Paradigm

A frame of reference which makes reasonably good sense of a complex subject, and enables us to think about it, to work with it and to get results which we can at least in part predict, and which we may be able to apply in useful ways. The word “paradigm” has been around some six centuries, in the sense of a story—an exemplar—retold many times to illustrate an argument or a group of ideas with something in common. It became a crucial idea in modern thought when, in 1962, the philosopher of science, Thomas Kuhn, published his book The Structure of Scientific Revolutions. in the context of the progress of scientific discoveries.^P2

The idea is important to Lean Logic for two reasons. It tells us something about how we as thinking people handle ideas which are too complex to work out from first principles. And it contains within it the idea of radical change: kaikaku is one of the key principles of lean thinking, and is central to our future at a time when circumstances of climate, economics and society require revolutions in our thinking and in our assumptions.

Let’s start with the first of those. When a complex subject is taught, there has to be an agreed frame of reference for it, and this forms the foundation of the profession consisting of people who are competent in it. But, of course, there is a problem here. The frame of reference that is being taught is, in many ways, certain to be wrong. It will be better than nothing, but it will contain anomalies and maybe even paradoxes—signs that there are mistakes in there. In some cases, the people working in the field are aware of them but have not worked out how to correct them; in other cases they suppose things to be true that aren’t: they don’t know what they don’t know. But they teach it as the truth, all the same. As Kuhn summarises, this all adds up to “a strenuous and devoted attempt to force nature into the conceptual boxes supplied by professional education . . . into the preformed and relatively inflexible box that the paradigm supplies”.^P3

So there is a sense here of making do with what we have, despite the imperfections, and the reward is a discipline—a story-so-far—which enables new entrants to learn quickly and to do useful work. And there is a stable basis from which specialists can then do the detailed research which Kuhn rather disparagingly calls “mopping up operations”, but which enable them “to investigate some part of nature in a detail and depth that would otherwise be unimaginable”.^P4

There follows a long-lasting phase of detailed study, or “normal science”—a period of steady progress—but this normality can be expected in due course to crack as it reveals more and more contradictions between the paradigm and actual observation. Scientists respond to these problems in a host of ways: they look again; they check their equipment; they develop what became known later as “auxiliary hypotheses” which solve—or explain away—the problem while leaving the paradigm itself intact. And yet, despite all these efforts to protect the paradigm, eventually the crisis is reached when there is no getting away from the horrible truth that it is significantly wrong, and the tension builds until something comes along to take its place. What follows is what Kuhn calls a “scientific revolution”.^P5

Those are tough and exciting times for people working in the field. The most-established experts who have been most successful in filling the gaps in the previous paradigm and polishing it up have most to lose, because much of their previous work now turns out to be irrelevant (Galley Skills). There is resistance; there are breakthroughs to be made, often by those who are relatively new to the field; there are difficulties in making the new, paradigm-busting work visible to those scientists who built their careers and mindsets around the paradigm that is now under attack. At the end of this storm and stress, you get a new paradigm. The profession converges on a new way of thinking, and there is a whole new supply of uncertainties and interesting questions to keep those in the field busy for a long time to come.

But if it’s wrong why put up with it?

Kuhn’s story of the progress of science did not meet with complete approval. One critic was Karl Popper, the great champion of the open society, scourge of ideology and defender of language from the unctuous, smiling emptiness which misshaped the twentieth century. As he constantly pointed out, evidence of being wrong is the only incontrovertible information we ever get about whether a universal hypothesis is right or not. It is like the “black swan” story: a theory that all swans are white is absolutely proved to be wrong by a single sighting of a black swan; but it is not proved to be right by a thousand sightings of white swans. That, for Popper, meant that when evidence that falsifies a theory comes along, it should be grasped with both hands, and the theory should be modified accordingly.^P6

An argument ensued. Popper’s central insight about falsifiability was beyond challenge. But sometimes it is more complicated than that, and here are two complications:

1. No evident alternative, yet. There is no case for rubbishing a hypothesis, even if it is not true, if at the same time it is useful and you have no idea what to put in its place.

2. More information needed. If you reinforce one hypothesis which isn’t fulfilling its promise (e.g., “Faith can move mountains”) with another one (e.g., “Clearly you don’t have enough faith”), there is of course the near-certainty of sinking out of sight into nonsense, and one of Popper’s most strongly-expressed arguments was about the need to avoid adding auxiliary hypotheses to the point that the statement could not possibly be proved to be untrue. And yet, the addition of an auxiliary hypothesis may be necessary: “Potatoes are perfectly safe to eat” is a dubious statement unless it comes with another statement, “You must cook them properly first”. That is, developing the substance of a hypothesis with supplementary information which sets out the circumstances in which it appears not to be false is an entirely reasonable thing to do.

There are other ways of protecting a paradigm (see “Digging In” sidebar), but it was with these two complications surrounding falsifiability that the Kuhn-Popper debate was concerned. Their significance, under fair conditions of scientific debate, is real, and this was recognised by another member of the cast in this story: the Hungarian philosopher and mathematician, Imre Lakatos. In retrospect, his contribution seems to be obvious, and it is a modification, rather than a contradiction, of what Kuhn wrote in the first place: don’t try to defend a whole paradigm, he argued; instead, decide, explicitly or implicitly, on its “hard core”—its central identifying principles. Around the periphery, there are invariably many things to be sorted out, and they may be substantial, but this can be done while leaving the hard core intact. As Lakatos pointed out, “All theories are anomaly-laden”; so scientists should be prepared to add revisions and auxiliary hypotheses as research and experience suggest.^P7

Eventually the hard core itself may crash, but if it is sufficiently small, it may endure for a long time or even indefinitely, for the less it says, the less likely it is that part of it is untrue. And that hard core provides the foundation for ongoing science—its collapse and replacement by a radical new research programme remains a possibility, but scientists searching for the piercing new insights which could bring such fundamental change will not find them unless they plod on within the existing framework until they hit an obstacle. You only realise that a door is locked when you try to open it. So, although the resilience of the hard core is not guaranteed, at least its expectation of life is longer than that of a whole paradigm; and there is less to lose, with only a relatively small set of principles at the core, which even Popper might accept, for now, as beyond refutation.

And yet, however comprehensive or lite the paradigm might be, the pattern of advance is essentially the same. It is like a river flowing gently down a cascade of rocks. The water comes to quite still pools, each with a lot of river-life going on in it, and it may stay there for some time, before tipping over the edge through turbulence to the next pool, large or small. The waters in the higher pools will have no knowledge of the lower ones; those in lower pools will have an existential memory of the higher ones, but will know they can’t be there any more: they are in a different ethic, a different paradigm. Some of their hard core is likely to survive, but who knows? Some shocks are extraordinary (Systems Thinking > Form > The Power Law).

The herringbone

Another way of visualising that pattern of rapid, turbulent change, followed by long periods of quiet, is Arthur Koestler’s model of the herringbone. Each line and arrow represents reasonably steady advance within the framework of one paradigm (or hard core); each change in direction represents the shock of a “scientific revolution”, followed by steady advance in the new direction/paradigm. Although there is no knowing how long this will last, it will undoubtedly be for much longer than the period of shock. And each shock undoes at least some of the careful construction that has happened under the protection of the preceding paradigm (each new line does not start from the tip of the old one), so it will, for understandable reasons, be resisted or postponed for as long as possible. In some cases, it might be quite minor—a setback; in other cases, it could be so great that the whole of the achievement of the current arrow-paradigm is lost. An extreme shock could ripple through some or all of the arrows representing the shocks and recoveries of the past.^P8

Importantly, such a pattern of progress is characteristic of the complex systems on which we depend, as we see with:

• Kaizen and kaikaku. The incremental improvement of kaizen is interrupted and set off in a new direction by the shock of kaikaku. This principle is at the heart of lean thinking.

• Resilience. To protect the resilience of a system there is a case for severely limiting elaboration (i.e., preventing excessive advance along the arrow). This may be expected to postpone the shock, to reduce its scale and the damage it causes when eventually it comes, and to aid recovery.

• Punctuated equilibrium in evolution. There is evidence of very uneven rates of progress in the evolution of new species and diversity, ranging from massive and rapid change to long periods during which the ecosystem and its resident species are fortunate enough to live in not particularly interesting times.^P9

• Gaia and Medea. This is the principle that the ecology (Gaia), so long as it is not damaged by gross interventions, will maintain itself and its stability for a long period, but not indefinitely. When the shock (Medea) comes, it will cross the threshold into a different form; this is the crisis which Gaia cannot prevent, and it will be followed by a new stasis and equilibrium, maintained by a new Gaia whose standards and values are utterly different from those of its predecessor.

You might think, on the basis of the above, that whatever we do—as members of a large complex system which we cannot control—there is no way of avoiding the paradigm shift which will finally destroy everything we have made and depend on. You might be right. But you might not—at least, not for a long time—because, as the presence of “resilience” in the list above suggests, there are things we can do to contain and limit the shock. We can, for instance, make it happen ourselves and on a small scale, on the principle of the controlled explosion; it is not beyond our wit, though it may be beyond our inclination, to make sensible judgments (Wheel of Life).

—

But the world which these judgments inhabit may be profoundly different from the one that Kuhn, Popper and Lakatos were debating. Kuhn describes a sequence of deep breakthroughs—Copernicus’ demonstration that the sun is at the centre of our solar system; Newton’s laws of optics and motion; Priestley and Lavoisier’s discovery of oxygen. But this paradigm-to-paradigm progression may have been of its time, characteristic of immature science, not (or not often) to be repeated. Maybe, in the case of a mature, large-scale science—whose many parts, once thought to be discrete, converge towards a single subject—the advances all take place around the edges. The larger the system, or the science, the longer its edge, and a long edge means plenty of space for locally-revolutionary changes to take place which have a big effect in their immediate vicinity, leaving the bulk of the subject unchanged.

If that is the case, then two implications come to mind. First, we may have a convergence between Kuhn’s large-scale paradigm—in the shape of modern science, its many strands persisting in essentially the same direction—and Lakatos’ smaller-scale hard core, in which, at the edges, there is also continuous movement and revision of the ever-expanding scientific universe. Maybe, in a mature scientific culture, we should not be looking for scientific revolutions, but for revisions of the detail. The big paradigm shifts, kaikaku moments, may be characteristic of sciences and social movements in the middle years of their evolution. The Transition movement, for instance, could perhaps reach that stage around 2015, when it has advanced far enough to learn the basics, but has hit a wall in terms of further developments. That is the point at which a breakthrough, a long time coming, and much resisted, could quite suddenly free the movement to move fast in a new direction.

And the second implication, in the strange way which is characteristic of this complex subject, seems both to confirm the first and to contradict it. While the scientific culture and establishment of our time is very sound, and while we should not be looking for revolutions except round the edges, it is also true that there is, even now, a profound stand-off between two scientific cultures. One looks for complex, big-science responses to our present condition—such as the convergence of genetic engineering with robotics, information technology and nanotechnology into a global agricultural establishment, and super-projects such as giant dams and high-speed rail links. The other looks for simpler, small-science responses in the convergence of observation, local conditions and the experience in the minds of people, with perceptive husbandry, local water management and functioning local rail systems, and with disagreement between local paradigms being welcomed as diversity; as an essential property of the planet’s future, if it is to have one.

The stand-off between these two paradigms is as profound as anything that Copernicus, Newton, Priestley and Lavoisier were part of. If big science were to “surrender” to small science, it would need, not just to retreat back down one of the arrows of the herringbone, but to climb down much of its length. It isn’t going to do that.

But supposing we change it to they? That means taking note of the System Scale Rule and recognising the essential need for local, grounded, rooted responses in all their diversity, applying modern science where local conditions call for it, and applying scientific brilliance and observation in local conditions. “They” means scientists; it may also mean applied technologies such as microchips, materials science, and the application of informed biology to soil fertility, including such critical problems as the conservation of phosphates at a time when it is not being imported any more. “They” could bring us science’s latest discovery: modesty—that is, science as a service to the people and to the ecology.

This is listening science, with an ethic of pull, a flexible call-and-response science that provides assistance when it is asked for, brings great competence and observation to small places and does not, with today’s large-scale certainties, kill off the local imagination needed to make a future. It will no doubt require a kaikaku moment to get us into that new paradigm—but, maybe, once we’re there, we will have stability: the requirement of shock after shock will no longer be built into the system.

Local resilience and scientific intelligence are going to need each other: there is nothing inconsistent there. The inconsistency lies between local resilience and ideology. And large-scale science is degenerating into large-scale ideology. Small-scale science applies the best that science can provide without imposing a whole paradigm; it is flexible enough to work within whichever paradigm is right for a particular place and time.

The word for this process of being guided by, and learning from, the paradigm you are in is “heuristics”. The possibility of being guided by a constructive heuristic means that we are not simply swept along the river like a cork; we can influence where we go, at least to some extent. Above all, we need to have a sureness of decision about which paradigm we want to be in. If, as Lakatos argued, we are in a “progressive” paradigm, we will have a positive heuristic, and may make some good decisions; if we are in a “degenerate” one, we won’t. And to give ourselves a chance of being in a progressive paradigm, we need both to think and to observe—that is, as the philosopher Alan Musgrave puts it, to sustain “a subtle dialogue between the theoretician (armed with his heuristic) and the experimenter (armed with the ‘facts’)”.^P10

The resilience of any large system requires the rise and fall of its smaller parts, on their different timescales—subcycles for subsystems. In terms of science and its paradigms, that means acknowledging the detailed local knowledge and small-scale research programmes which can succeed and fail, change and respond without denial or delay; and without making waves big enough to destroy everything else.

Lean Logic’s hard core consists of sustained and accurate alertness to the ecosystem and to its demands for change. That is the heuristic which must be obeyed, for postponement serves to deepen crisis from a series of small new beginnings to a large-scale ending.

Related entries:

Unfalsifiability, Genetic Fallacy, Humility, Harmless Lunatic, Ethics.