The complexity of complexity
A recent discussion on Twitter on complexity triggered me to write this post. Clearly, it is a subject that is being interpreted in many (3?) ways. Complex, chaos, simple, complicated, anarchy, all terms that are being compared in order to try to understand what they (should) mean. Some argue that you can use axes and create a spectrum, where all these phenomena can be plotted upon. Others disagree with the language used, or that these levels exist for complexity. And then there are other misunderstandings or misinterpretations. For example, complexity and Complex Adaptive Systems (CAS) are not exactly the same. We’re talking about the complexity of complexity.
Good for us humans, our thinking and behavior is quite complex as well. We are able to understand complex matter, albeit when looking back. We are used to think in linear ways, especially when we try to predict things to happen. In retrospective, we are capable of understanding things (events, behavior, etc.) that can be called complex. The most important attribute of complexity is non-linearity. Quite interesting finding, when looking back to understand phenomena it seems linear, looking ahead to the future, expect non-linear behavior. Is that complexity? No, it’s just uncertainty. Quite different things. And when looking back, uncertainty is gone, one outcome emerged in favor of many, at the time possible, outcomes.
Now I’ve almost lost myself in the above paragraph. Of course, complexity is related to uncertainty. However, the range certainty-uncertainty does not classify complexity, nor does predictability. In my view, complexity can not be classified, influenced or whatever. Complexity is an attribute of the behavior of a whole, where many actors are somehow involved and influence each other.
To me, complexity is not about systems. It’s about social phenomena. We can talk about the ‘problems’ of complexity and complex behavior, rather I’d talk about the opportunities. Dave Snowden understands this very well. Like I’ve said before regarding emergence, I’d like to say the same about complexity. It’s time to accept and embrace complexity, and to develop methods to get the most out of complex social phenomena or behavior. To be able to develop these methods it is important to understand complexity, however, I think we should not try to understand complexity fully. Our understanding will become better sooner or later, but we have to deal with it now. That’s inevitable.
Bas Reus' quest on self-organization and online collaborative spaces 
Bas,
you hit the spot as always. I agree that uncertainty is only one element of complexity: Another one is agreement/disagreement.
I have recently identified 5 major characteristics of social systems that contribute to their complexity:
1. All people have individual concerns, purposes, and circumstances.
2. All people in an organization make many unsupervised decisions, every day.
3. All people in an organization are connected in different ways to other people within and without the organization.
4. There is a close to infinite number of external and internal influence factors that shape the destiny of the organization.
5. Social systems have a strong urge to protect their integrity.
Based on that we have developed a meta approach to dealing with complex change: The Change Journey (http://www.changejourney.org).
By the way, Peggy Holman will publish a seminal book in fall: “Embracing Emergence” (http://patternsofchange.wordpress.com/). I have read it already. It’s very good.
“All models are wrong, but some are useful.” – George Box, 1969
This also applies to models of complex systems vs. other kinds of systems. They are all wrong, but some are useful.
Quite coincidentally I was just writing about this topic (the incompressibility of complex systems) for the conclusion of my own book, due to be out by the end of the year. 🙂
BTW, I disagree that complexity is about social phenomena. There is nothing “social” about cellular automata, learning classifier systems, dissipative structures, etc. All of them are complex systems.
One more great post, Bas.
I couldn’t more agree with you, our world is immersed into complexity, and emergence is like corks popping at the surface, not helping us understand, but giving us clues to follow.
We cannot connect the dots (or the corks) as complexity increases. This isn’t that new, as David Warsh, for examples, explains here: http://www.economicprincipals.com/issues/2008.05.25/319.html
I think that we cannot understand a system if fully immersed, we need to have a foot outside. Theories and postulates were created along history to try to ‘domesticate’ a rather mechanistic world, with mixed results, as we cannot step aside our world to observe it from the necessary holistic point of view.
Complexity is no exception of this principle, and may give us a few lessons to learn:
– we cannot understand the world as we cannot step aside. Let us embrace what’s happening
– theory building is a ghost of a past, mechanistic world. We now need a new cognitive toolset to approach our social environment
– by nature, complexity is nested. Trying to comprehend (a better suited word than ‘understand’) smaller, more discrete, phenomena, might help us to figure out the global complexity a LOT more than we think.
Cheers,
Thierry
Nice to see you distinquish between complex systems and complexity. I tend to agree with Ralph Stacey that, “To think in terms of system is to think in terms of formative causality which cannot encompass novelty or creativity.”
http://complexityandmanagement.wordpress.com/2010/03/30/organisations-and-the-concept-of-systems/
[…] a comment Go to comments Friend Scott Rogers (@jayhawkscot) sent me a link this morning, “The Complexity of Complexity“. A legacy from bi-gone days, when I studied Biophysics, though I still like to keep my eye […]
[…] }); }Friend Scott Rogers (@jayhawkscot) sent me a link this morning, "The Complexity of Complexity". A legacy from bi-gone days, when I studied Biophysics, though I still like to keep my eye on […]
Bas,
Thought provoking as ever.
Like you, I’ve begun to avoid the semantic debates – which we’ve stumbled into more than a few times, often from the starting point of your excellent blog. I always advocate clear definitions. But so often the words used to define our topics sit outside the mainstream. Hats off to the academic view. But as a practitioner, I need definitions that regular people (nurses, teachers, financial analysts, OD professionals) can understand.
I find that unpacking the human side of complexity is profoundly interesting. I agree, social ecosystems are not the only type of complex adaptive system. But if you’re wanting to attack complex social problems in new and creative ways (say “education” or “health care”), I think complexity science affords some powerful clues and new ways of thinking.
I very much like the idea of understanding patterns, trends and ‘simple rules’ (think “Boids”), moving beyond discrete cause-and-effect of closed loops systems (which have a place) to “dancing landscapes” and the like.
As Thierry says above, a new cognitive tool set is needed.
In complex social ecosystems, where I tend to focus, I’m finding that understanding the actors, the range of their behaviors (consistency vs. variability), and their prime motivators (priorities, survival schemes, cultural drivers, etc.) can be hugely insightful.
If anyone is interested in this line of thinking, let me know. We have a public domain team digging deeper on it.
Meantime, if you guys get “complex adaptive systems” (CAS) defined in layman terms, I will happily help you tweet and blog it into the mainstream.
Holger – you actually do this some justice with your bullets above.
To me, learning how to think and describe complex problems in new ways – an evolving complexity paradigm if you will – is foundational to moving this discussion forward.
Let’s talk more about it.
Bas – thanks as always, my friend. Keep shaking things up. New ideas emerge every time you do.
Chris @SourcePOV
Charlotte, NC
[…] too complex? 1 Friend Scott Rogers (@jayhawkscot) sent me a link this morning, “The Complexity of Complexity“. A legacy from bi-gone days, when I studied Biophysics, though I still like to keep my eye on […]
It does seem as if finding the words or symbols to describe complex adaptive social systems is getting to the top of the various convos , twitter chats and blog posts I’ve seen recently. I take them as data points that say if might be a problem ready to be solved.
Interesting post, but I have a quibble. You say. “looking back, we are capable of understanding things (events, behavior, etc.) that can be called complex” I guess it hinges on the precise meaning of “understand.”
i think it’s fair to say while we can construct a plausible narrative of the chain of the events that lead to the event in question, that might really me “understand.’ The many times human historical events are reframed is a good example of what I’m trying to point at.
Just sayin that if we can’t “understand” the past, there is much less hope that we will be able to “predict” the future.
Not to be the party pooper but “And when looking back, uncertainty is gone” This is hardly true. Why? We have a lot of ‘facts’ but we don’t always know what conclusions to draw from them. I’d contend that our inability to ‘forecast’ is directly tied to our inability to adequately evaluate what we’ve already seen. They’re all related. Michael alluded to this as well.
The crux of the problem lies with pattern matching, but the patterns we’re trying to match to are fractals. I’m not sure we’ve figured out how to not only do mental fractals, but I’m quite certain that we’ve not figured out how to account for the variabilities between them.
Now that’s at a deeper level. At a shallow level the strongest tool we have is intuition, but this is a tool that is shunned by our business culture, and yet it works perfectly for parents to protect their families. We are such fools.
@Rockapchen
“The crux of the problem lies with pattern matching, but the patterns we’re trying to match to are fractals” Nicely said.
I’ve been working with a couple of folks on twitter @spirosplidias, @openworld and @nedumar to try to define a language to capture precisely that reality.
In that context, folks might find this post interesting. http://ilnk.me/3015
Sure, I’ll try. By ‘not possible’ I mean ‘not exactly possible’. It makes sense to evaluate past events and learn from that. Very reasonable approach, a missed opportunity if you do not. Evaluating past events makes you more experienced I believe, and that will probably contribute to a better intuitive response in the future.
My point in my comment is that studying past events does not make a future event controllable.
Rotkapchen (and Michael), of course, you’re absolutely right about this. And as you tweeted, @johnt’s post about birthing and midwives makes your argument even stronger. Looking back does not make us understand, like textbooks (and too many business books as well) try to communicate.
How intiution works can not be learned or analyzed in retrospective, it only works during the moment you need it. And it gets better with every related experience, like the example of the firefighther in John’s post. The same example is used in Gladwell’s book ‘Blink’.
All (John’s included) our stories amplify that predecting how future events will unfold based on facts of the past is not possible. Foolish even to use your words.
Thanks all for sharing your thoughts!
Bas,
If you could clarify
“how future events will unfold based on facts of the past is not possible. and “Looking back does not make us understand, like textbooks (and too many business books as well) try to communicate.”
I just want to make sure I understand the words as written. They seem to say that studying history in the service of making better decisions in the future is not a reasonable approach. Do I understand your meaning?
To be clear, I agree that issues of timing are in principle not predictable with much accuracy. But if I take your words as written, it seems obviously wrong.
Do I misunderstand your intent?
Thanks for the clarification. I think we agree.
One way that I like to look at is having a prepared mind or organization to react quickly to new events. A soccer team might be good analogy. Certainly a goal is not “controllable.’ But there are well defined processes to increase the likelihood that one can take advantage or defend against events of “luck.”
Michael, love this : “increase the likelihood that one can take advantage or defend against events of “luck.””
Thx for putting this together Bas…yes let’s move on…create conditions, look for patterns and monitor complexity…
Snowden says:
http://www.cognitive-edge.com/blogs/dave/2006/10/communities_of_practice.php
“…you can not replicate the end point of an evolutionary process, but you can stimulate similar starting conditions. That stimulation can be as simple as making the tools available, or providing some initial stimulation or sponsorship.
You say:
“when looking back to understand phenomena it seems linear, looking ahead to the future, expect non-linear behavior…when looking back, uncertainty is gone, one outcome emerged in favor of many, at the time possible, outcomes.”
And the human bias is to not really think about the complexity due to “retrospective coherence”….looking back is supposedly easy to explain why things happened…but not possible to do the same when predicting the road ahead.
See snowden:
http://www.cognitive-edge.com/blogs/dave/2006/09/reality_avoidance.php
More here:
http://matt.blogs.it/stories/2004/03/15/daveSnowdenCynefinDynamics.html
“Unorder: It is the domain of “retrospective coherence”. When you look backwards everything makes perfect sense but there was no way, at the time, of predicting that particular outcome. What happens is that patterns of interaction stabilize and you can then understand how this occurred. But it could not be predicted in advance (Anti-terrorism: Emergence outcomes. Attempts to predict them can make you more not less vulnerable to threat.) Causality is apparent after the event.
Pattern of cause which become apparent, as particular assemblies of events repeat themselves, are not inherent properties of the system but coincidences which cannot be relied upon to happen again. Structured methods from business schools and management systems do not cope. Instead an approach based upon complexity theory (boundaries & attractors) is required.”
As Snowden says:
http://www.cognitive-edge.com/blogs/dave/2008/10/rendering_knowledge.php
“The way we know things is not the way we report we know things….Asked to describe how they made a decision after the event they will tend to provide a more structured process oriented approach which does not match reality. This has major consequences for knowledge management practice.”
Nassim Taleb calls it the Narrative Fallacy:
http://johntropea.tumblr.com/post/51471859/the-black-swan-nassim-taleb-p119-120
“When we look backwards we will always find a cause to make us feel better, we need closure, whether its true or not. “…we are explanation-seeking animals who tend to think that everything has an identifiable cause and grab the most apparent one as the explanation”
http://johntropea.tumblr.com/post/50373596/the-black-swan-nassim-taleb-p-63-69
“…narrative fallacy addresses our limited ability to look at sequences of facts without weaving an explanation into them, or…forcing a logical link…explanations bind facts together”
And how we can be more coherent:
http://www.cognitive-edge.com/blogs/dave/2008/05/coherence_and_uncertainty.php
http://www.cognitive-edge.com/blogs/dave/2008/05/coherence_revisited_and_it_won.php
Patterns not causes:
http://robpatrob.com/why-we-get-stuck-with-the-new-cynefin-as-a-he
Yes, no doubt humans are naturally prone to create a narrative to “explain” events. But i think it’s important to note that those narratives can be tested with evidence to get more and more accurate predictions.
It’s the process of creating hypotheses that can be disproved and then gathering evidence to disprove them.
A very messy process indeed. But in principle doable. Predicting the weather is a pretty clear example of a very complex system that with enough data and the theories to define the mechanisms of various cascades yields pretty good results. As the technology of observation allows finer distinctions and the definitions of the mechanisms get more precise, there’s no reason to expect those predictions won’t get more accurate.
Medicine is another example of better patterns yielding the precision that defines the mechanisms of the cascades from which events emerge. As the mechanisms are better articulated it becomes possible to craft interventions with the minimum of unintended consequences.
It seems to me we are now at the stage of using the same approach to understand social phenomenon. As we get better and larger data sets on communication exchanges, my hunch is that there are developing languages of words and symbols that define the mechanisms and the movements of communication exchange. Given the experience of the other sciences, i don’t see any, in principle reason, the same will not become more apparent in the social sciences.
Thierry,
Spot on…Don’t Connect the Dots, Watch the Noise
http://jbordeaux.com/dont-connect-the-dots-watch-the-noise
“There are thousands of scenarios considered daily and we do not know which scenario, which threat, which dot deserves our attention before the fact.”
“…control can be an illusion and adaptation preferred. We are starting to focus on nurturing networks and relationships;a recognition that certain systems are, by their very nature, non-linear, and they change their behaviors based on their starting points and the random events that might ensue, leading to emergent new behaviors that cannot be predicted.”
Paula,
re: intuition and decision making (http://reflexions.typepad.com/reflexions/2007/11/intuitions-role.html)
Click to access barthmarrs_km_world_2007_final.pdf
Click to access intuitions-role-in-making-decisions-090309.pdf
[…] https://basreus.nl/2010/06/24/the-complexity-of-complexity/ […]
To do justice to this item and the comments in a response would take me too long and I just don’t have the time right now!
So, if I can get away with posting a link to my blog “Complexity Facts” http://wp.me/P16h8c-eu that will, at least, spark some thoughts &/or comments.
One of the comments made reference to avoiding semantic discussions on the subject and I know exactly what he means. Complexity is, as the name suggests, a subject that lends itself to l-o-n-g, albeit interesting, discussions.
Ontonix have a clear, concise, definition and have developed a means of measuring system complexity that is sufficiently robust to be deployed in Healthcare, business, Air Traffic control, CAD, CAE, etc.
I put this and the host of complexity-based blog items out there in an effort to alert enlightened individuals that real progress has been made in the inter-disciplinary area of complexity and really welcome feedback/questions…but I only have limited time for semantic discussions only because I am genuinely too busy getting on with identifying scope for practical applications in keeping with “Our Mission” http://wp.me/P16h8c-f7
David
They can be helpful in emergencies such as illness or car repair bluehost renewal coupon if
facing such an urgent situation, owning a car may prove being an asset in more
ways than one.