Reinhardt Buys
BLC LLB (Pret) LLM (UCT)
Sonnenberg Hoffman & Galombik Attorneys, Cape Town
Defamation on the Internet | Pornography on the Internet | Hate speech on the Internet | Author biography
1. Defamation on the Internet

“Good name in man and woman, dear my lord,
Is the immediate jewel of their souls;
Who steals my purse steals trash; ’tis something, nothing;
’Twas mine, ’tis his, and has been slave to thousands;
But he that filches of me my good name
Robs me of that which not enriches him
And makes me poor indeed”

Shakespeare, Othello Act III Scene 3
1.1 Introduction

Defamation is a private and common law matter and its rules place severe restrictions on free expression. The law of defamation seeks to find a balance between the individual’s right to a reputation or good name and another’s right to free expression. Over the years, the law has reflected the constant tension between these two competing rights.1

  329

Being sued for defamation is a constant danger for ISPs and operators of bulletin boards and newsgroups. Every day there are hundreds of messages and postings which cross the boundaries between free expression and defamation. Each newsgroup is a public forum in which an Internet user can read and post messages on any particular area of interest. Topics range from the professional to the bizarre, such as the latest Elvis sightings. The potential for defamation is obviously high, as there is much use of personal abuse, attacks on other users’ characters and other matters which could well be defamatory.

There are a number of features unique to the Internet, which distinguish it from other mediums through which defamatory content could be published and which have led to the re-examination of the existing rules of defamation. These unique features include:

330

  • The global nature of the Internet raises jurisdictional issues and questions as to the applicable governing law. Plaintiffs may even have the luxury of “forum shopping” for jurisdictions where the law of defamation and awards for damages are most favourable.
  • The interactive nature of some of the services on the Internet makes it akin to a talk show or a village town-hall session. The opportunity for an immediate reply is unparalleled compared to any other media.
  • Accessibility is another feature of the Internet, which distinguishes it from traditional print or broadcast media. The relatively low cost of accessing the Internet means that every individual user has the potential to publish information to a world-wide audience.
  • The anonymity of Internet use allows users to post messages and send e-mail without having to reveal their true identity and users are therefore far less inhibited about the content of their messages. The use of anonymous remailers has obvious benefits, for example encouraging victims of rape, child abuse or AIDS to seek counselling.2 But it could also be used to spread pornography or defamatory messages over the Internet. Some writers even argue that the operators of anonymous remailers should be held accountable for constructive knowledge of civil and criminal violations over their systems.3
1.2 Online defamation in the United States of America
In Cubby Inc. v CompuServe4 the defendant was an ISP providing subscribers with access to a variety of special interest databases and forums. One of the forums focused on the journalism industry and provided specific information concerning broadcast journalism. One of the publications available carried false and defamatory statements about Cubby. The defendant did not dispute that the statements were in fact defamatory, but argued it acted as a distributor and not a publisher and could not be held liable for statements it did not know and had no reason to know of. It had no opportunity to review the contents of publications before they were uploaded into its computer data banks, relying on Smith v California,5 where the Supreme Court held that a distributor must have knowledge of the contents of a publication before it can incur liability for its distribution.
 
Whether a party is characterised as a publisher is largely dependent on how much editorial control is exercised over a publication. In Miami Herald Publishing Co. v Tornillo6 the Supreme Court held that the choice of material to go into a newspaper, and the decisions made as to limitations, size and content constitute the exercise of editorial control.

In the Cubby decision the court characterised the defendant’s product as an electronic library that provides a variety of publications and collects subscriber fees in return for access. The court found, furthermore, that the defendant had no more control over such publications than does a public library, book store or news stand and it would not be feasible for it to examine every publication it carries for potentially defamatory statements.

331

In Stratton Oakmont Inc. et al v Prodigy Services Company7 the defendant was an ISP that hosted a bulletin board where subscribers discussed stocks, investments and other financial matters. An unidentified person posted statements to the bulletin board about the plaintiff, a securities investment bank, stating that its president committed fraudulent acts in connection with a public offering of stock. The president was described as “100% criminal”, “soon to be a proven criminal” and the plaintiff was characterised as a “cult of brokers who either lie for a living or get fired.” Consequently Prodigy was sued for defamation based on its policy that it exercised editorial control and used board leaders, in this instance Charles Epstein, whose duties included enforcement of promulgated content guidelines. Prodigy also used a software screening program, which enabled board leaders to delete postings considered to be offensive or in bad taste. The plaintiff alleged that Prodigy was a publisher and not merely a distributor. Prodigy countered these claims by observing that although board leaders may remove postings that violate its guidelines, this was not the equivalent of acting as an editor. For legal authority Prodigy argued that it was akin to a distributor, much like CompuServe’s position in the Cubby decision.

In granting summary judgment to the plaintiff the court concluded that Prodigy was a publisher and should incur the level of liability associated with such status. The court distinguished the Cubby decision on two grounds. First, Prodigy held itself out to the public as a service that controlled the content on its bulletin board. Second, it had implemented this control through software screening programs and the use of discussion leaders to enforce its policies and guidelines.

US legislators have responded to the Prodigy judgment by enacting legislation to protect ISPs that take affirmative steps to remove, filter and edit material hosted by them from liability. Congress added to the Telecommunications Act of 1996 a new Title V, known as the Communications Decency Act of 1996. One source of protection for ISPs is found in section 230(c) of the Act, which provides as follows:

332

230(c) Protection for “good Samaritan” blocking and screening offensive material.

(1) Treatment of publisher or speaker
No provider or user of an interactive computer service shall be treated as the publisher or speaker of any information provided by another information content provider.

(2) Civil liability
No provider or user of an interactive computer service shall be held liable on account of – (A) any action voluntarily taken in good faith to restrict access to or availability of material that the provider or user considers to be obscene, lewd, lascivious, filthy, excessively violent, harassing, or otherwise objectionable, whether or not such material is constitutionally protected; or (B) any action taken to enable or make available to information content providers or others the technical means to restrict access to material described in paragraph (1).

A number of ISPs successfully used this section as a bar to defamation action. In Zeran v America Online Inc.8 the plaintiff became the target of a malicious hoax perpetrated through the Internet. Someone posted a message to AOL’s bulletin board suggesting that the plaintiff was marketing a line of apparel called “Naughty Oklahoma T-shirts” that sported messages in extremely poor taste about the Oklahoma City bombing. The plaintiff received a flood of intimidating phone calls and contacted AOL demanding that the posting be removed and a retraction published. AOL agreed to remove the posting but refused to publish a retraction. After the posting was removed, and after assurances by AOL that the source account would be cancelled, additional messages were posted which persisted for several days. The plaintiff subsequently sued AOL for defamation but the claim was successfully defended by reliance on section 230(c) of the Communications Decency Act.

Defamation claims against ISPs were also successfully barred by reliance on section 230(c) in, among others, Ben Ezra, Weinstein & Co. Inc. v America Online Inc.,9 Aquino v Electriciti Inc.10 and Doe v America Online Inc.11

Caden & Lucas Accidents on the Information Superhighway: Online liability and regulation at http://www.urich.edu/~jolt/2vi1/caden_lucas.html

333
1.3 Defamation and the Internet in South Africa
The delict of defamation is defined as the unlawful and intentional publication of a statement concerning another person which has the effect of injuring that person’s reputation in the eyes of right-thinking members of society.

The plaintiff in a defamation action has to prove that there was publication of defamatory matter referring to him/her. This must be done on a preponderance of probabilities. Once publication is established, the plaintiff must prove that the defendant was responsible for the publication. On the Internet this could lead to considerable difficulty for the plaintiff. Publication is attributed to the defendant if it was foreseen or reasonably foreseeable that an outsider would take cognisance of the defamation.12 Furthermore, not only the person from whom the defamatory remark originated, but also any other person who repeats, confirms or draws attention to it, is in principle responsible for its
publication. Proof of publication of defamatory matter referring to the plaintiff gives rise to two presumptions:

  • unlawfulness or wrongfulness
  • animus iniuriandi (subjective intention on the part of the individual defendant to impair the plaintiff’s reputation, with knowledge of unlawfulness)
The burden of rebutting these presumptions is on the defendant.13 This is done by proving a defence excluding unlawfulness, e.g. truth for the public benefit, fair comment, privilege occasion, consent, private defence or necessity. Defences excluding animus iniuriandi include mistake, jest, intoxication, provocation and insanity.
 
If a court concludes that the plaintiff has been subjected to an unlawful, intentional (only in the case of a non-media defendant) publication of defamatory words or conduct referring to the plaintiff that causes an impairment of reputation, the plaintiff is entitled to an award for damages. A person who anticipates that his/her reputation will be injured by defamatory matter may prefer to apply for an interdict restraining publication.14

The concise issue is therefore whether those who host content for others on the Internet could be held liable for the defamatory nature of such content and, if so, to what extent.

334

The right to free expression in the Constitution15 makes specific reference to the freedom of the press and other media, which is an essential element in any liberal democracy. From 1982 to 1998 the media were held to be subject to strict (no fault) liability for the publication of defamatory matter.16 In the much discussed decision of National Media v Bogoshi17 the Supreme Court of Appeal rejected strict liability of the press and media as incompatible with freedom of expression and set down a rule based on the objective reasonableness of the publication to replace strict liability. While the judgment of Cameron J in Holomisa v Argus Newspapers Ltd18 signified the high water mark in the recognition of the demands of freedom of expression, Hefer JA’s judgment in Bogoshi constitutes the watershed decision in the revival of common law emphasis on free expression, in particular of the print and electronic media.
 
For a publication to be reasonable still demands a high degree of circumspection on the part of editors and editorial staff, and includes factors such as the time and manner of the publication, the status or degree of public concern in the information, its political importance, the tone of the publication, the reliability of its source, the steps taken to verify the information and whether the person referred to has been given an opportunity to verify the information. The person allegedly defamed should also be given an opportunity to verify, comment or reply to the information.19
 
Because of different degrees of liability between the press and an individual, it would be necessary to distinguish which publishers should be regarded as “media” on the Internet. Some writers argue that the whole Internet should be regarded as “mass media”.20 This would be an oversimplification of the issue. From the case law it seems clear that “mass media” indicates some form of editorial control. Furthermore, the functions attributed to the mass media, to inform the public about every aspect of public, political, social and economic activity and thus to contribute to the formation of public policy,21 cannot be attributed to all forms of content on the Internet. Finally, “mass media” seems to imply a certain forum of publication and continuity. It could therefore be argued that online versions of newspapers and magazines and other online news services would be regarded as mass media for defamation purposes. The sender of a defamatory e-mail message or real time Internet chat containing defamatory content could hardly be referred to as “mass media”. Bulletin boards and newsgroups seem to be the grey area, and it would depend on the degree of editorial control over the content posted there to decide whether such a bulletin board should be regarded as “mass media”. A bulletin board where messages are posted without any editorial control or a newsgroup without a controller/operator to screen and delete messages would probably not be regarded as “mass media”.
1.4 What constitutes publication on the Internet?


335

Publication is the act of making known a defamatory statement or the act of conveying an imputation by conduct, to a person or persons other than the
person who is the subject of the defamatory statement or conduct.22 Although defamatory matter usually consists of written or spoken words, other conduct may be sufficient to convey a defamatory meaning, for instance a sketch, cartoon, picture or photo. Defamation on the Internet could therefore also occur in technologies such as video-conferencing, voice over the Internet, graphics on a web site and the like and is not restricted to written words.

 
Publication could also take place through omission.23 For instance, if the controller of a web site is aware that a defamatory posting was made to the site, there might be a legal duty to remove it to avoid publication.
 
The South African case law on defamation seems to take the approach that there can be no publication unless the person(s), to whom the statement was communicated, understood the meaning. If words are published in a foreign language or secret cipher, there would be no publication to readers who do not understand the language or cipher.24 A defamatory message that is encrypted would therefore only be published once it is successfully decrypted and understood. A defamatory e-mail message would only be published after it is opened or unzipped and read. There may, however, be sufficient publication to someone who does not immediately understand the meaning of the words used, but later discovers their meaning.25
 
Burchell argues that a genuine, bona fide mistake regarding the element of
publication should be excused. The possibility of sending an e-mail message by mistake or posting a message to a bulletin board by accident should be expected, as it takes no more than a single click on a “send” button and there is no way to retrieve such messages once they are sent.
 
Certain decisions indicate, and some writers argue, that liability for defamation depends on whether the plaintiff can prove that the defendant knew or ought reasonably to have expected that, in the circumstances, publication would be a likely result of certain conduct.26 It follows that defamatory messages on the Internet intercepted by others, decrypted without authorisation and accessed in abnormal ways, e.g. through hacking, would arguably not constitute publication.

However, writers agree that postings to a newsgroup, sending e-mail, making a web site available on the Internet, Internet relay chat, files transferred by file transfer protocol and video-conferencing, whether the content is in the form of text, graphics, audio or video, would amount to publication for defamation purposes.27

336
1.5 Where does publication take place on the Internet?
Publication takes place where the defamatory content is read, seen or heard and is completed when the receiver understands the content. It was held that every copy of a newspaper, book or written matter on a broadcast of television or radio technically constitutes a separate publication.28 The results of application of this theory to the Internet would be nothing short of chaotic. Apart from e-mail messages, the contents of a web site and postings to bulletin boards have potentially world-wide audiences. A possible solution would be to adopt a singe publication rule, as adopted by a number of states in the United States.
 
Plaintiffs could embark on forum shopping and sue where they have the greatest chance of success. After a decision by the French courts to award damages of 20 000 francs to the plaintiffs as a result of defamatory remarks on BBC Radio Guernsey (which broadcasts over parts of Normandy, France) a commentator stated:29
“We had been watching this case closely because of the implications it carries for broadcasters, Internet publishers and others involved in cross-border publication. The ruling shows it will be possible for wealthy litigants to sue where they think they have the greatest chances of success. France will be popular, because its privacy and some of its libel laws are draconian.” 
In a number of American cases the issue of jurisdiction in the case of Internet defamation was decided. In Edias Software International L.L.C. et al v Basis International Ltd,30 a defamation claim was based on Internet messages sent to customers of an Arizona company. At issue was whether the defendant, a European company with offices in Arizona, could be subjected to a specific jurisdiction. The court held that under a three-part test, the non-resident defendant must purposely avail itself of the privilege of doing business in the forum, the claim must arise out of its forum-related activities and the exercise of jurisdiction must be reasonable. In Playboy Enterprises v Chuckleberry Publishing Inc.31 the District Court held that US law could be applied to tortuous information originating from Italy and which can be accessed from the United States.

In Shevill v Press Alliance SA32 the European Court of Justice held that a plaintiff could sue either in the country of publication, pursuant to art 5(3) of the Brussels Convention, or in the country in which the defendant publisher was established. However, if the plaintiff chose to sue in the country of publication it could recover damages only in respect of the publication in that country. Otherwise it could recover for publication in all the Convention countries.

  337

In the Australian cases of McKain v R W Miller & Co. (SA) Pty Ltd33 and Stephens v Head34 the High Court laid down some principles which may arguably be applicable to resolving the issue of jurisdiction where Internet defamation is concerned. In order to recover damages for a defamatory statement published in another Territory or state (or overseas), the plaintiff must show that (a) the statement would have exposed the defendant to liability according to the laws of the jurisdiction where the action is heard, and (b) liability would also accrue under the law of the jurisdiction in which the statement was published.35

Although many countries have reciprocal enforcement of civil judgments, countries such as the United States are usually unwilling to enforce defamation judgments of other countries, because of the high degree of protected expression enjoyed under the First Amendment.

Potts & Harris Defamation on the Internet (1996) at http://www.cyberlibel.com/introduc.html
1.6 Liability of ISPs for defamatory content
To publish information on the Internet requires the participation of many different entities, including a host and access provider. As these will often have deeper pockets than the original author (who might even be anonymous), the extent of their liability for defamatory content is of great interest.
 
As a general rule everyone who participates in the publication of defamatory content may be liable for defamation. For example, where defamatory content appears in a newspaper, not only the author but also the editor, printer, publisher and owner can be held responsible.
 
Booksellers, libraries and news-vendors may also be liable for published defamatory content, but in Trimble v Central News Agency Ltd36 Stratford ACJ set out the principles governing the position of a newspaper vendor as follows:
“a newspaper vendor is protected if he proves (1) that he did not know that the newspaper at the time it was sold contained libel of the plaintiff; (2) that it was not by negligence on the vendor’s part that he did not know that there was any libel in the newspaper; and (3) that the news-vendor did not know that the paper was of such a character that it was likely to contain libellous matter, nor ought to have known.”

  338

The liability of the news-vendor is therefore tested by a negligence criterion and the liability of booksellers, newsagents, libraries and other distributors of published material is judged by the same standard.37 A standard of negligence places the burden on the distributor to make a reasonable check when in doubt, for example in Vizettely v Mudies Select Library Ltd38 the defence failed because the defendant library had overlooked a publisher’s circular requesting it to return the book.

If the ISP, whether as host, access provider or news collector, is regarded as being in a position analogous to a public library, news-vendor or bookseller, the appropriate fault element will be intention or negligence. If the publication of content on the Internet is found to be analogous to publication in the mass media, the parameters laid down in Bogoshi39 would be applied. In this regard, South African courts could be guided by the American judgments in Prodigy and Cubby.40 These two cases indicate that where there is a moderated discussion group online, there is potential liability for the moderator, the ISP hosting the discussion group or both.41 Rosenoer42 offers the following sobering assessment:
“What the courts [in Prodigy and Cubby] do not fully address ... is the fact that you cannot realistically screen for defamation. Nor do they recognise that an online system is simply not the same thing as a print (e.g. book or newspaper) distributor or publisher. In many cases, online systems provide the equivalent of a park and areas where people with common interests may congregate and set up their own soapbox. The online system may have wardens that ensure that people wait their turn to get on the box and that nude sunbathers do not scare the children, but they have no real ability to analyse content and determine whether a statement is defamatory. Forced to do so, they will simply not allow people to speak their minds – not a good thing in a society that values freedom of speech.”

Much (but by no means all) Internet content is created and distributed in unwrapped form, just as the author provides it. Therefore, in many cases it is no longer possible to distinguish between publisher and distributor. A typical example is online discussion forums, such as Usenet or bulletin boards, where members of the public can place messages directly onto a database accessible by anyone with the appropriate software. The wrapping function of the traditional publisher has disappeared, or at least has been reduced to the routine task of maintaining the organisation of the messages by topic, date and author. The discussion forum proprietor may choose to employ discussion leaders and whether or not to exercise any editorial control.43 It should be emphasised, however, that although the newsgroup example is used frequently, it is only one form of publication on the Internet. Many variations are possible, ranging from the permanent to the ephemeral, with different relationships between the parties involved. Some publishing on the Internet is more akin to hard copy publishing and advertising, with carefully crafted and vetted copy being placed on permanent web sites, or with electronic versions of well-known newspapers, magazines and journals.

  339

Interesting questions could arise as to the status of mirror sites, which act as a local duplicate host for a foreign site. This enables local users to obtain faster access. Such sites have features of both distributor and primary publishers in terms of the traditional analysis. They are distributors in the sense that they receive a third party’s publications, but in another sense they are doing nothing different from the original hosts that agreed to host a publication.44
 
A web site can provide access to content without any of the content being physically on the site at all. This could be done by a series of links, which when clicked on will take the user to the linked material. These links may be internal, leading to pages on the same site, or external, leading to material on another site elsewhere on the Internet. Some web sites contain only a collection of links to any number of particular subjects, for example a search engine.
 
What is the status and liability of such a site with regard to defamatory content contained on one of the linked sites? There is old English authority45 that drawing attention to an existing defamatory statement can amount to publishing it. In this case the defendant sat in a chair pointing out to passers-by a defamatory sign erected over a road. The author of the sign was unknown, but the defendant was found to have published it. This would seem to suggest that providing a link to another site could amount to publishing its contents. Much would depend on the facts to determine the status of sites containing links. For instance, was the link to the site as a whole or to a particular page or document on the site? The position of search engines would also be different because the owners or hosts of other external sites place the links on search engines there.

It would seem that the application of the law of defamation in South Africa to the Internet creates more questions than answers. Burchell,46 one of the most authoritative writers on defamation in South Africa, for example, asks the following questions: Communication over the Internet cuts across national and geographic boundaries, so where does publication take place? Where is the place of defamation (lex loci delicti) in multiple publication cases? On the question of extent of the defamation: is the plaintiff’s reputation lowered only in the eyes of users of a particular network47 or notional users of the entire network? Can a plaintiff bring an action for damages in one favourable jurisdiction and also seek damages with respect to other jurisdictions? Could the interactive nature of the Internet be an important reason for allowing greater leeway for publications on the Net?

  340

The potential liability of ISPs (providing whatever service) and Internet users is not yet clear. Solutions or guidelines will have to be identified soon to create certainty on the extent to which freedom of speech on the Internet should be limited to protect individual reputations.
1.7 Disclaimers to limit or exclude liability for defamatory content

A great number of web sites make use of disclaimers of legal liability relating to the accuracy of the content on the sites. Some disclaimers state that the owner of the site would under no circumstances be liable for the content on the site. Will such disclaimers protect the owners against liability for defamatory statements on their sites?

  341

The extent to which liability could be excluded or limited varies from country to country. Under UK law it is not possible, for example, to exclude liability for death or personal injury caused by negligence or for defamation.48 The validity of disclaimers is yet to be tested in South Africa. As far as defamation is concerned, the position in South Africa seems to be that a disclaimer cannot protect the owner of a web site against liability. The only defences available to a defendant are those creating an exclusion of either wrongfulness or fault. The courts would never protect defamatory content on a web site just because of a disclaimer of legal liability placed on such a site.

A disclaimer, in the light of defamation, could possibly be more useful if it sets out the editorial steps taken in relation to the content on the site, if any, or for example states that the site only operates as a distributor of electronic content. It would furthermore be prudent to include a user agreement in the disclaimer stating that a user accepts that the disclaimer and any legal matter between the user and the owner of the site would be governed by South African law. Finally, it is important that the disclaimer is brought to the attention of the user. It should preferably not be on a separate page on the site that could only be accessed optionally via an internal link hidden away at the bottom of the homepage. Preferable locations would be:

  • on the homepage
  • as a default entry page where a user should click on an “I accept” button before he/she can proceed
  • as a prominent link entitled “Disclaimer and use agreement” which should be placed on strategic locations on various pages of the site
Rosenoer Libel and suppression (1995) at http://www.cyberlaw.com/cylw1094.html
 
Sim Prodigy in Canada: Online libel revisited (1995) at http://www.mbnet.mb.ca/~psim/libel2.html
2. Pornography on the Internet

“The pride Comstock felt [in 1913] at having destroyed ’something over fifty tons of vile books [and] 3 984 063 obscene pictures’, most of which today would be likely to shock no one, should suggest a sober scepticism about any claim that the latest threat to decency [read “the Internet”] has finally crossed the line of the tolerable: what was once beyond the pale rests comfortably on today’s living room table.”

Laurence Tribe, Professor of Constitutional Law, Harvard University
2.1 Introduction

Much of the sophisticated technology that has made the Internet so powerful has been pioneered by the world of adult entertainment. Sales of pornographic magazines are down because cyberporn is such a big e-commerce market. It has led to world-wide discussion how to regulate the online porn industry. This in turn has led to extra concerns for ISPs worried about their liability for obscene, indecent or criminal content transmitted via their services, whether acting as hosts of the content, Usenet newsgroups or simply providing access.49

  342

In South Africa, the government and the majority of citizens have a heartfelt interest in keeping pornography and other objectionable content from children and unconsenting adults. Furthermore, and equally important, South Africans also have a national interest in protecting freedom of expression. In the past, faceless committees decided what adults might see, read and hear and their decisions were based on the perceived need to preserve an extremely conservative Calvinistic sexual morality. The arts, the publishing and film industries and ordinary people who view them as mature, responsible and often curious members of the human race have been subjected to a system of state censorship. Today, the Bill of Rights in the Constitution50 protects freedom of expression and although no right is absolute, South Africans must demand good reason from anyone wishing to limit their right to hear, see and read what they choose.
 
This section focuses on the regulation of online pornography. Of particular importance is the failed efforts of the American Congress to protect children against pornography, not only because the United States is home to most of the pornography found on the Internet, but also because of the great number of legal precedents available on the subject. The 1996 Films and Publications Act is examined as it applies to the Internet. Finally, suggestions are made on how the legislature in South Africa should proceed to successfully protect children and unconsenting adults, while not chipping away at the expression rights of others. A wealth of reading on this topic is available on the Internet and readers are encouraged to use the links provided in this chapter for a more extensive understanding of the issues.
2.2 The United States of America and online pornography
The right to free speech is protected by the First Amendment of the American Constitution. Generally, sexually explicit expression is protected while obscene expression is not. In defining the obscene, the Supreme Court moved from a view in which the obscene was unprotected by the First Amendment because it was utterly worthless,51 to an approach in which the obscene was unprotected if utterly worthless,52 to an approach in which obscenity is unprotected even if not utterly without worth.53 None of these attempts to define the “obscene” has turned out to be very successful. It has furthermore been stated that all expressions, even those having the slightest redeeming social importance, are to be protected.54

The current test for obscenity is found in Miller v California.55 All three parts of this test must be satisfied before the expression is unprotected by the Constitution:

  • Does the average person applying contemporary community standards find that the work, taken as a whole, appeals to the prurient interest?
  • Does the work depict or describe in a patently offensive way, sexual conduct specifically defined by the applicable state law?
  • Does the work, taken as a whole, lack serious literary, artistic, political or scientific value?

343

In the context of pornography as sex discrimination and the subordination of women, cities such as Minneapolis and Indianapolis have enacted ordinances that distinguish between sexually explicit materials that subordinate or degrade women and those that do not, banning only the former. Some of these ordinances have been drafted by prominent feminists such as Andrea Dworkin and Catherine MacKinnon and have been struck down as unconstitutional inter alia because the courts did not regard them as content-neutral, and amounting to thought control. Expressions that subordinate women are forbidden, while sexual expression that portrays men and women in positions of equality is lawful, no matter how graphic the sexual content.56 Representations of children engaged in sexual conduct57 and the private possession of child pornography58 have been banned.
 
Since 1996 there has been repeated Congressional attempts to regulate pornography on the Internet.
 
2.2.1 The Communications Decency Act

In February 1996 Congress enacted the Communications Decency Act (CDA) as Title V of the Communications Act, which was designed to protect minors from harmful material and control the free flow of pornography on the Internet.

344

Section 223(a)(1) of the CDA prohibited:

“Interstate or foreign communications ... by means of a telecommunication device ... [which allowed for] the transmission of, any comment, request, suggestion, proposal, image, or any other transmission which is obscene or indecent ...”

Furthermore, section 223(d) proscribed the transmission to a specific person or persons under 18 years of age of:

“any comment, request, suggestion, proposal, image, or other communication that, in context, depicts or describes, in terms patently offensive as measured by contemporary standards, sexual or excretory activities or organs ...”

In the subsequent and much publicised case of Reno v American Civil Liberties Union59 the Supreme Court held that the CDA violated the First Amendment. Justice Stevens concluded:

“We are persuaded that the CDA lacks the precision that the First Amendment requires when a statute regulates the contents of speech. In order to deny minors access to potentially harmful speech, the CDA effectively suppresses a large amount of speech that adults have a constitutional right to receive and to address to each other. That burden on adult speech is unacceptable if less restrictive alternatives would be at least as effective in achieving the legitimate purpose that the statute was enacted to serve.”
“Any content-based regulation of the Internet, no matter how benign the purpose, could roast the global village to roast the pig.”
Get the RENO v ACLU judgment at
http://www.ciec.org/SC_appeal/opinion.shtml

The importance of the judgment is that the Supreme Court confirmed that the Internet deserves the same high level of free speech protection afforded to books and other printed media, magazines and casual conversation. Therefore, for legislation regulating online pornography, it must be shown that the least restrictive methods are used to address the government’s concern of protecting children from inappropriate material.

345

Gahtan Free speech rights apply in cyberspace, CDA declared
unconstitutional at http://gahtan.com/alan/articles/cda-lawt.htm

2.2.2 The Child Online Protection Act
Resulting from the CDA being declared unconstitutional by the Supreme Court, Congress passed the Child Online Protection Act60 (COPA) in October 1998. With this Act Congress clearly learned from their previous mistakes in the enactment of the CDA and made provisions for definitions of the terms used and certain defences. Section 231(a)(1) provides that:

“Whomever knowingly and with knowledge of the character of the material, in interstate or foreign commerce by means of the World Wide Web, makes any communications for commercial purposes that is available to any minor and that includes any material that is harmful to minors shall be fined not more that $50,000, imprisoned not more than 6 months, or both.”

Section 231(b)(1) provides for the following defences:

“It is an affirmative defence to prosecution under this section that the defendant, in good faith, has restricted access by minors to material that is harmful to minors –
“(A) by requiring use of a credit card, debit account, adult access code, or adult personal identification number;
“(B) by accepting a digital certificate that verifies age; or
 “(C) by any other reasonable measures that are feasible under available technology.”

Material that is harmful to minors is defined in section 231(e)(6) as:
“any communication, picture, image, graphic image file, article, recording, writing, or other matter of any kind that is obscene or that –
“(A) the average person, applying contemporary community standards, would find, taking the material as a whole and with respect to minors, is designed to appeal to, or is designed to pander to, the prurient interest;
“(B) depicts, describes, or represents, in a manner patently offensive with respect to minors, an actual or simulated sexual act or sexual contact, an actual or simulated normal or perverted sexual act, or a lewd exhibition of the genitals or post-pubescent female breast; and
“(C) taken as a whole, lacks serious literary, artistic, political, or scientific value for minors.”
  346


Minors are defined in section 231(e)(7) as any person under 17 years of age.

 

Get a constitutional analysis of the COPA by the Centre for Democracy and Technology at http://www.cdt.org/speech/constitutional.html

Like its predecessor, the CDA, this Act soon ran into constitutional trouble when a Federal court in Philadelphia blocked its enforcement. In the decision of what came to be known as ACLU v Reno II,61 Judge Lowell Reed stated that:

“fears of prosecution under [the] COPA will result in the self censorship of... online materials in an effort to avoid prosecution, and this Court has concluded... that such fears are reasonable given the breadth of the statute. Such a chilling effect could result in the censoring of constitutionally protected speech, which constitutes an irreparable harm to the plaintiffs.”

“Despite the Court’s personal regret that this preliminary injunction will delay once again the careful protection of our children, I without hesitation acknowledge the duty imposed on the Court and the greater good such duty serves. Indeed, perhaps we do the minors of this country harm if First Amendment protection, which they will with age inherit fully, are chipped away in the name of their protection.”

The decision is an excellent and systematic review of how free speech jurisprudence is to be applied in cyberspace.

Get the ACLU v Reno II judgment at
http://www.cdt.org/speech/ACLUvsRENOdecision.shtml


2.2.3 The McCain Bill (S. 97)
While testimony was heard in the ACLU v Reno II preliminary injunction hearing, Senator McCain introduced a new Internet censorship bill that would require schools and libraries to implement filtering technology if they wanted to receive any government funding.

347


In Mainstream Loudoun v Board of Trustees of the Loudoun County Library,62 the court held as unconstitutional a library’s policy of preventing all patrons from using the library’s computers to access Internet sites containing child pornography, obscene material or material deemed to be harmful to juveniles. To give effect to this policy, the library had installed “X-STOP”, a blocking software package, on all the library’s computers. Applying a strict scrutiny analysis the court found that the sites at issue were limited public forums, and that the policy was neither necessary to further a compelling state interest, nor sufficiently narrow to achieve that end. Other, more narrow means were available to the defendant to achieve its goals, such as use of privacy screens or installation of blocking software on some but not all the computers. The court furthermore held that the policy was an impermissible prior restraint on free speech.

Get more detail on the Mainstream Loudoun judgment at
http://www.phillipsnizer.com/int-art124.htm


The American Library Association (ALA) opposes the use of blocking software in libraries. It argues that libraries have traditionally promoted free expression values by providing books and other information to users regardless of age or income. In a recent policy announcement, the ALA stated:

“Libraries are places of inclusion rather than exclusion. Current blocking/filtering software prevents not only access to what some may consider ’objectionable’ material, but also block information protected by the First Amendment. The result is that legal and useful material will inevitably be blocked.“63
Even if the blocking software is installed on computers only being used by minors, constitutional problems remain. Minors, especially teenagers, have a constitutional right to access many of the resources that have shown to be blocked by user-based blocking programs.64

Letter from the Centre for Democracy and Technology to Senator McCain with a list of alternative tools for protecting children online at http://www.cdt.org/speech/letter/cdtletter2McCain.shtml

Bilstad Obscenity and indecency on the Usenet: The legal and political future of Alt.Sex.Stories at http://www.ascusc.org/jcmc/vol2/issue2/bilstad.html

348

2.2.4 Organisations involved in freedom of speech on the Internet
 The Centre for Democracy and Technology (CDT) is an organisation fighting for the future of the First Amendment and free speech on the Internet. It believes that individual choice and individual control of access to information are the keys to protect free speech online. The CTD supports making user empowerment technology, such as filtering, blocking, monitoring and “Greenspace” tools, available to families who want to protect their children from viewing inappropriate material on the Internet. However, they oppose government attempts to set a standard defining inappropriate material or restricting adult access to constitutionally protected speech. The web site of the CTD hosts a rich source of up-to-date information concerning free speech issues on the Internet.

Get the CDT free speech site at
http://www.cdt.org/speech/

Other organisations involved in free speech on the Internet include:

American Civil Liberties Union (ACLU)
Benton Foundation
Computer Professionals for Social Responsibility (CPSR)
Electronic Freedom Foundation (EFF)
Electronic Privacy Information Centre (EPIC)
Media Access Project
People for the American Way
Progress and Freedom Foundation
Society for Electronic Access
Voters Telecommunications Watch (VTW)

Get the sites of these organisations from
http://www.cdt.org/links/national.html

Links to various discussion papers and other material of the European Parliament on harmful and illegal content on the Internet is available at http://www.cyber-rights.org/eu-watch/
 
UK regulation of child pornography, including legislation, statistics, police operations and background papers at
http://www.cyber-rights.org/reports/ukcases.htm
 
USA regulation of child pornography, including legislation, cases, FBI operations and recent developments at
http://www.cyber-rights.org/reports/uscases.htm
 
International regulation of child pornography, including Germany, Northern Ireland, the Netherlands, Finland, Belgium, Canada, the EU, France, Italy, UNESCO, UNICEF and other international developments at
http://www.cyber-rights.org/reports/interdev.htm

2.3 South Africa and online pornography

The two principal statutes in terms of which censorship was effected in the realm of indecency and obscenity were the Publications Act of 1974 and the Indecent or Obscene Photographic Act of 1967. Both these Acts have now been repealed by the Films and Publications Act of 1996. However, even before it was repealed, the Indecent and Obscene Photographic Act was invalidated by the Constitutional Court in Case v Minister of Safety and Security.65

  349

Other statutes also touch on the subject of pornography. Section 19(6) of the Sexual Offences Act66 and section 160(d)(i) of the Liquor Act67 have been applied to various forms of adult entertainment. At common law the offence of public indecency consists of unlawfully, intentionally and publicly performing an act which tends to deprave or corrupt the morals of others or which outrages the public sense of decency.68

2.3.1 The Films and Publications Act 65 of 1996
This Act, enacted by Parliament in 1996, seeks to provide for the classification of certain films and publications, outlaws child pornography and establishes a Film and Publication Board and a Film and Publications Review Board. Its constitutionality has not yet been scrutinised by the Constitutional Court.

Get the Films and Publications Act at
http://www.polity.org.za/govdocs/legislation/1996/index.html

In April 1999 the Act was amended by the Films and Publications Amendment Act 34 of 1999 to make specific provision for material available on the Internet.

Get the Films and Publications Amendment Act at
http://www.polity.org.za/govdocs/legislation/1999/index.html

350

In terms of section 2 the object of the Act is to:

“(a) regulate the creation, production and distribution of certain publications and certain films by means of classification, the imposition of age restrictions and the giving of consumer advice, due regard being had in particular to the protection of children against sexual exploitation or degradation in publications, films and on the Internet; and
“(b) make the exploitative use of children in pornographic publications, films or on the Internet, punishable.”
A “publication” is defined in the Act as, among others:
“(i) any message or communication, including a visual presentation, placed on any distributed network including, but not limited to the Internet.”69
Furthermore, a “visual presentation” is defined as:
“(a) a drawing, picture, illustration, photograph or image; or
“(b) a drawing, picture, illustration, painting, photograph or image or any combination thereof, produced through or by means of computer software on a screen or a computer printout.”70
It is therefore clear that all forms of pornography on the Internet are a “publication” for purposes of the Act. However, it could be argued that a pornographic video clip should rather be treated as a “film” than a “publication”.71
 
2.3.1.1 Production and possession of publications containing child pornography
In terms of section 27 of the Act, any person who knowingly creates, produces, imports or possesses material that contains a visual presentation of child pornography shall be guilty of an offence. Such a person, if found guilty, may be sentenced to a fine or to imprisonment for a period not exceeding five years, and in the case of aggravating factors, both a fine and imprisonment.72
 
“Child pornography” is defined in section 1 of the Act as:
  351
“any image, real or simulated, however created, depicting a person who is or who is shown as being under the age of 18 years, engaged in sexual conduct or a display of genitals which amounts to sexual exploitation, or participating in, or assisting another person to engage in sexual conduct which amounts to sexual exploitation or degradation of children.”
“Sexual conduct” is defined in Schedule 11 as:  
“genitals in a state of stimulation or arousal; the lewd display of genitals, masturbation, sexual intercourse, which includes anal sexual intercourse; the fondling or touching with any object, of genitals; the penetration of a vagina or anus with any object; oral genital contact; or oral anal contact.”
The prohibition on the possession of child pornography shall not extend to bona fide scientific, documentary or literary publications.73 To institute criminal action or conduct a search for child pornography, the written authority of the attorney-general concerned is required.74
 
The issue of possession creates a number of questions as far as it relates to the Internet. “Possession” is not defined in the Act. It is suggested that an ISP which knows about the nature of the content will be guilty of the offence. On the other hand, an ISP which merely provides access to the material would not be guilty, either of possession or importing the material, because of the knowledge requirement.
 
A further burning question is whether viewing child pornography on a browser would constitute possession or whether it requires a further positive act from the accused to establish knowing possession. The definition of “visual presentation” in section 1 of the Act confirms that an image on a computer screen is a visual presentation. This, however, does not answer the possession question. It is suggested that the possession and knowledge requirements of the offence are only established once the accused downloaded the child pornography onto a computer’s hard drive or onto a disk.
 
It seems that child pornography is not generally available through web sites on the Internet. It is produced individually and distributed through Usenet groups and Internet Relay Chat (IRC) services. Some Usenet newsgroups appoint moderators whose sole job is to screen all the messages before posting, but larger newsgroups such as alt.sex.stories (500 000 users a month) do not appoint such moderators. The problem of policing the Usenet and IRC is further restricted by the use of anonymous postings.
  352
2.3.1.2 Distribution of pornography
In terms of section 28 any person who knowingly distributes a publication which contains a visual presentation or a description of the following, is guilty of an offence:
  • child pornography
  • explicit violent sexual conduct
  • bestiality
  • explicit sexual conduct which degrades a person and which constitutes incitement to cause harm
  • the explicit infliction of or explicit effect of extreme violence which constitutes incitement to cause harm75
These materials are rated XX if visual or X18 if described and the offence does not extend to bona fide scientific, documentary, literary or artistic publications.
 
Furthermore, a person who knowingly distributes a publication that contains visual presentation, simulated or real, of explicit sexual conduct that, in the case of sexual intercourse, includes an explicit presentation of genitals,76 is also guilty of an offence. These publications are rated X18 and the offence does not extend to bona fide scientific, documentary, literary or artistic publications. Offences of this nature carry the same penalties as those for possessing child pornography.
 
It is suggested that ISPs which host material of this nature, would be punishable if it knew of the nature of the content, likewise a person posting material of this nature to any site would be considered to distribute it. As indicated above, the policing of the Usenet and IRC is extremely difficult.
 
However, section 24 creates exemptions in respect of the distribution of certain publications, but only if the publication has been classified by the Films and Publications Board as X18 and only in relation to adult shops, licensed in terms of the Businesses Act,77 and only “if such exhibition or distribution takes place within premises forming part of a building”.78 The “part of a building” requirement seems to exclude web sites, for the time being, but it is suggested that web sites should also be allowed to register as “adult premises”, given that technology is employed to effectively exclude children from accessing such sites.
2.4 Suggested solutions to regulate online pornography

International experience has shown that legislative regulation of the Internet is not always successful. The nation state as regulator is becoming increasingly irrelevant in cyberspace. Laws regulating online content run the risk of infringing constitutional rights and simply cannot keep up with fast-moving technology.

  353

2.4.1 Voluntary user self-regulation
Voluntary self-regulation puts the power of regulation in the Internet user’s hands and is said to be the answer to the intricate problem of content control. Free speech advocates and various organisations have long argued that self-regulation is the preferred kind of Internet regulation. Technology allows users to create their own regulatory principles, to implement and enforce them. The concerns that technology creates, could be addressed by applying technology in return.

The following self-regulatory options are available:

  • Content control options available through an access provider. In the US, three ISPs, America Online, Prodigy and CompuServe, have parental control features in their services that are available to all subscribers without additional charges. For example, America Online allows parents to block private e-mail to and from particular domain names, and access to chat rooms. Furthermore, parents may partially or entirely block access to Usenet groups.
  • User content control via software. As legislative attempts to regulate cyberporn failed in the US, private business stepped in with a wide range of software packages that allow users to filter, block and screen online content. In 1997 the White House called a summit meeting to encourage expression on the Internet to be rated by its speakers and to urge industry leaders to develop and deploy the tools for blocking inappropriate speech.

To enable blocking, filtering and screening software, Internet rating systems were introduced, where content is either rated by the speaker or by a third party.

  354

2.4.1.1 Rating systems
The Platform for Internet Content Selection (PICS) is a rating standard that establishes a consistent way to rate and block online content. In theory, PICS does not incorporate or endorse any particular rating system – the technology is an empty vessel into which different rating systems can be poured. Thus far three rating systems have been developed: SafeSurf, NetShepherd, and the de facto industry standard RSACi. Browsers can be configured to block content with negative ratings or not to accept any content or to block any unrated content. Search engines have also indicated that they may decide not to list sites that have negative ratings or that are undated.79

Of the rating systems in use some, like SafeSurf and RSACi, are self-rating. RSACi, developed by the same group that rates video games, attempts to rate content, e.g. sex and violence, according to objective criteria describing the content. For example, it rates violence from “harmless conflict” to “injured or killed”. Levels of sexual conduct are rated from “passionate kissing” to “clothed sexual touching” to “explicit sexual activity and sex crimes”. However, the context in which content is presented, is not considered, therefore educational material cannot be rated differently than other material. SafeSurf applies a complicated, more contextual, rating system. For example, sexual content is rated from “artistic” and “erotic” to “explicit and crude pornographic”. Other rating systems are third-party rating systems, for example NetShepherd.

355

The long-term effect of these rating systems caused alarm among free speech groups such as the American Civil Liberties Union (ACLU), and self-rating systems are criticised on the following grounds:

  • Self-rating schemes will cause controversial speech to be censored. Take the example of the Critical Path AIDS Project which has a web site that includes safe sex information written in street language with explicit diagrams to reach the widest possible audience. The Project does not want to rate the site as “crude” or “explicit”, but if left unrated, the site will be blocked as an unrated site. If the content is rated, the site will be labelled “pornographic” and blocked from view. As this example shows, the consequences of rating are far from neutral. What is most critical, is that the content in this example is entitled to the highest degree of constitutional protection. Self-rating has never been required from newspapers or books. There is simply no justification for treating the Internet differently.
  • Self-rating is burdensome, unwieldy and costly. Take the example of Art on the Net, a non-profit site that hosts online studios for up-and-coming artists where hundreds of them display their work. The vast majority of the artwork has no sexual content, although there is an occasional Rubenesque painting. The rating systems make no sense when applied to art – one’s masterpiece is another’s pornography. Art on the Net just does not have the time and staff to self-rate over 30 000 web pages. News-related sites are equally difficult to rate – should war pictures be rated as “violent” and therefore blocked from the view of teenagers? If a news article contains a quote of a curse word, is the word rated or the article as a whole?
  • Conversation cannot be rated. The burden of self-rating really hits home when applied to the vibrant conversational areas of the Internet. Most Internet users do not have web sites, but millions of people use e-mail and online discussion forums. A rating requirement would be analogous to requiring everybody to rate his/her telephone, dinner party and street-corner conversations.
  • Self-rating will encourage further government regulation of the Internet. As the self-rating system is established, government will follow with criminal penalties for content that is mis-rated and legislation to block out content that is not rated at all. This will force the speaker of constitutionally protected speech to be his/her own censor.

356

The criticism against self-rating systems begs the question whether the answer is third party rating. Some argue that content rating by an independent third party could minimise the burden of self-rating on speakers and reduce the inaccuracy and mis-rating problems. However, third party ratings pose serious free speech concerns, as it is virtually impossible for any independent organisation to rate the millions of web sites already in existence and the thousands created every day. Therefore the problem with unrated sites still remains.
 
2.4.1.2 Blocking software
With the explosive growth of the Internet and in the wake of recent court battles the pro-censorship lobby lost, the market has responded with a wide variety of user-based blocking software. Each user installs the software on a computer and may turn it on or off at will, so the home user, rather than the outside company, sets the defaults.

These products are, however, not free from serious free speech concerns:

  • Some of these software programs block content based on keywords, and result in such sites as www.middelsex.gov and www.superbowlxxx.com being blocked.
  • A wide range of subjects suited for minors is automatically blocked out, for example safe sex and AIDS prevention information.
  • Most of these products consider their list of blocked speech to be copyright protected and would in some cases not disclose the list or the criteria employed. The effect is that the software company ends up as an undemocratic, non-transparent censor, in some instances even blocking out sites that criticise their products.
The ideal blocking software package should allow users to switch it on and off with a password, allow users to delete the block on some sites and add some sites to the blocked list. The company that developed it should furthermore disclose the list of blocked sites and the criteria used to block those sites.

2.4.2 Upfront declarations in the domain name
Domain Name Server classification has been mentioned as yet another way of regulating pornography and to serve essentially as a declaration of what type of material is hosted on a site. Sex-related web sites could be given new domain names with the suffix xxx. Such an option would make it easier for those wanting to avoid such sites to do so and, on the other hand, make it easier for those who want access to such content to find the sites.

  357

2.4.3 International charter on Internet content regulation
While an international charter might be the long-term model with the most likely prospect of success as regards enforcement, it remains doubtful whether the parties would ever reach consensus because of their different approaches to free expression and equality. The EU and the Association of Southeast Asian Nations halted attempts to regulate Internet content because they could not reach consensus on what to regulate and how to do it.80
3. Hate speech on the Internet
3.1 Introduction
Democratic institutions, led by the media, have generally been successful in simultaneously protecting the right of the lunatic fringe to express their ideas of hate, while marginalising their agendas from mainstream society and culture by refusing to provide unchallenged access to their views. The Internet changed this – virtually overnight – and for low capital investment, hate groups have set up web sites offering everything from how to build a car bomb to promoting racial violence against minorities. No longer constrained by flyers and publications that previously circulated only to a few hundred hard-core believers, these groups have web sites generating hundreds of thousands of visits.81
 
The Simon Wiesenthal Centre and HateWatch, an organisation monitoring hate speech on the Internet, counted 354 hate sites in June 1999. These include:

Neo-Nazi sites 32
Anti-Semitism sites 14
Holocaust-denying sites 25
Christian identity sites 21
Anti-white (black racist) sites 4
Anti-gay sites 7
Anti-Christian sites 5
Anti-Muslim sites 2
Anti-Arab sites 13
White supremacist sites 46
White supremacist sites for women 9
Racist Skinhead sites 13
Ku Klux Klan sites 35
Church of the Creator sites 19

Get the Hatewatch web site at http://www.hatewatch.org
 
Get these hate sites through links from
http://www.hatewatch.org/onlinebigotry.html

358

Don Black, who was once national director of the Knights of the Ku Klux Klan and now runs his own hate site, states that:

“the potential of the Net for organisations and movements such as ours is enormous, ... we’re reaching tens of thousands of people who never before had access to our point of view.”82

Our society hardly needs bigotry, insults and threats in its public life. On the other hand, open and robust debate is a healthy feature of any democracy. People need to be able to vent their frustrations, anger and aspirations. This illustrates the possible tension between free expression and equality. The question is whether online hate should be regulated, and if so, how? The focus of this section is on models employed to regulate hate speech.

Cooper & King Spreading hate on the Internet at
http://www.wiesenthal.com/itn/oped09.htm
3.2 Regulation of hate speech in the United States of America
Although it is beyond the scope of this section to examine hate speech regulation in various different countries, the US is of particular importance because it is home to the vast majority of hate speech sites.
 
The US goes further than any other country in affording protection to hate speech. It is therefore hardly surprising that the vast majority of hate sites on the Internet are hosted in the US. In Smith v Collin83 the right of Nazis to march on a public street in a Jewish suburb, populated by Holocaust survivors, was protected. In Brandenburg v Ohio84 the right of the Ku Klux Klan to call for the expulsion of blacks and Jews in provocative language was protected. In RAV v City of St Paul85 a city ordinance through which a group of white youths was accused of burning a cross on the property of a black family, was declared unconstitutional because it was an impermissible form of content discrimination.

Words that are likely to make the person to whom they are addressed commit an act of violence, referred to as fighting words, have in a number of decisions emerged as an unprotected category of expression.86 Some argue that hate speech on the Internet could never be defined as fighting words, because an Internet user at a remote location is not likely to be provoked into imminent violence, as opposed to a person in a face-to-face confrontation.87 It should however be remembered that hate expression on the Internet will not always be limited to text, and that audio and video might soon be features of online hate expressions.

  359

Several reasons could explain the far-reaching protection afforded to hate speech in the US. The principle that limitations on expression may not be based on disapproval of the content of the expression is at the heart of US constitutional law. A concession on the banning of hate may lead to a slippery slope as far as all forms of unpopular expression are concerned and further limitations may be demanded.
 
Even organisations monitoring hate speech on the Internet feel that it should be protected:
“Does HateWatch believe in censoring hate sites or hate speech on the Internet? Absolutely not. We hold that once politically disagreeable even hateful speech is censored, even for ’good’ reasons, it imperils all speech regardless of content. We have never advocated censoring hate sites or hate speech.”88

The American Civil Rights Union and the Centre for Democracy and Technology also staunchly defend the right to publish hate expression on the Internet, arguing that content found on the Internet is not different from that found in any public library. Attempts to paint the Internet as a uniquely awful source of information for hate-mongers is overshadowed by the fact that, as in any public library, useful information far outweighs the troublesome material.89

Sharkey & Meyer The proliferation of Hate Speech on the Internet: What can be done? at http://wings.buffalo.edu/Complaw/CompLawPapers/sharkey.htm
3.3 Hate expression in South Africa
The protection of free expression in the Constitution90 does not extend to:  
“advocacy of hatred that is based on race, ethnicity, gender, religion, and that constituted an incitement to cause harm.”91
The distribution of hate speech is a criminal offence in terms of the Films and Publications Act.92 Section 29(1) of the Act states that:
“Any person who knowingly distributes a publication which, judged within the context –

“(c) advocates hatred that is based on race, ethnicity, gender or religion, and which constitutes incitement to cause harm, shall be guilty of an offence.”
  360

The offence does not extend to bona fide scientific, documentary, dramatic, artistic, literary and religious publications or to a publication which amounts to a bona fide discussion, argument or opinion on any matter pertaining to religion, belief or conscience.93 The offence carries a sentence of a fine or imprisonment for a period not exceeding five years.94

Only Internet service providers (ISPs) which know the nature of the content would be guilty of distributing hate speech. As possession is not an offence in terms of the Act, Internet users in South Africa may download hate speech from jurisdictions where it is constitutionally protected, but not distribute it further to others.

Get the 1996 Films and Publications Act at
http://www.polity.org.za/govdocs/legislation/1996/act96-065.html
3.4 Attempts at regulating hate speech on the Internet
Germany passed a law in August 1997 which criminalised Holocaust revisionism and neo-Nazi propaganda on the Internet.95 In response to legal threats from the German government ISPs in Germany have blocked access to a Dutch web site containing such material. The site allegedly hosted left-wing political content that is illegal in Germany but fully legal in the Netherlands. As a result of this action, not only the offending site, but also thousands of others that had nothing to do with the offending site, were unavailable to German Internet users. This futile attempt at censorship only drew increased attention to the offending material, resulting in its widespread availability on mirror sites around the world.

Section 13 of the Canadian Human Rights Act forbids the communication of hate messages by means of a telecommunication device. However, the Canadian Human Rights Commission was frustrated in its attempt to close down a site containing Holocaust denying material as the site was hosted in California and the material protected by the US Constitution.

  361

It is possible that more examples like these exist, but they all illustrate the same thing: technology precedes the law. Each legislative initiative to limit or censor free expression on the Internet can potentially be met with a superior technological development that will render the initiative nugatory.96
3.5 Regulating online hate speech in South Africa – alternatives to legislation
The government’s attempts with section 16(2)(b) of the Constitution to protect both groups and individuals from the harmful effects of hate could not be criticised. However, it is suggested that legislation would not be a successful vehicle to enforce this protection. Technological and jurisdictional realities of the Internet move the emphasis from legislation to other forms of regulating harmful expression that could be more effective. A move away from legislation as a regulatory model would not only be more effective, but also avoid potential problems in defining that grey area between hate speech and robust political expression. Furthermore, legislation has the risk of not being the least restrictive means available to restrict free expression on the Internet.
 
A number of alternatives exist to protect both individuals and groups against the harmful effects of hate expression on the Internet.
 
3.5.1 Self-regulation by the individual Internet user
Filtering software can be employed to filter content before it can be accessed. The Anti-Defamation League in the US has released a software program that
filters out hate speech sites, known as ADL HateFilter. The software blocks access to sites deemed anti-Semitic, racist and extremist based on reviews by staff members of the Anti-Defamation League.

The interactive nature of the Internet allows users to respond to content found on the Internet. When hate expression is encountered on bulletin boards or newsgroups the response could be immediate. If the hate expression is found as part of the content on a web site, a response could usually be sent to the webmaster. Those who feel strongly about hate on the Internet thus have the further option of engaging the speaker “right in the cyber battlefield of the Internet”.97 An ISP based in San Francisco was the target of an orchestrated mail bombing campaign from Spanish citizens because the ISP hosted the site of the Basque separatist movement. Although the ISP was forced to shut down temporarily, the controversial site is now hosted on a number of sites globally.98

  362

3.5.2 Regulation via codes of conduct by ISPs
In 1996 the Simon Weisenthal Centre petitioned ISPs such as America Online, CompuServe, Prodigy and others to refuse to host content that promote hate, mayhem and violence. The petition received mixed success, but at least one ISP banned hate groups. HateWatch believes that the market should be the regulator and encourage ISPs to have a “no hate page policy” as part of their contracts with users. Some hosts provide hotlines where users can report hate expression or other illegal content.
 
Regulation by ISPs, however, would be less desirable if it merely replaces the government as censor with the ISP.
Author biography  
Reinhardt Buys grew up on the farm Beerlaagte outside Grootvlei (Mpumalanga) and matriculated from the Hoër Volkskool Heidelberg in 1989. He completed BLC and LLB degrees at the University of Pretoria where he was also elected to the Student Representative Council in 1994. After completion of his articles at Weavind & Weavind Attorneys in Pretoria, Reinhardt completed an LLM degree at the University of Cape Town, focusing on electronic commerce and civil liberties in cyberspace. He currently heads the IT/Internet Law unit at Sonnenberg Hoffmann & Galombik in Cape Town and enrolled for a Postgraduate Diploma in Company Law at the University of Stellenbosch. Reinhardt frequently publishes articles on Internet law and Y2K-related legal issues.

 
362
  1. Burchell Personality rights and freedom of expression (1998) 179
  2. See further chapter 7
  3. Levine Legal pitfalls in Cyberspace: Defamation for Anonymous Communication in Cyberspace (1996) 96 Columbia LR 1526
  4. 776 F. Supp. 135 (S.D.N.Y. 1991)
  5. 361 US 147 (1959) (Back)
  6. 418 US 241 (1974)
  7. No. 031063/94 (N.Y. Sup. Ct.)
  8. Civ. 96-952-A (E.D. Va. 3/21/97)
  9. No. Civ. 97-485 LH/LFG (D. N.M. 1/3/1999)
  10. No. 984751 (Cal. Superior Crt, San Francisco City and County, 23/9/1997) (Back)
  11. Case No. 97-2587 (Fourth District Crt of Appeal, Fla, 14/10/1998)
  12. Neethling, Potgieter & Visser Law of Delict (1993) 320
  13. SAUK v O’Malley 1977 (3) SA 394 (A) 403
  14. Hix Networking Technologies v System Publishers (Pty) Ltd 1997 (1) SA 391 (A)
  15. Section 16 Act 108 of 1996 (Back)
  16. Pakendorf v De Flamingh 1982 (3) SA 146 (A)
  17. 1998 (4) SA 1195 (SCA)
  18. 1996 (2) SA 588 (W)
  19. National Media Ltd v Bogoshi 1998 (4) SA 1195 (SCA); Burchell Personality rights and freedom of expression (1998) 224
  20. Burchell Personality rights and freedom of expression (1998) 124 (Back)
  21. National Media Ltd v Bogoshi 1998 (4) SA 1195 (SCA)
  22. Burchell The Law of Defamation in South Africa (1985) 67
  23. Byrne v Deane (1937) 1 KB 818
  24. African Life Assurance Society Ltd et al v Robinson & Co. Ltd et al (1938) NPD 277 at 295
  25. Vermaak v Van der Merwe 1981 (3) SA 78 (N) (Back)
  26. Hall v Zietsman (1899) 16 SC 213; Martin v Kemlo (1909) 26 SC 457; Van Vliet’s Collection Agency v Schreuder 1939 TPD 265; Pretorius v Niehaus & Another 1960 (3) SA 109 (O); Boberg Annual Survey (1961) 78 SALJ 171; Burchell The Law of Defamation in South Africa (1985) 156
  27. Smith Internet Law and Regulation (1997) 82; Burchell Personality Rights and Freedom of Expression (1998) 128
  28. Pearce v Kevan 1954 (3) SA 910 (D) at 914
  29. Sunday Times (South Africa) 26/1/1997, citing The Telegraph, London
  30. 947 F.Supp. 413 (D AZ 1996) (Back)
  31. Playboy Publishers v Chuckleberry Publishing Inc
  32. [1995] 2 AC 18
  33. (1991) 174 CLR 1
  34. (1993) 176 CLR 433
  35. Smith Internet Law and Regulation (1997) 113 (Back)
  36. 1934 AD 43
  37. Burchell The Law of Defamation in South Africa (1985) 176
  38. [1900] 2 QB 170
  39. 1998 (4) SA 1195 (SCA)
  40. 776 F. Supp. 135 (SDNY 1991) (Back)
  41. Rosenoer Online Defamation (1995) http://www.cyberlaw.com/cylw0595.html
  42. Ibid
  43. Smith Internet Law and Regulation (1997) 83
  44. Ibid
  45. Hird v Wood [1894] 38 SJ 235 (Back)
  46. Burchell Personality Rights and Freedom of Expression (1998) 127
  47. See Mohamed v Jassiem 1996 (1) SA 673 (A)
  48. Chissick & Kelman Electronic Commerce: Law and Practice (1999) 34
  49. Chissick & Kelman Electronic Commerce: Law and Practice (1999) 40
  50. 108 of 1996 (Back)
  51. Roth v United States 354 US 476 (1957)
  52. Memoirs v Massachusetts 383 US 413 (1966)
  53. Miller v California 413 US 15 (1973)
  54. Ibid
  55. 413 US 15 (1973) (Back)
  56. American Booksellers Assoc Inc. v Hudnut 771 F 2nd 323 (7th Cir 1985), 475 US 1001 (1986)
  57. New York v Ferber 458 US 747, 102 Sct 3348 (1982)
  58. Osborne v Ohio 495 US 103, 110 Sct 1691 (1990)
  59. (1997) 2 Butterworths Human Rights Cases 405 (US Supreme Court) No 96-511, as yet unreported judgment, decided 26 June 1997
  60. 47 USC 231 (Back)
  61. Civ. Act. No 98-5591, 1999 US Dist Lexis 735 (ED Pa, 1 February 1999)
  62. No CV 97-2049-A (ED Va, 23 November 1998)
  63. Beeson & Hansen Fahrenheit 451.2: Is cyberspace burning? (1997)
  64. Ibid
  65. 1996 (5) BCLR 609 (CC) (Back)
  66. 23 of 1957
  67. 27 of 1989
  68. S v W 1975 (3) SA 841 (T); S v K 1983 (1) SA 65 (C)
  69. Section 1 Act 65 of 1996
  70. Ibid (Back)
  71. See the definition of “film” in section 1 of the Act
  72. Section 30
  73. Schedule 5
  74. Section 30(3)
  75. Schedule 1 (Back)
  76. Schedule 2
  77. 71 of 1991
  78. Section 24(1)
  79. Beeson & Hansen Fahrenheit 451.2: Is cyberspace burning? (1997)
  80. Cohen Cyber-Censorship: No tickets for speeding on the Information Superhighway (1998) SALJ 710 (Back)
  81. Cooper & King Spreading Hate on the Internet (1996) 1 http://www.wiesenthal.com/itn/oped09.htm
  82. Kanaley Hate groups love the Internet: Free speech flaunts its evil side with recruitment drives (1996) at A12
  83. 439 US 916 (1978)
  84. 395 US 444 (1969)
  85. 112 SCt 2538 (1992) (Back)
  86. Chaplinsky v New Hampshire 315 US 568 (1942); Cox v Louisiana (Cox I) 379 US 536 (1965)
  87. Sharkey & Meyer The Proliferation of Hate Speech on the Internet: What can be done? (1997)
  88. HateWatch Frequently asked questions about HateWatch (1999) http://www.eu.hatewatch.org/faq.html; see also Cooper & King Spreading Hate on the Internet (1996) 1 http://www.wiesenthal.com/itn/oped09.htm
  89. Sharkey & Meyer The Proliferation of Hate Speech on the Internet: What can be done? (1997)
  90. Section 16 (Back)
  91. Section 16(2)(c)
  92. 65 of 1996
  93. Section 29(4)
  94. Section 30
  95. Informations-undkommunikationsdienste-gesetz-IuKDG, 1 August 1997 (Back)
  96. Cohen Cyber-Censorship: No tickets for speeding on the Information Superhighway (1998) SALJ 710
  97. Sharkey & Meyer The Proliferation of Hate Speech on the Internet: What can be done? (1997)
  98. Cohen Cyber-Censorship: No tickets for speeding on the Information Superhighway (1998) SALJ 710 (Back)