"Anarchy is the Mother of the Order" or Gartner suggests what the EA 'must' do

Gartner's Press Release (Egham, UK, August 11, 2009) says: "Enterprise architects must adopt a new style of enterprise architecture (EA) to respond to the growing variety and complexity in markets, economies, nations, networks and companies". This is just great on-time observation. "Analysts advised companies to adopt 'emergent architecture', also known as middle-out EA and light EA, and set out definitions of the new approach" . Well, in my IT practice, all "light" approaches usually throw away the value and keep the form.

I think that this case is not an exception, unfortunately. I really do not think that this is a new approach, there were many attempts to compromise centralised control under the banner of local freedom and flexibility (or individual agenda). Service Oriented Enterprise (SOE) easily resolves the problems of business and IT flexibility via service composability, which requires very strong discipline as well as very high level of responsibilities for assigned work. At the same time, SOE promotes innovations by bringing regulated market inside the enterprise: if your idea is good technically and economically, you will get the resources for its development and maturity from your consumers even if this is not your area of business.

I would like to follow "seven properties that differentiate emergent architecture from the traditional approach to EA" according to Gartner.

1. Non-deterministic - this property requires Enterprise Architects to 'decentralise decision-making to enable innovation'. This extremely populist declaration mistakenly assumes that EA makes enterprise decisions (today). Sorry to disappoint the authors but EA only provides the enterprise governing policies, which, BTW, define who make the decisions, and, it is nor EA but Business. EA helps corporate Business to preserve architectural integrity and controls over the rest of IT. In many cases, the IT EA has less authority even than PMO.
From another hand, EA must make decisions at the enterprise level. The architectural teams at the LOB or BU levels must make decision at their levels in line with upper level decisions. I do not recall a case when an innovation in architecture was 'killed' based on the architectural ground; the stopper usually comes from managerial ambitions and economics (or manipulated economics).

Not all 'cool' ideas are always economically effective. Architects have to become more entrepreneurial and business argumentative to get a 'green line' to their ideas. Nobody is obliged to patronise them.

2. Autonomous actors - Please, read this nonsense: 'Enterprise architects can no longer control all aspects of architecture as they once did'. I dare to hope that some of EA still control, still can grasp the enterprise architecture with certain experience and qualification. Probably, the authors were not aware that architecture is about of abstraction. In the abstraction, very complex things appear simpler. If EA were strong before and did not allow delivery teams to mess with architectural integrity, EA may be well controlled and managed.

It is interesting to notice that the alternative Grtner's statement says "They [architects] must now recognise the broader business ecosystem and devolve control to constituents'. Well, I thought that EA is about 'broader business ecosystem' first of all and only then is about infrastructure and technologies. So, knowledge and responsibility for the architecture of an organisational unit does not necessitate to 'devolve control to constituents' all of a sudden; this depends on the level of the importance of the tasks, available skills, business information flows and so on. What would be a business value (yes, business value because technical value is not counted w/o related business value, c'est la vie) of delegating technology decision to the Project Architect who is not informed about the organisation's business objectives, needs and trends in the business domain?

3. Rule-bound actors - 'Where in the past enterprise architects provided detailed design specifications for all aspects of the EA, they must now define a minimal set of rules and enable choice'. I am afraid this is another mix-no-match case. EA must provide the detailed design specifications for all aspects of the EA, the LOB Architects must provide the detailed design specifications for all aspects of the LOB, the BU Architects must provide the detailed design specifications for all aspects of the BU, and the Project Architects must provide the detailed design specifications for all aspects of the Project. This is normal and effective practice. If an EA provide the detailed design specifications for all aspects of the Project, the architecture manager better be fired as well as all others who explicitly or implicitly allows such 'order'.

4. Goal-oriented actors - "Previously, the only goals that mattered were the corporate goals but this has now shifted to each constituent acting in their own best interests" - this is the absolute truth but... how this relates to EA? This is pure business managerial problem. The notion of 'conflict of interests' between the corporate and individual constituents is well known. I believe that properly formed Business Architecture can break this practice. But this is the subject of another BLOG.

5. Local Influences - '... No individual actor has data about all of an emergent system. EA must increasingly coordinate'. Certainly. And to coordinate, EA must control architectural process that incorporates individual actors, That's simple. For example, if an individual actor wants to design a technical solution using old but known technology or, right opposite, very new and not validated technology, EA or its lower level architectural level must stand against such unjustifiable 'innovations'. I've never read a story about a pilot who just graduated from the school and was assigned to drive an inter-continental flight; in 1998-2000 many school boys and girls who built a couple of Web pages claimed themselves Web Architects... and some managers believed them (for a while)

6. Dynamic or Adaptive Systems - "...EA must design emergent systems sense and respond to changes in their environment". I call this 'design for flexibility' Ability to adopt changes is the major success factor in modern market situation and EA, which does not do this, is simply a bad BA. I do not think that Gartner has found something really new in this case. EA always had to bridge Business to Technology; if Business changes fast, Technology must change even faster to survive as some researchers have found

7. Resource-Constrained Environment - "An environment of abundance does not enable emergence; rather, the scarcity of resources drives emergence" - wow! This is the Wisdom! At the higher abstract level, it sounds a bit shorter (as I said before): "Deficit is the mother of Innovation" In the context of previously described "architectural properties" by Gartner, this property may be interpreted something like this: if you want EAs to work better, revoke their decisions, authority and controls and give its to uninformed and irresponsible (not because they are 'bad' but because the business organisation doesn't relay on them) constituents with their own agenda. I think, this is the very recipe on how to rid of the EA without much blood and dust.