There was a time when people believed that the Internet was a realm that censors could not touch. But TCP/IP communications are carried by physical means; the people who originate them live in the real world and can be jailed or killed. Governments have shown that they can interfere quite effectively with Internet communications, just as they can jam broadcasts or tap telephone lines.
Access Denied provides an up-to-date account of how, and to what extent, countries around the world do this. This is a very good book when it sticks to the facts, though not nearly so impressive in its normative content and recommendations.
The book consists of three main parts: a set of six articles, a group of regional overviews covering the entire world, and a more detailed but less comprehensive country-by-country section.
We learn that different countries take a variety of approaches to preventing or blocking unwanted content. Some directly control ISP's or the gateways which they use to the Internet through a state monopoly. Others license service providers and can shut down those which don't comply with government demands. Others prosecute the providers of illegal content or the people who view prohibited material.
The technical approaches vary as well. Some block IP addresses, others go by domains, and a smaller number attempt live filtering by keywords in the content. Domain blocking can lead to the censorship of huge amounts of harmless content for the sake of keeping the population from seeing a few items.
Motivations affect what kind of content is blocked. These may include protection of public morals (however locally understood), preventing criticism of the dominant religion, and keeping critics of the government or opponents of the dominant party from being heard. My impression from reading the book is that silencing criticism and ridicule of the rulers is the most widespread motive.
Not surprisingly, the worst offenders are usually theocratic, socialist, or communist states. (I'm including China, which has become a fascist state but still uses communist rhetoric.) No country is completely untainted, though. Germany and France have a prohibition on "propaganda against the democratic constitutional order." Canada has a "hate speech" law. The United States has mandated filtering in libraries and schools, and it took a Supreme Court decision to stop the Communications Decency Act. Still, the United States continues to have the strongest protections for free speech in the world.
It's surprising, then, that the United States is also the dominant source of software which governments use to block free speech. Several American companies, such as Secure Computing, help governments to block content and "provide extensive lists of URLs managed by proprietary methods."
Strong as Access Denied is on facts, it's often weakened by its treatment of ideas. An egregious example occurs in "Filtering and the International System: A Question of Commitment." This article refers to the International Covenant on Civil and Political Rights (ICCPR) as containing a "broad unquibbling guarantee of freedom of expression," when in fact it mandates some forms of censorship:
Article 20
1. Any propaganda for war shall be prohibited by law.
2. Any advocacy of national, racial or religious hatred that constitutes incitement to discrimination, hostility or violence shall be prohibited by law.
In fact, the essay quotes that article verbatim, then proceeds to act as if it weren't there! Under that prohibition, saying that another country is a threat and war is necessary is plainly prohibited regardless of circumstances, and it can easily be stretched to justify bans on "hostile" statements about the official religion.
In "Good for Liberty, Bad for Security? Global Civil Society and the Securitization of the Internet," we have this astounding bit of bias:
To address these issues, we breakdown global civil society into three spheres of agency: civic networks, resistance networks, and dark nets. Civic networks refer to progressive environmental, peace, and social justice movements that are most typically associated with the term civil society.
If you aren't on the political left, Ronald Deibert and Rafal Rohozinski have written you completely out of "civil society" and into the shadowy fringes of resistance or outright criminality.
In Corporate Ethics on a Filtered Internet is a proposal which isn't in the same league with those two excesses, but is a bad idea that might get implemented. Jonathan Zittrain and John Palfrey suggest a voluntary corporate code of ethics with regard to governmental demands for censorship and disclosure. This is a reasonable idea, if well implemented. But they then propose that such a code, if successful, should be adopted as law. Voluntary codes and laws serve two different purposes; one specifies what is best to do, the other what people should be forced to do or not do. For example, some suggested language includes "a commitment to locating servers in places that are unlikely to result in the unethical, forced disclosure of user information." As a guideline, this is fine. As a legal mandate, the words "unlikely to result" are so vague that they would create a minefield for businesses.
The readers who will get the most value out of Access Denied are those who can put its factual material to good use. This includes people trying to alert the public of censorship practices, advocates of pro-free speech reform, and implementers of countervailing technologies.