Picture courtesy of Victor Nuno via Flickr
I’ve been reading around and about the subject of coherence for the last few days, mainly on public transport in Singapore between meetings and jet lag. I introduced the subject a week ago in the context of making decisions in a complex space. I quoted Thagard to the effect that A belief is justified not because it is indubitable or is derived from some other indubitable beliefs, but because it coheres with other beliefs that jointly support each other. I now want to extend that my looking at different types of coherence. I working here towards a formal method (including the possibility of using computational capacity) to rethink the nature of decision support systems so bear with me over the next few months as I work this through on the blog. Thagard proposes a taxonomy of coherence and I want to work through that using his language as a basis to organisational decision making.
Traditional accounts of rational decision making follow Aristotle in suggesting that you make a series of decisions, each one of which logically arises from its predecessor. It therefore follows that improving the information at each stage and a rigorous application of logic is what is needed. Analytic philosophy restricted this to linguistic analysis, phenomenology to conscious experience. Both rejected empiricism. Now with the growth of understanding from cognitive science not only is a degree of empiricism back in fashion, but the split between philosophy and psychology which started in the late nineteenth century is starting to be healed. Theories of coherence help us reconcile this various perspectives, so lets work through an example. The column on ethics is a summary of Thagard’s take on the left followed by my reflections and comments.
|Type||Summary of Thagard||My reflections|
|1||Deliberative Coherence||This gives priority to intrinsic goals, positive constraints are achieved if an action facilitates a goal, negative constraints arise from incompatibilities and contradictions. Tests include symmetry, degree of facilitation, goal priority and judgement.||Here we are dealing with explicit goals and potentially empirical evidence. Of course for this to be the case the granularity of the action must be compatible with the nature of the system. Fine granularity safe-fail experiments if complex, more elaborate fail-safe plans if complicated. We should expect for each action some assessment of expected outcome, and an evaluation of how those outcomes impact on overall goals together with a risk assessment. This is the domain of objective decisions without opinion.|
|2||Explanatory Coherence||Here decisions depend in part on an evaluation of empirical hypotheses and a complex evaluation of evidence. There will not be an absolute judgement on the basis of evidence, but there is a “soft constraint” that biases towards evidence as against opinion. I might for example accept the evidence that capital punishment deters criminals, but still oppose it||Here multiple goals and opinions are in play many of which may not be fully articulated. One can argue as to the facts, but an argument also has to relate to abstract values or principles from multiple interested parties. Here we will get ethical dilemmas where business goals for example may skirt the edge of ethiical behaviour. Duties to shareholders may conflict with duties to humanity.|
|3||Analogical Coherence||Is execution of a criminal analogically linked to killing a defenseless victim or to legitimate personal self defense. At the time of the invasion of Iraq was the west facing another Hitler and risking appeasement or was it the meaningless escalation that led to the WWI. These arguments are rarely convincing on their own but they contribute to the overall coherence of an argument||I think Thagard may be wrong here, in organisations analogies with past failure and success can often comprise the argument, the whole argument and nothing but the argument. However overall the point is well made. For humans to understand something it has to relate to something they already understand. Metaphor can also be a sophisticated form of argument here. Some of our techniques such as The future backwards are designed to identify the most common analogical points in an organisations past.|
Now this is a starting point. One of the features of the Cynefin framework is to make people thing differently about a problem, resolving issues in manners appropriate to the domain. Started to work on coherence is designed to expand this approach by distinguishing different types of coherence and creating different structures and forms to resolve issues before conflict arises at a whole decision level. Any comments or questions welcome, expect more on this