Category: complexity

Working in Complex Spaces from my Favorite Curmudgeon

Earlier this week Chris Corrigan pointed to a great blog post from one of my favorite curmudgeons, Dave Snowden (Dave, yes, I think a cacophony of curmudgeon’s is perfect!) on the heuristics of complexity.  Dave wrote in (Of tittering, twittering & twitterpating) the following:

We need to draw a fine line between legitimate experimentation and slipping into magpiedom and the legitimacy probably depends on the degree to which there is a coherent narrative around the core activity.

Aside from that I made a serious of points that apply more generally, as well as to the foresight community who were my primary targets. They included:

  • The whole success of social computing is because it conforms to the three heuristics of complex systems: finely grained objects, distributed cognition & disintermediation
  • I an uncertain world we need fast, real time feedbacks not linear processes and criticism includes short cycle experimental processes which remain linear.
  • The real dangers are retrospective coherence and premature convergence
  • Narrative is vital, but story-telling is at best ambiguous
  • Need to shift from thinking about drivers to modulators
  • You can’t eliminate cognitive bias, you work with it
  • Extrinsic rewards destroy intrinsic motivation
  • Messy coherence is the essence of managing complexity

I suggest you read the whole post for context (and humor).  I could ruminate on any of these, but I have my peeps coming into town starting today for the Seattle KM4Dev Gathering (Dave, we’ll still try and tempt you!) so I’ll just say this one is my focus for the week: “Need to shift from thinking about drivers to modulators.”

As we explore next week the practice of “managing knowledge” for international development, this could be a cracking good opener… Now I need a visual. Any ideas?

Monday Video – Cognitive Bias VideoSong

YouTube – Cognitive Bias VideoSong. Brilliant! Thanks to Irene Guijt for pointing it out!

This video by Mr. Wray, hits the winner bell on two fronts. First, it is a great overview on cognitive bias and second, it is in the form of a song. This brings me back to university days where the only way I could hold all the organic chemistry details I needed for an exam was to put them to a song. I’d sit in the back of the exam room (as a “w” that was easy) and quietly hum to myself. Crazy, but it worked for me. Enjoy and thank you, Mr. Wray! Your Advanced Placement high school students are lucky to have you. (And the rest of your vids are also pretty cool!)

A Gem from KM4Dev on Impact and Outcomes

There has been a great discussion on the KM4Dev mailing list the last 10 days or so about evaluation, impact and measurement. In the context of international development, this is critical. Why do something if it doesn’t make a difference. However, often we don’t do a very good job figuring out what does make a difference, let alone know why (causality.) Dave Snowden posted something that just rang the bell for me. I hastily copied it down to share here. The link to the web archive of the email discussion is at the bottom, if  you want to mine for the rest of the thread. Emphasis is mine.

The linear concept of input, leading to outputs, leading to outcomes which in turn leads to impact is I think at the heart of the problem, It implies (and I can see why people would want this) a causal chain that can be replicated.

However if the system is complex (in the sense of complex adaptive) then any input is a stimulus or modulator which influences but does not determine impact. That means we need to start measuring the sensitivity of a system to different stimuli, and the way in which some stimuli produce a disproportionate effect in that they catalyze other inputs. This is newly developing area which has not hit the development sector yet, but we are working on it in related fields, loosely termed modulator mapping. It also leads us to evolutionary representations (such as fitness landscapes) and measure based on stability of landscapes. In all those cases mathematics are simplified by representation and linked micro-narratives. There is no point in measuring anything if the results do not convince both donors and recipients alike to take action

All of that moves the “impact” agenda on. I didn’t confuse outputs and outcomes, I conflated them as the model means there is no real difference in what is measured in practice.

via Discussions.

I am now going to start paying attention to this idea of “measuring sensitivity of a system to different stimuli.” This relates closely to two projects I’m working on where I have been sensing this, but hadn’t had the words for it. Now I have a toehold. Onward!

Photo credit: