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 122

24, 25, and 26 in two positive-feedback loops and is pulse stable. This component
can be value stabilized by inverting any one of arcs [1, 2], [2, 4], and [3, 1]. The
fourth component is value stable and comprises vertices 16, 17, and 18 in 2 two-arc
negative-feedback loops.

4.4 dIscussIon

4.4.1 C of n s t r u C t i of n o f C of g in i t i v e m A p s
The idea of cognitive maps was easily understood and utilized by communities. This
may be a reflection of the fact that the maps are a much easier way of depicting their
perceptions, which in turn indicates that communities are aware of the high degree
of interrelationships among factors that influence the health and sustainability of
their agroecosystems. The use of a metaphor as a guide in the selection of concepts
to be included in the map was successful in all the communities.
Dividing the participants into smaller groups during construction of the cogni-
tive maps allowed for the expression of different perspectives and the active involve-
ment of most participants. The visual nature of the cognitive maps makes it easy to
engender debate, thus providing a basis for debate and consensus building among
the participants and the creation of a synthesis cognitive map. However, the end
product is a compromise between the various views and does not necessarily capture
all the divergent perspectives. In this study, relationships and concepts that were not
unanimously agreed on by all participants were left out of the final cognitive map.
Ideally, cognitive maps representing the most divergent of perspectives should also
be analyzed and the conclusions compared and offered for debate by communities.
Cognitive maps were largely in agreement with the findings of the initial village
workshops. This was remarkable given that the cognitive maps were drawn several
months after the initial workshops. Important discrepancies, however, were present
in some cognitive maps, especially with regard to the role of some institutions and
in the number and descriptions of some of the problems. In Mahindi village, water
quality was added as a problem in the cognitive maps. In Kiawamagira, distance to
schools was added, while crop and dairy productivity were mentioned as separate
issues. These changes were most likely due to a reevaluation of the problem situation
rather than an inaccuracy in the findings of the village workshops.

4.4.2 u s e o f s i g n e D D i g r A p h s
The analysis of the communities' cognitive maps using graph theory was constrained
by two key limitations. The first was that it was difficult to assign weights to relation-
ships between many of the concepts used in the cognitive maps. Even where this was
possible, communities found it difficult to apply mathematical ideas to concepts and
relationships that they perceived mostly in qualitative terms. The other constraint was
that it was difficult to state concepts and relationships in such a way that all the arcs
in the digraph reflect equal time lags. In this study, participants were made aware of
the need to state relationships in a way that makes the arcs have more or less equal
time lags, but many participants were unable or unwilling to put this constraint in