'Governance of controversial Internet content in the European Union.'
Simply put, the Internet is a network of computers. Loader defines the Internet as 'a collection of networks and gateways that use the IP protocol suite, and function as a single, co-operative virtual network' (1997:229). Encyclopaedia Britannica (1997) describes the Internet as a 'network connecting many computer networks and based on a common addressing system and communications protocol called Transmission Control Protocol/Internet Protocol (TCP/IP).The technical protocol referred to began on the 'Arpanet', initially sponsored by the US Defense Department's Advanced Research Projects Agency (ARPA), established in 1969 to provide a secure and survivable communications network for organisations engaged in defence related research (Froomkin, 1997). Researchers and academics in other fields began to make use of the network, and eventually the National Science Foundation (NSF), which had created a similar and parallel network called NSFNet, took over much of the TCP/IP technology from ARPANET and established a distributed network of networks capable of handling far greater traffic. NSF continues to maintain the backbone of the network (which carries data at a rate of 45 million bits per second), but Internet protocol development is governed by the Internet Architecture Board, while the Internet Network Information Center (InterNIC) administers the naming of computers and networks (Encyclopaedia Britannica, 1997). Thus has the TCP/IP protocol evolved into the standard for world-wide communication of digital information. The digitisation of information, that is, conversion into binary code, is the foundation of the entire information revolution. Both audio-visual information and text can be converted into this format (Saxby 1990, 1996; Gates, 1996). While the traditional telephone wire is the typical carrier of Internet services, amateur radio, cable television wires, spread spectrum radio, satellite, and fibre optics all have been used to deliver digital information using the same protocol.
The Internet is growing exponentially (Froomkin, 1997). In 1983 there were around 200 computers on the Internet's precursor, the 'Arpanet'. As of January 1993, there were more than 1.3 million computers with a regular connection to the system. In January 1996 there were about 9.4 million Internet host computers. Access doubles approximately every year. As Michael Froomkin says (1997): 'at this rate of growth, the Internet cannot help but penetrate deeply into the general population of industrialised countries.'
The two most successful Internet applications have been electronic mail, known as E-mail)-and the World Wide Web. The Web is a user friendly system for retrieving information in the form of pages. Written in HTML (Hypertext Mark-up Language), web pages contain hypertext 'links'. Browsing programs ('browsers') such as Netscape's Navigator or Microsoft's Explorer present a graphical interface which allows for easy navigation of the web. The hypertext links inserted by document authors refer the reader to other documents using addresses known as Uniform Resource Locators (URLs). A mouse click on a link refers the user to remote documents, images, sounds, or even movies that have been made Internet-accessible. A Web browser can retrieve data via File Transfer Protocol (FTP), Gopher, Telnet or newsgroups (discussion groups), as well as via the HTTP (Hypertext Transfer Protocol) used to transfer hypertext documents (Froomkin, 1997). E-mail - one of the original uses of the Internet - remains a vital application. An estimated 25 billion e-mail messages were exchanged in 1995 (Froomkin, 1997). In the landmark judgement against the Communications Decency Act (CDA) judges grouped Internet communications into six categories:
One-to-one messaging (such as 'e-mail')
One-to-many messaging (such as 'listserv')
Distributed message databases (such as 'USENET newsgroups')
Real time communication (such as 'Internet Relay Chat'),
Real time remote computer utilisation (such as 'Telnet'),
Remote information retrieval (such as 'ftp', 'gopher', and the 'World Wide Web')
(US District Court for the Eastern District of Pennsylvania, 1996)
The Internet is an open architecture, meaning that one computer system can connect to others simply by adhering to the TCP/IP standards. It is not necessary for users of one computer system to use the same hardware or software as another computing system with which they wish to communicate, or even to know the nature of the hardware and software in the other system (Perritt, 1995). It has a 'distributed architecture', in the words of Henry Perritt (1995). This means that different parts of the eventual bundle of information content and related services desired by users may be supplied by different Internet nodes, operated by different persons or entities (Perritt, 1997: Fig. 1). As a communication medium the Internet is non-hierarchical, interactive, and global. Anyone with access can freely publish material which is then available globally and indefinitely. So in contrast with traditional centralised media the Internet has a many-to-many aspect as a medium. It is interactive as opposed to passive.
In seeking to address the question as to what policy instruments are either desirable or feasible a clear understanding of the nature of the Internet is the fulcrum on which the answers must turn. Firstly, the Internet must be understood as a communication medium in some meaningful terms. Secondly, the technical infrastructure of the Internet must be understood in order to gauge which policy instruments are feasible.
On the first question: how does the Internet compare with the traditional media? Crucially do its communications fall into the private or public sphere? To what extent do analogies to print publishing, broadcasting and one-to-one communication by telephone or post apply? Print publishing is traditionally subject to very weak regulation (Commission, 1996b, annex 3) The answers to these questions impinge on what can be taken as desirable regulation, taking cognisance of privacy rights and questions of liability.
Jonathan Wallace and Mark Mangan, in the context of a libertarian argument against the ill-fated CDA, present a survey of communications revolutions from the invention of writing through printing to electronic communications. This yields fascinating insights. Each revolution prompted scares and the first attempts at regulation were inevitably misplaced and based on false understandings of the new medium (1996: 193-233). Wallace and Mangan, in conclusion, argue that the best metaphor by which to understand the Internet is the following: 'Cyberspace is a constellation of printing presses and bookstores' (1996:228). However this analogy fails to take adequate account of the more private elements on communication on the Internet. The closest analogy for E-mail, for instance, must surely be traditional mail. It is this private element which was emphasised in one of the judgements against the CDA. Judges maintained that 'the Internet may fairly be regarded as a never-ending world-wide conversation' (Cunningham, 1996). The range of types of information on the Internet renders both approaches limited, I believe. That range stretches from private text exchanges right through to moving images and sound events which approach the status of traditional television programming, for example the recent participation of a large audience in viewing pictures relayed from the Mars exploration mission. Interestingly, those with a strong civil liberties orientation tend towards the printing analogy. Like Wallace and Mangan, Britain's National Council for Civil Liberties (Bird, 1996) takes such a view in its response to the Commission's Green Paper On The Protection Of Minors And Human Dignity In Audio-visual And Information Services (hereinafter 'The Green Paper'). The NCCL is particularly concerned with discounting analogies to film and computer games, the latter as we shall see below being relevant in the development of the PICS standard.
The various Internet media, principally [the] world-wide web, comprise mainly a text message supported by images and layout, like the pages of a magazine. Sound, and especially video clips, are greedy of bandwidth and unlikely to become the dominant part of it. The material conveyed is therefore not fictional drama or dramatised documentary, and only a little of it conveys game-playing. It is generally news, factual, hobby and 'chat' material, most resembling a spread of newspapers and magazines - in fact it resembles a much wider spread of small personal/organisational newsletters, and particular 'channels' are consumed entirely by choice where one adult usually has strong control of the receiving 'set'. If it resembles anything else it would be individual phone & fax conversations (although it is the published aspect which causes most concerns). Film, and particularly computer games, are a very poor model for the Internet and bear little resemblance to it. (Bird, 1996:10)
There is much that is questionable here. Bandwidth, as I have argued above, is likely to increase with technological developments. The inconvenience of downloading large video or sound files is just that. Many Internet users put up with it. And while games are only one aspect of the Internet, content is not confined to the categories listed, but rather, runs the gamut of human interest. As I will argue in chapter 5 below, pornography in various forms is easily accessed.
The European Commission's 'Communication', I believe indicates a balanced understanding of the radically challenging nature of the new medium, rejecting the analogies based solely on publishing offered above.
A unique characteristic of the Internet is that it functions simultaneously as a medium for publishing and for communication. Unlike in the case of traditional media, the Internet supports a variety of communication modes: one-to-one, one-to-many, many-to-many. An Internet user may 'speak' or 'listen' interchangeably. At any given time, a receiver can and does become content provider, of his own accord, or through 're-posting' of content by a third party. The Internet therefore is radically different from traditional broadcasting. It also differs radically from a traditional telecommunication service. This constant shift from 'publishing mode' to 'private communication mode' - two modes governed traditionally by very different legal regimes - constitutes one of the main challenges of Internet regulation. (Commission 1996a)
One of the reasons why the question of analogies is so relevant is due to its impact on questions of liability for network and service providers. The traditional situation in the field of telecommunications is one of minimal liability, the provider of the means of transmission being seen simply as a neutral carrier. However, in such media as television and radio there has been a greater weight of legal responsibility involved. Essentially the development of digital technology is producing a phenomenon of 'convergence between telecommunications and audio-visual' media (Kaspersen, 1996), producing new quandaries in terms of allocating liability. The question of liability is given heightened importance due to the fact that, rightly or wrongly, network and service providers are seen as gatekeepers with some power over the content carried on their networks. Henrik Kaspersen of the Commission's Legal Advisory Board also points out (1996) the difficulties involved in pursuing sanctions against Internet users responsible for misuse of the medium. These include difficulties in locating and ascertaining the real identity of such users, difficulties finding the necessary evidence of their acts, and the difficulties involved in pursuing cases across national boundaries. Thus the role of network and service providers in the governance equation is a crucial one. Kaspersen maintains that:
Both for the purpose of the identification of the perpetrator and for the removal of the illegal information the co-operation of the network service providers is indispensable. ...Technically, all the data traffic with or by the user passes the system of the access provider. (1996:1)
An access provider is a commercial organisation which allows a user to have access to the resources of the whole network during a fixed quantity of time per week or month for a set fee. Unsurprisingly, the degree to which access providers should be liable is a controversial one. They can claim to be nothing more than the transporter of data over the network, denying any responsibility for its content and refusing any obligation to act in cases of misuse. Any policy concerned with the regulation of controversial Internet content must come to some consistent position on this question of liability.
Some of the theorists covered in the literature review have dealt with the liability issue from diverse perspectives. Henry Perritt (1997) contends that 'one of the by-products of jurisdictional uncertainty is that intermediaries become even more attractive targets than in the past for criminal prosecution and civil lawsuits'. He echoes Kaspersen in maintaining that 'Intermediaries (network access providers) may be easier to identify than originators of harmful content. Furthermore 'intermediaries' may be perceived as having greater financial resources and:
...may be more easily subjected to judicial jurisdiction because they have physical or virtual presence in many more places than content originators, making it easier to execute judgements against them or to use forfeiture as a criminal penalty. (Perritt, 1997)
Ultimately, according to analysts like Perritt (1997), Post (1995) and Reidenberg (1996) this subjection to liability puts access providers in a rule-making position. According to Perritt (1997), 'an intermediary who responds to a threat of civil liability by refusing to handle controversial material is in effect engaging in a rule-making function.'
Turning to our second question; how can the infrastructure of the Internet be understood so as to give clear indications on the technical limits of policy? Firstly the scale of accessible material is massive, meaning that scale in itself limits the feasibility of monitoring, in contrast with traditional media such as television. Secondly, the form in which information is transferred under Internet protocols makes it highly unamenable to monitoring. This entails a system known as 'packet switching', a method by which messages are subdivided into smaller standardised packets which are then routed independently to their destinations via an indeterminate number of intermediaries (Froomkin, 1997; US District Court for the Eastern District of Pennsylvania, 1996). This decentralised, anarchic, method of sending information appealed to the Internet's early sponsor, the Defense Department, which was intrigued by a communications network that could continue to function even if a nuclear war destroyed a large fraction of the system. Packet switching 'makes it difficult for anyone, even a government, to block or monitor information flows originating from large numbers of users' (Froomkin, 1997)
Thirdly, the Internet engenders a new space - 'cyberspace' - which presents regulators with an indeterminate terrain:
Cyberspace does not merely weaken the proposition that the effects of conduct taking place at a particular geographical location are felt primarily by persons at or near that location, it renders it virtually unintelligible; it does not merely render geographical boundaries ever more permeable, it obliterates them entirely, because geographical location itself is both indeterminate and irrelevant for transactions on the net. (Johnson and Post, 1997b)
There is no relationship between machine addressing (the location a computer in the virtual space created by the TCP/IP Protocol addresses and domain names) and the physical location of the machine and its user (Johnson and Post, 1997a, b). At present, while there is a convention of whereby many addresses include country signifiers (e.g. '*.uk', '*.ie'), no countries legally stipulate that machines within their geographic boundaries use such a signifier. Moreover, there are no registries that can prevent individuals holding such a country designator from assigning that designator to machines in other countries, nor is there a means of limiting the use of technical protocols such as the Telnet protocol to access the Internet from other Internet access points (with other addresses). Nor is the use of non-country-specific designators (e.g., '*.com', '*.org') restricted in any way (Johnson and Post, 1997b: note 6). In addition to the addressing problem location is rendered indeterminate by the features of some Internet interactions. Johnson and Post (1997b) instance distributed databases (such as Usenet newsgroups) which 'have no single place of existence at all, but are aggregates of information stored on a large (and constantly changing) number of individual machines at any time.' And location is irrelevant in the sense that network servers are equally accessible from everywhere. A Web site in any corner of the network can be accessed with equivalent transmission speed and message quality from any other corner of the network, and therefore the effects of whatever information is available at that site are felt simultaneously and equally in all jurisdictions, independent of their 'distance' from one another. Needless to say, this indeterminacy of location in cyberspace poses grave problems for those who would govern content. Besides the difficulties in locating transgressors there are problems which arise from the externalities of any policy implemented (which Post and Johnson call 'spillover'). This raises issues of legitimacy, since in such a situation many of those affected by a policy would have no say in its making.
Lastly, the availability of anonymous remailing facilities (Froomkin, 1997; Perritt, 1997) allows Internet users to avoid being held responsible for the contents of the messages they send. The effect is to make both censorship and information export restrictions 'nearly impossible to enforce'(Froomkin, 1997). Anonymous remailing is made possible by encryption technology. Digital information is highly amenable to encryption, a process whereby information is encoded using mathematical algorithms. The information is then only readable with the correct 'key', a password in effect. A process known as public key cryptography was invented in the 1977 by mathematicians Rivest, Shamir and Adelman. Their RSA algorithm did away with the need to exchange codebooks or keys before the sending of a message (Campbell, 1997b). The security of encrypted information depends to a large degree on the length of the key used, and RSA encryption with a key length of 112 bits (strong encryption) is apparently uncrackable. Any policy aimed at the regulation of communication on the Internet must face this issue. Some governments have been keen to either limit public access to strong encryption, or to oblige users to leave copies of their keys with 'Trusted Third Parties' (Campbell, 1997b). But the interests at stake for governments are complicated by the fact that strong encryption represents a 'holy grail' in terms of security of electronic commerce.
Given these elements in the information infrastructure presented by the Internet - volume of data, packet switching, indeterminate terrain and encryption supported anonymity - it's governance poses a huge challenge for would be regulators. Froomkin (1997) concludes that any effort to censor the Internet organised at the national level (or below) is likely to fail. He cites John Gilmore's pithy comment in order to summarise the situation: 'the Net interprets censorship as damage and routes around it.'
Daniel
Dunne.com