Thomas KUHN’s “The structure of scientific revolutions” has been hailed as “a brilliant original analysis of the nature, causes, and consequences of revolutions in basic scientific concepts.” I never took to this book, or its “big findings.” To me it is the work of an academic hedgehog: learned, ponderous, and…very highbrow. Here is my simplistic assessment.
- · First: that self-serving scientific elites sustain a scientific paradigm beyond its “use by date” is no surprise to me. Having a choice between status and truth, elites naturally will chose the former. The derisive term “academic” precisely describes the drifting of debate into erudite and polite pointlessness.
- · Second: that the discovery of a new scientific paradigm is the result of trial and error and of inspired guesses, rather than painstaking rational sifting of the evidence is not a great novelty either. The outcome of such recursive guessing is not a linear line of descent – the Great Chain of Ideas – but the rather haphazard “descent with modification.” Positing a Darwinian approach to scientific pursuits is a fundamental albeit small step forward: the real advance would have been developing the fullness of the metaphor.
- · Third: the issue of “incommensurability” is (in my admittedly ignorant view) an epistemological red herring piggy-backing on the Darwinian approach. In the material sciences, Popperian falsification dooms the “old” paradigm in the end, never to return. Copernicus falsified Ptolemy, and Einstein (to a degree) falsified Newton. If “incommensurability” truly existed, a reversion to Ptolemy would be conceivable. It is not. As long as falsification is the arbiter, there is accumulation of knowledge.
Kuhn’s analysis is deep – in a hedgehog sense. References to this work abound. Academic debates (predictably) raged for decades. Refutations slammed into reputations, relentlessly losing themselves on the shores of irrelevance. Widely used as a source of authority and quoted accordingly, the book had little impact on the practice of science or the formulation of science policies (if Kuhn’s insights shaped innovation policies or intellectual property laws, let me know).
Steven JOHNSON is a fox. He has no academic pretensions, or affiliations; his writing is lowbrow. His discussion of “Where good ideas come from” draws together material from much of the biological and social sciences to propose a new synthesis. I’m sure the work can be faulted on many points of detail (I can visualize the merciless cat-o-nine-tails-of-superficiality striking the author’s back). In Johnson’s view, ideas emerge from networks. As people and cultures interconnect, ideas emerge spontaneously from interaction among actors in the network and their environment. The market place, the meeting room in the labor, the coffee-houses, the city: all these small and large networks nurture ideas. The “solitary genius” is the exception.
In describing how good ideas emerge, Johnson’s closing metaphor is the “reef” – a platform which innumerable small and larger living beings use in infinity numbers of ways not envisaged by the reef’s originators. The reef is platforms within platforms, or platforms on top of platforms, where each platform takes the underlying one as “given” and uses it. Platforms are the perfect enablers. So are ideas: enabled and enabling at the same time, so that their origins get lost in the fuzziness of the network and its environment.
A liminary remark
Structures are bounded by their design – they are closed with respect to the surrounding context. Its constituent elements are endogenous, predetermined, and patent. Structures do not contain latent or emergent properties. Whether large or small, structures follow the same rules – the structures and rules do not depend on scale. Structures tend to be, or become, complicated.
Networks are open. The number of participants varies, and this affects the functioning of the network. The actors also interact with their environment. The links between agents, but also agents and the environment, are contingent: they vary in accordance with circumstances. The links themselves are simple – yet emergent properties – complexity – and sometimes singularities are the (surprising and unexpected) outcome. Networks may contain unforeseeable latencies. Networks change properties depending on scale: yet, they tend to self-organize in predictable patterns.
Where do good ideas come from?
For Kuhn, good ideas are endogenous to the mind – the work of an autonomous individual. Kuhn focuses on just one method for the emergence of ideas – the scientist’s inspired guess or ordering conjecture – and then follows the fate of scientific ideas as they weave their way through scientific circles.
Ideas, says Johnson, are the result of networks. From this insight, the whole book flows. If we look closely, networks are everywhere. Our neurons are networks, and so are our memories. Small and larger groups connect in social networks. Networks are open, hence they are shaped by the environment (and networks in turn shape the environment). Networks are not just jumbles: they have shared properties and patterns recurring time and again – at different scales. The properties tell us the conditions in which ideas emerge. He has identified seven: as en encouragement to engage with the main part of his book, here the main heads of chapter:
- · Adjacent possible: at any one time, possible new ideas lie just outside the boundary of what we have experienced and known. The adjacent possible is a kind of shadowy future, hovering on the edges of the present state of things, a map of all the ways in which the present can reinvent itself. Ideas are not intellectual meteorites coming from the inner space of the scientist/inventor, but grow from precursors (unless they are imported from afar, adopted, absorbed, and adapted). The trick is to figure out how to explore the surrounding edges of possibility. A gifted mind can connect detritus of old ideas and tinker with them, until something new emerges.
- · Liquid networks: In a solid state network, where all connections are preset, everything remains unchanged and no novelty can emerge. New connections require a “liquid” network, where experimentation and tinkering with new concepts is possible and (a) the full scope of the adjacent possible is explored and (b) the results retained. We need information spill-over. Immersed in a liquid network – historically the city has been the best example – individuals “get smarter”, because they are part of such a protean network. Modern avatars are meeting-rooms, corridors, water fountains that ease the flow and counter-flow of ideas. Often, it is not just a matter of exchanging, but of testing ideas against the intellectual environment – it matures the ideas by instilling confidence.
- · Slow hunch: Ideas often emerge as hunches, and need slow, repeated cultivation, dormancy, and re-elaboration in order to mature. It is more akin to a process of accretion and nurturing than linear solving of one problem after another. The environment will
- · Serendipity: inchoate ideas and hunches often mature when surprising new and often accidental connections are forged. This happens, e.g. during sleep, but also occurs in a social setting, like the meeting room, or the coffee house.
- · Error: being right keeps one in place, being wrong forces re-examination and exploration. Good ideas are more likely to emerge in an environment containing a certain amount of noise and error, for they dispel complacency and self-affirmation. In a “normal” setting, error tends to be dismissed (often rightly so). Transforming error into insight is the key to innovation.
- · Exaptation:rather than tinkering with existing concepts, a mature concept is borrowed from an entirely different field. In the field of ideas, concepts from one domain may migrate to another as a a kind of structuring metaphor – “putting the received wisdom on its head.”
- · Platforms: platforms are structures or ideas that enable other networks or ideas – the reef is a very good biological example. Twitter is an example in the cultural sphere. The key element is the add-on character of the platform: a clean interface allows the user to concentrate on his application, while ignoring the functioning of the platform. The result is a complex structure where each one understands a section, but no one designed the whole, or understands it.
Johnson’s close observation of the process leading to ideas allows him to observe that these patterns are scalar. The same patterns are found at the personal, the group, and social level (with different nestled structures therein, going from the firm to the whole society), but in different contexts. As context changes, so do policies. Degrees of freedom emerge one did not expect or imagine – the categorical one-size-fits -all prescription no longer obtains.
Johnson’s next step is to look at the historical record – he takes the long view – in order to verify his metaphor of idea as network. This (last) chapter to me is the most interesting, for it proves what I’ve known anecdotally. As one observes history from the distant past to current developments, individuals and the markets tend to yield to non-profit social networks. Today, ideas predominantly emerge in the social commons, where non-market criteria prevail.
From Johnson: in the 4th quadrant good ideas that where discovered or invented in a “non-market” framework, e.g. public research
What Johnson presents is a silent transformation in the process of creating new ideas. Viewed in this fashion, one can tailor policies to each specific context.
As good ideas increasingly seem to emerge in the commons, the conventional view that the commons have to be privatized in order to favor individual initiative – the enclosure of the commons – might be counterproductive. Enclosure runs against the very idea of network. In fact, companies favoring enclosure of their innovations increasingly find themselves isolated from the down-stream of innovation.
Johnson’s analysis immediately leads to a policy analysis and prescription. When enclosure no longer works, because it disrupts the networks, the alternative is to acknowledge the commons and manage them appropriately. Commons can be managed collectively and the “tragedy of the commons” is cheap ideology.
As an example: bounties have been a traditional method to divide up the returns from the commons. Patents were a strange form of bounty, in which a time-limited monopoly was granted (the government could not afford to pay the bounty). Note that it is a grant, not a right. The transformation of a bounty – a socially determined reward – into an unbounded exclusionary right is one of the great ideological feats of the last three hundred years. It was primitive accumulation on a grand scale. The rule might hamper future innovation by trying to encase a network into a structure. It is time to study innovation policies through the lens of networks.
 I’m quoting here the blurb on the cover of my 1968 edition.
 At that time it might have shattered the hagiographic self-portrait scientists propagated – the view of giants standing on the shoulders of giants in infinite succession. Insiders, however, knew differently. WATSON shattered the genre in 1968 (See James D. WATSON (1968): The double helix. Penguin, New York). Arthur KOESTLER, no academic insider, was denting the carapace from the outside (see: Arthur KOESTLER (1966): The act of creation. Pan Books, London). The mathematical formulation of the inspired guess – Bayes’ rule – had been in existence for over 200 years and gained recognition in Operations Research during WWII.
 The incommensurability of the “new” with the “old” paradigm in the short run is a trivial issue. Structural incommensurability is found in biological “descent with modification:” it is impossible to compare species as they emerge and establish a Great Chain of Being” culminating with man. Ideas, however, do get better (mostly).
 Steven JOHNSON (2010): Where good ideas come from. The seven patters of innovation. Penguin, London.
 The conventional topos of the autonomous scientist or innovator results from the telescoping of a slow process of interactions involving many encounters among actors in the network.
 “Command and control” is a rule applying (in theory) to a whole army as well as a platoon.
 Interestingly, he speaks of the “scientific community”; he implies that scientific ideas are bounded within the community. He does not seem to consider the possibility that members other than the community may be doing science.
 We often think of ideas as “structures”. This is just availability bias. Verbalization implies structure, but this is a requirement of language, not of the idea. In fact, most ideas are transmitted by “imitation” or “learning by doing” where experience leads to internalizing an idea. Experience is internalized emotionally for subconscious application.
 Darwin argued cogently “natura non facit saltum”. This applies to ideas as well. Another way to look at it is the recognition that astronomers have long debated whether the sun or the earth was at the center of our planetary system. The Copernican revolution, but not the general theory of relativity was the adjacent possible.
 The term has been first used in evolutionary biology. Feathers, which may have originally been a way to keep warm, turn out to find use for flight.
 See e.g. Clifford D. CONNER (2005): A people’s history of science. Miners, midwives, and “low mechanics”. Nation Books, New York.
 See Elinor OSTROM (1990): Governing the commons. The evolution of institutions for collective action. Cambridge University Press, Cambridge.
 Article 1, Section 8, Clause 8 of the US Constitution: “To promote the Progress of Science and useful Arts, by securing for limited Times to Authors and Inventors the exclusive Right to their respective Writings and Discoveries.” Its character as grant, rather than right, is patent (sorry for the pun).
 See e.g. Michael PERELMAN (2002) : steal this idea. Intellectual property rights and the corporate confiscation of creativity. Palgrave, New York.