The Real Meaning of Free Speech in Cyberspace

An invited talk for the conference "The Internet: Beyond the Year 2000", University of Toronto, May 1 1996.

Jeffrey Shallit

Electronic Frontier Canada


Department of Computer Science
University of Waterloo


  1. First-class versus second-class speech.
  2. Cyberspace as a forum for second-class speech.
  3. The role of established media.
  4. The real meaning of free speech in cyberspace.
Good afternoon. I'm Jeffrey Shallit, the Vice-President of Electronic Frontier Canada (EFC).

1. First-class versus second-class speech.

To help get at the real meaning of free speech in cyberspace, I'd like to propose a distinction between what I call first-class and what I call second-class speech.

First-class speech is books written by famous politicians or university professors, and printed by respected publishers, like the University of Toronto Press, McClelland & Stewart or Key Porter Books. First-class speech is newspaper columns written by renowned pundits, interpreting the news for us and telling us what to believe. First-class speech comes from them to us; they speak, and we listen. First-class speech is clean, organized, controlled, licenced. First-class speech is reassuring and comfortable. First-class speech is Canadian. There aren't many four-letter words, and there isn't much name-calling. First-class speech is the speech of the economic and intellectual elite. First-class speech is serious, thoughtful, informed, and, most important of all, protected by law. Although there are still a few closet censors and book-burners around, most Canadians are comfortable with making their own decisions about what books to buy and what books to read.

Second-class speech is everything else. It's the right-wing extremism of talk radio, the pornographer's blue movies, the communists' handbills that implore us to "Smash the State". It's citizen's band radio, the evangelists' pleas for more money, and TV shows like "Oprah" and "Geraldo". In second-class speech, respected pundits don't necessarily rule the day. Second-class speech is dirty, unorganized, uncontrolled, and unlicenced. Second-class speech is global. Second-class speech makes us blush. Second-class speech is disturbing and uncomfortable, because there are, from time to time, some four-letter words, and some name-calling. In second-class speech, the experts often take a back seat. Second-class speech means that we get to talk, and they -- well, if they don't listen, at least we still get to talk. Second-class speech is sarcastic, insensitive, and frequently ignorant.

Second-class speech is the speech of democracy at work.

But second-class speech doesn't get a lot of respect. Important conferences, such as this one, aren't usually devoted to extolling its virtues. And though almost all of it should be protected, second-class speech is, unfortunately, not always protected by law. Those who advocate strict government control of second-class speech in a misguided attempt to protect society should remember:

First-class speech isn't always right, and second-class speech isn't always wrong.

2. Cyberspace as a second-class forum.

Now, what's the relevance of this distinction for speech in cyberspace? I propose that sometimes an entire medium will be dismissed as a second-class forum. Here is a law of new media:

A new medium of communication will be regarded as second-class until it becomes legitimized by first-class speakers.

Film is a good example of this rule. Today, I imagine most of us regard the freedom to make and show movies as an obvious corollary of the Charter's guarantee of

...freedom of thought, belief, opinion and expression, including freedom of the press and other media of communication... [2]

But it wasn't always that way in North America. For example, in 1915, a US Supreme Court decision rejected the claim that movies were "expression", categorizing them as "shows and spectacles" instead. And a "show and spectacle", the Supreme Court reasoned, was clearly not entitled to any protection under the First Amendment [12].

Although recent Canadian movies such as "If You Love This Planet", "The Valour and the Horror", and "The Boys of St. Vincent" clearly represent important political expression, echoes of that early rejection of film as first-class speech can still be heard today -- for example, in the existence of provincial film censorship boards, such as the one in Saskatchewan that temporarily banned Dan Aykroyd's recent comedy, "Exit to Eden" [18].

Today the new medium is not film, but the Internet. It follows from this law of new media that speech in cyberspace won't get the respect it deserves. And it doesn't.

Let's look at a few examples. And where better to start than a Canadian university? I'm sad to say that some Canadian universities seem to have lost their intellectual compass; instead of ensuring robust debate on all issues, they have adopted the goal of avoiding offence at all costs. Frequently, the result is censorship, but the censors always refuse to call it by that name. After much research, I would like to propose the following observation, which I modestly call "Shallit's First Law of Censorship":

If someone says, "This isn't a censorship issue", you can be pretty damn sure it really is.

But back to my first example. Carol Walberg is a 49-year-old sociology student at the University of Guelph. In 1994, she moderated a "tasteless jokes" conference on one of the University's computers. Tasteless jokes did not pop up unbidden on the screens of users; rather, one had to explicitly join the conference, which was clearly labeled as such. Most of the jokes were copied from the Usenet newsgroup "alt.tasteless.jokes" which was freely available on other University computers. Nevertheless, a student complained, and Walberg became the target of an investigation by the University's Director of Security, Keith McIntyre. Walberg never received any written statement of charges against her, but was nevertheless ordered to appear at a meeting at the University's Security Office. During this meeting, Walberg was ordered to sign an apology written by the complainant, which was to appear in five local newspapers. She was threatened with expulsion from the University and a fine if she did not comply. To her credit, Walberg did not submit. And true to Shallit's First Law of Censorship, McIntyre was quoted as saying,

"[This] has nothing to do with censorship. They're making it into a censorship issue." [8]

The University eventually dropped its investigation, although Walberg never received any written decision clearing her. But all was not well at Guelph. Another jokes conference set up later received only a small fraction of the subscribers the earlier one had. Apparently, students were afraid of being targeted for investigation the way Walberg was. Although the University's actions resulted in a serious and irreversible chilling effect on speech at Guelph, the University never apologized or admitted wrongdoing.

This is a good example of how cyberspace was regarded as a second-class forum. The University would never have dreamt of taking a similar action had Walberg published a book entitled Tasteless Jokes. Indeed, Guelph does have a book in its library entitled Jokes: Form, Content, Use, and Function [17], which contains jokes very much like the ones that got Walberg in hot water. But you didn't see the Director of Security demand the removal of that book from the University library.

The second incident took place at the University of Waterloo, where I teach. Actually, it was not just one incident, but a series of on-again, off-again newsgroup bans and unbans, imposing censorship and then repealing it, stretching over seven years. Although most of the banned newsgroups have now been reinstated, Waterloo still bans the following newsgroups dealing with sexual topics, supposedly on the advice of legal counsel -- advice that has never been released to the scrutiny of the university community:      [d = discussion]
Yet, the university did not ban, or even choose to examine, books in its library and bookstore that contained material very similar to that in the banned groups. For example, the University of Waterloo library carries the book Caught Looking: Feminism, Pornography, and Censorship [4], which contains explicit scenes of sex and bondage. And, at the time of the ban, the UW bookstore openly sold Celeste T. Paul's book Women's Erotic Dreams [11], which contains explicit descriptions of sex with children and animals. Once again, this shows how first-class speech often seems to merit protection unafforded to second-class speech.

Even more difficult to explain is the fact that the following newsgroups dealing with sexual topics are still carried by the University of Waterloo:
Inconsistent? You be the judge. The fact that Waterloo bans "", but permits "" suggests to me that, in the eyes of the University, if you want to use the computer network to arrange sexual encounters, go right ahead. But if you want to write about them afterwards, you have to do it elsewhere.

Next, let's look at a third example where cyberspace was treated very differently from other media. When the Hemlock Society printed Derek Humphry's explicit suicide instruction book, Final Exit [7], there was some controversy, but no calls for censorship. Bookstores and libraries throughout Canada carried it with hardly any protest. Minors were free to peruse the book in bookstores, or read it at their local library.

But when the Right to Die Society of Canada opened up an information site on the world-wide web called "DeathNet" [3], all hell broke loose. "Suicide guru using Internet to tell teens how to die," read a headline in the Toronto Sun, despite the plain fact that the site didn't carry a single word of explicit suicide information [1]. Posturing politicians, such as Reform MP Keith Martin, opportunistically used the DeathNet site to speak in favour of a resolution demanding more government regulation of the Internet. An Internet service provider in Calgary, who provided Internet access for dozens of schools, terminated access to the DeathNet site for those schools. This censorship didn't prevent teens from gaining explicit suicide information -- since there wasn't any on the site to begin with -- but it did rob them of a valuable resource for learning about the debate on assisted suicide.

As Phil Kerby once remarked,

Censorship is the strongest drive in human nature; sex is a weak second.

Finally, let's turn to McGill University to see one more example of how the Internet is treated as a second-class forum. Earlier in this session we heard David Johnston, former chair of the Information Highway Advisory Council, discuss the work of that group. What he didn't tell you about, however, is his crucial, and I think, indefensible role in McGill's banning of the newsgroup "".

There's probably not a person in this room who hasn't heard about the gruesome crimes of Paul Bernardo and Karla Homolka. Shortly after Karla Homolka entered a plea, the sarcastically-named Usenet newsgroup "" was set up to discuss aspects of the cases. Soon, however, some small number of postings contained information that might have contravened a publication ban imposed by Judge Francis Kovacs. By my estimate, however, only about 1% of all the messages posted to the newsgroup were legally questionable. The rest, 99%, consisted of debates about the difference between US and Canadian law, the justification for the publication ban, etc. -- in other words, robust, open debate on issues of great public interest.

Last year, at the Net '95 conference in Ottawa, audience members heard Johnston put forth a revealing account of how, in his role as the Principal, he acted to censor this newsgroup at McGill University in 1993.

According to Johnston's own account, here's what happened. The censorship began with Vice-Principal Francois Tavenas' unilateral decision to shut down a computer in the electrical engineering lab because "there were details of the Homolka trial being published". Tavenas had apparently not bothered to consult any lawyer when he first pulled the plug. When Tavenas went and informed Johnston of his decision, Johnston -- to his credit -- expressed some skepticism about the need for such drastic action. Tavenas then left his office, and returned in fifteen minutes with a hastily-written opinion from the University's legal advisor supporting Tavenas' censorship. At that point Johnston agreed to remove the newsgroup entirely.

Now it is extremely rare for a Canadian university to remove a book from its library because of the threat of legal action. If you're at all familiar with universities the way I am, you know that if such a decision were made, it would almost certainly be made with great care and deliberation. Committees would be formed, there would be endless debate, and months later a recommendation would emerge.

But because speech on the Internet is treated as second-class, 15 minutes was considered adequate deliberation at McGill. This, despite the fact that something like 99% of the messages posted to "" did not contain any information banned by Judge Kovacs' order. Was there really no feasible alternative to banning the entire group? Was there no understanding that doing so would shut off debate on an issue of great public concern?

To show how inconsistently different media were treated at McGill, banned trial details were also available in a Washington Post article contained in the McGill University library [16]. The same legal advisor that advocated the censorship of an entire newsgroup based on 1% of the content stated confidently that merely having the newspaper on a shelf in the library posed no legal risk. But wasn't the real reason for this inconsistent treatment because the Washington Post was considered first-class speech, and the Internet was just second-class speech?

3. The role of established media.

As we've seen earlier in the DeathNet case, the established media share a large portion of the blame for the poor public perception of cyberspace as a forum for electronic speech. With a few happy exceptions, such as Eye's K. K. Campbell, Canadian coverage of the Internet has been, for the most part, uninformed and factually incorrect -- even hostile.

This brings me to a second law of new media:

Established media are scared of new media. And why shouldn't they be? An established medium consists largely of first-class speakers, while a new medium generally attracts second-class speakers. And a new medium represents all kinds of threats to established media: economic, political, and ideological. If I were a newspaper publisher, I'd be terrified that people would start getting their news directly from eyewitnesses, instead of filtered through years of journalistic experience.

In the case of cyberspace, the second law of new media has resulted in some abysmal and sensationalist reporting. A large number of the stories about the Internet fall into what I call the "Didja" category, as in "Didja hear the Internet could be used to harass people?", or "Didja hear that you could libel people on the Internet?", or "Didja hear that people actually look at pictures of naked women on the Internet?". For much of the coverage, the actual incidents aren't important. What is important is the apparently astonishing observation that computers can be used to commit the same old transgressions that take place in more established media.

When Marty Rimm, a student at Carnegie-Mellon University in Pittsburgh, published a "study" -- and I use that word generously -- in the Georgetown Law Journal [15] that concluded that pornography on the Internet had reached epidemic proportions, reporters at Time were quick to make it a cover story [5]. In their rush to exploit the sensationalist aspects of the "study", they didn't bother to investigate whether the study had any scientific validity, or whether Rimm was a credible source. But Time was soon forced to backtrack, when it was revealed that Rimm had a questionable track record, that his statistics were doubtful, and that his "study" hadn't received any serious peer review [6,10,20].

With the Internet on everyone's lips, reporters tried to find cyberspace connections even when there weren't any. When a 15-year-old Halifax boy tragically blew off part of his hand with a homemade bomb in July 1995, CBC Radio in Halifax started to convene a "panel of experts" who could discuss "bomb recipes on the Internet". Never mind that there was not a shred of evidence linking the kid's misfortune to the Internet. Never mind that The Anarchist Cookbook [13], which contains bomb recipes, is readily available in bookstores and libraries throughout Canada. (You can stroll down this afternoon to "The World's Biggest Bookstore" and pick up your own copy for $34.75.) CBC called EFC's President, David Jones, at McMaster University to ask if he could appear on the panel. But when Jones came up with a list of sources from the Nova Scotia Provincial Library where one could easily obtain recipes for explosives, CBC Radio suddenly lost interest in the story.

I think the moral is that you can't always believe what you read, even when it comes from first-class speakers. The Internet is a good way to teach your kids to be skeptical of all sources of information, and how to tell the good sources from the bad.

4. The real meaning of free speech in cyberspace.

Now we finally come to the title of my talk: the real meaning of free speech in cyberspace. The position of Electronic Frontier Canada, briefly, is this: if it's legal in one medium, it should be legal in another. Cyberspace should be treated as a first-class medium of communication.

At Electronic Frontier Canada, we don't want special treatment for cyberspace -- we want equal treatment. We think cybernauts shouldn't be treated as second-class speakers. If it's legal to print, we think it should be legal to distribute over the Internet, too. This applies to all speech that is legally protected -- and it includes most forms of pornography.

We'd also like to see electronic mail receive the same legal protection as ordinary mail sent by Canada Post. Your e-mail should be considered private and shouldn't be searchable by the authorities without a warrant.

The flip side to equal treatment is that if it's illegal in Canada to distribute in print, it should also be illegal for Canadians to distribute electronically. We don't make the absurd claim that anything and everything said in cyberspace is immune from legal action. Some may think of cyberspace an ethereal realm, existing everywhere and nowhere, but the people who use it are still physically located in space. At Electronic Frontier Canada, we wouldn't dream of claiming that the death threats somehow become protected speech just because they were sent via electronic mail.

We'd also like to see that the people who actually post messages with illegal content are the ones held responsible, not the Internet service provider that merely serves as a conduit for the message. We don't hold the telephone company responsible for harassing phone calls, or Canada Post responsible for chain letters. In the same way, we shouldn't hold the service provider responsible for illegal content.

Let's not forget, however, that the protections in the Charter for free speech apply only to the actions of government and its agents. When you sign up with an Internet service provider, or ISP, you agree to certain standards of conduct, such as not harassing other users. If you violate those standards, the ISP is within its rights to terminate your contract. As long as there are dozens of ISP's to choose from, I don't see this as a threat to free speech. Choice ensures, for example, that a fundamentalist Christian ISP can establish its own standards -- such as banning four-letter words, or not carrying newsgroups that deal with evolution or sexual topics. Choice ensures that if I don't like those rules, I can go elsewhere.

Some parents are still nervous about the Internet because of the presence of information they don't want their kids exposed to. Some demand additional regulation of the Internet to make it safer for children. But cyberspace is still developing. Poorly-thought-out regulations enacted now run the risk of either being completely ineffective, or choking the new medium before it has a chance to flower. Let's not repeat the same mistakes we made with radio and television. Remember:

Only a mature technology can be regulated effectively.

Let's take a moment to look at some of the proposals to prevent exposure to offensive material on the Internet and see which ones are reasonable.

First, there's mandatory rating systems. Some have proposed, for example, that every Web page be rated -- either self-rated, by the author, or rated by the government. Ratings might cover such things as violence or pornography. A rating of "P", for example, might mean that the web page contains material deemed pornographic. Web browsers could then examine these ratings and decide whether or not to display the pages on your computer.

But government ratings are clearly infeasible. For one thing, the sheer volume of material out there means the cost would be enormous, and the volume is growing at something like 10% per month. For another, it would put government, not parents, in the business of deciding what is acceptable for children to read.

Self-ratings are also impractical. There's no uniform standard that people can apply. Unless there is some penalty for classifying materials "incorrectly", there is no real incentive for people to rate their pages "appropriately". And, if there are such penalties, we again put government, not parents, in the position of telling us what is appropriate for children to read. Furthermore, mandatory self-rating means that hundreds of businesses and universities all across Canada will be wasting their time examining one web page after another, and writing on them, "This is not porn. This is not porn. This is not porn" over and over.

The proper solution is third-party ratings. The nice thing about third-party ratings is that they already exist and they provide a lot more useful information than simply "This is not porn." Third-party ratings can also tell you which web site to look for when you want current book reviews; which web site has the most comprehensive and accurate encyclopedia; and which web sites to avoid because their content is trivial, wrong, or offensive.

Third-party ratings are the right solution because instead of just one rating, there can be many. Fundamentalist Christians and atheists can each maintain their own ratings. No matter what your position on the political spectrum -- Reform, Conservative, Liberal, or NDP -- there will be a rating service commensurate with your desires. And parental control software can ensure that your kids only get access to web pages that are rated by someone you trust. In the same way that Underwriters Laboratories place a UL seal of approval on electrical appliances, some rating services will become known for their thoroughness and reliability. There's no one well-known enough to be the Siskel and Ebert of cyberspace yet, but it's just a matter of time. Third-party ratings accommodate the concerns of parents, and preserve the democratic goal free speech in cyberspace.

The real meaning of free speech in cyberspace is the same as the meaning of free speech in print, in pictures, and on radio and television. The real meaning of free speech is that a democratic society is built on the guarantees of the Charter of Rights and Freedoms:

...freedom of thought, belief, opinion and expression, including freedom of the press and other media of communication... [2]

A free society needs a free Internet. As John F. Kennedy once remarked,

"...a nation that is afraid to let its people judge the truth and falsehood in an open market is a nation that is afraid of its people." [9]

Cyberspace is the marketplace of ideas for the new millennium. Let's let individuals, not government, make their own choices about what is true and what is false. The real meaning of free speech in cyberspace is that cybercitizens should be -- indeed, must be -- treated as first-class speakers.


[1] K. K. Campbell, "The little DeathNet story that grew: why many people now believe that teens can login and learn how to off themselves", Eye, May 11 1995, p. 16.

[2] Constitution Act, 1982, R. S. C. 1985, Appendix II, No. 44.

[3] DeathNet.

[4] Kate Ellis, Beth Jake, Nan D. Hunter, Barbara O'Dair, and Abby Tallmer, eds., Caught Looking: Feminism, Pornography, and Censorship, The Real Comet Press, 1988.

[5] Philip Elmer-Dewitt, "On a screen near you: cyberporn", Time, July 3 1995, pp. 32-39.

[6] Philip Elmer-Dewitt, "Fire storm on the computer nets", Time, July 24 1995, p. 30.

[7] Derek Humphry, Final Exit: The Practicalities of Self-Deliverance and Assisted Suicide for the Dying, Carol Publishing, 1991.

[8] Eric Johnson, "Tasteless jokes -- no laughter at U of G", Guelph Tribune, January 4 1995, p. 1.

[9] John F. Kennedy, Remarks made on the 20th anniversary of the Voice of America at H. E. W. Auditorium, February 26, 1962.

[10] Peter H. Lewis, "Computer smut study prompts new concerns", New York Times, July 16 1995, Sect. 1, p. 11.

[11] Celeste T. Paul, Women's Erotic Dreams, Grafton Books, Collins Publishing Group, 1988.

[12] Lucas A. Powe, Jr., American Broadcasting and the First Amendment, University of California Press, 1987.

[13] William Powell, The Anarchist Cookbook, L. Stuart, Inc., 1971.

[14] Charles Rembar, in Censorship: For & Against, Hart Publishing Company, New York, 1971, p. 203.

[15] Marty Rimm, "Marketing Pornography on the Information Superhighway", Georgetown Law Journal 83 (5) (1995), 1849-1934.

[16] Anne Sewardson, "Unspeakable crimes: this story can't be told in Canada. And so all Canada is talking about it", Washington Post, November 23, 1993, p. B1.

[17] Christopher P. Wilson, Jokes: Form, Content, Use, and Function, Academic Press, 1979.

[18] "Saskatchewan reverses film ban", Globe & Mail (Toronto), October 14 1994, p. D1.

[19] Bomb recipes on the Internet, EFC archives.

[20] Donna L. Hoffman and Thomas Novak, The Cyberporn Debate.