Managing Complexity Through Conversation

Government increasingly manages complex requirements embedded in physical and adaptive systems, generally considered the two main types. Complex systems are characterized by self-organizing, non-linear and adaptive interactions (especially involving humans) nested in hierarchies or levels. Each is ordered by its own rules which interact to produce unexpected consequences, including rare and emergent behaviors. Government missions from ecosystem protection to financial market regulation exhibit these characteristics.

Government is appropriately increasing its use of data analytics to manage complex requirements. Data scientists simplify complexity and improve predictive capability by finding patterns and relationships that would otherwise go undetected. Better predictive capability means better program management, including more efficient resource use.

While there’s no substitute for computing power to understand some complexity, there are times when good ol’ fashion human communication will get the job done.

Humans are great pattern-finders and dot-connectors, especially of unstructured data. Neuroscience shows how our brains are wired to recognize patterns. We rapidly evaluate situations using pre-cognitive or pre-deliberative mental processes, and we impose patterns on fragments of data and information to make them coherent.

The easiest way to make use of this capability is simply to talk. You can start by talking to yourself but do it out loud. Voicing thoughts and “gut feelings” means putting words on things that might still be ambiguous. When our ears can hear what our mouth is saying, our brain can judge the fit between words and notions. That, alone, can clarify some things.

Talking with others adds advantages. Because intuition is commonly involved in pattern recognition and formation, and because intuition relies heavily on assumptions and implicit inferences, talking to others does two things. It allows you to check your assumptions and inferences, and it allows others to add information. Try this conversation sequence as a way to do both these things:

  • Describe the situation you think you face
  • Describe what you think should be done, and why
  • Describe how you think that will help
  • Ask for general reactions. Ask specifically what you might have overlooked or gotten wrong, or what could go wrong with your plan

This sequence invites everyone to voice and compare observations, judgments, assumptions, and inferences. There are structured techniques you can apply but even informal conversation – at a staff meeting, a brown-bag lunch – will enable people to pool what they know as individuals into a group knowledge. As the conversation unfolds, apply these guidelines:

  • The group will identify gaps in data and information. There’s a cost to collecting and analyzing data so get a sense of the value of additional increments of data before you launch a collection effort.
  • Look for patterns and relationships. Identify different types of relationships (cause, priority, supports the achievement of, hinders) and look at relationships in both directions.
  • Challenge each other to decide if the group is finding what it’s looking for, or seeing what it finds. One isn’t right or wrong, but you don’t want to confuse the two.
  • Look for something specific rather than broad, or keep decomposing broad ideas into more specific ones. Complicated explanations can hide information about assumptions, relationships, dependencies, etc. Strive for simpler explanations. If you oversimplify, you’ll know when you test the idea.

If conversation leads to a decision and action, talk all the way through implementation. Keep describing the situation you think you face…what you think should be done, and why…how you think that will help…what you’re missing or what could go wrong. Managing complexity is more like a game of chess or checkers than it is like rolling a bowling ball to see what happens to the pins. Continuous conversation will improve situational awareness as your plan meets reality and allow you to effectively adjust.

Using a team of subject matter experts as sensors and processors is a great way to manage complexity without computing power. Even when using computers for big data analytics, however, people need to talk about the data and the inquiry. This conversation sequence is an effective way to pair human and machine capability.

Leave a Comment