Thoughts about filtering

Just Give It To Me Straight:

A Case Against Filtering the Internet

T. A. Callister, Jr.

Whitman College

Nicholas C. Burbules

University of Illinois

Clearly, one of the most controversial and contentious issues surrounding the use of new information and communication technologies, especially in schools and libraries, has to do with whether or not we should filter students’ access to the Internet. The courts have decided that schools and libraries can be required to do so, but not that they must. In this article, we will argue that, with very few exceptions, they should not.

Before we make our argument, however, it might be useful for us to explain something about our backgrounds. We have both worked in institutional settings as teachers: one as a pre-school teacher and one as an elementary school teacher. We are now both professors of education whose interests and areas of expertise include the study of technology and education. Both of us hold strong positions about issues of civil liberties, free speech and expression, and the rights of students to learn and have the opportunity to learn. And finally, we are both fathers of young children. We want our children to experience success and happiness, we want them to have every educational opportunity, and, of course, we want to protect them, as best we can, from the harmful things in this world. So we understand full well the sentiments of parents and educators — especially those who have little experience with the Internet — who regard it as a strange and threatening environment, full of pornographers, child molesters, and political wackos. We want our children to have the educational benefits of the Internet, but to be protected from what is harmful or dangerous, and this is what filters promise to do.

And we should say up front that parents have every right to impose restrictions on what their own children view or do on the Internet at home, just as they have the right to limit what their children watch on television. What one family judges as entertainment, another judges as smut or a waste of time. But schools and libraries have wider responsibilities, in this as in other potentially controversial areas, to expose students to a broader horizon of ideas, experiences, and points of view; what counts as educationally worthy is a matter for public deliberation, and restrictions may be suitable here as well (children can’t check certain books out of the library, for example). But Internet filters, as we will see, work in a different way: they are indiscriminate, often arbitrary, and they remove decisions about what is and is not filtered from the domain of public deliberation, placing it in the hands of automated procedures and criteria developed by invisible and unaccountable programmers who, for commercial reasons, have a fundamental interest in erring on the side of filtering out more rather than less. From the standpoint of public education, this inevitably leads to abuses and anti-educational effects.

What is a “filter”?

Let’s start by looking at the metaphor of “filtering” itself. On the face of it, the idea seems perfectly reasonable. The “good” is allowed in and the “bad” is kept out. Just think of the many things we filter every day: we have oil, gas, and air filters in our cars, to keep out impurities; we use a filter to make coffee in the morning, to separate out the liquid from the bitter grounds; many people who still smoke use filters in the belief (encouraged by cigarette companies) that it is less harmful to them — though in this case it is what the filter lets through that is most dangerous. This range of examples should already start suggesting questions about how people decide what is “good” and what is “bad,” and whether such complex judgments can ever be entrusted to an automated technology.

But notice that filters work in other ways as well. The popular computer program Photoshop, which allows us to manipulate digital graphics, has a menu bar that contains filter. With this function the user can make changes to an image, like “sharpen,” “blur,” and “distort.” So, in this context, to “filter” means to change.

Another kind of filtering can simply be to separate and sort —not in terms of good from bad necessarily, but by rules and criteria put in place by an individual. Many email programs now allow the user to set up filters that automatically sort mail into various folders — perhaps personal messages go into one folder, work messages into another. Unwanted “spam” may be sorted out and transferred to the Trash. But what is “spam” to you may be very interesting to me. This kind of filtering alerts us to the importance of the criteria by which sorting and selecting is being done, and — when it is operating invisibly and automatically in the background — the dangers of finding out only after the fact that something important was lost because the filter misassigned it to the wrong folder. We will return to this idea in a moment.

Filtering the Internet sounds reasonable and benign, therefore, when we think of it in the first sense — removing the bad content, allowing only the good to pass through. But when we think of filtering in terms of changing or distorting information, or making invisible decisions about sorting and separating information, without our control or knowledge, it is less clear that this is what we want in schools.

Meanings of “protection”

Another key, unexamined term in this debate is “protection” itself. Because schools are generally thought to operate in loco parentis, it is easy to extend the justification of parents protecting their own children, to the expectation that schools must do the same. And, as noted, with the very young, who do not know what to make of certain upsetting things they may find on the Internet, who do not know yet how to make certain informed and sophisticated judgments for themselves, and whose educational interests are much more limited, filtering may in fact be justified. But this language of “protection,” so benign, so apparently well-intended, gets overextended to students generally. “Filters are there to protect students.” Who can argue against that?

But when we look at the situation through a different lens, it appears that protecting children may be less of a factor than protecting others in the educational realm. Filters are a way of protecting teachers from the upsetting nuisance of dealing with unpleasant or controversial topics in the classroom. Filters are a way of protecting school administrators from having angry phone calls (even lawsuits) from parents concerned over occasional instances where students go to “bad” places on the Internet. Filters are a way for adults generally to avoid the hassle of dealing with instances of student misconduct, after the fact, by attempting to forestall the act before it ever occurs. And, notably, filters are a way for filtering companies to protect their profits, which are based primarily on customer satisfaction. It is far less likely that products will be returned, or go unsold, because they filter too much, than because of a single instance of filtering out too little — and so the imperative is always to filter more rather than less.

Finally, the idea of filtering seems to imply protection from what may be harmful coming in. But filters are, if you will, two-way operations: they block what comes in, but by that very action effectively block questions or inquiries going out. They control not only the attempts of “dangerous” outsiders from reaching an audience of young people; they control and limit the attempts of young people to reach and out and explore the corners of this new learning environment. Once again, the merits and tradeoffs of doing so are something we can debate; but viewed in this light, the language of protection seems less paternalistic, and more about the protection of adult interests and concerns. How differently does this issue appear if we say, “Filters are there to control students?” Some of the same debates may arise, but now in a new light, and with some of the implicit assumptions at stake laid more bare. (Oops, well that last phrase will probably get this whole essay filtered.)

Meanings of “harm”

Another loaded term in this debate is what students are being “protected” from — harm, of course. Filters prevent them from encountering things that are judged to be bad for them. But what harms are we talking about here? Emotional distress, in some cases? Corrupting influences, perhaps? Encounters with people who would exploit them? These are plausible, but worst-case scenarios. More commonly, what is judged “harmful” is what makes adults uncomfortable, or what they simply consider “inappropriate” for the student. These judgments may or may not be justified, but this language — of “protecting” the young from “harm” — shields such judgments from scrutiny or question: of course we should keep children from things that could hurt them, but this assumes what needs to be demonstrated. Is something that adults consider distasteful, offensive, controversial, or upsetting necessarily harmful to kids, or does this simply mask the implicit value judgments being made under a more neutral sounding term? Of course none of us wants to harm children. But that begs the question.

We would point out here that “harm” is taken to obviously include pornography, obscene language, and so on, but not for example, pop-up ads or corporate sites aimed at marketing products to children (which could also easily be filtered). Should these be considered dangerous or potentially harmful to children? That might be an interesting conversation to have. But the fact that our society is preoccupied with the potential “harms” of purveyors showing pictures of naked breasts to children (which are routinely on display on beaches and in villages in many places around the world) — but not preoccupied with the potential harms of purveyors trying to sell them McDonald’s food — says much more about the values of our society than it does about what in fact might do children significant or lasting harm.

How do filters work (or not work)?

Filtering programs have clever names like NetNanny, CyberSitter, SurfWatch, CyberPatrol, and X-Stop — names that suggest either trustworthy and benign oversight (someone who cares for your children in your absence) , or tough, non-nonsense protective services (get them before they get you). They work through one or more of the following three strategies:

The first strategy uses key word or image analysis. The filtering software analyzes sites for words, phrases, or images that are deemed objectionable. Categories may include sexual or violent content — though this can go wrong. The Digital Freedom Network reported that the filters blocked the web site of the former Majority Leader of the U.S. House of Representatives, (and a staunch supporter of filtering), Dick Armey.[1]

The second strategy is content analysis by the software company. Many companies have their staff constantly reviewing sites and designating them as either suitable or not suitable for young people. It is also common for these programs to include features that allow the software “administrator” (a parent, a teacher, a principal) to program in additional words, phrases or web sites to be blocked. Of course, these are people with a point of view. According to Nancy Kranich, president of the American Library Association, one program blocked the main page for the Democratic Party, but not the Republican Party; Handgun Control was blocked, but not the National Rifle Association.[2]

The third strategy is self-identification. Some web sites voluntarily identify themselves as not appropriate for young people. This method is obviously fraught with problems. It is essentially asking the fox guard the henhouse. It also highlights one of the core questions in this whole debate — just what does count as objectionable, and who decides that?

In December 2000, Congress passed the Children’s Internet Protection Act. This legislation required schools and libraries to filter the Internet if they wished to retain their federal technology funding. The law was challenged on constitutional grounds by a coalition of plaintiffs including the ACLU and the American Library Association. In May 2002, The United States District Court of Appeals for the Third Circuit imposed a permanent injunction against enforcement of the law. The Court’s ruling, unfortunately, only addressed the law’s effect on libraries; public schools remain subject to the law’s provisions. [UPDATE]

Our concerns with filtering, however, have less to do with issues of constitutionality and legality than with our educational goals and purposes. After all, even if courts rule that it is legally permissible to require filters on computers, that does not mean it is a good idea to do so; or if it is acceptable for some young and vulnerable students, in some situations, that it is a good idea more generally. We contend that in the interest of providing students with an education that is democratic, intellectual, and personally meaningful, there are at least six reasons not to filter their Internet access in school.

Reason One: Filtering is anti-educational.

As we’ve said, parents have every right to filter their children’s Internet access at home. But schools and homes are different places. They have different norms, and to some extent, different values. At home, for example, we may teach our children to share and help each other. But that same activity, in the classroom, is often seen as cheating.

Schools are places where the education of children and young people is accomplished both explicitly and implicitly. Filtering is anti-educational in its explicit manifestation because it prevents students from accessing certain materials that they might find important, interesting, and relevant to their learning. Perhaps more important, filtering is anti-educational in its implicit messages about what adults think about education; it promotes a notion of education steeped in the importance of obedience and acquiescence, while compromising opportunities for independent student questioning and discovery. It manifests a distrust for students and in many cases an exaggerated sense of their vulnerability. As a result, filtering operates counter to what students need to learn in school — to discern, discriminate, synthesize, and evaluate. How can students learn to be responsible, to make good social and intellectual choices, if those choices are made for them by filtering the information they can and cannot access? It is difficult to teach young people self-control and judgment by denying them access to those things about which they need to exercise judgment.

An important aside is necessary here: Although it is most often framed in terms of protecting children, a proliferation of filtering software will have the general effect of censoring content for everyone. Our guess is that many of filtering’s most fervent advocates aren't drawing clean lines between children and adults — they don't want the sites they find offensive available to anyone, including adults. (Hence their lack of discrimination between schools and libraries, where most users are adult.) In this regard, it is also important to point out that a school’s Internet resources would not only be filtered for students, but for their teachers as well.

Reason Two: Filtering software does not work.

Filtering software does not work in the way it is advertised. In one sense, this alone should end the debate. Filtering software too often blocks legitimate sites, and often does not block the kinds of sites that it was intended to filter in the first place. There are hundreds of examples to be found on any number of anti-filtering sites on the Web,[3] many of them blocked, of course, but here is an extended example from the Censorware Project[4] to add to the one about Representative Armey.

The Utah Education Network (UEN) is, according to their web site, “a publicly-funded consortium providing Internet access and supporting educational technology needs for Utah's public and higher education institutions, public libraries, and state agencies.”[5] They are upfront in stating that Utah schools use filtering software (they call Internet filtering “Internet Content Management” — a euphemism that should make anyone immediately distrust it) and have done this for “many years.” The software used by the UNE at the time of the study we discuss below, was Smartfilter, a popular commercial filtering program.[6] The way the software worked is that it would examine each request for a web page from an individual computer and compare that request to an encrypted list of unacceptable Internet addresses determined by the Secure Computing Company, the maker of Smartfilter. Sites that were unacceptable (apparently determined by both human analysis and computer analysis) fell into the broad categories of: criminal skills; hate speech; drugs; gambling; and, of course, sex. The software also kept a log of each rejected request that listed the name of the requested site and the objectionable category into which it fell.

In 1998, the Censorware Project Organization (“Censorware” is what anti-filtering advocates call filtering software — not “Content Management”) was able, after much resistance from UNE, to obtain the logs for Sept. 10 through Oct. 10, 1998 generated by Smartfilter and kept by the UEN. Here, from their report, are a few selected examples of some of the things students and library patrons tried to get from the Internet, but were unable to get, because filtering blocked their access. These are actual requests for web sites or documents on web sites that were denied:

Under the category of Criminal skills:

The Krusty the Clown tribute page (from the television show, “The Simpsons”).

An e-zine (electronic magazine) about “modern Marxism.”

The Declaration of Independence

The complete texts of famous works including:

The Holy Bible

Moby Dick

The Book of Mormon

The Koran

The Complete Adventures of Sherlock Holmes

Dickens’s Christmas Carol

And many others

Under the category of Hate Speech

www.hatewatch.org — the best-known anti-hate speech site on the web

Under the category of drugs

Many sites discussing the debate over legalizing marijuana

The Earth First! environmental group

Corona.com (the beer company)

Under the category of Gambling

The History of Nevada website

Anything to do with a casino

The Instructional Systems Program at Florida State University (http://mailer.fsu.edu/~wwager/index_public.html)

Under the category of sex

The official “Baywatch” television show website

www.Birthcontrol.com

Dozens of news sites that contained any mention of the Starr Report on President Bill Clinton and Monica Lewinsky

www.mormon.com [7]

This same commercial software used in Broward County, Florida banned any sites having to do with vegetarianism or information on breast cancer.[8]

Admittedly, questions can be asked about the educational centrality of some of these sites (Baywatch?). But the issue here is about unintended consequences: you mean to go after X, but inadvertently pick up Y too. How does www.mormon.com get blocked for sex in Utah? How does the Declaration of Independence get blocked under the category of criminal skills? In some instances we can figure it out: the Florida State site was apparently blocked because the letters w-a-g-e-r (wager) were in the site’s address. The Declaration of Independence, Wuthering Heights? Other classic works? They’re are on the site wiretap.area.com which contains the text of hundreds of out of copyright books, governmental and civics materials, religious materials, etc. To avoid having to make careful discriminations, the entire site was simply blocked. The others? We don’t know. SmartFilter, like most filtering software programs, keeps its lists of unacceptable sites, and its reasons for blocking them, private.

Needless to say, a great deal is lost educationally when a student cannot access information about Marx, anti-hate speech, Wuthering Heights, or birth control. Moreover, the general pattern of what gets “accidentally” blocked tends to have a biased, partisan effect itself — it isn’t arbitrary or neutral. Sites that are in any way unconventional, controversial, or (by some standard) “radical” or “extreme” are more likely to get picked up by filters that may not be directly looking for those kinds of sites (because of words that appear on those sites, links they may have to other sites, etc.). Some parents, or groups, may be just as happy to see these sites filtered too — but this is not what the filtering software was intended for. Here is an example:

Consumer Reports[9] tested several different filtering software packages and found that two of the most popular programs blocked access to: the Citizens Committee for the Right to Keep and Bear Arms; to Lesbian.org; to the National Institute of Drug Abuse; and to the Southern Poverty Law Center. One starts out meaning to filter dangerous or obscene material, but ends up filtering substantive ideas, information, and points of view, and not just in a random manner, but by and large in a way that reinforces the safe and conventional, and disadvantages anything “on the edge.” Is this what we want our schools to do?

Although much of the discussion here and in the popular press has focused on examples of what gets wrongly blocked, filters also fail on the other side of the coin — what doesn’t get blocked. Out of curiosity, and in the context of writing this article, one of the authors put his favorite Web search engine into “family-friendly” (filtering) mode and typed in a crude synonym for breasts. The first site listed had just that. In another example, “One filter, at full settings, blocked a government brochure on the dangers of cocaine and let through a site describing in full detail how to make cocaine.” [10] Similar examples abound: Filters block too many things they should let through, and let through things they should block. They do not work as advertised. The use of filtering software instills in adults a false sense of security. It is like the example of cigarette filters, discussed earlier — you might feel better about smoking, but a lot of “bad” stuff still gets through.

Finally, filtering doesn’t work because it is so easy for savvy students to get around it. Especially when young people are pooling resources and sharing what they find, there is no technological solution that will prevent them finding something if they are determined to look for it. If you can get to Peacefire.com[11] (and if you can’t, your children or students probably can, even though many filtering programs block access to it), you can download a small bit of software that disables many popular filtering programs. As the Peacefire site used to state on the header of its web page: “It's not a crime to be smarter than your parents.” We need to be realistic — what we may not see going on is probably still going on. Thinking we can keep young people from sites we don’t want them to see simply by installing filters is whistling in the dark. If it works, it works only for the very young or the technologically naïve. The nature of the Internet is to expand access to information of all sorts, “good” and “bad”; and because its basic ethos is of openness, any attempts to filter, partition, or censor it will be met aggressively by skilled programmers and web site developers, somewhere.

Reason Three: Filtering is censorship.

At a deep level, the debate over filtering reflects conflicting ideas about liberal democracy and the importance of open public spaces (including the Internet). In this country, there is a widely held (but not unanimous) belief that in a liberal democracy people should be free to read pretty much whatever they want, and that if we err it is better to err on the side of allowing too much than too little. Most Americans mistrust delegating to authorities any judgments about what they may or may not read, think, say, or believe. They disapprove of the kind of censorship where individuals or small groups of like-minded people, who judge themselves wiser or more pious and pure than the rest, claim to see dangers in literature and educational materials that others cannot, and so try to have those materials withheld not only from their own children (which is certainly their right) but also from everyone else’s children. We are all probably familiar with many of the books they target: Huckleberry Finn, Catcher in the Rye, the Impressions reading series, anything written by Judy Blume, and now, of all things, Harry Potter. These are not the groups society should want adjudicating Internet filters too. In fact, for many young people in many parts of this country (and elsewhere in the world), the Internet is the only window they have to wider horizons of belief and possibility, beyond the tastes and prejudices of their own local community — which is exactly why there is such a struggle to limit, control, and censor it.

For those individuals and groups of people who find sex or sexism, witchcraft or indoctrination in secular humanism, hiding between lines of text or lurking in the recesses of the illustrations in children’s books, the advent of free and open access to the Internet is a nightmare. And as censors have always found, it is easy to begin with the examples that are most egregious, where the risks are most easily documented, and then gradually over time extend the criteria to include more and more that offends. On the Internet, as in any public space, there is certainly much to offend (although children can see obscene graffiti on the streetcorner too). But the notion that the Internet is somehow awash in pornography, child molesters, and bomb-making directions is an alarmist characterization that has been foisted upon parents and a public who, by and large, have had little direct understanding or experience with the Internet and so tend toregard it overall as strange and threatening.

The typical student turning on his or her computer and connecting to the Internet isn’t instantly bombarded with an enormous amount of unsolicited "information." Rather, users typically search for what they want — a specific piece information, a graphic, the best price for a new music CD, or even better, a downloadable MP3 file. Navigating the Internet is not like switching channels on a television set where a sudden change of the channel may bring something unexpected or shocking. Navigating the Internet is more like walking down a long corridor where all the doors are usually quite well marked. The two authors of this piece have logged many thousands of hours on the Internet and visited uncountable web sites. But on only a couple of occasions have we unexpectedly happened upon something pornographic or obscene. Our anecdotal experiences aside, according to the Online Computer Library Center, “adult content” exists on only an extremely small proportion of the Web (about two percent of free public sites, they claim, contain sexually explicit material).[12] So for those easily offended, a bit of free advice: If a link says, in large, flashing, red, capital letters: CLICK HERE TO SEE HOT TEEN SEX, and you don’t want to see hot teen sex — don’t click! The chances of a young person who is not looking for such material finding it accidentally is negligible — and this small risk needs to be weighed against the demonstrable shortcomings of filters.

Of course, many young people are eager to look for such materials, which reverses the metaphor of who is being filtered from whom. But as we have said, it is as difficult to prevent them from finding such material on the Internet, if they are determined to do so, as it is to prevent them from finding it in other venues. If we are to deal openly and realistically with their curiosity, we will need to come up with better strategies, which may include acknowledging and discussing their curiosities instead of seeking to ban them.

In the end, we believe, it is the responsibility of educators to provide students access to the greatest amount of appropriate educational material. It’s the idea John Stuart Mill had in advocating a free “marketplace of ideas.” It’s the notion that Marxism and, yes, even vegetarianism, are important topics for students to read, think, and argue about. Censorship is the tool of propaganda, indoctrination, small-mindedness, and ignorance; it is antithetical to educational opportunity, free expression, and intellectual inquiry in a democracy.

Reason Four: Filtering is deceptive.

Filtering, we have tried to show, is often capricious and unpredictable. But even if it worked perfectly, in the sense that it only filtered what it was intended to filter (according to someone’s definition of what deserves to be filtered), there would be another educational problem, intrinsic to the idea of filtering itself.

When filters do not let the user know that material has been filtered, or why — or when software companies refuse to release the list of sites they block, or the criteria by which they are selected — the absence or silence created by the act of filtering itself gives a misleading view of the domain of knowledge. When sites about certain topics are withheld from students, those students are prevented from exposure to and consideration of a range of ideas — ideas that have often been blocked arbitrarily. This point should be pretty obvious by now. The deeper problem is that students may be unaware of what it is they have been prevented from seeing. It is one thing to know some information exists and to be denied access — the book you can see but can’t check out. It is quite another thing for information to be hidden. Filtering prevents students from knowing enough to even have the opportunity to ask questions about what they have been prevented from seeing, reading, or thinking.

Here is a variation of an example we’ve used elsewhere:[13] A student searches the Internet for references to vegetarianism finds none. What can this mean to the student? Does it leave the impression that vegetarianism is not of sufficient importance to warrant any entry? That there are no vegetarians with web sites? Perhaps she is savvy enough to know that there is such a thing as vegetarianism and recognize that not finding it on the Web means that it has been filtered (goodness knows why). Does this leave an impression that there is something wrong with vegetarianism, and this is why it might have been blocked? The student can’t find out even enough to explore the question.

The absence of information, in this sense, conveys certain implicit messages itself. Nazi sites are blocked: does this lead the student to believe that there are no Nazis any more? This problem is exacerbated when the actual content is in no way wrong or dangerous, but has now been tainted as such because “if it is filtered, there must be a good reason.” This is why we say that filtering is deceptive.

Reason Five: Filtering is distorting.

Those who advocate filtering often seem to have a very simplistic conception of the nature of knowledge and understanding. Each fact or belief is something that can be evaluated as true or false, “good” or “bad,” and so filters should sort out those that are undesirable. But as we have argued, filtering is not only a matter of separating good from bad, but also of changing and distorting. Filtering — again, even when it is “effective” — presents to the student an incomplete, haphazard, obscured view of the world, a view fashioned by the vagaries of others’ judgments and the flawed heuristics of less than intelligent computer software.

The distorting effects of filtering occur because knowledge isn’t composed of discrete things that can be sorted or filtered; knowledge systems are linked, interdependent beliefs (just as the Internet is a hyperlinked system). When sites get blocked (even if appropriately so), other sites may be blocked because of their casual relationship with those subjects. A filter, in the course of blocking material of a sexual nature, for example, might also block access to related information about gender, women’s health care, or issues of equality in women’s sports. Baby pictures get blocked because the predominance of skin-toned pixels in the graphic tells the filter it is “nude.”

These filtering trends take the wealth of information available on the Internet, and then start chipping away: The filtering software company removes sites for reasons we can’t know because they won’t disclose them; the software blocks additional sites because it cannot understand the subtleties of ordinary language or the intent of the user (not everyone searching for “swimsuits” is looking for pictures of scantly clothed women);[14] the teacher, the principal, the school board, and the parent each have concerns that require additional restrictions, a little more of this, a little more of that. And what’s left? A very safe, but incomplete and distorted view of the world.

Deep knowledge and understanding, creativity, critical thinking, discernment, wisdom, and judgment are not about the accumulation of facts. They are about grasping the relationships between ideas, information, ethics, and culture. When students search the Internet, the sites they go to are not simply destinations; they are steps on the path to further discovery. When one door is closed, entire hallways of further doors may be closed off as well. It is not just that students go looking for “vegetarianism” as a topic; they also move through “vegetarianism” on their way to somewhere else — somewhere where they (and only they) may see a connection: nutrition and healthy diets, religious spirituality, meditation, animal abuse in the cattle industry, the chemistry of proteins and amino acids, and so on. If we close the door marked “vegetarian,” we may close off access to all those other possibilities. Even a “bad” site — on Nazism, say — may be an invaluable resource of “good” information: on history, perhaps, or the music of Wagner.

Reason Six: There are better solutions.

The problem, to the extent that there actually is a problem, of children accessing objectionable material on the Internet is ultimately educational, not technological. Attempting to restrict access to the wider Internet because a student might see a dirty picture is like closing libraries because some pervert once exposed himself in the stacks — it is the wrong response for the problem it attempts to solve.

We want to argue that free and open access to the resources of the Internet will create much less of a problem for schools and teachers than many people seem to fear. Children and young people using the Internet for educational purposes will not, typically, be flooded with pornography, cyberstalked by child molesters, or sucked into chat rooms on perversion. They will not learn how to build bombs or decide to become racist skinheads (and if they do, it will be for reasons ultimately having little to do with the Internet.). To put the point simply: good teachers in well-run classrooms know what their students are doing. Students in good classrooms are too busy and too involved in their education to have the time or interest to be looking at dirty pictures. If supervision is judged to be necessary, placing computers in more visible locations and moving around the classroom work perfectly well. But if students are spending large amounts of unsupervised time on computers with little educational purpose in mind, there is a deeper problem at work than any filtering technology can help with.

Acceptable Use Policies (AUP’s), although we are not generally fans of them, are another way of letting students know the boundaries of acceptable computer use, and the consequences of abusing them. The benefit of such policies is that they defer to trusting student choices and err on the side of allowing more, rather than less, access, until something goes wrong that merits restricting access or punishment. AUP’s can be written in an overly restrictive way too, and can constitute their own brand of censorship; but properly drafted, they can comprise the reasonable “rules of the road” for navigating the highways and byways of the Internet.

Continuing this analogy, students need to learn “defensive driving” on the Internet.[15] There are dangers and risks out there, and students need to learn how to recognize them and avoid them. But the paradox is that, as in driver’s education, the only way to learn how to recognize danger and avoid it is to be put in a situation where that danger is a real possibility. You can’t learn driver’s ed by just watching film, practicing with a simulator, or turning loops in a parking lot — you have to get out on the road, with supervision, where all the wackos and bad drivers are. The Internet is similar: educationally, once students get to a certain age, they have to be given the chance to make mistakes and bad decisions: this is an indispensable aspect of learning to make good ones. One of Plato’s great myths of education was that if you never expose students to the existence of certain ideas, it will never occur to them to be curious about those ideas. In actual practice, of course, the situation is often exactly the reverse. Filtering is based on the same Platonic illusion. It may make adults feel reassured, but it does little to limit the activities of really determined students looking for “bad” or dangerous things. It only restricts the naïve, the technologically inexperienced, and the “good” kids who accept the filter’s decisions without question — and as we have stressed repeatedly, these decisions are not to be trusted at face value.

There is an incredible wealth of educationally useful material available on-line. And there is also a lot of junk, and a certain amount of truly disturbing material. But ultimately, learning to make good intellectual and ethical decisions about what to make of it all is itself one of the most important educational aims of teaching with, and about, educational technologies. It is, on the whole, better to teach these skills by leaving important decisions and judgments in the hands of students, guided by their teachers and parents. The imposition of filtering removes that educational opportunity and compromises the vast educational potential of the Internet.

Conclusion

The Internet is now the primary way many teachers and students access information in their educational pursuits. For many young people, if they can’t access information in this way, they may not ever be able to discover it. To be honest, we suspect that the deeper issue is that many parents, and a few educators, do not want young people to be making these decisions for themselves, and don’t mind if their experiences and knowledge are limited to the bland, the conventional, and the mainstream. They don’t want their kids to become Nazi’s, but they also don’t want them to become vegetarians, atheists, or freethinking humanists. Ultimately, we believe, the filtering debate is not about pornography or bomb-making directions; it is about the reluctance of some adults to allow their children, and other people’s children, to have the free access to information that will allow them to come to their own conclusions about the world and their place in it.