2nd Blog Post by Judith Verweijen

The challenges of researching armed actors in Eastern DRC: #2
Balancing the analysis of discourses and practices

In general, human beings tend to justify their own acts, represent
reality in a way that suits their interests, and develop rational and
rhetorical strategies for coping with conflicting or “inconvenient”
information, like downplaying, ignoring, or twisting certain facts.
Therefore, any researcher seeking principally to explain human
behavior, and to see how discourses relate to social practice, will
need to complement analysis of what people say and how they think with
what people actually do (or are perceived to be doing by others).
While it is certainly of importance to understand people’s narratives
and worldviews, as well as their self-professed motivations and the
meanings they attach to their own observations of “reality”, we cannot
base inferences on the workings of that “reality” on research
subjects’ discourses alone.

Psychological research on verbal reporting (i.e. Nisbett & Wilson
1977) points to the unreliability of people’s own explanations for
their behavior. In their self-attributions, people tend to appeal to
common-sense notions and theories of causality,. For example, if it is
generally assumed that soldiers commit abuses because they receive low
and irregular pay, soldiers will tend to highlight that dimension when
explaining why they or their colleagues transgressed. What further
limits the explanatory value of self-attributions is that it is simply
impossible for a person to be aware of all the causes of her or his
behavior.  In my research, I found for example that soldiers, in their
own narratives, systematically downplay or ignore the influence of
horizontal social (peer) pressure, despite the fact that there is a
wide consensus in military sociology about the importance of peers,
especially the primary combat unit, in shaping combatants’ behavior.

Finally, there is the well-known problem of socially desirable answers
related to the positionality of the researcher, which further limits
the insight into social practice that narratives may yield. While
universal, these problems are perhaps particularly pronounced in a
context of high asymmetry between researcher and subjects in terms of
power and resources.. This is one of the factors contributing to what
Chabal (1996) has called the “politics of the mirror”, or the tendency
by its inhabitants to present an image of Africa that they think
outsiders want to see. For example, when talking about their relations
with civilians, Mai Mai combatants may alternately emphasize
internationally circulating discourses on civilian protection and IHL
or highlight the “magical” factors shaping their behavior, depending
on the context, their relation to the researcher and what they think
the researcher wants to hear.

All of this makes that we have to go beyond self-attributions and
narratives if we want to make inferences regarding the causes of human
behavior. This becomes particularly acute when entering the domain of
violence and human rights abuses. Few are the soldiers or officers who
readily admit that either they or their troops have committed abuses,
and when they do so, explanations are mostly located in “lay theories”
of causality and imbued with self-justification. This inevitably leads
to discourses of victimization which emphasize poverty,
marginalization, humiliation or the past atrocities someone has been
subject to. As we familiarize ourselves with our research subjects,
getting to know them “as human beings”, perhaps even starting to
identify with them or develop a certain sympathy for them, these,
perhaps natural, impulses of denial and self-justification create a
tricky situation for the researcher.

For sure, perpetrators of even the most horrible abuses can be highly
friendly, eloquent, caring, polite, magnanimous, intelligent, and
gentle individuals. Discovering this can be quite a shock: not only
does it turn out that the Butcher, the Rapist, or the Torturer is like
everybody else, we may actually start to sympathize with him or her.
However, the “de-demonization” euphoria following this surprise is
inherently dangerous, as it makes us susceptible to copying our
research subjects’ discourses of denial, justification or
victimization and attribute too high of an explanatory value to them.

Logically, those researching perpetrators develop coping mechanisms in
order to deal with the schizophrenia resulting from having to
reconcile the image of their research subjects “as normal human
beings” with the awareness of the abuses they have committed.
Internalizing perpetrator discourses can be one such coping mechanism.
Thus, we may come to think for example that commander X never gave the
orders for that massacre, but that the situation simply got “out of
hand”, or we may start to believe that what they did wasn’t that bad,
that the stories are exaggerated by political adversaries trying to
taint the commander’s reputation. We may even end up believing in the
necessity or inevitability of a certain act of abuse or the innocence
of our perpetrator “friend”.

It is not easy to learn how to simultaneously fully acknowledge the
realities of military abuse, extortion, and intimidation, and the
responsibility individuals bear for those acts, while also feeling
sympathy for those having committed these abuses. It is perhaps a
matter of recognizing that perpetrators are neither devils nor angels,
and that their acts are the product of the interaction between the
context they live in and their own decisions. Personally, I needed to
fall a few times in the “perpetrator narrative” trap before managing
to better circumvent it. For example, I really believed for a brief
moment that a certain rebel commander could never tolerate his troops
ambushing and stealing, and that it was plausible that these acts were
wrongfully ascribed to his troops or carried out by “uncontrolled
elements”, as he himself stated. However, when I started to
intensively analyze patterns of ambushing and robbery in his
deployment area, and talked to a number of former combatants, I soon
found out my analysis was wrong.

Making that mistake was a useful experience, as it opened my eyes to
the wider problem of the temptation to attribute too much explanatory
value to individuals’ narratives, without complementing an analysis of
the latter with a thorough investigation of social practices, by
consulting other sources, like key informants and secondary
literature, and by using other data collection techniques, such as
(participant) observation. This allowed me for example to see that
civilians can be active agents in military abuse, like when they
appeal to armed actors for settling their scores and arranging their
conflicts. This did not become clear from civilians’ own stories, in
which they are invariably portrayed as passive victims of the
military.

The discourses of research subjects and  the “lay theories” of
causality  they reflect are contagious, and researchers seeking
explanations for human behavior have to learn how to build up a
certain resistance to them. At the same time, we have to be careful
not to throw out the baby with the bath water, as narratives can
provide valuable clues for causality and motivations. However, they
have to be balanced with other data collection techniques and sources
and be critically analyzed in relation to relevant theoretical
frameworks and findings from other contexts. This constant balancing
act is not easy, especially when entering the moral and methodological
minefield of combatant behavior and military abuse. But I found that
guarding this equilibrium becomes slightly easier as research
experience grows. After all, walking a tightrope is something one has
to learn.

Judith Verweijen is a PhD Candidate at Utrecht University carrying out
sociological research in Kivu for her dissertation on the Congolese
military (FARDC) and the factors that shape their interaction with
civilians. Additionally, she conducts research on a variety of armed
groups in Uvira & Fizi territory ( South Kivu), exploring the
functions such groups have in their socio-economic and political
context. She is now in the 13th month of her field research, carried
out in three phases since January 2010