Internet Filtering: a personal perspective on the issue

..reprinted from a posting to an educational mailing list regarding the filtering issue
This posting was also printed in Impact magazine (published by the Association for Teacher-Librarianship in Canada) of February/March 1998.

Hi all. A brief rise from my usual silent lurking... I've gathered quite a volume of information on filtering lately, and I just thought I would share my views now that I've had time to digest everything I've read, including all the comments through this list.

I'm very protective of open access to materials and freedom of speech in all its various forms. Applied to the Internet, the best description I saw was "delicious anarchy"; one of the things I love about the Internet is how by it's very nature and scope it is out of the reach of those who control so much of our lives already. I fundamentally believe that every culture, every person, every idea has the right to a voice even if that means sometimes being subject to voices we don't always wish to hear. As individuals we have a right to make our own choices; hopefully as we grew up we were given the tools to evaluate and make appropriate decisions - appropriate being a very relative term in our world today.

That having been said, you can understand that I went into my research with a certain amount of bias although I made my best attempts to be objective, as any of us would. I looked at the types of Internet filtering that are available and how they work. I also considered the various environments in which Internet filtering could be/is being applied. One thing I noticed quite quickly is that filtering technology is advancing at breakneck speed and a number of the things I had previously understood about filtering were very out of date. Here is a short summary of what I found:

There are two main types of Internet filtering in use: hardware filtering and software filtering. Hardware filtering involves the actual configuration of routers to block the flow of information of specific types or from specific locations. Hardware filtering is most often used for security rather than "filtering" since it is extremely labour intensive to maintain.

Software filtering can be done either at the client level or the server level. At the client level, a piece of software is loaded and run at the desktop. To update the filtering mechanisms requires periodically loading disks or connecting to the vendor via an Internet site. In either case, it must be asked: "Can the user override the software if they feel it is necessary?"

Very often used by parents to lock up their machine at home (NetNanny is an example), client side filtering software isn't very practical in the case of a lab since every machine would need the software loaded and configured and continuous updates done locally to each station. Network versions are now becoming available that are of greater use in a networked environment. Advantages of client side filtering software include: it can be specific to an individual workstation and is easily implemented or discontinued at any time. Some of the disadvantages if client side filtering are that it is fairly easy for the local user to bypass/corrupt the filtering software, updates are often very labour intensive, and in organizations there are usually security issues for updates which require open access to the system by software vendors. There is also the need for additional space and memory on the local machine to operate another piece of software.

Server side filtering software runs on a server at a remote location. Since all access to the Internet anywhere runs through a server somewhere, it is simply a matter of directing all the access through a specific server where the software is loaded. Advantages of server side filtering include: elimination of the incidence of bypassing/corrupting the software, no additional space/memory required, and updates are done at a single location for any number of workstations; there is very little or no maintenance required at the desktop level. Used in combination with proxy technology, there is also an increase in bandwidth availability since visited sites are stored on the proxy server. If additional requests arrive for the same site, the proxy server does a quick check for an updated version of the site and if there is none since the stored version, the cached copy is sent rather than loading it again through the Internet.

Drawbacks of server side filtering software include: there may be a loss of control by individual users since they may not be able to disable the software (depends on the filtering software product being used), traffic jams at remote server(s) if they are unable to handle high volume of simultaneous access from a large user base. If the remote server is entirely controlled by the vendor, there is also a possibility that they will not be responsive to the various needs of particular environments (speed of updates, modifications to blocking/unblocking sites, etc.).

How does filtering software actually work? This is where things get really sticky. Various methods are employed by each brand of filtering software in varying degrees, and it is often the degree that defines how useful the filtering software actually is. To be of any value, every piece of filtering software in existence needs quite a bit of configuring to adjust how they block access - whether locally at the desktop or on the remote server. The methods the software use (and will need to have modified to match individual environments) include keyword blocking, rating systems, protocol blocking, and host blocking.

Being the oldest method, by now we've all heard or experienced the horrors of keyword blocking, so I won't go into depth here. It takes a list of words, which have been marked as "inappropriate", and doesn't allow anything containing those words to come through to the user. It is indiscriminate in it's blocking and is basically useless if you hope to do any research at all.

A number of organizations are in the process of developing content Rating systems for Internet sites. The site author would include ratings in the documents and filtering software would check the rating before delivering a site to the desktop. Some problems: will the authors be honest or diligent in including their ratings? Would it ever become governmentally controlled (horrors!)? Could it be used against individuals/cultures/groups to block their ability to publish their voice freely?

By invoking Protocol blocking, an entire service is inaccessible. If you don't want anyone chatting you block the IRC protocol. If you block the news protocol, you don't get news, nor can you view the archives of a newsgroup. It is a sweeping block that is more often used to manage access from a resource allocation perspective rather than a "filtering" perspective, but it is done, especially in schools.

Host blocking is by far the most popular and useful method invoked by filtering software. Entire hosts, pages, or sites are purposely selected for exclusion by an individual who then creates a list of these, which is used by the software. The biggest advantage here is that somebody has intellectually made a decision to include or exclude access based on a set of predetermined criteria. Questions that arise with host blocking are who is deciding, what is their background, what are the criteria, and who developed the criteria? How often is the list updated (sites can change/move/appear hourly) since it is extremely labour intensive? Anyone who currently maintains a simple list of links on a web page can easily imagine the amount of time it would require to keep a list for a piece of filtering software! And finally, host blocking is reactive rather than proactive and there's no way around it. The site can't become blocked until somebody finds it....

In a nutshell, that's how it all works. A condensed version to be sure, but hopefully that will give you an idea of what I looked at from a technical perspective. There is of course, the non-technical perspective: why is filtering software gaining popularity anyway? Doesn't it smack of censorship? Shouldn't we be fighting this tooth and nail to protect intellectual freedom?

Its popularity is due to a lot of bad press regarding the dangers of the Internet - all those evils lurking behind every click of the mouse. Parents are concerned, and rightly so. In less than three clicks I can download some of the most disgusting and dangerous material ever produced by the human race. I don't usually ever see it though because I'm a trained searcher who can usually recognize when something is going in that direction - I'm either ready for it or I change direction (let's be honest here!).

My general answer to whether it is a form of censorship is "yes", it does begin to infringe on intellectual freedom in relationship to the big global picture - and heaven help us if they ever succeed in their attempts to filter the Internet itself at higher levels. My specific answer is "no"; it is not infringing when used in certain environments and situations; it is responsible information provision.

Looking at our society in general, certain rights and responsibilities are selected/deselected for our young people - there are age restrictions for viewing films and videos, accessing certain establishments, purchasing certain items. It doesn't mean nobody can, it means certain individuals can't until specific criteria are met. I don't view that as harmful, but as part of growing up and becoming mature enough or having enough background to handle the responsibility and consequences. Society has deemed certain levels at which that happens (although I don't believe the levels set are perfectly accurate, they are a rough estimate to use as some kind of baseline). I'd rather have that than a free-for-all where anything is available anytime to anyone at any age for any reason.

Coming from a library background, the idea of selecting/deselecting information is not at all new; every library does it. The information collected in any library is acquired according to a preset policy that outlines the type of materials that should or should not be gathered. It isn't a judgment of what is "good" or "bad"; it is a matter of what is appropriate to that environment. Libraries are one of the biggest defenders of intellectual freedom and the freedom of expression - they may "deselect" certain materials, but they will defend to the death the right of someone to produce it.

In the same way, I don't believe the *Internet as a whole* should be "filtered". Nothing would cause me more fury than to suddenly have my general ISP tell me what was appropriate for my home, my culture, my life. If I want access, I want it to stay wide open - because I am an adult who still has that right. What I also believe however, is that wide open access to anything, for anyone, at anytime should not be available in certain environments.

As for the idea that we won't be able to do our research if our environment is filtered, let me say that I am convinced that there is no shortage of information on the Internet; we won't run out of answers to our questions just because some sites are not available, especially if we were able to override the filter when it really came down to it. Anyone right now who only uses one or two search engines are already cutting out much of what is available, and I rarely hear them complaining...

After the research I've done over the past month, I do feel that schools and libraries are two of the environments where implementation of filtering software is appropriate, but filtered access from these environments must be highly configured and modified to suit their needs - no piece of software can do that without a lot of input from the stakeholders involved. There has to be a lot of preplanning, a lot of cooperation, and a lot of understanding to implement and operate any form of Internet filtering option. Once implemented, there must be a strong commitment on the part of the organization to continually check to ensure the filtering is still meeting the needs of their environment and if it doesn't that they will take steps to quickly change the situation.

After many long hours, that's where I stand. I still love the "delicious anarchy" of the Internet, but I think we have a professional responsibility while young people are in our care to dish out that anarchy in gradual doses for their own safety and positive growth - this is not a new idea, just one being applied to a newer situation. I don't believe we should lose sight of how ordinary all of this really is.

Just as with most things I have an opinion on, it is always relative in some way and subject to change with the discovery of new information. Thanks to all of you for listening; hopefully I've tweaked a few thoughts ;)

Cynthia Beuselinck
November 1997

No portion of this document may be reprinted without permission of the author