Information Technology Reference
In-Depth Information
example, some were exposed to alternative income opportunities, modifying their
current and planned livelihood strategies [25].
These differences in norms were identified while working through the material,
with initial descriptive codes such as “cows”, “herd”, “livestock”, “land” and “plots”
applied, including the example statements above. Coding was carried out through
several iterations and refined into more analytical codes such as “worldview”, “tradi-
tional perspective” and “commercial perspective”, which also included the above
statements. In this way, and consistent with a grounded theoretical approach, a theory
has emerged from the data that some Maasai favour more “traditional” livelihood
strategies, investing in their herds as a means of coping with variable and uncertain
conditions, while others tend to invest spare capital in diversifying their livelihoods
with plots, employment and commercial activities. These worldviews are affected by
differential economic and political conditions and individual experiences of shocks
such as the 2009 drought. These findings indicate implications for the socio-cultural
norms and values of the system which will interact in complex ways with dynamics of
land-use change and climate change over time.
4.3
Agent Architecture
Based on the theory developed in the analysis of qualitative data, primary agents were
identified at the herder level and share common attributes such as ethnicity (Maasai)
and primary livelihood activity (pastoralism). Herder agents have a set of overarching
goals, which were derived from the primary and secondary data. These goals are as
follows, in order of importance:
1.
Meet subsistence needs e.g. through pastoralism, agro-pastoralism or a mix of pas-
toralism with other livelihood activities;
2.
Cope with stress and shocks e.g. maintain herd mobility, manage herd composi-
tion, herd-splitting, increase herd size; and
3.
Maintain socio-cultural functions e.g.
osotua
(a practice by which Maasai will gift
stock to each other when in need, in a relationship that is reciprocal over time),
knowledge or information transfer through social networks, social capital between
members of the same age-set, clan and kin groups.
While all herders share these common goals, the agents are differentiated in their
motivations and options by criteria which include clan, relative wealth, opportunities
for alternative strategies and worldview.
To capture the role of different worldviews identified in the analysis, a Beliefs-
Desires-Intention (BDI) agent architecture has been developed [26]. This approach
allows for agent actions to be mediated by their particular worldview or belief sets,
which are derived from the environment, from their learned experiences and through
their interactions with each other. The ABM is implemented in Netlogo and uses an
existing BDI and communication library to manage beliefs and intentions and pass
messages between agents [27].
Under this protocol, all agents have declared variables beliefs and intentions. Be-
liefs are composed of
belief-type
, e.g.
traditional
or
commercial
, and
