Criteria for languages?

John Cowan cowan at
Wed Dec 2 04:44:57 CET 2009

Caoimhin O Donnaile scripsit:

> But most Scots speakers don't "switch" to English in the way
> that most German dialact speakers are capable of switching to
> Hochdeutsch.

Well, they do when they take up the pen (or the keyboard); it's not an
accident that in Switzerland, Standard German is called _Schriftdeutsch_.

Saying what is Scots in detail, of course, is a hard problem.  I've been
talking with the _Dictionary of the Scots Language_ folks about producing
a Scots wordlist in the public domain.  They aren't opposed in principle,
but the problem is that the dictionary doesn't contain any of the words
shared with English.  This is a natural consequence of making the major
dictionary of a language a bilingual dictionary.  Just leaving the shared
words out isn't an option: the word "diglossia" is certainly Scots --
you can hardly talk about the position of Scots in Scots without using
it -- but the DSL understandably doesn't include it.

If we had a corpus of works in pure Scots, we could use that to figure
out which words are at least presumptively Scots, but we don't and can't.
So when "right" appears instead of "richt" in a supposedly Scots document,
is it a typo or a borrowing from English?  "Typo" is the safe assumption,
but then we have the problem of taking an English wordlist and filtering
out all the words that have Scots cognates -- or is it all of them?
What if most of the time the English word is used and the native Scots
word is deeply obscure?

John Cowan              cowan at
Any day you get all five woodpeckers is a good day.  --Elliotte Rusty Harold

More information about the Ietf-languages mailing list