Even if you just study a Western European language you'll end up with extra characters. A bit further afield and you end up with alternative alphabets or for example Chinese characters.

No it isn't.

I'm working on an application now where many of the variable names use Norwegian specific characters and the bulk of the comments that I write are in Norwegian. Now that we have Unicode there is no reason why non-keyword text should be in English or in a Latin or Latin derived alphabet unless you expect people who can't read that language to maintain the code.

Do you have Norwegian standard libs for your application? Or is it just your business logic that's in Norwegian?

Fact is most of the code you find anywhere uses English keywords like "while" or "except", and the standard libs will use English names like "socket". The occasional non English code language is an oddity.