Template talk:DisplayTranslations

From Wikipedia, the free encyclopedia

To preserve issues that were raised on the original page for this effort I am copying comments here and may refactor the thread if I can find time.Buddhipriya 21:48, 22 February 2007 (UTC)

Contents 1 Non-English languages 2 Yes, IAST doesn't work 3 Replies to most recent questions 4 Brilliant work so far! 5 Some confusion in the documentation about the role of Unicode versus IAST 6 The Sanskit page may be renamed 7 Brain twister 8 Case study: Used to display the term skandha 8.1 Variant 1: Omit the Devanagari row 8.2 Variant 2: Use IAST transliteration method 8.3 Variant 3: Show Devanagari and IAST versions together 9 Replies to most recent questions

[edit] Non-English languages

I would like to open a discussion on the addition of non-English languages to Buddhist terms. My point is that all Buddhist articles are becoming illegible because many people are currently adding non-English words even in non-Roman script in the texts: I can understand the reason for adding Pali and Sanskrit terminology (in translitteration, not in Devanagiri etc.), as these were the original languages of the scriptures, but people are now adding Chinese, Tibetan, Japanese, Sinhalese, Korean to name but a few. Should we not limit all this illegible and totally superfluous stuff; for example, the Japanese terms should be placed in the Japanese version of Wikipedia and not here, unless the terms are typically of Japanese origin such as Zendo etc. Otherwise we could go on and translate every Buddhist term here in Dutch, German and French as well....??? rudy 17:21, 18 February 2007 (UTC)

rudy, there seems to be a very good reason for people adding terminology in different languages. I did that myself many times with Tibetan terms. I also feel good when I am able to find and add Sanskrit original term that was translated into Tibetan.

I know some Tibetan terminology and from that knowledge: very very often regular translators who know Tibetan much much better than me, do not, I repeat, do not know the correct meaning of a particular term. Did I mention that it happens often? Also there seems to be a lot of meaning in etymology of Tibetan terms. People who devised the terminology, who transated the original Sanskrit, knew what they were talking about. --Klimov 19:17, 19 February 2007 (UTC)

I certainly acknowledge the reason for adding terms in the 'original' languages of the scriptures: Pali and Sanskrit, and to a lesser extent Chinese and Tibetan, but what's the use of any original script in an English text? Even most scholars of Sanskrit barely use Devanagiri, but use the transliteration, just like the Wylie transliteration makes a 'legible' Tibetan etc. For example on the Dharma page I read:

"Dharma (Sanskrit: धर्म) or Dhamma (Pāli: धम्म) in Buddhism has two primary meanings:"

"In East Asia, the character for Dharma is 法, pronounced fǎ in Mandarin and hō in Japanese. The Tibetan translation of this term is chos (Tibetan: ཆོས་; Lhasa dialect IPA: [tɕǿʔ])."

Similarly, the first sntence of the Three Jewels page:

"The Three Jewels, also rendered as Three Treasures, Three Refuges or Triple Gem (Sanskrit: Triratna त्रिरत्न, also Ratna-traya रत्नत्रय ; Pali: Tiratana तिरतन , Tisarana तिसरन; Chinese: 三寶 or 三宝, Sānbǎo; Japanese: Sambō or Sampō; Sinhalese: Teruwan තෙරුවන්)"

Where is the end to adding original scripts, dialects and pronunciations, should we add Thai, Korean, Burmese, Vietnamese, Mongolian, and what have you in their original script and with pronunciation?rudy 23:06, 19 February 2007 (UTC)

Just discovered on the page Bodhisattva:

"In Buddhist thought, a bodhisattva (IPA pronunciation: [ˌbɑ dɪ ˈsæt və]) (Pali: bodhisatta; Tibetan changchub sempa (byang-chub sems-dpa');Simplified Chinese: 菩萨; Traditional Chinese: 菩薩; pinyin: púsà; Korean: 보살 bosal ; Japanese: 菩薩 bosatsu; Vietnamese: Bồ Tát; Thai: พระโพธิสัตว์) "

Is this how we want to start an article?rudy 23:09, 19 February 2007 (UTC)

My personal favorite is the lead for Tanha:

Taṇhā (Pāli: तना) or Tṛṣṇā (Sanskrit: त्र्स्ना) means "thirst, desire, craving, wanting, longing, yearning".

Which, if you don't have a Devanagiri-enabled browser (like the one I'm currently using), looks like:

Ta_hā (Pāli: ___) or T___ā (Sanskrit: ________) means "thirst, desire, craving, wanting, longing, yearning".

Not very enticing for a general WP reader.

Worse still, many of the devanagari versions are completely wrong -- as the two examples above. I would delete the lot of them. And why give devanagari equivalents for Pali ? --Stephen Hodge 02:23, 20 February 2007 (UTC)

translations
Pali:	dhamma
Sanskrit:	dharma
Chinese:	法
Tibetan:	ཆོས་
v • d • e

I'd like to suggest creating a sidebar template that allows people to put in the various translations along with standard links to something somewhere which will inform WP readers how to install the font needed to view the associated translation. It could be something relatively simple like the one to the right here. (Perhaps the links to the appropriate code pages should be on the referenced language-specific pages?)

Then, within the article text itself we could use the coding recommended by rudy perhaps. Would this meet both needs? Larry Rosenfeld (talk) 23:56, 19 February 2007 (UTC)

This may have come up before, but in case it has not, here is a LANG template that shows up often for Sanskrit: ([[Sanskrit]]:{{lang|sa|गणेश}}; {{IAST|gaņeśa}}). Regarding transliteration methods for Devanāgarī I hope we can agree on IAST if this has not already been settled. Some on this talk page have questioned why this matters, but for those who work with the Sanskrit materials the issue is important. Also, it is good to differentiate Sanskrit (a language) from Devanāgarī (a writing system). Sanskrit can be written in various writing systems such as Bonji, Gujarati, IAST, etc., in addition to Devanāgarī. And Devanāgarī is used to write multiple modern Indic languages. Confusion often arises on other pages when Sanskrit terms are used due to differing transliteration methods. Buddhipriya 00:30, 20 February 2007 (UTC)

See also the earlier discussion in aspects of this topic above under the heading "No need for BOTH Sanskrit and Pali transcriptions, especially for words like 'sutra' and 'nirvana'".--Stephen Hodge 02:34, 20 February 2007 (UTC)

Thank you for drawing my attention to the topic above. I am wondering if shifting some of the detail to footnotes may help with handling some of the terms. In a footnote one could show it in Klingon if you want to. For the terms that are key technical terms, others have already noted that those terms (e.g. dharma) will often be articles in their own right, where sidebars can be used for language variations. Regarding Sanskrit and Pali, would it be true that most of the Sanskritists working with these materials would be likely to have a copy of Edgerton's Buddhist Hybrid Sanskrit dictionary on their reference shelf? I just looked at how that is set up and he seems to have the Sanskrit and Pali variants pretty well covered. So the specialists do not need to see both versions, unless there perhaps is a key difference that is not obvious. From the average reader's perspective I am guessing that they are most likely to see simplified English transliterations of Sanskrit terms (using only the Latin-1 character set) in most of what they are reading, OR they may see IAST if they are reading an academic transliteration. I do not think you need both IAST and Devanāgarī *if* there is agreement to use IAST as a standard for transliteration, and the IAST LANG template is used. I do not know Chinese but I think it is kind of fun to see the characters because they remind me that these concepts may be subtle and hard to pin down. Buddhipriya 03:19, 20 February 2007 (UTC)

Here is a small tweak to the suggested sidebar box code to use the IAST template for Sanskrit. It makes clear which of the various transliteration systems are being used. The code change is: {{IAST|''dharma''}} and it looks like this. Buddhipriya 03:28, 20 February 2007 (UTC)

translations
Sanskrit:	dharma
v • d • e

I understand your desire for the IAST template and, if I were to create this hypothesized template, I'd certainly keep in mind to include it (only for Sanskrit?). I'd like to discuss designing the new template in terms of two areas: the hypothesized template's name and its invocation within WP articles.

In terms of its name, I'm torn between making a generic template (with perhaps gray, white, black colors) or a Buddhist-specific template (for instance, using the color scheme already demonstrated here). If the former, would a title such as "TranslationList" be good enough. If the latter, how about something like "BuddhistTranslations."

Secondly, in regards to invocation, I'm tempted to make something generic and infinitely expandable, so that it could be invoked like this:

{{BuddhistTranslations| Pali="khandha" | Sanskrit="skandha" | Klingon="kro-lak" | English="aggregate,<br>heap" }}

However, if I'm gonna code this, then I personally have a couple of problems/limitations with such including:

• I don't know enough about the wiki-coding language to know if I could parse and process an infinitely long parameter list (i.e., does wiki-code have a DO-WHILE feature?)
• I don't know enough about the wiki-Parser to be sure that it would provide the wiki-code both an infinite-varieties of parameters (such as, "Klingon") and their associated values ("kro-lak")
• I don't konw if the wiki-Parser would allow the article to be associated with the identified language. For instance, if we had a translation into "English," we would want to pipe this to English_language, not necessarily English.

Soooo, a simpler template to code would be one where we knew ahead of time the allowable languages to be represented in the table, such as Pali, Sanskrit, Mandarin, Japanese, Tibetan, English, etc. (I'd be happy to pursue a sample of existing WP Buddhism articles to create the list and additional languages would be easily added to the template as needed.) This would also allow for a standard ordering of the translation (e.g., based on chronology, though I would need help from you all to tell me what that order should be!). If a parameter (such as, Pali="khandha") is not provided, then it would not be shown in the transcluded sidebar. I could probably code this in one lunch period.

In summary, my questions are:

• What title should we use for the template?
• Should it be generic or Buddhism-specific?
• Should it allow an infinite variety of languages or just a fixed (e.g., Buddhism-specific) set? (If the former, it might take me a while to figure out how to do it. [Anyone else of course should feel free to pick up this baton :-) ])
• If Buddhism-specific, what order should the languages be presented in?

What y'all think? Thanks, Larry Rosenfeld (talk) 16:16, 20 February 2007 (UTC)

Thank you so much for your service on this, by the way. I am deeply ignorant of Wiki coding methods and am unable to help much on that. I am wondering if the guidance of some Wikicoding Bodhisattva can be invoked? Perhaps someone has already got a template that does this, or which can be sightly modified. I would keep the title simple, such as "Translations". I would make it generic. If you figure this out it probably could be used elsewhere. I would allow it to take any additional language through the addition of a LANG parm, the same way pages get LANG indexed in unexpected ways. At the same time, note that it probably should not be used to replace or re-invent the existing LANG indexing methods for page titles, which are invisible to the naked eye. In cases where the page is for a technical term, such as dharma various Wikibots will do their own magic to provide unexpected translations such as those that are embedded now at the bottom of the Dharma page:

[[ko:법보]] [[id:Dharma]] [[it:Dharma]] [[he:דהרמה]] [[lt:Dharma]] [[nl:Dharma]] [[ja:法 (仏教)]] [[no:Dharma]] [[pl:Dharma]] [[pt:Dharma]] [[ru:Дхарма]] [[sa:धर्म]] [[simple:Dharma]] [[sk:Dharma]] [[fi:Dharma]] [[sv:Dharma]] [[th:พุทธธรรม]] [[vi:Pháp (Phật giáo)]]

Buddhipriya 17:48, 20 February 2007 (UTC)

I just noticed that those bot-provided LANG tags make the ususal identification of the LANG "sa" with the writing system "Devanāgarī". e.g., the bot wrote: sa:धर्म. Just a pet peeve. Buddhipriya 17:56, 20 February 2007 (UTC)

Thanks for the quick feedback. And I appreciate your points. I'll definitely keep this in mind as a "next thing to work on" (after I'm done cobbling together in a sandbox a half-baked Buddhism and psychology article). Good point too that the WP processor somehow takes [[AA:XXX]] and translates it into http://AA.wikipedia.org/wiki/XXX. (I wonder how it knows to parse the string in this manner -- perhaps the two-letters followed by a colon [":"] signals the processor?) I don't know if the code that does this parsing and concatenation function is exposed to general WP editors, but I'll try to find out. (Might you or anyone else be able to point me in the right direction?) Thanks again! Larry Rosenfeld (talk) 18:14, 20 February 2007 (UTC)

Translations
Pali :	dhamma
Sanskrit :	dharma
English :	law, teaching, mental object
Mandarin :	法
Tibetan :	ཆོས་
view • talk • edit

Well, for what it's worth, at User_talk:Larry_Rosenfeld/sandbox3 I created a hideous little template that can take up to 12 parameters (that is, six language-text pairs) and displays a trivial "translation" sidebar. The transcluded sidebar to the right was invoked with:

{{User_talk:Larry_Rosenfeld/sandbox3 | Pali | ''dhamma'' | Sanskrit | ''dharma'' | English | action | Mandarin | 法 | Tibetan | ཆོས་}}

Obviously, if this was a template as we discussed it would be invoked as something like:

{{Translations | Pali | ''dhamma'' | Sanskrit | ''dharma'' | English | '''law, teaching,<br>mental object''' | Mandarin | 法 | Tibetan | ཆོས་}}

This is just spit on the wall. Is it really what we want? Any suggestions for improvement? Anyone know how to insert a "do while" or "goto" into the template to allow an infinite number of parameters? (Also, note that for simplicity the template code adds IAST to everything -- otherwise, the template code would have to compare the first parameter of a pair to "Sanskrit" -- six times. Let me know if this is unacceptable.)

Any feedback is welcome. (Though Sartre might say that even no feedback is feedback, oui?) Larry Rosenfeld (talk) 00:24, 21 February 2007 (UTC)

It's Chinese, not Mandarin ! Mandarin is a (rather dated) term for a spoken form of Chinese (= putonghua, I think). When using Chinese characters for Buddhist terms, without indicating the pronunciation, it might be best to use the term "CJK" (Chinese-Japanese-Korean) if the traditional form of the character in question is used. If a simplified form is used, then that form tends to be specific to Chinese -- the Japanese simplified characters tend to be different from their Chinese counterparts. I'm not sure about Korean usage.--Stephen Hodge 00:52, 21 February 2007 (UTC)

Translations
Esperanto :	Achtung
Chinese :	法
Sinhala :	සිංහල
Klingon :	kro-LAK
PurposefulError :	oops! (my bad)
view • talk • edit

Thanks, as always, for the lesson! If I may, though, the current template is OBLIVIOUS to what the language name is being passed. For instance, it could be coded as:

{{User_talk:Larry_Rosenfeld/sandbox3 | Esperanto | Achtung | Chinese | 法 | Sinhala | සිංහල | Klingon | '''kro-LAK''' | PurposefulError | ''oops!''<br>(my bad) }}

As long as there's an existing [[{{Language}}_language]] article (e.g., Chinese_language). Though, this might pose a problem with CJK as there's no CJK_language article or redirect — although "East Asian" would work as there is an East_Asian_language redirect. Make sense? Need to change? Thanks again & GTG, Larry Rosenfeld (talk) 01:14, 21 February 2007 (UTC)

FWIW, I just created Template:DisplayTranslations. Hope you all don't mind, but each person whose name is in this section was mentioned in the intial edit's Edit Summary. Thanks for your all's help. Hope you all might find this template useful. Please feel free to modify or ask me to try to modify. Thanks, Larry Rosenfeld (talk) 18:54, 21 February 2007 (UTC)

Translations
Sanskrit :	bodhisattva
Pali :	bodhisatta
Tibetan :	changchub sempa (byang-chub sems-dpa')
Chinese :	菩萨 (simplified), 菩薩 (traditional), púsà (pinyin)
Korean :	보살 (bosal)
Japanese :	菩薩 (bosatsu)
Vietnamese :	Bồ Tát
Thai :	พระโพธิสัตว์
view • talk • edit

P.S. Just to give the template a quick try, I used it in the following articles: Skandha and Upadana. To the right is an example of how I think this template might look if inserted in the Bodhisattva article but, in case I got it wrong, I didn't want to insert it into such a high-traffic article that I have not participated in developing. So, I include this here in case anyone else wants to motify or use it. FWIW, this sidebar to the right was generated with:

{{DisplayTranslations | Sanskrit | bodhisattva | Pali | bodhisatta | Tibetan | changchub sempa<br>(byang-chub sems-dpa') | Chinese | 菩萨 (simplified),<br>菩薩 (traditional),<br>púsà (pinyin) | Korean | 보살 (bosal) | Japanese | 菩薩 (bosatsu) | Vietnamese | Bồ Tát | Thai | พระโพธิสัตว์ }}

Another thing I am ignorant about, but would like to less ignorant about, is Tibetan. I do not work with Tibetan and have noticed that on XP systems the procedures outlined by the standard {{IndicText}} tag do not result in correct display of Tibetan because the fonts are apparently not part of the same install. The procedures to install Tibetan Unicode fonts as given in Tibetan language and Tibetan script seem not to reference any pre-installed Microsoft fonts and point only to third-party fonts. Perhaps this is better in Vista, as it is with Indic fonts (in Vista it all is preinstalled). I mention this because I enjoy seeing the funny-looking Tibetan characters which look to me like Klingon. But the point for this thread is that if we are to include Tibetan script, perhaps we can include a good link on how to install Tibetan fonts that addresses the fact that the Indic script procedure does not do the job well. Buddhipriya 18:12, 20 February 2007 (UTC)

Here is a test of the template using a Sanskrit word that requires IAST. The word dharma only requires characters in the Latin-1 charset so it is not the best demo. You may want to set a standard of putting the English term first. There will be disagreement on the English translation for some words. Also, I just looked at the code for the template, and while I do not know anything about Wiki coding language it looks to me that the IAST template is being inserted for languages in addition to Sanskrit, which is not correct. Since IAST is a transliteration method for Devanāgarī it is only applicable for languages that are written using that writing system. Here is the Tibetan line that shows an IAST template:

<tr> <td align=left style="border-bottom:1px solid Peru"> [[Tibetan_language|Tibetan]]:</td> <td align=center style="border-bottom:1px solid Peru"> {{IAST|ཆོས་}}</td></tr>

Buddhipriya 20:02, 21 February 2007 (UTC)

Translations
English :	faith
Sanskrit :	śraddhā
Pali :	bodhisatta
Tibetan :	changchub sempa (byang-chub sems-dpa')
Chinese :	菩萨 (simplified), 菩薩 (traditional), púsà (pinyin)
Korean :	보살 (bosal)
Japanese :	菩薩 (bosatsu)
Vietnamese :	Bồ Tát
Thai :	พระโพธิสัตว์
view • talk • edit

Yeah, this is why I wrote above:

(Also, note that for simplicity the template code adds IAST to everything -- otherwise, the template code would have to compare the first parameter of a pair to "Sanskrit" -- six times. Let me know if this is unacceptable.)

The IAST template -- which is very nice by the way, thanks for alerting me & others to it -- doesn't seem to harm non-Devanagari text, but you're probably right that I should restrict it to Devanagari words. What's the best way to test for this? I guess I could simply test, based on the Devanagari article, to see if Sanskrit, Hindi, Marathi, Sindhi, Bihari, Bhili, Konkani, Bhojpuri, Nepali, Nepal Bhasa, Kashmiri or Romani is the identified language -- though this might be more trouble than it's worth? (Perhaps I could create a macro to return TRUE if the identified language is one of these or FALSE otherwise?) Also, there's the problem if someone identifies the language as "Sanskrit" and then uses non-Devanagari (e.g., Roman) characters to represent the Sanskrit word. Any suggestions -- for instance, are the ASCII values of the Devanagari code set significantly incompatible with non-Devanagari words?

Secondly, regarding English words, I agree that there might be disagreement about what word is best -- or a multiplicity of possible translations; this is why, for instance, on both the Upadana and Skandha pages I offer multiple English words in the sidebar. I'd encourage users doing this. As for making English the standard first translation, I intuitively can understand this desire but think it might be confusing if for instance there is no good English translation (like for Bodhisattva) or if the article title is non-English (like Nirvana), etc. So, at this point, personally, I'd like the template's users to make this decision.

Thanks for your continued help, Larry Rosenfeld (talk) 22:38, 21 February 2007 (UTC)

Sorry I missed your comment in the original post about the fact that all used IAST. I must read these things more closely. The Template:IAST is only intended for use with Sanskrit, so using it with other languages may not break it but it may confuse some poor bot somewhere in the universe who is looking for Sanskrit. The IAST notation has implementations both in ASCII and in Unicode, but those implementations are not inter-operable for characters that are not in the Latin-1 character set. Unicode is designed to be interoperable and transparent for most applications in the Latin-1 range. Unfortunately most of the good stuff in IAST is outside the Latin-1 range. Is that what you asked, or am I just rambling? Buddhipriya 23:02, 21 February 2007 (UTC)

So, for example, if I understand you correctly, if I was to change all the Template:IAST invocations to Template:Unicode invocations, this would not solve the problem of displaying all Devanagari correctly? (I was thinking that this might work for Devanagari as well as have more general applicability for non-Devanagari languages.) Larry Rosenfeld (talk) 23:08, 21 February 2007 (UTC)

Actually, I just ran a test. The following:

IAST: {{IAST|स्कान्धास}}; Unicode: {{Unicode|स्कान्धास}}

results in:

IAST: स्कान्धास; Unicode: स्कान्धास

So, would you see a problem with my changing the IAST invocations in the template to Unicode invocations? Thanks again for your help! Larry Rosenfeld (talk) 23:14, 21 February 2007 (UTC)

We are communicating in real time and I am not caught up to your last post. Correct, it would not solve the problem to assert Unicode. (Yes, we have no bananas today.) There are multiple transliteration systems for Devanāgarī. When you use the IAST template you are asserting that you are using the IAST transliteration method. By using the unicode template you assert that you are using Unicode, but who knows what transliteration method? Wiki is a Unicode-compliant loka, which is good. So in Wikiloka when you use the IAST template by default you are asserting "this is IAST notation in the Unicode glyph set".

If it is any solace to you, I just had to do some reverts earlier today of a well-intentioned person who replaced IAST encoding with Unicode templates. This is also a thorn in the side of the Hinduism project. So you do not suffer alone.

As we said, the ASCII and Unicode implementations of IAST are inter-operable for characters that are not in the Latin-1 character set. Consider the word dharma. That word only uses characters in the Latin-1 range, so that word would be transparent across ASCII and Unicode. But the word śraddhā includes two IAST symbols that are not in the Latin-1 range (ś and ā) and those two symbols have different numerical representations in ASCII implementation and in Unicode implementations. So a person who has only ASCII fonts on their system may understand that it is IAST notation, but may lack the Unicode fonts to display it. What's worse, there are multiple ASCII implementations that are not inter-operable amongs themselves. Best to stick to Unicode (the Wiki default), and mention that you are using IAST. Buddhipriya 23:02, 21 February 2007 (UTC) Buddhipriya 23:33, 21 February 2007 (UTC)

I am trying to think this through with you, so let me parse it for myself as follows. Let me just consider the Sanskrit parm first. Each language I am sure has similar problems (where are the Klingon speakers when you need them?) For Sanskrit, since Wikiloka is Unicode by default, you do not need to specify Unicode if you specify IAST. In the example just given above there is an incorrect use of the IAST template because it shows Devanāgarī.characters, not IAST characters:

IAST: स्कान्धास

Devanāgarī characters are not a valid parameter for the IAST template. Only IAST characters are valid parameters. So the correct version of the IAST template would be:

IAST: skāndhāsa

Let me get this point up and then I can consider the nested issue for a list. Buddhipriya 23:45, 21 February 2007 (UTC)

My turn for an edit-conflict :-) :

Man, every time I try to do something good on this page I find myself getting in way over my head ;-)

Please let me try to better understand:

1. Are you saying that this has little or nothing to do with Devanagari but has to do with non-Latin-1 Roman characters (like ś and ā)?
2. Is it correct to infer that, although Template:IAST and Template:Unicode appear to have essentially the same actual function (that is, converting text to Unicode, right?), that this is just the way things are now and at some time Template:Unicode may not be compatible with Template:IAST?
3. You are recommending that, instead of changing the Template:DisplayTranslations code, I simply document that it makes use of Template:IAST?

Did I get any of these right? If not, am I banned from editing this Talk page?

Thanks for taking pity on the slow & dumb, Larry Rosenfeld (talk) 23:52, 21 February 2007 (UTC)

I think we will both be banned if we keep this up much longer. Your questions:

1. Right, it has nothing to do with Devanagari.

2a. Template:IAST and Template:Unicode do not have the same function.

2b. Actual practice on most of the Hinduism pages is that is you use Template:IAST it is assumed that you are using Unicode. While it might be true that someone uploads ASCII IAST, it probably would not display well and the person would be chastised for doing it.

2c. I do not know what the practice is with other languages regarding use of Template:Unicode.

3. Perhaps the best idea is to take the easy path and do as you suggest, just document that it makes use of Template:IAST for words in Sanskrit. You may also want to document that we are throwing water on the idea of showing both IAST and Devanagari for the same term, because they are identical. IAST is a lossless method for representing Devanagari that looks less strange to the average person.

Regarding loops, if you were to try to generalize this (which may be overkill):

Is it true that to fully-specify the representation of each language a loop would need to include elements of this form:

• ISO Language Code
• ISO Language Name
• the function to display the word may need to optionally invoke an encoding method: (encoding method|term being translated)

A Sanskrit example which includes an encoding method using an existing template:

(LANG|sa) (Sanskrit) (IAST|skāndhāsa)

A Klingonese example:

(LANG|tlh) (Klingonese) (FSK|Qapla')

FSK = Federation Standard Klingon, a Latin-1 encoding system that works in both ASCII and Unicode, which I just made up.

Can someone who knows Chinese or Tibetan give an example of this structure for their languages?

Buddhipriya 00:18, 22 February 2007 (UTC)

I created a new template (really a "macro"), Template:Lang2iso that takes the ISO Language Name and converts into an ISO Language Code (based on ISO 639-1). This then enables Template:DisplayTranslations to use the Template:Lang (instead of Template:IAST)... I think. Let me know if I'm getting warmer or if I just need to get some sleep. Thanks! Larry Rosenfeld (talk) 06:29, 22 February 2007 (UTC)

Okay. I got some sleep and I infer there's still a problem because Template:Lang does not display Sanskrit using Template:IAST (nor Devanagari using Unicode, which is something I could definitely use). I've tried to pass IAST-related functionality to Template:Lang (e.g., {{Lang|sa|{{IAST{{!}}samsara}}}}) but this seems to create errors with what is displayed. I'm certainly open to discussing this further ... perhaps using Template_talk:DisplayTranslations would be a better place to free up this talk page from this thread? However you'd like to proceed. Larry Rosenfeld (talk) 17:57, 22 February 2007 (UTC)

Your suggestion to continue this analysis on Template_talk:DisplayTranslations is great and I am sure will relieve much suffering by readers of this page. I will try a test of the template on a page and will post my results there. Please, can't someone from the other language groups give it a test too? Buddhipriya 21:41, 22 February 2007 (UTC)

[edit] Yes, IAST doesn't work

Please check if I am using the template correctly.

I have confirmed your statement that embedding IAST into the template makes it break. You can see the test at User_talk:Buddhipriya#Template_test. Since I know nothing whatsoever about Wiki template coding, I cannot provide much helpful analysis except to guess that the embedding of a template within a template raises issues. In the test I did, the template works correctly through the completion of the first pair (Pali) and fails on the second pair (Sanskrit) when it hits the embedded IAST template. If I were you I would probably either give up on trying to embed the IAST template or else get help from someone on how to nest templates.

A different issue is that in working with the template I found it confusing that the fact that the template works with pairs of attributes is not enforced structurally in the template code. Right now all the individual entries are separated in the same way, with the vertical line. What if someone does an edit that accidentally adds only one of the required members of a valid pair? That would throw off the entire list. I will set up another test of that now. In the template I think the user should be given a visual clue of how to form the pairs using something like {{DisplayTranslations | (Pali | dhamma) | (Sanskrit | Saṃkaṣṭacaturthivratam) | (English | bogus term, testing,<br>Name of a monthly ritual) | (East Asian | 法) | (Tibetan | ཆོས་ )}} instead of the current {{DisplayTranslations | Pali | ''dhamma'' | Sanskrit | {{IAST|Saṃkaṣṭacaturthivratam}} | English | '''bogus term, testing,<br>Name of a monthly ritual''' | East Asian | 法 | Tibetan | ཆོས་ }} Buddhipriya 22:11, 22 February 2007 (UTC)

Excellent set of tests! Thanks so much for taking time to do it. Questions, if I may:

(1a) You wrote, "If I were you I would probably either give up on trying to embed the IAST template...." Honestly, given my other obligations, I'm inclined to "give up" now; however, only if this template is useful the way it currently is would I want to give up. Is this template of use without the IAST function? If not, I'll keep trying to generate a solution.

(1b) One solution that comes to mind is to write a specialized version of the Template:Lang that tests for "Sanskrit" (maybe "Template:LangWithSanskritTest"?); however, the template would then lose it's genericness and expanding it similarly for other languages could become onerous. So, at face value, I'm disinclined to pursue this. (Would you counsel me otherwise?)

(2a) The second issue you raise is: "I found it confusing that the fact that the template works with pairs of attributes is not enforced structurally in the template code." Frankly, I agree. However, after exploring different options and looking at a variety of other templates, I assessed that this kind of parameter passing has been done elsewhere (though, honestly, I can't recall specific templates off the top of my head) and I was at a loss for finding an alternate solution. So, I'd like to suggest we agree this is not a pretty aspect of the template's interface but until we come across (or someone advises us) of an alternate interface, this issue will have to be backburnered for now. (Item (2b) below offers a different possibility though....)

(2b) If this were a specialized template, e.g., only dealt with five to ten languages, then we could simply create parameter names such as "Pali", "Sanskrit", etc. and then the template could be invoked as:{{DisplayTranslations| Pali=dhamma |Sanskrit=dharma | English='''law, mental objects'''}}. But, with all the IS0 639 Language Names, this might get cumbersome. Although, I could probably copy-and-modify the Template:Lang2iso switch statement so that a test is made for any parameter with an ISO 639-1 Language Name and, if it exists, then a table row is printed out -- this would also allow for language-specific print outs (e.g., IAST for Sanskrit). It's kind of tempting. The downside is that it is programmatically "ugly" (brute force!) and not upwards scaleable (that is, any new languages would be ignored -- although, the default could simply be that, if a language is not recognized as IS0 639-1 compatible, then the table line is simply displayed as is ... though this would conflict with the idea below in (2c)....).

(2c) Your idea of somehow signalling that one item from an intermediary pair has been dropped is an interesting one. A number of solutions present themselves, mostly dependent on time. One of the easier solutions might be to test the first parameter to see if it is a valid ISO 639-1 Language Name (639-2, etc., would take too long for me to code now; to check for a valid ISO 639-1 tag though, I simply need to change the "default" return value of Template:Lang2iso from "en" to something like "0" [zero] and then, if zero is returned, I know the passed Language Name was not known) and, if the Language Name is not known, the template won't display the language pair. In the case where one intermediary parameter is missing, this would cause all entries after the faulty pair to not be displayed. A more sophisticated piece of code would require that, if a parameter is found not to be a valid Language Name, then the next parameters are sequentially checked until a valid Language Name is found; once it's found, then the table continues processing as it usually does. This would be easy to do if there were some kind of loop capability in wiki-markup (e.g., a goto or do-while), but I haven't been able to find one despite spending a couple of hours rummaging around wiki-functions and meta:ParserFunctions help. The significant downside of such a solution, in my mind, is that if someone passes a Language Name that might be meaningful in the real world but is not in ISO 639-1, then that Language's translation will be dropped from the table. This possibility is highly problemmatic to me. Perhaps were you suggesting something in between such as: If the first parameter is not a known Language Name, it should be flagged somehow (e.g., precede it with the words "Unknown Language: ")? Here again, though, if the "unknown language" is a valid language (perhaps one with a variant spelling, like "Devangari"), then this flag would be detrimental to the resultant table's display. We got ourselves some options. I could go either way here. Do you have an explicit recommendation about flagging unknown language names?

If we were in a room with a blackboard, this would take 15 to 30 minutes to iron out :-( . Thanks for the laborious but fruitful Q&A sessions. Larry Rosenfeld (talk) 23:26, 22 February 2007 (UTC)

[edit] Replies to most recent questions

1a. Q: Is this template of use without the IAST function? A: Yes.

1b. Q: Would you counsel me otherwise? A: No. It would be insanity to try to customize it by language.

2a. Suggestion to backburner template format for pairs of attributes. A: Sure, why not?

2b. See answer 1b. You have had too much coffee.

2c. See answer 2b. Please go take a short walk outside, then tackle this again.

Great work so far, keep it up.

Buddhipriya 23:50, 22 February 2007 (UTC)

Fresh air good :-) We're on opposites sides of the U.S., I think. Hope the skies are clear where you are now.

FWIW, I ran the following tests:

1. <span class="Unicode">'''śraddhā'''</span> generates: śraddhā
2. {{Lang|sa| '''śraddhā''' }} generates: śraddhā
3. {{Lang|sa| {{IAST|'''śraddhā'''}} }} generates: {{{2}}}
4. {{Lang|sa| 2={{IAST|'''śraddhā'''}} }} generates: śraddhā
5. {{Lang|sa| {{IAST{{!}}'''śraddhā'''}} }} generates: {{IAST|śraddhā}}
6. {{Lang|sa| <span class="Unicode">'''śraddhā'''</span> }} generates: {{{2}}}
7. <span lang="sa" xml:lang="sa">'''śraddhā'''</span> generates: śraddhā
8. <span lang="sa" xml:lang="sa"><span class="Unicode">'''śraddhā'''</span></span> generates: śraddhā

And, to make a long story short, numbers (3) and (4) identified what's possibly a curiosity with the current Wikipedia parser. In short, to the best of my understand, (3) and (4) should be equivalent; the only difference is that number (3) implicitly makes {{IAST|'''śraddhā'''}} Template:Lang's second parameter and number (4) — by using "2=" — explicitly states that {{IAST|'''śraddhā'''}} is the second parameter. Nonetheless, Number (4) displays correctly while Number (3) does not. So, I changed this template's code so that all invocations of Template:Lang use "2=" to explicitly identify the second parameter. (This might be a wiki-bug?)

BUT, THAT WASN'T GOOD ENOUGH! Compare these two invocations of this code:

{{DisplayTranslations|Sanskrit| {{IAST|'''śraddhā'''}} }} generates:

Translations

Sanskrit:	{{{2}}}
view • talk • edit

{{DisplayTranslations|Sanskrit| {{IAST{{!}}'''śraddhā'''}} }} generates:

Translations

Sanskrit:	śraddhā
view • talk • edit

That is, using the pipe (|) character as part of a passed parameter causes the wiki-parser to incorrectly parse the template's internal expression; THUS, the "pipe escape character" ({{!}}) has to be used. Egads! I'll try to document this in this template's introductory notes.

Good enuf? (I sure hope so.) Regardless, Buddhipriya, I really want to thank you for your support, prodding, probing and encouragement. Thanks so much, Larry Rosenfeld (talk) 04:34, 23 February 2007 (UTC)

[edit] Brilliant work so far!

You may have uncovered some secret of Wikicoding here that rivals the Davinci code. Playing with your new version I discovered quite by accident that it can make links to the foreign language editions of Wiki if you embed the language code with the second parameter. Click on this sample for Hebrew, Arabic, or Chinese and you go to the Wiki for that language. If you click the Tamil label you go to the overview of Tamil for some reason. This is an interesting error that I have gotten to through some accident. I will look at it again tomorrow.

Translations
Simple English :	Ganesha
Sanskrit1 :	गणेश
Sanskrit2 :	Gaṇeśa
Arabic :	ar:غانيشا
Chinese :	zh:格涅沙
Hebrew :	he:גנש
Italian :	Gaṇeśa
Japanese :	ガネーシャ
Tamil :	விநாயகர்
Russian :	Ганеша
Telegu :	te:వినాయకుడు
view • talk • edit

In this sample there are two entries for Sanskrit because the LANG tag by default confuses Sanskrit with Devanagari just like most people do. Buddhipriya 05:38, 23 February 2007 (UTC)

Genius! Nice test! And very nice find Buddhipriya! We'll have to document that (preceding the text with the ISO Language Name) as a "feature" in the intro!

Tentative TBDs [To-Be-Decideds]:

1. Expand number of table entries! I see from your Ganesha example, the table will need at least 30 entries. I'll make it 40 for good measure. (I'll do this over lunch.)
2. Handle an odd number of parameters. I think I'll add a test so that if the second ("tranlated text") parameter of a pair is unreadable, instead of generating {{{2}}}, the template will either generate a blank or not print the row or perhaps generate an English-language message such as "missing value". (Another lunch-time activity, assuming I get lunch today :-) )
3. Other error handling? (For instance, after some sleep, I'm still ambivalent about flagging unrecognized Language Names, for the aforementioned reasons, as your use of "Simple" and "Sanskrit1" possibly show the utility of ignoring such....)
4. Other TBDs?

Thanks again O Wise & Generous one! Larry Rosenfeld (talk) 13:17, 23 February 2007 (UTC)

I will reply to your questions in detail in another post. I wanted to focus on one point, which is that the Hinduism project may be settling on an approach for handling Devanagari in Hinuism-related articles. See this thread. The use of Devanagari is very inconsistent across the Hinduism articles. I think opinion has reached critical mass to replace all Devanagari with IAST in a gesture toward English reader sensibilities. I mention this because for the Translations template, the "Sanskrit" entry alone perhaps could be tested on some real pages to show people some alternative ways of doing it. I think that I will set up a test demo of this that will look something like:

Translations
Simple English :	Ganesha
Sanskrit :	गणेश Gaṇeśa
Chinese :	格涅沙
Japanese :	ガネーシャ
view • talk • edit

Note that in this test the "Sanskrit" entry shows the word both in Deva and one in IAST.

Is there any way to let the user modify the title of the box? For example, could the user add a parm to have the title of this box read: "Ganesha's name in other languages"? Buddhipriya 18:40, 23 February 2007 (UTC)

This demo also proves that you can use a construct like Chinese | [[zh:格涅沙 | 格涅沙]] to make the translation be a link to the foreign language Wiki without needing to show the LANG code in the visible table. Much more attractive. Buddhipriya 18:50, 23 February 2007 (UTC)

Replies to your question about other error handling, I am not sure what to say. Perhaps putting the template into live test use on a few pages would uncover real-world issues.

Regarding the use of language names, if I understand your code now you use a name like "Sanskrit" and then look up the LANG code "sa". Since the LANG codes are less subject to error would it be good to allow the user to provide a LANG code as the first argument? For instance, what would happen if you test the language name "Saṃskṛtam" which is the word for "Sanskrit" in Sanskrit? Buddhipriya 18:59, 23 February 2007 (UTC)

Yes, this has crossed my mind. In my mind, I've rationalized that this is the English-language WP and thus it is not unrealistic to assume that an English-language Language Name will be parsed. I could see the value though in assuming someone might want to pass a ISO 639 alpha code -- for one thing, it's shorter. Perhaps another thing to add to the TBD list? (Just where is that list anyway? ;-) ) Larry Rosenfeld (talk) 06:32, 24 February 2007 (UTC)

I'm not sure what it is, but I keep finding myself having edit conflicts with you :-) (I was gonna document that, instead of passing [[ar:غانيشا]] one could also pass [[ar:غانيشا|غانيشا]] .... :-) )

Per your request -- excellent idea, BTW -- let's add this to the TBDs:

• Make some table features configuable such as the title. (Anything else, e.g., background colors, foreground colors...?)

FWIW, one reason I've been sitting on this type of suggestion is that it could get tricky. In particular, to the best of my understanding, either the table-feature-configurable parameter would have to go last or all the other parameters would have to be numbered. For instance, let's say someone wants to create the following table:

{DisplayTranslations | Pali | dhamma | Sanskrit | dharma }}

To add a "title=" parameter, I think it would have to be specified either last as in:

{DisplayTranslations | Pali | dhamma | Sanskrit | dharma | title=Translations<br>of<br>Dhamma }}

Or the other parameters would need to be explicitly numbered as in:

{DisplayTranslations | title=Translations<br>of<br>Dhamma | 1=Pali | 2=dhamma | 3=Sanskrit 4=dharma }}

So, as long as users understand and are willing to abide by this convention, it should be pretty easy to implement. In fact, during my lunch period (right now), I'll start with trying to implement this.

Larry Rosenfeld (talk) 19:00, 23 February 2007 (UTC)

Whoa! Turns out I was wrong. Configurable parameters are way easier to implement than I thought. It took me about 30 seconds to implement the title= param and the other parameters do NOT need to be numbered. I'll see what else I can make readily configurable and then introduce it to this template within the next 20 minutes or so.

Thanks again for the excellent ideas! Larry Rosenfeld (talk) 19:09, 23 February 2007 (UTC)

Okay, the following configurable parameters are done:

title - the title of the template (Default: "Translations")

bgcolor - the in-row background color (Default: Cornsilk)

textcolor - the in-row text color (Default: black)

bordercolor - the border's color (Default: Peru)

headercolor - the header & footer's background color (Default: Wheat)

headertextcolor - the header & footer's text color (Default: Maroon)

If you hate the names, we can change them :-) I'll try to get to the other TBDs tomorrow. (Perhaps I can squeeze one in later.) Thanks again, Larry Rosenfeld (talk) 20:06, 23 February 2007 (UTC)

Where is the documentation on how to write these things, by the way? Buddhipriya 19:28, 23 February 2007 (UTC)

I generally start with Help:Template and meta:ParserFunctions. Thanks again! GTG! Larry Rosenfeld (talk) 20:06, 23 February 2007 (UTC)

All of this is wonderful, thanks. I will sit tight and wait for your next move. Let me know when you think it is stable enough to put on a live test page. Buddhipriya 04:49, 24 February 2007 (UTC)

Let 'er rip! (FWIW, I've already inserted this table into Skandha, Upadana and Sravaka.) One thing just to underline, at this time, I'm thinking it's best to let someone pass erroneous parameters and then see the erroneous display and then fix their invocation accordingly; nonetheless, I'm still open to more elaborately exploring options for error-checking here. Thanks again for all your hard work and excellent feedback. Larry Rosenfeld (talk) 06:32, 24 February 2007 (UTC)

[edit] Some confusion in the documentation about the role of Unicode versus IAST

I made changes to the documentation to the Unicode and IAST issues, splitting them into two separate sections. Please check over what I have done and revert or rewrite as you think best. Please test that the code actually works, I have not had time to do good testing.

Here I will crosspost something I just put on the Unicode template talk page to try to clarify the same confusion there:

Using the Unicode template to display Devanagari is no longer as much of an issue as it once was due to better support for Unicode on all new computers that have been sold for at least the past two years. On a practical basis the Unicode template is rarely used on the Hinduism pages that use Devanagari. The IAST transliteration method can optionally be shown via the IAST template if IAST is used. But IAST and Unicode address completely different issues and should not be used in the same sentence, except like I just did.

The IAST template and the Unicode template do different things. IAST is one of several incompatible transliteration methods for Devanagari. So using the IAST tag specifies which of the alternative methods is being used. Also note that IAST is a transliteration method for a writing system (Devanagari) which is used for multiple languages such as Hindi, Sanskrit, etc. Sanskrit is a language that can be written using various writing systems, such as Bonji, IAST, Devanagari, etc.

The template I see most often on the Hinduism pages for Sanskrit is some variant of this: ([[Sanskrit]]:{{lang|sa|गणेश पुराणम्}}; {{IAST|gaṇeśa purāṇam}}) which displays:

(Sanskrit:गणेश पुराणम्; gaṇeśa purāṇam)

Notice that no explicit Unicode tags are used. The LANG tag argument is just the raw Unicode character value. As explained in the documentation the whole Unicode thing is now a passing issue as this loka becomes fully Unicode-complaint.

Most computers I use – both at home and at work – are about five to ten years old. (The one exception is my wife's new laptop.) If I don't use the {{Unicode}} template, most of the time all I see is a string of empty boxes. (Relatedly, when I look at sa.wikipedia.org pages, most of what I see are thousands of empty boxes.) So it is not at all "debateable" for me; for me, it is a matter of basic usability. Larry Rosenfeld (talk) 04:46, 24 February 2007 (UTC)

No problem, just rewrite to explain that. Buddhipriya 04:52, 24 February 2007 (UTC)

Changed "debateable" to "waning." You da best! Larry Rosenfeld (talk) 06:55, 24 February 2007 (UTC)

[edit] The Sanskit page may be renamed

See the discussion and join in at this thread. Buddhipriya 05:20, 24 February 2007 (UTC)

FWIW, I could go either way on it -- as long as there's a redirect page named Sanskrit_language, this template is content :-) Larry Rosenfeld (talk) 06:57, 24 February 2007 (UTC)

[edit] Brain twister

How do you say L'Chaim in Chinese? Buddhipriya 05:26, 24 February 2007 (UTC)

[edit] Case study: Used to display the term skandha

Now that you have gotten some live samples up, let me use one from the Skandha page to raise a couple of usage details. I will build up the comment via a series of edits, of which this is the first. First, here is the code in use now:

Translations of Skandha
Chinese :	五蕴
Devanagari :	स्कान्धास (Sanskrit), खान्धास (Pali)
English :	aggregate, mass, heap
Japanese :	五蘊
Pali :	khandha
Sanskrit :	skandha
Vietnamese :	Ngũ uẩn
view • talk • edit

Very attractive layout, nice colors!

Notice that the layout refers to "Devanagari" as a "translation". Since Devanagari is a writing system, and not a language, I think it would be better to not cite it on the left side, but rather to refer to it as one of the ways you could write the word in the two Indic languages that use it below (Sanskrit and Pali). Let's see what that would look like:

[edit] Variant 1: Omit the Devanagari row

Translations of Skandha
Chinese :	五蕴
English :	aggregate, mass, heap
Japanese :	五蘊
Pali :	khandha
Sanskrit :	skandha
Vietnamese :	Ngũ uẩn
view • talk • edit

This line of code was snipped: | Devanagari | {{Unicode{{!}}स्कान्धास}} (Sanskrit),<br>{{Unicode{{!}}खान्धास}} (Pali)

[edit] Variant 2: Use IAST transliteration method

Currently the display tranliterates the Sanskrit and Pali using Simple English as a method. Lets re-tranliterate the Devanagari terms you cited for the Sanskrit and Pali languages into IAST. I will use IAST and will not change your spellings, just tranliterate what you wrote in Devanagai:

Translations of Skandha
Chinese :	五蕴
English :	aggregate, mass, heap
Japanese :	五蘊
Pali :	khāndhāsa
Sanskrit :	skāndhāsa
Vietnamese :	Ngũ uẩn
view • talk • edit

Note the appearance of the pesky IAST character for long a: (ā)

Zooming in on the code for the Pali and Sanskrit we discover that one uses the IAST pipe and the other doesnt. A discrepancy!

| Pali | khandha | Sanskrit | {{IAST{{!}}skandha}}

If we want to signal that we are using IAST for the Pali, we could recode this as:

| Pali | {{IAST{{!}}khāndhāsa}} | Sanskrit | {{IAST{{!}}skāndhāsa}}

Is this the best term to use in those languages? Let's look it up. One source could be Franklin Edgerton's Buddhist Hybrid Sanskrit Grammar and Dictionary (ISBN 81208-0997-1) which you may have. Edgerton is handy to have because he covers both Sanskrit and Pali variants of terms. He also is a better source than regular Sanskrit dictionaries such as Apte or Monier-Williams because he specifically cites Buddhist spins on these words. The main term would be in Sanskrit, so we look up and find (p.607) "skandha" (not skāndhāsa), and the cross-reference to Pali "khandha" (not khāndhāsa). Is this right? Let's look it up in Apte's The Practical Sanskrit-English Dictionary (ISBN 81-208-0567-4) and yes, it is the same (p. 1003), but here it is in Devanagari because that is what Apte uses (स्कंधः) As is Apte's style, he uses anusvara for the "n" and also shows the word in nominative case. Apte gives us no Pali version.

Perhaps the Devanagari you started with was wrong?

[edit] Variant 3: Show Devanagari and IAST versions together

Lets use Apte's version for the Sanskrit line, and show two different ways to write it, one in Devanagari and one in IAST. This time if we show off and use the IAST characters ṃ and ḥ (skaṃdhaḥ) our impressive display of scholarship will only confuse the average reader. So let's use "skandha", transforming the anusvara back to "n" and dropping the case ending. Amazingly, this is what you started with.

Translations of Skandha
Chinese :	五蕴
English :	aggregate, mass, heap
Japanese :	五蘊
Pali :	खंधः (khandha)
Sanskrit :	स्कंधः (skandha)
Vietnamese :	Ngũ uẩn
view • talk • edit

But now we have the same Sanskrit word written using two different writing systems (Devanagari and IAST) together so the reader can compare them. Actually, here the IAST is a simplification of the Devanagari and not an exact romanization. We have cheated a bit in the interests of reader comprehension. While we are at it, let's do the same thing for the Pali version. In the code I use the IAST tag around the Pali romanization, just like the Sanskrit romanization.

To me, this table is more clear. What do you think? Buddhipriya 23:53, 24 February 2007 (UTC)

Looks great! And thanks for the many interesting and well-articulated lessons. (Honestly, last week, I didn't know what IAST was and couldn't spell from memory the word "Devanagari.")

FWIW, if I may, I did something similar to what you've demonstrated here on the Three Jewels and Sravaka pages though I was ambivalent as to whether to put the IAST or the Devanagari in the parentheses. (On the Sravaka page I put the Devanagari in parentheses. On the Three Jewels, I just left the two encoding methods side-by-side.) In addition, when I first inserted this template on the Skandha page, I did put the Devanagari under "Pali" and "Sanskrit" titles (similar to what you have above) but then thought that whoever introduced the terms into the article might prefer having Devanagari separated out; that, plus, I know there's an ISO code for Devanagari (deva), so .... Regardless, I like what you have. Please change the Skandha page accordingly (or, if you prefer, I'd be happy to do so). Good work!

This re-introduces a topic, I think, that you've touched on before. I see at least two components:

1. how should we order the languages in the table?
2. when using IAST and Devanagari, which (if either) goes first and which goes in the parentheses.

In regards to (2), it looks like you prefer "Devanagari (IAST)". That's fine with me. I consider you an authority on this. When it comes to Devanagari (as indicated above), I'm pretty clueless. FWIW, my preference with Pali has been to put the Romanized letterings first because that's how I see it when I read the Pali Canon and contemporary Theravada commentary. But, honestly, I don't have a strong preference.

Regarding (1), I've chosen a different method for each article I've inserted this template into so far. I recall using three different methods:

(a) alphabetical (e.g., Chinese, English, Japanese, Pali, Sanskrit, Tibetan)

(b) quasi-chronological (e.g., Pali, Sanskrit, Chinese, Japanese, Tibetan, English)

(c) language groupings (e.g., English, Pali, Sanskrit, Chinese, Japanese, Tibetan)

Frankly, if I were smart enough, I think I'd prefer (b) -- although, there could be some confusing outcomes, such as, technically, should "Tibetan" go between "Chinese" and "Japanese" (given when Buddhism was first established in each nation)?

Also, I think the ordering I used on the Three Jewels article (English, Pali, Sanskrit, Sinhalese, Chinese, Japanese, Vietnamese) is a failure mostly because the sidebar is entitled, "Translations of the Three Jewels" and then provides four "translations" in the English row: Three Jewels, Three Refuges, Three Treasures, Triple Gem. But, of course, "Triple Gem" is not a "translation" of "Three Jewels" -- they're both translations of ti-ratana. So, I think in this case, clearly, staring with Pali/Sanskrit would make the rationale for the multiple English translations more sensible. What has caused me to pause in making this change is that the article is entitled, "Three Jewels," and I've felt that it was appropriate to keep this title for the DisplayTranslations sidebar. If I were to put Pali/Sanskrit first, I'd rather the sidebar be entitled, "Ti-Ratana" (or Tiratana, etc.). (Am I overthinking and second-guessing enough for you?) More to the point, are you ready to spell out your recommendations regarding these two questions yet or perhaps we should have more time to experiment on a variety of other articles before imposing some guidelines?

FWIW, if you're inclined to do your scholarly magic to any other DisplayTranslations sidebar I've inserted, please feel free to do so. I give you carte blanche. (My only request is that, on WP Buddhism articles, the sidebar's color scheme be related to that used by Template:Buddhism :-) )

THANKS AGAIN for the excellent elucidations & guidance. Best wishes, Larry Rosenfeld (talk) 01:52, 25 February 2007 (UTC)

[edit] Replies to most recent questions

You raised several points, I will build up a reply here with a series of short edits.

1. Regarding making changes to the usage of the template on live pages in the Buddhism topics, I would prefer that you take care of that for now, as you understand what you are trying to accomplish with the end result for the user. Once the template is fairly stable I may try it on some other pages.

2. English should come first, because the purpose of the table is to help English readers. The title of the box is in English, for example.

3. Regarding putting Devanagari or IAST first, I don't know which is better. Does it really matter? If the purpose of this is to help the English reader, they can't read Devanagari anyway. If the purpose is to show lots of funny-looking marks (like the Chinese and Tibetan, which to me are attractive squiggles) then I would put the funny marks first.

4. Regarding what order to list the languages, does it really matter? It may make sense to use one order or another depending on what word is being translated, showing the earliest source word first. For words that originated in Indic texts, putting the Indic languages first might make sense. Most of the Mahayana Indic material was written in Sanskrit first, but most of the Hinayana Indic material was written in Pali first (I think that is true, I would need to look it up to be sure). Some of the terms related to Buddhism in China and Japan probably have no exact counterparts in Indic languages if those concepts developed subsequent to the migration of Buddhism from India to other countries. Perhaps the people who are working with the various languages would be able to tell what the best order would be. I only work with Sanskrit materials, so I really don't know.

5. Regarding the actual words used in the tables, be sure you get people to double-check everything, even if you are an expert in the languages yourself. It is easy to make mistakes. I have not had time to review the pages you mention. But in your post here you mention that "Three Jewels" is a translations of "ti-ratana". Are you referring to the Sanskrit term "tri ratna" or some other term? "ti-ratana" is not Sanskrit, perhaps it is from some other language. I would need to look it up, I don't really know. Buddhipriya 17:57, 26 February 2007 (UTC)

FWIW, ti-ratana (or tiratana) is Pali. I know some Pali. If I have questions about a Roman-lettered Pali word, I look it up in the on-line PTS Pali-English Dictionary. While of course there will be typos, in general I would like to suggest that there's no need to question my use of Romanized Pali. GTG, Larry Rosenfeld (talk) 18:24, 26 February 2007 (UTC)