This was followed by a PhD which he completed in 1967; the research topic was the construction and estimation of econometric models intended to predict patterns of industrial development.

In 1968, he returned to South Africa where he taught applied economics at the University of the Witwatersrand for two years after which he moved permanently to the United Kingdom where he lived from 1970 onwards.

They became friends and colleagues and together they developed the theory of complex responsive processes of relating as a way of understanding human organisations and their management.

Leaders, policymakers and managers are bad at forecasting because it is impossible to predict the consequences of actions, a conclusion supported by the complexity sciences.

However, in moving to the second, responsive processes phase, of his work he dropped the diagram and now argues against its use because it is easily interpreted in a way that collapses the paradox of certainty and uncertainty and so sustains the dominant discourse on management while using an alternative jargon of complexity.

This phase drops the application of complex adaptive systems to organisations, arguing that it is invalid to simply apply the natural sciences to human action.

Instead, the complexity sciences are regarded as a source domain for analogies and when these analogies are transferred to the domain of human action they need to be interpreted in a manner that takes full account of the attributes of human agents, namely, that they are conscious, self-conscious, emotional, often spontaneous, often thoughtful and reflective beings who have some degree of choice over what they do.

Human agents are basically interdependent, they respond to each other and their choices and intentions play into each other producing unpredictable, emergent patterns over time.