Nov 192013

The information and communications technology (ICT) world is constantly evolving in complexity. As computational technology advances, it allows for the building of more capable systems, architectures, and solutions. We’ve added so much agility to the behavior of systems that many now consider them as complex adaptive systems, suspected of developing their own intelligence. I would rather call it “stochastically featured” because — due to their omnipotent presence, internal complexity, and strength of interrelationship — we are not able to predict their deterministic behaviors in definitive ways.

I have seen a certain pattern in the science world in which a decomposition path is utilized to understand the “pieces” but — as Albert-László Barabási described in his book Linked — people then often fall into an understanding trap of not being able to reconstruct the pieces into proper “wholes” due to links, which have been lost or missed during subsequent decompositions. I have observed the same pattern in technology management, where we focus on analysis until we understand the pieces, then we reconstruct the solution based on such an understanding. But what a pity — over and over again, this does not work.

To manage complexity is not to understand it in details, but to control the whole. Porting it into the ICT world means focusing on architecture and synthesis instead of systems and analysis. Enterprise architecture, as a philosophy, makes this control feasible.

Wrong Practices or Wrong Views?

During the last decade we have developed many management practices to aid us in dealing with IT problems. But, despite efforts of practices implementation and maturation, the problems tend to continually slip from our control. Where is the solution? Are these practices outdated? Do we need some new silver bullet? Or maybe humans are too weak to handle the IT world and cannot control it anymore?

My opinion is that none of these statements is true. We have focused so much of our attention on practices that the subject of our effort — the IS environment itself — has evolved unnoticed. It became too complex for us to handle using these practices in their entirety. Thus, we must consciously decompose these practices by subject, but in a way that guards us from the trap. We can continually exercise the practices as long as we incorporate an EA philosophy into their bodies. This is where we will gain both control and efficiency.

I suppose all IT managers at one time or another scream, “But we did that already!” As long as organizations only embed EA philosophy in “this or that” ICT management practices, they will make no significant gains. Such philosophy should become the most fundamental technology management tool practiced, and we should perceive it as the bond among various ICT-related tasks and endeavors. This is the view I present, and hopefully prove, in my recent Executive Report.


Sebastian Konkol

Sebastian Konkol is a Senior Consultant with Cutter Consortium's Business & Enterprise Architecture practice. He is an enterprise startup and strategy consultant and the owner of the consultancy firm TwisterSolve.


  6 Responses to “IT Is Complex, Not Unmanageable”

  1. Sebastian,

    your posting follows the basic argument of reductionism by reducing everything into its pieces as opposed to holism. There are always areas where reductionism is working, but then again there are some where it does not work. The argument has so far run for over 100 years now. An example in ICT where the reductionism you are promoting does not work are for for example wicked problems ( where just by the definition it will not work.

    Since I have spent many years on these arguments on both sites I think the best is to say it depends, rather than to try to come up with the silver bullet.

    • Cay,

      I would not dare to tell, that there is anything close to silver bullet. Neither divide-and-conquer, nor look-at-relations approaches are perfect. In practice, of course, it depends.

      The complexity is growing. Methods used successfully for dealing with ICT problems can no longer be successful due to rise in complexity. That is the message.

      I have spent some years on studying complexity in human world – social networks. I can see many similarities. My goal is to convince the report readers to have a look at the other approaches, as well. I am too young to judge on silver bullets. :-)

      Thank You very much for Your comment!

      Sebastian Konkol

  2. avatar


    While I fundamentally agree to your assertions, your post seems to make EA extremely limited to just IT. Sure complexity in IT may be important, but EA is (and should be) used to tackle business complexity, which is a much bigger issue leaders are grappling with.

    In your post you allude to taking a systemic view of the enterprise, even though you don’t seem to mention it explicitly.

    Check out the book “A Systemic Perspective to Managing Complexity with Enterprise Architecture” ISBN: 9781466645189. I just delivered a plenary session at the Federal EA Conference in Washington DC. A hundred copies of this book were distributed to Government Enterprise Architects. So, there is already acceptance and traction to “systemic enterprise architecture”.

    Best wishes to your report.

    • I agree – 100% or more, if possible. :-) EA is capable of providing solutions to much wider scope of issues, that these located in the core of IT. In my report I focused on IT management issues, which solutions can be guided and augmented by EA.

      Thank You for the book reference!


  3. I think reducing IT problems into pieces will increase its complexity, not while decomposing and solving it but when it is meant to get the original answer by re combining the pieces’ resultant.

    • Sure! I do not try to convince, that reductionism is appropriate way of dealing with complexity. I say, that while decomposing problems into smaller and smaller pieces we loose the links, which makes the original problem hardly possible to solve based on the sum of individual pieces’ solutions.

 Leave a Reply

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <s> <strike> <strong>