Towards adding argumentation information to systems maps and systems complexity to argument maps.Posted: February 23, 2015
This brief exploration assumes that discussions as well as any systems analysis and modeling are essentially part of human efforts to deal with some problem, to achieve some change of conditions in a situation, — a change that expected to be different from how that situation would exist or change on its own without a planning intervention.
1 Adding questions and arguments to systems diagrams.
Focusing on a single component of a typical systems diagram: two elements (variables)
A and B are linked by a connection / relationship R(AB) :
A ———R———> B
For convenience, in the following these elements are listed vertically to allow adding questions people might ask about them, and hold different opinions about the possible answers.
A What is A?
| What is the current value (description) of A? (at time i)
| How will A change (e.g. what will the value of A be at time i+j)?
| What causes / caused A?
| Should changing A be a part of a policy / plan?
| If so: What action steps S (Sequence? Times? Actors?) and
| What Means / resources M will be needed?
| Are the means actors etc. available? Able? Willing?
| What will be the consequences KA of changing A?
| Who would be affected by KA? In what way?
| Is consequence KAj desirable? Undesirable?
| Q: Is A the appropriate concept for the problem at hand?
| (and the questions about A the appropriate questions?)
R(AB) What is the relationship R(AB)?
| What is the direction of R?
| Should there be a relation R(AB)?
| What is the (current) rate of R? (Other parameters? E.g. strength)?
| What should the rate of R be?
B What is B?
. What is the current state / value of B?
. Should B be the aim / goal G of a policy / plan?
. Are there other (alternative) means for attaining B?
. What should be the desired state / value of B? (At what time?)
. What factors (other than A) are influencing B?
. What would be the consequences K of attaining G?
. Who would be affected by K? In what way?
. Is consequence KBj desirable? Undesirable?
. Q: Is B the appropriate concept for the problem at hand?
. (and the questions about B the appropriate questions?)
Most systems models and diagrams do not show such questions and arguments – it is my impression that they either assume that differences of opinion about the underlying assumptions have been ‘settled’ in the respectively last version of the model, or that the modeler’s understanding of those assumptions is the best or valid one (on the authority of having constructed the model?). They thereby arguably discourage discussion. They also do not easily accommodate the complete description of plans or policies, assuming a kind of ‘refraining from committing to solutions’ attitude of just ‘objectively’ conveying the simulated consequences of different policies while limiting the range of policy or plan options by omitting the aspects addressed by the questions and arguments.
2 Adding systems complexity information to argument maps
Typically, the planning discourse will consist of a growing set of ‘pro’ and ‘con’ arguments about plan proposals; any decision should be based on ‘due consideration’ of all these arguments. In the common practice of discussion (even in carefully structured participatory events) the individual typical planning argument can be represented as follows:
“Plan P ought to be adopted and implemented
Implementing the plan P will have relationship R (e.g. lead to) consequence K, given conditions C
Consequence K ought to be pursued (is a goal G)
Conditions C are present.
This argument, in which several premises already have been added that in reality often are omitted as ‘taken for granted’, can be represented in more concise formal ways , for example as follows:
D(P) (Deontic claim: conclusion, proposal to be supported)
FI((P –R—>K)|C) (Factual-instrumental premise)
D(K) (Deontic premise)
F(C) (Factual premise)
The argumentative process, in the view of Rittel’s ‘Argumentative Model of Planning’, consists of asking questions (in the case of controversial questions, ‘raising issues’) for the purpose of clarifying, challenging or supporting the various premises. This serves to increase participants’ understanding of the situation and its complexity, which from the point of view of the ‘Systems Perspective’ may be merely ‘crudely’, only qualitatively and thus inadequately represented in the arguments in a ‘live’ discussion. Some potential questions for the above premises are the following:
D(P) Description, explanation of the plan and its details:
Current condition / situation?
Causes, necessary conditions for problem to exist, contributing factors?
Aims / goals?
Other possible means of addressing problem?
Q: wrong question: wrong way of looking at the problem?
Implementation details? Steps, actions? Sequence?
Actors / responsibilities?
Means and resources needed? Availability? Costs?
FI((P –R–>K)|C)) : Does the relationship hold? Currently? Future?
R(P,K) Explanation: Type of relationship?
(Causal, analogy, part-whole, logical implication…)
Existence and direction of relationship? Reverse? Spurious?
Strength of relationship?
Conditions under which the relationship can exist / function?
D(K) Should consequence K be pursued?
Explanation / description of K: details?
What other factors (than the provisions of plan P) affect / influence K?
Other (alternative) means of achieving K?
F(C) Are the conditions C (under which relationship R holds) present?
Will they be present in future?
What are the conditions C?
What factors (other than those activated by plan P) affect / influence C?
If conditions C are NOT reliably present,
what provisions must be made to secure them? (Plan additions?)
These questions, (which arguably should be better accommodated in systems diagrams) can be taken up and addressed in the normal discussion process. Their sequence and orderly treatment representation, especially to provide adequate overview, can be improved, and could be significantly improved by better representation of the variety and complexity of the additional elements introduced by the questions raised.
This is especially true with respect to the question about Conditions C under which the claimed relationship R is assumed to hold. A more careful examination of this question (i.e. more careful than the common qualification ‘everything else being equal’: what IS that ‘everything else’ – and IS it ‘equal’?) will reveal that there are many conditions, and that they are interrelated in different, complex ways, with behaviors over time that we have trouble fully understanding. In other words, they constitute a ‘systems network’ of elements, factors and relationships including positive and negative feedback loops – precisely the kind of network shown in systems diagrams.
Thus, it must be argued that in order to live up to the sensible principle that decisions to adopt or reject plans should be made on the basis of due consideration (i.e. understanding) of all the pro and con arguments, the assessment of those arguments should include adequate understanding of the systems networks referred to in all the pro and con arguments.
The implication of the above considerations is. I think, fairly clear: Neither does common practice of systems modeling or diagramming adequately accommodate questions and arguments about model assumptions, nor do common representations (issue and argument maps) of the argumentative discourse adequately accommodate systems complexity. Which means that the task of developing better means of meeting that requirement is quite urgent; the development of effective global discourse support platforms for addressing the global crises we are facing will depend on acceptable solutions for this question. But this is still a vague goal: I have not seen anything in the way of specific means of achieving it yet. Work to do.