No earlier issue with same topic
Issue
Previous article
Article
SIGCHI Bulletin
Vol.30 No.1, January 1998
Next article
Article
No later issue with same topic
Issue

Interactive Systems for Supporting the Emergence of Concepts and Ideas

A CHI 97 Workshop

Ernest Edmonds, Thomas Moran & Ellen Do

This workshop was concerned with the fact that current computer application systems are best at dealing with well-defined materials rather than in helping users create new concepts. For example, it is easier to use computers to draft precise drawings than to quickly sketch new configurations; it is easier to craft polished documents than to jot ideas and play with them. A body of research is beginning to accumulate that explores systems, such as pen-based sketching applications, to support the user in dealing with ill-defined concepts and materials. The key point of concern is that a person needs be able to easily create a visual representation, even for abstract and verbal ideas, and then respond to it perceptually and to interact with it in order to discover new arrangements and shapes suggesting new ideas. Thus, new concepts emerge from the concrete materials of the representation.

The research question of the workshop was how interactive systems can aid users in quickly creating and manipulating representations and whether they can support the discovery of new relationships, structures, and meanings in the materials. This is clearly an important new direction for the development of computer system design.

The most interesting thing about the workshop was the highly interdisciplinary nature of the participants, who came from fields ranging from dance (Beth Adelson), cognitive science (Kumiyo Nakakoji), communication (Fred Lakin), architecture (Ellen Do and Mark Gross), and music (Doug Rieken). A large fraction of the participants were also designers of one kind or another. This reinforced the workshop proposal that emergence is an important component of design and design thinking, a key theme in the CHI research community.

Workshop participants differed over the manner and extent to which computational tools should support emergence. Andruid Kerne, for example, described his `Information Collage Machine' that generates visual/spatial combinations of text and graphics, where emergence happens in the mind of the observer. Other systems, such as Tivoli (Tom Moran, Patrick Chiu) and VIKI (Frank Shipman), have built-in structure recognizers for generic arrangements of freeform drawing and text. In the Electronic Cocktail Napkin system (Mark Gross, Ellen Do) recognizers can be programmed and reprogrammed by the end-user.

A related theme was the idea of informal and semi-formal interfaces. The idea is that for a system to support the emergence of new concepts (in a conceptual or creative phase of work), it may be most appropriate not to structure the human-computer interaction. This flies in the face of many of the human-computer interfaces of the past and present, which have relied on highly structured interactions to ensure that the computer `understands' what the human is up to.

A heavy emphasis on freehand sketching emerged from the workshop discussions, both in the morning and afternoon sessions. Participants gave ample examples of `sketches' how they use sketches in their particular domains and argue that every activity has a sketch component of some kind. The recurrent theme was that formal and structured artifacts do not always work. Quick and easy sketches: (1) allow experimentation and exploration of design options, (2) reduce information to a more abstract level that emphasizes overall structure instead of details, (3) relax constraints, (4) permit uncertainty and ambiguity, and (5) allow visual reminding and seeing things from the ambiguity. In short, sketches afford emergence.

There was a consensus that, although participants came from different disciplines, all were engaging in design of one kind or another. Ernest Edmonds commented that design can be thinking with the hands, using hands to make representations such as physical model sketches. Other participants (Gyuri, Erez Pedro, Scott Chase, Chuck Burnette, Mark Gross) supplied examples about design in terms of drawing, looking, and inferring. We noted that the typical designers' technique of tracing over drawings focusses attention on previously draw objects and their refinement, finds new ideas from old things, and reorients thinking. An interesting point was that drawings contained the history of the thinking as well as the latest thoughts.

Another significant theme in the workshop was spatial reasoning. Spatial information (for both text and graphics) seemed to be a form to trigger emergence. Mark Tapia and Scott Chase described the idea about shape emergence as decomposition of entities of a shape, or extracting recognized precedence from the current shape. Frank Shipman, Tom Moran, and Patrick Chiu argued that grouping of objects encourage new reading and stimulate perceiving actions in a personal desktop organization or a meeting white board. Ernest Edmonds and Kumiyo Nakakoji went on describing the KJ method that encouraging participation of a group of people in a team by writing down ideas on cards, putting them on the table, and letting everyone take turns moving, rearranging, or adding comments to the cards. The process of arranging and grouping to allow emergence of ideas helps collaboration through interaction and consensus building.

What, then, is the difference between sketching and arranging? In the session chaired by Frank Shipman and Mark Gross, participants discussed sketching again: they argued that sketches were used for more than just arranging, but it also allow flexibility of doing things, reduction of complexity, etc. Kari Kutti concluded that, in the extreme case when sketchy bubble diagrams are merely representing abstract ideas, sketching out the bubbles with different configurations is similar to arranging objects. Tom Moran and Patrick Chiu, in their Tivoli system on the Liveboard, have combined sketching and arranging and linking, along with command gestures. Ellen Do and Mark Gross's Electronic Cocktail Napkin and James Landay's Silk system also allows gesture recognition.

Information was also deemed necessary in helping seeing emergence. Fred Lakin described his experience as a facilitator for group discussions: providing information to the viewer (of public political debates or other kind of events) helps new ideas emerge. Information could be provided as a case of precedence (Chuck Burnette), semantic analysis of the information (such as color for Kumiyo Nakakoji's eMMa), hypertext (Frank Shipman), recognized context (Mark Gross), reasoning (e.g. building codes, Gyuri Juhasz), or history and explanation of the design process (Tammy Sumner).

Emergence happens usually after a long period of evolution in a design process. It is sometimes a background process, a change of attitude or getting information about a new aspect or attitude for perception and interpolation or shifting context and paradigm. It sometimes involves revisiting old ideas and seeing the breakthrough or overtracing drawing to see new possibilities emerges.

Emergence is driven by perception -- which is to say that the perceiver applies structure to a set of experienced phenomena. Perception is driven by expectation -- you see what you expect to see -- but to build human-computer interfaces that support emergence, you must model these expectations and how they structure perception. Hence the development of emergence-supporting interfaces will likely call on the expertise of cognitive scientists, information designers, and of course domain experts. One interesting episode is that Fred Lakin, Mark Gross, and Scott Chase were standing around discussing the term `emergence' itself. Fred advanced the view that emergence is, among other things, a functioning of the perceptual system -- a figure emerges from the ground and that is the basis of seeing. Mark then said, "Yeah, emergence is in the eye of the beholder". They all looked at each other for a moment and then said in unison, "No! emergence is the eye of the beholder!".

It was fruitful to see the concept of emergence stretched and applied and examples found in a wide variety of domains. The workshop concluded by reviewing 30 pages of easel notes spread over four walls and identifying five summary propositions:

  1. The phenomenon of emergence is ubiquitous. It should be supported in everyday tools.
  2. Ideas emerge over time, often last over long periods of time, the idea is often not recognized as new until later in time.
  3. Rough sketching is an important representation for affording emergence.
  4. Systems should enable user to see things in different ways by providing multiple representations and suggesting alternatives.
  5. Representations are used both as particular languages for individuals and to mediate collaboration among group of designers.

More information is available at:
http://bashful.lboro.ac.uk/chi-wshop/

About the Authors

Ernest Edmonds is a full Professor and Executive Director of the LUTCHI Research Center at Loughborough University, UK.

Email: e.a.edmonds@lboro.ac.uk

Thomas Moran is a Principal Scientist and Manager of the Collaborative Systems Group at Xerox PARC, USA.

Email: moran@parc.xerox.com.

Ellen Yi-Luen Do is a research associate at the Sundance Lab for Computing in Design & Planning, University of Colorado, and a PhD candidate at Georgia Tech.

Email: ellendo@cc.gatech.edu
http://wallstreet.colorado.edu/Napkin

No earlier issue with same topic
Issue
Previous article
Article
SIGCHI Bulletin
Vol.30 No.1, January 1998
Next article
Article
No later issue with same topic
Issue