The Australian Computer Society is the professional association for those in the computing and information technology fields. Established in 1966, the ACS has over 15,000 members and on a per capita basis is one of the largest computer societies in the world. Activities are announced in the Usenet newsgroup 'aus.org.acs'. Information is available via e-mail from 'info@acs.org.au' or Gopher at URL: gopher://acs-gopher.mit.csu.edu.au:1605/00/acs, courtesy of Charles Sturt University.
This paper contains proposed Society policy, but is subject to further discussion and ratification. Copyright in this document belongs to the ACS. The Society welcomes the copying and dissemination of the document, provided that it is not done so for profit, and that evidence of the authorship and this copyright notice are copied and disseminated with it.
Tom Worthington
Director of the Community Affairs Board
Australian Computer Society Inc.
G.P.O. Box 446, Canberra ACT 2601
Tel: +61 6 247 4830
Email: tomw@adfa.oz.au
Fax: +61 6 249 6419
This paper's purpose is to establish the policy of the professional body of information technologists in Australia, the Australian Computer Society, in relation to the public interest in network services.
The elements of a 'national information infrastructure' currently being debated include the long-standing public switched telephone network, the more recent analogue mobile telephone services, the new digital mobile telecommunications services, and television signal reticulation, including terrestrial and satellite broadcast and narrowcast, and cable TV. Improvements in those technologies are generating significant pressure for change. Added to that is a new element: the explosive growth of the Internet, and its rapid maturation from a restricted electronic message transmission service, for a limited range of locations, with an arcane user interface, to a vehicle for the provision of usable and useful services to individuals in a wide range of occupations and locations, and all for very low cost to the end-user. Associated with the Internet's development is the rapid emergence of a de facto dispersed 'electronic library'.
The paper commences with introductions to key elements of the emergent technologies, and to the economics and politics associated with them. The nature of an information infrastructure (II) and information services is outlined, and the use of metaphors, and especially the 'information superhighway', is discussed. Public fears about what may go wrong with the II initiative are collated, the role of government considered, and necessary policy measures identified. Government actions are vital, in relation to clarifying the public's requirements, surveying the terrain and ensuring orderliness in the process. On the other hand, it is important that government does not place itself in the driving seat and unnecessarily stultify a dynamic process.
This paper was originally placed on-line in plain text format using the AARNET FTP server, in 1994. Public access to that copy was lost after AARNET was transferred to Telstra. This HTML version is URL: http://www.acs.org.au/president/1997/acsnet/acsnet.htm and was constructed from one prepared by Paul English of the University of Western Sydney. The text version is available at URL: http://www.acs.org.au/president/1997/acsnet/acsnet.txt as well, as is an abstract URL: http://www.acs.org.au/president/1997/acsnet/.
While considerable changes have occurred in the on-line area in the three years since this paper was written, no changes have been made to the text. The observations made in 1994 have stood the test of time.
It should be remembered that in 1994 the Internet was not known to the general public. The general view of IT professionals was that the Internet would be quickly replaced by OSI technologies, such as X.400 e-mail. Much has changed since then.
A new ACS home page is available at: http://www.acs.org.au/ in place of the Gopher service.
The mailing list link@wombat.anu.edu.au has moved. Details are available from the new home page for LINK at: http://ningaui.anu.edu.au/link/ The importance of LINK to the establishment of on-line policy in Australia has yet to be fully realised. It was only well after this paper was finished, I discovered that policy advisors on both sides of politics, as well as staff of several federal agencies, had been reading the drafts and discussion on LINK. In this way the content of the paper went directly into the policy process, even before it had been completed or submitted. I later described this process as the Internet conspiracy.
It should be noted that this paper, dated 17 May 1994, is marked "draft", but no later version was produced. The paper in this form was as submitted to the ASTEC Working Group on Research Data Networks; the Broadband Services Expert Group; the Bulletin Boards Task Force; and the Senate Standing Committee on Industry, Science, Technology, Transport, Communications and Infrastructure.
Tom Worthington MACS
President of the Australian Computer Society
24 August 1997
There is currently very active discussion about the direction of telecommunications policy in Australia. The matter is especially complicated by the convergence of a number of hitherto fairly discrete areas, viz. telephone, television and data. There is also a great deal of excitement being generated by the initiative of the Clinton/Gore Administration referred to as the 'National Information Infrastructure' (NII). It is vital that information technology (IT) professionals be active in the policy formulation process. The Australian Computer Society is one of the most important and largest of the associations which represent those professionals.
This paper represents the policy of the Australian Computer Society on the subject of information infrastructure and network services in Australia. It has been prepared on behalf of the Society by its present and immediate-past Directors of Community Affairs. The paper commences with an outline of current and emergent Internet services. It then identifies and discusses social and economic considerations relating to access to such services, and presents a set of policy proposals.
The Information Infrastructure (II) movement has been primarily driven in the United States, but other advanced nations are considering related initiatives, Australia among them. In the United States, the main driving forces are the progressive inter-weaving of:
The capacity of the fibre-optic cables used by the local and long-distance telephone operators is sufficient that the telecommunications companies ('telcos') are capable of competing with cable TV companies. The cable TV companies may soon be in a position to offer additional services which will be to some extent competitive with services offered traditionally by telcos. Satellite and terrestrial broadcast and narrowcast capabilities are increasingly competitive with particular cable-based services. The Internet is increasingly able to provide (near-gratis) services comparable to some of those offered by the telcos.
Added to this competitive cauldron is the fact that the telcos are very heavily regulated at both local and long-distance levels, whereas the cable TV companies are subject to somewhat less intervention, and the Internet has developed only during the last decade, and is largely free from government interference. The very large and influential telecommunications and entertainment companies are lobbying very hard for changes in, and reduction of, the regulatory regime.
The existing Australian infrastructure, economics and politics differ in important ways from those in the United States. Australia has a single provider of local telephone networking and services, Telecom/Telstra. It has two long-distance operators, Telecom and Optus (later to be joined by a third). It has three mobile telephone services providers, Telecom, Optus and Vodafone. But whereas 60% of the population of the United States has access to television via co-axial cable into the home, cable TV is at present virtually non-existent in Australia. Licences have been issued and a service is under development, but its viability and coverage are yet to become clear. Satellite services exist and are being extended.
Many different transmission technologies are available and becoming available. They are outlined in Exhibit 1.
Exhibit 1: Competing Transmission Technologies
- fibre-optic cable, which has high bandwidth, and is largely in place in Telecom's trunks, but which will be very expensive and slow to lay between exchanges and central business district locations, let alone all community and domestic sites;
- terrestrial microwave, which is used for some transmissions as an alternative to long-haul cable;
- analogue cellular mobile telecommunications, which provide connections sufficiently reliable for low-grade voice services, but not for other forms of transmission, and is only economic in densely populated areas;
- digital cellular mobile telecommunications, which provides better-quality connections capable of supporting more than just voice, but it is also only economic in densely populated areas, and conversations cannot currently be reliably sustained when a device crosses successive cell-boundaries;
- geo-synchronous (high-orbit) satellites, which have a wide footprint, particularly suitable for broadcast services, and relative stability once placed in orbit, but are expensive to establish and result in a transmission delay that is too long for some purposes; and
- low-orbit satellites, which have a narrower footprint, move relative to the earth's surface, have lower stability, and are dependent on more sophisticated tracking and control mechanisms. On the other hand, they do not have as significant a transmission delay.
Discussion about these capabilities has sometimes been couched in such a manner as to suggest that one particular technology will be the winner, and the others losers. In respect of particular services, that may well turn out to be true; but there is a wide array of services, and the diversity of both demand and technology characteristics is so great that most of the available transmission media are likely to find niches which they can dominate.
The brisk developments in these various technologies have resulted in progressive redefinition of the businesses in which the various companies are engaged. A telco used to concentrate on telephone services, but fax traffic has become very important to its customers and its revenue generation; and the bandwidth it has available enables it to now contemplate offering additional services more widely, some of which edge into what was previously the entertainment sector. Meanwhile, cable TV companies are recognising the demand by consumers for more than just choice among multiple concurrent film broadcasts. They want the ability to provide feedback up the line, and it is claimed that they want 'videos-on-demand'.
If cable TV companies deliver these capabilities, they are capable of offering alternatives to some telco services. Computer users, in the meantime, have hired basic telecommunications capacity from the telcos, strung these links together into the Internet, and now, in effect, offer services which directly compete with those of the telcos. The popularity of the Internet has forced commercial providers of electronic mail services to provide gateways to it. Hence several hitherto distinct kinds of corporation are in the process of rapid convergence.
An important outcome of these developments in technological capability has been an expectation of ubiquitous computing and communications. A person's primary place of work is increasingly only one of the locations from which he or she seeks electronic services, and to which the infrastructure must reach. Many people work at least partly from home; airports make space available in which travellers can make connections; plans are in place to provide connections in locations such as newsagents, the foyers of office-buildings and shopping centres, and taxis; and new-model aircraft are being designed to provide mid-air digital connection to terrestrial networks.
The Internet's growth has been so explosive, and it is sufficiently threatening to telcos, cable TV companies and value-added network providers, that a brief introduction is warranted.
Technically, the Internet is merely that collection of networks which are inter-connected using the a particular family of standards referred to as 'TCP/IP'. It is a highly non-hierarchical, 'democratically' structured, collaborative arrangement. Seemingly ironically, it emerged from the ARPAnet, which was sponsored by the U.S. Department of Defense to link the enormous number of private sector and university-based researchers working on defense-funded projects. In fact the irony is superficial: the military places a high priority on robustness, and dispersed, non-heirarchical networks exhibit greater resilience than centralised 'star-topology' networks. A brief overview of the Internet is provided in Attachment 1. AARNet (the Australian Academic and Research Network) is a segment of the Internet which is located in Australia. It was established, and is presently run, under the auspices of the Australian Vice-Chancellor's Council (AVCC), and has to date been funded primarily by the Universities, the CSIRO, and the Australian Research Council.
The Internet has spawned an extraordinary explosion in creativity. Vast amounts of inter-personal communication are undertaken using it, and large numbers of documents are accessible from servers located not only in North America, but also in many other countries throughout the world. Outside North America, Australia is one of the most vigorous participants, both in terms of usage and contributions.
Initially, access to data on the Internet required considerable technical capability; but the days of obscure services such as 'ftp' and 'archie' are quickly receding. Contemporary services are based on:
With professional-quality services now available, the kinds of documents which are accessible via the Internet are no longer limited to abstruse computer science papers, discussions about pop-groups, films and life on the net, games software and university student polemics. For example, university library catalogues are accessible over it; specialist collections of scientific papers and data can be located using it; government committees make their discussion papers available over it; lobby groups prepare their submissions to government committees on it; and government agencies in places as remote as Washington DC, California, Wisconsin, Tasmania and now in Canberra, publish reports, Bills, submissions and proceedings on servers connected to it. To sustain some semblance of order and accessibility within the Internet's constructive anarchy, programs busy themselves constructing directories of participating sites and people, and registers of documents.
The services are maturing beyond structured data and text. Sound, graphics and images can also be transmitted. Text messages can be delivered to fax machines, circumventing the telcos' much more expensive direct-dialled fax services. Synchronous conversations are being supported, particularly in text, but also using sound. Video (moving-image) transmission is being trialled in many locations, and may well make the 'video-phone' widely available before the telcos can launch 'official' telephone-with-picture services. If it were left alone, the Internet, or rather corporations exploiting it, may well mature into alternative value-added telecommunications service-providers, in direct competition with telcos, and perhaps cable TV operators, in respect of at least some of their mainstream services.
In a decade's time, the Internet may prove to have been just an experiment which leads to the inter-network of the early twenty-first century. Nonetheless, its exponential growth, the scale it has achieved, the services it provides, and the vibrancy of the electronic communities it has spawned, have attracted the attention of telcos, cable TV companies, and governments.
Development of the Internet protocol was funded by the U.S. defense establishment, and undertaken by scientists to satisfy their own needs. The resultant connectivity, and data and software resources, were originally accessible primarily by scientists and other researchers. As Internet services matured, access to them became more desirable. Companies and government agencies have increasingly secured access not only for their research staff, but also for executives, managers and operational staff. Indeed, the volume of traffic generated by non-academic users overtook that from academics as long ago as 1991. It is clear that companies providing telecommunications-based services can no longer ignore the Internet, and are no longer ignoring it.
The design of a networked nation needs to be such that users are unaware of the technical features upon which they depend. Instead, they need to be able to think in terms natural to them, yet still gain access to, exploit, learn more about, and contribute to, their electronic environment. In order to deliver this environment to users, it is conventional and useful to distinguish two levels: the infrastructure. and the services which depend on it. This section distinguishes between the two concepts.
Infrastructure is, according to the Macquarie Dictionary, "a basic framework or underlying foundation", upon which the remainder of a structure can be built. The distinction between foundation and superstructure is somewhat arbitrary, but infrastructure is generally depended upon by at least one, and generally several, other elements. The interpretation used in this paper is that an information infrastructure comprises the set of inter-related but separately addressable needs listed in Exhibit 2.
Exhibit 2: Information Infrastructures
- high-capacity backbones and associated switching mechanisms, which can support the transmission of diverse kinds of data associated with many different services with significantly varying characteristics, and which provide a high level of reliability, resilience, redundancy and robustness
- ubiquitous tails between the backbones and the users, both connection-based (i.e. using various kinds of cabling) and wireless (i.e. using some part of the electromagnetic spectrum). These may need to be, for cost reasons, of relatively narrow bandwidth
- broadband tails to support particular uses which involve high-volume, high-priority and/or synchronised data-flows, such as those for video-conferencing and multi-site visualisation work. These will generally be expensive, and hence likely to be restricted to locations which can justify the investments involved
public terminal equipment, including both general-purpose work- and play-stations and specific-purpose terminals, in public locations such as libraries, shopping centres and transport interchanges
A wide diversity of services has already been delivered in particular contexts, and many additional services, variants and combinations will emerge. The information infrastructure must make all of these available in order to bring about a networked nation. The list in Exhibit 3 is meant to be illustrative rather than exhaustive, and is roughly structured according to the tail-capacity needed to support it.
Exhibit 3: Indicative List of Information Services
- low tail-capacity requirements
- electronic mail
- electronic newspapers
- electronic encyclopaedias and other reference material
- classified advertising
- business data transfer
- community bulletin boards
- security services and utility meter reading
- public information services (such as information kiosks)
- data transfer for working from home (telecommuting/teleworking)
- education services including remote classroom teaching (text-only)
- home banking
- home shopping (text-only)
- moderate tail-capacity requirements
- multi-media mail, including video (i.e. asynchronous messaging)
- modest-quality video-telephone (i.e. synchronous conversations)
- high quality audio
- interactive television
- interactive multimedia (i.e. on-line video games)
- video conferencing (including 'virtual meetings')
- education services, e.g. remote classroom teaching(including video)
- home shopping (including image and/or video)
- high tail-capacity requirements
- videos-on-demand (i.e. the electronic video shop)
- high definition TV (HDTV)
- medical services including remote diagnostics using imaging
- linkages among high-performance computers
In general terms at least, it is desirable that the infrastructure be capable of supporting any and all services, despite their potentially different requirements. For some services, however, special features will be essential, in particular those outlined in Exhibit 4. It appears likely that different carrier technologies will have and retain competitive advantage over others in relation to services with particular 'special feature' profiles. This implies either multiple infrastructures, or multiple backbones within a single, integrated national information infrastructure.
Exhibit 4: Special Features
- high priority. These depend on virtual connections being established and sustained between pairs, or among sets, of communicating locations, to support real-time data transmission
- synchronicity between related streams of data, typically video and sound
- high degree of robustness and resilience
- especially high end-to-end data transmission capacity
- high degree of security
To facilitate commercial provision of services, it is important that the infrastructure support convenient forms of network-based value-transfer. Current approaches based on insecure credit-card numbers need to be initially supplemented by encrypted transmission of value-data, and in due course replaced by more sophisticated alternatives such as electronic cash based on chip-cards. This requires the incorporation of appropriate features into the infrastructure.
A considerable amount of analogical thinking will inevitably be used in policy discussions, because the issues confronting government regulatory agencies and parliamentarians are new and unfamiliar. Particularly in the United States, but increasingly also in Australia, the II is commonly referred to as the 'Information Superhighway' and equivalents in other languages.
Metaphors need to be carefully chosen, to avoid the risk of the regulatory decision-making being made on the basis of serious misunderstandings. There are many problems with the term 'superhighway'. For one thing, it's a little long for common usage, and the term 'infobahn' has emerged to address that deficiency. More importantly, however, 'highway' has many negative associations, ranging from traffic jams, accidents and speeding tickets, to tollways. It implies massive investment, and a swathe cut through the countryside.
What's needed in a metaphor is that it be graphic, attractive, and contain many useful analogical features and few unfortunate ones. Its efficacy must be judged from the perspective not of the well-informed specialists, but of those people who need to grasp the new idea. In this case, the audience is non-Internet users, the general media, officers and executives in relevant regulatory agencies and key Parliamentarians.
With current technologies, an II comprises both backbones or trunks (to carry large volumes of traffic between major conurbations) and tails (to connect individual sites to the backbone). The 'highway' notion corresponds to the backbones, but fails to encompass the tails. This has led some commentators to use grotesque extensions of the metaphor, such as 'on-ramps' and 'cloverleaf interchanges'. Moreover, there are several senses in which railroads would be a better analogy than main roads, because once a message is on the net it submits to active management by it, rather than being a free agent. If road or rail transport imagery is to be persevered with, then the tails to industrial and commercial sites and to homes would be by-roads or branch lines. Hence it would be more appropriate to compare the II with the complete road transport infrastructure, or the entire rail network.
An alternative approach might be a reticulation model, along the lines of the way in which supplies of water, electricity and gas reach us. This has the deficiency that these models tend to be designed for one-way rather than two-way flows.
But these ideas still miss an important element of the II of the very near future. By-roads, branch lines and pipelines imply physical connections, whereas wireless tails are already technically feasible, and becoming commercially available in such forms as GSM digital mobile networks. Analogue services support only voice, but digital technology will also enable such transmissions as fax, data and email. Using portable devices such as laptops and hand-held 'personal digital assistants' (PDAs), people will soon be sending and receiving email from park-benches and taxis. To convey the idea requires images like broadcasting and narrowcasting. The poverty of the 'superhighway' idea becomes even more apparent when it is appreciated that backbones can also be wireless, in particular through satellite and terrestrial microwave technologies. To express wireless backbones and tails in terms of road, rail or pipeline metaphors is challenging in the extreme.
A number of alternative notions are popular among Internet and sci-fi aficionados. 'The Net', 'the Matrix', and 'the Web' all have appeal, but lack the power to convey the essential difference to newcomers. 'Cyberspace' may be technically misleading, but its virgin nature enabled William Gibson, Bruce Sterling and others to paint meaning on a blank canvas. It still seems unlikely, however, that an idea popularly defined as 'shared hallucination' will be saleable to hard-headed public servants, parliamentarians and Cabinet Ministers.
Another idea that has circulated is the use of ecological and geographic idioms, including arable land, pastoral land, parkland and wilderness areas, perhaps augmented by storehouses, libraries and delivery trucks. A further notion, less appealing at first sight, but meaningful nonetheless, is of a 'superfungus', whose spores travel invisibly. There is scope for a more comprehensively connection-less image, such as the 'data-field' (not as in 'paddock', but as in 'magnetic field').
Finally, perhaps a metaphor should be chosen which is driven not by the technical way in which the II works, but rather by the social possibilities it delivers. For example, images such as electronic 'collective', 'community' and 'society' may better convey the real meaning, or, in Australian terms, 'the bush telegraph'.
Realistically, though, it takes a great deal to launch a new metaphor, and the II will doubtless continue to be discussed with the help and hindrance of transport-related metaphors. One small refinement would be to replace the unhelpful prefix 'super' with some safe, catchall, and preferably very open-ended alternative, resulting in such alternatives as the short and philosophically appropriate 'meta-bahn' and the more romantic and stylish 'cyber-strada'.
This section's purpose is to argue the critical importance of the public's interest in the information infrastructure debates. The term 'the public' is a vague generalisation. It is used here to refer to the non-corporate participants in society, including individuals, informal groups and communities of common interest, not-for-profit incorporated associations, and small businesses. It is used in distinction to large corporations, including government agencies and authorities, government business enterprises and large joint stock corporations.
The Internet and the culture surrounding it cannot be understood on the basis only of corporations seeking to exploit the resources it gives access to. The net has long been a playground for individuals, and a significant proportion of the traffic has been of little direct benefit to their employers. It has been of great indirect importance, however, as it has resulted in the foundation of electronic communities, and an explosion of creativity.
Access to networked resources is quickly assuming the shape of a public utility, i.e. a service which needs to be available to all, on an equitable basis. In the same manner in which advanced western societies have recognised clean water, electricity, refrigerators and the telephone as facilities which the majority of people should have access to in their normal work and home environments, so too is network access becoming a reasonable expectation of a resident in a civilised, information society.
Unlike many utilities, however, an information infrastructure must support a great diversity of services. This is partly due to the degree of openness and cultural variation, but also because change and creativity are dependent on diversity.
Closely associated with the rise of the Internet has been the re-definition of the notion of 'library'. Documents are increasingly coming into life in electronic form, and are increasingly available via the Internet. Even in respect of the vast collections of important documents which exist only in printed form, increasing numbers are being scanned into electronic form and made available over the net.
The librarianship profession is already in the process of adaptation by reducing the emphasis on its curatorship function, and accentuating its 'knowledge navigation' expertise. The significance of this is that libraries have for many years been the means whereby equitable access to many kinds of knowledge, and to literature, have been provided to people in all walks of life. Equity of access to the new, dispersed electronic library must be sustained.
Another necessary condition for public access to network services is the widespread ability to use them. Public book-libraries could only be of benefit where social programs were instituted to promote literacy in the population. The notion of literacy, and the focus of literacy programs, need to reflect the new electronic environment.
The whole population needs skills in order to operate computer equipment and use the basic application software. Moreover, the nature of the media makes new styles of communication necessary. For example, replying to email demands care with the selection of addressees; and commenting on documents, and on other people's comments about documents, requires some form of linking arrangements or text to ensure that recipients understand to what the comment relates.
The people who wrote the Australian Constitution at the end of the nineteenth century recognised that equitable, universal access to transport and telecommunications services was fundamental to the cohesion of the federation. A century later, equitable, universal access to the much wider range of networking services is at least as critical to the cohesion of our information-based society, as roads and telegraph and telephone links were to the agrarian and emergent industrial society of 1894.
A further important aspect of telephone services, recently re-asserted by the regulatory authority Austel, is the freedom from monitoring of and interference with conversations. The cluster of rights associated with freedom of thought, freedom of communication and freedom of assembly must be embedded in the conception and architecture of the information infrastructure.
The pattern of the public switched telephone network (PSTN) has always been 'symmetrical', i.e. all parties have comparable capacity to send and receive signals. The pattern of cable television has, on the other hand, tended to be 'asymmetrical', or 'broadcast' in style, i.e. suppliers have high sending capacity, and drive their products and services outwards, whereas consumers have little or no capability to send messages, and adopt a passive role, choosing among the supplier-decided alternatives placed before them.
The new information infrastructure could be easily conceived and developed using the asymmetrical model, supplier-driven, with consumers passively selecting among 500 largely similar channels (as the popular image of next-generation cable TV has it). This would be greatly against the public interest. The model to be preferred is one more nearly symmetrical, in which consumers have at least the ability to provide feedback, and even to initiate.
With sufficient outgoing capacity, people can actively participate in decision-making, and can contribute to the pool of available materials. A measure of the success of the infrastructure will be the extent to which it enables appropriately motivated and organised individuals to be service providers, performing electronic publishing and maintaining collections for access by others.
This capability can be applied in other contexts as well. Of particular importance is the use of the information infrastructure to encourage the public to exercise their democratic rights and responsibilities.
The justification for the public interest being the primary determinant of the II's architecture might be criticised as being idealistic, because it seems to be based on 'renaissance mankind' notions, e.g. "modern people will at last shake off their primitiveness and innate savagery, because they will be enabled by information technology to be educated, and to fully participate in their democratic society".
The public interest argument is, however, not based just on social idealism. The notions of social equity, equal opportunity, and assistance for the socially and educationally disadvantaged have become well-accepted during recent decades, and parties of the right do not appear inclined to turn the clock back: these principles have become established underpinnings of contemporary Australian society.
Another element of the public interest argument for access to networked services can be advanced at the political level. Genuine democracy, if it is indeed an aim of Australian society rather than a mere slogan, demands that information infrastructure comprise inherently democratic structures, provide access to information to the population as a whole, and deny large protected spaces within which narrow interests can arrange resource allocation and public policy to suit their own interests.
Finally, there is a crucial economic argument. Countries with high standards of living and hence high labour costs must harness their relatively well-educated population's creativity. This is encapsulated in the phrase 'the clever country'. Linked with this idea, 'distance education' must become much more than just a means to attain geographical equity for far-flung populations; it must become a way of life for a workforce continually preparing itself for the next change in market demands. Universal access to network services is a pre-condition for any society which seeks to sustain its well-being in a world in which an increasingly large proportion of countries are becoming clever.
There are also potential benefits to be gained from increased international telecommunications linkages, because if a greater proportion of the population have the opportunity to interact with people in other countries, improved mutual understanding and tolerance may result. Moreover, Internet services are designed to be robust, and are therefore appropriate for third-world countries whose underlying telecommunications may be, and may remain, fairly limited. There is therefore scope for contributions by advanced nations like Australia to under-developed and newly industrialising countries.
This section has canvassed the following propositions:
From the perspective of the general public, serious concern exists about the shape of the infrastructure which may emerge from the current wave of enthusiasm. This section seeks to summarise those concerns.
The debates have been dominated by the Clinton Administration's 'National Information Infrastructure' initiative. The discussions that are proceeding apace in Washington DC are very interesting, and the material that is becoming available from United States sources is highly valuable. There is a danger that assumptions may be made in Australia that the nature of the debate, and the conclusions reached, in the United States are directly relevant to Australia's needs. There are significant differences between the situations in the two countries, and Australia needs to determine its own priorities.
At the other extreme lies another danger. In pursuing what it perceives to be its own best interests, Australia may adopt too parochial an outlook. The PSTN's inter-connections with the telephone systems of other countries and access to film libraries and 'newsfeeds' controlled by organisations outside Australia are just some of the important elements of Australia's inter-connectedness with the rest of the world. The Internet is increasingly important, and is inherently superordinate to national jurisdictions. It is vital that Australia conceive of its information infrastructure as being a component within the emergent 'global information infrastructure', and ensure continued two-way access to and from people, and data and processing resources, throughout the world.
There is a danger that the agenda may be captured, or at least excessively influenced, by large organisations of a variety of kinds:
It is, of course, appropriate that the information infrastructure debates provide appropriate opportunity for these organisations to seek to protect their interests. Policy decisions must, however, reflect the interests of the Australian public above all, and the interests of particular corporations only where they are shown to make a vital contribution to the public interest. The ethos of the Internet favours the general good of its electronic citizens, not sectional self-interest, and this ethos also needs to permeate the networked nation. For the public to have confidence in the information infrastructure, the decision-making process about its architecture needs to be undertaken with the maximum practicable degree of participation and visibility.
There is a real fear that charged services will not be complementary to existing services, but will swamp them. The Society's stance is that commercial services must be able to develop, and even be encouraged to develop, but the II's architecture must be conceived and implemented so as to enable the non-commercial, mutual-service ethos of existing communities to be sustained.
This implies formal recognition within the II's architecture of both for-profit and not-for-profit sectors, and segments within sectors, the avoidance of biases against non-commercial electronic communities, and the ability of not-for-profit communities to keep themselves free from commercial incursions.
The worst-case scenario would be an attempt to create the II by a government agency or a tightly regulated monopoly in the form of a government commission. Nearly as inimical to the public's interest in a diverse set of competing services would be tight government regulation of the shape which the infrastructure was to adopt, because this would result in a lack of diversity in carriers and in carrier-technologies. An example of such a mistake would be for the government to pick one or more winners among ISDN, ATM, fibre-optic versus various kinds of copper cable, geosynchronous-orbit satellite, low-orbit satellite, analogue cellular, digital cellular, etc. technologies.
It is vital that government not intervene in relation to the choice among carrier technologies without very good reason, in particular unless and until it is clear that:
Although it may seem wasteful for multiple carrier technologies to be implemented, there are good reasons why each of several different approaches may find a sufficient market to justify the investment.
A particular concern is that the architecture might reflect the hitherto common pattern of corporate-sender / passive-consumer. This was consistent with technologies which intrinsically supported only broadcast approaches to telecommunications (often because the capacity would support only a small number of senders). It is not consistent with the new environment, in which both two-way communications and consumer-initiated selective transmission are economically feasible. The architecture must facilitate participation by the public in both the provision and consumption of services.
In the United States, the first round of actions by government agencies in the information infrastructure issue has included moves to ensure that data flows can be readily eaves-dropped. In Australia, some of the first suggestions made have been for the censorship of data flows on the Internet, due to the scope for it to be used to disseminate pornographic and racist materials.
It is inconsistent with the notion of a free society to commence with a design requirement that the content of communications be subject to the purview of government agencies. That would be equivalent to designing public spaces like parks and street-corners in such a manner that conversations could be monitored, requiring all mail to be left unsealed to facilitate inspection, requiring prior submission of all published materials to the Chief Censor, and outlawing the use of vague language and unauthorised dialects on the telephone (or, indeed, anywhere else).
Controls over content should be post facto only, i.e. published material which comes to the notice of regulatory authorities should be prosecuted if it offends the law, and the law should be modified to ensure that the standards applied to other kinds of communication are also applied to network-based communications. But no pre-emptive powers should be built into the infrastructure. Pre-justified court-issued warrants should be necessary before tapping of network traffic is undertaken. The design of the infrastructure should not be subverted in order to facilitate interception.
Similarly, there is a concern that the provision of services might be permitted only subject to prior licencing of the service and/or the service-provider. This would be a boon to those corporations which were granted licences, because the barriers to entry by competitors would be established and maintained by the State rather than by a cartel of existing suppliers, freeing them at once of both the costs involved and the threat of trade practices intervention. But licensing would stultify creativity, and reduce diversity of supply, leading to higher prices. Some industries will undergo a revolution as a result of application of the information infrastructure (not least of them the publishing, entertainment and news industries), to the great benefit of the public, although not necessarily of the corporations presently active in those industries. Again, controls should be exercised post facto over services which breach the law; the force of the law should not be used to pre-determine which corporations will profit from the new order.
The availability of access to the infrastructure may be differentially advantageous to some kinds of organisations and people at the expense of others. The primary grounds which may intentionally or unwittingly become bases for discrimination are:
In many cases, access inequities are likely to compound existing discrimination.
One of the primary ways in which bias can come about against the public in general, and against particular classes of people, is through the selection of the basis for charging for the information infrastructure, connection to it, and its use.
It was argued in Clarke (1993) that the costs of networks are essentially stepped-fixed-cost in nature, with negligible variable costs per message. This has the implication that virtually all costing (i.e. attribution of costs to particular uses of the infrastructure) is arbitrary, in that different parties can argue quite rationally why another party should bear the greater share of the total costs. The greatest powers of persuasion lie with the corporations. There is accordingly a significant risk that the public may directly bear a significant share of the costs.
One approach which would be inimical to the public interest would be to 'front-end load' the charges to consumers, small groups and communities, incorporated associations and small business. This might involve high connection fees, high software or set-up costs for each small organisation, or high annual subscriptions. This would have the highly detrimental effect of discouraging a great many people from ever connecting to the information infrastructure, or from gaining access to particular services.
It is vital that entry costs to the public be kept very low. For most mainstream services, usage costs should also be kept to little or nothing. For other services, it may be appropriate for charges to be levied on a partial cost-recovery basis. In the case of 'luxury' services (such as what the French refer to as 'messageries rose'), even a full cost-recovery user-pays approach may be appropriate.
A related concern is that small organisations and people who are in remote locations may be forced to bear a large proportion of the costs involved in the outreach of the infrastructure and the services to them. As argued earlier, this is contrary to the principles on which the Commonwealth of Australia was founded. Any per-transaction or data transmission volume charges which are levied should be distance-independent, at least for traffic within Australia, as is already the case with data transmission via the public packet-switched network, Austpac.
7.9 A 'Tower of Babel' of Standards
In the event that proprietary protocols were to dominate, then inter-connectivity among services providers, and inter-operability among services, would be restricted. This is largely overcome in the Internet context. It is highly desirable that, once each new carrier-technology and each new service stabilises, standards be generally open and non-proprietary. Standardisation may require encouragement through government policies, and in extreme instances even mandating through legislation.
In the event that proprietary protocols were to dominate, then inter-connectivity among services providers, and inter-operability among services, would be restricted. This is largely overcome in the Internet context. It is highly desirable that, once each new carrier-technology and each new service stabilises, standards be generally open and non-proprietary. Standardisation may require encouragement through government policies, and in extreme instances even mandating through legislation.
There is a fear that the enthusiasm of security, law enforcement and other control-oriented agencies to use the II as a means of monitoring the population will unduly influence its design. The Society supports the ability of the Courts to provide appropriate agencies with the legal authority to intercept and monitor flows. The Society does not support the treatment of interceptability and monitorability of traffic as a requirement of the information infrastructure. To do this would not only warp the architecture, but would also create political power even more dangerous than the criminal activities it would assist in detecting and prosecuting.
This concern about monitoring of network traffic is particularly pronounced in the United States, because its National Security Agency is seeking to sustain its ability to intercept communications by exercising control over the use of cryptographic techniques. This has led to the current controversy over the so-called 'Clipper-Chip' for telephones and 'Tessera-Chip' for data transmission. There have been no public signs of Australian security agencies seeking to control the use of cryptography, and the Society welcomes that restraint.
This section considers the role of government generally, then proposes a specific agenda for government in relation to firstly the information infrastructure, and secondly services available over that infrastructure.
Historically, it has been normal for infrastructure to be established, and in many cases also maintained, partly or even entirely by government agencies. Ports, rail and road transport corridors, water, electricity and gas, health and education are common examples. In addition, in most countries of the world, the provision of basic telecommunications infrastructure is undertaken by government agencies.
In the context of the United States and the Internet, Reinhardt (1994) argued that: "The government is not planning to dig a trench from New York to San Francisco, fill it with fiber-optic cables, and call it a data highway. Rather, the information highway will be privately built, owned, and operated: the Feds will encourage its development only through research funding, standards efforts and changes in regulations". The United States has its own particular approach to the provision of infrastructure, which has had both significant successes and abject failures. Most countries recognise a rather larger role for government, and in many a substantial degree of cooperation exists between the public and private sectors. Australia is presently positioned closer to the United States approach than that of, for example, European countries and Japan.
The conventional presumptions in Australia in the 1990s are that government has the responsibilities to:
The Society expresses concern that, in its efforts to reduce its interventionism, the Commonwealth Government has withdrawn so far that it is failing to provide leadership, moral encouragement and substantive stimulation on some important matters. Because it is so vital to Australia's future, it is imperative that the Commonwealth Government adopt a very positive stance in relation to the Information Infrastructure.
Policy cannot exist in a vacuum, but must be targetted at a particular setting. The proposals in this section are formulated specifically to address the circumstances which exist in Australia in the mid-1990s. Australia is an advanced nation, in relatively very calm transition from a strongly anglo-saxon-celtic towards a multicultural society. The distribution of income and wealth are less widely spread than in many nations, and there is a substantial welfare net, which may become increasingly important if the growth in productivity continues to outstrip the growth in demand, and unemployment remains high, or grows further. Secondary education is free to age 17 or 18, a considerable proportion of young people complete that stage, and tertiary education and training are very accessible.
Organisations and individuals throughout Australia are sophisticated users of information technology generally. In a number of areas, Australian organisations are among the world's leaders, including in the finance, mining and aviation industries, and in taxation, social security, customs and government budgettary management. The existing telecommunications infrastructure is among the best in the world in terms of availability, quality of service and cost. The means whereby the telecommunications infrastructure is provided has been progressively changed from a government commission along the lines of the European PTT model, by way of a government business enterprise, to a blend of monopoly, duopoly and open market, with further evolution planned.
Australia's population is heavily concentrated in a few urban centres, with small numbers scattered over the remainder of a vast continent. It is difficult to achieve equity in such a context, regionally, and even racially.
The Society argues below for a series of policy measures in relation to the infrastructure itself, and services offered over it.
There are areas within the information technology industry in which market failure is occurring; for example, one of the authors has called for constructive government intervention in relation to the application of electronic data interchange (EDI) to both purchasing and international trade. This is because both are regarded by all parties as being highly desirable, and even inevitable, but the breakthrough is not occurring, because the benefits do not accrue to the parties which must make the investment.
There has been, and continues to be, a great deal of energy and willingness on the part of not-for-profit organisations and altruistic individuals to provide data, software and services via the Internet. There appear to be many emergent opportunities for corporations to sell services on or via the Internet, especially to other corporations, but to some degree also to individuals.
There are, however, serious signs of concern at the level of the infrastructure itself. Very substantial long-distance and international capacity is installed, but is currently priced very high in order to ensure profit from the investment. These prices are falling only very slowly. There appear to be key areas in which monopoly conditions still exist, and hence direct intervention may be required. The conditions, however, are not such as to require governments to assume responsibility for the entire undertaking.
Consideration is needed as to whether AARNet represents the basis of the information infrastructure, or merely a prototype. Strengths and weaknesses of the Internet generally are well-described in CPSR (1993). The weaknesses are addressable, albeit in some cases slowly. It is therefore tenable to make either choice. Moreover, it does not appear to be unduly wasteful of resources if the decision is left to the marketplace. It would therefore be beneficial for both AARNet and alternative services to be allowed to develop in parallel. Hence basic telecommunications links should be readily available to intermediaries and end-users.
The underlying principles on which the remainder of the policy proposals are based are expressed in Exhibit 5.
Exhibit 5: Cornerstones of Government II Policy
- the central elements of the information infrastructure require substantial public funding, at least in the early years of development, in order to ensure that the manner in which the costs are allocated does not inhibit use, nor bias developments against the interests of the Australian public
BUT
the further development of the information infrastructure needs to be facilitated by the encouragement of competition among many suppliers of diverse technologies
Some features of an entirely market-driven infrastructure are unlikely to satisfy the needs of Australian society, and some intervention is needed. Important public interests which need to be protected, and which are likely to require government action, are identified in Exhibit 6.
Exhibit 6: Public Interests in Need of Protection
- universality of access to a basic (but nonetheless substantial) level of service:
- for consumers, community groups, incorporated associations and small businesses;
- for sectors which are at least in part dependent on government funding, and in particular:
- education;
- research;
- health; and
- libraries;
This refers not only to the possibility of connection, but also to the practicality and costs involved;
- diversity of connection products and modes, and of carriers;
- distance-independence of transmission costs; and
- civil rights, and in particular:
- freedom of access to information;
- freedom of communication; and
- freedom from monitoring.
To ensure that the market addresses these specific matters, the government has a range of approaches at its disposal. These are expressed in Exhibits 7a to 7d inclusive, in terms of leadership, stimulation, coordination and measured, not precipitate, regulation.
Exhibit 7a: Governmental Leadership Measures
- establishment, and clear and consistent communication, of the policy direction in relation to the exploitation of information technology, consistent with, and as a means of achieving, socio-economic objectives
- targeted funding for the central elements of the infrastructure, especially during the early years of the II initiative
Exhibit 7b: Governmental Stimulatory Measures
- grants for research into relevant information technology. These should not be restricted to the existing Cooperative Research Centre for Research Data Networks, but should be openly available
- grants for community-based applications and prototypes. The United States Government's NII program has made $26 million available in this category in 1994 and anticipates making about $100 million annually available during the next few years. An equivalent Australian fund would make an additional, earmarked $2 million available per annum, rising to $10 million
- grants for research into implications of relevant information technology and its applications
- stimulation of government agency participation, through the careful structuring of incentives to provide services via the infrastructure, and to use services available over it
- stimulation of appreciation by Parliamentarians of the importance of the information infrastructure, including secure connection of the Extended Parliamentary Network to it, and the provision of outreach services by the Parliament to the Australian public
- reversal of the trend towards 'user-pays' for government data. This is stifling research, and is counter-productive to an informed society. In any case, much of the information is of the nature of a 'public good', and hence it is unlikely that cost-recovery can be effective or economically justifiable. Data gathered under compulsion of law, or using public funds, should be freely available to the public
- careful examination of the role of intellectual property law in innovation, and amendments to the law to ensure that creativity is not hampered
Exhibit 7c: Governmental Co-ordinative Measures
- sponsorship of a public policy advisory body representing all interested parties, including commercial interests, user organisations of all kinds (especially the research and education sectors which have driven the development of Internet services), regulators and public interest groups
- sponsorship of well-publicised regional public interest summits, to provide a focal point for public education, the fleshing out of policy, and discussions about key aspects of the networked nation
- sponsorship of the development and use of open, non-proprietary standards in order to achieve inter-connectivity and inter-operability. Rather than through the blunt instrument of mandating particular protocols, it would be preferable to predicate government agency participation on the use of nominated standards
Exhibit 7d: Governmental Regulatory Measures
Adaptation of existing laws to deal with the new networked environment, and maintain an effective balance between the interests of originators and owners of data and services, and users of data and services. In extreme and well-justified cases, this might include licensing. Particular areas in which activities may prove to be necessary include:
- the removal of obstacles to diversity among carrier-technologies, carriers and services
- the removal of obstacles to network-based value-transfer. Change is likely to favour value-added network operators over financial institutions. This is likely to require amendments to the law, and adaptation not only by banks, but also by such organisations as the Reserve Bank, the Treasury and the Australian Payment Systems Council
- where monopolies remain or emerge, or physical dangers arise, the encouragement, and in extreme cases, imposition of operating standards
- the clear specification in law of community standards in relation to such matters as pornography and the incitement of racism
- the refinement of existing enforcement regimes to ensure that the community's standards are sustained
Some kinds of what might be referred to as 'high-science' research need high-capacity tails, and synchronous transmission. These requirements are somewhat in conflict with the relatively low capacity and largely asynchronous needs which dominate the rest of the research, education and library sectors. The response in the United States has been the National Research and Education Network (NREN), which is to be a government-funded backbone specifically for this class of usage. Consideration could be given to the funding of such a network for Australian 'high-science' research.
Another segment which has a requirement for high-capacity tails is distance education. An initiative in this regard is under development under the auspices of the Department of Employment, Education & Training (DEET). It is highly desirable that a pluralistic approach be taken to the emergence of the nation's information infrastructure, provided that individual initiatives do not undermine the need for a much broader range of access to services, and an adequate degree of commonality and inter-operability is designed into them.
At the level of services, government again does not need to, and should not, adopt a general role of services provider, nor of services planner, nor of services regulator. The market appears to be, subject to a few provisos, sufficiently enthusiastic and dynamic to ensure appropriate development of services, and access by Australians to services available in other countries.
Individual agencies, however, should be active participants both as users and as providers of network services. Of especial importance are such services as electronic data interchange (EDI - structured business transactions transmitted electronically rather than using hard-copy documents). In this area a number of agencies of the federal government are already leaders, including the Australian Customs Service, the Australian Taxation Office, the Department of Finance, and Purchasing Australia; and considerable progress has also been made in some State Governments, especially in N.S.W. Public access to government-maintained databases must also be facilitated; the Australian Government Publishing Service's Government On-Line Directory (GOLD) is one example of such a service.
Care is needed in regard to the pricing policies adopted by agencies which offer services or data of interest to corporations, other government agencies, researchers, and the public generally. The recent trend towards 'user-pays' pricing has resulted in instances of data priced at an estimate of full cost, which has been above what the market would bear, and has significantly reduced its availability and use. It is important that the same policy be adopted as by the Clinton Administration in its NII initiative, viz. gratis availability of basic data, and pricing of the remainder at the cost of dissemination only, i.e. without any attempt to recover the costs of collection and maintenance.
Once again, some intervention is likely to be necessary in respect of some specific matters. In particular, action is likely to be necessary in relation to socially desirable services which are not emerging due to market failure, including:
These may be dealt with through targetted funding and subsidies.
In addition, law enforcement is of course necessary in relation to services which break the law. Legislative amendments to ensure clarity of the law, and, in extreme and well-justified cases, licensing, may also be necessary in respect of some kinds of services provided using the networked nation's information infrastructure.
A networked nation is emerging. It is critical that the public's interest in the services, and in the underlying infrastructure, be reflected at all stages in its development and use. Specific, targetted government actions will contribute to the process. Excessive governmental involvement in activities which will develop satisfactorily anyway, would, on the other hand, retard and distort the available services and the underlying infrastructure.
The government must recognise the importance of its multiple roles:
Australia's information technology academics have placed Australia among the top few nations in relation to the implementation of networking. This key national asset needs to be recognised and exploited for the economic and social benefit of all Australians. Australia's IT professionals, through such bodies as the Australian Computer Society, need to continue to promote the country's expertise to our Parliamentarians, community leaders and the community as a whole.
Note: The authors have been particularly assisted by materials from the mailing list link@wombat.anu.edu.au, the October 1993 policy statement by Computer Professionals for Social Responsibility of Washington DC, Tony Barry of the A.N.U. Centre for Networked Access to Scholarly Information, and Eric Wainwright, Deputy National Librarian. Attachment 1: An Overview of the Internet and AARNet (from Clarke 1993)
The Australian Academic and Research Network (AARNet) is a means for inter-connecting the networks of the organisations making up the Australian research community. It uses the TCP/IP (Transmission Control Protocol / Internet Protocol) suite of standards, although it also supports smaller volumes of DECnet, Appletalk and X.25 transmissions. It functions as the Australian segment of the worldwide 'Internet', which is the set of networks inter-connected using TCP/IP.
TCP/IP networks use a connectionless, packet-switching mechanism. This means that there is no committed connection between the two end-points of a transmission, and that data does not travel in a single, unbroken stream from sender to receiver. Instead, the data is broken into packets (of about 200-2,000 bytes) and sent from node to node along the network.
Packet-switched networks are distinctly different from connection-based networks such as the Public Switched Telephone Network (PSTN) and on-line sessions between workstations and servers. These dedicate a series of arcs to make up a connection, and commit them for the duration of a session or call, rather than just for the time it takes to send a single message. The packet-switching approach denies priority to messages within pre-arranged conversations, and as a result real-time transmissions such as voice and video are likely to sound and appear jerky to the receiver. In return, packet-switching makes much more efficient use of the arcs and nodes in the network, provides much greater total throughput, and is accordingly much cheaper for a given volume of traffic.
Each node on the inter-network (roughly speaking, that means the gateway processor on each organisation's internal network) co-operates with the remainder by acting as a switch or exchange, and passing data along towards its target. In order to do this, some of the nodes act as 'name-servers', and maintain a set of routing tables which determine where, depending on its ultimate objective, each packet has to be sent next.
The Internet emerged during the 1980s from the Defence Advanced Research Project Agency's ARPANet. Since AARNet was implemented in 1990, it has grown at a dramatic rate, initially in terms of the number of corporate networks it connected, and continuously in terms of the volume of data it carries. It is essentially a non-hierarchic network, i.e. there is no central controlling point (although there is a central service point for management and trouble-shooting). Its functioning depends on co-operation among many processors and hence also among the organisations which own them.
To the end-user, AARNet is almost transparent. It provides an extended address-space of machines and users, which comprises the entire worldwide Internet. It has high resilience to downtime on individual nodes and arcs, and very high reliability and efficiency. The delivery time for a message is well under a second anywhere within Australia, and usually under a second anywhere in the world. The speed and convenience of data transmission results in greater immediacy in human communications, and has unleashed an explosion of new ideas and services. Attachment 2: The New Electronic Support Technologies (from Clarke 1994a)
In order to impose a limited degree of order on a dynamic field, this section somewhat arbitrarily distinguishes between the tools and services which came to prominence during the decade of the 1980s, and those which have recently emerged or are presently emerging.
During the 1970s, computing had involved singular resources controlled by a priestly caste. During the 1980s, this evolved to a position in which large numbers of inexpensive standalone microcomputers (commonly but imprecisely called 'personal computers') were much more readily accessible by people who could directly benefit from them, in a much less controlled environment. Where 'dumb' terminals had previously offered, to approved users only, pre-defined and restricted functionality, these machines provided power for a modest investment to anyone who wished to capture, store, process and publish data.
What has been usually referred to as 'the marriage of computing and telecommunications' enabled hitherto independent micro-computers to talk with one another, and (by emulating dumb terminals) to talk to mainframe equipment. Progressively, users' own microcomputers were linked with centrally controlled macro-processors (mainframes, 'mini-computers' and mid-range machines) to perform cooperative tasks in networks of dispersed machine intelligence. The old hierarchy of a central mainframe with slave terminals was inverted, and the users' workstations became the primary focus, with local, corporate and wide area networks placing 'servers' at the 'client's' disposal.
Accompanying these developments were great improvements in the learnability and usability of software development tools, of utility software and of applications. Scrolling displays and exotic command languages which were inimical to convenient use by non-specialists were replaced by menu-driven and icon-based human-computer interfaces. For the arbitrariness of the old ideas has been substituted consistent machine behavior, enabling the intuition users develop from their experiences with the first few functions they meet to be generalised to the learning of the remainder. The combination of these various factors has brought about the dispersion and democratisation of computing.
A wide range of tools very important to research were delivered or significantly improved during this phase. The management and analysis of structured, numerical data were better supported, and access to databases was facilitated by the alternative technologies of on-line access to remote databases and directories, and local optical storage (CD-ROM). Within free-text databases, the navigability improved, through the growing availability of so-called 'hypertext', i.e. the ability to roam along the inherent multi-dimensional paths in a document, rather than being limited to the linear sequence of the text (Bolter 1991). The concept has since been generalised to 'hypermedia' for compound documents, i.e. those which contain data in more than one format (Marcus 1991).
Document preparation was facilitated and researchers became less dependent on document preparation staff, not only in the sense of word processing, but also in regard to the preparation of diagrams and the incorporation of diagrams and images into text. Even more emphatically, the scope for refinement and revision was increased, with outcomes whose quality (at least potentially) was higher in terms of both content and presentation.
The long-standing dependence of academics on professional and relatively expensive design, production-editing and printing services was demolished. Desk-top publishing enables a team of a lone academic and a single moderately skilled support staff-member to produce conference proceedings, journals and monographs of a quality and at a cost directly comparable to, and in many cases superior to, those of the traditional publishing houses, and to do so much more quickly. The functions in which publishers retain significant advantages have been reduced to such areas as marketing, distribution and capital access.
During this period, communications between researchers were greatly enhanced. The telefax was the most explosive single change during the 1980s, but there was a vast increase also in the number of people connected to national networks such as Bitnet (U.S.A.), JANet (U.K.), EARN (Continental Europe) and ACSnet (Australia). The various networks were then linked via what began as a domestic United States initiative, but has since become the world-wide Internet.
What was originally a very basic and at best semi-reliable email service for unformatted text, has been progressively enhanced. Mailing-lists enable multiple copies of messages to be transmitted to many recipients with a single command. Mailing-lists can be stored centrally, rather than copies being maintained on multiple machines. Updates may be undertaken centrally, or the power to insert or delete one's own address may be vested in the individual, using emailed commands and so-called 'list-server' software.
A complement to the mailing list is the bulletin-board or news group, which stores messages in a database, displays data about the date, time, sender and subject, and enables individuals to read such messages as they choose. Because of the explosion of mail and news-groups, posting to some of the more influential mailing-lists and news-groups is not freely accessible, but is moderated by kind souls who dedicate much time to encouraging a sufficient volume of useful traffic, and filtering out the less constructive messages.
Remote login capabilities at local and distant servers enable access by pre-authorised users to specialised processing capabilities and databases. In addition, limited access is provided to all comers, to such openly accessible data as the catalogue systems of hundreds of university libraries, and the vast volume of material stored in areas which are labelled (less than clearly) 'anonymous ftp'. This is a form of publication which is at once very cheap and very open-ended, because anyone can use any form of computer-based aid to locate it, and copy from it.
By the beginning of the 1990s, the tools and services outlined in this section had been comprehensively prototyped by computer scientists and other devotees. The developments were by no means confined to the United States; for example, reflecting their deep-seated fear of isolation from the world mainstream and strong desire to overcome 'the tyranny of distance', Australians generate a significant volume of traffic over the Internet. For reviews of contemporary Internet services, see Benedikt (1989), Scientific American (1991), Kehoe (1992), Krol (1992) and LaQuey (1992).
No longer was information technology just about computing qua computation; it was about communication as well: "[an organisation] is constituted as a network of recurrent conversations ... computers are a tool for conducting the network of conversations" (Winograd and Flores 1985, pp.168, 172). The 1980s had seen the establishment of the basis for more widespread use of these tools, the development of more sophisticated services, and the customising of tools and services to the needs of particular disciplines, workgroups and individuals.
With many of the important breakthroughs in hardware and communications already achieved, developments in the early 1990's have been primarily in the software area. The implicit assumption of previous decades had been that individual users worked in isolation, or they cooperated with others but only by sharing the same database. A recurrent theme in the 1990s is the support of collaborative work between individuals, among members of localised workgroups, among widely dispersed workgroups, throughout organisations and even across organisational boundaries.
On-line access capabilities are now moving beyond library catalogues to full-text and image storage and retrieval. The data are increasingly available at the researcher's own workstation, rather than only on a specialised terminal in a library. Statistical databases and collections such as satellite images of weather and landforms are also becoming more widely available over the Internet. Meanwhile significant progress is being made in assisting researchers to locate material they need; in particular, Archie and Veronica provide centralised and frequently updated lists of files around the Internet; the Wide Area Information Service (WAIS) provides convenient but powerful search facilities, and a prototype distributed hypertext search product called World Wide Web (WWW) is also available.
Access to specialised computational capabilities on remote machines is also being facilitated through such initiatives as the U.S. supercomputer network, NFSnet. The emergent National Research and Education Network (NREN) in the United States and the Research Data Network (RDN) in Australia are preparing the ground for high-bandwidth telecommunications superhighways, and linkages across the oceans.
Weaknesses in worldwide email facilities are being quickly overcome. One of the greatest difficulties - finding the addresses of people you wish to communicate with, will be addressed during the next few years through what are commonly referred to as 'directories', both of the so-called white-pages variety (keyed by individuals' and institutions' names) and yellow-pages (organised by the class of service offered or research undertaken).
Beyond text-only mail is the ability to transmit formatted (so-called 'binary') documents. Provided that sender and recipient(s) have the same or compatible software, data can be sent in any format whatsoever (e.g. those of particular word processing, diagram-drawing, image-processing, desktop-publishing, or spreadsheet packages), and compound documents containing multiple media-forms can be supported. The products to produce and read these documents are increasingly including capabilities to attach annotations to any location within them, and thereby facilitate fast turnaround among co-authors, and among authors, referees and editors. Many discussions of these capabilities appear in the literature under the term 'computer-supported cooperative work (CSCW)', e.g. Grief (1988). Many documents will continue to be under the editorial control of one person, but conventional notions of document authorship and editorship are being complemented by alternative forms, some democratic and some intentionally anarchic.
Another emerging capability is workflow, which can be conceived as the creation of an intelligent virtual mailroom and file-register, such that documents route themselves between individuals' work-trays. Although this notion's primary applicability is in transaction-processing environments in business and government, it could be usefully applied to aspects of research work, including the receipt, editing, coding and analysis of data, and the management of conference papers and journal articles.
Researchers in group decision support systems (GDSS) distinguish a 2x2 (place and time) matrix of multi-person discourse, negotiation and decision processes. The participants may meet in the same place or remain dispersed, and the interactions may occur in the same time-frame (synchronously) or at different times (asynchronously). Each of the four combinations requires different forms of electronic support. Many of the emerging tools are of potential benefit in research as well as in business and government; for example much faster development of ideas can be facilitated by electronic conferences (which support different-time / different-place communications, and enable sharing of text, tables, graphics and voice data), and video-conferences (which provide same-time / different-place communications, with 'real-time' sharing of the various kinds of data, plus synchronised moving image and sound).
Another potential application is to enable doctoral supervisors who are remote from one another to 'meet' from time to time as an interactive panel, rather than restricting their operations to multiple one-to-one interactions with the candidate. With the trend from entirely government-funded to dual-funded research programmes, the importance of interaction between programme sponsors, directors and staff is greater than ever before, and GDSS tools have much to offer in this area also. A more exciting contribution is the facilitation of brain-storming by, for example, electronically-supported anonymous delphi techniques.
A related development is 'visualisation' (McCormick et al 1987, Cox 1990). Initially, this is enabling solo and co-located researchers to impose alternative models on data through iterative attempts to explain and replicate patterns. With the increasing availability of high-bandwidth networking, same-time / different-place experiment-ation will be possible in what amounts to a high-tech/high-science 'jam session'.
A further emergent notion with potential research applications is the 'digital persona'. This is a person's data shadow, or image, as projected out into the net; for example, each individual may use multiple identities to reflect different aspects of their 'real' personality, such as the enthusiastic researcher, the social advocate, the sceptic and the 'angry young man'. Ideas and reactions which would be inappropriate in one role may therefore be given free rein via one of the other personae. This is therefore a means whereby members of a community can gain access to all of the varied and often mutually inconsistent thoughts of outstanding individuals.
There are many other applications of the notion, however. In its more active sense, a digital persona encompasses the idea of an 'agent', which was conceived by computer scientists to perform functions on a person's behalf on a local workstation, a local server, or elsewhere in the net. One simple application is a mail-filter which vets a person's incoming mail, classifies and prioritises it, and perhaps discards low-interest messages, or places them in a 'read-only-if-something's-missing' pile (Loch & Terry 1992).
A more interesting class of active persona or agent is the so-called 'knowbot'. Because no central directory exists for the Internet as a whole, it is necessary to search multiple directories when trying to find a particular user or a particular data-file. Prototype programs are in use which adopt modestly efficient search patterns in order to find desired data from anywhere in the net. A further development is the notion of a news-gatherer or environmental scanner, which wanders the net, browsing through accessible files, and sending back references to or copies of items which appear to satisfy search criteria nominated by the 'real' person. A more sophisticated persona could 'learn' criteria through feedback gathered about the usefulness of earlier dispatches from the individual, and perhaps also from his work-group or a wider reference group.
The richness of emergent ideas is so great that individual researchers, and even whole research teams, are unlikely to apply more than a few of them. Selection and use of those most appropriate to the needs of particular individuals, teams and disciplines will be facilitated by 'researcher workbenches'. These incorporate a user interface supporting access to the various services; tutorial, on-line help and reference material; significant customisability for particular disciplines, research areas and personal preferences; and extensibility to additional and new applications. The most important feature which a workbench offers is, however, integration, both between functions and among local and remote team-members. Prototype workbenches have been reported in the legal area (Greenleaf et al 1992), and exist on the Internet (e.g. gopher). Although there are signs that mature products will soon begin to be offered commercially, the new ideas will continue to emerge from collaborative efforts among researchers themselves.
By way of recapitulation, the kinds of services which are accessible by way of the Internet can be classified into the following groups:
machine access to processing services, in particular the use of spare capacity on under-utilised machines.