Vous êtes sur la page 1sur 38

Terror and Guerrilla Warfare in Latin America, 1956-1970

Author(s): Timothy P. Wickham-Crowley

Source: Comparative Studies in Society and History, Vol. 32, No. 2 (Apr., 1990), pp. 201-237
Published by: Cambridge University Press
Stable URL: http://www.jstor.org/stable/178913
Accessed: 18/11/2008 01:42

Your use of the JSTOR archive indicates your acceptance of JSTOR's Terms and Conditions of Use, available at
http://www.jstor.org/page/info/about/policies/terms.jsp. JSTOR's Terms and Conditions of Use provides, in part, that unless
you have obtained prior permission, you may not download an entire issue of a journal or multiple copies of articles, and you
may use content in the JSTOR archive only for your personal, non-commercial use.

Please contact the publisher regarding any further use of this work. Publisher contact information may be obtained at

Each copy of any part of a JSTOR transmission must contain the same copyright notice that appears on the screen or printed
page of such transmission.

JSTOR is a not-for-profit organization founded in 1995 to build trusted digital archives for scholarship. We work with the
scholarly community to preserve their work and the materials they rely upon, and to build a common research platform that
promotes the discovery and use of these resources. For more information about JSTOR, please contact support@jstor.org.

Cambridge University Press is collaborating with JSTOR to digitize, preserve and extend access to
Comparative Studies in Society and History.

Terror and Guerrilla Warfare
in Latin America, 1956-1970

Georgetown University

Now is the timewhenterrorwill be usedagainstthepeasantsby bothsides,butwith

--Che Guevara'
Most of the extraordinarywaves of terror which have swept many Latin
American societies since 1970 have occurredin guerrilla-basedinsurgencies
or even civil wars. Because of the massive body counts producedduringthese
confrontationsbetween revolutionariesand counterrevolutionaries based in or
linked with a government, human rights organizationshave issued a long
series of reportsabout terror-especially that which has been carriedout by
incumbentregimes and death squads-and which has been supplementedby
the expos6s of the guerrillasthemselves. Amnesty International,the Human
Rights group in the Organizationof American States (OAS), and Americas
Watch have been the major internationalactors documenting the wave of
terror. Many independentnational groups, such as El Salvador's "Socorro
Juridico" and other human rights organizationslinked with church bodies
have undertakenthat more perilous task at home.
Terroragainst the civilian populationdid not begin twenty years ago, as it
had pervadedthe insurgenciesof the 1950s and 1960s throughoutLatinAmer-
ica as well. The "bone-heaps" of the earlier wave of terrordo not match
those of the later surge in size, to be sure, and I will try to accountfor those
differences later in this paper. Nonetheless, the very obscurityof the earlier
terror morally and intellectually compels us to address it. The earlier Latin
American guerrillastruggles-more so than the "Prolonged PopularWars"
characteristicof guerrillawarfaresince 1970-were "Wars in the Shadows,"
as one authorput it,2 and the time has come to bring light to those shadows,
lest they be forgotten.
My purposehere is not solely historiographical,even thoughmany sources
here are known to and employed by the few chroniclers of these events;
I Emesto "Che" Guevara, The
CompleteBolivian Diaries of Che Guevara and Other Cap-
tured Documents, Daniel James, ed. (New York: Stein and Day, 1968), s.v. "End of May,
2 RobertAsprey, Warin the Shadows: The Guerrilla in History, 2 vols. (GardenCity, N.Y.:
Doubleday, 1975).
0010-4175/90/2120-0596 $5.00 ? 1990 Society for ComparativeStudy of Society and History


instead, my aim, which is ultimatelyboth sociological and theoretical, is to

employ the informationthatwe can retrieveaboutguerrillawarfarein orderto
arrive at a better understandingof the causes, concomitants, and natureof
terror in guerrilla wars. As Robert Merton has cautioned us, we should be
quite sure, however, that we are accounting for facts, not their deceptive
counterparts,before we seek to explain any social phenomenon. Since so
much of the reportingon terroris ideologically freightedand tendentious,that
caution will carry special weight herein.
I will reporton six cases here:(1) Fidel Castro'sown insurgencyagainstthe
Batista regime from 1956 to 1959, which began with a boat landing in the
Oriente Province and ended with a triumphalmarch into Havana;(2) The
Venezuelan insurgencyof the Communist-backedArmed Forces of National
Liberation (FALN), and the Movement of the RevolutionaryLeft (MIR),
against elected Acci6n Democrdticagovernmentsin the 1960s; (3) the Guate-
malan insurgencyof the 1960s, involving the Communists(PGT) and one or
two-there was fission and refusionhere-guerrilla groups, the Thirteenthof
November RevolutionaryMovement (MR-13) and the Rebel Armed Forces
(FAR); (4) three different Colombian insurgentgroups posing revolutionary
challenges to the coalition National Front governments in the 1960s: the
Moscow-linked Colombian RevolutionaryArmed Forces (FARC); the Cas-
troist Army of National Liberation(ELN); and the Maoist Army of Popular
Liberation (EPL). All four of the preceding guerrilla movements secured
substantialpeasant supportin their areas of operationduringthe insurgency.
The remainingtwo nationsexperiencedinsurgenciesin which the guerrillas
secured at best moderate(and at worst no) supportfrom regionalpeasantries:
(5) Peru, in which two left-wing political splinter groups unleashed four
simultaneousfocos (guerrillabands) in the Andes in 1965 againstthe elected
governmentof FernandoBelaundeTerry:the MIR (Movementof the Revolu-
tionaryLeft) and the ELN (Army of NationalLiberation)(all four were elimi-
nated in less than a year); (6) finally, therewas Che Guevara'sfoco in eastern
Bolivia, which providedat best a weak challenge to the popularlyelected and
peasant-supportedregime of General Rene Barrientos. At the end of the
insurgency, in October 1967, Guevarawas wounded,captured,and killed.


Government Terror
We will consider terror to be certain acts forbidden by the rules of war.
Among these are: (1) beating, killing, robbing, bombingor otherassaultson a
civilian population, including relatively unusualitems such as forced reloca-
tion; (2) beating, torturing, or killing or combatantswho have indicated a
willingness to surrender;(3) the use of weapons which do not sufficiently
discriminateamong combatantsand others. Such weapons include germ war-

fare, nuclear weapons, and punji stakes. The latter were used by the Viet
Cong in Indochina, where a high percentageof casualties were the result of
people falling into pits filled with these sharpened, dung-covered stakes.3
First, some preliminaryconceptualdistinctionsmay help us to sort out the
empirical accounts to follow in our analysis. Eugene Victor Walterhas done
the yeoman work in developing a systematic theory of terror. Walter dis-
tinguishes between a regime of terror(by a government)and a siege of terror
(by the opposition to a government).Eitherof these encompassesthe process
of terror, which has three distinct elements: (1) the violent act itself; (2) the
victim of the violent act; and (3) the targetof the violent act. This last element
is fundamentalto any system of terror,for the basic aim of terroris not to kill
individuals but to frighten entire social groups.4 Under these criteria, the
Cubanpeople certainly sufferedintensive terrorfrom 1953 to 1958, although
the extent has been overestimatedby Fidelistas. Hubermanand Sweezy ap-
parentlyfirst gave the total of 20,000 deaths duringthe Cuban insurrection,
which was a figure that they attributedto Castro. This numberwas quickly
converted-a typical occurrence in the building of revolutionary my-
thology-into the killing of 20,000 innocent civilians by Batista.s These
figures appearto have no factual basis, and there is good reason to lower the
total numberof deaths by a factor of ten. A list of the war dead publishedin
Bohemia on 11 January1959 (after Castro's victory) counted 898 dead, with
over half of these being combatants. These figures exclude the deaths of
peasants, which probablynumberedseveral hundred.6Estimatesof hundreds
or perhapsabout a thousanddeaths due to Batista's terrorare also supported
by comments made by Fidel Castro and other Batista critics during the war
itself.7 The figure of 20,000 apparentlyis a giganticballoon blown up by anti-
Batista emotions.
Batista's terror was especially evidenced by the faithfully fulfilled "no
prisoners" rule. Following the Moncadaattackof 26 July 1953, Castrohim-
self was only saved from summaryexecution by a lieutenantwho knew him
from the university. Later, both an attemptedlanding of guerrillareinforce-
ments and a naval uprising at Cienfuegos in 1957 ended in the same fashion:
3 See James E. Bond, The Rules of Riot: Internal Conflict and the Laws of War (Princeton,
N.J.: Princeton University Press, 1974), for a full discussion of the rules of war and possible
applicability to guerrilla warfare.
4 Eugene Victor Walter, Terror and Resistance (New York: Oxford, 1969), Chs. 1, 2. For
some less formal observations about the social functions of terror, see Rogger Mercado, Las
Guerrillasdel Peru (Lima:Fondo de CulturaPopular, 1967), 160-1, and H6ctorB6jar, "Ne pas
surestimerses forces," Partisans (Paris), 38 (July-September 1967), 111.
5 Leo Hubermanand Paul Sweezy, "Cuba: Anatomy of a Revolution,"
MonthlyReview, 12
(July-August 1960), 29.
6 Hugh Thomas, Cuba: The Pursuit of Freedom (New York: Harperand Row, 1971), 1044;
Boris Goldenberg, The Cuban Revolutionand Latin America (New York: Praeger, 1965), 144.
7 Fidel Castro Ruz, RevolutionaryStruggle: Vol. 1 of The Selected Works Fidel Castro,
Rolando E. Bonachea and Nelson P. Vald6s, eds. (Cambridge,Mass.: MassachusettsInstituteof
Technology Press, 1972), 375, 399; Thomas, Cuba, 972, 999.

All those who surrenderedwere shot on the spot.8 Batistalackedthe airpower

with which to inflict severe casualtiesupon the SierraMaestrapeasantry,but
deaths did result from his bombing of the region. These are, however, unim-
portant relative to the number of deaths caused by direct troop attacks on
peasants. Many of these attacks did not even have the pretext of seeking
information,as those armyunits whose commanderswere afraidto go into the
guerrilla zones would simply attack peasant villages on the outskirtsof the
zone and then report the number of "guerrillas" killed. One particularly
brutalofficer in the Sierra Maestraregion, Lt. Casillas, who literally drove
the peasantryfrom the village of Palma Mocha, kept humanears in a box to
show to visitors.9 Batista took a small page from the book of strategy by
Spanish General Weyler during Cuba's War for Independencein the 1890s,
when he, Batista, also forced the evacuationof several hundredSierraMae-
stra peasants(Weyler's programhad aimed at 500,000 forced relocations).10
As in almost all of the cases underreview, much of the governmentterror
took place as the torture of urban cadres, innocent victims, and peasants
transportedto town jails and prisons. Those who lived to tell such tales can
relate grisly stories. Hayd6e Santamariawas shown the extractedeye of her
brotherin an attemptto make her informon the M-26 (Castro'sTwenty-Sixth
of July Movement) operationsin Santiagode Cuba.1 While we will usually
restrict our examples to the terror that takes place in rural settings, these
broaderaspects should be understood.
The Venezuelancase providesus with a greatmany examples of terror,but
the sheer numbershould not overly impressthe reader.Dead men indeed tell
no tales, and the high survivalrateof Venezuelanguerrillashas led to a fairly
extensive literatureof interviews and memoirs. The exceptional numberof
surviving guerrillas suggests that terror against guerrilla combatantsthem-
selves was considerablymore muted than in Cuba, even though it was still
common against the peasantryduring army sweeps throughguerrillazones.
An author very unfriendly to the governments of Romulo Betancourt
(1959-64) and Raul Leoni (1964-69) accuses them of causing "more than
200" peasantdeaths due to such tactics,; while anotherhostile source claims
that more than 1,000 persons were killed overall under Betancourtalone,
suggesting again a very strong imbalance in favor of urbanterror.The vast
bulk of the slayings of the peasants-a variety of sources would place the
numberbetween 100 and 300-were in the states of Falcon and Lara.12
8 Ram6n L. Bonachea and Marta San Martin, The Cuban Insurrection, 1952-1959 (New
Brunswick, N.J.: Transaction, 1974), 134-7, 151-2.
9 Bonachea and San Martin,The CubanInsurrection,270, 192, 276; GiinterMaschke, Kritik
des Guerillero (Frankfurt,Germany:S. Fischer, 1973), 85; Thomas, Cuba, 920.
10 Thomas, Cuba, 335; Ram6n BarquinL6pez, Las Luchas Guerrilleras en Cuba, 2 vols.
(Madrid:Plaza Mayor, 1975), vol. 1, 14-15; Newsweek, 17 June 1957, 60.
11 Such tales can be told only when thereare survivors.This and other stories can be found in
Carlos Franqui,The Twelve (New York: Lyle Stuart, 1968).
12 Luigi Valsalice, Guerriglia e Politica: L'esempio del Venezuela(Florence, Italy: Valmar-

Guerrillazones were bombed regularlyin Venezuela, usually just before

the armybegan sweeps or encirclementcampaignsin the regions. While some
of the peasantsallegedly died in such bombings, evidence is fragmentary.The
army would at times force the evacuation of the peasants from villages and
then bomb the area. Luigi Valsalice describedsuch evacuationsas "more or
less voluntary."13An example of the latteris seen in the following talk given
by an army officer in a village in Falc6n:
All the peasants have to leave the area because we're going to bomb
these lands. Whoever remains will be burnt to death. So get ready to leave, you
old bastards.

and later:
Youworthless,you'reall guerrillas.Getoutof here.We'regoingto burnit, to bomb
Valsalice also reports forced evacuations in Lara State.14
There is little doubt that torturetook place duringthe 1960s in Venezuelan
prisons, especially that carried out by national police agents of DIGEPOL
(General Directorate of Police), as occasional congressional investigations
revealed. DIGEPOL's public reputation was so tarnished that President
Rafael Caldera was forced to reorganize the agency after taking office in
A torturesession might begin with a warningspeech such as the following
delivered in a DIGEPOLjail cell in Falcon:

tina Editore, 1973), 135, n. 12; EfrafnLabanaCordero,TO-3: CampoAntiguerrillero(Caracas:

Ediciones Barbara, 1969), 110 ("more than 200"); Angel Raul Guevara, Los Cachorros del
Pentagono (Caracas:Salvador de la Plaza, 1973), 8 ("more than 1,000" dead [?]); Francois
Maspero, ed., Avec Douglas Bravo dans les Maquis Venezueliens (Paris: Francois Maspero,
1968), 52 (the interlocutorwith Bravo here could well be Regis Debray); Jean Larteguy, The
Guerrillas (New York: World Press, 1970), 195; Teodoro Petkoff, "Pre-Election Climate in
Venezuela (An Interview with ComradeTeodoro Petkov [sic])," WorldMarxist Review (En-
gland), 11 (April, 1968), 30. Betancourtstatedonce (see Daily Journal[Caracas,21 June 1963]),
that accordingto "certain delirious, cheap literature"the governmentwas staging "mass mur-
ders" of peasants, and rejectedsuch allegations. One shouldbe open-mindedon such mattersand
not accept these data uncritically,since the guerrillasare the usual source. This need for a critical
eye is a point forcefully driven home by the name of the original publisherof Cabieses's work
(see note 15)- "FALN publications"; see Klaus Lindenberg, "Zur Krise der Revolutionaren
Linken in Lateinamerika:Das Beispiel Venezuela," Vierteljahresberichte-Forschungsinstitut
der Friedrich Ebert Stiftung(FFES), 33 (September 1968), 288.
13 For
bombingreportsin the news, see El Nacional (Caracas),3 February1963, 4 April 1963
(both Falc6n), 30 October 1964 (Trujillo), 17 November 1964 (Falc6n), 9 June 1965 (Lara).
Bombing clearly frightened nearby peasants even when not striking their villages; see, for
example, El Nacional, 6 April 1963, and Valsalice, Guerrigliae Politica, 126. Also see Timothy
F. Hardingand Saul Landau, "Terrorism,GuerrillaWarfareand the DemocraticLeft in Venezu-
ela," Studies on the Left, 4 (Fall 1964), 125.
14 Guevara, Los Cachorros del Pentagono, 164, 166; Valsalice,
Guerriglia e Politica, 135.
15 Examples abound in Manuel Cabieses Donoso, iVenezuela, Okey! (Santiago, Chile: Edi-
ciones del Litoral, 1963). One DIGEPOLkilling of a Communistby torturewas revealed by the
Judicial and Technical Police (PTJ);see J. Carrera"The Murderersof Alberto Lovera Will Be
Punished," WorldMarxist Review (England), 9 (June 1966), 46-47.

You know who I am, don't you? To you I'm called CaptainVegas, you fucking
commie.Lookat me, you shittycommie.I amCaptainVegas,I'll be theterrorof you
faggots.Lookat me well so you'll rememberme yourentirelife.16
The tortureswhich followed, at least in a well-equippedDIGEPOLheadquar-
ters in Caracas, might consist of suction devices appliedto the fingers, elec-
tric shocks, steel rods placed with pressurebetween the fingers, a large paper
clip-type device for pulling and pinching the skin, and such other torturesas
ice picks under the fingernails, dunking the head in unflushedtoilets, lemon
juice in the eyes, standing with the arms over the head for hours, beatings
with rubbertubing, and hot candle wax applied to the skin. 17 (In ruralareas,
techniques by army patrols were cruder, but these will be discussed.below.)
The routinegoal of torturesessions was informationon the FALN guerrillas.
After the torturesessions or beatings, especially if a court appearancewas
required,the following advice might be given to those who might potentially
Carefulwhatyou say. No one has beatenyou. Andif you denyyourwrittendeclara-
tions, we'll takeanother,moreseriousjauntto the camp.18
It is perhapsmore importantthat the "hidden" terrorin ruralareas should
be placed on the historical record. Even allowing a substantialdiscount for
propaganda,the recordof ruralterroris a brutalone and is admittedby even
such a skeptic as Valsalice. Occasional reports of abuse and torture even
reached the orthodox press and the halls of Congress.19
In Falc6n, FALN leader Elias Manuitt alleged that peasants had been
publicly torturedin the Sierra del Coro, while his comrade Douglas Bravo
describeda raid on a village by thirtycommandosin which every woman was
raped, including one over forty years old, whose violation was carried out
under the eyes of her husbandand children. He also alleged that a gang rape
of a sixteen-year-oldpeasantgirl by twenty-five soldiers took place at Pueblo
Nuevo in 1963.20 As we shall see below, the common incidence of such
assaults on female members of "heretic" populationsin guerrillawar bears
closer theoretical scrutiny.
In Lara, Angela Zago's hamlet was raidedby armytroops, from whom she
16 Guevara, Los Cachorros del Pentagono, 32. The translationattempts to capture flavor
ratherthan exact denotation. Lists of killing and torturevictims of government-whatever their
accuracy-can be found in Cabieses, Venezuela,Okey!, 269-76, and EduardoVicente, "On the
FALN," Studies on the Left, 5 (Winter 1965), 99.
17 Cabieses, Venezuela,Okey!, 203; WalterH. Slote, "Case Analysis of a Revolutionary,"in
Frank Bonilla and Jose A. Silva Michelena, A Strategy for Research on Social Policy
(Cambridge,Mass.: MassachusettsInstituteof Technology Press, 1967), 304.
18 Guevara, Los Cachorros del Pentagono, 118-20.
19 Daily Journal (Caracas) 4 February 1966, 9 March 1966, 3 June 1966; El Nacional
(Caracas)4 May 1962, 13 August 1965.
20 James D. Cockcroft and Eduardo Vicente, "Venezuela and the FALN Since Leoni,"
MonthlyReview, 17 (November 1965), 36; Maspero, Avec Douglas Bravo, 53.

was forced to hide; but the events related to her by the peasantswho experi-
enced the terrorwere grisly indeed. Peasantswere tried to jeeps and dragged
on the ground;others were beaten to death or had their hands cut off; women
were raped;huts and grain stores were burnt. A small boy was allegedly shot
for refusing to sell meat pies to a soldier. One woman (who had only grudg-
ingly accepted the guerrillas in the area) was taken into the center of the
hamlet, disrobed, and stomped by soldiers for thirtyminutes. More than fif-
teen soldiers raped one pregnantwoman. Anotherpregnantwoman was also
rapedand miscarriedas a result. Zago sums up the results: "What the govern-
ment wanted it got: The peasants are terrified."21
More fragmentaryreportsare available from other areas, including Miran-
da, Portuguesa,Trujillo, and the llanos (plains states), in which the residence
of a guerrilla was blown to pieces with grenades without first attemptingto
see if it was occupied; and three women were rapedby NationalGuard(Fuer-
zas Armadas de Cooperaci6n, or FAC) troops. When one of the victims
lodged a complaintat the local FAC post, the captainreplied, "No sir! Take
these 100 Bolivares and buy some new clothes; and take care what you say,
because the next time we'll kill them. And you're not to say anythingabout
Even if one takes these reportswith much skepticism, the recordis still one
of agonizing humanmisery. This is neitherto say thatthis activitywas official
policy, nor that it always occurred,nor thatthe guerrillasdid not employ it for
propagandapurposes. Two illuminating comments were made by guerrilla
leader "El Gavilan" (Jose Diaz) when he came down out of the Larahills in
August, 1965. He indicated not only that he had been treated well by the
soldiers who capturedhim, contraryto his expectations, but also that such a
public statement might cost him his life.23 Governmentsare apparentlynot
alone in wishing to control informationflows duringguerrillawar, and guer-
rillas also obey the rules of propagandacontrol.
The contrastof the terrorby the Venezuelan governmentwith that carried
out in Guatemalais strikingfrom a methodologicalpoint of view. There is far
less specific informationin the lattercase, even though the terroristicexecu-
tions of Guatemalanpeasants ran into the thousands, as opposed to a few
hundredin Venezuela. Indeed, the Guatemalanexperience in 1966 and 1967
clearly standsout as the most brutalregime of terrorimposed upon a peasant-
21 Angela Zago, AquiNo Ha Pasado Nada (Caracas:El Sobre, 1972), 111-46, 149-50, 156-
7, 170.
22 Maspero, Avec Douglas Bravo, 53; Cabieses, Venezuela Okey!, 204-5; Norman Gall,
"Teodoro Petkoff:The Crisis of the ProfessionalRevolutionary-Part II: A New Party," Ameri-
can Universities Field Staff Reports-East Coast South America Series [hereafter called
AUFSR-ECSA], 17:9 (August 1973), 6. On the llanos events, see the memoir by Antonio
Zamora, Memoria(s) de la Guerrilla Venezolana(Caracas:Sintesis Dosmil, 1972), 88, 99, 125
23 El Nacional (Caracas), 27 August 1965.

ry in all the time period underreview, even dwarfingBatista's terroragainst

the Cuban guajiros (peasants of the Sierra). The estimates of deaths in this
terror run from 589 to as high as 6,000 for a one- or two-year period.24
Furthermore,there is striking agreement between the government and the
guerrillason the magnitudeof the terrorand the fact that very few of those
killed in the army campaignin ZacapaDepartmentin 1966-67 were actually
guerrillas.The guerrillasin 1968 claimed that 3,000 civilians had died in the
previous four years, while the armyreported2,000 dead-but that only forty
to fifty guerrillashad been killed-in the one year of the Zacapacampaign.25
This campaign was supervised by Colonel Arana Osorio, who later cam-
paigned for the presidency in 1970, promising to turn all Guatemala into
"anotherZacapa." The FAR guerrillasfavored his candidacy, apparentlyon
the groundthat the more repressivethe countrybecame, the betterwould be
their chances for success.26 By 1972 Aranamay well have been carryingout
his campaign promises, as the Communists claimed that 2,000 were dead
after one to two years of his government.27
Guatemalasaw the first large-scale emergence of right-wingterrorgroups
in Latin America, which killed at least 100 persons in April 1967 alone and
may have killed 4,000 persons up to 1974. As Aguilera has demonstrated,
they were activated under extremely patriotic and virulently anticommunist
symbols and were linked to certainlandownersandbusinessmen,as well as to
the police and the army. The elements of vigilantism in these groups-as
opposed to those elements of official (though sub rosa) sponsorship-were
summedup in a speech by a right-wingcongressman:"The regime shouldn't
find it remarkablewhen the bourgeoisie organizes itself in order to take the
law into its own hands." One other intriguingcharacteristicof these death
squads was the presence of "numerousex-guerrillas" in their ranks.28
A second aspect of terrorin Guatemalawas the bombingof guerrillaareas.
At first the United States restrictedthe military's use of eight B-26 bombers
and fifteen F51D Mustangs to Zacapa, but by 1971 these planes were avail-
able wherever they were needed. The use of napalm was apparentlyquite

24 Several estimates are gathered by Gabriel Aguilera Peralta, La Violencia en Guatemala

como Fen6meno Politico (Cuemavaca, Mexico: Centro Interdisciplinariode Documentation,
1971), p. 4/18. Also see Robert F. Lamberg, Die Castristiche Guerilla in Lateinamerika:
Theorie und Praxis eines RevolutiondrenModells (Hannover,Germany:Verlag fir Literaturund
Zeitgeschehen, 1971), 65.
25 Luis MercierVega, Guerrillas in LatinAmerica:The Techniqueof the Counter-State(New
York: Praeger, 1969), 129; Norman Gall, "The Legacy of Che Guevara," Commentary,44
(December 1967), 42.
26 Fritz Ren6 Allemann, Macht und Ohnmachtder Guerilla (Munich:R. Piper, 1974), 359-
27 Francisco J. Prieto, "The CommunistRole in Guatemala," WorldMarxist Review (En-
gland), 15 (September 1972), 27.
28 Allemann, Macht und Ohnmacht, 184-6; Aguilera, La Violencia en Guatemala, pp. 4/21
to 4/25.

liberal and not only for use in burningfields: The guerrillasreportedthat five
peasants found dead near Rio Hondo were burnt beyond recognition by
Terrorappearedin armyand police sweeps throughoutZacapa. One partic-
ular twist, also reportedin Venezuela, was that local feuds or vendettaswere
grafted onto the terror by the army. One peasant would point out a local
"enemy" as a guerrilla collaborator,whom the army would promptly ex-
ecute. In the Guatemalancase, the heavy elements of distrustand suspicion
were accentuatedby the presence of local militaryoutposts or comisionados,
which apparentlytook on the aspects of a vast spy network after 1954-a
feature accentuatedduring and after the Zacapa sweep of 1966-67.30 In the
Guatemalancase, then, the totalitarianelements of government and right-
wing terrorreachedtheir highest peaks for the periodpriorto 1970. It should
be no wonder, then, that the governmentsucceeded in eliciting collaboration
from peasants under such extraordinaryconditions, especially when such
terror was wedded to simultaneouscivic action campaigns in the same re-
gions. (The government's employmentof both terrorand civic action would
escalate yet furtherfrom 1978 to 1984.)
The broad outlines of terrorin Zacapa have been described by the FAR
guerrillas and others. Mano Blanca (White Hand), the leading right-wing
terroristgroup, apparentlywas the cutting edge or at least a majorparticipant
in the terror. The army came into the most revolutionaryvillages, gathered
and shot peasantleadersin fronteveryone, threateningto execute morepeople
if the villagers did not collaboratewith the authorities.This kind of activity
continued for more than half a year, and the peasantcasualties numberedin
the thousands. It was Walter's "regime of terror" at its most brutal.31
Army terrorneitherbegan nor ended with Zacapain 1966-67. In February
and March of 1964, military bombing of Izabal guerrilla areas allegedly
resulted in seventeen peasant deaths; and army sweeps in the same area
routinelyemployed tortureof peasantsfor information.On the day following
a guerrillavisit and talk ("armed propaganda")to the Kekchi Indianvillage
of Panz6s in Alta Verapaz, the army raided the town. One pregnantwoman
was gang-rapedby thirtysoldiers, and five peasantswere executed at a nearby
hacienda in which the guerrillas had paid a similar call. (That same town
would suffer a notorious massacre by the military in 1978.) The guerrillas

29 Susanne Bodenheimer, "Inside a State of Siege: Legalized Murderin Guatemala," Ram-

parts, 9 (June, 1971), 52; Eduardo Galeano, "With the Guerrillasin Guatemala," in Latin
America: Reform or Revolution?, James Petras and Maurice Zeitlin, eds. (Greenwich, Conn.:
Fawcett, 1968), 374.
30 Georgie Anne Geyer, "Guatemalaand the Guerrillas," New Republic, 163 (4 July 1970),
18; RichardN. Adams, Crucifixionby Power (Austin:Universityof Texas Press, 1970), 271-2;
Camilo Castano, "Avec les Guerillas du Guatemala," Partisans, 38 (July-September 1967),
31 Castano, "Avec les guerillas du Guatemala," 152.

workedmore carefullyafterthis tragedy,but the terrorby the governmentstill

occurred. In a similar sequence of events at the village of San Jorge, nearthe
town of Zacapa, anotherpregnantwoman was gang-rapedand her womb cut
open. After this the guerrillasreturnedto the village to execute the man who
had pointedout guerrillacontactsto the army. The most brutalemploymentof
terror up to 1966-67 was probably during OperationFalc6n in Zacapa in
Septemberand October 1965, which was carriedout much more quietly than
the campaigna year later. The guerrillasreportedthatat least fifteen peasants
innocent of guerrillacontacts were taken out to a quarrynear San Lorenzo.
There they were beaten and torturedfor information.One man had his penis
cut off and stuffed in his mouth. Among the techniques applied were the
placement of a suffocating oilcloth hood over the head (capucha); being
hoisted up by one's arms after they were tied behind one's back; beatings on
the head and testicles; the breakingof a victim's arms;and outrightshootings.
To top off these grisly tales, an Americanmissionaryreportsthatthe govern-
ment employed Orwellian techniques "to cover the traces of bombings and
massacres, [reroutingrivers] as well as razing forests and bulldozing vil-
lages."32 Again, even giving a heavy discount to such stories-and the
Guatemalanreports have high rates of confirmationby nonguerrillas-one
must conclude that rural Guatemalansexperienced governmentterrorin its
most extreme form.
In Colombia, a full understandingof the extent of violence during the
specific period of the guerrillamovementsis only possible with knowledge of
the awesome amount of violence that occurred during La Violencia from
roughly 1948 to 1964. After fifteen years of careless or hyperbolicestimates
of the dead-which had ranged as high as 500,000-Paul Oquist employed
more careful techniques to arrive at the conclusion that more than 200,000
died duringLa Violencia.33Much of the brutalityoccurrednot only within the
peasantry but also in the acts of the army, police, and pajaros (literally
meaning the "birds") againstpeasants, almost all of the victims being affili-
ated with the Liberal Party (the piajaroswere vigilante groups which were
32 C. Tzul, "Reaction Rampantin Guatemala," WorldMarxist Review (England), 7 (Sep-
tember 1964), 87; Donn Munson, Zacapa (Canoga, Calif.: Challenge, 1967), 160-4, 172-3,
194-5, 64-65; NormanDiamond, "Why They Shoot Americans," TheNation, 206 (5 February
1968), 167.
33 For a range of earlier, ever-higherestimates, see Germin GuzmanCampos, OrlandoFals
Borda, and EduardoUmafia Luna, La Violencia en Colombia, 2 vols. (Bogota: Tercer Mundo,
1962-1964), vol. 1, 287-93; RichardGott, Rural Guerrillasin LatinAmerica(Harmondsworth,
Middlesex, England:Penguin, 1973), 272; James Petras, "Revolution and GuerrillaMovements
in Latin America: Venezuela, Colombia, Guatemala, and Peru," in Petras and Zeitlin, Latin
America, 334; Regis Debray, "Problemsof RevolutionaryStrategyin LatinAmerica," New Left
Review, 45 (September-October 1967), 40; WalterSchump, Las Guerrillas en AmericaLatina:
El Principio y el Fin (Buenos Aires: Punto Critico, 1971), 16. Russell W. Ramsey, "Critical
Bibliographyon La Violencia in Colombia," LatinAmericanResearchReview, 8 (Spring 1973),
10, entry no.32, notes a ster critiqueof early, high estimates. Paul Oquist, Violence, Conflict
and Politics in Colombia (New York: Academic, 1980).

informally created by the Conservative Party and whose attacks on Liberal

peasants were ignored by the army and Conservativepolice).
Many allegations of later army violence againstthe peasantry,particularly
in the guerrillaareas and formercenters of La Violencia, were gatheredin an
unfortunatelyunderdocumentedwork by a group of left-wing Colombians.34
Official governmentfigures also suggest thatarmyterrormay have continued.
From 1 January 1966 to 31 March 1967-a period of heavy guerrilla ac-
tivity-the army reportedthe following death totals in the counter-guerrilla
campaign: 154 in the military, 29 guerrillas, and 384 others. Similar data
appeared in the annual Memorias submittedby the Ministry of Defense to
Congress. From these data it is very difficult to separatethose peasantswho
were intentionallykilled by bandits from those slain by soldiers or by guer-
rillas, or even from those caught in cross fires. One suspects that the army
contributedmore than its shareof peasantdeaths.35Still, the most persuasive
negative evidence may come from the guerrillasthemselves, for one of their
members candidly admitted that the Colombian army had "changed com-
pletely. It no longer frightens, it wants to reassure." If he was right, then
terrorwas not routinelyapplied in Colombiaand was insteadlargely replaced
by Civic Action.36
Army terroragainstthe peasantryin the guerrillaperiodwas apparentlythe
most intense during the 1964-65 campaigns, which were intended to bring
quasiautonomous "peasant republics" back under federal authority. There
are reportsof armytroopscuttingoff the heads and handsof the deadpeasants
for identification purposes-if not out of sheer barbarism.The army used
aerial attacks with napalm and other bombs, which were followed by heli-
copter landings, as standardprocedures; at the same time ground troops
employed ever-tighteningcordons sanitaires aroundtowns, graduallycreat-
ing free-fire zones outside of the circles. One guerrilla during this period
alleges thathundredsof peasantsdied due to such indiscriminateactionby the
military.37Throughoutthis entireperiodthe governmentcontinuedto referto
the operation as an anti-banditcampaign, but this label is not persuasive.
Guzman and his collaboratorslisted violent events in Colombiaby municipio

34 Comite de Solidaridad con los Presos Politicos, Libro Negro de la Represion: Frente
Nacional 1958-1974 (Bogota: Editorial Graficas Mundo Nuevo, 1974). This work is a good
example of the use of political blindersin moralevaluation:No guerrillaviolence againstpeasants
is reported, except for one peasant who was ajusticiado ("judged," the euphemism for ex-
ecuted). The two verbs employed to describe killings by governmentare consistently different
from the words used for those slain by guerrillas.
35 See Mercier, Guerrillas in Latin America, 236-7, for official casualty figures.
36 Larteguy, The Guerrillas, 69.
37 Specific allegations of violence against peasants in guerrilla zones appear in Comite de
Solidaridad, Libro Negro, on 69, 70, 83, 87, 88, 90, 119, 151; Jacobo Arenas, Diario de la
Resistencia de Marquetalia (Bogota?: Abejon Mono, 1972), 27-28, 70-72, 89. See also Jaime
Velbsquez Garcia, Contrainsurgenciay GuerraRevolucionaria(Bogota: Tinta Roja, 1974), 20-

for the 1958-63 period, and the core areas of the peasantrepublicswere not
prominent in degrees of banditry or violence; indeed, there is substantial
evidence that they were created as havens from violence. For example, the
municipioof Marquetalia,in Caldas, which gave its name to the most impor-
tant peasant republic, only had one violent event in that period.38
Personal testimony providing particularexamples of governmenttorture,
beatings, and killings was given before the ColombianCongressin November
1964 (thatis, at the beginningof the guerrillamovementsproper).Among the
techniques reportedwere the placement of a grenade in a prisoner's mouth
and threateningto pull the pin; faked firing squads;punching, kicking, and
walking on prisoners;electric currentappliedto the genitals, hands, and ears;
burning with cigarettes; and outright execution.39 It is hard to believe that
such tortures-aimed at finding ManuelMarulanda(also known as Tiro Fijo,
or "Sure Shot"), who was later the commanderof FARC-ended after the
guerrillas became mobile.
Estimatesof the scope of terrorin Peru vary widely, and one suspects that
the presence of inflation in the body counts is similarto that noted above for
Cuba and Colombia. The highest number,takenprobablyfrom a MIR propa-
gandasheet and propagatedfurtherby Victor Villanueva,claimedthe govern-
ment was forced "to massacre 8,000 peasants" to suppress the guerrillas,
while Petras suggested over a thousandsuch deaths, yet a MIR radio broad-
cast spoke only of hundredskilled, and the MIR's newspaper "El Guer-
rillero" on 5 September 1965 spoke only of "dozens" killed in Junin, the
area with the heaviest fighting and bombing. Clearly most of the dead peas-
ants were there and in Bejar's area of Ayacucho, where he lodged the infla-
tionary charge of "genocide." Gall, our best chronicler of the Peruvian
events, suggests that hundredsof CampaIndianswere killed in Junin alone.
Overall, an estimateof "only" 300 to 1,000 deathsfrom governmentterroris
a best guess, which is still bloody enough indeed for a short six-month
military campaign.40Indirectevidence also indicates that hundredsof peas-
ants died at the hands of the army. For example, the guerrillas, including
those in Piurawho never fought a battle, only numbered100-150 at best; yet
by mid-August, before any serious guerrilla losses were ever incurred, the
governmentclaimed to have killed 100 or even more guerrillasin the Satipo
and Pucuta areas of Junin alone.41

38 Guzman et al., La Violencia en Colombia, vol. 2, 301-26.

39 Colombia:An EmbattledLand (Prague:Peace and Socialism, 1970), 19-28.
40 For a reproductionof the pamphlet, see Mercado, Las Guerrillas del Peru, 230-2. For
Villanueva's use of the figure, see Sara BeatrizGuardia,Proceso a Campesinosde la Guerrilla
'TupacAmaru' (Lima: Compafia de Impresionesy Publicidad, 1972), 15; Petras, "Revolution
and GuerrillaMovements," 349; Lamberg,Die CastristicheGuerilla, 118; "Peru:Entrevistaa
dos Guerrilleros," Pensamiento Critico (26 July 1967), 192; Gall, "The Legacy of Che Gue-
vara," 39.
41 Lamberg,Die CastristicheGuerilla, 117; La Prensa (Lima), 14 August 1965. Notably, by

The bombing of guerrillazones in Peruby the armywas the most intensive

seen in Latin Americanguerrillawarfareuntil the Nicaraguancivil war of the
late 1970s. Mercier even uses the phrase "saturationbombing" to describe
the air force's style of attack, which was certainly not precision bombing,
since it intensely shelled the Mesa Pelada in Cuzco througha dense fog from
mid-to-late September. Guerrillazones in Junin were bombed in mid-July,
then much more intensively between 5 and 15 August; and napalm was
droppedin the areas of Pucuta, Ajospampa,and RosarioPampa. Bombing in
Junin was renewed in late September.42Interestinglyenough, there were no
news reportsof air force bombings of Bejar's ELN front in Ayacucho, which
was apparentlyliquidatedthroughthe use of groundtroops alone.
Finally, three majorguerrillaleaders-Javier Heraud(PuertoMaldonado,
1963), Luis de la Puente Uceda (MIR-Cuzco), and Maximo Velando Galvez
(MIR-Junfn)-were all apparentlykilled by the army after surrenderingor
trying to do so.43 Similar fates overtook several Venezuelanguerrillaleaders
(perhapsincluding Fabricio Ojeda) and Che Guevarain Bolivia as well.
Our descriptionsof terrorin ruralareashave largely come from all sources
except the people who experiencedit directly:the peasantsthemselves. In this
case we are fortunateto have exceptionaldata:the transcriptsummariesof the
trials of the peasants who were chargedwith collaboratingwith or participat-
ing in the Tupac Amaruguerrillafront in Peru (Junin). What did these peas-
ants say? Pablo Torres Cordovareportedbeing threatenedwith death by the
Investigative Police (PIP) if he would not sign a deposition indicating his
presence at the sites of various guerrilla raids (he had indeed joined the
guerrilla band but said he was tricked into doing so). Evaristo Pahuacho
Valverde testified that he was beaten by the authorities;that his father had
been beaten so that he died as a result; and that this two female cousins,
thirteenand ten years old, were rapedby the police underhis very eyes. The
judge asked GuillermoLoardoAvendafiowhy he had desertedthe guerrillas,
and he wryly replied, "Because I saw how the police burnedhouses in the
village, the reprisals taken against the people," and that he had decided to
desist if government violence continued. In his deposition, he claimed that
guerrillas massacred people, but he later repudiatedthe document. When
asked if he hadn't thought that he would be attackedafterjoining the guer-
rillas, Loardoreplied, "I didn't think at all. I didn't think that police burned
houses, stole cattle, rapedwomen." Moises SuarezMercadoclaimed he was
torturedwhile in jail and that the Civil Guard had killed his farm animals.

late Septemberthe governmentclaimed that only fifty guerrillashad been killed in the area in the
previous two months; see La Prensa, 28 September 1965.
42 See La Prensa, various issues,
especially 7 August 1965; Guardia,Proceso a Campesinos,
20; Mercier, Guerrillas in Latin America, 84; Norman Gall, "Peru's Misfired GuerrillaCam-
paign," The Reporter, 36 (26 January1967), 38.
43 Mercado, Las Guerrillas del Peru, 55-60; Gall, "Peru's Misfired GuerrillaCampaign,"
38; Guardia,Proceso a Campesinos, 21.

When asked if the police had injuredhis family, he replied, "No, but my
countrymen, yes." Miguel Matensio Torres also said that he was torturedin
jail to force a confession to the charges against him. A somewhat more
educatedpeasant (?) and MIR member, Jose MirandaBalbin, admittedin his
testimony to serving as a guerrilla contact only after a month of torturein
which he was hung up and beaten into unconsciousness.44
These Peruvianpeasantsalso had to worry, as in Venezuela, of the possible
consequences of retractingtorture- and threat-induced"confessions" once
they entered public courtrooms, since they knew that they would not always
be in the presence of legal protections. Nonetheless, some exchanges were
remarkableand even injected elements of macabrehumor into the proceed-
Prosecutor: "Have they [the jailers] mistreatedyou?"
Peasant: "Only verbally."
Defense lawyer: "What did this verbal mistreatmentconsist of?"
Peasant: "They hung me up and beat me."
Anotherpeasant, afterthe prosecutorexposed inconsistenciesin his testimony
concerninga tripto Cuba, exculpatedhimself by suggesting, "I'm sick in the
head. I forgot because they beat me a lot."45
We come now to the Bolivian experience. Despite Che Guevara's com-
ments recordedat the beginning of this article, the Bolivian guerrillawar was
almost completely devoid of terroron either side-an issue to which we shall
returnin our theoreticaldiscussion below. The government, for its part, felt
assuredof the supportof the peasantry,while the guerrillaswere too mobile,
too weak, and too fragmentedto be able to derive any practicaladvantages
from terror.Allemann arguesthatreprisalsagainstthe peasantsby the govern-
ment occurred "only exceptionally" in Bolivia, but he provides no exam-
ples.46 Guevara's diaries did note the bombing of guerrillaareas (in a very
sparsely populatedzone) on 24 March, 28 March(napalm), and 3 September
1967. On 5 September, he also noted that soldiers coerced nearby peasants
into giving them informationabout the guerrillas.
Those guerrillaswho deserted or surrenderedwere in some cases the vic-
tims of sadistic excess. For example, two desertershad theirsilhouettestraced
on a backdrop with bullets (wounding one); and a guerrilla who tried to
surrenderfollowing the Vado del Yeso ambushin late August and had his arm
literally shot to pieces afteran exchange of invective with a soldier, was killed
shortly thereafter.47The most famous case is that of Guevarahimself. After
44 Guardia,Proceso a Campesinos, 41-66, gives the trial evidence. Similar reportsof peas-
ants who had been killed by troopsappearin a privateletterfromGuillermoLobat6nto Luis de la
Puente on 20 June 1965; see Mercado, Las Guerrillas del Peru, 154-5, or Gott, Rural Guer-
rillas, 421-3.
45 Guardia, Proceso a Campesinos, 44-45, 66-70.
46 Allemann, Macht and Ohnmacht, 238.
47 Luis J. Gonzalez and Gustavo Sanchez Salazar, The Great Rebel: Che Guevara in Bolivia
(New York: Grove, 1969), 167, 171-2.

initial reports of a death in combat, various investigative reportersrevealed

that he had been shot in a schoolhouse in La Higuera on orders from the
government in La Paz.

Guerrilla Terror
"Killjust one andfrightenten thousandothers."

There is no doubtthatthe terrorused by governmenttroopswas vastly greater

than that used by the guerrillas. This fact should not be confused with the
position that terroris generally an unimportantelement of guerrillaactivity.
This thesis emerged in its boldest outlines from the Vietnam War, when Viet
Cong sympathizersarguedthat the NLF was successful simply because it had
won the "hearts and minds" of the populace. The Vietnamesecase providesa
most curious dais from which to propoundsuch a thesis, for the Viet Cong
unleashedupon the civilian populace a siege of terrorof such dimensionsthat
it has not been subsequently approachedby any other guerrillamovement.
Perhaps50,000 deaths were the direct result of Viet Cong terrorin a decade
and a half of civil war. One can not claim thatthose civilians who were killed
were strictlygovernmentofficials (not thatsuch an argumentwould legitimate
such killings). In the period after 1968, when data were finally kept on
nonofficial deaths as well, 80 percent fell in the nonofficial category.48
A certain type of argumenttries to legitimate insurgentterroragainst non-
combatantson two grounds. First, it is argued that the insurgentsmust use
terrorin orderto survive early regime assaultsupon their insecuremovement.
Second, it is arguedthat such terroris highly selective and used only against
government officials, informers, deserters, and local criminal elements, in-
cluding "social" criminals such as landlords.49Despite those assumptions,
some guerrillas have succeeded virtually without terror (notably in Cuba),
while those guerrilla movements which do employ terror can see that the
range of their victims broadens in scope (notably in Cambodia). Are the
guerrillas not bound by the same combatant-noncombatantcategories to
which the army is expected to conform?Bond sums up the objection nicely:
"Why an internal struggle for power should transformwhat is otherwise

48 On Viet Cong terror,see GuentherLewy, America in Vietnam(New York:Oxford, 1978),

272-9, and Douglas Pike, Viet Cong: The Organizationand Techniquesof the National Libera-
tion Front of South Vietnam(Cambridge,Mass.: MassachusettsInstituteof Technology Press,
1966), especially 238-51. Deaths from terror,mostly aerial, by the United States were probably
about 150,000; see Neil Sheehan, "Should We Have War Crimes Trials?," New YorkTimes
Book Review, 28 March 1971, 1-3. Peter Bergerhas also rejectedthe "heartsand minds" view
of the Vietnam war; see his "Indochina and the American Conscience," Commentary,69
(February1980), 33.
49 For a debate focussed on my second point, in this case concerning El Salvador's recent
killings, see Phillip Berryman, "Another View of El Salvador," Dissent (Summer 1982), 356
and "Gabriel Zaid Replies," 359.

murder into a legitimate act of war, without the conditions that normally
justify killing in war [that is, a fire-fight engagement]is not clear."50
The consistent reappearanceof guerrillaterroragainst fellow countrymen
in the most diverse settings is due directly to two factors: the claim by the
guerrillasthatthey constitutelegitimateauthorityand the high degree of their
vulnerabilityto "informationleakage."51 This second point means simply
that many guerrillas can only survive as sub rosa organizationsand must
controlthe informationflow aboutthem if they areto survive;yet, pace Eqbal
Ahmad,52 survival provides insufficient moral grounds for guerrillasto kill
noncombatants,just as "suppressionof guerrillas," "regime survival," or
"defense of the fatherlandagainst communist subversion" do not provide
such grounds for the military. From such legitimations are "dirty wars"
created. In general, attempts to legitimize the killing of civilians contain
elements of a moralized Whig history, that tendency to view certainpresent
results-and some putative future ones-as providing historical and moral
validation for past or present evils. History is indeed often written by the
victors of such wars;yet the tales of the losers andthe dead also have much to
tell us.
Guerrillaterroris generally far more selective than governmentterror.For
one thing, socialist guerrillasare often fundamentallymotivated by a moral
vision of a betterworld, which precludesterroristactions as inconsistentwith
such a vision. There are, of course, always exceptional cases. One of the
more interestingis Che Guevara,who behaved in a highly principledfashion
in Bolivia, while writing the quotationthat began this article, as well as the
following: "In fact, if Christ himself stood in my way, I, like Nietzsche,
would not hesitate to squish him like a worm."53 Such sentimentsare, how-
ever, rare; and one should not underestimatethe role of moral principles in
Second, guerrillaterroris, death for death, far more effective than govern-
ment terrorin eliciting compliance from the peasantry.Studentsof criminal
deterrence have argued that quick and sure punishment, more than severe
punishment, maximizes deterrenteffects. If we visualize guerrillaand gov-
ernmentterrorin these terms, we may understandwhy minimal terrorcan be
so effective. A peasant may collaborate with the guerrillasyet still have a

50 Bond, The Rules of Riot, 89.

51 Among the more unlikely outcroppings were the Vend6e revolt in eighteenth-century
France and a twelfth-century revolt of an English Earl. See Walter Laqueur, Guerrilla: A
Historical and Critical Study (Boston: Little, Brown, and Co., 1976), 27; and John Beeler,
"XIIth CenturyGuerrillaCampaign," Military Review, 42 (August 1962), 44-45.
52 Eqbal Ahmad, "RevolutionaryWarfareand Counterinsurgency,"in National Liberation:
Revolution in the Third World, Norman Miller and Roderick Aya, eds. (New York: The Free
Press, 1971), 159-62.
53 Dolores Moyano Martin, "A Memoir of the Young Guevara," The New York Times
Magazine (18 August 1968), 62.

good chance of escaping the indiscriminatecombing of an area by army

troops. Guerrillas,unlike the army, tend to establisha more or less permanent
presence in an area and will have the informationnecessaryto sort out actual
"offenders" (informersand spies for the government)from the faceless peas-
antry.54Thus a peasantinformerto the governmentis more likely to be found
out than one who collaborates with the guerrillas. Apologists for guerrilla
violence who argue that insurgent terror is selective fail to recognize that
terrordoesn't have to be indiscriminateto be highly effective. In addition,the
public natureof either governmentor guerrillaterrormaximizes the empathic
immediacy of the deterrenton the target population, the local villagers.
Third, guerrilla terrorcan be so effective because it is combined with a
programto promotepeasant interests. This wedding of power to beneficence
is crucial to understandingthe magnetic power of guerrilla movements.55
Newby sums up this "deferentialdialectic" nicely by notingthat, "The great
benefactorcould also be the greatpersecutor."56This fundamentalelementof
guerrillapower allied with benefactionscannotbe ignoredin analysis of their
relations with the peasantry.
Informationon the guerrillas'terroris sketchybecause they attemptto hide
actions incongruent with their professed values from the public, a stance
common to social movements in general. An internalmemo of the Venezue-
lan guerrillassuggested a halt to the deathpenaltyfor traitorsto the cause "to
obtain the correct image among the people and gain their support," since
guerrillasshould practice "justice not vengeance";57the memo also proposed
the quiet punishment of informers, again to avoid adverse affects on the
cause. Rarelydo guerrillasannouncetheirexecutions, except when the victim
is locally or nationallyinfamous. Hence, much of our informationon guerrilla
terror comes from former members, peasants who are willing to talk, and
newspaper accounts which are supplied either by governments or local
The Colombian ELN provides our only case in which an ex-member re-

54 Daniel Glaser, Social Deviance (Chicago: Markham, 1971), 58, cites a study by Tittle
which shows that certaintybut not severity of punishmentis inversely relatedto crime rates. See
as well the discussion and accompanyingreferencesin MarshallB. Clinard,Sociology of Deviant
Behavior (New York et al.: Holt, Rinehartand Winston, 1974), 346-50. One Eurasianofficer
noted of earlier Indochina that "the French destroy at random because they don't have the
necessary information";see Nathan Leites and CharlesWolf, Jr., Rebellion and Authority:An
Analytic Essay on Insurgent Conflicts (Chicago: Markham, 1970), 109.
55 See Timothy P. Wickham-Crowley, "The Rise (and SometimesFall) of GuerrillaGovern-
ments in Latin America," Sociological Forum, 2:3 (Summer 1987), 473-99, for an expanded
treatmentof this and related themes.
56 HowardNewby, "The DeferentialDialectic," ComparativeStudiesin Society and History,
17:2 (April 1975), 164.
57 Ralph H. Turner and Lewis M. Killian, Collective Behavior (Englewood Cliffs, N.J.:
Prenctice-Hall, 1957), 337; El Nacional (Caracas),29 August 1965;Daily Journal (Caracas),29
August 1965.

vealed the degree of guerrillaterror.He still consideredhimself a revolution-

ary when he was executed by the ELN in Bogota for publishingthe expose. In
La GuerrillaPor Dentro, Jaime Arenasoutlinedthe deathtolls attributableto
the ELN's "ajusticiamientos" (executions) of peasants. Over a period of
more than four years, the ELN killed forty-foursoldiers or police and lost the
same number of their own people in battle, but killed fifty-seven others,
including civilians and members of the guerrilla band itself.58 Unlike the
corporealatrocitiescommittedby banditsunderLa Violencia, however, guer-
rilla terrortypically occurredas forthrightexecutions.59
Arenas's reporton the Colombian ELN is particularlyrich and thorough,
and we lack anything quite like it for other guerrillamovements. This fact
raises some critical questions. Can his findings be generally applied to other
guerrilla movements which simply lack a Khrushchevto reveal the hidden
crimes of their Stalin? Have other guerrillamovements also killed a peasant
for each dead enemy, being merely more successful in hiding the fact? Be-
cause the ELN exhibited some unusual characteristics,I would offer a cau-
tious "no" as an answer. Sharpintellectualversus peasantconflicts split the
movement, with the latter group mostly following the authoritarianperson-
ality and behavior of ELN chief Fabio Vasquez Castaiio, who ordered the
bulk of the killings.60
Fidel Castro and RautlCastro (in their respective Cuban fronts) and Che
Guevara(in both Cuba and Bolivia) held degrees of power similarto thatheld
by the ELN's Vasquez, yet there is little or no terrorassociated with these
guerrilla movements. It is true that Fidel Castro tried and executed bandits,
rapists, and criminalsin the Sierrabut the "trial" procedureswere carriedout
with care by all accounts (probably more so than the [in-]famous stadium
trials of 1959). An internalrebel deathpenalty for desertion, insubordination,
and defeatism was apparentlynever used. Aside from this, a Sierranhacienda
foreman was executed for arrangingan hacendado's grab of land from some
peasants;Eutimio Guerrawas executed for becoming a double agent for the
army, that is, for espionage; and three peasantswere executed as army infor-
mers in mid-1957 (the difference between the last two examples is the dif-
ference between legitimateapplicationof the rules of war and ad hoc justifica-
tions). Barquinlists others executed by the rebels and argues that some were

58 Jaime Arenas, La Guerrilla Por Dentro (Bogota: Tercer Mundo, 1970), 199-204.
59 La Violencia's atrocities are now legend both for their number and for their grisly and
creative character.For an account of one haciendaraid punctuatedby rape and dismemberment,
see Evelio Buitrago Salazar, Zarpazo the Bandit: Memoirs of an Undercover Agent of the
ColombianArmy (University, Alabama:University of Alabama Press, 1977), 61.
60 Arenas, La Guerrilla Por Dentro, passim. For more informationon Vasquez's importance
in the ELN, see Mario Men6ndezRodrfguez, "Colombia: !Al Ataque!," Sucesos (Mexico City),
1777 (24 June 1967), 21-22. This was the first of a series of four or five weekly articlesin which
the journalist reported on his visit to the ELN's foco. The reports are very informative, if
uncritically laudatory,especially of Vasquez.

simply noncooperativepeasants falsely categorized as bandits, especially in

Raul Castro's front. There is evidence from Raul's treatmentof enemy sol-
diers and bandits that he had far fewer moral scruples about killing than did
his brotherFidel.61
In Bolivia, guerrillaterrornever went beyond the planning stages, as the
sole referencesare the entriesin Guevara'sdiary, such as the one at the end of
April: "The peasantbase has not yet been developed althoughit appearsthat
through planned terror we can neutralize some of them; supportwill come
later." Otherthan this, guerrillaresponses to a distrustfulcivilian population
never went beyond threatsand occasional seizuresof suppliesor food, usually
in response to some offense such as informing. Despite this, the peasants
regularly received the guerrillas in fear or even "panic-strickenterror."62
The guerrillasused more terroragainstthe peasantryin Peru. The Peruvian
Ministry of War accused the guerrillasof having barbarouslyexecuted peas-
ants who might have suppliedinformationto the authoritiesand also listed six
civilians-four of whom being peasants-killed by guerrillas in the six-
month guerrillacampaign.63The charge that the guerrillasin Junin shot and
killed two capturedmembers of the Civil Guardafter applying burningtor-
tures to their bodies (the guerrillas denied this) is particularlydifficult to
evaluate. Photos of two corpses with massive bums on various partsof their
bodies were prominentlydisplayed by the authoritiesand newspapersduring
the antiguerrillacampaign, and the army produceddoctors who claimed the
burs occurredprior to shooting.64
While on trial, several Peruvianpeasants reportedthat the guerrillashad
used threatsto induce their cooperationwith the insurgents. On 23 October
1965, the army asserted that the MIR guerrillasof Cuzco and Mesa Pelada
killed three peasants who refused to help them while the guerrillas were
desperatelyfleeing the army's relentless pursuit. The guerrillas, as might be
expected, "reversed the charges" and blamed the deaths on the army. This
same guerrilla group reportedly shot four deserters in the previous month,
which is not inconsistent with their history of desertions and the scattered
reports that the foco was being literally torn apart by internal dissension.
Members of this same foco apparentlyexecuted Indians near the Apurimac
river when they refused to help build rafts to cross the water. The man who

61 Eresto "Che" Guevara, Reminiscences of the Cuban RevolutionaryWar (New York:

Monthly Review Press, 1968), 65; Bonachea and San Martin,The CubanInsurrection, 100-1,
190, 197, 368 (n.80); Barquin,Las Luchas Guerrilleras, vol. 1, 364.
62 Eresto "Che" Guevara, The CompleteBolivian Diaries, 151-2, 170-1, 178-9 (see the
entries on the analysis of the month for April, 20 June, 3-7 July, all in 1967).
63 Peru, Ministeriode Guerra,Las Guerrillasen el Peru y su Represi6n(Lima?:MinisterioDe
Guerra, 1966), 386.
64 ArmandoArtola Azcarate, iSubversion!(Lima: EditorialJuridica, 1976), 64-65; Guardia,
Proceso a Campesinos, 14; see La Prensa (Lima) 3 August 1965, for the allegationsandevidence
about torture.

broughtnews of the event also reportedthe tortureof his two sons.65 There
was also a ratherbizarrereportthatguerrillasin Concepci6nProvince(Junin)
beheaded six membersof a peasant family, aged five to forty-two years old.
At this time Guillermo Lobaton's guerrillas were desperatelyfleeing army
pursuit, and Lobat6n himself died while fighting two weeks later.66
The Guatemalanguerrillasused still more terrorthan in Peru but selected
their targets more carefully: hacendados and their foremen, informers and
suspected informers,as well as villagers whose collaborationwith the armed
forces led to the deaths of innocentpeasants. Luis Turcios Lima of the FAR
defended executions by arguing, "We only respond to violence with vio-
lence. We deal out revolutionaryjustice." Indeed, one accountof the ajusti-
ciamiento of a villager was quoted directly from a proguerrillasource by
Munson. The guerrillaspounded on the offender's door with the command,
"Open in the name of revolutionaryjustice." In general, there is, however,
little evidence, as one critic has proposed, that the the use of terrorby the
guerrillas paralleled the atrocities by the government which we cited
By far the best documentationon guerrillaterroris that which is available
for Venezuela. In this case, I do not believe the amount of documentation
misleads us as to its extent. Venezuelanguerrillaswere faced with the task of
gaining supportfrom a peasantrythatwas largelyorganizedby and supportive
of the Accion Democrdtica (Democratic Action, or AD) party, which held
governmentalpower from 1959 to 1969. One might expect thatterrorin such
a milieu would be especially directedagainstAD peasantleaders,just as most
of the early Viet Cong terrorwas directedagainstlocal officials appointedby
the Diem governmentof South Vietnam. Our expectationshave been largely
borneout by reportson guerrillaexecutions. NormanGall reportsthe guerrilla
execution of a local AD peasantleader, Rodolfo Romero, in October 1964 in
Falc6n. "With his hands tied behindhis back, the guerrillashung him from a
tree by his armpitsand threw broken bottles in his face to make him bleed.
They read an execution decree accusing Romero of betrayingthe cause of
national liberation, then shot him as the whole community watched." In

65 Guardia,Proceso a
Campesinos, 38-63; Ministeriode Guerra,Las Guerrillas en el Peru,
67; Mercier, Guerrillas in Latin America, 185; Mercado,Las Guerrillasdel Peru, 199-201; La
Prensa (Lima), 25 September 1965, 14 October 1965 (the latterfor the man's reporton the raft
66 The New YorkTimes, 24 December 1965; Manuel Castillo, "Las Guerrillasen el Peru,"
Estudios (Buenos Aires), 581 (April 1967), 165.
67 Munson, Zacapa, 115 (Turcios' quote), 111; Georgie Anne Geyer, "Guatemalaand the
Guerrillas," The New Republic, 163 (4 July 1970), 18. For some referencesto executions, see
Lart6guy,The Guerrillas, 88; RobertRogers andTed Yates, "The UndeclaredWarIn Guatema-
la," SaturdayEveningPost, 239 (18 June 1966), 33; Adolfo Gilly, "The GuerrillaMovementin
GuatemalaII," MonthlyReview, 17 (June 1965), 11.

addition, Alexander reportsthat Romero was beaten, his teeth knocked out,
and his fingers and toes brokenbefore being shot-all before the eyes of his
wife and small children. Gall also reportedon the FALN's kidnapand killing
of four peasants in Trujillo State.68
The numberof civilian deaths-almost all peasants-attributable to rural
guerrillas in the Venezuelan literatureis at least twenty-nine and perhapsas
high as fifty. The commentby a radicalleftist thatthe victims of guerrillascan
be "counted on one's fingers" thereforeseems a bit of an understatement,
especially in view of the FALN's recordof massive urbanterror.69At times
the executions by guerrillas took on grisly aspects, as in the one already
described. Two uncooperativepeasant leaders were killed and their bodies
hacked to pieces on one occasion, and in another instance the guerrillas
allegedly shot the victim, stabbed him seven times, and stuffed his mouth
with propagandaleaflets. The allegationsby the governmentaboutthe "bru-
tal and insane" rape of many'women by the guerrillasreportedno details,
however, and are largely unpersuasive.70
Venezuelan guerrillasdid not restrictterrorto noncompliantpeasants. On
27 November 1966, a school for childrenfrom the United States was machine
gunned;a formergovernmentofficial from the Social SecurityAdministration
was kidnappedand killed in early 1967; many peasantswere kidnappedand
later released in addition to those simply executed; and at least two guer-
rillas-and probably many more-were executed for trying to desert the
focos.71 In the urbanareas, the FALN also killed 160 policemen from 1960 to
1966, who were not generally chosen for specific crimes but ratheras repre-
sentatives of government. Critics, radical sympathizers, and ex-guerrillas
agree that this tactic eventually generatedrevulsiontowardthe FALN among
poor barrioresidents, since most of the police came from these same barrios.
Finally, and in reference to our previous discussion of DIGEPOL's use of
torture, an instance of "sweet revenge" occurred in September, 1964: A

68 Gall, "The ContinentalRevolution," 4-5; RobertJ. Alexander, The CommunistParty of

Venezuela (Stanford, Calif.: Hoover InstitutionPress, 1965), 94-95. Two elements of further
interestin this case: (1) Romero was accused by some of "fingering" local enemies to the army
as guerrillacollaborators,and (2) thirtyfamilies abandonedtheirfarmplots in the areain the next
few days. On "official" targets, see Valsalice, Guerriglia e Politica, 221; El Nacional (Car-
acas), 14 January 1965, reportedthat an ex-guerrilladirected the army to the graves of four
69 Venezuela, Oficina Centralde Informaci6n(OCI), Six Years of Aggression (Caracas:Im-
prenta Nacional, 1967?), 36-49; Valsalice, Guerriglia e Politica, 141, 216 (n.23), 221; Daily
Journal (Caracas), 19 August 1964, 18 November 1965; El Nacional (Caracas),28 March 1962,
18 August 1964, 14 September 1964. Guevara,Los Cachorros del Pentagono, 8 (quote).
70 Venezuela, OCI, Six Years of Aggression, 42-43, 48-49; Daily Journal (Caracas), 18
November 1965.
71 Venezuela, OCI, Six Yearsof Aggression, 64; Alphonse Max, Guerrillas in LatinAmerica

(The Hague: Interdoc, 1971), 33; El Nacional (Caracas), 7 July 1964.


guerrilla deserter reported that three captured DIGEPOL agents had been
torturedwith hot irons and executed by his formerguerrillacompanions.72


My approachto terrordiffers in fundamentalways from some trends in the

literature.First, I reject as unscholarly those analytic attemptswhich enter
into collusion with the combatantsand choose to describethe killing of human
beings in the euphemisms commonly used by guerrillas for their various
actions. Bank robberiesbecome "expropriations"and the shooting of peas-
ants becomes "ajusticiamiento" or the application of "revolutionaryjus-
tice";73 yet the moral legerdemainhidden by such phrases pales beside the
pernicioususe of circumlocutionin military, state department,and think-tank
circles. In such circles the mass murderof Guatemalanpeasants was de-
scribed as the eliminationof the FAR's "peasantinfrastructure"and "popu-
lation resource." In Vietnam, the military referredto village bombings as
deprivingthe enemy of its "populationbase." In Leites and Wolf's systems
analysis of insurgency,killing guerrillafightersbecomes "destroyingR's [the
Rebellion's] outputs." Removal of a "population buffer" is the label that
they attachto massive programsfor uprootingandrelocatingvillages. The use
of a cordon sanitaire, which strictly controls all of the movement of people
and supplies into and out of an area (hence martiallaw in extremis), becomes
"raising R's input costs."74 Robin Williams has shown how euphemism
tends to be used as a way of achieving sanction for acts of violence that
otherwise could not be legitimated.75It is inappropriatefor a scholar to get
caught up in legitimizing terms used by the participantsin the conflict.
Furtherprefatoryremarksabout causal and ethical analysis are required.
Both sociologists and the public consistently confuse causal analysis of be-
havior with the absolutionof responsibilityfor such behavior. (A recent and
striking example occurred in the public discussion of the rape and vicious
beating of a female New York jogger apparentlyby a group of youths on a
"wilding" spree. All of the issues that I present here have been previously
raised about that event.) Thus, when sociologists claim to explain patternsof

72 Valsalice, Guerrigliae Politica, 169;TaltonF. Ray, ThePolitics of the Barrios in Venezue-

la (Berkeley:Universityof CaliforniaPress, 1969), 130-6; NormanGall, "TeodoroPetkoff:The
Crisis of the Professional Revolutionary-Part I: Years of Insurrection,"AUFSR-ECSA, vol.
16:1 (January, 1972), 15; Daily Journal (Caracas), 13 September 1964. More evidence on
Venezuela appearsin the following section.
73 Lamberg, Die Castristiche Guerilla, 40; Bonachea and San Martin, The Cuban Insurrec-
tion, 91.
74 Victor Perera, "Guatemala:Always La Violencia," The New YorkTimes Magazine, 13
June 1971, 71; Sheehan, "Should We Have WarCrimesTrials?", 1-3; Leites and Wolf, Rebel-
lion and Authority, 36-37.
75 Robin M. Williams, Jr. "Legitimateand IllegitimateUses of Violence: A Review of Ideas
and Evidence" (Paper prepared for the Behavioral Studies Research Group of the Hastings
Center, May, 1979), 9-18.

behaviorsuch as deviance or crime, both they and the generalpublic are aptto
confuse such theories about the behaviorof people in group settings with the
ethical evaluationof such conduct. Such a mixtureis invalid for two reasons.
First, there are elements of the ecological fallacy in such an admixture,which
is the errorof identifying the propertiesof groups with those of their respec-
tive individualmembers. Causal sociological analysisconsists of probabilistic
statementsabout the incidence of such behaviors, including statementsabout
the incidence and intensity of such behaviorswithin differentgroups, neigh-
borhoods, communities, or societies. Theory must also explain cases of non-
occurrence, and "poverty" can explain neither the stealing of food by a
particularhungrychild nor the theft of a television by a particularunemployed
teenage looter, since most poor people do not steal or loot. In contrast,ethical
analysis is concernedprecisely with the evaluationof such individualbehav-
iors. To conflate these two modes of analysis is a variantof the ecological
The second objection, which appliesjust as well to terroristicacts commit-
ted during guerrilla war, was stated eloquently by John Dewey over half a
century ago. Dewey argued that one should never try to lay total blame on
society or the criminal (read: soldier or guerrilla) for crime (read: terror),
because this implies:
An unrealseparation of manfromhis surroundings, mindfromthe world.Causesfor
an act alwaysexist, butcausesarenot excuses.Questionsof causationarephysical,
not moral except when they concern futureconsequences. . . . Society excuses itself
by layingblameon thecriminal,he retortsby puttingtheblameonbadearlysurround-
ings, the temptationof others,lackof opportunities,
andthepersecutionsof theoffi-
cers of the law. Bothareright,exceptin the wholesalecharacter
of theirrecrimina-
Dewey adds that good intentions alone do not meritper se the estimation of
"good," for the consequences of an act also fix its moral quality.76
The purpose of this introductorypreface is simple: The readershould not
understandmy following attemptto build a theory of terroras an attemptto
exculpate those who commit atrocities in guerrillawar. Readers may make
judgments based on the informationpresentedhere, but I cannotbe the Virgil
to their Dante in this ethical realm, in which we are in principleall equals and
claims to moral virtuosity inevitably smack of hubris.


Element One: Support Systems in Guerrilla War

Moder conventional war, which has grown increasingly "capital intensive"
over the past few centuries, clearly defines the combatants,civilian popula-

76 John Dewey, HumanNature and Conduct (New York: Carlton

House, 1922), 18, 44-45.

tions, and support systems. The support systems include arms, factories,
transportationnetworks, and the supply lines which connect the sources of
supplies with the combatants. The combatant-noncombatantdistinction is
perhapsthe centralissue. Under the laws of war, primarilyderiving from the
Geneva Convention, the combatantsare to be distinguishedfrom noncomba-
tants by standarduniforms with a distinctive insignia visible at a distance.
Attempts are also to be made to keep the civilian population distinct from
supportsystems, but the mere presence of a civilian populationnear an arms
factory does not, under the Laws of War, always protect that factory from
attack. Under "Rules for the Limitationof Dangers Incurredby the Civilian
Population in Time of War" (1956), Chapter 11, Article 6, it is stated that
attacks on the civilian populationas such are prohibited,but "should mem-
bers of the civilian population . . . be within or in close proximity to a mili-
tary objective they must accept the risks resulting from an attack directed
against that objective." Such attacks, however, are forbiddenif the military
gain is not proportionalto the degree of destructioninvolved. Furthermore,
the presence of "individual combatants"within the civilian populationdoes
not immediately renderthe latter subject to attack.77
In situationsof guerrillawarfare-Vietnam serving as the clearest case-
the distinction between combatant and civilian population is intentionally
blurredby the guerrillafighters. This was accentuatedin Vietnam when they
actually located themselves within villages, fortifiedtheirpositions, and then
fired at army patrols from within the villages. This tactic often produced
village bombings and terroristicsweeps as a response, with drastically re-
duced regard for the combatant-noncombatantdistinctions by the soldiers
from the United States, which in extreme cases could and did produce a
massacre of 150-500 civilians like that in My Lai.
In fact, the tripartitedistinction of the combatant,the noncombatant,and
the support and supply system is typically blurredin guerrilla war, unlike
conventional war. Does the villager who carries potatoes to the guerrilla
camp, as in Peru, constitute a military target?What about the peasant who
lodges a guerrillafor the night (a common occurrence)?Or the peasant who
serves as lookout for the guerrillas?Or peasants who regularlyprovide tor-
tillas and otherfood for guerrillas,as in Guazapa,El Salvador?Or those who,
as in Guatemala and later in El Salvador, serve on sporadic or permanent
peasantmilitias?In El Salvador'sguerrillawar, for example, evidence strong-
ly indicates thatthe peasantryhas often mixed with the guerrillasin ways that
makes them very hardindeed to distinguish.78Those featuresonly accentuate

77 Bond, The Rules of Riot, 201, 214, for reportson the rules of war.
78 See the commentsby JoaquinVillalobos in MartaHarnecker'scollection of interviewswith
CentralAmericanrevolutionaries,Pueblos en Armas (Mexico, D.F.: EdicionesEra, 1984), 173-
232; the reportof a brief visit by PhilippeBourgois, "What U.S. ForeignPolicy Faces in Rural
El Salvador:An Eyewitness Account," MonthlyReview, 34:1 (May, 1982), 14-32; and for a

in high relief the features found in bas-relief in those nations in which the
peasantssupportedthe guerrillamovementsdiscussedhere:Cuba, Venezuela,
Guatemala,and Colombia. (Even in El Salvador,a distinctioncontinuedto be
made, however, between those who supportedthe insurgencyand those who
incorporarse, became full-fledged soldiers.)
Terroragainst civilians is apparentlya far more regular, even "natural,"
concomitantof modem guerrillawarfarethanof modem conventionalwarfare.
Highly suggestive evidence along these lines comes from the difference in
behaviorbetween the combattroopsfrom the United Stateswhile they were in
Europe(including Germany)duringWorld War II and the way in which they
conducted themselves in some areas of Vietnam, as Susan Brownmiller's
sequentialaccountsof the two wars suggests, despiteherown (ratherdifferent)
intent and conclusions. Other suggestive evidence comes from the Vietnam
War itself, in which the militaryforces of the United States seemed to employ
terrormorecommonly in insurgentareasof SouthVietnamthanin the enemy's
homelandof North Vietnam.79I submitthatthe disparityin the actions of the
American combat troops-in degree, not absolutes-derived from the dis-
tinctive types of army, civilian, and supplynetworkcomplexes thatthey faced:
They engaged in lesser degrees of terrorwhen they confronteda conventional
army that was controlled, supplied, and directedby the state, with a civilian
population which still resided at home and far from the fronts; the troops
employed greater levels of terror, at times quite intense, when they instead
faced a guerrilla army intimately tied into the civilian peasant population,
especially in the Mekong Delta and (perhaps)in certaincoastal areas.80
Terror is particularlycommon in guerrilla warfare because there is an
aggregation and mixture of combatant, noncombatant,and support system
into a very small social and geographical space. The natureof the support
system is fascinating, for it consists in large part-although not completely-
of the peasantryitself. Whetherthe peasantryacts willingly or not, there is
often a very deep social and geographicaloverlapin guerrillawarfarebetween
the support system (the source of military intelligence, food, supplies, and
recruits)and the civilian population,with a large overlapbetween the civilian
populationand the combatantsas well, often in the form of peasant militias.
The contrastwith conventional war may be illustrated(see Figures 1 and 2)
with Venn Diagramsin which each circle indicatesa set or a "population"of
elements. The intersectionbetween the two circles indicates that certainele-

medical doctor's reportof a longer period in administeringmedical care in the Guazaparegion,

see Charles Clements, Witness to War: An American Doctor in El Salvador (Toronto et al.:
Bantam, 1984).
79 Susan Brownmiller, Against Our Will: Men, Womenand Rape (New York: Simon and
Schuster, 1975), ch. 3. For more on the selectivity of Americanterrorin Vietnam, see Lewy,
America in Vietnam, ch. 11; and Tom Wolfe, "The Truest Sport: Jousting with Sam and
Charlie," Mauve Gloves and Madmen, Clutter and Vine (New York: Bantam, 1977), 24-58.
80 Brownmiller, Against Our Will, 73-78, 86-113.

FIGURE 1. Social forces in conventional warfare.

SOURCES: See text.

ments are members of both sets; in this case, civilians and combatants, or
combatantsand partof the supportsystem, and so forth. The shaded areas in
Figure 2 can also be understoodto represent,not just support,but also areas
of potential reprisalsagainst the civilian population-that is, the deeper and
more thoroughthe overlap between the guerrillacombatantsand the civilian
population, the more likely that the government would engage in terror
against the civilian population in guerrillazones.
Two types of comparative evidence strongly suggest that the depth of
"system overlap" within guerrilla zones is correlatedwith the intensity of
government terror. During the period covered in this essay, we can clearly
detect two areasin which the local peasantswere not subjectedto terrorby the
military: La Convencion, Peru (although the nearby, isolated foco site was
certainly bombed) and Bolivia during Guevara's insurgency. In both cases,
the peasantry had given clear indications that they were indifferent to the
insurgency, if not actively hostile to it. In one Peruviancase in Cuzco, several
local peasant leaders abandonedthe insurgencyafter a brief period of mem-
bership and later guided the army to the destructionof the campsites on the
Mesa Pelada. In Bolivia, the nation's peasantry,especially in the Cochabam-

2. Social forces in guerrillawarfare.

See text.

ba Valley, were very strong backersof and voters for the Barrientosgovern-
ment; their nationalconfederationdeclaredsupportfor him in mid-1967; and
the leader of the peasant union in Guevara'slocale was also a pro-Barrientos
man.8' In contrast, we encounteredterroralmost everywhereelse in our four
other cases, and in all those cases we have clear evidence of strong peasant
supportfor the guerrillamovements in certain areas of each nation.
The second type of comparativeevidence comes from the "second wave"
of insurgenciesin Latin America, after about 1975. Every one of these move-
ments, except perhapsin Nicaragua, establishedbases of peasant supportin
the countryside far strongerthan the pre-1970 movements had done. Hence,
the peak numberof guerrillacombatantswas routinely in the thousands, not
the hundreds seen before 1970: perhaps 10,000 in El Salvador; 7,000 in
Colombia; 6,000 in Guatemala;and at least 5,000 for Peru's Sendero Lumi-

81 See Wickham-Crowley, "Guerrilla Governments," 481; and Idem, "Winners, Losers,

and Also-Rans: Toward a ComparativeSociology of Latin AmericanGuerrillaMovements," in
Power and Popular Protest: Latin American Social Movements, Susan Eckstein, ed. (Berkeley:
University of CaliforniaPress, 1989), 157. Some excellent detail on Barrientos'ssupportamong
the Bolivian peasantry,including the 1966 electoraltallies, can be found in Jean-PierreBernard,
et al., Guide to the Political Parties of South America (Harmondsworth,Middlesex, England:
Penguin, 1973), 135-46.

noso.82 Those numbersare sure signs that peasant supportfor the insurgen-
cy-and hence the overlapof our threesystems-has been greatersince 1975
than it was before 1970. The correlateof that greatersupporthas also been
obvious: In nations with majorruralinsurgenciessince 1975 (includingNic-
araguaas well), the centralgovernment,and often the death squads linked to
it, have responded with levels of terror which have taken at least 10,000
civilian lives (perhapsfewer than that in Colombiaand more than that-over
50,000 lives each-in El Salvadorand Guatemala).
Element Two: Legitimate Authority and "Heresy"
The above correlation strongly suggests the second necessary element in
understandingterroragainst civilians in the context of guerrillawarfare.As I
have arguedelsewhere, drawingparticularlyon BarringtonMoore's Injustice,
governments and their underlying civilian populaces typically iron out an
"implicit social contract" which unites the rulerand the ruled into a system
of interlockingrights and duties. The main duties of the populace are obe-
dience to appropriatedirectives and the renderingof surplusto supporttheir
governors;the duties of the rulersare to defendthe region, providepolice and
conflict resolution, and arrange for the material security of the populace.
When those rights and duties are perceived as fulfilled and intact by both
parties, we may speak of legitimate authority.The peasantpopulace in many
regions of Latin America has, however, clearly transferredits loyalty to
guerrilla "governments" in many instances, precisely because the central
governments (or informal landlord "governments") do not protect, defend,
and benefit the locals; indeed, they may do just the reverse.83
Notwithstandingsuch a peasant viewpoint, the authoritiesview the peas-
ants' allegiance to and supportof guerrillagovernments-and the consequent
existence of dual power or a counter-state in the nation as a whole-as
tantamountto the denial of the centralgovernment's "contractual"claim to
the obedience of the populace, that is, traitorousactivity. They have become
heretics, not from the churchmilitantbut the body politic. The sheer number
of guerrillasand the depthof theirsupportby the peasantsin certainregions of
these nations have made them something more: a "cancer" in the body
politic, something to be rooted out and destroyed througha kind of "sur-
gical" terror.
82 The
figure of 10,000 for El Salvadorhas been widely disseminated;for the 7,000 Colom-
bian guerrillas,see the LatinAmerica WeeklyReport, 31 August 1984; for the 6,000 Guatemalan
guerrillas, see James Dunkerley,Power in the Isthmus(London:Verso, 1988), 483; for Sendero
Luminosoof Peru, I am indebtedto privateconversationswith Michael Smith and also to Cynthia
McClintock, "Peru's SenderoLuminosoRebellion:OriginsandTrajectory,"in Eckstein,Power
and Popular Protest, 63.
83 Wickham-Crowley, "GuerrillaGovernments,"passim. BarringtonMoore, Jr., Injustice:
The Social Bases of Obedience and Revolt (White Plains, N.Y.: M. E. Sharpe, 1978), ch. 1,
especially 20-22.

The right-wingnatureof the regimes-or at least the right-wingsentiments

in the officer corps of the military-encountered by guerrillas might well
contributehere to the intensity of the terrorthat the peasantryhas suffered.
"Right-wing" radicalismhas ideological featureswhich distinguishit in part
from leftist radicalism.The most distinctiveof those featuresis, it seems, the
exaltationof violence per se (ratherthan the grudgingacceptanceof violence
as a painful but necessary means, which is so common on the left); and
another is the emphasis on cultural and biological purity-the most recog-
nized outcroppingof this being, of course, the Nazi and Japaneseemphases
on such issues in World War II.84
If terror is the response of a government to a decline of-and an open
challenge to-its authorityin the body politic, then we would expect to find
guerrillaterroragainstthe peasantryin similarcircumstances.When the guer-
rilla movements claim that they constitute legitimate counter-states, new
governments in microcosm-they are sharply challenged, usually through
the populace's unwillingness to cooperate with the "legitimate directives"
of the guerrillas. The evidence from both Peru and Venezuela on the
timing of guerrillaterroris consistent with the view that any "government"
responds with terror when its very authorityto give orders is substantially
In Peru, guerrillasare alleged to have killed uncooperativepeasantsat the
end of the brief insurgencyjust as the insurgents,their "counter-state"clear-
ly in disarray, were fleeing from the armed forces and the most in need of
succor. In Venezuela, the evidence is ratherclearer. In the countryside,there
was virtually no guerrillaterroragainst the peasantryduringthe years 1962
and 1963, as the insurgentsbegan to build up strong supportin certainrural
areas. However, the governmentchargedthatthe guerrillaskilled twenty-nine
to fifty peasants in the period from May 1964 to November 1965-the very
period in which the guerrillaswere being flushed from theirruralstrongholds
and when, especially in Lara State, there were clear signs of declining sup-
port and cooperation from the peasantry. The linkage of guerrillaterrorto
challenged "authority" is even more distinct in urbanareas. The guerrillas
called on the populace not to participatein the December 1963 election,
which constituted the first opportunity for a consecutive electoral trans-
fer of power in national history and was a severe challenge to the claims
by the FALN guerrillas that they were a legitimate counter-state. (In
their offices in Havana, the FALN issued memos on stationerywith a mast-
head that read, "Reptblica de Venezuela. Fuerzas Armadasde Liberaci6n
Nacional.") The revolutionariesdid more than exhort: In their preelection
wave of terror,twenty-twoto thirty-fourpersonswere killed andover 100 were

BarringtonMoore clearly assertsthe first principle, and the Nazi diariesthathe quotes from
strongly suggest the second; see Injustice, 413-34.

wounded.85One might draw latter-dayparallels to the SalvadoranFMLN's

antielectoral "campaigns" as well. Shortly prior to the 1989 elections, for
example, they threatened forty more mayors into forced resignations-a
choice that the officials found preferableto the lethal alternative. Sendero
Luminoso ("Shining Path") has consistently done similar things in the
Peruvian Sierra. These mutually exclusive claims to authoritymay help us
better to understand why guerrilla war is not fought like conventional
war and why terror, especially by incumbent governments, is so common
and often so massive. In conventional war, one expects to defeat the
enemy's soldiery and to make the enemy's citizenry acquiesce in that con-
quest, not hail it; however, in guerrilla war, the enemy is not just a mili-
tary foe but a renegade and heretic who rejects such claims to exclusive
In summary, governmentterroragainst the civilian populationin guerrilla
war is common because of the conflation of guerrilla, noncombatant,and
supportsystems, as well as the lack of informationon the partof government
soldiers necessary to sort out these categories. In addition,the intensityof the
"regime of terror" results from the view that the guerrillas-and all who
assist them-are renegades who reject an authority that defines itself as
The Incidence of Rape by GovernmentTroops
One of the most common featureswhich we have encounteredin our descrip-
tions of government terror in guerrilla warfare is rape. (In contrast, in a
decade of research, I have encounteredbut one or two such persuasive alle-
gations against leftist guerrillas.) How can we understandthe intensity and
regularityof rape, and often gang rape, of female civilians under the condi-
tions of guerrilla warfare?We can quickly dismiss the argumentabout the
licentious soldier, which views rape as the result of some form of pent-up
sexuality demanding release in any opportunity.Decades of systematic so-
ciological study of rape have increasinglydemonstratedthat rapeis a political
act ratherthana sexual one, and thereis no reasonto challengethatinterpreta-
tion here, as we shall see.87 While some scholarshave indeed trackeddown
some distinctive psychological profiles that differentiateboth rapists and the
men prone to rape from the generalpopulace, such psychological differences
85 Colonel Edward F. Callanan, "Terror in Venezuela," Military Review, 49
1969), 49-56; The New YorkTimes, 23 November 1963 and 2 December 1963; Gall, "Teodoro
Petkoff I," 16; Gott, Rural Guerrillas, 210.
86 For more thoughts on renegades, see Robert K. Merton, Social Theoryand Social Struc-
ture, 3rd. ed. (New York:The Free Press, 1968), 209-11, 323; CharlesTilly, FromMobilization
to Revolution (Reading, Mass.: Wesley-Addison, 1978), 106-14, deals analytically with the
way in which the acceptabilityof an acting group, ratherthanjust the qualityof its actions, affects
the way in which repressive forces treat it.
87 Edwin M. Schur, Labeling WomenDeviant: Gender, Stigma, and Social Control (New
York: Random House, 1984), 145-56.

are ill-suited to explain the recurrenceof rape in particular, well-defined

social situations, such as the counterinsurgencyoperationsin ruralareas and
rural villages under considerationhere. Moreover, Susan Brownmillerhas
made a strong, although not completely convincing, argumentabout the es-
sentially "ordinary"characterof the rapist. Just as HannahArendtpointedto
the "banality of evil" when consideringthe Nurembergdefendants,Brown-
miller has suggested thatone doesn't have to be a psychologicalmonsterto do
monstrous deeds.88
Confrontedwith the near-ubiquityof rape in counterinsurgencyoperations,
one can take two theoretical paths. One can argue, as Gerda Lerner and
Brownmiller do, that the rape of conquered females is as old as systematic
warfare itself, dating back in Lerner's estimation to the second millennium
B.C. If this were true, then our analysis could stop here; yet I am not con-
vinced. For one thing, the ancient evidence is extremely weak on the central
question, which is essentially a statistical one, as to the variations in the
frequencyof rape;indeed, thatevidence is incapableof addressingthe issue at
all or explaining the absence of rapeby guerrillasoldiers. Furthermore,while
Brownmiller clearly supports the general thesis that rape is naturalto male
warfare, her own rather more nuanced analyses of rape in warfare in this
centurysuggest a greatdeal of diversity, as she herself comes close to conced-
ing, if uncomfortablyso. The Germans, Japanese, and North Africans (the
last while participatingin the Europeanreconquest)seem to have been more
prone to commit rape than other armies. Certain soldiers from the United
States, in particularthose fighting in the Mekong Delta and coastal areas
dominatedby the Viet Cong in Vietnam, were apparentlymore proneto rape
than United States military personnel in that area in World War II.89
For those seeking to understandvariationsin culturesproneto rape, Peggy
Sanday and Rodney Stark have turnedup some highly instructivefindings.
Stark has statistically demonstratedthat higher rates of rape occur in those
states and provinces of the United States and Canada in which there are
relatively more men living in households with no women present. Roughly
similarresults show up in the city of Seattle, where districtswith higherratios
of unmarriedmen to unmarriedwomen have more rapes. (He also shows that
a substantial, positive, zero-ordercorrelationbetween rape rates and porno-
graphic exposure-indexed by Playboy circulation-disappears when con-
trolling for the relative number of male-only households.) Paralleling this
finding in her earliercross-culturalstudy, Sandayfound threeclear correlates
of societies which were more proneto rape:(1) a heavy stresson interpersonal
violence; (2) an ideology of male dominanceand toughness(which she appro-

88 Brownmiller, Against Our Will, ch. 6.

89 Gerda Lerner, The Creation of Patriarchy (New York: Oxford, 1986), 80-81;
Brownmiller, Against Our Will, ch. 3.

priately encoded by the label of "Macho"); and (3) the presence of male-
segregatedsocial institutions(the last echoing Stark'slaterstudy). She found
no significant correlationbetween rape and "sexually repressive" societies.
Sanday argued, in summarizing those findings, that rape would be more
common in those situationsin which there is a clear social disruptionof the
harmonybetween men and their naturalenvironments,which, her argument
suggests, should naturallyentail closeness to a variety of women.90
Those "disruptive" elements of separateness and social distance from
women dovetail with another increasingly consensual finding among so-
ciologists who study rape: Rape is likely to be more prevalentwhen women
are "objectified," including the proverbialtreatmentof them as nothing but
sex objects. Men who live in sex-segregated environmentsand those who
stress sharpmasculine-feminine differencesare precisely those lacking in the
nuanced and intimate contacts with women-including mothers, sisters,
friends, and sociosexual intimates-which would lead them to treatwomen in
subtler ways, not as objects. Such men seem to be the prime candidatesfor
To speak the obvious yet heretoforeunspoken:The barrackslife of soldiers
heavily emphasizes all three of the elements that Sandy found to be presentin
rape-pronecultures;hence, we would expect soldiers in generalto be particu-
larly prone to rape when they come into contactwith the civilian populationin
a context of war (thatis, "interpersonalviolence"). The aggravatingfeatures
of guerrilla warfare with regard to the incidence of rape lie in the extraor-
dinarilyintense "objectification" of the civilian populacethattakes place. To
reiteratemy earlierpoint, the civilian supportersof the guerrillashave placed
themselves, from the regime's perspective, outside of the body politic, as its
"enemies"; worse thanthat, they have declaredthemselvesto be hereticsand
renegades. The recurringuse of official language that utterly dehumanizes
and categorizes all of the civilians in guerrillazones-who become nothing
but "communists," "reds," and "subversives," to use the three foremost
terms-is an indication of the symbolic political excommunicationthat has
taken place; for they are no longer viewed as one's fellow Cubans, Guatema-
lans, Peruvians, and so forth. Those who have moved "beyond the pale" of
the body politic therefore suffer the double objectificationof being traitors,
and not simply the citizens of a conqueredenemy province.
In contrast to the behavior of governmenttroops, there is very little evi-
dence that left-wing guerrillashave committedrape against civilian women,
even in areas which have not provided support for the insurgency. Two
contributoryelements here are of an ideological nature:left-wing radicalism,
90 Rodney Stark, Instructor'sResource Book with Demonstrationsand Activities to Accom-
pany Sociology (Second Edition) (Belmont, Calif.: Wadsworth, 1987), 46-47, 51-53, 109-13;
Peggy Reeves Sanday, "The Socio-CulturalContextof Rape:A Cross-CulturalStudy," Journal
of Social Issues, 37:4 (Fall, 1981), 5-27, especially the results on 23-24.

unlike right-wing radicalism, has not generally glorified violence per se (al-
though the writings of Frantz Fanon give us pause here);91and the strong
egalitarianelements in all socialist ideologies, including the stress on com-
radeship, also militate against rape. Both of these left-wing ideological ele-
ments are the opposites of Sanday's first two findings about rape-prone
cultures (that is, they are the opposite of emphases on interpersonalviolence
and on male dominance). The last of Sanday's elements is relevant to the
soldier-guerrilladistinctionas well: The guerrillatypically lives in very close
cooperation with the local peasantry, including peasant women, and hence
does not have the utterly sex-segregated life of the soldier in the barracks.
Moreover, and perhaps the most sociologically importantfactor in differ-
entiating the two armies with respect to rape, the guerrillaarmies began to
include women in the 1950s and 1960s; and by the 1970s and afterwards,the
level of this integrationrangedanywherebetween 20 to 40 percentof the rebel
armies in Nicaragua, El Salvador, and Peru.92If there was any social struc-
tural feature of the guerrilla armies, as opposed to an ideological one, that
should limit the tendencytowardrapeby male guerrillas,it would be the close
cooperation with female guerrillas.
Peasants typically find themselves caught between the pressuresexerted by
both contendersfor power, and there is little doubt that they react with fear
and terrorat the prospect and reality of fire fights, bombings, and military
roundups. When possible, flight from an area is a logical course of action.
Such a step involves an enormoussacrifice in lost crops and changedlives yet
still often occurs. Indeed, the massive flight and emigration from the war
zones of Nicaragua, Guatemala, and El Salvador since 1975 serve only to
confirm that flight is a prime option for those affected by war. On a clearly
reducedscale, we have previouslyobservedsimilarphenomenain Venezuela,
as well as the en masse desertionby Peru's CampaIndiansfrom their villages
after bombing attacks and by Cuban squattersfrom the SierraMaestra, who
similarly ran away from Batista's terror.93
Yet peasantssubjectto terrormight insteadstandand fight. Moore suggests
thatone of the prerequisitesof revolutionaryactionby the lower classes is that

91 FrantzFanon, The Wretchedof the Earth (New York: Grove Press, 1966).
92 On leftist radicalism, again see Moore, Injustice, 420-33; on the breakdownof the gender
of the guerrillas, see George Black, Triumphof the People: The SandinistaRevolution in Nic-
aragua (London:Zed, 1981), 323-4; for El Salvadorand Nicaraguaboth, see StanfordCentral
America Action Network (SCAAN), eds., Revolution in Central America (Boulder, Colo.:
Westview, 1983), 383, 416-7; for Peru, McClintock, "Peru's Sendero Luminoso Rebellion,"
original manuscript,and a private communicationfrom Michael Smith.
93 Gall, "The ContinentalRevolution," 5; Minsterio de Guerra,Las Guerrillas en el Peru,
60; La Prensa (Lima), 23 June 1965; Guevara,Reminiscences,94. In the Peruviancase the army
reportedthat the mass exodus was due to the "excitement" caused by the guerrillas.

the prospect of flight or escape must seem impossible.94 Various guerrilla

groupshave agreed, at times indicatingthatpeasantsfoughtbecausetherewas
no other way out. Peasantresistanceis also fueled by moralindignationin the
face of government terror; for terror violates the fundamentalnorms that
supposedly govern the relationshipsbetween ruler and ruled.95
Most observers now recognize that governmentterrorleads to new waves
of recruitsfor guerrillabands. When the Bolivians requestedextensive mate-
rial assistance from the United States in 1967 to fight Che's guerrillas, an
official from the United States said, "We are certainlynot going to supply the
means for the Bolivian army hotheads to start bombing and napalmingvil-
lages or even suspected guerrilla hideaways. Civilians would inevitably be
killed and we have a long experience thatthis inevitablyproducesa streamof
recruits for the guerrillas."96
Government terror did indeed produce new recruits for guerrilla move-
ments in Latin America, and they may have been extremely zealous fighters.
Batista's terrorespecially drove the people of Santiago and the Sierras into
Castro's fold, as even Castro's critics have confirmed. Barquin wrote that
Batista's campaignof terror"sealed the alliance" between the guerrillasand
the squattersof the SierraMaestra,and Draperrefersto the "universalrevul-
sion" released by Batista's wave of violence.97 In Venezuela a guerrilla
suggested there was probably a new recruit for every woman raped by sol-
diers. After a particularlybrutal raid on a village in Lara State (described
above), the guerrillas received a whole new wave of recruits; and similar
events occurred in Colombia.98
How could such recruits be used? The outrage that created such peasant
warriors(in Moore's terms, this was an injectionof "iron in the soul") was
difficult indeed to contain within the organizationalboundariesof the guer-
rillas. One Venezuelan insurgent argued that "The campesinos want us to
fight to mitigate repression against them, but we cannot because it would
disrupt our plans for organizational consolidation."99 In Guatemala such
94 Moore, Injustice, 125; PeterA. Lupsha, "Explanationof Political Violence: Some Psycho-
logical Theories versus Indignation," Politics and Society, 2 (Fall 1971), 89-104.
95 Wickham-Crowley, "GuerrillaGovernments."
96 Gall, "The
Legacy of Che Guevara," 32 (quotation).Similar thoughtsare also expressed
in Lt. Colonel Harold R. Aaron, "Why Batista Lost," Army, 15 (September 1965), 71; Lt.
Colonel Neal G. Grumland, "The FormidableGuerrilla," Army, 12 (February1962), 65; and
ChalmersJohnson, "Civilian Loyalties and GuerrillaConflict," WorldPolitics, 14 (July 1962),
652. Benedict Kerkvliethas also suggested that governmentrepressiontransformsmere protests
into political rebellion; see The Huk Rebellion: A Study of Peasant Revolt in the Philippines
(Berkeley: University of CaliforniaPress, 1977), 260-2.
97 Daniel Friedenberg, "Notes on the Cuban Revolution," New Republic, 138 (17 February
1958), 15; Barquin, Las Luchas Guerrilleras, I, 319; Theodore Draper, Castro's Revolution:
Myths and Realities (New York: Praeger, 1962), 126.
98 Zago, Aqui No Ha Pasado Nada, 173-4; Regis Debray, Strategyfor Revolution, Robin
Blackburn,ed. (London: JonathanCape, 1970), 103.
99 El Nacional (Caracas), 31 August 1962.

cases were especially frequent (unsurprisingly, given the levels of terror

there). One peasantproudly relatedhis recruitmentafter the army had raided
his town and burneddown his hut. Anothersummedup his own battlemorale:
"I fight well because I hate well." One fighter, Rocael, had seen his uncle
torturedand his penis cut off and stuffed in his mouth. Luis Turcios Lima
summarizedthe attitudesof such recruits:"It was also significantto see with
what rage those campesinos fought whose families had been assassinated."
As in Venezuela, the guerrilla leaders tried to mold the battle lust of the
peasants to the perceived organizational,military, and political needs of the
guerrillas.Peasantattitudeswere ones of "desperateangerwantingto kill and
be killed killing," but "they have to be calmed down" and the "spirit of
classical peasantrevolt has to be channeledinto the painstakingand scientific
strategy of the conscious revolutionary."100
One variation on this theme is that when terror produces a temporary
contractionof guerrilla-peasantcooperation, it portendsa longer-termcoop-
eration with the insurgentsand alienation from the authorities;hence, Bejar
claimed that Peru's 1965 wave of terror in Ayacucho produced "an un-
bridgeable abyss" between peasants and governmentauthorities.While the
evidence is thin for Peru, Japan's "Three-All" policy in 1941 in ruralChina
("Kill all! Bur all! Destroy all!") did indeed producethat temporaryguer-
rilla contractionwhich was followed later by prolongedpeasant supportand
movement expansion. 01
When terror, however, is combined with civic action or other benefits, it
may breakthe back of the insurgency. The guerrillamovement, as a counter-
state in formation, is obliged to protectthe peasantryfrom such violence; and
a lack of protection may lead peasants to withdrawtheir support. We now
have recordedinstances of such "peasant withdrawal,"on a greateror lesser
scale, with the Chinese Communistsin the 1930s and 1940s and with certain
areas of Guatemala,Colombia, and Peru-the latterhappeningboth in 1965
and the 1980s during the Sendero period. Under such conditions of with-
drawal, it would not be surprisingto see the emergenceof an attitudesuch as
"a plague on both your houses."102

100 Adolfo Gilly, "The Guerrilla Movement in GuatemalaI," Monthly Review, 17 (May
1965), 24-25; Munson, Zacapa, 194-5; Schump, Las Guerrillasen AmericaLatina, 55; A. P.
Short, "Conversationswith the GuatemalanDelegates in Cuba," MonthlyReview, 18 (February
1967), 37.
101 Guevara, Reminiscences, 192-4; Hector Bejar, "Bilan d'une Gu6rilla au Perou," Par-
tisans, 37 (April-June 1967), 98; B6jar, "Ne Pas SurestimerSes Forces," 111; Lt. Colonel John
W. Woodmansee, Jr., "Mao's ProtractedWar:Theory and Practice," Parameters, 3:1 (1973),
102 See
Wickham-Crowley, "GuerrillaGovernments," 492-3 for more details on all but the
final case; for some evidence that Sendero Luminoso's supportin the Andes has fallen because
they failed to protectthe peasantry,see McClintock, "Peru's SenderoLuminosoRebellion," 90.


The preceding descriptions and explanations of terrorcould clearly be ex-

tended past this period to encompass the far more widespread terror in
post-1970 guerrillawarfarein Latin America. Tens of thousandsof innocent
civilians have perishedundersuch terror;indeed, that many have been killed
by governmentsand death squads in both El Salvadorand Guatemalasince
1975. The documentationof this second, more violent wave of terroris far
superior to that of the earlier period, in good part due to the efforts of
Amnesty Internationaland other independentmonitorsof humanrights vio-
lations.103Moreover, in at least two cases-El Salvadorand Peru-the guer-
rillas themselves have also steppedup the degree of terrorby taking the lives
of well over a thousand-perhaps many thousands-of civilians in their own
waves of terror.104 I lack the space to pursuethese more recent comparisons
in greater detail. Instead I simply note in conclusion that, as if to confirm
(tragically)this perspectivedown to the goriest details, guerrillawarfarehas
once again given rise not only to even more widespreadterrorbut also to
argumentsand excuses on behalf of terrorthat exactly parallelthose so regu-
larly offered by governmentsand insurgentsin the precedingtwo decades.105
Plus (a change, plus c'est la meme chose.

103 For information

by Amnesty International(A.I.) on a variety of cases, see their The
Republicof Nicaragua (Nottingham,England:The Russell Press, 1977); a capsule summaryis in
"Guatemala:A GovernmentProgramof Political Murder," in Guatemala in Rebellion: Un-
finished History, JonathanFried et al., eds. (N.Y.: Grove, 1983), 139-45; and a summaryof
their findings for the early period underRios Montt, in LatinAmerica WeeklyReport (London),
15 October 1982, 11; for Guatemalain 1983, see as well the summaryof the Americas Watch
findings, in "Exterminationin Guatemala,"New YorkReviewof Books, 30:9 (2 June 1983), 13-
16; two summaries of A.I.'s reports on Peru and El Salvador appear in, respectively, Latin
America WeeklyReport, 23 September 1983, 10; and Latin AmericaPolitical Report (London;
same periodical), 21 March 1980, 12. Independentchurch authoritiesalso criticized and docu-
mented governmentterroras well, especially in Nicaraguaand El Salvador. For more detail on
governmentterror,mostly from left-wing sources, see, for Nicaragua:Doris Tijerino,Inside the
Nicaraguan Revolution, As Told to Margaret Randall (Vancouver:New Star, 1978), 165-75;
Latin America Political Report, 29 June 1979, 194-6; New YorkTimes, 2 March 1977, Section
2, p. 1. For Guatemala, see George Black et al., Garrison Guatemala (New York: Monthly
Review, 1984), 94-97; ConcernedGuatemalaScholars, Guatemala:Dare to Struggle, Dare to
Win (San Francisco, Calif.: Solidarity, 1982), 67-72; and Proceso (Mexico City), no. 412 (24
September 1984), 40-43, for an extensive list of massacresand body counts. For El Salvador,
see Vicente Navarro, "Genocide in El Salvador,"MonthlyReview, 32:11 (April 1981), 1-16; on
reportsof U.S. governmentbackingfor terror,see Allan Nairn, "Behind the DeathSquads," The
Progressive, 48:5 (May 1984), 20-29, and ShirleyChristian,"El Salvador'sDivided Military,"
AtlanticMonthly, 251:6 (June 1983), 50-60 on the militaryand civilian links to terror.For Peru,
see CynthiaMcClintock, "Sendero Luminoso:Peru's Maoist Guerrillas,"Problemsof Commu-
nism, 32:5 (September-October 1983), 30-32.
104 Guerrillaterroralso has expandedin scope, with the insurgentsusing far morekidnapping,
extortion, bank robbery, attackson economic and utility targets, and interdictionsof road traffic
than in the past. All of these have occurredin El Salvador, at least a few in other locales. On
Sendero's terrorin Peru, with the estimatedcivilian deaths well into the thousands,see McClin-
tock, "Sendero Luminoso," 19, 32; Mario Vargas Llosa, "Inquest in the Andes," New York
Times Magazine, 31 July 1983, 18-23 et passim (which also documents governmentterror);

(LatinAmerica) WeeklyReport, 18 January1985, 10; and 13 January1984, 10-11. Widespread

agreementon the Sendero's wave of terroris not sharedin the literatureon El Salvador, where
left-wing scholars routinelydeny such features. For suggestive evidence, includingreferencesto
the guerrillas' self-recordedkillings of thousandsof civilians, see Penn Kemble, "The Liberal
Test in El Salvador," The New Republic, 14 March 1981, 18-19; R. Bruce McColm, El
Salvador: Peaceful Revolutionor ArmedStruggle? (New York:FreedomHouse, 1982), 23-24,
43; Mark Falcoff and Robert Royal, eds., Crisis and Opportunity:U.S. Policy in Central
America and the Caribbean (Washington, D.C.: Ethics and Public Policy Center, 1984), 199-
200, 269; Michel G. Englebert(Interviewer), "Flight: Six SalvadoransWho Took Leave of the
War," The Progressive, 47:3 (March 1983), 38-43, includingvictims of both sides; and (Latin
America) WeeklyReport, 16 March 1984, 1; for a moregeneralview, see GabrielZaid, "Enemy
Colleagues: A Reading of the SalvadoranTragedy," Dissent (Winter 1982), 13-40.
105 This is clearest in Guatemala and El Salvador; in the former, the government under
General Efrafn Rios Montt (1982-83) legitimated terror by the army (not the death squads)
against civilians as justifiable attackson "subversives"; for a typical defense (that is, by mini-
mizationandjustification)of the terrorby the guerrillasin El Salvador,see Berryman,"Another
View of El Salvador," and Zaid's critique, in "Gabriel Zaid Replies."