International Conference On Computer Communications

Washington, D.C., October 24-26, 1972

Luncheon Address By H.J. von Baeyer

Conflicts In Computer Communications

It was indeed a great honour for me to be invited to present the luncheon address at this important convention. When we were discussing a subject and a title for this talk, one of our program co-ordinators made a joking reference to the predominance of the letter "C" at this conference. The word "conflict" came up, and that immediately fascinated me as being a subject of great relevance and importance to all of us.

The diversity of many of the conflicting views and attitudes was impressed on me during our recent work at the Canadian Computer/Communications Task Force. So it seemed to me not only fascinating but also a highly stimulating and pertinent topic for this occasion.

There are conflicts: Conflicts in the theories of what computer/communications should be and do; Conflicts in the visions of its possible future; Conflicts in its problems, and Conflicts in the ideas on how to deal with those problems.

Computer/communications is today the subject of study and concern the world over. Some people see it as just another application of communications technology - for example, the recent Communications issue of "Scientific American" has little to say about computer/communications beyond a description of the developing terminal technology which provides new interfaces between communication networks and human users. Seen that way, the functions of computers and communications are performed by two independent industries which support each other but have really little in common. Public policy issues on questions such as the degree of competition and regulation in the industry can then be decided separately for each field.

Other people, however, consider computer/communications as a totally new entity, one opening up all the thorny issues which stem from the promotion of widespread access to sources of information. These issues include the problems of protection of privacy of the individual, and protection of copyrights; of autonomy and national sovereignty; of changed methods of education and money systems, and a host of other puzzling matters typical of the post-industrial society. This view suggests that public policies in the fields of computers and communications must be integrated because the two have merged into a new entity.

So from the outset we are faced with conflict in the very definition of the field and its functions.

We often speak of the marriage of computers and communications, and relate the offspring of that marriage to the unique characteristics of the union. The Canadian government publication "Instant World", for example, says:

"Some computer experts forecast that the marriage of computers and communications systems, if it can be successfully consummated, may generate, within the next few decades, social changes more profound than those of the past 200 years."

The French refer to it as a marriage of convenience, while Maurice Karnaugh, in the recent Computer/Communications issue of the IEEE transactions, claims that a multi-billion dollar industry is emerging as, in his words, "the unplanned offspring of two unwed and dissimilar parents".

So you see we have diverse views about the union of computers and communications. Questions are continually being raised about the effectiveness, even the value, of proper planning in that union. And questions are raised about the degree of control we actually have over its development. Can we assume, with some hope of success, that we will be able to shape the future of our lives by proper planning of computer/communications and by wise use of its potential? Or must we accept that we can do very little to control what is to come, that we are helplessly exposed to the vagaries of unknown future events? These opposing views express a far more fundamental conflict in commonly held beliefs about future events, a conflict which existed long before computer/communications came on the scene. Herman Kahn's book "Things to Come" makes the point, as follows:

"in our attempts to elucidate even relatively simple and straight-forward policy issues, we have been struck by the fundamental importance of the persistence of an ancient dispute between the Augustinian and Pelagian views of man."

In the 5th century, St. Augustine publicly accused the British theologian Pelagius of the heresy of believing that man was basically the master of his fate and could be good through his unaided efforts. St. Augustine believed the opposite -- that man was fundamentally sinful and could achieve salvation only through God's Grace. If you are Pelagian you believe that man can improve both himself and his surroundings. If you are Augustinian you cannot accept the view that any secular program is going to change people's fundamental weaknesses.

As Kahn points out, the Augustinian rejects the ideas that things are going to get much better in this world, or that they can be improved by the efforts of governments, professors, or policy analysts. The Pelagian believes just the contrary -- that the future will be affected by such efforts, and can be affected to the betterment of society and the individual.

So the next time you begin to discuss with someone the future promises of computer/communications, make sure that you first find out whether he belongs in the Pelagian or the Augustinian camp -- unless you are the same, do not hope to come to an agreement within the next few hundred years.

Personally, I would call myself a modified Pelagian. I say "modified" because I am skeptical about relying too much on the accuracy of the expected end results of long term planning. In my own life time there have been so many examples of totally unexpected technological developments, and so many unpredictable effects created by such developments; in fact my own batting average as a prophet has been so low that I now look on planning essentially as the establishment of a working hypothesis, based on severely limited information, and always ready to be adjusted to realities. Subject to these constraints, however, I do believe we must plan and we must try to keep order in situations which would otherwise be fragmented and possibly chaotic. From this it follows that I hold with the Pelagians that our thinking today will help to shape the future for the better. I trust I am not the only Pelagian here.

Coming back to computer/communications, and assuming that one tries to control future events through planning, we are immediately faced with a dilemma: computer/communications has an inherent, dichotomous capability, either of leading to a more enlightened future, better self-fulfillment, new heights of awareness and common understanding, or of being used for enslavement, loss of privacy and depersonalization, an Orwellian future. Do we believe that through planning we can steer the utilization of this technology towards its potential benefits, avoiding the pitfalls and dangers? Or are we afraid that in spite of well intentioned planning, a computer controlled tyranny may evolve? Or do we dare to have faith that we can follow a middle path, being able to react positively, if and when necessary? These are conflicts in fundamental attitudes, and it is important that we are aware of them. Now, let us turn to more specific conflicts:

As I mentioned before, we had in Canada a Task Force on computer/communications. For a year and a half we studied Canadian conditions with the object of developing recommendations for government policies in this field. Because Canada's population is spread over an immense area with widely separated centres of activity, many of our correspondents were deeply concerned with the costs of data communications. They pointed out that potential benefits from remote access to computers and of resource sharing would not be obtainable unless communication costs dropped substantially. However, opposed to this, we had strong representations from the more remote areas of the country and from those regions with yet little industrial development. They argued that low communication costs would prevent the build-up of local computer facilities and would lead to centralization of computation and data storage at a few highly industrialized centres in the country. Here we have an obvious conflict -- on one hand the potential value of decentralized network operations using many dispersed, localized facilities for common benefit; on the other hand the economies of centralization, though accompanied by a loss of local capabilities, which in turn would result in reduced employment opportunities and reduced adaptability to local problems and conditions. It is obvious that the extent of this conflict is determined by the degree of autonomy of the various network participants. Low communications costs, even if they were zero, would certainly not have induced the American universities to establish one centralized computation centre instead of participating in the ARPA network. However, the idea that low communication costs may favour centralization of computers could well be of interest to industrial enterprises with many geographically distributed subsidiaries or branch plants. This was demonstrated to us when several local EDP managers expressed the fear that reduced communication costs would lead to centralizing of computation and data storage at company headquarters, as opposed to operating in a more self reliant, decentralized mode.

This conflict between centralization of computer operations, and increased emphasis on local autonomy was shown to be due to economic factors -- in particular the relative costs of computation and communication. Similarly, a conflict exists between shared use of computer facilities in a computer network and increased emphasis on local autonomy. We can trace its roots to some of the basic characteristics of the communications process. It is by now well known that improved access to, and improved spreading of information does not create a greater unlformity of thinking across countries and around the world. In a paper on "Policy Problems of a Data Rich Civilizationl", Harold D. Lasswell said back in 1965:

"The modern communications revolution has been unable to universalize the outlook of mankind. The new instruments have been utilized by the managers of the information media to overcome localism. However, the chief gainer from reduced localism has been, not a common world perspective, but intermediate attitudes of a more parochial character."

Without doubt there will be some common thinking induced by the use of common information banks. At the same time, one must also expect a greater polarization of opinions and attitudes, based on better knowledge of other people's views, and leading to a crystallization of one's own views. The rapid increase in the number of independent member states of the United Nations has been related to such a dispersive effect of better communications. The rapid increase of more and more clearly defined communities of interest, often with diametrically opposed aims, may also be traced to the ready availability of information and control of its flow. In this respect, we see both the withholding and the dissemination of information as instruments of power. As a consequence, in the specific field of computer/communications, there are two conflicting concepts. On the one hand, networks crossing institutional, political and geographical boundaries; on the other hand, increased parochialism and insistence on autonomous operation. There is a strong desire to share skills and information with others. There is an equally strong desire to do one's own thing, to solve one's problems by one's own solutions, and to be in control of one's own operations. These are opposing forces which can be found in many computer operations, small and large, private and public, national and international.

Conflicts also occur in the design itself of computer networks. Features that are desirable from a communications point of view are not necessarily compatible with what is desirable from the computer operator's point of view. Purely technical characteristics present a bewildering range of choices in methods and procedures, which are debated between professionals the world over, nationally and internationally. Synchronous versus asynchronous transmission, switching with or without storage, the choice between multiplexing, polling or loop operation are a few examples. On a more general level, a common source of conflict is caused by the fact that the network designer would prefer to have the types and programs of all computers on the network as uniform as possible -- whereas the individual computer operator would reject such constraints and select those characteristics and conditions which are most suitable for his particular requirements. Since the very concept of a network is that of bringing together diverse skills and characteristics, it would be disastrous if the network designer yielded too much to the technical pressures of unification and standardization. Ideally, one may think of combining the two concepts of individuality and communality. This would postulate systems where the individual participants have complete freedom to develop their own approaches and information bases, but where, at the same time, special information and special skills are made mutually available without imposing undue constraints through excessive standardization or operational control.

Technical solutions for approaching such an ideal may become available. But for a long time they will probably be confined to communities of interest which have a genuine desire to cooperate. Outside such communities one is again faced by the conflict between common interest and parochialism, which hinders the development of generalized large scale integration. But at the same time we see unquestioning acceptance of the benefits provided by widespread networks in highly specialized areas covering undisputed, quantifiable information of a scientific or technical nature, and in highly specialized communities of interest. Examples are the efforts of world organizations in such fields as health, agriculture, meteorology and others, and the systems of airlines or the press. Common interests in these areas outweigh parochial tendencies. These systems represent binding forces of increasing strength in an otherwise highly fragmented and decentralized world. Once the common bases are established in areas of quantitative and generally undisputed information, it may be hoped that the developing level of common understanding will allow us to search for mutually beneficial solutions to problems where hitherto outright confrontation was the only possible approach. This calls for a certain collective rationality; a pragmatic ethic as expressed in the philosopher Immanuel Kant's categorical imperative: "Act in such a way that, if others acted in the same way, you would benefit thereby." The chances that our rationality will subject itself to such an ethical principle appear to improve, as we learn more about building information systems which combine the freedom for doing one's own thing with the possibility of sharing with others.

So we have seen that wherever we look there are conflicts in our chosen field of computer/communications right up from its very definition. They exist at the technological level, between the desire for standardization, and the desire for technical diversity. They exist at the operational level between the benefits of network operation and the attractions of local autonomy. They exist between the issues of centralization and of dispersion. And, most important of all, they exist in the beliefs about the potential long term effects of computer/communications on society. We arrive at a point where we can ask in the form of a riddle: What can be very good and also very bad? What tends to spread out and also to concentrate? What is diverse and also unified? What can bring together and also divide?

And the best candidate for an answer would be "computer/communications".

Very much depends on our attitudes and dreams what we do with computer/ communications as an instrument of mutual relationships. Too often underlying motivations are hidden or left undefined -- as so beautifully expressed in that little story used by Professor Weizenbaum in a paper on the impact of the computer on society: A lady dreams of being raped and begs her attacker to be kind to her. He answers "It's your dream, lady." Weizenbaum then goes on to say that we must come to see that technology is our dream and that we must ultimately decide how it is to end. May this convention contribute to that ultimate decision and give support to the modified Pelagian view that our work of today can serve to prepare us for a better future.