enterprise-architecture-three-worlds1

Enterprise Architecture’s Three Worlds

Enterprise Architecture (EA), as a discipline, is about creating models of the enterprise that may be used, normatively or prescriptively, to guide or inform decision-making processes in an enterprise. The models that EA activity produces capture tacit knowledge from across the enterprise and make it explicit – and available across the enterprise.

The notion of ‘tacit knowledge’, first introduced by Michael Polanyi in the 1950s, refers to knowledge that is difficult to articulate, express and write down – or otherwise communicate. How do you describe the knowledge of how to ride a bike? Or turn a good-looking chair-leg in teak? Or paint a good Turner-like landscape? [Chippendale and Turner both knew something that very few others did, or have since – something that is very difficult to, metaphorically, write down.]

The Oxford philosopher Gilbert Ryle, in his 1949 book, The Concept of Mind, drew the distinction between ‘Know-How’ knowledge and ‘Know-That’ – or ‘Propositional’ – knowledge. The relation is obvious: ‘Know-How’ is often Tacit, and ‘Know-That’ can easily be made Explicit. The problem is that what is important in enterprises is Know-How, not Know-That.

Reading a PRINCE2 manual does not a project-manager make, nor reading the TOGAF manual an Architect. Know-How comes from doing – from experience and practice. Know-That can be bought from the open-market relatively cheaply – you can read it in a book.

But, whatever the details of theory, it is clear that knowledge, especially important tacit Know-How knowledge, is ‘localised’, and ‘relative’ to communities within enterprises. This is important for Enterprise Architecture because often decision-making processes are significantly separated from the operational processes and may themselves be localised in some sense.

In the management area, the concept of Tacit and Explicit knowledge was popularised by Nonaka and Takeuchi. They suggested that Tacit Know-How and Explicit Know-That in enterprises evolve in a continual spiral of learning and experience. The spiral passes through phases of Socialisation (Tacit-Tacit Learning), Externalisation (Tacit-Explicit ‘codification’), Combination (Explicit-Explicit ‘rationalisation’) and Internalisation (Explicit-Tacit ‘interpretation’) – the ‘SECI’ model.

The SECI model can be applied to personal, individual knowledge – but it can also apply to teams – small and large, departments and even whole divisions of enterprises. There is an emergent Know-How that groups of people develop through the experience and social learning of working together over time. This Know-How is often Tacit – cannot be easily written down – and shared by the people in the team, through their having ‘common understanding’ – the same mental models in their heads, familiar shared language, routines and expected behaviours and common assumptions and presumptions.

Max Boisot, in his book published in 2000, Knowledge Assets – Securing Competitive Advantage in the Information Economy suggested a similar cycle operates at the level of firms and industries. Boisot characterised knowledge in three dimensions – its position on the Abstract-Concrete scale, the degree to which the knowledge is codified, and the level of diffusion of the knowledge. Tacit knowledge in a small group may be identified as ‘Uncodified’ and ‘Undiffused’ whereas common, industry-wide or wider Explicit knowledge is ‘Codified’ and ‘Diffused’. Boisot’s theory may be regarded as the macro-level, or top-down, counterpart to Nonaka and Takeuchi’s bottom-up micro-level theory.

The Nonaka-Takeuchi and the Boisot theories of how knowledge is created and diffused through enterprises, industries and social groups of varying size have both been questioned and challenged on both empirical and theoretical grounds.

There is a body of theory around organisational learning and decision-making that suggests there can be a significant difference between ‘true and accurate’ models of the enterprise and its context and the localised models used in decision-making.

But, whatever the details of theory, it is clear that knowledge, especially important tacit Know-How knowledge, is ‘localised’, and ‘relative’ to communities within enterprises. This is important for Enterprise Architecture because often decision-making processes are significantly separated from the operational processes and may themselves be localised in some sense.

The question is, if decision-making processes are localised and know-how is localised and relative to particular communities within an enterprise, how do you know that the decisions are good, bad or indifferent? How do you know whether the mental models people are applying in decision-making are a true and accurate reflection of the reality in an enterprise? How do you know whether the best information has a) been used as input to the decision-making process and b) been properly and correctly understood in its detailed implications? How do you know that decisions made in one part of the enterprise will not have severe adverse consequences for another part of the enterprise?

There is a body of theory around organisational learning and decision-making that suggests there can be a significant difference between ‘true and accurate’ models of the enterprise and its context and the localised models used in decision-making.

March and Simon talked of ‘Bounded Rationality’ – that is localised knowledge and understanding. Argyris and Schon theorised about the difference between ‘Espoused Theories’ and ‘Theories-in-Use’. Cognitive Load Theory, and notions of Cognitive Frames and Cognitive Filters describe how managers are simply unable to assimilate and process the available information in their decision-making. When compounded with communication difficulties in large enterprises and the obfuscation effects of increasingly large volumes of data and raw information – ie ‘noise’, it is a wonder that decision-making processes in enterprises can be coherent at all.

There is a similar situation in the Philosophy of Science. Scientific knowledge production is a social activity – and science knowledge may be localised to particular communities within a science. When applied, how do you know the right knowledge – a true and accurate reflection of Reality – good theory – is being used?

Karl Popper, arguably the greatest Philosopher of Science of the 20th Century created a ‘framework’ for understanding how ‘objective’ knowledge is obtained from ‘subjective’ knowledge. He divided reality up into three distinct worlds: World 1 – the world of physical objects and events, World 2 – the world of our subjective impressions, perceptions and intuited mental models, and World 3 – the world of Objective Knowledge – or rather objective knowledge codified according to common conventions in information. [See Tanner Lectures on this.].

So Steven Weinberg’s book – Gravitation and Cosmology is a particular presentation of the theory of General Relativity – but the theory is not his subjective view, nor that particular presentation, but is an objective description and mathematical model of the reality of gravity. Gravity is a World 1 thing, Steven Weinberg’s perception of it is a World 2 thing but the theory of General Relativity is a World 3 thing.

The link to Enterprise Architecture is clear: 1) Enterprise reality – the technological and social structures in the enterprise and their interactions are World 1 things. 2) The intuited mental models and small-group managerial perceptions are World 2 things. 3) The coherent set of consistent explicit models – including Reference Models and Principles – are World 3 things.

What makes the EA models ‘objective knowledge’ – true and accurate, and independent of personal beliefs or perceptions – is that they are derived from subjective World 2 perceptions by processes of hypothesis or conjecture and then validation and verification in discussions. So socialised, verified and validated EA models are a solid basis for making good Enterprise Engineering decisions – just as scientific knowledge is the solid basis for decisions – particularly design decisions – in other types of engineering. The coherent models we usually call the Enterprise Architecture are World 3 objects – and an objective theory of the enterprise, not dependent on the views of any individual, or inscribed into any particular presentation. This is the philosophical basis for the practice of Enterprise Architecture – and the reason why EA distinguishes ‘Models’ from ‘Views’.

From this perspective, Enterprise Architecture is the model-making part of a scientific approach to enterprise management – just as, in Science, every theory has one or more models associated with it. From this perspective so-called ‘Scientific Management’ – meaning the management theories of F.W.Taylor and Henry Ford and others – looks very much like a 100-year-old, Victorian, ‘classical’ proto-science (Taylor developed his theories during the 1880s and 1890s). A proto-science about to be swept away by the modern, technology-enabled, 21 st Century scientific revolution in the business of enterprise management.

Sign up to our newsletter for free

 

 

 

 

The other lesson from the philosophy of science is that, if the above analysis is correct, the ‘classical paradigm’ of Tayloristic management won’t be swept away in a small number of years. Rather, it will gradually fade from the scene of management education and practice – as adherents to the old paradigm ‘leave the field’. According to Thomas Kuhn and his sociological theories of change in Science, a new paradigm, such as Enterprise Architecture, comes to be orthodox as the number of adherents to the older paradigm decline. The question is – how quickly?





There are no comments

Add yours

This site uses Akismet to reduce spam. Learn how your comment data is processed.

x
freshmail.com powered your email marketing