1. Further research possibilities

The information in this publication is maintained in computer files as summarized in Section TZ. The project to date has been, and should continue to be, a data-collection and presentation exercise. The existence of an updated data base of this kind should also facilitate some types of research which have hitherto been almost impossible.

The data collected will not, for example, contribute directly to research using quantitative models, although it may suggest some problems and relationships for inclusion in such models. A precondition for conventional model building is a minimum of quantitative information on the dynamics of the relationship between two or more selected levels or quantities in the system. The "problems" become evident by interpretation of the results of the quantitative analysis. In the absence of quantitative information, or where the latter is vulnerable to criticism, no other systematic analysis has been possible.

One use of the data collected in this project will be to test whether results can emerge from analyzing the networks of relationships as networks in which qualitative rather than quantitative values are attached to the links between the nodes. The readily available tools of graph theory and topology, for example, have not yet been applied with computer assistance to such data (see Section TZ). For example, it should prove useful to conduct computer comparisons of the degree of isomorphism of a network of interrelated problems and the corresponding network of agencies (or treaties, or disciplines) which purport to focus upon them. If the functional interlinkages, particularly communication channels, of the latter do not correspond to the linkages (or degree of structural complexity) of the problems, then it is probable that problem complex is uncontained, and uncontainable, by the programmes of the agencies as they are currently implemented. It would seem that in a world system characterised by a number of relatively complex networks on which information is largely unavailable or inadequate for numerical analysis, such techniques could be used to identify and analyze clusters and critical points for action.

2. New kinds of information

Since this volume is integrated with the Yearbook of International Organizations, any improvements will be closely related to improvements to that publication and could comprise inclusion of:

• other organizations (including subsidiary bodies);
• other kinds of relationship between organizations;
• other kinds of relationships between strategies and organizations, or problems, or values, or understandings of human development.

Ideally, given the emerging possibilities of CD-ROM and Internet (World Wide Web) access, "users" of this information will take on some functions of editors and information providers. Faced with entries on which they are able to articulate improvements to the presentation, or to the relationships with other entries, facilities will be offered to enable them to propose specific amendments. In this way, if they so choose, they effectively become an extension of the editorial research team.

At present the editorial team endeavours to extract such improvements from the documents of international organizations. This is a relatively slow process. In the future, use of the information and improvement to it will take place simultaneously. The maintenance of the information will therefore have a greater number of auto-correcting and self-sustaining features built into it.

Such an approach would have the considerable advantage of facilitating input of information on entries and relationships which may be obscure to many but vital to some. Such information is often difficult to locate and summarize and, with budgetary constraints, may otherwise easily be treated as of lower priority.

4. Quality of information

An obvious concern is the manner in which the quality of the information can be controlled in such a relatively open system. Clearly there is a need to separate changes advocated by highly authoritative specialists from those at the other extreme and from casual, or even irresponsible, users. Equally there is a need to ensure inputs from the seemingly unqualified who have had direct experience of the issues and whose insights may cast a different light on the matter from those of specialists blinkered by disciplinary mandates.

One function of the editorial team could then be to attribute "expertise codes" to distinguish between a variety of levels of input to which users can control their exposure. Hotel and restaurant guides, with their five star to one star qualifiers, suggest one approach.

In this environment, "entries" and "relationships" then effectively become dialogue arenas to which user/providers are contributing -- as in many Internet conference groups. Alternatively, users could be connected to separate discussion conferences on the themes of specific entries (or referred to sites where such conferences are ongoing).

Clearly the intent would be to create a decentralized learning environment, possibly one in which the boundary between the Encyclopedia and other information environments is creatively porous.

5. Capturing the dynamic

Inclusion, where appropriate, of claims made by an organization for the importance of its programmes with respect to particular world problems and the various obstacles to any increase in their effectiveness.

Inclusion, where appropriate, of counter-claims by competent critics (such as other international organizations with the same focus) concerning the ineffectiveness of a particular organization's programme activity. Such information could of course only be included in situations where there is some degree of unanimity on the content of the counter-claim, and where information could be included from the organization in defence of its position.

3. Display possibilities

This project will succeed to the degree that it can render transparent the complexity it attempts to map and on which continuing feedback and dialogue is sought. This poses the problem of developing a satisfactory form of display. Although the hardware exists and some software has been developed to handle network structures in three dimensions and colour, this work has hitherto been confined to engineering design, architecture and chemistry, and its potential for handling the great complexity of social structures is poorly recognised. The new virtual reality markup language (VRML) may prove vital to the challenge (see demo on this site).

An advantage of holding the relationships in computer files as components of directed graphs is that such graphs can be plotted (in colour) by computer with appropriate labelling and choice of projection. As a descriptive device for highly complex structures, apart from permitting relatively sophisticated analyses, a graph representation can be transformed into much more iconic forms than those conventionally used to describe psychosocial structures, and is thus more comprehensible. Detailed problem "maps" can therefore be produced, printed and bound into "atlases" - the argument being that people (whether students, executives, researchers or policy makers) have at least as much need for such visual devices to orient themselves in the social system as they have for road and other maps (see Section TZ).

Comprehension will be made easier by on-line computer graphics devices (discussed in Section TZ) with display screens to permit the user:

• to interact with that part of the network the user chooses progressively to explore,
• at the level of display complexity the user is prepared to tolerate, with the ability to call up textual explanation, use computational power, or activate a parallel display whenever required.