Meta-question about FLEx

[Claire Bowern]

Thanks for this Beth, good to hear.
My own test of Flex 7 on a netbook was not very successful. On an eeePC
frrom 2010 it was definitely faster than Flex 6, but still too slow to use
day to day. It also required all other programs to be closed (so no
switching back and forth between Elan and Flex). I'm talking about delays
that add 30-40% to the time needed for a query.
My understanding for how flex could work 'collaboratively' was that the
files could be hosted on a shared drive which could then be accessed by
multiple users (though not simultaneously). Reports I've heard from people
who do this is that "it usually works from computers in the same lab, but
not any further." I was never able to do any bulk editing without the
program crashing (this on a project originally developed in Toolbox).
Regarding the theoretical embedding in the program. The semantic domains,
for example, reflect a culturally specific framework which differ in
different cultures. There were also aspects of the grammar coding that
reflected a model of word formation that is quite specific to a particular
theory. I can't be more specific because I'm working on a mac and don't have
FLEx installed anymore (I tried it for a week or so on my Bardi materials,
with a view to seeing if it would be appropriate for my field methods class
and for interlinearising my Bardi texts, but the performance was well short
of what I needed for both tasks). I also remember FLEx asking me to make
choices about categories in the grammar early on in setting up a project
that I would have been unable to make if I hadn't been studying the language
for ten years.
More generally, the model of dictionary development is, I think, really
unfortunate, especially for linguists working on revitalization projects.
It's so hard to produce dictionaries that people will actually use with this
model! For non-fluent speakers, they spend all their time in the
English-Language section of the dictionary, not the Language-English
section. But that's the side of the lexicon that's least developed.
Claire

On Tue, Oct 11, 2011 at 9:28 PM, Beth Bryson <Beth_Bryson at sil.org> wrote:

> I'll try to add some information in reference to some of the comments
> below.  Many of the answers are, "It's coming."
>
> 1. Interoperability.  I may not be understanding this, but the goal is that
> the LIFT format would be one that a number of tools could import and export.
>  And it is possible to interact with the FLEx data structure using Python.
>
> 2. Open Source.  The project is technically open source, although in
> practice that hasn't happened.  If someone wants to participate in
> developing it, there is a very steep hill to climb, but the possibility is
> there.
>
> 3. Time alignment.  Efforts are underway to modify the model for texts to
> include a place to store the time alignment information.
>
> 4. Collaboration.  It is currently possible for more than one user to use
> FLEx at the same time over a network.  (Maybe Clare's comment was about how
> well this works, rather than the existence of the possibility?)  And efforts
> are underway to allow remote syncing of the database.  It is currently
> possible to do this remote syncing using WeSay, and between WeSay and FLEx.
>  Work is underway for more than one FLEx user to sync in that way.
>
> 5. The attempt was to be as theory-neutral as possible.  The goal was to
> make it "easy for linguists".  If there are specific places where it feels
> particularly theory-specific, have you interacted with the developers about
> how those places could be more neutral?
>
> 6. OS.  There is a beta version of FLEx on Linux, and there are users
> currently using that.  There have been discussions about whether to try to
> make the Linux version also work on Mac.
>
> 7. Lightweight.  It isn't lightweight, but certainly FW 7 requires far less
> disk space and processing power than FW 6 did.  It is being used
> successfully on Netbooks, and recently there was an experiment in running it
> on a version of Linux booted from an SD card on a Netbook.
>
> This is not to say that the concerns raised are not concerns.  But there
> are at least partial answers to most of the points raised.
>
> -Beth
>
> On Oct 11, 2011, at 4:25 PM, Claire Bowern wrote:
>
> Given that we're talking about FLEx, I have a more general question for
> list members. FLEx is billed by those who like it as the new standard for
> language description software. But at the same time, this program seems to
> fly in the face of many of the ideals for language description software that
> fieldwork technologists have been advocating. I'm thinking of articles like
> Bird and Simons' "7 dimensions of portability for language documentation".
> Flex is not interoperable, it's not based on open source, it doesn't
> preserve material on import (e.g. time-alignment), it's impossible to use
> collaboratively, and it builds the descriptive model into the software in
> the name of 'ease of use for non-linguists', so it's not theory-netural
> (whatever that means anyway).
> Does this mean that we've just given up on goals like interoperability,
> open access and collaboration? Or are we recognizing that the ideal software
> tool for language description just doesn't exist, and FLEx optimizes enough
> of the process that we are willing to live without the more abstract and
> intangible goals like interoperability?
>
>
>


-- 

-----
Claire Bowern
Associate Professor
Department of Linguistics
Yale University
370 Temple St
New Haven, CT 06511
North American Dialects survey:
http://pantheon.yale.edu/~clb3/NorthAmericanDialects/
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://listserv.linguistlist.org/pipermail/resource-network-linguistic-diversity/attachments/20111013/b304329b/attachment.htm>