by Samuel E. Trosow**
The filtering of public access computers
in public libraries has long been controversial. Pressures to limit
access to Internet content arose in the 1990's when libraries began
providing Internet access, and librarians have generally defended unfiltered
access on intellectual freedom grounds. Filtering has generated more
controversy in the U.S. than in Canada, (1)
but recent events in London suggest the issue may be heading north of
At its May 2007 meeting, the London Public
Library (LPL) Board adopted an Internet
Policy Review Project, (2) the purpose
of which was "to review the balance between filtered and non-filtered
computers to determine an appropriate balance of filtered and non-filtered
The management recommendation identified
factors to be studied including "an individual's experience in
the library in terms of unintentional exposure to visual images not
appropriate in a general library setting" and "the steps the
library can undertake in order to mitigate risk of exposure to such
images for its customers and itself."
report also indicated that LPL had "received negative comments
on an infrequent but regular basis from customers at Central and at
Branch locations about these types of incidents," and that LPL's
"mission statement and value promise assures customers that we
will provide a welcoming environment for all people, such as families
and children, and pays attention to the individual's experience in the
library." While the stated intent was to experiment with the balance
between filtered and non-filtered terminals, the default for the five-month
review period was to filter all but a few of LPL's terminals, as well
as wireless access. Terminals in the children's areas had been filtered
Public opposition to the plan quickly surfaced,
(3) and in June two trustees, Gina Barber
and Nancy Branscombe, sought to rescind the May action. The chair ruled
their motion out of order and it was matter was deferred to the September
meeting where there were additional staff reports
(4), more presentations from filtering opponents, and a presentation
from the filtering vendor.
While the Branscombe/Barber motion to rescind
was defeated 6-2, there was much more discussion around the table and
some good questions were asked of the filtering vendor. In November,
despite submissions from CLA
(5) and the
Faculty of Information and Media Studies (6)
at Western, the Board voted 5-4 to continue the project indefinitely.
In June 2008, the LPL Board received a
staff report (8) along with submissions
asking for additional information, (9)
that the Board refer the project to a third party evaluation, (10)
as well as a letter from a public interest law clinic advising the Board
of several constitutional issues that they need to consider. (11)
The board is now seeking their own legal opinion (something they probably
should have done a year ago) and there will be a further report to the
Board in the fall.
What distinguishes the London filtering
controversy from others is that the impetus for filtering came from
LPL management. Throughout the controversy, LPL downplayed the tensions
between filtering and intellectual freedom, and the important tradeoffs
involved. In other similar situations, librarians have typically opposed
filtering demands, and have sought moderation and compromise. For instance,
at the Ottawa Public Library in 2003, demands for filtering coming from
staff were resisted by management, and a compromise was reached whereby
adult users choose a filtered or non-filtered session.
How would the Canadian Courts respond to
a challenge to Internet filtering in a public library brought under
the Charter of Rights and Freedoms? Unlike the United States, where
Internet filtering has been directly addressed in several cases, (13)
Canadian courts have yet to directly address the issue.
Since blocking access to lawful Internet
content is a prior restraint of protected expression, how might Canadian
courts respond to a challenge of LPL's filtering under the Charter of
Rights and Freedoms? Under the precedents that have arisen in other
expression cases, a court would likely find filtering to be a violation
of expression rights under Section 2B. The outcome would turn on whether
filtering is justified under Section 1 as a reasonable measure "prescribed
by law as can be demonstrably justified in a free and democratic society."In
cases involving obscenity, (14) hate-speech
(15) and child pornography,
(16) Canadian courts have found the expressions as within section
2B, but that limitations were justified under section 1. There seems
to be little question that a court would also find Internet filtering
as limiting expression protected by section 2B and the inquiry would
turn to justification under section 1.
An initial issue under section 1 is whether
filtering is even a measure prescribed by law. The blocking decision
is delegated to a proprietary computer algorithm, and the public lacks
access to underlying decision rules about how it works, and attempts
to obtain this information through applicable Access to Information
legislation has been resisted. (17)
Assuming LPL could overcome this problem,
they still have the burden of justifying the limitation on expression
under what is commonly known as the Oakes-test.
Under the Oakes test, a court first looks
at the importance of the objective of the limiting measure, which must
be of sufficient importance to warrant overriding a constitutionally
protected right or freedom. The documents before the Board identified
the underlying objective as reducing the risk of unintentional exposure
of customers to images, on computer screens in the library, that are
not appropriate in a public space. While this objective is rather general
and open-ended, it's likely that a court would find this to be a reasonable
objective and they would go on to the next step which looks at the means
chosen to reach the objective.
There are three aspects to this second
prong of the Oakes test. First, there must be a rational connection
between the objective and the limitation, it cannot be arbitrary or
capricious. Second is the minimal impairment test, which considers whether
other reasonable ways to satisfy the objective that would have less
impact on the right being considered. Finally, a court would balance
the objective and the means employed to reach the objective for proportionality.
It is in these latter stages of analysis
that the London measure would face its greatest challenges. As the record
before the Board indicates, the incidents were admittedly infrequent.
A review of documents obtained under MFIPPA confirmed not only the infrequency
of these complaints, but the fact that they could have been dealt with
under existing policies. LPL will also have difficulty passing muster
under the minimal impairments test, as various less restrictive alternatives
had been suggested and were not adequately considered by the Board as
alternatives. A court would likely find that filtering adult terminals
and the wireless environment to be a prior restraint infringing section
2B of the Charter, which is not justifiable under section 1.
However, beyond the impact of a legal challenge,
there are deeper implications of subsuming questions of intellectual
freedom within the rhetoric of customer service. Embracing a strong
version of the customer service paradigm, stressing a reactive stance
to all public concerns no matter how unrepresentative or unreasonable,
cuts at the heart of the meaning of public library service and sets
the stage for the erosion of important library values. Where values
of intellectual freedom and customer service collide, the latter may
well prevail unless decision-makers have a strong grounding in, and
appreciation of, the values of the profession and the overriding importance
of intellectual freedom. (19)
this is an expanded version of an article first published in OLA Access
Magazine, Summer 2008)
**Associate Professor University
of Western Ontario, Faculty of Law and Faculty of Information &
(1) For general background information
on filtering controversies in the U.S., see http://www.ala.org/ala/oif/ifissues/filtersfiltering.cfm
. For resources on filtering in Canada, see
http://www.ualberta.ca/~aschrade/internet/resources.htm See also
Alvin Schrader, "Internet Filters: Library Access in a Cyberspace
http://www.ifla.org/faife/papers/others/schrader.htm and Ann Curry
& Ken Haycock, "Filtered or Unfiltered?" (2001) 47:1 School
(3) See letters to LPL Board from Samuel
Harris and Toni
Samek (June 2007).
(7) Another trustee (City Councilor David Winninger)
joined Branscombe and Barber in opposition, but a fourth (Joanne Tilley,
who represents London District Catholic School Board on LPL) voted no
because she wanted all terminals filtered without exceptions. The adopted
measure retains a small number of unfiltered machines.
(8) Public Computer Use and Internet Access Policy Update
(9) Letter from Samuel
Trosow (June 2008).
(10) Letter from Bill
Irwin. (June 2008).
(11) Letter from David
Fewer, Canadian Internet Policy and Public Interest Clinic (June
(13) See Reno v ACLU
http://www.ciec.org/SC_appeal/decision.shtml (reaffirming that communications
over the Internet deserve the highest level of Constitutional protection);US
v ALA http://www.supremecourtus.gov/opinions/02pdf/02-361.pdf
(upholding Children's Internet Protection Act against ALA challenge).
See also http://www.ala.org/ala/oif/ifissues/issuesrelatedlinks/courtcasesagainst.cfm
for links to other US cases.
Butler 1992 CanLII 124 (S.C.C.).
v Keegstra 1990 CanLII 24 (S.C.C).
(16) R. v Sharpe 2001 SCC 2 (CanLII).
(17) A series of records requests pursuant to the provincial
Municipal Freedom of Information and Protection of Personal Privacy
Act (MFIPPA) have been served on the LPL and are now on appeal in the
Office of the Ontario Information & Privacy Commissioner. (request
1, response 1, request 2, response2, request 3, response3)
(18) R. v. Oakes, 1986 CanLII 46 (S.C.C.),  1
S.C.R. 103. Cases applying the Oakes-test to section 2B challenges include
R.v Butler 1992 CanLII 124 (S.C.C.)(upholding obscenity provisions of
Criminal Code); R. v Keegstra 1990 CanLII 24 (S.C.C.) (limiting harm
cause by hate-speech a legitimate interest); Thompson Newspapers v Canada,
1998 CanLII 829 (S.C.C.) (banning publication of opinion polls within
three days of election not based on legitimate interest); RJR v Canada
1995 CanLII 64 (S.C.C.) (rational connection and legitimate objective
exists for curbing cigarette ads); Little Sisters Book and Art Emporium
v Canada 2000 SCC 69 (CanLII) (administrative failures in applying the
Customs Tariff Act were sufficient to grant relief and additional guidelines
were established). Little Sisters maintains an archives of documents
relating to their ongoing controversies with the government at http://www.littlesistersbookstore.com/index_court.html
. (The page was classified by Netsweeper as pornography even though
it contained no images, so it would have blocked at LPL. It was last
tested at http://www.netsweeper.com/Support/Test%20A%20Site
on April 14, 2008 with a result of 23 pornography. As of June 20, 2008,
the site has been re-classified and is no longer classified as 23 pornography.
While it is good that Netsweeper has corrected this obvious mistake,
it is not clear how this classification was assigned in the first instance
and why it remained in tact for so long.
(19) LPL Board members Branscombe and Barber have been
awarded the 2008 CLA Award for the Advancement of Intellectual Freedom.