Talk:تاء مربوطة

I wonder when final t become unpronounced in Arabic. It seems some Hindi borrowings preserve it and some don't. —Aryaman (मुझसे बात करो) 12:16, 2 August 2017 (UTC)
 * It also shows up quite consistently in Arabic borrowings in Malay, e.g., , , , etc. I wonder if it would be better to somehow conditionally make it transcribed in etymologies. Wyang (talk) 12:31, 2 August 2017 (UTC)
 * That would be too much unnecessary work imo, having to detect the script equivalent of final -t etc. Why not just add (t) or (t) in the transliteration by default? That way those who can't read the Arabic script understand that -t doesn't spontaneously appear in the descendants (as I used to think). —Aryaman (मुझसे बात करो) 14:06, 2 August 2017 (UTC)
 * I agree. I think it may be more desirable for the transliteration module to return (t) for tāʾ marbūṭa when it is invoked by etymology templates, such as or . Wyang (talk) 08:27, 3 August 2017 (UTC)
 * I wouldn't transliterate it. If it's really necessary, a form with ʾiʿrāb could be provided, e.g. (pronounced "šarikat" only in the construct states), nominative indefinite . Compare the situation with Ukrainian, which often alternates  and . E.g., a Ukrainian lemma  sounds very different from its Russian cognate  but "о" re-appears in the inflected forms -  (genitive or accusative singular, etc.). --Anatoli T. (обсудить/вклад) 00:35, 4 August 2017 (UTC)
 * I think using tanwīn to force display of tāʾ marbūṭa is not accurate, as borrowings in other languages which preserve the final -t are generally directly loaned from the historical pronunciation of the non-ʾiʿrāb form, not the tanwīn forms. The easiest solution would be to transliterate the tāʾ marbūṭa as (t), when (and only when) it is working in templates invoking Module:etymology. The current romanisation in such etymologies is non-ideal, as there is dyschronicity in the etymology chain ― using a modern pronunciation to explain borrowing from a historical pronunciation, and leaving the false impression that the -t in recipient languages was an innovation therein. Using as the Arabic etymology may be desirable if the nunation is also preserved in another language (which seems to be very rare). Apart from the Malay ones above, entries which may benefit from the modified tāʾ marbūṭa display include: 🇨🇬, 🇨🇬, 🇨🇬, 🇨🇬, 🇨🇬, 🇨🇬, 🇨🇬, 🇨🇬, 🇨🇬, 🇨🇬 (currently manually forced to show a -(t) transliteration), etc. Wyang (talk) 01:15, 4 August 2017 (UTC)
 * I am well familiar with the situation with the loanwords from Arabic with the letter and it's the question people often ask - why loanwords have -t? But it's not always retained. I wrote the usage notes in ة a while ago. --Anatoli T. (обсудить/вклад) 01:22, 4 August 2017 (UTC)
 * Hi. Since you mostly built these modules for Arabic, what's your opinion? . --Anatoli T. (обсудить/вклад) 01:24, 4 August 2017 (UTC)
 * I feel like we've had this discussion before. I definitely disagree with having one transliteration in etymology sections and a different one elsewhere; that's a massive can of worms. Benwing2 (talk) 03:17, 4 August 2017 (UTC)
 * Yes, it was discussed a few times. One place is Wiktionary_talk:About_Arabic. One idea was using ẗ. --Anatoli T. (обсудить/вклад) 03:31, 4 August 2017 (UTC)
 * If a single romanisation is to be used, I definitely feel like supporting the use of either (t) or (t). I recall having great trouble trying to find the etymology when creating 🇨🇬; and when I finally found the page, the same was romanised as both mušāraka and mušārakat on the page, and the -t form wouldn't show up when it was used in  ― very frustrating for one of the Arabic illiterati. Wyang (talk) 04:05, 4 August 2017 (UTC)
 * "-a(t)" was used but then it was abandoned. Don't remember whose idea it was but it was agreed that it looks annoying, especially because it's not optionally silent. Currently, in [ʾiḍāfa]] constructs the pronounced -t is added manually (as in Benwing2's example below) and in combinations - noun + adjective, it's removed, e.g. . --Anatoli T. (обсудить/вклад) 04:25, 4 August 2017 (UTC)
 * I always thought loans into Persian, Turkish, Hindi, etc. with the -t are loans from the construct state. In other words the -t is found mainly in words that are frequently found in the construct state. For example, 🇨🇬 comes from phrases like . --WikiTiki89 03:04, 4 August 2017 (UTC)