Proposal for reordering and creating a new translation relationship.

Summary:

Currently the translation  looks like this https://gcdnb.pbrd.co/images/EQHZGUrHJb8b.png
 That is:
 $originalauthor wrote $work
 $translationauthor translated $work
This proposal will replace that with the following
 $originalauthor wrote original languagework of $work
 $translationauthor translated $work

Problem: 
Right now we cannot easily distinguish between works an author has *written* and works that are *based on* (including translations) of a work an author has written.
This leads to several problems:

* We cannot display these relationships separate in the interface (e.g. tabs)
* Semantically, these are actually separate (but adjacent) concepts, but we are serving them to the end user programmatically and visually as if they are the same (api, databases, future users)
* Adding this data is confusing for users (and it's been annoying the stylemonkey (aka apekattquest-monkeypython (me)) for aages)

Proposed solution:

We already have the "provided story for" relationship:
https://bookbrainz.org/relationship-types/author-work
 Author provided story for  Work
(Indicates an author wrote the original story that a work is based on.)

This is for a link between an author and a work based on a work of theirs; adaptations, comics, abridged and so on.

However in our work we've found that having a separate relationship specifically for *translations* is fortuitous.

So, I propose creating this, sub-relationship:

 "Author  wrote the original language work of  work"

suggested revers:
 "Work  was originally written by  Author"
 "Work  was written in the original language by  Author"

Now, this wording is a bit clunky, but so far, it's the clearest way we've come up with that states plainly how this is different from the other relationships.
If you've got a better suggestion, we'd love to hear it!


Justification:

Doing this will fix and clarify many problems
 
* api (fetching translations is hard without this, because the alternative is to fetch a lot more data which is inefficient)
* logically and semantically correct (an author has (usually) not *literally* written the words of the translated work)
* database (it will be much easier to order this data into corresponding tabs in the new tab view, if the data in the database has this underlining structure that defines them differently)
* easier for users (it should be clearer for users (especially once we update the docs) how to enter the data if we have semantically correct terms for "links between the original-work author and translated-work")