Machine Translation

Michael Everson everson at evertype.com
Thu Sep 10 22:44:31 CEST 2009


On 10 Sep 2009, at 16:56, Peter Constable wrote:

> Michael, this is you at your worst in the role of Reviewer. You're  
> reacting just the same way as you did when we had requests for tags  
> for "Latin American Spanish": it doesn't fit your understanding of  
> linguistic realities, and so you think it's out of scope. You're  
> wrong!

I disagree. I also think that your comment is not civil.

Any piece of text can be translated from one language to another.
It can be translated well, or badly.
It can be translated by a person, or by a machine.
That is an authorship tag.
There is no way of knowing whether a bad translation was written by a  
person or a machine.
There is no way of knowing whether a good translation was written by a  
person or a machine.

This machine tag could not distinguish between an accurate translation  
or an inaccurate one. It has no a priori rationale but to be some sort  
of vague "caveat" and since a translation might as well be accurate as  
inaccurate, what good would such a caveat yield?

> The whole purpose of IETF language tags is to be used in processes.  
> There is no reason to tag content unless some process somewhere is  
> going to do something with that information.

Yes, the tags are used *in processes*. The tags tag *entities*,  
however; they do not tag *processes*. We will not have a tag <human>  
to indicate that a human wrote text, or translated it, will we?

> I'm not at all convinced that there are reasonable use cases that  
> warrant a subtag (or extension) for indicating MT content, but I  
> really object to you ruling this out from the outset.

This really angers me, Peter. I said "I foresee little likelihood".  
This is not "I promise to reject". It is my present opinion (I do, in  
fact, foresee little likelihood that reasonable use cases can be  
presented); I am entitled to hold this opinion, and I am entitled to  
express it if I wish.

Am I not?

I think you are out of line insulting me in your first sentence, and  
out of line in saying that I said what I did not say.

Michael Everson * http://www.evertype.com/


More information about the Ietf-languages mailing list