Recently many of my conversations have focused on accountability. Sometimes the context is management of a complex project in an unpredictable landscape. Sometimes it is demonstration of outcomes and impacts for international programs to shift massively complex patterns such as nutrition and livelihood. Sometimes, it is simply a prickly relationship between a supervisor and a technical professional. Sometimes it is a new resident in the White House and his retinue.
Unpredictability makes accountability a problematic concept in complex systems, especially ones that are recognized to be adaptive or self-organizing. If you don’t know what an outcome will be, how do you hold self or others accountable to produce a result?
At the HSD Institute, we think of three different kinds of accountability in complex systems.
One is the traditional kind of accountability. It is possible—even advisable—when a system is in a relatively stable state. Examples in my experience include safety regulations, assembly line production, auto repair, logistics, and construction. Even as I write these, I am aware that practitioners in each of these fields would challenge the predictability of their domains, nevertheless:
• Change is slow.
• Parts are tightly coupled to each other.
• Causality is linear or simply nonlinear.
• Boundaries are relatively impermeable and clear.
• Diversity is limited.
• Degrees of freedom are low.
• Part, whole, and greater whole are constrained.
In such situations, one can be relatively certain of outcomes—at least in the short term. It is possible to articulate roles, responsibilities, and objectives and to expect them to remain constant over a specified period of time. In such conditions, one can be held accountable for outcomes.
The second kind of accountability emerges in the process of active self-organizing. From my point of view examples include functional mergers and acquisitions, service delivery, collaborations, cross-functional teams, and effective governments. In these situations:
• Change is unpredictable—sometimes moving quickly and sometimes moving slowly.
• Parts are loosely coupled to each other.
• Causality is complex and nonlinear.
• Boundaries are acknowledged, but they are permeable and sometimes fuzzy.
• Diversity is acknowledged in a small number of differences that make a difference.
• Degrees of freedom are variable across the system.
• Part, whole, and greater whole are mutually influential as patterns emerge and disappear across the entire system.
Accountability does not disappear in these systems, but traditional accountability isn’t possible. Instead of being held accountable to outcomes, individuals and groups can be held accountable for learning, shared meaning making, and directional movement. Are individuals and groups learning new things? Is shared meaning being constructed and/or maintained? Is the trend-line of processes and products moving toward a desirable goal?
The third kind of accountability is even more problematic. It arises when systems appear totally unorganized or random. Immediate response to disasters, transition times in economics and politics, crowds, new technologies, emerging markets, many conflicts are examples of random dynamics in human systems. Also, as I list these, I know there are people who see order and predictability in each of these situations. Patterns, and the categories that describe them, are always relative. In such situations:
• Change is so fast that patterns cannot be discerned over time.
• Parts are uncoupled from each other.
• Causality appears to be absent.
• Boundaries are nonexistent or so numerous that they are meaningless.
• Diversity is unbounded.
• Degrees of freedom tend toward the infinite.
• Interaction of the parts is so random that a system-wide pattern of the whole cannot appear and/or be maintained over time.
Though it may be hard to imagine, accountability is important even in human systems tending toward the random. In these situations, people can be held accountable to explore and share. In the same way that an ant colony spreads out in random patterns, finds a juicy spot, then returns to share the news, people hunt and gather. In random systems, people must be held accountable to gather and disseminate information. This behavior over time increases the coherence of system-wide understanding and action.
So, as I work with clients to improve performance, outcomes, and impacts, I try to help them distinguish which kind of accountabilities are possible in their environment, which kind fits their mission and vision, and which ones represent a reasonable investment in the continuing improvement of humans and their complex systems.
What can you learn when you apply this three-part definition of accountability to yourself and your colleagues?
Comments (3)
Glenda,
The way you are describing accountability sounds suspiciously like conventional management, which is knowledge about the business, the documents and electronic records (artifacts), and the intellectual assets (individuals and their work) that is within the organization's control or influence. If accountability is knowledge, then the definitions that you describe would seem to fit within the Cynefin framework. The first case fits the Known realm and the last is Chaotic.
Perhaps I should say that this accountability is knowledge that management SHOULD have, because some managers are pretty clueless. They may be clueless due to complexity (the CEO of any large company), or to poor communication skills, or to disembling (knowledge corruption). These behaviours and circumstances seem to align with the disordered center of Cynefin.
Cluelessness seems to beget cluelessness, corrupt knowledge grows, and companies and entire marketplaces spiral down into disorder. It clearly takes things other than knowledge to escape; things we call leadership, trust, collaboration, etc. But these are all processes - communications over time to increase knowledge. It is very easy to get lost in the recursion, and it seems that every time I look at knowledge I find more questions than answers.
Posted by tonyj | December 12, 2008 4:18 PM
Posted on December 12, 2008 16:18
Thank you very relevant to my work with voluntary organisations who are being stifled by demands for outcomes. Increasingly creativity is being expressed in meeting demands for proving outcomes rather than off thedoing the work. Yet the work is too complex and fuzzy-edged to be ameanable to the linear processes that are imposed. Your piece explains the issue very well.
Posted by James Ashdown | December 12, 2008 7:13 PM
Posted on December 12, 2008 19:13
The real challenge--for me as a leader and as a consultant--is to consider these dimensions simultaneously and each in the context of others. I can no more focus solely on predictable outcomes than I can focus solely on emergent patterns or broad exploration of the territory. To be effective, I have to do them all at the same time, with attention to fitting the options I can imagine with the reality I experience.
Often I find that proponents of complex adaptive responses are leary of acknowledging the facets of work that can be predicted and controlled. I find that stance as limiting as the one that sees only the linear, causal measures of accountability. Experience demands a both/and approach.
The other challenge is to see the options not as dichotomies but as continuua. Though our discourse often focuses on distinction, we need to be able to move with some agility across the complex landscape that lies between the two.
I don't see this a different in kind from traditional management. Certainly there are aspects of the tradition that are adaptive, else it would have died out long ago. Rather I see it as a reframing that allows for more, and more adaptive options for action.
When I worked with IT shops as they were moving from massive mainframe aps into the world of rapid prototyping, object-oriented architectures, and iterative methodologies, I found it helpful to talk about a new disciple. It isn't that the new approaches are undisciplined--the argument of many traditionalists. In my view we must address new disciplines and multiple disciplines to meet challenges of today and tomorrow.
I am doing some work with The National Implementation Research Network (NIRN) http://www.fpg.unc.edu/~nirn/. Their practice is at the intersection of evidence-based research and real-world implementation. They have lots of insight into this murky space between the objective, positivist knowing and the adaptive, interpretive action.
Posted by glenda eoyang | December 13, 2008 8:17 PM
Posted on December 13, 2008 20:17