Location references

[Paul Chandler]

I'm a medievalist working largely with critical editions of Latin texts. I used to use R.J.C. Watt's Concordance program but it was not updated for new operating systems after 2016 and has now become unavailable. It allowed the user to encode subdivisions of the text (e.g., book|chapter|verse; title|page; author|work|division1|division2 -- it was very flexible). I haven't found anything else that quite does this.

AntConc only gives a reference to the file in which the target word or phrase is found. I'd find it valuable to have a location reference of the kind that are found in the traditional concordances to, say, Shakespeare (Play.Act.Scene.Line) or the Bible (Book.Chapter.Verse).

In my case I've been working on a complex 14th-century text divided into ten books of eight chapters each; in my edition I also numbered the paragraphs. Watt's program allowed me to encode these divisions as |b, |c, and |p and returned a location reference in the form 7.3.22. I suppose I could divide the work into 80 files named 1-1 to 10-8, which would give the book and chapter references, but it seems a bit of a kludge

A Watt concordance on a more complex text I was working on allowed encoding author|work|division1|division2|division3|division4 and gave results like Anon.JVC.Pref.10, and revealed whether the search target(s) were used by which authors and where in which work. 

I understand that this functionality is not very important in much corpus linguistics, but it was a basic function in traditional concordances. I'm sure many potential users would find it valuable, as I would.

Could some functionality like this be incorporated into a future version of AntConc? Please forgive me if I am asking for something complex. -- Paul Chandler

--

Revd Dr Paul Chandler, FRHistS

Holy Spirit Seminary  |  Banyo. Qld. 4014

[Laurence Anthony]

Hi Paul,

This is probably a very useful feature to have for some AntConc users. At the moment, AntConc has two features that get you somewhere near. First, you can load a corpus as a set of rows in a table. If you do this, each row will automatically get its own file id, and can be seen in the results later. Perhaps you could load your book as a kind of table, and get the reference information that way. Second, AntConc allows you to load a table of metadata associated with file in the corpus. So, if you combine the row approach, with a metadata table, I think you could probably achieve exactly what you want. But, loading metadata tables is a little clunky in AntConc. If you search in the Discussion Group, you'll see a discussion of this that I had with Martin Wynne, where I also provide a demo:

The exact thread is here:

https://groups.google.com/g/antconc/c/Si5EdVrIddE/m/_xf8BR3pAAAJ

I suggest you start with a very short demo text (of just a few lines) and see if you can achieve what you want. I think it's possible.

Laurence.

###############################################################
Laurence ANTHONY, Ph.D.
Professor of Applied Linguistics
Faculty of Science and Engineering
Waseda University
3-4-1 Okubo, Shinjuku-ku, Tokyo 169-8555, Japan
E-mail: antho...@gmail.com
WWW: http://www.laurenceanthony.net/
###############################################################

--
You received this message because you are subscribed to the Google Groups "AntConc-Discussion" group.
To unsubscribe from this group and stop receiving emails from it, send an email to antconc+u...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/antconc/a919ccb4-1a45-42d7-acc1-d5094a051c02n%40googlegroups.com.

[Marius Amado-Alves]

"Multilevel" annotation of text is extremely complex specially when combined with corpus linguistics. It's been a hot issue for sometime. I'm working on a model and implementation but not expected to be ready before a good number of months yet. It's a polytheoretical architecture where multiple annotators of any nature -- human or machine -- can contribute, and queries across any subset be done. The Holy Grail, basically.