User:Prosfilaes undid my rfd, although it's not allowed to do so, until it's discussed. Here you go - my reason for rfd: not English, not in Roman script, attestability doesn't make this word English. --Anatoli 02:15, 30 September 2011 (UTC)
The rfd concerns the English section, of course. --Anatoli 02:17, 30 September 2011 (UTC)
Not English. This is called w:code-switching. Just because a foreign word is inserted unchanged into an English text doesn’t make it suddenly English. —Stephen(Talk) 09:26, 30 September 2011 (UTC)
Yes, this seems to be humorous code-switching, as the English word was obviously available to people writing these sentences. Lmaltier 20:26, 30 September 2011 (UTC)
Before this discussion is going off-topic, yes, delete that English section. -- Gauss 22:43, 30 September 2011 (UTC)
Is drive-by tagging permitted? Just tossing a tag on an entry, without actually starting the RfD, isn't doing the job, IMO.--Prosfilaes 07:33, 1 October 2011 (UTC)
Should be completed, of course, but people sometimes just toss on a tag. What I do when I find an incomplete one like that is complete the process. But that’s just me. —Stephen(Talk) 07:56, 1 October 2011 (UTC)
I can't think of any CFI reason to delete this. The only non-CFI reason I can think of is "common sense". Abstain. Mglovesfun (talk) 08:36, 1 October 2011 (UTC)
Delete Yes, common sense shows that people using it used the foreign term intentionally and humorously, not because they needed it to express their idea. But this reasoning should not lead to delete foreign words used in the language because they are felt as the most appropriate to express something, such as kimono or bouillabaisse, even when they are very rarely used in the language. Lmaltier 16:12, 1 October 2011 (UTC)
Double standards, Lmatier? Why a Cyrillic word in English is not OK but Thames in Mandarin is OK? Or you think you have a better understanding of its usage in Mandarin? --Anatoli 01:52, 2 October 2011 (UTC)
Delete per SGB unless it can be shown (which I doubt) that the word is used as English and not as Russian (or other) in English.—msh210℠ (talk) 17:50, 2 October 2011 (UTC)
They're not parallel; Москва is unquestionably a real word, and will still have an entry if the English section is deleted. If Thames河 is a real word, then we need to have an entry under some language header.--Prosfilaes 07:29, 3 October 2011 (UTC)
Thames河 is SOP, and we will still have both parts even if the sum is deleted. If we decide to keep it despite it being SOP and conscious code-switching, it would be inconsistent not to keep Москва, which is just conscious code-switching but not SOP. The two are parallel; the argument for keeping Thames河 (either where it is, or at Thames#Mandarin) is that it is conscious use of the term in a Chinese sentence; this argument, if valid, also saves Москва (which is consciously used in an English sentence). - -sche(discuss) 07:46, 3 October 2011 (UTC)
That's right, 河 just means "river", which is always attached to river names but river names are always written in Chinese characters, not Roman, no matter how little known they are, if they written in Roman or other scripts, they are not in Chinese, very much like Москва is not English, even if it's inside an English sentence. (Moskva is more attestable (not only as river but city), international airports often use Moskva, meaning Moscow.) Thames河 represents the code-switching in the actual the name of the river. There are also examples of "Москва市" on the internet (Moscow city).--Anatoli 08:01, 3 October 2011 (UTC)
I think the script difference is a bit of a red herring. The key, as Prosfilaes notes, is whether a word brought in from a foreign language is used as the source language, or as the host language (i.e. the language of the majority of the utterance; for "至yeah" this would be Cantonese, for "Москва" in this discussion it would be English, for "Thames河" in this discussion it would be Mandarin).
Proper nouns are generally so specific that they only signify their referrant -- the "Thames" in "Thames河" only refers to the same "Thames" that is signified in English. Moreover, as others have noted, "Thames" in Mandarin contexts is almost always spelled out in hanzi. Both points argue that alphabetically spelled "Thames" in Mandarin contexts is being used as English.
That said, if a term takes on a meaning specific to the surrounding host-language context, as in "至yeah", a much stronger argument can be made that it has been adopted into the host language, regardless of the script used. In this case, "yeah" used to signify "trendy" is quite different from usage in English, and points to the adoption of "yeah" into Cantonese, no longer as an English word, but now as a specifically Cantonese term.
As a side example, "Washington" in English is generally a proper noun, indicating either the state or the capital of the US. The word was adopted into Navajo in a way that made it into a more general-use term, which now means Washington, DC, the federal government, or a government in general. In this case, "Wááshindoon" is used not just as the English place name, but as Navajo, with meanings specific to usage in Navajo contexts.
I hope this helps tease apart some of the arguments, vis-à-vis scripts, and use in vs. as a language. Since we deal with the written forms here at WT, I think we sometimes get sidetracked by the visual representations, and perhaps lose sight of some of the other aspects of language. -- Cheers, Eiríkr ÚtlendiTala við mig 17:46, 3 October 2011 (UTC)