[Standards] Fwd: Standards in the view of LINDAT/CLARIN (CZ)

[Dieter Van Uytvanck]

Dear standards committee,

as agreed in Vienna we have distributed the position paper to all the
centres and the centre committee. There were 2 reactions:

- one from Brian McWhinney (Talkbank/CMU):

We have no Centre Committee representative for CMU/TalkBank, but I think
that the recommendations of the standards committee make entirely good
sense.  However, it is very helpful to have the incorporation of
standards be accompanied by the provision of tested tools for
implementing and using those standards.  Conversion of CMDI data to
OAI-PHP format for harvesting would be an example.

- one from the Lindat colleagues (see forward below, will be included in
the minutes of yesterday's meeting at
http://www.clarin.eu/content/centre-committee)

best regards,
Dieter

-------- Forwarded Message --------
Subject: Standards in the view of LINDAT/CLARIN (CZ)
Date: Tue, 21 Apr 2015 19:07:27 +0200 (CEST)
From: Jan Hajic <hajic at ufal.mff.cuni.cz>
To: Dieter Van Uytvanck <dieter.vanuytvanck at mpi.nl>, Lene Offersgaard
<leneo at hum.ku.dk>
CC: Pavel Stranak <stranak at ufal.mff.cuni.cz>, Eva Hajicova
<hajicova at ufal.mff.cuni.cz>

Dear Dieter, Lene,

as agreed on today's call, I am forwarding our common view on the
standardization effort.

Thank you,

Best,

-- Jan

Position on the Standards recommendation

We find we can agree with most of the Position Paper and indeed, some
action is needed. We however disagree that the action needed is for
"CLARIN BoD to ensure the use of these recommended standards in the
national consortia." BoD enforcing compliance of the national consortia
who will in turn enforce compliance of users / researchers is not the
course of action we want to support. Above all because experience shows
us that most of the successful standards in our field come from the
bottom-up initiatives and cannot be predicted and planned. We have all
seen many top-down standardisation attempts with lots of effort and
expenses, often ISO standardisation, and in the end almost no traction.
ISOCat is just one of many, we could probably all name many more.

So we believe that Clarin should remain benevolent in leaving it up to
discretion of authors and requests of user community ( = professional
researchers) what formats are used for producing, archiving and using
resources. What Clarin can do and what we would wholeheartedly support
is encouragement and propagation of the standards we like: Write Q&As
for novice researchers on how to record and store data, how to annotate,
etc. Argue, why it is good to use this format or that and what
methodology. Suggest good and reliable libraries. But ultimately leave
the choice to the – now well informed – researchers. They are
professionals too and we should respect their choices about their data.

In summary, we argue for the action to be producing howtos, tutorials
and other materials in support of what we believe being good standards,
and publishing and spreading these. But let us also respect that we
cannot and should not attempt "to ensure the use of these recommended
standards".

For LINDAT/CLARIN:

Jan Hajic
Pavel Stranak