The Illusion of Complexity

Introduction from the Editor

Complexity is not a feature of the world that can be modelled, or managed. Complexity is, rather, a function of our lack of understanding, or the way we have attempted to understand something—something which, once understood, can be seen to be fundamentally simple. Complexity needs to be filtered. The illusion of complexity, which continually renders our actions ineffective, results from the application of simplistic, sweeping midlevel abstractions to the idiosyncratic details of real-world situations. The radical implications for management and decision-making are the focus of this brief introduction to our contrarian view of complexity.

—The Editors

---

The Illusion of Complexity

A few years back, I had a pleasant lunch with a friend of mine at a restaurant in London’s West End. As we started on our soup course, Richard, who runs a leading management consulting firm he founded, asked me, “You hear a lot nowadays about how the world is changing faster than ever before. Is that your impression, James?” 

I had to think about it for a few moments, my spoonful of gazpacho suspended midair in front of me. Finally, I replied, with a shrug:  “Nope,” and downed my spoon.  To which Richard replied with a shrug of his own, “Yeah, me neither.”  

Richard then seemed deep in thought, and after a minute or so broke his silence to ask, “Ya know, James, you also hear a lot about how our world is far more complex than the world of 30 years ago.  Do you agree?”  

Well, I didn’t have to think about that one for longer than a New York nanosecond, which is the interval of time between the traffic lights turning green and the car behind blowing its horn. “More complex? No, of course not!” I laughed. To which Richard responded again, much more slowly now, “Yeahhhhhhhhhh, …me neither.” And then he let out an almighty, world-weary sigh that seemed to say, “James, I think I’m getting too old for this game, …or maybe too smart—like, man, wtf’s the world coming to? Why do people even think like this?”
 

Why the World Appears More Complex Than Ever 

The corporate managers of a few decades back were actually dealing with a greater degree of complexity than the computer-bound managers of today—complexity which they had skillfully and resourcefully rendered simple, obvious and intuitive, often deploying a far more varied skillset than contemporary managers realize they still have available to them today. 

I’m referring here to a dim and distant time, back in the fabled 1980s, when spreadsheets were literally big pieces of paper filled-in with a bit of pencil by nerdy accountants, and when a manager who was found sitting in front of a computer screen staring at it for more than five or ten minutes would be referred by his colleagues to Occupational Health, and if found fit and well would be hauled up before his boss and challenged on his apparent dereliction of managerial duty.  

I exaggerate here, of course. But only a little. Certainly, it was a different world—a more complex and nuanced one it seems to me, and one more easily, effectively and efficiently managed, though managed by very different, more varied and usually idiosyncratic means. In fact, the complexity or richness I’ve just averted to might better be described in terms of “thick” as opposed to “thin” descriptions as the Oxford philosophers used to say:  “A large, skinny light-roast decaf cappuccino with no chocolate on the top, in a ceramic cup” being a ‘thicker’ description than “coffee.”

No wonder industrial productivity hasn’t increased in generations, despite all our advanced technological tools to assist us. In reality, we have available to us a smaller range of more standardized tools, and a more abstract, more standardized and less versatile language for describing what we’re dealing with.

The world of 30 years ago to which my friend Richard referred in his second question was also a world in which senior managers still used to read serious scientific books relevant to their work, not just airport-bookstall, quick-fix business books. It was a time when denizens of the C-suite still read such sophisticated thinkers as Stafford Beer and Norbert Wiener, and when most of them still knew who Gregory Bateson was—and if you barely know who any of those three are yourself, don’t worry, you’re in good company, in fact you’re among the overwhelming majority nowadays. 

And nowadays the overwhelming majority are also the overwhelmed majority, and that’s for a good reason, as we shall see.  

Nowadays people are in fact, increasingly, only addressing themselves to an impoverished, abstract, surrogate world, in place of the infinitely rich, concrete detail of the real world in which a few of us still live, a world vastly more abundant in possibility. 

All abstractions are merely metaphors, and they limit us to the few possibilities available within the selected metaphor. 

If you have just a comparative handful of business-school metaphors at your disposal along with a handful of conventional MBA-standard-issue metrics to go proxy for the infinitely re-describable real world, then you are living in an artificially much simpler world than the real world, and you just have to manage that simpler, surrogate world because that’s called doing your job.

But the price of this artificial reduction in complexity is, at least way too much of the time, not being able to figure out what’s really going on, and that’s because you have abstracted out of account (= already deleted) all the valuable information that would easily tell you.

If suddenly things appear so much more complex to us nowadays, it is because we address ourselves to far less in the way of a range of variables and possibilities.  The world isn’t fundamentally more complex than the world of thirty years ago.  Nor is it the world itself that has changed.  Rather, it’s something else that’s going on.   Our thinking has become more abstract, regimented, simpler and less sophisticated, which makes the things we are trying to understand seem to be more complex, like trying to find your way around New York as a pedestrian with only a map of the New York subway system to guide you.

Recovering Our Sense of Wonder

Now let me tell you something that will amaze and awe you, and fill you with wonder. 

It’s this:  If you take a rock-solid ice cube out of your freezer and put it outside on the pavement in the sunshine, it will melt after a while into a puddle of water which eventually will partly seep into the ground but mostly will disappear into thin air.  Isn’t that incredible?  Doesn’t that fill you with wonder?  

Well it should.  It’s an amazing thing.¹  Scientists now understand quite a bit—though not everything, by a long chalk—about the changing states of water and the effect of temperature, and so on, but I would ask you, dear reader, to please experiment with adopting this mindset of absolute wonder, and if only for the remainder of your time reading this article.

In fact, it’s never a bad idea to look at familiar things with a sense of astonishment. Because everything is a miracle, really. 

Or at least to a philosopher, or to the philosophically minded, it should seem that way.  Plato in the Theaetetus tells us that philosophy begins in wonder, and Aristotle, in the Metaphysics concurs.  Aristotle also tells us that what is most obvious is what is most elusive, and Lichtenberg asked that God grant the philosopher insight into what lies in front of everyone’s nose. 

Francis Bacon’s great Project for creating the wealth of nations (which succeeded handsomely and effectively moved us out of the middle ages and into the modern world) was founded on the imperative to explore the most humble and trivial matters scientifically, and to invest capital accordingly.  

Think too of Freud’s quest to understand what he called the flotsam and jetsam of human life—dreams, slips-of-the-tongue and other errors, jokes, taboos and so on—soon recasting our whole notion of consciousness.  Or consider Einstein’s almost childlike ‘thought experiments’ challenging our everyday notions of simultaneity, soon unlocking the mysteries of space and time.  Nothing in this world is beneath your sense of wonder! Not only for those of a philosophical cast of mind, but above all for those of us inclined to look at things scientifically.

What I want you in particular to wonder about is that anyone with any intelligence and any experience of the world should regard complexity as a feature of the world at all, ever—and worse, that anyone should go around devising complex, rationalistic methods of addressing what are, once understood, trivially simple matters. 

And this is indeed something that otherwise perfectly intelligent people do all the time, and nowadays more than ever. They even sic fancy computers and even AI on all this alleged complexity, which is like siccing an American Pit Bull Terrier on a fairyfly.

Ah, yes, fairyflies. You know those teensy little flies, much smaller than the pesky tiny mosquitos known in the southern states of America and elsewhere as “No-see-’ems”?  No, I’m talking about flies much smaller—much smaller even than the tiniest of newly-hatched fruit flies, and usually too small even to see at all, many of them smaller than a poppyseed.  The smallest of the fairyflies, the Kikiki huna, at 0.15mm (150 microns) in length, is around the size of a single-celled paramecium and holds the title as the world’s smallest known flying creature. 

Now, being a philosopher, I do wonder at fairyflies on one of the rare occasions when I am lucky enough to espy one. I wonder at the ingeniousness and infinite intelligence of their design, engineering, and construction, along with their sophistication, and incredible functionality.  For I’d wager they are a trillion trillion times more advanced in design and functionality—and complexity—than anything human engineering or medical science have ever achieved or likely ever will achieve.  

Never mind the philosophy student Hamlet’s, “What a piece of work is Man!”  Heck, what a piece of work is the fairyfly!  But when you see one of these, or better, perhaps when swatting an ordinary housefly, I bet you don’t start babbling about what extraordinary complexity and ingenuity lies before you. When a housefly or an ant is pestering you while you’re eating outdoors, you just swat the blasted thing. There’s no complexity in that.

Or again, your car is incredibly complex as a piece of engineering. But if an equally complex piece of engineering runs into the back of you and dents your bumper when you stop for a red light, none of this engineering complexity exists for you at that moment—you’ve got a pretty simple situation that your respective insurance companies can haggle over.  Complexity is in part a function of our purposes, and describes the relationship between our understanding and the object of our understanding.  And in fact, complexity is not a feature, it’s a bug.

Complexity is Only Our Lack of Understanding

Complexity is not a feature of the world. 

It is not a feature of the territory, but a feature of the territory-as-mapped.  It’s a relationship between our understanding, or lack of it, on the one hand, and what we think we need to understand right-now-this-minute.  

To understand something is to understand it in terms that, for you at least, are simple.  Ronald W. Clark, in his biography of Einstein, relates this anecdote:

To [the physicist Louis] de Broglie, Einstein revealed an instinctive reason for his inability to accept the purely statistical interpretation of wave mechanics... Einstein, having a final discussion with de Broglie on the platform of the Gare du Nord in Paris, whence they had traveled from Brussels to attend the Fresnel centenary celebrations, said “that all physical theories, their mathematical expressions apart, ought to lend themselves to so simple a description ‘that even a child could understand them’.”²

When you understand something it is simple.  If you yourself do not find it simple then you do not really understand it.

The fairyfly is unimaginably complex to me, who can’t imagine how such an amazing thing with its extraordinary capabilities could be created on such a microscopic scale: 15 hundredths of a millimetre in length and it flies and eats and avoids predators and breeds?  But to God no doubt it is pretty simple—“elementary my dear Watson” in the general scheme of things, compared to human beings at least. 

As for humans:  Scott Adams quipped in Dilbert Newsletter No. 60 that, just as in silicon-chip design the designer’s signature is microscopically etched into the chip, so once we eventually completely decode all of the human genome including all the so-called “junk DNA,” we may find amongst it a signature which reads something like, "I am Kaloopah, from the star system Nebulon IV. I have sent this evolution program into space as my eighth grade science project." 

How to Handle Complexity:  the Way of Science

So in the new epistemology, and in our own approach to change and so-called complexity, rather than attempting, McKinsey-style, to match the perceived complexity of some big issue with a correspondingly big and complex intervention (so complex it needs managing), and rather than attempting to model the situation in a complex and ultimately ad hoc map made up of interconnected names of abstractions, we ourselves do something fundamentally different.

We do quite the reverse. We filter, filter, filter, thus reducing the complexity we have to deal with in a thoroughly ‘non–reductionist’ way.  This contrasting approach is illustrated dramatically by the “holiday leave” case related in “Why Creating Lasting Change Need Only Ever Take You a Few Hours… Part I.”

And this filtering is always carried out specifically in relation to the client’s specific purposes. Are we entomologists studying wasps, or picnickers being pestered by them? Are we automotive engineers or has our car just been pranged?

This is precisely the way complexity is properly handled in science. The method of science is not to model complexity, but to filter it. In science, we seek to find the key to complexity, the simple account of what are at first apparently obscure and complex miscellaneous phenomena, yielding an account which renders them transparent and readily explicable.  And we do this not by taking complexity for granted and seeking to gear ourselves up for it, but rather by questioning it.

The faith at the heart of science is that there is always unity to be found in diversity, simple pattern running throughout the apparently random. We seek to understand the complex by finding the level at which it is fundamentally simple.

In seeking to understand something not-yet-understood, I am seeking to remove some of my doubts, to clear up a mystery, to see how it all fits together. In science, the doubts, the mystery, the pieces of the jigsaw I am trying to fit together, are always specific, and not merely general; for this is the point from which we begin any serious scientific inquiry. 

I can be intrigued and even awed by the various phenomena of light, say, but unless and until I can pose specific questions—for instance, why it is that my shadow is longer in the late afternoon than it is at midday—until that point I have not yet taken a single step on the road to understanding. 

And when I at last land on the principle of The Rectilinear Propagation of Light (the law of physics that says light travels in straight lines) and then (perhaps in a diagram) show quite simply that what I’ve been struggling to explain, on the basis of this principle now has to be this way (i.e. my shadow must be longer at teatime than at lunchtime given the height of the sun and the simplest of simple, primary-school geometry), at that point I have arrived at a necessarily simple understanding of a number of related phenomena, and not only an understanding of the phenomenon of my lengthening shadow. 

For I can use that selfsame simple understanding (Rectilinear Propagation) equally to reduce the apparent complexity of a whole range of apparently complex optical phenomena, which were not at first evidently related.  

The thing to notice, though, is that filtering has taken place at a number of levels. 

First, I did not stop short at, “O the vast complexity of the countless phenomena of light!” and then seek somehow to ‘model’ that vast and bewildering complexity.  Rather, I started by asking some specific questions about selected aspects of the complexity that I did not yet understand (e.g. the question about the changing length of my shadow), and I sought to account simply, somehow, for what at first seemed puzzling (remember that only a few centuries ago there was as yet no explanation). 

And so I restricted my inquiry to what I specifically needed to know.  Far from seeking first to encapsulate the vast complexity of optical phenomena into a complex model and only then asking questions about it, I began directly with what specifically I wanted to account for. 

Second, to arrive at the principle of Rectilinear Propagation, I would have needed to consider which amongst the potentially limitless dimensions or aspects of light were relevant to the observed pattern which interested me. This too involves a filtering process. Even if I knew that light had a frequency and a wavelength and so on, I could filter all these aspects out of account—indeed I would have needed to. This second filtering procedure was mapped out in detail by Francis Bacon, the father of modern scientific method, as early as 1620 in his Novum Organum. 

I would begin by asking a series of questions designed to delineate the exact limits of the pattern, for example, questions about whether ‘stronger’ sunlight produced a shorter shadow (for indeed the sun seems stronger at midday), or whether in fact it mattered not one jot or tittle whether it was a day with ‘intense’ sunlight or a rather overcast day. 

In this way, I could filter out a large number of dimensions as irrelevant to what I was interested in. By filtering out the dimensions that are irrelevant to my particular purposes, I can drastically reduce the complexity involved without eliminating any relevant information whatsoever. 

The scientific approach to complexity is thus not to ‘model’ it, but to filter it, not once and for all purposes, but each time for each different kind of purpose, according to the specific question we are seeking to answer now.  New question, new filter.

Abstraction, Complexity, and Change

On our own approach, each major change we seek to bring about in an organization requires us to stop and expend a considerable amount of time and thought in reducing the apparent complexity, not only in order to save ourselves and our client a great deal of wasted time and effort thereafter, but so that we can deal with the real world, with its wealth of possibilities, and not with some artificially dumbed-down surrogate world.

What we have found, over some four decades of dedicated pure and applied scientific work, is that there is not a single major change even in organizational culture and behaviour, and not one major and intractable corporate problem, however complex, whose complexity cannot be successfully filtered, reduced in this way, to yield some simple, immediate actions that can achieve the whole of the desired result.

The filtering process requires us first, however, to come down from the usual airy heights of abstraction at which we normally think and manage, and to characterize situations instead in unusually concrete terms, at a fine level of detail. And this means we have to get very, very specific. 

Consider, for example, a manager talking about her organization. She may refer to such a motley collection of midlevel abstractions as, say (to choose some abstractions more or less at random): customers, orders, suppliers, production, stock holdings, sales, distribution, staff attitudes, marketing, manufacturing, planning, objectives, motivation, operations, targets, customer orientation, information technology, strategic business units, human resources, decisions, flexible working practices, organizational structure, business processes, market penetration, trade unions, appraisal, the finance function, the accounts department, inter-functional relationships, alignment . . . and so on and so on and so on. 

The thing to note about all these abstractions—and any abstractions like them—is first that they are abstractions, and second that they have no meaning whatsoever for, say, the physicist, or the biologist, or the cybernetician. The physicist has no laws referring to strategic business units per se.  Neither classical physics nor quantum physics have concerned themselves with trade unions or mission statements or customer service.  If the phenomena remain cast in such abstractions, hard science will have nothing to say about them, now or in the future. 

At the very least, we will need to begin by climbing down the ladder of abstraction to identify the specific people, places, and things (medium–sized dry goods), and the specific concrete, physical actions which are covered by the abstractions with which we set out.

As a term of art, more than four decades ago, we coined the term ‘video descriptions’—descriptions that do not go beyond what could be recorded on a video camera with a soundtrack. A video camera cannot record ‘staff attitudes’ or ‘flexibility’, ‘proactive supervision’ or ‘self–managing teams’, ‘unthinking compliance’ or ‘obstructive behaviour’ or ‘alignment’, to take just a few examples. These ‘things’ are far too abstract. We want to know what they’re called when they’re at home.

If we had the desired flexibility amongst the work force, what would we see and hear on a video recording that would be different from what we would see on a video of what happens now instead?  

We need to collect the purely objective, non-negotiable, hard facts of the concrete situation at the exceedingly low level of video description, stripped of all negotiable interpretations, free of all unwarranted assumptions however dearly–held, and minus any attributed meaning. A video description of a situation, restricted to direct, uninterpreted observables, is by definition a description all parties could agree upon, irrespective of the particular assumptions they are bringing to bear. 

Only once the facts have been collected at this exceedingly low level of abstraction, can we recast them in terms that enable them to be understood scientifically.  The point is that when, on our new understanding of complexity (the perspective of the new epistemology and Minimalist Intervention), we address problems concerning business units or the finance function or customer service or inter-functional relationships, or what have you, we invariably need to begin by unpacking these abstractions. 

We need to identify their content at a much lower level of abstraction—that of concrete, physical reality, of specific times and places, and of specific individuals who have names and addresses, personal idiosyncrasies, and patterns of behaving and of valuing that are quite unique to them as specific, individual people. We deal with these detailed, concrete, purely idiosyncratic patterns and with nothing else in order to catalyze the desired change.

The abstractions in which managerial situations have initially been framed are not typically abstractions that are readily amenable to rigorous scientific investigation and intervention conducted in a methodical, predictive way. 

Those sweeping, one-size-fits-all abstractions that are the stuff of MBA degrees and management consultants’ toolkits, and which nowadays falsely appear to make up the subject matter of decision-making in the C-suite, are not only illusions of limited usefulness, but they completely obscure the very details where all the real possibilities for creating rapid, permanent change are to be found.  And they make very simple, readily tractable matters unnecessarily intractable and complex.

And by the way, going back to Richard’s first question: 

Suppose you are trying valiantly to follow the gist of a Spanish-language podcast. It’s only going to seem to get faster and faster if you understand barely a word of Spanish yourself. If you don’t understand the world because you are looking at it through the wrong lens, a lens conferring no genuine understanding at all but only the substitute understanding of a simplistic, surrogate world, and if in consequence the world seems terribly complex to you and unfathomable, and getting more complex and unfathomable all the time, it’s also going to seem to be flying past you faster than ever before.

---

© Copyright 2022 Dr James Wilk
The moral right of the author has been asserted

---

1. I believe I owe this point, and this rhetorical way of making it, to an old lecture by Heinz von Foerster.

2. Ronald W. Clark, Einstein: His Life and Times (1972), p. 418