FC: D.Burt on filtering sw in libraries (he likes it)

[Declan McCullagh <declan@well.com>]

===

Date: Tue, 10 Nov 1998 12:37:28 -0800 (PST) 
Content-Type: text/plain;
charset="us-ascii" 
From: Filtering Facts <burt@northwest.com> 
Subject: My
testimony from this morning is on-line 
X-UIDL:
d08722ed8bfdfa29432ecb4f20981402 



Testimony before the National Commission on Library and Information Science,
November 10. 1998
http://www.filteringfacts.org/nclis.htm
 

Thank you very much for inviting me and being willing to allow me to
testify via
speakerphone. My name is David Burt and I am a practicing librarian. I am
currently
employed as the Information Technology Librarian at the Lake Oswego,
(Oregon) Public
Library. 

I have been an active American Library Association member since 1991. In
August,
1997, I started Filtering Facts, an organization dedicated to protecting
children in
libraries, because I was deeply disturbed by the positions taken by my
librarian
colleagues at the American Library Association regarding filtering the
Internet access of
children. 

I'm sure we're going to hear many times today that filters simply don't
work. I do not
believe the evidence supports this view. Unfortunately, there are no good,
scientific,
studies to tell us how well the filters in public libraries work, so we have
to rely on
anecdotal evidence. There are plenty of colorful, often humorous anecdotes
about the
websites that filters have blocked unintentionally. But there is no evidence
to suggest
that the filters in use today by public libraries block more than a few
dozen, or perhaps a
few hundred sites by mistake. The "exposes" of filters by the anti-filtering
activists
themselves actually support this claim. In December 1997, a group called "The
Censorware Project" examined the filter CyberPatrol's list of blocked sites,
and found
about 60 sites that were clearly blocked inappropriately. In the Loudoun
County case,
the plaintiffs claim about 100 sites were inappropriately blocked by the
filter XStop. The
defense claims the number is far less. Let's get some perspective here: the
precise
number of individual websites is not known, but it is widely believed to be
in the millions.
Fifty or one hundred sites out of millions are a tiny fraction of one
percent of the entire
Internet. It would follow then that public libraries would report few
complaints from
patrons about incorrectly blocked sites. Indeed, a survey I conducted last
year of 24
public libraries that filter showed an average of 1.6 complaints per month.
If a child using
a filtered terminal encounters an incorrectly blocked site, most libraries
have a policy of
overriding the filter upon request. This helps to frame the choice libraries
make by
filtering or not filtering for children: a library can either require
children to ask permission
to see a tiny fraction of 1% of the entire Internet, or they can expose
children to 70,000
pornography sites. 

Because of this lack of reliable data, I'd like to suggest that this
commission take the
lead in producing better data. I think that conducting a study that could
tell us what we
need to know would be pretty straightforward. Such a study would involve
writing a
special computer program that would run on Internet workstations in several
public
libraries that either filter for all patrons, or just for all minor patrons.
First, the program
would record the address of every website that every patron visited.
Second, the
program would record the address of every website someone tried to access,
but was
blocked by the filter. Third, the program would record if the filter were
overridden in any
of the cases where a patron encountered an inappropriate block. With this
method we
could actually get a reasonable idea of: 1) What exactly are patrons being
prevented
from viewing in libraries that filters, 2) How often are patrons prevented
from viewing
websites they want to access, and 3) When a patron encounters an
inappropriately
blocked website, how likely are they to ask to see it. 

I'd also like to make a second suggestion to this commission. That
suggestion is that
this commission embrace a compromise solution to the problem of the
availability of
hard-core pornography in libraries. The compromise would be that all minors be
required to use filtered Internet access, with the parent having the option
to grant their
own children unfiltered access. This solution first appeared in the Boston,
Mass. Public
Library, after a much-publicized controversy when the Mayor of Boston
ordered filtering
on all Internet terminals. In Boston, the compromise has proved to be both
successful
and popular. 

In Boston, the librarians keep a list of which children have obtained their
parents
permission to access the unfiltered Internet. Some libraries have even
automated this
policy. The Dayton-Montgomery Public Library in Ohio has worked with the
company
N2H2 to produce a customized filter that interfaces with the library's
database of patron
records. All of the terminals are filtered, and when an adult wants to have
an unfiltered
Internet session, the patron simply enters their library card number, and
the system
determines if the patron is an adult or a child. A patron may have their own
child's card
authorized for unfiltered access.

Let me say at this point that this isn't my ideal solution. I believe that
all minors' access
should be filtered, and that it shouldn't be the library's function to
provide children with
pornography under any circumstances. But I'm willing to compromise. This
isn't a
solution that will satisfy everyone, but I believe it will satisfy most
people. So I'd like to
ask this commission to endorse this type of compromise, and to ask the
American
Library Association to promote this solution as well. I believe that if this
compromise
solution were promoted by ALA and widely implemented by public libraries,
there would
be no need for future legislation.

Thank you.

QUESTIONS FROM THE COMMISIONERS:

Question 1:
What is your position on mandating federal funding to filtering for minors
as opposed
to mandating that each library only submit a written policy?

BURT:
Yes, I can address that. I supported that legislation. I don't think that
federal funding
should be used to provide pornography to children against the wishes of
their parents. I
am in favor of tying funding to filtering and I think the problem with the
local control option
is that according to the American Library Association, 85% of public
libraries already
have policies and that includes libraries like the Los Angeles Public
Library and the New
York Public Library that have reported extensive problems with children
accessing
pornography. So, this would simply ratify the existing situation and not
really do anything
to address the problem.

Question 2:
Mr. Burt, I read with interest your comments in the New York Times when you
evaluated 5 different filtering products, including CyberPatrol. Everybody
I've ever
talked to and some of our witnesses this morning have said any kid knows how
to get
around all of the filtering systems. What is your comment on that?

BURT:
Well, it depends on the type of filtering. If it's being done at the server
level, not at the
local workstation PC but at the server level, it's much more difficult for
the kid to break
into it. If the filter is actually running on the PC that the kid is using,
a smart kid, that has
some information about how to disable the filter, and is determined to do
so, yes will
probably be able to disable most filters. I don't know that we see a lot of
evidence of
kids disabling the filters in libraries. I have not read a lot of reports of
that, and I think
that's because the use of these terminals is typically monitored. A
librarian would
observe a child disabling the filter and they would be disciplined for doing
that. I think
this also points out that filtering is only a tool. It's only part of the
solution and it has to be
done in conjunction with a policy, with monitoring children's behavior. You
don't just
simply put up a filter and then just ignore the child. As a computer
professional, that's not
how I treat my workstations, because I like to keep an eye on them and see
what people
are doing, and make sure that people aren't abusing them.

Question 3:
So that at our library, if I assured the board of trustees that the filter
was placed at the
server level that, we could feel more secure that it would be more difficult
to get
around?

BURT:
Yes.

Question 4
Why is it that everybody pooh-poohs filters, if they are as effective as you
say they are
when placed on the server level?

BURT:
Well, I think it has to do with two things. First off, the first filters
that came out 3 or 4 years
ago relied on word blocking technology, and people, rightly so, made a lot
of fun of
them. They blocked out things like breast cancer and Fred Couples and that
sort of
thing. The second thing is that people tend to ridicule filters for
ideological reasons.
When you ask the people who are so against filters, "Well, what would a
perfect filters
look like? What would an acceptable filters look like to you?" they
typically answer that
there is no such thing, that no filter would be acceptable to them. So I
think their reasons
for opposing filters are more ideological than technological.

Question 5:
At our library, which is a large urban public library, if we place a filter
on a server, which
one would you recommend?

BURT:
There are several good ones: CyberPatrol, WebSense, Bess. There are a
number of
pretty good filters out there for that purpose.

Question 6:
Reading your comments in the New York Times, you say "CyberPatrol: probably
the
best overall home filter." That leads me to believe it's more appropriate
for a home
PC filter rather than a server. Is that inaccurate for me to read it that way?

BURT:
In general, that's a true statement that these products are intended more
for home use,
although this is changing: WebSense is more aimed at business, Bess is more
aimed
at schools. The thing with CyberPatrol is that it has about a dozen
categories that you
can select. I think "sex acts" is the one that most libraries select, and of
course they
would not select the "hate groups" category or the kind of thing that is
political speech
that you would not want to block, and I would not support blocking even from
children in a
public library. 

Question 7:
Can these filters be set to block chat room access?

BURT:
Yes they can.

Question 8:
But I'm told that there's all different kinds of chat rooms, for instance
the chat rooms
you get to off of a webpage?

BURT:

That's correct. There is a separate think called "Internet Relay Chat" that
chat rooms
traditionally ran on. But you're correct, as of late there is such a thing
as web-based
chat, as well as web-based e-mail and Usenet and other Internet services are
offered
over the web. So then the challenge becomes, rather than just simply
blocking out all of
chat, for the filtering company to hunt and find web-based chat sites and
block them as
web-based resources. That's a little more difficult than just simply
blocking chat.

Question 9:
Would we be able to block chat on specific webpages?

BURT:
Yes.

Question 10:
To what extent to any of the witness feel that domain restrictions, such as
the ".XXX"
proposal would apply to webpages?

BURT:
I think that's an outstanding idea. I think that would go a long way toward
solving the
problem. As I remember, and I'm not a lawyer, that Justice O'Conner
suggested the
potential for zoning of the Internet, and I think that's really the
long-term solution is
through zoning, so I think the .XXX domain is an excellent idea.

*****************************************************************************
David Burt	President, Filtering Facts
Website: 	http://www.filteringfacts.org
E-Mail:  	David_Burt@filteringfacts.org
Phone/Fax:	503 635-7048


--------------------------------------------------------------------------
POLITECH -- the moderated mailing list of politics and technology
To subscribe: send a message to majordomo@vorlon.mit.edu with this text:
subscribe politech
More information is at http://www.well.com/~declan/politech/
--------------------------------------------------------------------------