Rohan makhecha jw3GOzxiSkw unsplashIn recent times I’ve been engaged in a series of interesting exchanges with Dr Mike Jackson OBE, hereinafter referred to as Mike.  I have a visiting chair at Hull University where he is Emeritus Professor and as well as a fair number of mutual friends including Yasmin Merali and Gerald Midgley.   I have been provisionally scheduled to give the annual MC Jackson Lecture in 2023, Peter Senge gets the slot in 2022, so there will be some interesting contrasts to be made.  This year will see Carlo Rovelli is leading a symposium on the work of Alexandr Bogdanov whose work in systems has been much neglected so I am looking forward to that.

I should also make clear at this point my gratitude to, and respect for, the work that Mike has put into understanding not only Cynefin but the wider fields of complexity and systems.  It makes an exchange both interesting and rewarding and allows for a non-homogenising understanding of the wider field.  We share concerns about the rejection of all Systems Thinking by Stacy, and the, at times arrogance of the agent-based modelers of what I call Computational Complexity.

Now the exchanges, while interesting, have led to a certain amount of bafflement on my part as a large part of my responses have been along the lines of but that isn’t what I am saying and that isn’t what Cynefin is about and variations on that theme.  When this happens it is usually a result of the way one or other party is framing the problem and/or the way the idea is being communicated.  Two recent events resulted in a breakthrough for me at least, the light dawned and while I don’t yet hold said light in the palm of my hand I think I am getting there.

The two events were Mike’s review of the EU Field Guide on Linkedin and a lecture he gave online for the Cranfield School of Management.  I had tweeted a contrast between the field guide and an almost parallel publication of the GAPPS framework for leadership in complexity.  The former talks about what you do collectively, the latter fits within the pattern of identifying leadership qualities.  I’ve always found that type of approach to be interesting but lacking in pragmatism or (from a complexity perspective) too focused on supposed agents and insufficiently focused on the system and interactions with the system.  Mike picked that up and I think our views are in harmony and well summarised by him as follows:

… from the CST perspective, and the point has also been made from a complexity theory viewpoint by Dave Snowden (on LinkedIn), the list gives the impression that competencies are static, and that homogeneity is desirable. In the midst of a crisis, brought on by complexity, the appropriate leadership qualities are likely to be emergent rather than fixed and those dealing with the crisis better served if they display a diversity of competencies

CST is shorthand for Critical Systems Thinking which is as much Mike’s as anthro-complexity and naturalising sense-making are mine and both of us are passionate advocates for our views.  Another post from Mike seeks to take Mazzucato’s excellent work on Mission Economies and use it as evidence for the critically (sic) of CST and I may do something similar.  In the context of that work.  I also agree with the intent of Mike’s closing question as the grand challenges presented clearly require a post-capitalist economy.  We have a lot in common, including studying at Lancaster, but also some significant differences.

To return to the main storyline.  I promised to write a response to Mike’s review here on the blog and this post is it, but that is to some extent now a postscript to the main event.  I was going to listen to the Cranfield lecture anyway, but I did so with particular attention and Mike placing Cynefin front and central in the Scientific approaches to Complexity was good to see, but then he suggested that strange attractors should be thought of as metaphors as they don’t manifest themselves readily in social systems.  His lecture lists Cynefin (I have to teach him how to pronounce but, but I do like the new Yorkshire variant) and recognises my ‘people are not ants or termites’ criticism of much of Santa Fe.  But he goes on to say (correctly) that I still adhere to a naturalising perspective.  His problem with that is that he thinks that there are issues in suggesting that issues “intentionality and meaning, purposeful and power can be captured using natural science perspectives”.  Those two statements, along with my puzzlement about some of the statements in the review created a lightbulb moment.  What we have here is a long-standing misunderstanding between the Systems Thinking and Complexity Communities around the whole idea of enabling constraints and a misunderstanding of the matching of natural phenomena, understood scientifically, to the humanities (of more interest to me than social science per se).

So this post is going to be in two sections, the first on the way that science is used in Naturalising sense-making in general and anthro-complexity in particular; the second dealing point by point with other issues raised in the review.   There is also a post-script as to why I am posting here not on LinkedIn.

Science as an enabling constraint

Let’s start with a quote from Mike’s review, a point he repeats in the lecture

The Cynefin version of complexity theory is a ‘naturalising’ approach that seeks to be relevant to social- or anthro-complexity. It wants to bring ‘good science’ to bear to understand how humans interact with each other and engage with the world. Snowden is critical of existing science-based variants of complexity theory when they reduce the complexity exhibited by humans. Humans are not the same as ants, birds or crystals, he insists. Any complexity theory worth the name, and seeking to address anthro-complexity, must take account of human identities, values, intentions, and cultural practices. It is a difficult feat, I will argue, to remain ‘scientific’ while embracing those features of human systems that have been subject to multiple interpretations in the social sciences, leading to the paradigm wars with which other forms of complexity theory have had to become engaged.vv

Here Mike is picking up on Checkland’s suggestion of the limitations of science and I am more or less in agreement.  I’ve had some bitter arguments with people from the Santa Fe Institute (not all of them and a lot agree with my criticism) who think it is only a matter of time before they can model everything of significance about human behaviour.  I think that is a fundamental error.  At the same time, social science has a significant issue with verifiability, and schools of thought can emerge, achieve dominance and survive without any requirement for experimental or other validation.   I’ve always found the debate between critical realism and social constructivism stale for example, the former defining itself in opposition to the latter.  But schools of thought can easily emerge based on interpretation of cases or of a field that then takes on a life of their own, precocious becoming pernicious forms of autopoiesis.  But the humanities and many aspects of social science are powerful in their ability to make sense and to make meaning in understanding human systems.   DeLanda’s reinterpretation of Deleuzian epistemology using the idea of assemblage has been a key aspect of the Cynefin work from our early days on the DARPA GENOA II project, but so has Deacon’s work in Symbolic Species which finally challenged (conclusively in my view) Chomsky’s theory of language.  Many people have seen in SenseMaker®, our approach to distributed ethnography a manifestation of the research approach advocated by Bourdieu.  I am proud, that on the promoting of Zhen and John, that I was able to establish Derrida’s use of Aporia into the language of the Field Guide and more generally.   I’ve also taken the idea of exaptation from Gould, Odling-Smee, and others and created methods and tools based on a key concept in biology, namely the radical repurposing of a trait that evolved from one function to something completely different.  Mary Douglas’s work in Anthropology, and on one memorable occasion Mary Douglas in person was a part of the design of several Cynefin methods.   I also worked with Frank Land and some of the early pioneers of Cybernetics several decades ago, learned a lot, got some great ideas but also decided to move on.

So it is plain wrong to suggest that Cynefin, or more specifically the field of Naturalising sense-making (one of the five identified approaches to sense-making) limits itself to the natural sciences.  Sense-making is defined as how do we make sense of the world so that we can act in it and the naturalising element is a reference to a school of Philosophy that seeks to base epistemology, and more recently ontology into a stronger relationship with the natural sciences.  The great advantage of the natural sciences is that knowledge arises, not just from developing an explanation but from testing that explanation and critically your experiments being validated by other scientists.  So Naturalising Sense-making starts in a field by asking what we know from a natural science perspective, accepts that as an enabling constraint, and then goes on to develop practice accordingly.  The purpose of an enabling constraint is to disrupt, by creating a gradient, what would otherwise be an equiprobable state.  Social science, without that constraint, can, and has, asserted more or less anything with varying degrees of coherence,  So let’s take three examples of how this has worked:

  1. We know from work on inattentional blindness that people only see what they expect to see, so we realise that running a workshop is likely to reinforce that heuristic (it’s not a bias in the main it has an evolutionary advantage).  We also know from a body of work that Participative Action Research (firmly in the ST tradition) has been criticised for privileging the culture of the facilitator/designer and that is before we start on issues of epistemic justice.  So we accept that as a constraint and develop a tool, MassSense, which is designed to identify the 17% who have seen a gorilla on the final X-Ray and bring attention to such views.  This was also part of a body of work developed when I was challenged by John Poindexter in DARPA days to find a way of resolving the issue of creating an objective approach to abduction.
  2. Again the fact that art and music preceded language in evolution gives us a key insight.  Something that probably developed as a means to communicate swiftly moved into more abstract forms – we now have evidence that Neanderthals created abstract cave art.  In evolutionary terms, it appears that abstraction, removal from the concrete is one of the ways that we have developed the ability to create novel connections between things, which in turn accounts for the extensive use of exaptive practice in the evolution of technology.  So in our work on radical repurposing the exapt phase of the Field Guide we look at abstract signification to break entrained patterns of thinking.  Checkland’s work also recognised this (but without the underlying science) and I respect that, but the fields of semiotics and aesthetics are far more advanced now, and to an extent then.   I was studying aesthetics a few hundred yards away from where Checkland worked at Lancaster University and that was in the early seventies.
  3. We are currently looking again at some of the practices of psychotherapy, but not from theories of Freud and Jung, but from modern Cognitive Science and the work of people such as Andy Clark (who bridges Science and the Humanities) with a post-cartesian understanding of human consciousness.   It is no surprise that a lot of people in the field are both scientists and philosophers as the former is creating the discoveries that the latter can interpret.  Distributed consciousness and fuller development of Clark’s ideas of scaffolding will, I predict, produce a more radical change in organisational design than anything we have seen to date.  The process here is to take practice and to validate it against modern science rather than 19th Century thinking that at times verges on the mystical.  If it contradicts what we know then we have to look for Placebo or Hawthorn effects in order to determine use or rejection; if coherent then we can amend and scale.  Knowing the why allows us to scale the what.

The fact that Cybernetics had addressed issues of complexity is without question.  But equally, humans found all sorts of ways to work with gravity before Newton came along, after that everything was different.

So a naturalising approach to sense-making gives us three benefits at least

  1. It can break sterile debates and differences within the social sciences and can give new insight and understanding.  Strange attractors are far from a metaphor, they are real phenomena in human systems.  Once you put the science together with assemblage theory and the narrative concept of a trope you start to realise that this is something you can both measure and monitor; SenseMaker® was designed for that as well as to create better training datasets to handle the Stochastic Parrots issue but that is a story for another day, and SenseMaker® does a lot more besides.
  2. It gives us an alternative to case-based approaches to deriving theory (Weick etc) which is closer to Physics.  The equations of theoretical physics create hypotheses that are tested by experimental physicists.  We have done a lot of that with the Cynefin work, starting with natural science and then experimenting.  This is important during times of uncertainty as it avoids the correlation-causation error that is a part and parcel of inductive approaches.  Uncertainty requires abductive approaches more than inductive ones.
  3. Faced with multiple options it allows us to test hypotheses for coherence to identify what are legitimate paths of investigation and/or experimentation, it thus reduced the cognitive load of decision making

There is another key quote in Mike’s review relating to mass engagement and human sensor networks so again let us start with a quote from Mike’s review

Geoffrey Vickers, for example, argued that the components of human systems, active individuals using ‘appreciative systems’ to attribute meaning to their situation, makes it impossible to study them using the natural scientific approach. Following Vickers’ insights, and drawing upon hermeneutics and phenomenology, Checkland rejected any attempt to understand problematic social situations in scientific terms, and developed ‘soft systems methodology’ as an approach that works with different perceptions of reality and facilitates a systemic process of learning that can lead to purposeful action in pursuit of improvement.

I think I have already answered this but I wanted to make what we do clear.  We developed an approach, using concepts of abstraction (see above) by which human actors could attribute meaning to their experiences in a quantitive form.  It is a part and parcel of my humans are not ants criticism of the computational complexity folks.  We are currently extending that to include high abstraction searches using hermeneutics.  Vickers and Checkland currently stated a problem at the time, they set a question and restriction and trying to work around that was a driver for my work in DARPA.  Their statement of the problem was critical, but things have now moved on we have approaches that are not subject to the process limitations we see in Checkland, but probably would not have happened without his initial insights.  Those theory-initiated methods and tools can also represent in statistical terms multiple perceptions of reality and by carrying the narrative with the statistics allows the more stories like these, fewer like those approach to Vector Theory of change.  It doesn’t make it all scientific. but it radically increases the scientific element and the ability to operate at scale and in real-time: action and interpretation are entangled at fractal levels.

Overall this also links in to the concept of democratising sense-making to return to an earlier post. That extends the ideas inherent in the various bodies of work on epistemic injustice (Humanities again, we use and draw on it extensively) and also argues against a reliance workshop/discourse approaches common in Systems Thinking practice. The very linear approach of getting people together, discussing what should happen within a framework, and then assuming that agreement will translate into action has been a real issue and still is. In our indigenous work where the action is intertwined with thinking it wouldn’t be understood. The very culturally specific orientation of Systems Thinking is something we have been seeking to overcome in the wider sense-making work. It’s not that that workshops are not important, or that articulation of intentionality and meaning does not have its place, but it is a part of a wider approach. Now again I’m happy to be pointed to material from a CST or any ST background which doesn’t do this. But naturalising and democratising sense-making are also a way to shift away from the its all about perception error which is argued by many cyberneticians. As we say to post-modernists (when we are not talking about giving a flying duck) is that reality exists, live in it. Yes perceptions matter, but human knowledge can exist independently of perception and increasing the friction between perception, knowledge and reality is a key aspect of this approach to sense-making.

On the field guide review

This is not as important as the above section so I will write it as a series of bullet points roughly by paragraph sequence

  1. Mike points out that the field guide has a certain coherence but that in consequence, it means that “other complexity and systems perspectives receive little attention”.  To that, I would point out that this is intended as a Field Guide not an academic review of the field.  It identifies a series of activities and types of action that are needed and most of the fields that Mike mentions, including his own, could make a contribution to that practice.
  2. There is also this rather odd phrase that Mike uses both in the review and the lecture where he talks about “complexity theory’s promiscuous crawl through social theory”  Now it’s a neat phrase with its implications of multiple dalliances without commitment but I do think it is the wrong word to use.  In effect Complexity Theory (and it is not the only science involved) is providing us with a radical new way of thinking about some of the old dilemmas in both the humanities and the social sciences.  The fact that ontologically incompatible states can co-exist (that is a simplification but this is a blog post) enabled Alicia Juarrero’s brilliant re-examination of the problem of intentionality, better expressed as When is a wink a blink? and some emerging work is starting to give us new insight into Free Will and The Hard Problem.  Too much of social science is still stuck in the understanding of science expressed in the work of its founders Compte and Saint-Simon and the paradigm shifts in Physics, Chemistry, and Biology have long passed that by.  Complexity theory is allowing us to re-examine (a better phrase than promiscuous) and a lot of the sterile debates of social science (see above) go away.
  3. Mike does a good job of summarising the essence of the field guide in a long paragraph and despite his concerns, I don’t think that he loses the dynamism of the guide.  I do think he misses a key aspect of the guide, and I admit we didn’t make it clear at the time, namely the use of “(and Chaos)” in the title. All of Government strategy, as well as industrial work, is about managing as menege more than menage and the Guide has much wider applicability than a crisis.  The two stages after the pivot, exapt and transcend are key to creating downstream resilience.  The Guide is the starting point of a journey to a very different way of understanding organisations and society in general.
  4. I agree with Mike on the criticality of works such as Burns and Stalker’s seminal The Management of Innovation first published in 1961 and still relevant today.  Mike is also right that complexity theory has lacked a body of methods, and a coherent methodology (my rephrasing) compared with Systems Thinking and the Field Guide was certainly intended to rectify that and is the culmination of a Theory informed practice approach that I and colleagues have been engaged in for the best part of three decades.
  5. I disagree that Mingers Appreciate-Analysis-Assessment-Action process is the same as the Field Guide’s Assess-Adapt-Exapt-Transcend process.  I know Minger’s work and would have acknowledged it if I had used it.  I think fitting one onto the other is a little artificial and loses the exapt-transcend idea which is all about a phase shift.  But I appreciate Mike making the link and I will pick that up in subsequent writing
  6. Mike is completely right in suggesting that Ralph Stacy in his current manifestation would not see the Field Guide as being anything to do with complexity.  The earlier Ralph Stacy would have seen it as valuable, or at least I think he would.  The one time we had extended interaction in Lecce there was a lot of agreement but more disagreement.  In his latest manifestation, Ralph seems to condemn as systems thinking anything which is in the very restricted field of not-Ralph and in conferences in both Liverpool and Germany has adopted some of the worst aspects of Old Testament Prophets.  One of these days his successor Chris may be released to have an intelligent conversation with the rest of the community but for the moment he risks ex-communication. Ralph is a convert from economics and there is nothing so dangerous as a convert.
  7. Whether CST echos Cynefin or vice versa is a moot point, but both agree context-free solutions will not work in a context-specific work and there I am doing some key work on semiotics to create a different type of pattern language that will be announced at the end of April, the banner picture gives a hint of what is coming there.  that will allow CST processes to interaction with Cynefin ones.
  8. The main critique from CST is that the Field Guide could have simply used existing methods from Systems Thinking which “seem to offer the most obvious and proven resource”.  Now there are three general responses to this (i) if they did why haven’t they after many decades in other than isolated pockets, and (ii) we had all sorts of ways of managing Gravity before Newton came along but after that it was different, no one denies that ST has tried to address the same issues as CAS (iii) there is no reason why some of those methods should not be brought in.  The Field Guide provides an overall activity-based structure and doesn’t mandate practice.  The assessment process I am currently building specifically focuses decision-makers on using what works.
  9. Like most people in the field, I have a lot of respect for Beer’s viable system model but I have long argued that if he had been born a couple of decades later he would have been able to use the new insights and understanding from science (not just complexity science) and would almost certainly have produced something different.  There is nothing static in this field and nostalgia for past insightful work should not hold us back.  My view is that VSM is too instrumental and over-focused on design with insufficient attention to generating informal networks.
  10. The same is true for Checkland’s idea of rich pictures – I used a lot of soft-systems methods when I was working in Datasciences and in the early days of IBM. Similarily with the various curated forms of narrative collection – curation is an issue both from a research and a scaling perspective. Checkland’s rich-pictures were pioneering at the time showing flows with graphic facilitation at various levels of artistic capability. All good stuff and it triggers me and others to start thinking more generally about the field. Then we engaged in the big DARPA programs on weak signal detection and put frankly those approaches, inspirational as they were, just didn’t scale.  That was when I first developed the ideas of distributed ethnography and more recently started to build ideas of epistemic justice. We also moved on to more sophisticated approaches to semiotics and aesthetics in sense-making, informed and constrained by a growing body of natural science. I’m currently looking at the various work on the olfactory nature of human sense-making and you can expect new methods based on that early next year.  Journaling and distributed decision-making using high abstraction metadata are not a part of the sources Mike quotes, they are different, and with due homage to the past, it is time to move on.  I’ve responded to this point more broadly in the earlier section – see the response to Vicker’s quote
  11. Mike’s suggestion that I have invented a “whole new technical language of concepts” is partially true, although a lot less so than with many others.  But I suggest this means he hasn’t read or appreciated Heidegger’s point that while we think we are the masters of language, language in effect controls the affordances of meaning that are available to us. In fact, the number of new words is limited to what is necessary to shift understanding – exaptation, aporia, enabling constraints.  The list of new words is not long and contrasts well with Cybernetics which has created a whole esoteric and elitist language.  Checkland was a good antidote to that but we can do better.
  12. Mike makes the statement “it is shared appreciations, values, and intentions, at the level of meaning, that actually leads human beings to act in consonance” and does this in the context of rejecting the idea of strange attractors as anything other than a metaphor.  Here he is rather hung out on his own petard, as ideas of affordances and assemblages in the humanities (not natural science) mesh well with the idea of strange attractors.  They provide a both/and approach to ideas of compulsion other than Mike’s either/or.  Again this falls back to the fundamental misunderstanding (and I am sure I am partially responsible) of our use of natural science as an enabling constraint and not taking account of the fact that Cynefin draws on the social sciences and the humanities.  Indeed there are academic papers in Germany that say Cynefin is a modern form of Daisen and Heidegger is hardly a natural scientist.
  13. So Mike’s statement that “The attempt to understand anthro-complexity with concepts and tools drawn from the natural sciences acts as a significant constraint on the argument of the field guide” is at best incomplete, at the worst misleading.  The insights in the Field Guide owe as much if not more to the humanities as they do to natural science, but they working within the enabling constraints of natural science.
  14. Again we get a sort of straw man fallacy when Mike says “Dave Snowden needs social theory to really get to grips with social complexity”.  Well, I and colleagues have drawn heavily on Deleuze, Derrida, Bateson (as an anthropologist), Bourdieu, and many others.  So the statement is false unless Mike means engagement with the body of social theory that makes up his own CST approach (which is strongly advocated in the review).  I can respect that approach, I have learned from it, I don’t reject it but I think to paraphrase Lincoln that it is inadequate to the stormy present.  We have drawn very broadly on social theory to get to grips with anthro-complexity.  It is notable that most of our non-science sources come from the continental, not the Anglo-Saxon tradition in philosophy.
  15. There is a fundamental issue that arises out of this.  Too many systems thinking practices are based on workshops, study, and linguistic agreement to think holistically, act differently, or the like.  Put bluntly that can work in the moment (when the success is often reported) but it has issues with sustaining change at scale  The approach has had several decades to develop but the field of human understanding and meaning-making in the sciences and humanities moves on and we all need to move with it.

A final point here, and an open question to Mike.  He says “the easiest way he can improve the field guide is to recommend systems approaches which have already translated the insights of the different epistemologies offered in social theory into practical methodologies”.  Well as I have said the Field Guide does not mandate or prescribe and the assessment process will leave open a pathway to multiple methodologies.   We focused the guide on practices that could be implemented and used quickly, not workshops and discussions to agree on what should be done.  If Mike wants to list some of those from his own tradition I will happily go through and comment.   But I’ve already explained why I didn’t take forward the soft systems approach which is one example he gave. But I am happy to engage there or around other methods.  Also, the pattern language alternative we are developing would allow those methods to incorporated and I’m happy to work with him on that.

Post-script on the cybernetics fanboys

Mike recognises the ST and CAS backgrounds are different and is happy with robust debate and is prepared to engage in a meaningful way.  What I have found on LinkedIn of late is that there is a small coterie of cybernetics fanboys (and they are all boys) who obviously feel threatened by Cynefin.  Their comments range from patronising to occasionally interesting.  For some their tendency to ad hominem attacks shows though, and back door one to one libelous communications with other participants really breaks any normal ethical standards.  There is no debate in an open unmoderated forum where people choose to mob against the unfamiliar and you give them too much credibility by responding time and time again with the same dialogue.  The intent is pretty clear to try and drive people away from a post or use it for self-promotion.  All are welcome to comment or ask questions, but if I get the same sort of behaviour as I have seen on LinkedIn I will simply block the contributors.  Few of them would survive Wikipedia where behavioural norms act as an enabling constraint.

Also, the above material is important (at least to me), and placing it here gives it more permanence.


The opening picture of a light bulb in a human hand is an original photo by Rohan Makhecha Banner picture of a Settlers of Catan game in play is by Galen Crout both on Unsplash

< Prev
Next >

Flexuous flight

I first used the illustration to the left last November when I was updating the ...

Further Posts