GrowingPains

It Takes at Least Two Tools to Tango

Quite a while ago, I mentioned that communities need less tools, more process. Community interaction processes tend to be very complex, leading to context-specific requirements which often cannot be met completely by any particular tool. Vice versa, many tools are often used for a very different purpose than what they were originally designed for.

I recently experienced a wonderful example of how relatively straightforward tools can be combined into "distributed applications" to provide very useful functionality. I am one of the three co-chairs of the First International Pragmatic Web Conference (see the Patterns for the Pragmatic Web-entry on this blog for an explanation of what the Pragmatic Web is about). Since we live and work far apart, we tried to write the Call for Papers by sending Word-files around. However, as it is such a new paradigm, confusion abounded and versions kept flying back and forth between the three of us. It was clear that just using e-mail + Word processors did not fit our collaborative needs.

We decided that, additionally, we needed voice contact. However, the phone would not do, since the three of us needed to discuss simultaneously. Phone conferences are still relatively expensive, so we decided to use Skype instead. This voice-over-Internet application allows up to five people to discuss for free. Additional benefits are that the sound quality is much better than that of the phone, and that one can talk using a microphone. This means that one's hands are free and one can type while talking.

Okay, so now we could talk, but we were still not there yet. We were considering three versions of the document: one previously written by one chair, one by another chair, and the third one the version-in-progress which contained the modifications that we were making as we spoke. This was too much, the poor mind's eye was getting very confused by all those almost-but-not-quite similar texts. Instead, we came up with the following solution:

 

• All of us opened one of the  previous versions in Word for reference.
• One of us was the editor, and opened the version of the second chair as the starting point of the version-in-progress.
• We now started editing this version paragraph-by-paragraph, line-by-line. Each time, the editor would copy the paragraph currently being discussed, and paste it into Skype's chat-window.
• All three of us would look at this paragraph, and compare it with the same paragraph in the Word-version.  We then proposed and discussed modifications a few lines at a time.  The editor kept track of these changes in the version being edited. Having agreed orally upon only a few lines, it was already getting difficult to see the bigger picture. Whenever that was the case, the editor copy/pasted the now modified paragraph into Skype-chat again. We would then orally make a few more changes, to be copy/pasted into chat by the editor again.
• This continued until we were satisfied with the full paragraph. We then moved on to the next paragraph, until we were, at last, happy with the final Call for Papers.

So, what exactly happened? Using a few standard functionalities (Word for editing text; Skype-Conference for group voice discussion;  Skype-Chat for keeping track of the rapidly changing "focus-text"), as well as a few simple rules (all authors focus on same paragraph; all authors compare the current focus-paragraph with the related Word-paragraph and make comments; the editor changes lines on which consensus has been reached in the version-in-progress; the editor pastes modified paragraphs into chat whenever there is too much confusion) a very complex group authoring process had been successfully supported by ICT. This is the first time I have been able to deal with such a daunting collaborative task online instead of face-to-face. It really was an amazing experience. So, who needs all these very expensive, proprietary collaborative platforms? Perhaps the future of online work is not so much 'less tools, more process', but really 'lean tools, more process'?

NB A much more sophisticated analysis of the co-evolution of tools and process is described in (De Moor and Aakhus, 2006).

References

- A. de Moor and M. Aakhus (2006). Argumentation Support: From Technologies to Tools. In Communications of the ACM, 49(3):93-98.

Community Modelling with System Dynamics

This blog is about the evolution of virtual communities. One major problem with studying (virtual) communities is that by nature their examination requires a longitudinal approach. A community cannot be "declared", but needs to grow slowly over time. This means that collecting empirical data to detect meaningful patterns  is very difficult to do, requiring a time horizon which is far too long for average research projects and decision making processes. As a result, much research is anecdotal, or focuses on single aspects instead of the full system.

Many people cannot visualize the exponential, non-linear effects of interventions in complex systems like virtual communities. Much real-world behaviour is very complex because of dependencies between variables and feedback loops in which the output of a system component has an effect as an input in the future. Delays and accumulations add further complexity. As a result, people make incorrect decisions with a short-term focus. The long time horizon associated with communities and the lack of experimentation results in wrong decisions about which interventions are most likely to benefit a community. Should, for instance, budget be spent on installing an advanced community memory system or in appointing a number of human facilitators?

System dynamics is a methodology particularly suited to analyze such complex, large-scale, non-linear, partially quantitative dynamic systems (Sterman, 2000). System dynamics models can be used to generate and analyze simulations of very complex, realistic behaviour. However, they consist of combinations of only a few very simple conceptual building blocks, notably stocks, flows, feedback loops, and delays.

I am fascinated by system dynamics as a much-needed research instrument for community research. There are many ways in which it can be applied. One way in which it might be made relevant to our field I described in (De Moor, 2005). I will definitely get back to you on this in future posts, since I believe system dynamics will help us to better think through, validate and link theory and practice.

References

- A. de Moor (2005). Towards a Design Theory for Community Information Systems. In Proc. of the 11th International Conference on Human-Computer Interaction (HCII 2005), Las Vegas, July 2005, Lawrence Erlbaum Ass., Mahwah, NY.
-J.D. Sterman (2000), Business Dynamics: Systems Thinking and Modeling for a Complex World, Irwin McGraw-Hill, London.