On 18 May 2021 at 17:54 Douglas Clark <clarkdd@gmail.com> wrote:

<snip>

Wikipragmatica will not need less people, it will require the same number and skills, plus more skills as an unabridged thesaurus of concepts will provide new insights into human communication.

This may be labouring the point. But when I joined the list, I commented it was interesting as a place to discuss both language and computation.

Two further use cases where the AW approach might gain traction would be abstracts of papers, and patents. I'm sure these areas have both had plenty of attention from the point of view of translation.

The problems to do with having adequate technical vocabulary are being addressed by Wikidata. Disambiguation, for example of acronyms used as index terms for papers, is already on the agenda. The issues of sentences not just having assertoric force can be seen in those contexts in a relatively controlled way.

I mentioned multiple choice questions. There is a different kind of point there, namely the lack of a de facto standard, which is essential for reuse. (Use of questions online is hardly the problem - Wikimedia places a high value on reuse of content.) 

What gets called "scholarly communications" is behind my raising the Wikijournals. Here again there is an issue of standards, often posed as "scholarly HTML", because of the lack of uniformity in the way learned journals are put online. There is a whole raft of issues that comes up when you say "write in AW code, then produce HTML later" as a publishing model. Actually, Wikipedia experience in evolving a style guide can help understand the philosophical logic side of that. Definitely a hard area.

My take on text mining is certainly influenced by the couple of years I spent working with the ContentMine group. I wouldn't expect those insights to be common ground.

Charles