|  | >When does a word pass from being of colloquial usage into common usage?
When sufficient numbers of people use it for a sufficiently long time that
it is not considered a fad ... bearing in mind that certain colloquialisms
are generally considered in poor taste and therefore are never really 
accepted as common usage.
>What are the precise measures (if any) that are used to determine the 
>transition?
None ...
>Is any one authority (for English language at least) recognised as wielding
>final judgement in this respect?
No, English has NO governing authority, such as l'academie francais.  However,
acceptance by one of the major dictionary editors ... notably the Oxford
University Press (publishers of the Oxford English Dictionary) in the UK,
and the publishers of Websters in the USA are generally considered the
prime 'selectors' for any given word's acceptance into the language.  Although
there are many more dictionary publishers all with slightly different ideas
as to what is acceptable in the language.  Note also that dictionaries
publish foreign words in the dictionaries as foreign words that are used
by English speakers on a regular enough basis that the word's meaning should
be available to a dictionary searcher ... but it's inclusion does not really
make it English!
>Seemed strange that this topic had not already been raised when so many 
>discussions revolve around the concept.
It's not really that strange given that there are no custodians of the
language ... which is what actually allows us the kind of discussions that
go on in here.
Stuart
 | 
|  |     	English has probably a higher turnover of words than any other
    language, and a word is "English" when either a significant number of
    English speaking persons understand it, or it is the only word
    available that seems to express the meaning.
    
    	My (Oxford) dictionary has "troth" as archaic, and doesn't even
    suggest that "fianc�" might not be English. I suppose that this means
    that English now has accented characters.
    
    	It doesn't mark "ellipsis" as not English, but "bazaar" which is
    probably understood by more people is there but described as Persian in
    origin.
    
    	We has met the authority, and he is us.
 | 
|  |     The Japanese Monbusho (which I think is Ministry of Education) dictates
    which characters can appear in certain kinds of documents, while other
    words have to be spelled out.  However, most publications ignore the
    rules, and especially professional or scholarly publications use a lot
    more characters.
    
    Certain additional characters are allowed for use in names.  When these
    regulations were first announced, a lot of people suddenly found that
    their names were illegal.  However, they kept using their names, and
    some parents rebelliously continued giving their children such names.
    
    So, both sets of regulations have been changed, and both are scorned.
    
    (Unfortunately, the Mombusho has a lot better enforcement powers in
    most other areas, such as preventing history books from teaching what
    Japan did during World War II.)
    
>    	My (Oxford) dictionary has "troth" as archaic, and doesn't even
>    suggest that "fianc�" might not be English. I suppose that this means
>    that English now has accented characters.
    
    I have occasionally seen the words cooperate, coordinate, etc., with
    a dieresis over the second o.  I believe this conference already has
    notes distinguishing this dieresis from an accent, but I still can't
    see the difference.
    
    However, my Oxford dictionary doesn't have a word that is spelled
    f - i - a - n - c - square box.
 | 
|  | >    However, my Oxford dictionary doesn't have a word that is spelled
>    f - i - a - n - c - square box.
    
    I suspect that most readers of this file don't know what you're talking
    about here.  I presume that your VT382 terminal is set normally to
    interpret characters with the sign bit set as a DEC_Kanji character
    rather than a Latin-1 character.  However, a byte with the sign bit on
    followed with one with sign bit off (a Latin-1 character followed by an
    ASCII character in Latin mode) is in DEC_Kanji a user-defined character
    to be downloaded from the host.  However, if no such UDC exists in the
    host system's on-demand loading character table, a square box is
    displayed.
    
    (If on the other hand, regular readers of this file know very well what
    you were talking about, but I don't know that because I'm an infrequent
    reader, then I offer my apologies for the presumption.)
 |