typological interface for Dogon languages project

Tue Nov 25 18:17:02 UTC 2008

On Thu, 06 Nov 2008 19:02:19 +0100, Heath Jeffrey  
<schweinehaxen at HOTMAIL.COM> wrote:

>
> The website for our Dogon linguistics project (West Africa) is primarily  
> focused on lexicon, but I have added a page including several  
> "typological discussions", briefly going over major grammatical features  
> of the Dogon languages we have studied to date, and indicating  
> differences among these languages (micro-typology). Hopefully,  
> typologists will find an informal family-level interface useful.
8< snip >8
> I would welcome feedback about our site, and inquiries about Dogon  
> languages including fieldwork possibilities (email to me directly).  
> Separately, it might be good to have a general discussion on the lingtyp  
> list about how best to construct such language family sites.

Some ideas about the requirements and possibilities for such platforms can  
be found in the following documents, which discuss the grammar authoring  
platform GALOES

http://home.medewerker.uva.nl/s.nordhoff/bestanden/ALT2007.pdf
http://home.medewerker.uva.nl/s.nordhoff/bestanden/The%20grammar%20authoring%20system%20GALOES.pdf
http://home.medewerker.uva.nl/s.nordhoff/bestanden/Growing%20a%20grammar%20with%20GALOES.pdf

There is a write-up of the values and requirements having an influence on  
the design of web-platforms for grammatical descriptions, which should  
appear in the next issue of Language Documentation & Conservation. I can  
send the draft to interested readers. It lists the following maxims, which  
could be relevant (and could be followed or violated):

(1) Every step of the linguistic analysis should be traceable to a  
preceding step, until the
original utterance of a speaker is reached
(2) Every phenomenon described should be sourced using an actual utterance
(3) More sources for a phenomenon are better than fewer sources
(4) The context of the utterance should be retrievable
(5) A GD should incorporate provisions to incorporate scientific progress.
(6) The GD should present state-of-the-art analyses.
(7) The GD should present both historical and contemporary analyses
(8) Layout should be automatic as far as possible
(9) A GAP which provides templates is better
(10) A GAP which does not interfere with the creativity of the author is  
better
(11) A GAP which does not require the writers to be present at the same  
place is better
(12) A GAP should show which collaborator has contributed what
(13) A GAP which can be used both online and offline is better
(14) A GAP should provide the author with regular automated backups
(15) A GD which has table of contents, index, full text search is  
preferable
(16) A GD which does not require internet access is preferable
(17) A GD should permit the reader to follow his or her own path to  
explore it
(18) A short path between two related phenomena is better
(19) A GD which is similar to other GDs known to the reader is better
(20) The GD should present the data in a didactically preferred way
(21) The readers should be able to know that they have read every page of  
the grammar
(22) The relative importance of a phenomenon for a) the language and b)  
language
typology should be retrievable
(23) The quality of a linguistic description should be indicated
(24) A grammatical description should not change over time to make  
reference to it easy
(25) A GD should be available in several languages, among others the  
language of wider
communication of the region where the language is spoken
(26) The data presented in a GD should be easy to extract and manipulate

A first implementation can be found on  
http://www.galoes.org/slm/Introduction . This site contains a hypertext  
grammatical description  of Sri Lankan Malay (or a beginning thereof ...).  
The structure of this grammatical description was chosen in order to  
facilitate cross-linguistic comparison. I actually had typological  
comparison in mind when I thought about the structure of the site, but an  
interface for genetic comparison as alluded to by Jeffrey could also be  
implemented, I guess.

The following pages contain interfaces for doing cross-linguistic work  
based on the LDS questionnaire and WALS:

http://www.galoes.org/
http://www.galoes.org/trees.html

The idea is to select a language and an LDS question and to be presented  
with all pages of that language's description that might provide  
information relevant for that question. Currently, you can only select SLM  
as a language and LDS as a questionnaire, but I hope that in the future,  
more languages and possibly more questionnaires can be added.

I would very much appreciate feedback on the aspects of such a web-based  
grammatical description, both by field workers and by typologists.

Best
Sebastian Nordhoff

> In principle the technology we've developed could be used for other  
> language families, but developing a new site would require some labor. I  
> myself would like to do something along these lines for Songhay and  
> perhaps Berber. The creator and administrator of the Dogon site is  
> computational linguist Steve Moran (stiv at uwashington.edu), and he is  
> better positioned than I am to answer technical inquiries.
> --Jeff Heath (U Michigan, Ann Arbor)

-- 
Using Opera's revolutionary e-mail client: http://www.opera.com/mail/