How elephants, icebergs and leverage points stand in the way of serious systems thinking

By Michael C. Jackson.

michael-c-jackson_2026
Michael C. Jackson (biography)

Why is serious systems thinking about complex situations hindered by suggesting that it has something to learn from viewing complex situations as though they are like elephants to blind men, or icebergs, or have leverage points?

Elephants

In the ancient parable of the blind men and the elephant, various blind men touch different parts of an elephant and conclude that they are encountering a spear, a fan, a wall, a rope, a tree, or a snake, depending on which part they handle. The person touching its trunk, for example, concludes he is dealing with a snake. The purported lesson is that systems thinking needs to avoid looking at issues from a partial perspective and needs to be sure it gains an understanding of the whole system.

There are at least five problems with this parable in the context of complexity.

  1. C. West Churchman (1968) was an eminent systems thinker who interpreted the parable very differently and in a manner far more appropriate to dealing with complexity. He argued that the teller is arrogant in assuming that they can adopt a ‘God’s eye view’ which enables them to see the whole situation of interest. This is impossible in the case of complexity where the teller needs to accept that they are part of the situation and offering just another perspective on the nature of that complexity. Those who assume super-observer status are mistaken; they are also dangerous.
  2. Churchman (1968, page 35) also states that perhaps the teller is being rather blind in taking “the skin of the elephant [to] really represent the dividing line between the elephant and its environment.” Doesn’t the boundary depend on the purpose of the observer? From a complexity perspective, there is much more to elephants than their bodies, ears, trunks, etc. They can be animals in a zoo, part of the timber and logging industry, elements in a conservation project, part of the ivory trade and more.
  3. A related point is that when dealing with complexity, partial views of the situation cannot be aggregated to provide a more perfect picture of the whole. They represent genuine alternative engagements with a reality that is too complex to grasp. For example, a student might see the UK National Health Service (NHS) as a potential career opportunity; a manager as somewhere to exercise her communication and control skills; a nurse as providing a rewarding profession; a cleaner as offering a job; a patient as a mess they must negotiate to get the treatment they need. They are not seeing parts of the same neutral object but offering different, partial, but valid descriptions of the world.
  4. Dave Snowden (2023 and 2026) argues that the elephant example describes a relatively stable object, the behaviour of which we can largely predict, and which can be relatively easily trained to fulfil certain purposes. It hides from us the dynamic interactions that occur in complex systems. Here, the parts and whole are rapidly changing during the process of interaction. There is no pre-existing object to be discovered. The interactions, including those in which we are involved, lead to emergent properties. We cannot tell in advance what these will be. We need to engage with the complexity, evaluate the possibilities, and test alternative strategies to learn what works and what doesn’t work.
  5. Finally, the elephant described in the parable is pretty much a fixed entity. We can change it a little from the outside, perhaps by cutting its tail off. But we cannot radically change it to better serve an existing purpose or to align with a different purpose. This kind of change is very different from the changes we can potentially make, alongside others, to organisations such as the National Health Service.

Icebergs

An iceberg, of course, has most of its bulk below the waterline and, therefore, invisible to the observer on the surface. The lesson is that there is a need to pay attention to the underlying causes that give rise to the ‘events’ more easily observed. In the usual representation, the bit above the water is labelled ‘events’, while lying further and further beneath the water are ‘patterns and trends’, ‘structures’, and ‘mental models’. There is an implication that the more fundamental causes of what happens can be found at the deepest level, ie., ‘structures’ generate ‘patterns and trends’ and ‘mental models’ shape ‘structures’. There are significant issues with this metaphor when it is employed to suggest how we should address complexity.

  1. Many of the problems are the same as those discussed in relation to the elephant metaphor. The observer is credited with a ‘God’s-eye-view’, standing outside the situation of interest and able to identify and act on the true cause of the behaviour they witness. And, as Blomsma and Lenton (2026) argue, icebergs are rather large and immutable objects which are likely to be highly resistant to change. It is certainly not clear how we might engage with them to effect a beneficial transformation in their behaviour.
  2. Further problems arise stemming from the metaphor’s adherence to the notion that you need to ‘dig deep’ to get to the true explanation of events. This, in Morin’s terms (2008) is a clear indication of the reductionism encouraged by ‘simple thought’. It is reinforced by the implication that there is a hierarchy of causes, with ‘mental models’ as the most fundamental. This suggests that all problems have a cognitive basis. Instead, in complex situations, causal relationships are circular and recursive, existing in mutually influencing relationships. And, of course, there are likely to be numerous causal factors involved at any one time.

Leverage Points

The idea of the leverage points metaphor is that, if you can understand the architecture of a system, whether mechanical or social, you can find an appropriate leverage point to effect the change you want. In particular, you may be able to discover a point where a small, targeted expenditure of energy will produce a substantial change in a direction you approve.

Many of the problems are the same as those discussed in relation to the elephant and iceberg metaphors. It requires a super-observer capable of standing outside the system, understanding its structure, and knowing exactly where to intervene to get the hoped-for outcomes. In ‘complex’ situations, too many elements are interacting, and the norm is to find unpredictable non-linear relationships between the elements. It is unrealistic to think you can discover one decisive point to intervene and predict the consequences of doing so.

Conclusion

When I make the case for eradication of elephants, icebergs and leverage points from the discourse of systems thinking, I am often told that, while these metaphors are simplistic, they do offer beginners an easy route in. I take a contrary view – that it is cruel, disheartening and confusing for novices to be led in the wrong direction and then asked to retrace their steps. Promoting accessibility by replacing the ‘complex thought’ necessary to engage with complexity with the ‘simple thought’ exhibited in the three metaphors is a way backwards, not forwards.

What do you think?

To find out more:

This i2Insights contribution is a shortened version of “Doing Serious Systems Thinking Means Getting Rid of Elephants, Icebergs and Leverage Points” published 0n 14 October 2026 on LinkedIn at: https://www.linkedin.com/pulse/doing-serious-systems-thinking-means-getting-rid-jackson-obe-pswoe/

References:

Blomsma, F. and Lenton, T., M. (2026). Metaphors for Agency: Getting Past the Paralysis of Complexity and Accelerate Transformative Change. Preprint, SSRN (Social Science Research Network). (Online): https://ssrn.com/abstract=7301480 or http://dx.doi.org/10.2139/ssrn.7301480

Churchman, C. W. (1968). The Systems Approach. Dell Publishing: New York, United States of America.

Morin, E. (2008). On Complexity, Hampton Press: Cresskill, New Jersey, United States of America.

Snowden, D. (2023). The Pachyderm Post. Cynefin Company website. (Online): https://thecynefin.co/the-pachyderm-post/

Snowden, D. (2026). There Is No Effing Elephant. Cynefin Company website. (Online): https://thecynefin.co/there-is-no-effing-elephant/

Use of Artificial Intelligence (AI) Statement: Artificial intelligence was not used in the development of this i2Insights contribution or the work on which the contribution is based. (For i2Insights policy on artificial intelligence please see https://i2insights.org/contributing-to-i2insights/guidelines-for-authors/#artificial-intelligence.)

Biography: Michael C Jackson PhD OBE is an Emeritus Professor at the Centre for Systems Studies, University of Hull, UK. His latest book is “Critical Systems Thinking: A Practitioner’s Guide” (Wiley, 2024).

16 thoughts on “How elephants, icebergs and leverage points stand in the way of serious systems thinking”

  1. Thanks Mike – and David’s references too – for this discourse on systems thinking, metaphors, complexity, elephants, ice-bergs and leverage points. I’ve been there, done that, but not got the T-shirt!

    I understand your frustrations: I share some of them. What do I think?

    A parable tends to have a moral, ethical or spiritual lesson within it if reflected upon fully. As far as I understand, it’s not sensible to shave off certain bits of it to suit one’s argument. There’s nothing wrong with the parable (or the poor elephant), but it can be misappropriated in the context of complexity when people refer to a prospect of seeing the whole – or ‘a whole system’ (an impossibility), or by remaining ignorant of the wonderful dynamic, uncertain and unpredictable characteristics of complexity within which we all reside.  As serendipity would have it, Luke Craven has just penned #50 ‘There is no elephant, but there is a king’.

    https://pigontracks.substack.com/p/50-there-is-no-elephant-but-there

    It’s worth a read.

    As for metaphors more generally, it’s interesting in itself that people are seeking more appropriate metaphors for complexity. The Blomsma and Lenton paper you reference develops an interesting nautical metaphor consisting of flap rudders, the fleet and flow. (I have a personal preference for the allegory of Plato’s cave, which is another story…). Gareth Morgan’s substantial body of work around the use of metaphor includes an observation about the effectiveness of metaphor: “Effective metaphor is a form of creative expression that relies on constructive falsehood as a means of liberating the imagination” (Morgan, 1980, pg 612). And again “encourages a spirit of critical inquiry and cautions against excessive commitment to favoured points of view”. So use of metaphor should help us to be less restrictive in our thinking and is not intended as a way of locking in a form of dangerous certainty about how things are.

    I wrote a piece for the Operational Research Conference (OR60) in 2018 about the significant loose ends of systemic leadership. I identified one of these loose ends as ‘systems thinking’: as an empty phrase which has in effect been hollowed out. I identified the need for a counter-movement ‘to protest loudly against this form of superficiality.’ I do think things have moved on apace since then in terms of ‘serious systems thinking’, even though I prefer the term ‘systemic thinking.’

    Contextual exploration is important, so I do see such shared exploration as the order of the day. It’s not about the nouns of systems, metaphors, elephants, ice-bergs, leverage points and organisations, but how we choose to organise and arrange ourselves dynamically and collectively in a form of systemic inquiry, learning and adaptation within this non-static and continuously ‘becoming’ universe.

    Unfortunately, we can never neatly put everything we don’t like on the discard pile – these things persist and we somehow need to focus less on who we believe is right/who is wrong, and more on the potential of liberating the collective imagination around what is absent, by widening thinking together, being more aware of and honest about our limitations, recognising the creativity of dissent, and thus letting go or setting free an improved human capacity for dynamic, systemic inquiry in a complex universe.

    Morgan, G. (1980). Paradigms, metaphors, and puzzle solving in organization theory. Administrative science quarterly, 605-622.

    Reply
    • Can’t disagree with that Cathy. As you know, I am having a go at simplistic uses of the three metaphors. That said, why start with metaphors that require so much hard work to make them say anything relevant about complexity?

      Reply
  2. Thanks for the short but thought provoking post! I generally agree with the sentiment of this post but I would want to caution that the post does not engage with “steel man” versions of the criticized concepts. For example, one could use the elephant story to tease out very similar points made in this post. Also, arguably the points of the iceberg metaphor and also the leverage points concept could be made differently. With the iceberg one could focus on the relative “weight” of different aspects in trying to make sense of a situation (i.e., directly observable vs. “hidden” aspects). Then it seems like it could be useful heuristic rather than a misleading metaphor. Similar with the leverage points. The way Donella Meadows introduced the concept the key point was not to identify any particular leverage point in a specific “system” but to offer an heuristic that points towards general mechanisms that can change the behavior of systems in general. To sum it up, I think that the strong connection to “super-observer capable of standing outside the system” is not inherent to these metaphors and concepts themselves but actually coming from a specific framing and presentation that can be avoided.

    Reply
    • Although Meadows talks about the greatest leverage point as moving between paradigms, she actually remains stuck in the system dynamics paradigm and its concept of leverage points.

      Reply
  3. Hrm.. I’m definitely still fairly novice in the complexity area. I read Dave’s Pachyderm posts a while back, and I definitely don’t have the feeling of having been led in the wrong direction or having to retrace my steps. I find the kinds of points you’re making here just provide more nuance, and build on the understanding I had before in, generally in an interesting/exciting way.

    I tend to think of these kinds of metaphors as useful blunt-instruments for getting kick-started in knowing what to look out for. They aren’t useful for providing a permanent structure to hang your thoughts on, but they are useful for providing a temporary structure for getting your thinking out of the rut it’s stuck in.. I’m very happy to hold a weak/partial model of the world loosely, and to be open to overturning or updating it later.

    Re: I think the leverage points critique is less convincing for me.. I mean, I think your points are solid, but in my reading of the point of Meadow’s essay wasn’t “Understand the whole system and then find the single best place to act”, but “there are diverse possible ways to act on (parts of) a system that can have different effects”, and that being able to recognise these can help you look out for plausible paths of actions, and might help you estimate their relative efficacy (acknowledging that you’re always going to have only a partial and possibly misleading understanding of what the actual result will be). Even if I’m reading that wrong, the catalogue that Meadow’s essay presents seems pretty useful for making people aware of the diversity of ways of acting (or of being acted upon).

    Reply
    • Why start with ‘blunt instruments’ rather than ‘systemic perspectives’ that have stood the test of time? (see my Critical Systems Thinking, A Practitioner’s Guide’).

      Meadows produces a good list but always constrained by system dynamics thinking. She urges mixing across paradigms but never escapes the limits of system dynamics.

      Reply
  4. [Re-posting my comment from LinkedIn as requested by Gabriele]

    Thanks for this, I enjoy a good debunking and agree that metaphors are powerful and should be used with care. I personally think all three still have value, but I’d love to hear of the authors (or anyone else’s) preferred metaphors for this topic

    [Mike and Gabriele replied on LinkedIn pointing me to the response to Susanne. I definitely didn’t have the same reaction as Susanne, but read with interest Mike’s post on pragmatism and the two Dave Snowden posts linked at the bottom.

    I wouldn’t say any of those posts put forward preferred metaphors for complexity or systems thinking – terms which I wield cavalierly as I’m totally unlearned in these domains. In fact Dave’s recent post has a fairly clear admonition to not apply metaphors to describe systems. The point is well made but as a fan of Doug Hofstadter, I reckon it’s just about impossible to avoid metaphors. Using them with care, tentatively, in certain contexts etc, is still hopefully doable. As an untrained outsider, I come to this discussion with a very different set of experiences and I’d probably be far more sensitive to people using metaphors I don’t like in my line of work!

    Thanks for sharing your ideas all, much appreciated.

    Reply
    • As I mentioned, my book ‘Critical Systems Thinking, A Practitioner’s Guide’ (Wiley, 2024) sets out 5 ‘systemic perspectives’ that have proved themselves useful in the face of complexity.

      Reply
  5. Thanks Mike for again emphasising that you can’t learn to be a systemic thinker using a reductionist mindset. I agree the learner is hampered by the use of static models that may be used to highlight a systems characteristic but then have to be unlearnt (harder than learning) to build a judgement free holistic mindset. In many ways this static sampling practice is an example of our laziness as teachers and learners to engage the unknown. I have discovered that inviting learners to focus on a complex issue in their own life that they share with others and then engage with the others in a wide ranging conversation about the issue leads to the recognition of new insights and innovative ways to manage a complex situation. The challenge for the learning facilitator is to create an environment where listening non-judgementally to others is more important than telling. The ‘technical’ aspects of systems thinking and its language can then be learnt retrospectively as the individual and group reflect on the conversation as its outcomes.

    Systems thinking is best learnt experientially not from a text book.

    Reply
    • I like that – asking learners to reflect on a complex issue in their own lives. They would be right to be dismissive of you if you suggested the elephant, iceberg, etc. metaphor could provide the solution for them.

      Reply
  6. Does the hindrance lie in the metaphors or in adopting a God’s-eye view? How might we benefit from them if we were to first acknowledge that an understanding of the whole system is outside of our purview?

    Reply
    • Even with that understanding, I don’t think you would benefit much from these metaphors in the face of complexity. They are misleading. You can benefit from well-developed ‘systemic perspectives’ that have proven their worth over time – please see my ‘Critical Systems Thinking, A Practitioner’s Guide’ (Wiley, 2024).

      Reply
  7. O.k., that’s interesting, but a DEEPLY frustrating article, because the author only takes away our toys, our mental models, but offers nothing in the way of didactics, analysis or intervention in their stead. So, I want to challenge the author to write the sequel and offer something useful, not just leave readers in a frustrating voice. Thanks!

    Reply

Leave a comment

This site uses Akismet to reduce spam. Learn how your comment data is processed.