10

Clandestine Political Violence

Conclusions

Research on political violence, which has bloomed in recent years, has offered
numerous explanations of the structural preconditions for, organizational char-
acteristics of, and individual predispositions toward the development of clandes-
tine political violence. In particular, literature on terrorism has linked it to a
broad list of pathologies. Although rich in case studies, which are sometimes of
good quality, terrorism studies as a field has been strongly criticized for the
inadequacy of its sources of information, as well as its lack of theorization and
predominant orientation toward counterterrorism. Social movement studies,
which instead offers concepts and theories for the study of the evolution of
repertories of protest, has only rarely addressed violence, especially in its more
radical forms. This volume aimed at filling this gap by focusing on what I have
defined as clandestine political violence. In this conclusion, I first synthesize my
empirical results and then discuss some potential extensions of these results.
In considering clandestine political violence as an extreme form of violence
perpetrated by political groups active in the underground, I have looked at the
field of social movements for inspiration. Following that literature, I have
considered the evolution of violence as embedded in social and political conflicts,
influenced by the political opportunities available for elites and challengers as
well as the material and cognitive resources available to contenders. Building on
some important innovations in the area, I suggested that violence could not be
satisfactorily explained by looking exclusively at structural conditions.
Clandestine forms of violence, in particular, are embraced by tiny minorities
that react with radicalization to conditions that lead others toward moderation.
Without claiming to explain clandestine political violence, I singled out some
causal mechanisms that I found at work in the different cases, linking macro,
meso, and micro levels of action. Rather than emphasizing teleological or
behavioral explanations, I have built on what Charles Tilly (2003) called a
relational perspective. In all my cases, radicalization processes happened during

282

Downloaded from https:/www.cambridge.org/core. University of Sussex Library, on 03 Mar 2017 at 14:29:18, subject to the Cambridge Core terms
of use, available at https:/www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139043144.011
 Conclusions 283

the interactions of various collective actors, including social movement orga-

nizations as well as state institutions. These actors engaged in changing relations
of cooperation and competition.
I have also emphasized the constructed nature of the process, as contextual
conditions are filtered through the perceptions of such conditions – the “attri-
bution of opportunities” in McAdam, Tarrow, and Tilly’s (2001) language. In
all my cases, I noted how clandestine groups perceived that the spaces for
peaceful protest were closing down and also that conditions conducive to violent
action were opening up.
Moreover, I have stressed that violence developed during long-lasting pro-
cesses: conditions were not only preexisting but also formed in action. In this
sense, violence is an emergent phenomenon, reproducing in action the condi-
tions for its very development. Transformative events feed escalation through
the constitution of radical identities, as well as by forcing actors to take positions
(Sewell 1996).
My attempt at global comparisons – which I defined as based on a small-N
comparison of critical cases, chosen on the basis of a most-different research
design – brought me to the search not for root causes but for causal mechanisms,
as chains of interactions during which structural conditions are filtered and
produce effects but are also themselves transformed. No doubt, the degree of
economic development, democratic qualities, and national cultures vary broadly
among the historical cases I have analyzed, which include illustrations of left-
wing, right-wing, ethnonationalist, and religious types of clandestine political
violence. However, my analysis has also shown some similarities in the mecha-
nisms that underlie the radicalization of political conflicts and sustain clandes-
tine political violence. I have defined these mechanisms as competitive
escalation, escalating policing, organizational compartmentalization, action
militarization, ideological encapsulation, networking into clandestinity, and
militant enclosure, and I have devoted a chapter to each concept.
Escalating policing is the first mechanism I singled out. Protest, as a challenge
to public order, normally brings about interactions between protestors with
police, who are charged with defending the public order. Strategies of protest
policing, however, vary broadly. The police can privilege the right to demon-
strate over disturbances to public order and thus tolerate minor violations, or
they can strictly enforce law and order. They can rely on softer or harder tactics
when they intervene, using persuasion or force. Their intervention can be not
only more or less brutal but also more or less selective. As Tilly (1978) suggested,
regimes are distinguished by a variable mix of facilitation of some groups and
forms of participation and repression of others.
In all of the cases studied, in a process of double diffusion (della Porta and
Tarrow 2012), the radicalization of the forms of protest interacted with repres-
sive styles that were not only brutal but also often diffuse, hitting not only violent
militants but also nonviolent ones. Tactical interactions developed through
reciprocal adaptation to innovative turns, such that each party’s choices were

Downloaded from https:/www.cambridge.org/core. University of Sussex Library, on 03 Mar 2017 at 14:29:18, subject to the Cambridge Core terms
of use, available at https:/www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139043144.011
 284 Clandestine Political Violence

influenced by those of the adversary (McAdam 1983). Violence spread when the
state was perceived to have overreacted to the emergence of protest – as was the
case in Italy, when the student movement and the labor movement protest
signaled a growing intensity of conflicts; this process was even more evident in
Franco’s Spain, when labor protests met ethnic revival, and in the authoritarian
regimes in the Middle East, which reacted strongly to the so-called religious
awakening.
In all of the cases, in fact, everyday experiences of physical confrontation with
police brought about an image of an unfair state that was ready to use brutal
force against its citizens. The more the repression was perceived as indiscrimi-
nate, the greater the people’s solidarity with – or at least the tolerance of – the
militant groups: this was the case, in particular, in Franco’s Spain and under the
authoritarian regime of Mubarak in Egypt or the Israeli occupation in Palestine.
These types of police actions delegitimized not only the police but also the state,
which the police claimed to serve. Moreover, perceptions of injustice increased
when the state was seen as taking sides, repressing some groups’ violent behav-
iors but tolerating the violence of others – as was the case in Italy, where the state
was seen as supporting the radical Right. In action, repression created subcul-
tures sympathetic to violence, often resuscitating old myths; thus violence itself
started to be perceived as a resource in the internal competition within the social
movement family. Escalation was facilitated by not only indiscriminate but also
inconsistent repression. For the Italian right-wingers as well as for the Islamists
in Saudi Arabia, feelings of disconcert were created by what was considered as a
betrayal by a state that had been seen as somewhat supportive. In all cases,
repression was perceived as unjust (Gamson, Fireman, and Rytina 1982).
In fact, repression produced transformative events (Beissinger 2002; della
Porta 2008b; Sewell 1996). Left-wing, right-wing, ethnonationalist, and reli-
gious militants alike recalled brutal charges of demonstrators or the killing of
comrades as fueling intense emotions of identification with a community of
fighters and the designation of the state as an enemy. Repression created and
recreated martyrs and myths, which justified violence as defense and/or revenge.
Read within a broader narrative of oppression and resistance, heavy repression
was framed as an indicator that there was no other way out. This led to, at the
same time, mistrust in peaceful means of protest and also confidence in the
effectiveness of violence.
Furthermore, the interactions in the streets were then embedded in broader
relations that involved various actors: from political parties to interest groups,
and from social movement organizations to opinion makers. Protesting and
policing became bones of contention, resulting in intense debates on the limits
on protest rights and on forms of police repression. In short, those debates
addressed the metaquestion of the meaning of democracy, in particular the
development of civil rights versus law-and-order coalitions (della Porta 1995).
These conflicts often led to legitimation of some forms of protest and stigmati-
zation of violent ones.

Downloaded from https:/www.cambridge.org/core. University of Sussex Library, on 03 Mar 2017 at 14:29:18, subject to the Cambridge Core terms
of use, available at https:/www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139043144.011
 Conclusions 285

However, the process of depolarization had different timings, twists, and

turns in different cases; the differences were related not so much to the class,
ethnic, or religious identities of the violent actors but rather to the characteristics
of the political regimes they addressed. In fact, the escalation of violence went
much farther in nondemocratic countries. In authoritarian Franco’s Spain and in
the Middle East, hard repression was usually unable – at least in the short term –
to demobilize protests, contributing instead to a radicalization that was all the
more challenging because it went along with increasing popular support, which
often included a sort of temporary territorial control of some areas by the rebels.
However, as the Spanish case indicates, violence can develop as opportunities
are opening up, within a process of liberalization and transition, as a means of
negotiation and/or as a reaction to the perceived frustration of hopes for a
quicker and deeper democratization process.
I also focused on the mechanism of competitive escalation. Social movement
studies have linked radicalization to the development of protest cycles. Whereas
waves of protest often bring about a normalization of once-unconventional
forms of protest, as well as a civilization of its form, an immediate outcome is
often the development of some violent forms of action. These tend to change
along the cycle: primarily occasional and defensive in the beginning, they
become increasingly organized and ritualized. Toward the end of the cycle,
although the number and size of protest events decline, clandestine forms of
violence develop. One of the reasons I identified for this escalation is organiza-
tional competition within dense milieus of social movements, social movement
families (made up of social movements that share some general orientations and
are often allied), and also broader social movement sectors involving a plurality
of social movement families.
The protest cycles from which the clandestine organizations examined grew
originated not only from within existing collective actors but also through
internal contestation: the PCI for the Italian left-wing groups, the MSI for their
right-wing counterparts, the PNV for the ETA, and the MBs for the Islamist
groups were such original actors, which the emergent social movements con-
tested as too tame, if not traitorous.
Although protest cycles bring about the emergence of large numbers of social
movement organizations that tend to cooperate with one another during the
peak of protest, the decline of mobilization produces conflicts about the best
strategies and tactics to be used to overcome the perceived crises. On the left as
on the right and in the case of the ETA and Islamist fundamentalism, we saw
internal strategic struggles, which ended with the more or less rapid and intense
radicalization of one part of the group in question and the moderation of
another. Additionally, cycles of protest stimulated the emergence of counter-
movements, often involving physical conflicts between militants of different
fronts.
In all of the covered cases, experimentation with violent tactics emerged from
attempts to outbid the other groups – attracting the residual militants – and

Downloaded from https:/www.cambridge.org/core. University of Sussex Library, on 03 Mar 2017 at 14:29:18, subject to the Cambridge Core terms
of use, available at https:/www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139043144.011
 286 Clandestine Political Violence

through small everyday adaptations to the tactics of adversaries. Organizational

competition influenced the radicalization processes. The choice to use radical
forms and action was a sort of “slight product differentiation (offering margin-
ally different goals) and, especially, tactical differentiation” (McCarthy and
Zald 1973: 6). Addressing different constituencies, various movement organi-
zations targeted their strategic choices to make themselves more attractive to
their audience. Especially during the declining phases of mobilization, violence
became a trademark, designed to attract attention in the radicalized movement
groups.
In these adaptations of practices, activists were slowly socialized to violence,
while at the same time the organizations adapted their structures – for instance,
through the creation of marshal bodies, devoted first to defense but later to
attack. In all cases, structures that specialized in violent repertories developed
slowly, during fights with political adversaries and the police, until occasions of
violence or repression, in particular, pushed their members underground.
Whereas in the Italian case street fights involved left- against right-wingers, in
the Basque Country ethnonationalists competed (although not necessarily physi-
cally) with a class definition of the conflicts, and in the case of the Islamists, harsh
(sometimes physical) struggles pitted left-wing social movement organizations
against nationalist and religious ones.
The activation of militant networks is another mechanism I have singled out.
Rather than psychological characteristics and beyond heterogeneous social
background, what militants who joined clandestine organizations had in com-
mon was their belonging to networks of friends-comrades. Social science liter-
ature on social movements (and other subjects as well) has time and again
stressed the role of social networks for recruitment in contentious politics. The
cases examined fit well in Doug McAdam’s model of recruitment to high-risk
and high-cost activism (1986: especially 68–71). According to this model, fam-
ilies or other socialization agencies play a role in making individuals receptive to
certain political ideas; when individuals who have thus become politically sensi-
tive encounter political activists, they are then motivated to become involved in
an initial low-cost and low-risk activism. Under conditions of “biographical
availability” – that is, for instance, for young people – “these ‘safe’ forays into
activism may have longer-range consequences … for they place the new recruit
‘at risk’ of being drawn into more costly forms of participation through the
cyclical process of integration and resocialization” (McAdam 1986: 69). For the
activists in the groups discussed herein, as for the Freedom Summer activists
McAdam studied, this process involved an increasing integration into activist
networks that formed a “culture of solidarities,” as “a cultural expression that
arises within the wider culture, yet which is emergent in its embodiment of
oppositional practices and meanings” (Fantasia 1988: 17). Cultures of solidarity
emerge “in those moments when the customary practices of daily life are
suspended and crisis requires a new repertoire of behavior, associational ties,
and valuations” (ibid.: 14). For the activists, the reality of everyday life – the

Downloaded from https:/www.cambridge.org/core. University of Sussex Library, on 03 Mar 2017 at 14:29:18, subject to the Cambridge Core terms
of use, available at https:/www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139043144.011
 Conclusions 287

“reality par excellence,” in which “the tension of consciousness is highest”

(Berger and Luckmann 1966: 21) – created the conditions for a gradual accept-
ance of violence as a political means.
If networks are important for most types of activities, the challenge is to
specify which networks are conducive to commitment in clandestine organiza-
tions. In this text, the four types of clandestine violence differ in terms of the
relevance of the specific movement milieus – leftist, rightist, ethnonationalist,
and religious, respectively – from which they derived their recruits.
However, there were also similarities. In all cases, the organizations recruited in
groups that had already undergone a process of radicalization, and blocs of
recruits often came from a process of splitting within the most radical wings of a
social movement family. Intensive affective relations were typical of these nets. In
all cases, networks were not only exploited but also produced by the radical
groups. As Elisabeth Wood (2003) observed on different forms of political com-
mitments, social networks are created in action. This also explains why, in all four
types of clandestine organizations, I found two (or even more) generations.
The first generation grew inside long-standing social movement traditions,
nurtured in the red subculture of the Left, in nostalgic milieus in the case of
neofascists, in nationalist communities in the case of the ETA, and in specific
religious enclaves in the case of Islamic fundamentalists. Family ties were partic-
ularly relevant, sometimes materially but more often symbolically. Recruitment
or, better, the foundation of clandestine organizations proceeded mainly within
the political environments in which the militants had been socialized at a very
young age.
Militant networks developed in small and radical groupings in which political
commitment and friendship mutually strengthened each other. Relays were
constituted in action and influenced the perceptions of the activists as well as
intensifying their emotional attachment, and therefore the peer-group pressure.
Grievance and greed were not (predominant) preexisting predispositions; rather,
they increased in action.
Once active, the clandestine organizations themselves contributed to the
constitution of radicalized milieus, from which they then recruited a second
generation of activists. When the founders were arrested or forced into exile,
new and more radical recruits took over their roles. They grew during the
escalation to which clandestine organizations contributed and were socialized
to violence very early on, almost skipping nonviolent forms of politics. These
second generations were described as more violence prone.
Once underground, clandestine organizations underwent a process of implo-
sion, in which interactions with the outside were reduced. To illustrate this
process, I singled out mechanisms of organizational compartmentalization,
action radicalization, and ideological encapsulation.
In terms of organizational compartmentalization, the organizational struc-
tures of the four types of clandestine groups showed some differences as well as
some similarities. Clandestine organizations of each type took inspiration from

Downloaded from https:/www.cambridge.org/core. University of Sussex Library, on 03 Mar 2017 at 14:29:18, subject to the Cambridge Core terms
of use, available at https:/www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139043144.011
 288 Clandestine Political Violence

the organizational repertoires (Clemens 1996) that characterized their social

movement families, adapting them to a hostile context. This produced a different
balance of hierarchical and network structures, as well as different functional
internal divisions that reflected, for example, the importance of the factory for
the left-wing groups or of religious issues for the Islamists. In this sense, ideas and
interests interacted in decision making. In addition, collective structures tended
to reflect the organization’s size in terms of members and resources.
However, the groups exhibited a similar cross-type evolution toward compart-
mentalized structures. Faced with rising repression and declining support, the
clandestine organizations became more hierarchical, giving up on attempts to
host large meetings and increasing the (formal) power of a few leaders. Like in
organized crime, however, the aspiration to effective centralization and hierarch-
ical control met with a reality that was more centrifugal. To begin with, clandes-
tinity imposed a reduction of the role of organizational structures open to
sympathizers and of the role of legal (rather than clandestine) militants, while the
various cells became increasingly independent from one another and also difficult
to control from above. In this process, the very difficulty of intervening in social
conflicts increased the relevance of the military over the political organizational
bodies. But being military oriented did not mean being obedient: in fact, in all of the
cases factions formed from and fought against one another, in a never-ending
process of splitting into ever-tinier units and, sometimes, bloody internal purges.
Similar remarks apply to the analysis of the next mechanism: action militar-
ization. In this case as well, the comparative analysis has shown that, when
choosing targets and forms of action, clandestine organizations followed some
normative preferences. Action strategies were indeed not only assessed instru-
mentally but also constrained by group norms. The shift from actions against
property to actions against people, and then from wounding to killing, produced
tensions among the left-wing radicals. In the radical Right, massacres of ran-
domly selected people were stigmatized by a new generation of militants. In the
ETA, this was the case with bombings in public places, especially when issued
warnings did not succeed in avoiding victims. For the Islamist groups as well, the
killing of civilians, particularly fellow Muslims, raised opposition. In all groups,
creating victims among people considered as innocent (or at least noncommit-
ted) led to internal criticism – as did violence used internally, to punish with-
drawal or “betrayal.” Suicide missions, even if they might have been effective,
were in fact used by only a few groups in extremely radicalized conflicts; for most
groups, such tactics were not even in the realm of possibility.
However, I also noted a process of growing detachment from action aimed at
propaganda and an increasing tendency toward action oriented to mere orga-
nizational survival. As repression increased organizational isolation, as well as
the acceptance of violence, the so-called repressive apparatuses (police, army,
judges, and so on) became the main targets of clandestine action. The logic of
action became increasingly military and decreasingly political. In fact, the nor-
mative constraints against the most brutal forms of action were overcome as a

Downloaded from https:/www.cambridge.org/core. University of Sussex Library, on 03 Mar 2017 at 14:29:18, subject to the Cambridge Core terms
of use, available at https:/www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139043144.011
 Conclusions 289

result of the militants’ search for a certain type of reputation as soldiers and
heroes. In a vicious circle, however, this reputation started to damage rather than
advantage the clandestine groups. Often (in the case of the ETA and especially in
Egypt), territorial control meant attempting to force the population into certain
types of behavior, as well as at extracting resources, and this in turn reduced
support. As they became increasingly isolated, the clandestine organizations
tended to target the very social and political groups they had previously tried
to attract: the Left, the Right, the Basques, and Muslims, respectively. The pace
of the process interacted with the degree of radicalization of existing conflicts
and its cultural effects in terms of tolerance for violence.
These developments also interacted with still another mechanism: ideological
encapsulation. As aforementioned, political violence was normatively justified,
as radical beliefs were at the same time preconditions and, especially, effects of
violent actions. In general, all narratives described a path from a glorious past, to
a long decadence, and then to a rebirth. Dichotomous visions, a sense of moral
superiority, and essentializing thinking all developed in action. Justification
follows escalation, which is only to a certain extent strategically planned.
Rather than emerging from preexisting ideologies, violence developed with
repression and competition.
As for the previous phenomena, we noted some differences in the narratives
initially adopted by the clandestine organizations as they embedded their dis-
courses in the broader cultures of the social movements they wanted to address. So
the Italian Left stressed resistance and revolution; the Right revived the fascist
spirit; the ETA built on the Basque mythology; and the Islamist groups went back
to specific trends in the interpretation of religious texts. All types of clandestine
organizations, however, shared a certain path toward a narrative that became less
resonant with those of the social movements they wanted to influence. Adapting
their discourse to the organizational compartmentalization and action militariza-
tion, they changed their definition of themselves from (effective) soldiers to
(defeated) martyrs. In all four types, the self-justification became ever more elitist,
depicting an image of heroic – if not successful – fighters. And to justify ever more
cruel forms of action, clandestine organizations constructed an image of an
absolute evil, whose cognitive borders grew ever broader.
In a vicious circle, the more isolated the organizations became, the more they
withdrew from attempts at bridging their frames with those of activists in
potentially sympathetic environments, developing instead a self-contained and
self-referential narrative. Whereas it was initially justified instrumentally as the
only way out against a powerful adversary, violence then increasingly became an
existential response to a hostile environment – from consequentialism there was
a move to deontological justification. Therefore the Marxist-Leninist, neofascist,
exclusive nationalist, or Islamist fundamentalist ideologies – which had been
available for ages – were not the direct causes of the waves of clandestine
violence. Rather, they were twisted and transformed through the process. Even
the language changed, becoming less understandable from outside.

Downloaded from https:/www.cambridge.org/core. University of Sussex Library, on 03 Mar 2017 at 14:29:18, subject to the Cambridge Core terms
of use, available at https:/www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139043144.011
 290 Clandestine Political Violence

Similarly, once the groups went underground, members’ commitments were

kept through mechanisms of militant enclosure. In some cases, the group slowly
transformed from a source of support to a prison: “Participants in some social
movements find an oppressive and stifling side to close-knit personal relation-
ships. So-called cultural free spaces sometimes become prisons from which some
participants would like to escape but cannot because they lack the courage to
defy the group censure and ostracism that would follow” (Gamson 1992: 64). A
sort of affective enclosure followed the everyday experiences of the militants in
their environment. Political socialization in clandestine organizations involved
in fact “a transformation in the reference system, that is, a gradual process of
individual exit from the majoritarian culture and integration in a political
counterculture with divergent norms, values and loyalties, and a particularly
rigid pressure to conform” (Jäger and Böllinger 1981: 232). It has been observed
that the clandestine organization
provides back-up when other support is eradicated. Within the group several psycholog-
ical variables become essential: solidarity, complicity, and reality perception. Group
membership means obligatory acceptance of a certain system of values and norms;
deviations from these are punished. The group dynamics of complicity catalyze actions
that any one member would hardly have been able to accomplish alone and that he has
difficulty understanding later on. He acts within a set of mutual expectations and role
assignments. Being forced into the underground, the group lives in isolation and, [its
members] working in close cooperation, evolves new models for the interpretation of
reality that acquire a binding character for the group members. (Rasch 1979: 82)

Affective ties among friends-comrades in the underground are notable for their
intensity,1 as participation requires broad changes in the individual’s value
system and behavior, or, in other words, alterations that
resemble primary socialization, because they have radically to reassign reality accents
and, consequently, must replicate to a considerable degree the strongly affective identi-
fication with the socializing personnel that was characteristic of childhood. They are
different from primary socialization because they do not start ex nihilo, and as a result
they must cope with a problem of dismantling, disintegrating the preceding nomic
structure of subjective reality. (Berger and Luckmann 1966: 157)

Thus “people and ideas that are discrepant with the new definitions of reality are
systematically avoided” (ibid.: 159).
In addition, group identification increased with level of risk. McAdam’s
Freedom Summer participants described their experiences as a sort of “ecstasy,”
a “sense of liberation,” a feeling of being “finally at home,” and a “transcendent
experience”: “These people were me and I was them” (McAdam 1988: 71).
Similarly, for militants of the underground, as for those in other high-risk secret

1
Movements are said to differ from one another in terms of “the emotional tenor, presence, affective
climate, or demeanor enacted and communicated by movement organizations” (Lofland 1985:
219).

Downloaded from https:/www.cambridge.org/core. University of Sussex Library, on 03 Mar 2017 at 14:29:18, subject to the Cambridge Core terms
of use, available at https:/www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139043144.011
 Conclusions 291

societies (Erikson 1981), the excitement of shared risks strengthens friendships.

In fact, the need for secrecy tends to become the most important determinant of
the organization’s structure and strategy (ibid.).
Although at the beginning of their careers militants recruited in the under-
ground were embedded in broader networks, affective and cognitive dynamics
interacted to produce a progressive encapsulation of the small circles of friends-
comrades. From the affective point of view, identification with a group of
“heroic” peers increased at the same pace as the severing of other ties outside
the group. For individuals in the underground, the comrades active in their own
organizations became their only source of material and emotional support. This
does not means that life in the underground was only rosy, funny, and exciting:
small-group dynamics also brought tensions and breaks, increasing stress in not
only political but also everyday life. The experience was certainly intense, leav-
ing little space for other relations – which were in fact actively discouraged.
From the cognitive point of view, the (increasingly tinier) groups became the
only trusted source of information for their members. Justifications for violence
were built on not only emotional identification with dead or imprisoned com-
rades but also the construction of a special vision of external reality. At the
individual level, this reflected the increasing encapsulation of the organizational
narrative that I have already described. Within totalitarian types of institutions,
as in clandestine ones, cognitive closure produced an increasing detachment
from shared visions as the pressure of cognitive coherence deterred members
from opening up alternative channels of communication.
These mechanisms developed in part as reactions to external challenges
(see J. H. Jackson and Morgan 1978), but in progressively hostile environments,
strategic choices were limited. In fact, perverse effects, neither planned
nor foreseen, strongly influenced the evolution of these organizations, whose
decline was often related to the very choice of clandestinity. The unanticipated
results of this and subsequent choices, addressing different problems of survival,
reduced the range of options available to the group. In general, tactical
transformation required adaptation in the organizational structure as well as
symbolic changes, which in turn had unpredictable effects that required new
adaptations. As I had observed with reference to the Italian and German left-
wing radical groups,

unable to avoid arrest and alienation from the external reality, they were drawn deeper
and deeper into a sort of spiral in which each successive turn further reduced their
strategic options. The very condition of clandestinity drew the organization into a kind
of vicious circle in which each attempt to face problems at one level produced new
difficulties at another. As a result, the organizations had to abandon externally oriented
aims for a “private war” with the state apparatuses. That is, operating illegally as they
did, the militants of the armed struggle could not appear at the site of social conflicts, and
this physical distance led to a kind of psychic distance as well. It reduced the terrorists’
capacity to pursue effective propaganda strategies. Abandoning their propaganda efforts,
they concentrated their energies on their struggle with the state and became increasingly

Downloaded from https:/www.cambridge.org/core. University of Sussex Library, on 03 Mar 2017 at 14:29:18, subject to the Cambridge Core terms
of use, available at https:/www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139043144.011
 292 Clandestine Political Violence

involved in their private war, an obsession that isolated them still further. And the more
isolated they became, the weaker was their capacity to escape repression. (della Porta
1995: 135)

These internal dynamics became increasingly relevant once the radical groups
went underground. Even if they continued to try to adapt their strategies to the
changing external reality, the moves they had made to clandestinity drastically
reduced their range of possible choices and further weakened the group’s sense
of reality. In fact, the very choice of clandestinity forced them into a losing
military conflict with the much more powerful state apparatuses. Clandestinity
by definition entails material and psychological isolation, and the distinctive
spiraling pattern of radicalization and isolation characteristic of semi-illegal
groups only accelerated when the groups went underground. As their action
became increasingly brutal, most radical groups lost the (large or small) external
support they had received when they were first organized. Each successive turn
in this spiral reduced the groups’ strategic options, making them a prisoner of
their own version of reality. Entrepreneurs of violence thus unleashed a force
they could not control: embracing violence, they cultivated the source of their
own dissolution.
Whereas environmental conditions were relevant at the onset, however, there
was also an agency power of radical organizations, which themselves repro-
duced the resources for their survival. In a vicious circle, as groups adapted to
radical resources in their environment, they helped to perpetuate them. The
process developed through trials and errors, advances and retreats, in which
groups experimented with different forms of action and organizational formats
and then justified them through frames of amplification. In addition, as skills for
violence were gradually formed, those who possessed those skills then played a
role in spreading them.
Radicalization can in fact be seen as a good example of the evolution of
“absurd” chance processes or vicious circles, characterized by spirals of neg-
ative feedback that produce different effects from what was planned. In these
processes, participants operate based on a self-constructed image of reality,
gambling on the results of the choices made (Neidhardt 1981: 245, 251–2).
The final outcome of their actions results from a chain of actions and reactions,
based on miscalculations of the moves of the different actors: “This circle of
actions and reactions forms a routine until a more or less chance event ruptures
the pattern and produces a qualitative jump, the group debates its possible
choices, and in this crisis some members decide to go underground” (della
Porta 1995: 111). In vicious circles, negative feedback loops can actually
produce results that are the opposite of those expected (see, for instance,
Masuch 1985: 14–5; see also Merton 1957). The choice of clandestinity evolved
gradually and during long processes; it was only in part premeditated, and not
irreversible. Radicalization was therefore the result of not only strategic choices
but also unplanned internal dynamics. It showed not only attempts at strategic

Downloaded from https:/www.cambridge.org/core. University of Sussex Library, on 03 Mar 2017 at 14:29:18, subject to the Cambridge Core terms
of use, available at https:/www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139043144.011
 Conclusions 293

action but also their limits. Collective choices also emerged as short-sighted,
featuring some short-term advantages but also disastrous consequences in the
medium and long term. Semimilitary units were created to organize violent
practices, but they then privileged military action and split from the main
organization. Choosing clandestinity, they avoided immediate repression but
also reduced their capacity to speak to their constituency, increasingly reducing
their contacts with the outside. For these specific organizations, the tendency
toward moderation was hindered by reliance on ideological incentive and
strong group solidarity (Zald and Ash 1966).
The timing of this trend toward dissolution was influenced by the environ-
mental conditions that influenced the degree of isolation of the underground
groups. The higher the support for the use of violence in the environment, the
slower the process of implosion tended to be. At the same time, however, the
presence of resources for violence in the environment – sometimes including a
territorial control over some areas – pushes clandestine organizations toward the
use of brutal forms of violence.
Throughout the presentation of these various mechanisms, I have noted
similarities as well as some differences. Although I stated that I am not interested
in lawlike statements, some caveats must be mentioned if one is to move beyond
the internal validity of the description of the selected cases to address the study’s
external validity – that is, its potential generalization to other cases within a
broader geographical and historical range.
First, although I tried to make the most of my past fieldwork on several of the
cases covered, the research design is a cross-national historical comparison, based
in good part on published materials. As such, it inherits all the richness and
challenges of historical sociology, as I have mixed (various amounts of) firsthand
empirical evidence with secondary analysis of the literature. Even though I tried to
cover as many studies as I could, I encountered obvious difficulties in obtaining
complete information on all of my cases. More focused empirical fieldwork in the
future can enrich the analysis of each of the mechanisms I singled out.
Second, although I have selected crucial cases of left-wing, right-wing, ethno-
nationalist, and religious types of clandestine political violence, their represen-
tativeness is certainly questionable. In particular, most of my cases are
European, and, although they address different waves of protest, they still
cover no more than a half century. In the course of the volume, I have system-
atically compared the Italian and German left-wing clandestine groups but
barely mentioned the Weather Underground in the United States or the
Japanese Red Army, which seem to share some of their characteristics
(Zwerman, Steinhoff, and della Porta 2000). I have also introduced occasional
references to works on the American radical Right (Wright 2007), the Irish
ethnonationalist conflict (Bosi 2006; Waldmann 1998; White 1993), and the
Algerian conflicts, all of which appear to confirm the presence of some of the
mentioned mechanisms. However, more systematic comparison is clearly
needed before one can claim too much about the generalizability of my results.

Downloaded from https:/www.cambridge.org/core. University of Sussex Library, on 03 Mar 2017 at 14:29:18, subject to the Cambridge Core terms
of use, available at https:/www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139043144.011
 294 Clandestine Political Violence

Third, the research focused on cases of conflict radicalization that ended

in clandestine political violence. Future research that compares positive cases,
in which clandestine violence developed, to negative cases, in which clandestine
violence did not develop, notwithstanding the presence of some escalation,
would represent a further step in our knowledge. As observed by McAdam,
Tarrow, and Tilly (2001: chap. 7), many revolutionary situations do not actually
develop into revolutions. In parallel, of the many self-determination movements,
only a few escalate into violent forms (for example, in 2006, 28 out of 168 such
movements escalated, according to data from the Center for International
Development and Conflict Management, reported in Hewitt, Wilkenfeld, and
Gurr 2007). And there are many situations in which harsh repression of intense
cycles of protest and organizational competition help the spread of some violent
repertoires, but events stop short of the foundation of clandestine organizations.
Among the cases I have studied, the contestation over the use of locally
unwanted land in Italy in the 2000s (della Porta and Piazza 2008); the transna-
tional protests in the early 2000s (della Porta 2007); the French riots of the past
decades (della Porta and Gbikpi 2011); the activities of the radical Right in Italy,
Germany, and the United States (Caiani, della Porta, and Wagemann 2011); or
the politicization of religious conflicts in Italy in the 2000s (Bosi and della Porta
2010) could be good candidates for a systematic comparison of a universe of
positive and negative cases.
Rather than exhausting the study of clandestine political violence, I hope in fact
that my work can stimulate it. In particular, during a time in which every new
wave of clandestine political violence is presented as peculiarly evil and irrational,
I believe that my analysis has provided enough evidence on the existence of similar
mechanisms to stimulate comparative analysis among different types of violence.

Downloaded from https:/www.cambridge.org/core. University of Sussex Library, on 03 Mar 2017 at 14:29:18, subject to the Cambridge Core terms
of use, available at https:/www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139043144.011