Use of Systems Theory, Directed Graphs, and Pulse Process Models in an Adaptive Approach to Agroecosystem Health and Sustainability - Integrated Assessment of Health and Sustainability of Agroecosystems - page 123

their maps. The digraphs are therefore most useful for short-to-medium-term analy-
sis that can be updated iteratively. Using geometric analytical techniques as well
as sensitivity analysis, useful insights can still be obtained from the less-detailed
digraphs derived from these cognitive maps. However, the conclusions from these
analyses are less detailed than would be with complete and detailed digraphs. In
an action research process, cognitive maps can be reevaluated and updated in each
action research cycle as the local theory develops. In this way, more details—both
structural and numerical—can be added to enable detailed analyses.
In this study, an additional quality—impact—of the relationships between ver-
tices was included. The rationale for this was that the purpose of community goals
and objectives was to minimize the negative impact of problem situations. Because
of the interrelationships among factors, minimizing negative impact involves trade-
offs because some of the solutions may in turn generate negative impacts. The
goal-seeking behavior of the system can therefore be seen as maximizing the posi-
tive impact of community goals through changes in structure and the application of
pulses to the system. In addition, the digraph assumes that the system dynamics are
linear. Although this is not the case for most of the relationships, the linear model is
likely to be a suitable generalization of the processes over the short term.
4.4.3 g e of m e t r i C A n A l y s e s
Income, farm productivity, and human health were consistently among the factors that
had high cognitive centrality. This would be expected for agricultural communities in
a largely subsistence farming system. The inclusion of labor as an important system
output and its export as a coping strategy, however, may be an indication that subsis-
tence was heavily constrained. This conclusion was further augmented by the exis-
tence of several sources in all the digraphs, many of which related to infrastructure
and other cash-economy-dependent components. While some of these may be due to
cognitive imbalances, they also indicate (largely) a perceived dependency on exter-
nal intervention. A good example was in Mahindi village, where use of water tanks
was seen as ameliorating water shortage. An agent external to the community had
donated these tanks to some households. No connection was made between income
and availability of tanks. Surprisingly, although many of the sources were recognized
as public goods that should be provided through a taxation system, participants in all
communities were unanimous that those connections do not exist in reality.
In nearly all communities, the rank of community goals based on their total
impact was different from the ranking during the initial workshops, suggesting
that either the geometric process was not confluent with the community's cogni-
tive processes or the impact of goals was not the predominant criterion for ranking.
Unexpectedly, the ranking of community goals based on their total degrees was
remarkably confluent with that in the initial village workshops, suggesting that the
complexity of interrelationships may have been an important ranking criterion. It is
therefore not surprising that all communities reconsidered their ranking after draw-
ing the cognitive maps. It would have been useful to provide the communities with
the results of the geometric analysis during this reevaluation. Unfortunately, this was
not possible in this study.