The Field Guide › Paper
The founding textbook of cybernetics, and the source of the one idea from that tradition that organisations most need: the Law of Requisite Variety. Ashby's formulation is that only variety in the regulator can force down the variety produced by disturbance, so a system can be controlled only to the extent that the controller commands at least as many distinct responses as the situation can present states. The book builds this carefully: a formal account of machines, feedback and stability, then variety as a measurable quantity, then regulation itself, including the case Ashby calls the very large system, the one too big to specify fully and which must therefore be regulated without ever being wholly known. The opening chapter is the surprise for a 1956 text, arguing that science had largely confined itself to systems that are simple or reducible, that the dogma of varying one factor at a time is fundamentally impossible where altering one factor immediately alters others, and that cybernetics exists to take complexity seriously as a subject in its own right. For this map the value is that requisite variety puts a formal floor under arguments usually made on moral grounds: a narrow and homogeneous group of decision-makers cannot regulate a varied organisation, not because that is unfair but because it is arithmetically impossible, and an organisation whose only sanctioned channel is the leader has a regulatory variety of one. Voice, dissent and difference become capacity rather than courtesy. The limits are worth stating plainly. This is a book of mathematics and biology rather than management; Ashby's examples are thermostats, organisms and Markov chains, and every organisational reading is downstream interpretation rather than his own claim. It is also the point where the tradition is most often flattened, since variety is a precise term (the number of distinguishable states available) whose popular management use drifts into a loose synonym for diversity of background, which sounds similar and argues something quite different.