symbols; spelling.

Since we are talking here about all types of writing systems, and not just those for which "spelling" and "letters" are relevant and applicable, the definition needs to be extended to cover in general these other instances of writing systems.

Thus, the attempt to capture the two major senses- "the study of writing systems, their rules and relations" equates to the name of the field of study (senses 1 & 2, above), and "describes the set of elements and rules themselves" equates to the object of the field of study (sense 3).

In this, the usage of the term "orthography" follows that observed for probably all other "-graphy" or "-ology" words, in that the same term stands for both the "study of", and the "thing studied"; for e.g.:

• there is more to Geography than the study of maps
• the Nullarbor Plain has an exceedingly flat geography

Indeed, the definition currently given in the orthography article ("the set of rules of how to write correctly in the writing system of a language") could itself probably do with a rewrite/update/expansion. --cjllw | TALK 07:18, 2005 Jun 5 (UTC)

Except the "ortho" in "orthography" means "straight" or "correct" (the word literally means "straight writing"). The comparison with geography is not really apt, because the analogous extension meaning for "orthography" would be the general study of straightness or correctness, not the general study of writing. Orthography refers to the rules for spelling, or the study of the rules for spelling, but not for the study of the letters or symbols that are used to spell. There is no general, generic name for the study of writing systems other than "the study of writing systems". The only suitable candidate, graphology, was unfortunately co-opted by pseudoscientists. Maybe we could coin a new term graphography. But the use of "orthography" to refer to the general study of writing sytems is not something that you'll find many linguists agree with. Nohat 08:05, 5 Jun 2005 (UTC)

Thanks. I had wanted to avoid using the term "graphology" because of its connotations, although several sources such as Encarta give "the study of writing systems and their relationship to the sound systems of languages" as a definition. Likewise, the cognate term "grammatology" as used by I.J.Gelb is now associated primarily with the work of Derrida. The term "graphemics" is used in several directory listings, but is perhaps a little obscure.

A better parallel might be with epigraphy, which is used to refer to both the study of inscriptions, and the inscriptions & rules themselves ("calendrical information features prominently in Mayan epigraphy").

However, I take your point; it may well be too much of a stretch then to associate orthography with study of all aspects of writing systems, although the term is frequently used to address writing systems such as Chinese, Mayan and Harrapan (Indus) scripts where "spelling" is not really an applicable concept. When used in this context, it applies not only to the conventions of romanised transliteration, but refers to the structure and rules of the signs themselves.

Perhaps a reformulation along the lines of In the study of writing systems, "orthography" refers to the method and rules of observed writing structure, and their study? The merits or otherwise of writing system terminology could maybe be explored elsewhere in the article.--cjllw | TALK 02:26, 2005 Jun 7 (UTC)

[edit] Syllabaries vs. abugidas

I've always found this thing confusing. As I see it, in syllabaries, every syllable is represented by a grapheme, with usually no visual correspondence between related syllables. Whereas, in abugidas, there's a base syllable which is "modified" by markers, such as those for vowels in Devanagari. So why is it a given that abugidas have to have an inherent vowel (acc. to the article)? Why can't it be that the base grapheme in an abugida represents only the consonant? Or if such a writing system is termed something different, what would it be? Ambarish 29 June 2005 06:57 (UTC)

That's a bit of a philosophical difference: are we simply not bothering to write /a/ (as most vowels are treated in an abjad), or do we have a bunch of true /pa ta ka/ etc. syllables that are modified for other vowels? I guess you could ask the same thing about tone in the Latin alphabet: if the three tones of a language are written á a à, does that mean that we simply aren't bothering to write mid tone, or does a have an inherent mid tone that is modified by adding diacritics? Not so straightforward.

It's probably best to consider how the script is conceived of by the people who use it. Being raised with a (true) alphabet, it's easy for us to see abugidas as alphabetical. However, Ethiopic (the prototypical abugida) is learned and treated as a syllabary: each CV combination is learned and read as a separate syllable; similarities between syllables are simply convenient mnemonics. Thus the old label "alphasyllabary": alphabetic in conception, but syllabic in use. Something similar occurs with Hangul: its conception may be featural, but it's learned and used just like an alphabet.

I don't know how Indian abugidas are conceived of by their users. However, given the natural human bias toward syllabaries, I wouldn't be surprised if each basic letter were considered an inherent syllable by most users. kwami 20:23, 2005 July 16 (UTC)

[edit] Grapheme vs Character

This article treats the terms grapheme and character as synonymous, but the article on graphemes does not mention "character" at all. Please see my comment on Talk:Grapheme. I am mainly interested in providing as much disambiguation as possible between this usage of "character" and character (computing). Thanks —mjb 9 July 2005 03:37 (UTC)

[edit] Map

Very cool map.

Mind if I add a few details? Probably should have the Nko and Vai scripts (Nko at least is quite vigorous, and Vai hanging on), and perhaps the Afaka script of Surnam. Don't worry, I won't replace the original map. There's no room for the other scrips of S & SE Asia, except in Indonesia; but last I knew, Javanese and Balinese were learned by kids in school, but weren't actually used for much except fun and some temple inscriptions, rather like the Arabic alphabet in Turkey. We should probably stick to scripts used as the main medium by a community.

Can anyone think of anything else that's missing, or needs modification? I'm a little puzzled by the distribution of Tifinagh and Tibetan, for example. (Should probly post this on German Wikipedia.) kwami 20:34, 2005 July 16 (UTC)

As I've written on the image talk page, I strongly dislike the use of "Wikipedia" as sample word. It's not only navel gazing, it's plain wrong that the name is "Wikipedia" in all languages using latin script. --Pjacobi 20:39, July 16, 2005 (UTC)

Navel gazing, yes, though personally I don't have a problem with that. But can you think of any word that would be the same in all languages using the Latin alphabet, that we could use instead? Or any other inter-ethnic alphabet? The Chinese isn't what "Wikipedia" looks like in Taiwan, either, nor the Arabic in Iran, nor the Cyrillic in Bulgaria. kwami 21:28, 2005 July 16 (UTC)

The script name or "Unicode" come to mind. --Pjacobi 21:31, July 16, 2005 (UTC)

Neither of those would be universal: Unicode/Unikod, Юникод/Уникод, 统一码/通用碼, etc.; or Latin/Latynse/llatinu/łaciński/latinka/lateinische/latinalaiset/llatí/Latín/latino, etc.

BTW, I don't see your comments on the image page; the discussion page there appears blank. kwami 21:53, 2005 July 16 (UTC)

The shape of Chinese-speaking and Mongolian-speaking areas are quite...odd. I understand this is a simplication. In this case, Chinese-speaking area can be simplified to be within the territory of China and Taiwan. As of now, it extends into Far East Russia... Mongolian would be in Mongolia and Inner Mongolia. As of now, it seems to only be in Inner Mongolia. --Menchi 23:51, 16 July 2005 (UTC)

These are writing areas, not speaking areas. Is the Mongol alphabet now widely used in Outer Mongolia? I'd thought it was still primarily used in Inner Mongolia, and to a lesser extent, Xianjiang. The northernmost bits of Chinese are I believe Manchuria and Xianjiang, which would be right, except that maybe Xianjiang should be Arabic? And then there's Tibet... Granted, this isn't a very precise map! kwami 01:47, 2005 July 17 (UTC)

[[Image:WritingSystemsoftheWorld(test).png|thumb|right|Test image (without labels yet)]] (this image is now superseded by the one on the main page)

How's this for a start? It would be nice to get some feedback before I go to the effort of adding the scripts. Same color scheme as the other. It has all the national borders; the internal borders in China are still sloppy, but the rest should be reasonably precise. The new script in South America is Njuka; in West Africa, Vai and Nko; in Morroco, Tifinagh; in Algeria, the Latin is for Kabyl. I'm uncertain about ex-Soviet Central Asia and Mongolia.

Any factual errors? Suggestions? (Personally, I think it would be nice to give Devanagari a separate color, but need to dig up a reference map.) Orange could maybe be reserved for abugidas, maybe red for syllabaries (kana, Yi, Njuka, Cherokee, Vai), etc. kwami 11:09, 2005 July 18 (UTC)

my suggestion would be to colour-code related scripts, by system or by genesis (e.g. Chinese and Japanese in a similar colour, alphabets descended from Greek / Aramaic in similar colours or colour code for abjad / abugida / alphabet / syllabic / ideographic ) dab (ᛏ) 12:02, 18 July 2005 (UTC)

Maybe I could stripe Japan to be both syllabic and Chinese/logographic? Also, Zhuyin is used in Taiwan for aboriginal languages, so should perhaps add that (not a lot of room, though). kwami 20:18, 2005 July 18 (UTC)

Okay, updated a bit: Cree in Canada, Arabic & Tuareg in Mali & Niger, Yao in China, Formosan in Taiwan, Devanagari. Syllabaries now red, abugidas orange, misc. alphabets/abjads green. China's still a mess.

And Santali, now distinguish alphabet from abjad from featural (Hangul), and 2nd location for Cherokee.

Okay, all web-safe colors now; brought the color of Cyrillic into line with other alphabets (blues = alphabets, greens = abjads, oranges = abugidas, red = syllabaries, yellow = Chinese); added a couple more minor scripts. Note: the minor script labels are there so you know what I've included; they are not intended to be the final display.

As you can see, I don't know how to enter Arabic text in Photoshop. Can anyone help?

Two things we need to decide:

• Should Mongolia still be considered Cyrillic, and
• Do we want to keep "Wikipedia" as the exemplar, or should we use the name of the script? If the script, I'd suggest using the form in the most wide-spread language using the script: the English word for "Latin", the Russian word for "Cyrillic", the Arabic for the Arab abjad, the Hindi for Nagari, the Simplified Mandarin version for Chinese. Not sure I can find all the names of the scripts in the scripts themselves though. (But then, I'd have to make up equivalents of "Wikipedia" for several of them too!)

kwami 01:58, 2005 July 20 (UTC)

hi. I only have a personal bit to offer in regards to Mongolia: about 6 yrs I met some Mongolians and the one person I asked said that they used primarily Cyrillic, but they were taught the old script in school. He could write in the script, but he really to think a bit about it & wrote it pretty slow (maybe like remembering how to do math or chemistry from high school). He was about 20-25 yrs old at the time. Maybe it is taught more now to children? That's my limited knowledge... – ishwar  (speak) 03:40, 2005 July 22 (UTC)

There are other scripts like you describe, Ishwar, such as Javanese, which I'm not indicating. However, I believe Mongolian is being officially revived, which gives it much more importance. How about I ring Mongolia in dark blue for 'other' alphabet? (There are a lot of scripts in S China/SE Asia, such as Frasier, which aren't making it in, due to lack of space in that part of the map.)

Also, given the lack of comment, I've decided to use the script names instead of 'Wikipedia'. This should also now be close to the final script boundaries. Added the Alaska script, which was still in use in the 1970s, and Batak. Should Javanese and Balinese be included? Perhaps Ethiopic should not be so extensive, if Oromo's switch over to Latin? Please comment, or let me know if you see any errors. kwami 07:50, 2005 August 1 (UTC)

Okay, have added many of the labels. The ones without Unicode support (or only multiplane Unicode support) still aren't done, and the Arabic is still screwed up. Can anyone help? (I'm using Photoshop.) Also, several of the scripts are actually the names of the languages. This might remain the case in India, due to lack of space, but often I simply don't know the name of the script in its language. Any comments would be welcome. kwami 23:42, 2005 August 25 (UTC)

Arabic's in, Hebrew's corrected, minor scripts added. It will take a while before I can dig up fonts for the remaining scripts: Mandaic, Nko, Vai, Ol Cemet', Balinese, Javanese, Batak; might try to throw in Pollard & Sign Writing. And I have no idea which glyphs to use for Nu Shu.

What about the layout? kwami 12:22, 2005 August 26 (UTC)

Nearly done: I found a Mandaic font, but am having trouble with it. I also plan to change Mandaic from abjad to alphabet. Other than than, I'm just missing Alaska (Yup'ik) and Nü Shu. Might touch up a couple others, but this is basically done. Any objections to replacing the image on the main page?

kwami 07:44, 2005 August 27 (UTC)

Yup'ik is the only script left on the map to illustrate. One person suggesting hash-marking biscriptal areas (Xinjiang? Inner Mongolia?), and another would like better screen resolution. Does anyone have a higher res blank map of the world? I'll go ahead and put this on the main page. kwami 10:10, 2005 August 28 (UTC)

Kwami, very nice work, the map is admirably informative and comprehensive. I particularly like your solution to indicate dual script-use areas; for clarity, perhaps this could be mentioned below the left-hand Legend.

I've only a couple of minor comments:

• in the Legend, "other abjads" rather than "others abjads"
• the untranslated scripts of the Indian subcontinent are crowded, but not much can be done about that. Perhaps either use a numerical superscript against each which can relate to your listing of them below (but might be difficult to see), or alternatively rephrase your listing along the lines of "The untranslated scripts in the Indian region are (anticlockwise, starting from the entry above Nagari)- Gurmukhi,... etc"
• in some places where the script label is meant to refer to a smaller region surrounded by other script types, perhaps a line drawn from the label to the region could be used to clarify which is meant (eg in the Middle East/Caucasus); but this may make it too crowded;
• Chinese label is not translated.

Once again, thanks for your excellent work!--cjllw | TALK 01:02, 2005 August 29 (UTC)

Thanks. Will check out adding some lines & superscripts, and will add a key for biscriptal areas. kwami 07:38, 2005 August 29 (UTC)

Kwami, I hate to rain on your parade, but I think the old image was much more attractive. In particular:

1. The font used for the Latin script doesn't look nearly as good. This kind of image is much more suited to a sans serif font than a serif one because serifs make the text harder to read. Serifs make linear text in paragraph form easy to read but they just make a distraction when they are used as labels. Also, a font with a narrow letter-width, like that used in the old image is better because it makes much more efficient use of the space.
2. In the old image, all the script examples were of the same word (the admittedly navel-gazing Wikipedia), but the benefit of this was that the relationship between similar scripts was much clearer. All the indic scripts, for example, show their relationship with each other much more clearly whent the word is the same. Many of the script examples, like Thai and Lao are too short to be very enlightening. Also, the uniqueness of Cherokee is completely opaque in the new image. It just looks like the letters CWY.
3. The text is proportionally smaller and so harder to see. When both the old and new maps are viewed at the same time, in the new map, the text is much harder to see.
4. Bright red for syllabaries draws unnecessary attention to them.
5. The anti-aliasing of the Latin text is inferior in the new image compared to the old image. It looks pixelated.
6. The "blurry circle" approach to a legend looks messier than the enclosed squares, which is a much more standard way of doing a color legend
7. The introduction of national boundaries adds additional unnecessary clutter.

In all, I recognize that the new image is better in the sense that it is more accurate and contains more information. However, I think that it is substantially aesthetically inferior to the old one. Whereas the old image exuded the quality of being made by a professional illustrator, the new image is clearly the work of an amateur. Nohat 02:51, 29 August 2005 (UTC)

Some good points, Nohat. I've doubled the resolution, which should help with several of them. (This map was 2/3 the size of the prior one, which was unfortunate.) If the anti-aliasing is still a problem, I can try changing it - if you know Photoshop, perhaps you have a suggestion. A sans-serif font and square key are easy enough to arrange. I can expand several of the samples to the name of the script instead of the language. As for using the same word for everything, that has its problems too (I have no idea what I'd do for Nü Shu!), and people objected to it as well. "Alphabet" is common to Latin-Greek-Cyrillic; maybe I can add "akshara" to all the Indic fonts, or just use that word alone?

I'll see what I can do for another color for the syllabaries. Maybe brown? There aren't all that many distinct web-safe colors. (None of the colors in the old map were, for example.)

However, I see the national boundaries as a plus, and that was one of my motivations for creating a new map. It's nice to know where the things are. kwami 07:38, 2005 August 29 (UTC)

You really don't need to worry about web-safe colors. If you are downsampling the entire image to 8-bit color (to be "web-safe"), then that is probably causing the lousy anti-aliasing. Almost no one uses a computer these days that is limited to 8-bit color, and those who are are accustomed to the vast majority of the web being dithered. Just upload a version with full 32-bit color—it will look fine even on an 8-bit display, but it will look great on 32-bit displays, as opposed to now, where the anti-aliasing just looks badly dithered, reminding me of the web circa 1994.

As for font choice, I implore you to avoid Arial/Helvetica. They're widely regarded as ugly and passé (albeit ubiquitous). I'd also avoid any of the "web" fonts like Verdana and Trebuchet because they don't hold up well when resized to arbitrary point sizes, which is what happens when they get embedded in images. I think the old image used Myriad condensed (actually it seems to be a custom-generated Multiple Master version of Myriad with a width about halfway between regular and condensed) and it looked quite nice. But any of Futura, News Gothic, Univers, Gill Sans, or Stone Sans would look just as nice, particularly if a condensed variety is used. Contact me via e-mail if you need information about how to obtain or generate a particular font.

Also, I'd really recommend taking a cue from the old image and putting the Indic scripts in the Indian Ocean (where they belong ;-) with a key indicating where in India they are used. Same with putting the SE Asian Indic scripts in the Pacific. It was much less cluttered that way. In fact, in general, I'd say it looks much cleaner to have as many of the example texts in the ocean with lines pointing to the area of use. You could put those west African scripts in the mid-Atlantic. It's a much more "even" use of the available space, especially now that there are all those national boundaries adding additional clutter. As it stands, the oceans are these vast seas of blank space with all the information—colored regions, national boundaries, not to mention the example texts and script names—all crammed onto the continents. It just makes the whole image seem kind of unbalanced. But don't overdo it: Chinese, Latin, and Cyrillic can stay inside the larger countries. However, if you choose not to heed my advice concerning the placement of labels, please at least fix the glaring grammar error in the explanation of the placement of Indian scripts.

I'd say that for non-ideographic systems, you'll want at least 5 characters of sample text to give an adequate "flavor" of the system. If you put the example texts in the ocean, then you'll have more space for longer samples.

I suppose I can live with the national boundaries if you fix these other problems. It sure would be neat if the map were clickable.

Cheers! Nohat 08:42, 29 August 2005 (UTC)

A couple other things:

• "Others abjads" in the key should be changed to "Other abjads"
• You don't really need the parens around the Latin names of the scripts. (More needless clutter). Also, try to make the Latin names of the scripts larger so they can be read more easily.
• I don't think the outlining of outer Mongolia and Japan is working out that well. I'd really recommend a diagonal stripe, like what's typically done with time zone maps. See [1] for an example. Nohat 08:51, 29 August 2005 (UTC)

I wanted to try placing the names in their actual locations to see if it would work out. I haven't had any other complaints, but you're right, it would make things much easier to have them in the oceans. Also, with a sans-serif font for the labels, the parentheses would no longer have any function.

Can we make clickable maps in Wikipedia?

It might be possible using some clever CSS. Maybe something along the lines of what was done on Template:CSS IPA vowel chart. Nohat 21:12, 29 August 2005 (UTC)

Thanks for the font suggestions. Don't worry; I detest Arial and Helvetica for Latin (though Arial's not bad for many of the Indic fonts).

I originally had diagonal lines, and thought they looked horrible. kwami 19:12, 2005 August 29 (UTC)

Technical: What program are you using? Perhaps it would be a good idea to load an editable copy (separate layers) to Commons, so that variants for other language Wikipedias can be easily created. --Pjacobi 09:59, August 29, 2005 (UTC)

Photoshop CS (8.0). I have anti-aliasing set to "sharp". I was planning on posting, but the doc's 4.3MB. Can I also post fonts? Otherwise I don't know how much good it will do. Some of the fonts are uncommon, and I had to create a font of my own for symbols that Photoshop wouldn't support in the original font (like Vai), or which I didn't have a font for (like Nü Shu), or where the only available font contained errors (like Afaka/Njuka, which is missing the letter jo/ju!). kwami 19:12, 2005 August 29 (UTC)

Size wouldn't be a problem, but fonts may be problematic, if they are not PD or GFDL. Perhaps the Commons version can be just the English map with all text in latin script stripped.

BTW: Looking for Njuka, I got the impression, the more common spelling is Ndyuka, e.g. [2].

Pjacobi 20:53, August 29, 2005 (UTC)

For the non-Latin scripts you could just render the layer and not worry about uploading fonts. You could also try making EPS files of the non-Latin script pieces and including those. You could do that using the "make outlines" feature of Illustrator or somesuch. Nohat 21:08, 29 August 2005 (UTC)

This map is fascinating to me, but the representation of the Mongol script is odd. The Mongol script is written from left to right. So mongγol and bičig should be reversed. Also, I think mongγol üsüg is better than mongγol bičig if you mean "alphabet" rathar than "document." --Nanshu 22:22, 29 August 2005 (UTC)

Thank you for catching that! I had simply assumed the direction of writing was as in traditional Chinese, without verifying. (Once upon a time I knew the difference!)

As for üsüg, I assume that it is spelled "uysuk", with the same final "k" as in bičig, correct?

Also, do you know the situation of the Mongol script in Mongolia? Is it correct to portray that country as biscriptal? kwami 00:22, 2005 August 30 (UTC)

The text for "Khmer" in the Khmer script is wrong. I see that it was typed with Unicode, but it is not displaying correctly because of issues with the Uniscribe processor. Please update the Uniscribe processor or place it in the correct folder(s) on your computer. Please see the image to the right for the correct way it should look like (I used a different font though). --Dara 07:39, 3 November 2005 (UTC)

Thank you. I'll hopefully be able to upload an updated map within a week. However, I've already replaced the Khmer example with the longer name, which I took from the wiki page on the Khmer alphabet. I'll see what I can do about the error, but I don't think my computer will take any more updates. We'll see... kwami 08:00, 3 November 2005 (UTC)

Since Khmer only recently got Unicoded, computers with old Uniscribe processors cannot properly render the script. So it is necessary to update or else you might see a lot of dashed-circles for vowel diacritics and such. Read about the Uniscribe processor here. The sign below that looks like a cross is called the Coeng Sign (it has no meaning in Khmer). It is suppose to convert consonant symbols typed after it into their subscript forms. Instead of ខ្មែរ (Khmer), you can use ខេមរ (Khemara) which does not use a subscript consonant. It is a more formal way version of Khmer. --Dara 02:27, 4 November 2005 (UTC)

The map is affecting text layout; the page looks rather odd. I tinked with placing the map elsewhere in the Lead, but to no avail. Can one of the map's contributors or creators address this? Thanks. -Fsotrain09 17:46, 27 April 2006 (UTC)

[edit] Graphemics

Hi, I wonder why there is no such Article or Category for this one. Would it be a subcategory of Linguistics or Semiotics? If the latter, I guess Linguistics has to be categorized beneath Semiotics either. --Trickstar 12:30, 1 September 2005 (UTC)

[edit] Proto-Elamite and OES

Since there's been some disagreement as to the proto-Elamite and "Old European" scripts, and Daniel & Bright's book is expensive enough that small libraries can't afford it, I'll copy the relevant sections here.

[edit] The Proto-Elamite Script

(Robert K. Englund, Asst. Prof. of Near Eastern Studies, Free University of Berlin. Research includes the decipherment of proto-cuneiform and administrative forms of 3rd-millennium BCE Mesopotamia)

The ideographic writing system conventionally called Proto-Elamite was developed and used in western and southern Persia at the end of the fourth through the beginning of the third millennium B.C.E., a historical phase generally considered to cerrespond to the Jemdet Nasr and the Early Dynastic I periods in Mesopotamia (Le Brun 1971; Damerow and Englund 1989: 1-4). The region of Persia designated "Elam" in later Mesopotamian cuneiform sources lent its name by association to the language spoken there; Old Elamite/Old Akkadian bilinguals employing the partially deciphered linear Elamite and Old Akkadian cuneiform date this language of unknown linguistic affiliation (Reiner 1969) no earlier than ca. 2300 B.C.E. "Proto-Elamite" is the name used for the writing system of the earliest documents from the region—texts on clay tablets which are assumed to represent a precursor of Old Elamite (Hintz 1975; Meriggi 1971: 184-220; André and Salvini 1989). The earlier language has not, however, been identified; the phonological structure of the archaic script is thus entirely unknown. However, contextual analyses and the formal similarity of Proto-Elamite documents to better-understood proto-cuneiform tablets from Mesopotamia dating to ca. 3200-3000 B.C.E. make possible a substantive assessment of the ideographic nature and the fields of application of the indigenous Persian writing system.

History of decipherment

Since the first archaic texts were discovered at the turn of the twentieth century, some 1500 Proto-Elamite tablets have been published, the great majority excavated at Susa on the Kerkha river east of Babylonia, but including in smaller numbers tablets found in sites reaching to the southeast across to Shahr-i Sokhta on the Afghanistan border (Damerow and Englund 1989: 1-2). The tablets are administrative documents, to the near total exclusion of either literary or lexical texts.

Syllabic sign readings adduced from an assumed link between Proto-Elamite and the ostensibly related linear Elamite (see above) have not lead to successful decipherment of the archaic script. A prelinimary graphotactical analysis of the Proto-Elamite texts has also met with only modest success (Meriggi 1975: 105, 1971: 172-84; Brice 1962-63: 28-33; Gelb 1975). To be sure, scholars have with mixed success established some graphic and semantic connections between Proto-Elamite and proto-cuneiform, the first writing stage of which predates that of Proto-Elamite by some 100 years (Langdon 1928: viii; Mecquenem 1949: 147; Gelb 1962: 217-20; Meriggi 1969: 156-63; Damerow and Englund 1989: 11-28). However, a lack of necessary philological tools, above all a dependable sign list purged of redundant sign variants, continues to hinder progress in this work.

Basic characteristics of Proto-Elamite script and texts

A priliminary study of the entire text corpus suggests that the Proto-Elamite sign repertory was comparable to that of proto-cuneiform, using less than 1000 individual signs and thus in the range of logo- or ideographic writing systems (Dumerow and Englund 1989: 4-7). Superficially, a large number of signs seem entirely abstract—which, considering the probability that the script developed explosively during the Jemdet Nasr Period (ca. 3050-3000 B.C.E.), suggests that its developers consciously chose geometric and other nonpictorial shapes and introduced them into conventional usage. The extent to which pictography may have been represented in a dead script is, however, difficult to discern.

The first serious work on a formal description of the Proto-Elamite texts was done in the 1960s and early 1970s (Brice 1962-63, 1963; Meriggi 1971-74; Vaiman 1972). Proto-Elamite documents were written in a linearized script from right to left, in lines from top to bottom. The first signs on a Proto-Elamite tablet generally express the purpose and acting person or institution of the text, followed by individual entries, without the formal arrangement of the tablet into the columns known in proto-cuneiform (see FIGURE 13). Each entry normally includes an ideographic notation representing persons/institutions or quantified objects or both, followed by a numerical notation. That all entries in Proto-Elamite texts seem to contain a numerical notation suggests they represent more the structures of a system of bookkeeping than the division of a spoken language into distinct sentences or comparable semantic units. Continuing analysis of the Proto-Elamite numerical system (see FIGURE 14), which derived from the systems developed earlier in Mesopotamia, has been a powerful tool in recent semantic identifications of a number of signs and sign combinations, including those for animals, for grain products and, it seems, for humans ([lots of refs follow]; see FIGURE 15).

[edit] The Old European script

In The First Civilizations, Peter T. Daniels.

"Forerunners" of writing

The full range of scripts of the ancient Near East are or appear to be related to each other by immediate adaption or by direct, conscious influence. There are, in addition, remains of two recording devices that have been hypothesized to underlie Sumerian cuneiform, the earliest true writing system: the Vinča signs and the Near Eastern clay tokens.

The Vinča signs

The Vinča culture, found in the central Balkans and dating to 5300-4300 B.C.E., is named for the Serbian site southeast of Belgrade, Yugoslavia, where it was initially excavated during the first third of the twentieth century (Gimbutas 1991: 62-70, with bibliography, describes the culture and places it in its Balkan and wider contexts). Numerous objects—what Gimbutas notes as "religious items only" (p. 308)—bear graphic marks that look as though they might be elements of a script (pp. 308-21); often an object displays a series of such marks. According to an analysis by Winn (1973/1981), there are 210 signs; 30 are core signs, with the remainder being variants and combinations (Gimbutas, figures 8-1 and 8-2). Gimbutas supposes (as Winn apparently does not) that these marks are a writing system that records the pre-Indo-European language of her "Old European" civilization. She goes so far as to (facetiously) compare Old European signs with those of Linear A (her figure 8-22) and the Cypriote syllabary (figure 8-23; for both, see SECTION 7). Had Gimbutas not included the proviso "for our amusement" (p. 320), she would have committed the oldest fallacy in the study of writing systems: the comparison of shapes alone without attention to sound values. Researchers must, therefore, not be misled by such charts into retrojecting the Greek values to a putative Old European language. Moreover, no claim seems to have been put forward that any recurring sequences of signs have been identified, and the first step in any linguistic analysis is the identification of strings that are the same or partly the same. It seems, then, most improbable that the marks represent a language, either logographically or phonetically. Thus on the current evidence it is not possible to recognize a Vinča writing system. A sober study of these and related materials, e.g. the "Tartaria tablets," by a scholar of the Aegean scripts—not cited by Gimbutas—is Masson 1984.

That's all that the 900-page tome has to say on the "Old European Script". kwami 07:39, 23 October 2005 (UTC)

Tnx for the clarification and above fragments. That's very interesting. --Piotr Konieczny aka Prokonsul Piotrus ^Talk 01:54, 30 October 2005 (UTC)

Yeah, I'd have my doubts that proto-Elamite were even a writing system, rather than an accounting system, if the sign inventory didn't equal that of cuneiform. The Vinca signs don't even have that going for them, and the lack of repetition might be more indicative of religious or magical incantations than writing, like the abrafax (abracadabra) tradition in Judaism, since human language is full of repeating elements. kwami 19:00, 30 October 2005 (UTC)

[edit] Question on external link

I don't understand the relevancy of the link "A Typographic Outcry: a curious perspective". Can someone elaborate? The link also exists in Logogram and Chinese character.

[edit] SignWriting

What kind of writing system would SignWriting be? It's a writing system for sign languages (as is HamNoSys). Should these be included one way or another in this article? (I have no expertise on this, just knowledge of the existence of these writing systems.) --Jadriaen 22:04, 20 December 2005 (UTC)

It seems to be a featural system rather than phonemic, because the symbols are depictions of the hands and face. I don't believe it follows any rules as to representing what is contrastive. However, I would imagine that in practice only meaningful elements would be included, or that the script would evolve that way over time. You'll have to find someone who really knows how the script is used to know for sure. kwami 22:44, 20 December 2005 (UTC)

Could Sign language#Written forms of Sign Languages be of some help? --Jadriaen 12:14, 21 December 2005 (UTC)

I wrote much of that, and I fear it may turn out to be a rather ignorant account. Part of the problem is that I don't know how the tab-dez-sig-ori elements of SLs correspond to phonemes vs. features; another is that I don't know how much redundant material is commonly included in SignWriting, or how systematic it is. That is, if you have several people transcribe a video of someone signing, will they come up with the same sequence of SignWriting glyphs? kwami 12:29, 21 December 2005 (UTC)

[edit] New article to merge?

Somebody created a new article on Earliest writing, reporting on 5000-year-old pottery shards from Harappa and on the Chinese tortoise-shell findings. Could somebody with a knowledge of the field look over, I have the feeling that piece could do with a bit of expert attention, and might possibly be merged here (or at Writing#Writing_in_Historical_Cultures, depending on what happens to that merge proposal.) Lukas 14:19, 12 January 2006 (UTC)

redirect here. The article has some potential, but it treats proto-writing rather than actual writing. If the material in the history section here gets out of hand, we can branch out into History of writing. dab (ᛏ) 14:48, 12 January 2006 (UTC)

Just want to say i think writing and writing system should merge as entries, though writing should re-direct here or there should be a seperate history of writing section. -Jan 27, 2006, megan

All of this should be merged to History of writing.-- _{Piotr Konieczny aka Prokonsul Piotrus | talk}  17:40, 1 September 2006 (UTC)

[edit] Invention of writing - Rongorongo

Don't you think that some mention about Rongorongo should be included. It's independent from Sumerian and Chinese scripts --xRiffRaffx 12:48, 29 September 2006 (UTC)

Rongo-Rongo is one of those writing systems that gets blown all out of proprotion to its importance. It is an inverted boustrophedoun writing system. There have been periodic claims about its decipherment, but none have held up. There are two theories about its origans. One that it was created because of European sailors visiting Easter Island. The other is that it was created and forgotten before the first wave of European sailors visited Easter Island. 64.40.60.88 21:41, 29 October 2006 (UTC)

[edit] General Properties

"Contrast this with other possible symbolic systems such as information signs, painting, maps, and mathematics, which do not necessarily depend upon prior knowledge of a given language in order to extract their meaning."

Information signs, paintings, and maps I agree can frequently be meaningful to people regardless of their background, but math? I totally disagree with that. Mathamatical expressions are every bit as arbitrary and abstract as written language. Tell me, who would know be able to figure out a basic calculus expression such as:

without any mathematical education? Even something as elementary as 2 + 2 = 4 isn't in any way intuitive.

--amRadioHed 00:48, 1 November 2006 (UTC)