Concept Tracking in Conferences
- / -
One way to explore future possibilities for electronic and non-electronic conferencing is to consider the implications of various possible marriages between modes of information:
1. Text and Data: The classic separation between text processing and data processing has severely impeded the evolution of conferencing. A fruitful marriage would allow users their current freedom of expression but would also enable them to navigate more effectively through the maze of messages. Various approaches could be taken. An outline facility would structure lengthy communications so that users could explore them to different depths using an onion-skin approach. Of particular interest would be to code such levels to indicate their relevance to the core message (e.g. background or context, argument or justification, precedents, counter-arguments, action implications, explanatory or learning mode material, anecdotal illustrations, etc). Archiving could then be done selectively, gradually reducing to the core concept only. On the other hand, a hypertext facility would obviously empower users in new and interesting ways.
The issue in both cases is how to code levels of the text and embed hypertext links in the text as part of the message generation process. This is an extension of the classic problem of how to motivate authors to provide abstracts. The long-term solution is to shift the focus of attention from the text to the representation of the knowledge implied by the text. A transitional solution is to develop what might be called a "text compressor" or "concept processor" based on artificial intelligence procedures.
As has been repeatedly noted, the desk-top publishing revolution and its conferencing parallel will more than overwhelm a saturated readership. Desk-top readers do not accomplish what we would like their name to imply. They do not help us to filter and comprehend the content. Some form of text analysis and restructuring by a concept processor is required to mine the conceptual ore from what needs to be dumped or filed at a lower priority level. The most practical approach would to provide users with a minimum facility which they could adapt and tune to their personal idiosyncrasies. Users could of course view and edit the structured product generated from their own outgoing communications. Such a processor might usefully be related to the need for machine-assisted translation.
2. Data and Graphics: There are two challenges here:
(a) enabling a group of users to address the emerging articulation on a shared
map (possibly with personal overlays, etc);
(b) escaping the conceptual straitjacket of packages based on a directed
graph or tree structure in order to use an associative structure (on which
alternative tree structures can be temporarily imposed).
It is worth noting that a heroic attempt was made to do just this by Stafford Beer and Gordon Pask at the first international conference of the Society for General Systems Research (London, 1979) before the PC era. Both concept maps and participant network maps were produced and used to orient discussion. Such experiments would be infinitely easier now and many refinements could be incorporated. Stafford Beer is currently extending this work with new protocols for specialized conferences based on tensegrity structures. Mind mapping software is now available on PC -- but anchored to a single initial concept.
3. Comment: The absence of such tools is an indication of the priorities of conferencing at this time. Questions such as the following need to be asked:
(a) Why is it that participants in a conference have experienced no need
to represent the conceptual structure which they are collectively attempting
(b) Is it that participants are satisfied with the schematic representation
in an agenda or programme? Or is it that they prefer a discursive mode in
which the structure is implied or left ambiguous?
(c) Why is it that in the academic analysis of social networks almost no
attention has been devoted to the graphical problems of representing complex
networks -- despite the extensive manipulation of data on them.
(d) Why is it that in the current enthusiasm with hypertext, no effort is
made to provide the user with a map of the hypertext pathways between the
set of frames? It is almost as though a hypertext stack was designed like
a rat maze, which the user has to explore like the rat, without any sense
of perspective. Learning is the process whereby the rat builds up its own
mental model. The map of the relations in a relational database is not considered
as valuable information to orient new forms of inquiry or modification of
the pattern of relations. It can be argued that it is that map which constitutes
knowledge, in contrast to information.
There is every possibility that users have different preferred cognitive modes (possibly under different circumstances) and that it remains important to cater flexibly for those who feel constrained by particular structures.
One possible reason for the relative lack of interest in conferencing systems in the international community is that in the present form they do not reflect the dynamics of factional interaction. The action is perceived as being elsewhere. Even the texts produced can be viewed as conceptual shells discarded by a dynamic beast that has moved elsewhere. The consensus-mania pervading explicit conferences forces the real, tension-filled, business of factional wheeling and dealing into other arenas -- if only the corridors and bars outside meeting rooms or in one-to-one messaging. This clearly suggests the need for handling the public-private interface more flexibly, veiling and unveiling explicit structure when appropriate. The conferencing of the future may yet prove to be a conceptual dance of the seven veils !