Wikipedia talk:Naming conventions (writing systems)

From Wikipedia, the free encyclopedia
Jump to: navigation, search

Naming consistency[edit]

archived from Wikipedia talk:WikiProject Writing systems

section renamed

A bunch of scripts were changed from X script to X alphabet, for consistency with a few scripts, like Devanagari, Hebrew, Arabic, etc.

I think we should revert all of these (except maybe Proto-Canaanite and Ugaritic), but would like some sort of consensus. VIWS talk 23:47, 30 July 2011 (UTC)

I'm fine with moving them all back to 'script', but if we do that, we need to move Hebrew alphabet, Arabic alphabet, Devanagari alphabet, etc. as well, which may take some doing. I don't much care which word we use, as long as we treat minor/regional scripts equally with the major (from an English POV) scripts. It's unjust to say that Hebrew but not Aramaic, or Nagari but not Bengali, is an "alphabet". — kwami (talk) 00:03, 31 July 2011 (UTC)
I think that we should consistently name abugidas, syllabaries, "ideographies", and undeciphered systems as "scripts" or even "writing" when appropriate for undeciphered, pasigraphic, asemic, etc., with the Semitic and European scripts to "alphabet", as well as any other scripts that segment C/V/C or Cv/Cv, but not CV/CV or CV(C)/CV(C). In other words, an "alphabet" is a "script" with segmentable consonants and either segmentable vowels, or understood, but not inherent vowels. What do we think? VIWS talk 00:21, 31 July 2011 (UTC)
As we had it before my moves, "script" has two meaning in our titles: style (cursive, clear, square, grass, block, italic script) and funny squiggles made by obscure peoples. With one or two exceptions I hesitated over, it currently AFAIK only has the former meaning.
I'm not sure what your "Cv/Cv" notation means: I would interpret it to mean abugida, but you specifically exclude abugidas. Why? They're just as much alphabets as abjads are, if not more so, since they're fully segmental while abjads are not.
I was using the "/" to indicate the verbal element corresponding to a writing segment, so a small "v" would be an understood vowel, opposed to an inherent or explicit vowel. Maybe Cv/Cv is better. I just think there is a distinct divide between the alphabetic/abjadi tradition and the Indic 1abugida tradition that should be respected, even if you assume the Brahmi-from-Aramaic hypothesis. I personally would not name caligraphic styles "X script" - I think "Fraktur", "Carolignian miniscule", and "Nastaliq" are precise enough terms for those writing styles, for example. VIWS talk 01:29, 31 July 2011 (UTC)
They also fit the definition you just gave for "alphabet" better than adjads do. It would also lead to some oddities: Kashmiri, for example, is written in the Arabic alphabet script, yet we wouldn't be able to call it an "alphabet" because the vowels are written, making it an abugida. So, if we add vowel diacritics to a consonantal alphabet, the result is no longer an alphabet! — kwami (talk) 00:54, 31 July 2011 (UTC)
That would make it alphabetic, rather than abjadi. I think Arabic should proably be named "Arabic alphabet", but I disagree that the Bangla script should be "Bengali alphabet". I think they are of different quality. Now, I'm sure that you can find an odd example of one type of system being used in another way, but that doesn't change the general nature of the use of that script.

I'm mostly concerned about the slippery slope of "katakana alphabet" and the like, but I believe abugidas are of a different quality than abjadi/alphabetic writing systems. VIWS talk

What slippery slope? Katakana is a syllabary. It isn't segmental at all, apart from a couple oddities (soku-on, etc.), so it has no relevance here.
Fraktur, sure. But we can't rename the clear script "clear".
So, Brahmic abugidas are not alphabets, because they're distinct from the Semitic tradition? Would Ethiopic abugidas then be alphabets, because they are in the Semitic tradition?
I see the dividing line between alphabets/abugidas, which are fully segmental scripts, and abjads, which are defective segmental scripts. (Using 'defective' as a technical term.) Saying an abugida is not an alphabet is like saying Vietnamese is not an alphabet because it marks tone with diacritics rather than with letters the way Hmong does.
"That would make it alphabetic, rather than abjadi". No, that would make it neither. If you write consonants with letters, and mark vowels with diacritics, you have an abugida. Kashmiri is an Arabic-based abugida, just as Brahmi may have been an Aramaic-based abugida, or Amharic is a Ge'ez-based abugida, and by your convention not an alphabet.
If you're only worried about kana, that's a non-issue. No-one claims they're alphabets unless by "alphabet" they simply mean writing. There's no connection to abugidas. But I don't see why we should start making arbitrary choices about which segmental scripts qualify as alphabets: either only full/true alphabets do, in which case we move Arabic and Hebrew to 'script', or they all do, in which case Bengali and Thai stay at 'alphabet'. Anything else strikes me as a personal take on what's close enough to normal to be called 'normal'. — kwami (talk) 02:03, 31 July 2011 (UTC)
No. I'm saying that "alphabet" is only consistent with alphabetic and abjadi scripts. That's it. I don't believe abugidas are appropriate to call an alphabet. Script is neutral, if less than elucidating. Like I said, I want to get consensus on this issue, but I'd prefer not to have my proposal - which is insanely clear - not to be misconstrued: and I quote myself: "I think that we should consistently name abugidas, syllabaries, "ideographies", and undeciphered systems as "scripts" or even "writing" when appropriate for undeciphered, pasigraphic, asemic, etc., with the Semitic and European scripts to "alphabet"." Period. If a caligraphic or print style is consistently refered to in plain English as "script", I don't think that should call into question every other item that we call a script as well. On the one hand, we have consistency, on the other exclusivity. I argue only for consistency. I think that refering to syllabaries as "scripts", but abugidas as "alphabets" doesn't make sense. I think the dividing line is syllabic segmentation. VIWS talk 03:42, 31 July 2011 (UTC)
"Syllabic segmentation". Abugidas are not "syllabic". They are segmental scripts. (A segment is a consonant or a vowel; the only syllabic segments are vowels, syllabic nasals, and the like.) If it is not acceptable to call them alphabets because they don't treat vowels as separate letters, then it is not acceptable to call Hebrew and Arabic alphabets either. If we can agree to move all abjad articles to 'script', then I have no problem moving the abugidas as well. I also only argue for consistency: either all segmental scripts are "alphabets", or only "true" alphabets are "alphabets", but not Western scrips are "alphabets" and Eastern scripts are not.
BTW, what about Wadaad's writing? Is that actually its name? — kwami (talk) 05:27, 31 July 2011 (UTC)

Wow! Serious bad faith in moving more articles while this is being discussed. I'm extremely disappointed. VIWS talk 07:15, 31 July 2011 (UTC)

I have, unfortunately, been forced to request administrative assistance in preventing more moves until a consensus can be reached. I ask for the opinions of everyone, so that we can get as many opinions as possible. VIWS talk 07:46, 31 July 2011 (UTC)

Nearly all of those were move you agree to. The few (4?) that weren't are merely consistent with the rest.
You haven't been "forced" to do anything. And which "more moves" would that prevent? AFAIK the articles are now consistent. — kwami (talk) 07:59, 31 July 2011 (UTC)
Quite frankly, I don't know what other moves you would make. That's the problem. You moved 20 more articles without discussing it. Nobody has any idea what the heck you are going to do next, least of all me. If we can't rely on good faith, we have to rely on people with power, which means requesting admins to step in. I'm sorry you can't understand that moving all those articles without discussing it was an incredible breach. I'm sorry I can no longer trust you to act in good faith. VIWS talk 08:14, 31 July 2011 (UTC)
In general, a script is language-neutral and a writing system is language-dependent, where an alphabet is part of a writing system, but not every writing system employs an alphabet. That’s sometimes non-obvious to decide, because many scripts are used (almost) exclusively with one writing system (and language).
Some linguists, e.g. influential Daniels/Bright 1996, use script and writing system as synonyms and then need the term adaptation when a script is used outside its original writing system.
To qualify as an alphabet, a signary must be a fixed set (of letters) and it needs a conventional collation order. Some use alphabet as a hyperonym, some as a cohyponym or antonym to abjad. (Others don’t use the terms abjad andabugida at all.) — Christoph Päper 10:13, 31 July 2011 (UTC)
So do you have any sort of opinion on a naming convention for Wikipedia writing system articles? VIWS talk
I only have experience with languages that use alphabets; but I have been concerned for a while with the double usage of the word alphabet.
E.g. the English language uses a Latin alphabet that is not the same as the Latin alphabet; and the phrase the Latin alphabet itself means (a) "the set of letters used to write the Latin language" or (b) "the union of the sets of letters used to write all languages which use an alphabet derived from the Latin alphabet in sense (a)".
The situation is not as confusing with the Cyrillic alphabet. It is a collection of letters, not all of which are used in the writing system of any one language, and there is no Cyrillic language. So, the set of letters used to write the Russian language is the Russian alphabet, and that is a subset of the Cyrillic alphabet.
Would it make sense to call the set of letters used in the writing system of a particular language an alphabet and the larger collections of letters a script? That way the Russian and Ukranian alphabets would be subsets of the Cyrillic script, and likewise the English, German and Latin alphabets would be subsets of the Latin script. (I think this distinction is line with what Christoph Päper said above.)
Yes, that’s more or less what I tried to say, see also another section. By the way, some linguists, e.g. Coulmas, use roman for the script and Latin for the alphabet. — Christoph Päper 18:47, 31 July 2011 (UTC)
Roman script can also mean Roman typeface, but maybe that wouldn't be a problem. — kwami (talk) 21:07, 2 August 2011 (UTC)
A set of letters used only in the writing system of a single language would be both and alphabet and a script. If another language later used an adaptation of this alphabet, then there would be two alphabets and an enlarged script.
I don't have anything to say about those writing systems which are not alphabets. —Coroboy (talk) 11:33, 31 July 2011 (UTC)
That's an interesting idea, naming all writing systems "script", and saving "alphabet" for national alphabets. I kinda like it. How would you deal with situations, like "Arabic alphabet/script", where the two are less easily distinguishable than, say, Latin? VIWS talk
I rather like it too. It fixes a gap that's been bothering me for a while. I wonder how well we'd be able to defend it, but with good refs it should be okay.
So we'd have the Persian and Arabic alphabets of the Arabic script. Or the Persian and Arabic abjads of the Arabic script, if we prefer. Of course, with historical adjads and modern Brahmic, it may be tricky trying to decide when things are distinct enough to count as separate scripts. — kwami (talk) 12:05, 31 July 2011 (UTC)
Ease of navigation.
What is the purpose of having X [where X is alphabet/script/abugida/etc] after a language? To distinguish the writing system(s) from, in most cases, the related language(s) itself(themselves), yes? If I was typing in a search on Wikipedia (or indeed Google) for Turkmen writing system (not knowing whether Turkmen is written in Cyrillic or Arabic or Roman since independence, or historically) I'd be much more likely to search for "Turkmen script" as a neutral default rather than Turkmen alphabet (I expect all three of those to turn up as redlinks since it's unlikely that Turkmen has it's own script, but I'd still search first for "script") I can sort of see the distinction of saying "Danish alphabet" when someone is making a distinction of the extra marks in many alphabets, but wouldn't Danish script be just as good? The real pain is if the user has to guess whether a given language is on Wikipedia as an alphabet or a script. In ictu oculi (talk) 05:19, 1 August 2011 (UTC)
I agree that not all readers would know beforehand whether a language uses an alphabet/abjad/abugida etc.; but the title of an article should still be precise. A search for Turkmen script should be redirected to the article Turkmen alphabet. —Coroboy (talk) 07:14, 1 August 2011 (UTC)
I think redirects are an inevitability, no matter what is decided. The question is how to standardize the nomenclature so that someone who visits a writing system article will likely be able to guess the name of the next article. If you were to visit "Gurmukhi script", and you wanted to compare with something in Tifinagh, will the name of the current article guide the visitor to the right choice? Or will they end up at an article about the Tifinagh language (is there one? I know that Tuareg uses the Tifinagh script.) instead of the Tifinagh writing system? Or will they end up at a page for a people? In other words, are people going to be helped or hindered by the naming conventions? That's the question.VIWS talk

Tree example[edit]

I came here because I found an abugida moved to a name "X alphabet" [1], and always thought the two to be distinct. Extending Coroboy's proposal, I would like to add a tree:

An alphabet for me has the connotation of being an ordered set of characters, but that may be wrong. Script does not. I don't no about abugida etc.

Yes. For example, the German and Finnish alphabets both contain the letters ‹ä› and ‹ö›, but the alphabet ordering is quite different. Both of these alphabets use a selection from the "Latin script". —Coroboy (talk) 07:14, 1 August 2011 (UTC)

I strongly suggest to use the term alphabet not for things that are not "true" alphabets. If an abugida is not an alphabet then it is confusing to call an abugida article "Xyz alphabet". Bogdan Nagachop (talk) 13:32, 31 July 2011 (UTC)

Terminology varies; abugida and abjad are not at all consensual terms, but at least they are quite unambiguous (unlike alphabet). There’s segmental script, of course.
In your tree I would usually use singular “script” where you chose plural “scripts” and sometimes I would use “alphabet” (or a similar term) instead of “script”, e.g. for Arwi. Often you would have “Culture script” (includes transliteration) and “Language alphabet” (may include transcription) or “Language writing system” (contains alphabet/syllabary/signary(s), orthography, partially phonology). — Christoph Päper 18:47, 31 July 2011 (UTC)
I guess for Brahmic you would leave the plural? That is because the sub-elements almost share no glyphs, while the ones in the Latin group do share? If a system like Arwi only extends something it would not be a new script? Maybe better to document all the namings in a NC than in a talk page, I started Wikipedia:Naming conventions (writing systems). Bogdan Nagachop (talk) 19:53, 31 July 2011 (UTC)

Has everyone visited Wikipedia:Naming conventions (writing systems) and its talk page? VIWS talk 08:36, 1 August 2011 (UTC)

Yes, comments there may be useful. It illustrates how several aspects of our current naming are somewhat disordered. — kwami (talk) 21:36, 2 August 2011 (UTC)

Codifying the Coroboy proposal[edit]

I'd like to take a stab at codifying User:Coroboy's proposal from Wikipedia Talk:WikiProject Writing systems. This should be taken as my interpretation of how to apply those principles, not Coroboy's interpretation or intention.

In these descriptions, L is a language name or script convention, S is a script name, C is a writing style, X is any modifying adjectives, and l/s/c/x means a term is only included if it is necessary to distinguish - eg English Latin alphabet is redundant, so we list it as "L s alphabet" to get English alphabet, but German Cyrillic alphabet.

National and Notational systems[edit]


I'd like to get these in line with the above examples, with redirects from the "unmarked" terms, so:




  • other - conventional name for large ideo/morpho/logo-graphic scripts

Caligraphic and typographic styles[edit]

Why not 'C script' as currently at Insular script, Nastaliq script?
I think is solves two issues: 1)sometime later, a writing style and script name are going to conflict. 2) by having different naming conventions, it allows semi-automated editors to differentiate between vastly different types of articles in the Writing systems project.
ISO 15924 gives a code to Fraktur, i.e. regards it as script. But I don't see why this is not just a different style of Latin. I think X (script) is confusing, if WP is of the opinion these are not scripts but styles. And if it is a script it should be Fraktur script. Clever idea to use X (script) vs X script, but I think it is confusing ;-) Bogdan Nagachop (talk) 21:40, 3 August 2011 (UTC)
Yes, regardless of the ISO, we have sources below speaking of 'a fine fraktur hand'. Whether we can get away with that as a title I don't know, but IMO I agree that it what it is. On the other hand, many of the Indic 'scripts' are about as distinct as italic and fraktur hand, or kufic and nastaliq hands, but I think we would have problems calling them 'hands'. OR in any case; I doubt we'd have sources backing us up. — kwami (talk) 23:02, 3 August 2011 (UTC)
Actually, we have two topics, fraktur hand and fracktur typeface, analogous to italic hand and italic typeface. Currently the fraktur article is mostly about the typeface; we could maybe split off an article on the hand. — kwami (talk) 23:21, 3 August 2011 (UTC)
Bogdan, you have to realize what ISO 15924 is meant to codify. It is based on a system of tagging library records for language (I believe, ISO 639), and is intended to extend that purpose to the script. At one point in time, there was considered a need to specifically mark two Latin scribal traditions for library purposes. The fact is that 15924 is not some perfect end-all, be-all of classification. It is fundamentally pragmatic, and responds to needs that are not necessarily the same as ours. So please don't confuse having a 15924 code with anything other than having been recognized for a specific purpose, much as you should not mistake encoding in Unicode as a fundamental identity of a script. VanIsaacWS 08:46, 4 August 2011 (UTC)

Script entities[edit]

  • Mark (other) - for marks limited to a single writing system


  • L alphabets - conventions for writing a language in multiple scripts
Not going to be acceptable to call it 'Hindi'. Currently at Hindustani orthography.
It was an example, and Hindi is the only language I could think of at the time that I knew had two separate orthographies. How about:
    • Mongolian alphabets - for Mongolian orthographies of 'Phags Pa, Cyrillic, Mongolian, Tibetan, etc.
The problem is in calling it Hindi. Hindi is a language cluster; the standard language within that cluster is actually Urdu. But most Hindus will have an absolute shit fit if you call their language "Urdu", and many Muslims will have the same if you lump Urdu in with Hindi. It would need to be 'Hindi-Urdu alphabets' or st. But yes, 'Mongolian alphabets' is fine. — kwami (talk) 23:09, 3 August 2011 (UTC)



That's my best shot for now. VIWS talk 01:57, 1 August 2011 (UTC)

I like the general approach, though there will be problems with some specific cases. (I adjusted a few that stood out—sorry, prob'ly shouldn'ta done that. I took it as a working draught.) My concern is that we don't use the term "alphabet" to push value judgements, as we did before I made the moves to 'X alphabet'. — kwami (talk) 21:30, 2 August 2011 (UTC)
There is no value judgement. There is no cabal trying to keep the little man down. It was just how naming conventions evolved. I think it has more to do with the fact that people who are interested in writing systems create articles for smaller, less well-known scripts, so they ended up with the more specific naming convention. The ones that got created early on by the less knowledgable ended up with "common" names, which means that the term "alphabet" gets bandied about rather liberally. VanIsaacWS 23:00, 2 August 2011 (UTC)
They were once all at 'script'. That was our convention: 'alphabet' was used only for Greek-type scripts. (We just never wrote it up as a guideline.) But in the case of Hebrew and Arabic people objected, and got them moved to 'alphabet'. So no, there is no cabal, but popular scripts have lobbies to 'upgrade' them to 'alphabet', while obscure scripts are left to languish. The end result is the same. — kwami (talk) 02:15, 3 August 2011 (UTC)
  • Pan-Nigerian Alphabet - alphabet does not need capital a. -DePiep (talk) 20:38, 16 August 2011 (UTC)

Kwami's version[edit]

I think that's probably too much detail, and in some cases makes minor decisions which do not have global impact and therefore can be left to the articles. Here's a condensed proposal, though I left out a few things like numerals. — kwami (talk) 22:04, 2 August 2011 (UTC)

Well, right now the project page has approximately that level of detail, so I mimicked it - actually, I copied it and added or deleted articles to illustrate points. VanIsaacWS


The term 'script' is used with four meanings:

  1. A general segmental writing system, as opposed to its local instantiations:
  2. In the plural for a family or geographic group of such scripts; 'writing systems' conveys the same idea
  3. A calligraphic style; in many cases 'hand' may be used instead, and may avoid confusion with other uses of the term
  4. A non-segmental writing system, especially one which is logographic, mixed, or of unknown character

'Alphabet' is used for national instantiations of a segmental script, with a defined sorting order

The terms abjad and abugida, although often used in the text, are considered jargon and so are inappropriate for the title.

If an article conflates the script and an alphabet using that script, as at Georgian alphabet, then the name should reflect the preponderance of the coverage of the article: 'script' if it details the various alphabetic instantiations of the script, but 'alphabet' if it's largely limited to the dominant alphabet. Of course, these topics may eventually be sufficiently developed to split them into separate script and alphabet articles.


Exceptions may be made where a common name exists, and 'script', 'alphabet', or 'syllabary' is not necessary:

Note: "hieroglyphics" is deprecated.

Exceptions may also occur where a different technical term is widely used:


Modifiers may be used for subtypes or other cases of disambiguation:

What about...[edit]

What about the term "hand" for the writing styles, eg "Carolignian miniscule hand", or "Nastaliq hand". I just think we're going to run into a conflict somewhere if we go with "Nastaliq script". (Yes, I know there's an apostrophe somewhere in there, I just don't know exactly where it should go.) I'm not sure it works with "Fraktur hand". Maybe I'm splitting too many hairs, but I think we're moving in on a consensus, but I'd like to have a stable consensus for about a week before we start screwing around with all the article titles. Fortunately, I got approved recently for WP:AWB, so I will probably be able to automate a lot of this moving when it comes time. VanIsaacWS 23:08, 2 August 2011 (UTC)

Yes, I added in that suggestion above in the hope that it might be a viable convention. However, in some cases it clearly isn't: "seal hand", for example, isn't gonna fly, not least because it isn't handwriting. But "Nastaliq hand" does work: [2]. "Fraktur hand" can work too: [3][4]
Actually, in the case of Chinese I do see "running hand". There is even "seal hand",[5] but I'm dubious that would get far without a challenge.
BTW, moving things w AWB isn't much faster than moving them by hand, unless I'm missing something. — kwami (talk) 01:59, 3 August 2011 (UTC)
Yeah, I started looking through the documentation for AWB, and it doesn't look like they have move functionality yet. I was going to use the Writing system assessment of importance=low to help generate a list of letters that have non-standard names, and move them en-mass, but it looks like I may need to find a bot operator to help. VanIsaacWS 02:28, 3 August 2011 (UTC)
There is a move option, but if I remember right all it does is fill in the reason for you. You still have to move each individually, just as with edits, and I don't think it will even automate script → alphabet in the name. — kwami (talk) 02:42, 3 August 2011 (UTC)

I like to see your listings. I am not happy with Greek preference over Latin. What do you think to always use "X (Y letter)" in case there is "X (letter)" anyway?

Regarding "Brahmic scripts (or perhaps Brahmic alphabets)" - the latter would collect the instantiations the former the scripts, so this would not be synonymous. Bogdan Nagachop (talk) 21:42, 3 August 2011 (UTC)

The Greek preference over Latin is simply because if you are looking for the letter lambda or alpha, it's probably the Greek version you want, not the Latin one. None of the "regular" Latin letters have names that conflict with Greek letters, so it's basically a rule that X (letter) should always go to the regular Latin or Greek letter. VanIsaacWS 22:21, 3 August 2011 (UTC)
Correct, 'Brahmic alphabets' would not be synonymous. It's been a while since I've looked at that article.
I'm undecided about Nagachop's letter suggestions. I think Ng (Arabic letter) is superior, because otherwise we wouldn't know we're talking about a letter: an "ng" could be a relative of the "oud" or something, for those who know nothing of Arabic phonology. In other cases the greater dab would be unnecessary, though perhaps still helpful.
Given all the details involved in letter names, that is IMO really a second topic. And numeral systems a third. I don't think we need consensus on those in order to move forward with the scripts. If consensus is easy it's probably not necessary, and if it's difficult we are probably creating problems for ourselves later, since we'll almost certainly overlook something vital. We could always add a section on glyphs and on numerals if we find reason later on. — kwami (talk) 22:58, 3 August 2011 (UTC)
Ng (Arabic) could be a word like Abd (Arabic). But if the article is about a character of the script, then better not let the people think it has to do with the language. Bogdan Nagachop (talk) 03:23, 4 August 2011 (UTC)
Agreed. — kwami (talk) 22:50, 4 August 2011 (UTC)

Honing in on a consensus[edit]

Script names[edit]

I think we are seeing an emerging consensus on "X script" for all of the scripts out there, eg "Latin script", "'Phags Pa script", "Jurchen script"; and "X alphabet" for national alphabets, notation systems, and other symbol collections for specific languages or use, eg "Belarusian Latin Alphabet", "NATO phonetic alphabet", "Russian alphabet" (implied Cyrillic), "American manual alphabet", etc. The plural forms "X scripts" and "X alphabets" should be used for describing a family of scripts, and the collection of script conventions for writing a given language or in a particular medium, eg. "Brahmic scripts", "Semitic scripts", "Belarusian alphabets", "Semaphore alphabets", "Manual alphabets". If anyone objects, or wishes to add any addenda, please add a dot with your contribution. If we have a consensus for a week, I say we implement it. VanIsaacWS 00:37, 4 August 2011 (UTC)

Are you proposing we return to 'Phags-pa script because it's multilingual? — kwami (talk) 03:50, 4 August 2011 (UTC)
No. I am proposing 'Phags Pa script because it is consistent with all the other script names. Naming things based on the number of different languages that use a particular script seems like madness from my perspective. 'Phags Pa is a distinct set of glyphs that are used together to represent language in written form. Hence, it is a script. If, on the other hand, we were writing an article about how a distinct set of glyphs are used together to represent a specific language, then it would be, eg, the Mongolian 'Phags Pa alphabet. VanIsaacWS 08:23, 4 August 2011 (UTC)
PS, I am sure there are many script articles that will contain the information on how a script was used to write the major language it was used for. Just because a script article does not yet merit a content fork for the language representation aspect does not override the naming convention for scripts. VanIsaacWS
  • comment 1 Undecided for the alphabet names, but script names seem fine. One set of scripts one finds at ISO 15924#List of codes which ideally matches the items in Category:Scripts with ISO 15924 four-letter codes. There are also pages like the ones in Category:Unicode blocks. Arabic alphabet has some blocks inside the article. What will happen with Template:List of writing systems - Under alphabet it has Cyrillic, IPA, Latin. Cyrillic per the new definition is not an alphabet. IPA is part of the Latin script. BTW, NATO phonetic alphabet, is actually only used by the NATO and it is not phonetic. Seems like a double error. All in all, for the individual scripts we have a clear standard that we would use as basic reference, namely ISO 15924, correct? Fraktur being one of the deviations. I am already curious how the result of the reorganization will look like. Bogdan Nagachop (talk) 03:11, 4 August 2011 (UTC)
None of this has to do with the classification of scripts. There is no new redifining, there is no effort to change content. The Latin script is still alphabetic, as is Cyrillic. We are only trying to hone in on a consistent nomenclature for script articles that allows for reliable navigation and proper disambiguation, not overturn the technical classification of writing systems. I'm sorry that the NATO phonetic alphabet is named that way, but they call the "Alpha, Bravo, Charlie, Delta, Echo..." systems phonetic alphabets for some damnable reason - and it's not just the NATO version. The templates will all be updated with the new links, but I don't think that any content should be changed without a separate discussion at template talk. Like I said, this is only about the standardization of article names. Unicode blocks have standardized names that should not be altered, unless a naming convention along the lines of X (Unicode block) were to be decided. VanIsaacWS 03:40, 4 August 2011 (UTC)
(edit conflict) The NATO Phonetic Alphabet is not a script, so wouldn't be covered by our convention. Yes, it's a misnomer, but a pretty entrenched one.
Unfortunately, we do need to handle them, because semaphore systems, telegraph codes, braille representations, "phonetic" alphabets and the like are all under the purview of the Writing systems WikiProject. Whether we like it or not, we are the keepers of those articles.
IPA isn't really Latin. Part is Greek (ɸβθɣχʋʎɛɑ), and a bit is even Arabic (ʕ, though filtered through Latin). I think it would count as a distinct script. Or semi-distinct, anyway.
This comes up every once in a while, but I think there are very important reasons to treat IPA as a Latin alphabet, not the least of which is that it is based on the Latin typographic tradition, not Greek or Arabic.
Also, we need to consider what to do when s.t. is both an alphabet and a script. IMO we should go with 'alphabet'; it's not clear to me if, say, Kannada–Telugu, Mon–Burmese and Thai–Lao would be the scripts, with Kannada, Telugu, Mon, Burmese, Thai, and Lao being instantiations of those. Calling national forms 'alphabets' would avoid the ambiguity, and also cut down on use of the ambiguous term 'script'. — kwami (talk) 03:50, 4 August 2011 (UTC)
I think we need to stick with script unless the article specifically deals only with the use of a given script for writing a particular language. Otherwise, we are moving essential content around when we discover that another language uses a particular script, and that can't be good. I believe that except for the large ideo/morpho/logo-graphic sets, basically every writing system in ISO-15924 should have an article name ending in "script". There may be several instances where X alphabet redirects to a section of a script article that deals with how a script is used to represent a given language, but that's ok by me. VanIsaacWS 08:39, 4 August 2011 (UTC)
So we'd need to move Greek alphabetGreek script and Hebrew alphabetHebrew script, as well as Armenian, Coptic, Estrangelo, Gaelic Latin, and Phoenician. — kwami (talk) 08:48, 4 August 2011 (UTC)
That would be my preference. The article "Greek script" would contain information, not just on the modern script, but to all of the historical forms (polytonics, archaic letters, historic Doric and Ionian forms). Then you could have two other articles "Greek alphabet" and "Greek polytonic alphabet" about how the Greek script is used to represent modern and ancient forms of Greek, respectively, as well as a possible "Greek Linear alphabet" for the use of Linear B to write pre-Macedonian Greek, if such content ever needed to be forked from the "Linear B script" article. The same for Hebrew, Armenian, Coptic (this may be an edge case whether to unify with Greek script), Estrangelo, and Phoenician. Gaelic Latin, refering to the insular type, should probably take the nomenclature of a "hand", rather than a script, whatever standard we settle on for those stylistic variants. VanIsaacWS 10:22, 4 August 2011 (UTC)
Well sure, if we have separate articles for the script and alphabet, they should be called just that. But generally we won't. We'll have one article that combines the two. Our conventions should keep that in mind. Since the emphasis tends to be on the alphabet itself, IMO that should be the title. Cf. Thai alphabet, which is about the script as it's used in Thai, with highly language-specific tone rules; Georgian alphabet, which is primarily its application to Georgian. If articles for the other alphabets of the Georgian script are ever written, the current article is likely to be kept for the Georgian alphabet. Of course, the article could be rewritten, with Georgian alphabet as just one section, but as you've noted, our goal here is not to rewrite the articles, but to name them as they are currently written. — kwami (talk) 10:38, 4 August 2011 (UTC)
I guess I consider the script to be the most important level of the hierarchy. It's the presence of all of the technical stuff about code pages, letter-form evolution, and typographic traditions that seems to be the heart of these articles to me. How a language is represented by them seems almost parenthetical to me. VanIsaacWS 11:11, 4 August 2011 (UTC)

──────────────────────────────────────────────────────────────────────────────────────────────────── I am for splitting articles, so topics are clear. I also think it is the most important level. A script is a collection of characters. Then, what we call alphabets, deploy some or all of the characters and associate them with a more restricted pronunciation. E.g. in the Mapudungun alphabet, h is used for . But most of the Mapu letters follow common variation. But this all is a little bit floating, e.g. the shapes Р/Ρ/P on my screen are exactly the same, but belong to Cyrillic/Greek/Latin. Maybe the script articles can include possible pronunciations of the characters in the future. Vanisaac also mentioned Coptic vs Greek. I think each script that has an entry in ISO 15924 should get a script article. So we refer to an official ISO standard and can then avoid debates on how we think the stuff should be arranged. So even Fraktur would get a script page. But the article can state that x, y, z sources regard it as a variant of Latin. Same for Coptic. Bogdan Nagachop (talk) 17:00, 4 August 2011 (UTC)

(@ VI & Nagachop) But that's not the business of naming conventions to decide. If you wish to split the articles, or create new articles, you are of course welcome to do so (assuming the results are sufficiently well developed to satisfy other editors), but we shouldn't tell people to do so here. Most of the articles concern themselves almost entirely with the alphabet, with the script elements being parenthetical, often limited to a note in the intro that it's used to write local minority languages, but without going into any detail, or even clarifying if those minority languages actually have their own alphabets. (Sometimes languages are simply written ad hoc in the majority alphabet, without any formal distinction.) That may not be the proper weight for the articles in your opinion, but the solution is to actually expand or split the article, not to pontificate in the naming conventions. Our job is to advise people on how to name articles as they are currently written, not on how to write the articles or what they should cover. That belongs in the MOS.
Also, the ISO standard is hardly the end-all of philological research. The World's Writing Systems, for example, treats Kannada and Telugu together, as it does Thai and Lao. There are numerous minority Tai alphabets in the same family, arguably a single script, and historically we speak of the Kannada-Telugu script as well. They look somewhat different, but no more so than the various hands of Latin, Arabic, or Chinese. I don't think we should advocate blindly following some bureaucratic convention when facts on the ground may indicate otherwise. Although the ISO may be a convenient starting point for scripts, just as it is for languages, IMO we shouldn't follow it just because it's the ISO, any more than we do for languages. Take a look at our language articles: some have no ISO code, and some combine several ISO codes together, because that's what the lit supports.
BTW, I just noticed that Ethnologue uses the phrase 'Latin script' when describing the writing systems of small languages which have their own Latin alphabets, exactly what we're proposing here. — kwami (talk) 22:47, 4 August 2011 (UTC)
I never said that you have to content fork those X alphabet contents, only that we could do that, and that I believe the naming convention of X script allows the most leeway. For instance, let's pretend that we've named the article on the Greek script "Greek alphabet", and it contains the combined contents of a Greek script and Greek alphabet pages. Now someone wants to add information on a hypothetical Turkish Greek alphabet to the page. By having the page named "Greek alphabet", the *Turkish Greek alphabet doesn't actually fit there, even though it logically belongs there. If, on the other hand, it was named "Greek script", then you can have any number of national alphabets that use that script listed and described on that page, without any conflict between the article name and the contents. That's the big quesiton. Are we going to have to dynamically rename pages as content is added? Naming all scripts, by default, as X script keeps us from having to do that. VanIsaacWS 23:37, 4 August 2011 (UTC)
Malayalam alphabet would go back to Malayalam script? Category:Brahmic scripts maybe needs clean up. I already removed Category:Philippine scripts and Category:Indonesian scripts since they are not all Brahmic. @ISO 15924 relevance - their groupings of characters may end up in Unicode/ISO/IEC 10646 and then be exposed to lot of people. So I think to have individual articles might be fine, even if they only state that the definition is not supported by many people. @MOS vs NC - yes, some of my thoughts related to MOS. But the article name relates to the content, so some mentioning of article content could be ok. But it doesn't matter much. There is a lot of clean up work waiting. I created WP:NCWS. Coro, Kwami, Vanisaac and me, we are now already 4 editors supporting "X script" for the basic articles. Several X alphabets have only been moved by Kwami last WE, so moving them back to X script should be fine? I would like to clean up the Brahmic stuff. @Vanisaac, I think we could also create new articles for Latin script, Arabic script. Bogdan Nagachop (talk) 23:43, 4 August 2011 (UTC)
I have no problem with creating article forks if necessary, but it should be done because we are going to be making a distinct content choice regarding the separate articles, not just to have articles that correspond to our hierarchy. I just think there is a distinct difference between general script properties like encoding, glyph evolution, glyph variants, writing styles, etc., and the use of a script to represent a given language. The fact that these are contained in the same article in many instances just means we need to figure out our default name, and which content should be forked out of the original article. I happen to think that the "script" articles are the prefect breeding ground for "alphabet" content, which can then be forked into separate "alphabet" articles when that content, by itself, is beyond, say, start class. Does that seem like a reasonable perspective?
Yeah, I think category:Philipine scripts is pretty limited - I can think of three, and I don't even know if they have separate articles. I know there are separate Unicode encodings, but I tend to think of them collectively as simple typographic variants of each other. I think I'm getting off topic. I'd definitely open a thread on Wikipedia talk:WikiProject Writing systems to discuss it, though. Just in case people are using that category for something. VanIsaacWS 04:24, 5 August 2011 (UTC)
Your comment on Philippine scripts is what I was thinking. Is there just one, with various hands developed into alphabets by different nations? I don't know, but that would affect our naming.
I have the opposite view on the default name. If 90% of the material in the Georgian X article is on the Georgian alphabet, then the article should be called 'Georgian alphabet'. Other Georgian-script-based alphabets will most likely be given separate articles anyway, not lumped in with the Georgian alphabet. That, at least, is AFAIK how our articles generally develop. We shouldn't name an article based on what we judge it might become, but on what it is. — kwami (talk) 06:07, 5 August 2011 (UTC)
No no, I don't think this is about what an article might become, but whether its name actually reflects both the present content and welcomes the content that rightly belongs there. Right now, when I look at "Greek alphabet", its content is almost entirely about the Greek script. There's a separate article, "Greek orthography" that I think is what we are looking for with our Greek alphabet page. Same thing with the "Georgian alphabet". Basically none of that article is on how the Georgian language is represented by the Georgian script. It has the development of the majescule, ecclesiastical, and modern miniscule cases. It discusses ligatures and obsolete letters, transliteration, and Unicode encoding. Of those sections, transliteration is the only one that may belong in a "Georgian alphabet" article. The rest is on the script, regardless of language. Note, however, that the Greek script article has a brief introduction to minority and historic langauges that used the Greek script. That's exactly the kind of content that our script pages should have, but an article named X alphabet is much more hostile to that content than X script. I think that's a fact worth considering. VanIsaacWS 10:03, 5 August 2011 (UTC)

──────────────────────────────────────────────────────────────────────────────────────────────────── Sure it's worth considering. But if it's background info for the intro, then that's not the topic of the article and therefore not the proper name. Any alphabet article is likely to have some background info on the script, esp. if the script is predominantly a single alphabet, and I don't see how calling in an 'alphabet' would make the article hostile to such info, though if expanded enough we'd prob'ly want to split it off. But these are the kinds of article-specific decisions that should be made on each article's talk page; they have little to do with us. — kwami (talk) 12:56, 5 August 2011 (UTC)

I think a test implementation could help. I started with Arabic script, the table there (Arabic script#Table) lists Nastaliq, Shahmukhi and Jawi. These are scripts by themselves? Then there would be Arabic scripts as a group. What about Perso-Arabic. If there is at least on Arabic Latin alphabet (naming like in Uyghur Latin alphabet) then there would be several Arabic alphabets (Arabic the language). On the other hand there are already several Arabic alphabets listed, like Belarusian Arabic alphabet. Which category is higher up, "Arabic script" or "Arabic scripts"?Bogdan Nagachop (talk) 17:31, 5 August 2011 (UTC)
I think we need to go with VI's idea of default script/alphabet. Without a dab, "Arabic alphabet" can only mean the Arabic script. Any other Arabic alphabet would need to be specified (Arabic Cyrillic?), but that one does not. That's basic application of COMMONNAME.
I'd say "Arabic script", sg. Whether Jawi is considered a script or not, the Arabic script is still conceptually a single entity. We shouldn't move our articles between sg. and pl. every time s.o. argues that some instantiation is a script in its own right. That would simply create chaos, with no benefit that I can see.
I don't know about Jawi being a script. Maybe 'Jawi variant of the Arabic script'? But we could have 'Jawi script' as a subcat of 'Arabic script', why not? 'Song birds' are a subcat of 'perching birds'. But AFAIK Jawi is just an alphabet. — kwami (talk) 07:32, 6 August 2011 (UTC)
Jawi article says it is also used in the Philippines, that may mean for another language than Malay. I hope we can cleanup Category:Scripts with ISO 15924 four-letter codes, some thoughts at Template talk:Infobox writing system#Automatic categories by ISO category. Bogdan Nagachop (talk) 01:59, 7 August 2011 (UTC)


At this point, it appears we do not yet have a consensus on some of the more peripheral matters, such as letter names, so let's keep working on that separately. VanIsaacWS 00:37, 4 August 2011 (UTC)

Yeah, I'd just as soon get the scripts done, and do letters, digits, typefaces, etc. later if needed. It may not be needed.
Which version should we work with? Are we each wedded to our own, or can we consolidate them? — kwami (talk) 07:15, 4 August 2011 (UTC)
I altered the NC page [6]. I suggest to keep the observations of July for a while so we don't forget anything. In the top I added what we agreed to and tried to say what a script with respect to that NC can be. This ISO reference so far is only a proposal by me. Bogdan Nagachop (talk) 17:19, 4 August 2011 (UTC)


Just for future discussion:

  • Pe (letter) - about the Semitic letter -> Pe (Semitic letter)
  • Pe (Cyrillic) -> Pe (Cyrillic letter), to make clear it is not about any non-letter concept in the Cyrillic script
  • Pe (Persian) -> Pe (Persian letter)

Is Pe (Arabic letter) or Pe (Arabic script letter) ambiguous? Bogdan Nagachop (talk) 23:04, 6 August 2011 (UTC)


The following discussion is closed. Please do not modify it. Subsequent comments should be made in a new section. A summary of the conclusions reached follows.
(moved to project page; propose revisions in a new section)

Articles on writing systems typically consist of a proper or other identifying name combined with a broad typological classification of the script, such as the following:


The term 'script' is used with four meanings:

  1. A general segmental writing system, as opposed to its local instantiations:
  2. In the plural for a family or geographic group of such scripts. 'Writing systems' conveys the same idea:
  3. A calligraphic style. In many cases 'hand' may be used instead, and may this avoid confusion with other uses of the term 'script':
  4. A non-segmental writing system, especially one which is logographic, mixed, or of unknown character:

'Alphabet' is used for national instantiations of a segmental script, usually with a defined sorting order:

The terms abjad and abugida, though often used in the text, are considered jargon and inappropriate for a title.

If an article conflates a script and the dominant alphabet using that script, as Georgian alphabet does, then the name should reflect the preponderance of the coverage of the article: 'Script' if it details the various alphabets that use the script, but 'alphabet' if it's largely concerned with the orthography of the dominant alphabet. These topics may of course be developed sufficiently to split into separate script and alphabet articles.


Exceptions may be made where a unambiguous conventional name exists, and thus 'script', 'alphabet', or 'syllabary' is not necessary:

Note: "hieroglyphics" is deprecated.

Exceptions may also occur where a different technical term is widely used:


Modifiers may be used for subtypes or other cases of disambiguation:


Names should include some indication that the article concerns a glyph and not a word.

The above discussion is preserved as an archive. Please do not modify it. Subsequent comments should be made in a new section.


Few recent comments above, so I'm moving my suggestions, trimmed a bit, down as a draught, and adding a pared-down version of the glyphs. We shouldn't over-dab, I think. That would go against our general NC's. — kwami (talk) 23:59, 15 August 2011 (UTC)

Looks good to me. VanIsaacWS 02:06, 16 August 2011 (UTC)
First quick remark: OK in general line. Fraktur leads to a dab page. Would like to cross-check with ISO and Unicode usage (Cyrillic cleaned up nicely), but not now. Unfortunately, the ISO 15924 templates are deleted yesterday. -DePiep (talk) 14:24, 16 August 2011 (UTC)

Just added zeta so people don't try adding '(letter)' to every single article using this page as justification. (Actually, we might want that, but it would be a new convention and could be discussed here in the future.)

I think Fraktur is covered by WP:primary topic; the other article can be handled with a hat note.

What do you mean by "Cyrillic cleaned up nicely"? — kwami (talk) 14:28, 16 August 2011 (UTC)

At Cyrillic [7], I concluded: no dab involved, not "Cyrillic language" involved, and both ISO 15924 and Unicode alias names are without any specifier. This might be a rare exception in the list of writing systems. -DePiep (talk) 20:31, 16 August 2011 (UTC)

I've moved the Perso-Arabic article to Persian alphabet, since it deals primarily with just that. But many of the incoming links deal with Pashto or other Persian-derived alphabets, so perhaps we should split off a Perso-Arabic script article? — kwami (talk) 17:27, 16 August 2011 (UTC)

The article seems to deal with the Persian alphabet and has links to many of the other alphabets that are Perso-Arabic based. If the treatment of some of the minority alphabets starts to get large enough, I'm sure we can branch the article, but I don't think this article will ever naturally fork the "Persian alphabet" content. That says to me that it's probably the right name. VanIsaacWS 17:41, 16 August 2011 (UTC)

Okay, I've moved some articles and categories in accordance with this, to see if there would be any opposition since not many people watch the NC pages, and the only potential issue was with Cyrillic 'script'. Not clear if it was an actual objection or just a desire for discussion, but I moved our example down to plain 'Cyrillic'. Moved draught to WP space and linked from the main NC page. Closing the draught above so that we can keep track of where we came from; we can always open it again below if need be. — kwami (talk) 18:48, 16 August 2011 (UTC)

  • re "'Alphabet' is used for national instantiations" -- Can we make that like Alphabet is language specific. That might be closer to the situation. -DePiep (talk) 20:35, 16 August 2011 (UTC)
  • At the moment, most of the {{ISO 15924 templates}} and Unicode related script lists are spoiled. I cannot check names. -DePiep (talk) 20:56, 16 August 2011 (UTC)
Yes, 'national' wasn't right. But sometimes languages share alphabets, don't they? How about 'specific applications'? Currently also mentions individual lanugages. — kwami (talk) 21:07, 16 August 2011 (UTC)
If an alphabet is specific to more than one language then it’s still language-specific. The point is that scripts are not, they’re language-agnostic. — Christoph Päper 22:40, 16 August 2011 (UTC)
How's that? — kwami (talk) 23:44, 16 August 2011 (UTC)
If I understand Crissov well, the description should cover the plural: Alphabet is specific to one or more languages. Still, in the singular phrasing, to me it reads like covered as well. Somehow, the singular one does not exclude the plural (it does not exclude another language). If I am correct in this, we can use the simpler option. Either way: "language" is preferred to "nation". -DePiep (talk) 08:08, 17 August 2011 (UTC)

I really like the note about the Georgian alphabet vs. Georgian script part. I think it captures the balancing act that I have been advocating far better than I have ever phrased it. VanIsaacWS 15:08, 17 August 2011 (UTC)

Guideline status[edit]

Could someone point me the way of any discussions held to whether this should be marked as a guideline? Grandiose (me, talk, contribs) 19:57, 17 August 2011 (UTC)

What, exactly, would that do? VanIsaacWS 20:40, 17 August 2011 (UTC)
"New proposals require discussion and a high level of consensus from the entire community for promotion to guideline or policy"; I realise this is a naming convention but it basically fits under guideline (at least the notification bot thinks so). I just wanted to check whether this had been met, or whether it needs to be discussed now. Grandiose (me, talk, contribs) 21:13, 17 August 2011 (UTC)
If you think it's necessary, then do it, but this isn't an isolated discussion. It actually started over on WikiProject Writing Systems, and was moved over here to facilitate the discussion and pull all the activity off the WikiProject talk page. It really is only a convention that is being agreed to by the major contributors to Writing Systems articles. I certainly don't have any intention of treating this as a policy that would be brought to an RfC or anything like that. It's merely a convention that we are going to implement because we have some VERY inconsistent naming in the project right now, and we need to get some sort of understanding about where we want to go in the future. I can't speak for anyone else, but I don't intend to treat it as a policy, I intend to treat it as a convention. VanIsaacWS 21:46, 17 August 2011 (UTC)