“Evolution is cleverer than you are”Orgel’s Second Rule
In 1972, around the time I was giving up piano lessons, the evolutionary biologists Stephen Jay Gould and Niles Eldridge published a landmark paper, ‘Punctuated Equilibria: an Alternative to Phyletic Gradualism’. They argued that the fossil record didn’t show a steady and continuous process of evolution, but one characterised by long periods of relative stasis punctuated by rapid speciation events. They were mainly interested in the the long bits where nothing happened, but twenty years later, in ‘Speciational Evolution or Punctuated Equilibria’, Ernst Mayr took the view that it was the ‘morphological discontinuities’ – the big, sudden changes – that were more curious.
The issue of whether or not the dinosaurs were almost wiped out by a massive asteroid impact may therefore be beside the point. What the fossil record seems to show is a whole raft of speciation and extinction events with no obvious exogenous causes. In the late Permian, for example, 96% of all marine species and 70% of those on land vanished almost overnight, apparently without warning, and with as yet no convincing explanation.
All this may have remained a curiosity, fascinating no doubt to palaeontologists but of little relevance to business professionals, had it not been for an ontological turn which evolutionary biology was to take towards the end of the 20th Century. Researchers such as Richard Dawkins and Daniel Dennett were looking to abstract the mechanisms of evolution; to show how biology was effectively following what Dennett characterised as a kind of algorithm. What emerged was an explanatory theory of evolution that was ‘substrate neutral’. Its logical structure comprised a design space, schemata, interactors, modules, environment, constraints, competition and fitness landscape. It’s recursive algorithm provided for variation, replication and selection as it searched this landscape for fit ‘designs’. Computer simulations built around these rules were to exhibit emergent behaviour that closely resembled biological evolution. Biology slotted neatly into this theoretical framework, because that’s what it was for, but in principle the algorithm could provide insight into the behaviour of other complex adaptive systems. Such as economies.
At the same time as evolutionary biologists were driving their body of theory in the direction of greater generality, complexity theorists were deriving results that would in return shine a light on the problem of punctuated equilibrium. In particular, work by Duncan Watts and Stephen Strogatz on ‘small world networks’ had shown a tendency for some populations to self-organise into webs of densely connected clusters with sparse connections between them (see our article Small World Networks, together with Watts’ Six Degrees: the Science of a Connected Age and Strogatz’s Sync: the Emerging Science of Spontaneous Order). For many purposes a network like this can be very efficient – maximising connectivity across the network while keeping the total number of connections in check. However, Sanjay Jain and Sandeep Krishna at the Indian Institute of Science have fingered exactly these emergent structures as being the likely culprits behind punctuated equilibria in biological ecosystems (see ‘Large Extinctions in an Evolutionary Model: The Role of Innovation and Keystone Species’).
In a small world network, some nodes represent ‘super-connectors’, with disproportionate importance in reducing average path length across the web. The equivalent in an evolving biological ecosystem is a ‘keystone species’. Jain and Krishna noticed that a simulated ecosystem would generally be resilient to random changes, including loss of species, but every now and again a variation would hit a keystone species, propagate throughout the network, and bring the whole thing crashing down.
Early 20th Century attempts to transfer the idea of evolution by natural selection outside the field of biology generally yielded results that ranged from disappointing to horrifying. In part this was because the theoretical structure of evolution wasn’t sufficiently well understood to provide such a basis – it was more of a loose metaphor. But in some quarters it also acquired a wholly specious ‘moral’ dimension that equated ‘fit’ with ‘good’. Hence Eugenics. What the likes of Dennett had done was different. They’d established a robust theoretical framework into which biological evolution could slot. But the theory was substrate neutral and, in principle, could be applied to other, less squishy organisms.
One such substrate may be technology. In practice, technologies exist not in isolation but in a web of economic and technical interdependencies that look a lot like small world networks. There are also the equivalents of keystone species. Rebecca Henderson and Kim Clarke distinguished between modular and architectural technologies in ‘Architectural Innovation: The Reconfiguration of Existing Product Technologies and the Failure of Established Firms’. Architectures in particular can have the capacity to propagate waves of disruption through big chunks of the technology ecosystem, subjecting it to its own forms of punctuated equilibria.
In The Origin of Wealth, Eric Beinhocker sets out a measured argument for the application of the substrate-neutral evolution algorithm to physical and social technologies distributed about a ‘fitness landscape’. The idea of a ‘roughly correlated’ fitness landscape is fundamental to evolutionary systems. The ‘coordinates’ of a position on the landscape represent a particular possible design (in the universe of all possible designs), and the ‘height’ of the landscape at that point represents ‘fitness’. In the case of technologies, Beinhocker identifies fitness with the degree of acceptance and imitation of a design.
The contours on the fitness landscape look a bit like the Lake District, but without the rain. If you move a little from where you are (tweaking an existing design, in other words), you might move up or down slightly in fitness terms. If you cast out further, with a more radical innovation, you might find a new mountain. To complicate matters, the hierarchical architectural/modular nature of technological designs means that ‘distance’ on the fitness landscape is a function both of the extent of change and of the number of variables which are changed. By their very nature, architectural innovations tend to spur change on more dimensions than modular ones.
Technological history shares common features with other evolutionary systems, such as cascades where one innovation triggers a raft of others, which bear comparison with what Giovanni Dosi identified as ‘technological trajectories’, as well as punctuated equilibria (see our article The Toilet Test). The evolutionary model also helps to put in context well known phenomena, such as product life cycles (‘S-curves’), disruptive innovation, and cooperation strategies reflected in Nash equilibria and positive sum games.
In particular, it’s not surprising that an established organisation would find it hard to make long jumps on the fitness landscape when they’re already at or near a local peak. If you’re already on high ground, an incremental movement might take you higher, in which case you’ll keep going, or a little lower, in which case you’ll change tack. But if you cast out in a radically different direction – taking a big jump somewhere else – you might end up, unpredictably, on an upslope or in a deep valley. And the odds favour the latter.
Beinhocker goes a step further than this though. Characterising a business as a collection of physical and social technologies ‘glued together by strategy’, he constructs an evolutionary model of commerce based on a fitness landscape selecting amongst the universe of ‘business plans’. Economic evolution is thereby painted as a product of three distinct but closely interlinked and ‘co-evolving’ design spaces: physical technology, social technology and business plan.
In biology the unit of evolutionary selection is the gene, a kind of schema involved in encoding certain heritable traits. In the corporate jungle the equivalent schemata are the intrinsic elements of a business plan (which might be tacit or explicit) that can provide “a basis for differential selection between businesses in a competitive environment”.
The whole process works like this:
- What Joel Mokyr dubs a ‘superfecundity’ of business plans – meaning an over-abundance on which selection can work its magic – are differentiated through the ‘deductive tinkering’ of agents.
- Business plan schemata use strategies to bind physical and social technologies together.
- The equivalent of gene transcription is implementation of business plans by ‘readers’ – management teams in this case.
- Selection occurs at several levels – individual minds, problem-solving groups, organisational hierarchies and, ultimately, the market.
- Successful business plan features are replicated and rewarded with influence over more resources – by businesses, customers and financial markets.
The value of markets as a resource allocation mechanism in this model isn’t quite the same as the ‘efficiency’ claimed by traditional economics, but as an effective distributed processing algorithm that gets ‘the right signals to the right people’.
The evolutionary model of complexity economics can provide a useful perspective through which to visualise and communicate about change, but there are also a number of potential practical consequences for business professionals.
Most obviously, it’s likely to make sense to locate decision-making close to the market itself. More radically, it may be effective to trial a raft of experimental business plans (‘real options’) before making commitment decisions. The evolutionary perspective also implies that the main purpose of strategic planning exercises is not to create some kind of specious multi-year budget (see our article There’s No Such Thing as Long Term Strategy), but to create ‘prepared minds’ within the organisation. Problem solving approaches need to recognise the intrinsic complexity of strategic decisions (see our position paper, Strategy as a Complex Systems Problem). Finally, a critical priority of C-Suite teams should be to create an effective ‘selection environment’ within the organisation, though such mechanisms as vision/mission/values, culture, reward mechanisms, capital allocation processes, and so on.