It started with a simple question posted to the NewsLib list. The ensuing
discussion hit a collective nerve and sparked a debate that went to the
heart of news database integrity.
The question I asked was "how do you handle corrections of your Web
edition." The answer, generally, was "we don't." Most libraries have
nothing to do with the Web editions, most of which are managed by non-news
people who don't "get it" or don't care. That's a whole debate in itself.
Fortunately most Web editions aren't archived for long and so the damage
may be limited. Is the story libelous? Change it or kill it.
Laws governing how corrections or retractions must be handled vary
from state to state, and that works (pretty well) when few papers
circulate widely. But what happens now that errors, once limited in
circulation, are disseminated beyond states' borders and become national,
even international, in the damage they do or the misinformation they
create? There does not appear to be any case law --- yet. But when even
the smallest paper gains national circulation simply by transmitting its
stories to a commercial distributor, the issues of error and correction
become national ones.
The real debate that emerged was not about Web editions, but about the
news databases we create and maintain daily for our own staff and, through
vendors, for searchers worldwide. There were clearly issues we needed to
sort through. And so here we are.
Cathy Tierney and Michael Jesse, Tim Rozgonyi and I are have carried carry
on what we started. Also contributing is Bruce Oakley whom I "met" early
last year when he was researching database quality and corrections,
mentored by Barbara Semonche, who put us together. His extensive research
looked at the problems we all face and the work we do to maintain database
integrity.
Most of us work hard to produce a clean, credible database
meeting all the requirements of legality. Further, we try to follow up on
how that work is presented by the vendors to whom it is sent. Vendors
have told me that they, in turn, try to screen for poor work and clean it
up, even appending corrections if necessary.
In fact, there are representatives from several database services here
today: UMI/DataTimes Publishers Services (who join AP in feeding us this
morning), Lexis-Nexis, Dow-Jones, MediaStream, NewsBank, Dialog. They want
to hear what we have to say and possibly put in a word.
Few would argue that databases should be clean and accurate (certainly not
librarians or database managers). What surprised me most was how much
disagreement there is about how corrections should be handled. Many say
they would never - NEVER - change anything in a database that alters the
way it appeared in the paper. But just as many - just as many -- say they
sometimes, or even usually, correct errors.
Each of the panelists here today will present a point of view some of you
will applaud - and some will dispute. We hope to start a dialogue that
will continue beyond this day and this room. We're not here to prescribe.
After each of us has made some opening remarks, we'd like you all to join
the discussion.
Barbara has produced a Web site we've been building on - and will continue
as long as there's interest.
We're here to open the debate wide and examine the issues, to generate
some thought, and not incidentally to educate - to try to reach those for
whom database integrity is not a priority or even an option. News
librarians, confronted by indifference or the quaint notion among upper
management that electronic means automatic, struggle against the
impression that human intervention is not required in database quality
control, that somehow "magic" happens and everything turns out all right.
In fact we all know the "magic" is sweat and toil and eternal vigilance.
And, finally, that we have the capability to control what happens to the
quality of the databases we construct.