Standardizing on Oxford English

Hello,

I noticed the Documentation/SubmittingPatches file reads:

May I ask why? US English is highly idiosyncratic, illogical, and used by a minority of the English-speaking population of the world (see https://en.wikipedia.org/wiki/American_and_British_English_spelling_differences). Since British English has its own idiosyncrasies, why not use Oxford English, the most international English that is used by millions of the world? It is used by practically every international organization (such as the UN, ISO, IEC, BIPM, NATO, etc.), taught in practically every school in non-native-English-speaking countries (and even native-English-speaking ones), and used by myriad publications (e.g., Nature) and people around the world. Given the inherently international nature of the Git project, it makes complete sense to follow suit.

Regards, Varun

On Fri, Jun 05, 2020 at 12:34:21AM -0500, Varun Varada wrote:

Traditionally, the "GNU C" locale expects that the character set used for all output should be in "us-ascii" -- a choice dictated by historical encoding standards. The most logical language to use when writing in something called "us-ascii" is "US English." :)

This has little current day relevance, but projects with code histories spanning decades tend to stick with whatever the expectation is for the "C" locale as their default language.

You could try being Canadian and sit on both of those chairs at the same time. Then you can use both "colour" and "authorization" in the same text. :)

I think that since git was "born" in the US (courtesy of a Swedish-speaking Finnish immigrant), it makes sense for it to continue to use US-English as the internal default. There's already a way to cause it to output your preferred version of English by setting your locale appropriately.

-K

On 2020-06-05 at 05:34:21, Varun Varada wrote:

I should point out that many of your arguments about U.S. English are true of English in general. As a native U.S. English speaker who also knows Spanish and French, I can confidently say that even French, which many find difficult, has a mostly regular correspondence between letters and sounds, and, overall, a reasonably consistent set of rules for verb conjugations, albeit with many irregular verbs. English, in any form, has none of that. It is, as languages go, highly irregular.

I didn't write the text in question, but I suspect the reason is practicality: most open source projects use U.S. English, and most contributors to Git are able to write the U.S. variety. It's hard for me personally to write Oxford English because I have never written or spoken it, and when I need to consult a reference, the one I have is from the University of Chicago, not Oxford. I suspect many Canadians and second-language speakers from at least parts of the Americas are more likely to be familiar with the U.S. variety than Oxford or British English, although I don't know for certain.

This isn't a defense of U.S. English (after all, I wrote the first paragraph), but just an acknowledgement of the way things are. This project is all about practicality rather than purity; to quote from CodingGuidelines:

   Again, we live in the real world, and it is sometimes a
   judgement call, the decision based more on real world
   constraints people face than what the paper standard says.

On Fri, 5 Jun 2020 at 18:24, brian m. carlson <sandals@crustytoothpaste.net> wrote:

Agreed, as I'm a native US English who knows French as well. But I guess my point is that out of all of the varieties, Oxford English is the most prevalent, international, and etymologically correct, which is why I suggested it.

The reference for Oxford is the first spelling on lexico.com, which is a very widely-used resource. Canadian English is essentially identical to Oxford except for the -yze set of words, for which Oxford maintains the etymologically correct -yse endings. And second-language speakers learn Oxford by and large, though many from Brazil apparently end up just moving to the US to learn English, where they necessarily learn US English.

However, I never found the guidance of doing what other people are doing a convincing one, especially when the alternatives are more logical/convincing/"better". Though I do recognize Konstantin's point that the project is decades old.

While I somewhat sympathize with the sentiment, from the text, it seemed like things were in a mixed state, so I wanted to suggest picking the standard that makes the most international sense. As the usual guidance goes: when faced with two choices of relatively equal difficulty, choose the one that makes the most sense long-term.

Varun