W Ross Ashby was a psychiatrist who, through books such as Design for a brain, was one of the pioneers of the development of systems theory (qv) in the 1950s. A particular branch of systems theory was ‘cybernetics’ – from the Greek ‘steering’ – essentially the theory of the ‘control’ of systems. This was, and I assume is, very much a part of systems engineering and it attracted mathematicians such as Norbert Weiner. For me, an enduring contribution was ‘Ashby’s Law of Requisite Variety’ which is simple in concept and anticipates much of what we now call complexity science. ‘Variety’ is a measure of the complexity of a system and is formally defined as the number of possible ‘states’ of a system of interest. A coin to be tossed has two possible states – heads or tails; a machine can have billions. Suppose some system of interest has to be controlled – for simplicity – a machine, a robot say. Then the law of requisite variety asserts that the control system must have at least the same variety as the machine it is trying to control. This is intuitively obvious since any state of the machine must be matched in some way by a state of the control unit – it needs an ‘if ….. then ……’ mechanism. Suppose now that the system of interest is a country and the control system is its government. It is again intuitively obvious that the government does not have the ‘variety’ of the country and so its degree of control is limited. Suppose, further, that the government of a country is a dictatorship and wants a high degree of control. This can only be achieved by reducing the ‘variety’ of the country through a system of rules. This can then be seen as underpinning the argument for devolution from ‘centre’ to ‘local’ – a way of building ‘variety’ into governance. So we begin to see how a concept which appears rooted in engineering can be applied more widely.
We can now take a bigger step and apply it to ‘knowledge’, and, specifically, to the knowledge required to make progress with a research problem. The ‘problem’ is now associated with a ‘system of interest’ and the ‘requisite knowledge’ is that which is required to ‘solve’ the research problem. The application of the law of requisite variety can then be interpreted as relating to the specification of the toolkit of knowledge elements needed to make progress and the law asserts that it must be at least as complex as the problem. We might think of this as the ‘RK-toolkit’. It seems to me that this is an important route into thinking about how to do research. What do I need to know? What do I need in my toolkit? It forces an interdisciplinary perspective at the outset.
Consider, as an example, the housing problem in the UK (cf. Real challenges): what is the requisite knowledge – the RK toolkit – which would be the basis for shifting from building the current 100,000 new houses p.a. in the UK to an estimated ‘need’ of 200,000+ p.a.? We can get a clue from How to start: there will be policy, design and analysis elements of the toolkit. Elementary economics will tell us that builders will only build if the products can be sold, and in turn this means can be afforded – basic supply and demand. Much of the price is determined by the price of land – so land economics is important; or if land price is too high for elements of need to be met, there may be an argument for Government subsidies to generate social housing. Alternatively, prices could be influenced by the cost of building and this raises questions of whether new technology could help – and this brings engineering (and international experience) very much into the toolkit. Given that there is likely to be a substantial expansion, geography kick in: where can this number of houses be built? This is in part a question of ‘where across the UK’ and in part, ‘where in, or on the periphery of, particular cities’. Or new ‘garden cities’? All of this raises questions for the planning profession. The builders are part of a wider ecosystem and land owners and the government regulation of land, through taxation or whatever become part of the research task.
So there are challenges for all of us who might want to work on this issue – academia, divided by discipline; the professions, functioning in silos; the land owners, developers and builders; and government, wanting to make progress but finding it difficult to coral the different groups into an effective unit. In this case, the RK-toolkit can be assembled as a knowledge base, but this sketch shows that an important part of it is a capacity to assemble the right teams.