Assignment Instructions: Write a 3-5 page (750-1250 words) essay, answering the following question:
¡, IÍ " ,,, fJ.' ,, .: (Í ,; e] 11 'li )\
l. E x
p la
n a
tio n
s o f P
o litica
l V io
len ce
C onflict ... is a them
e that has occupied ti1e thinking of m
an m ore than any other,
save only G od and !ove. A
natol R apoport,
F ights, C
arnes, and D ebates
T H
E IN
S T
IT U
T IO
N S
, persons, an d
policies of rulers have inspired th
e violent w rath of th
eir nom inal subjects throughout the history
of organized political life. A survey of the histories of E
u ro
p ean
states
an d
em pires, spanning tw
enty-four centuries, show s that
th ey
averaged only four peaceful years for each year of violent dis-
ff f
turbances.1 M o
d em
nations have no b etter record: b
etw een
1961 an
d 1968 sorne form
of violent civil conflict reportedly occurred in ll4
of the w orld's 121 larger nations an
d colonies.
2 M ost acts of
group violence have negligible effects on political life; b u
t sorne have b
een enorm
ously destructive of h u
m an
life an d
corrosive of political institutions. T
en of th
e w orld' s thirteen m
ost deadly con- flicts in th
e past H>O years have b een
civil w ars an
d rebellions; 3
since 1945, violent attem pts to overthrow
govem m
ents have b een
"r'
m ore com
m on than national elections. T
h e counterpoise to this
grim record is th
e fact that political violence has som etim
es led
to th e creation of n
ew an
d m
ore satisfying political com m
unities. T
h e consequences of th
e A m
erican, T urkish, M
exican, an d
R us-
sian revolutions testify in different w
ays to th e occasional benefi-
cen ce of violence.
In this study political violence refers to all collective attacks
1 P itirim
S orokin, Social and C
ultural D ynam
ics, V ol. III: F
luctuations o f Social
R elationships, W
ar and R evolutions (N
ew Y
m k: A
m erican B
ook C o., 1937), 409-
475. T w
elve countries and em pires w
ere studied over the period 500 B .C
. to 1925 A
.D ., none of them
for the entire span. O nly "im
portant" disturbances w ere re-
corded, i.e., those m entioned in standard histories.
2 B ased in part on data for 114 polities reported in
T ed
R obert G
urr, "A C
om -
parative S tudy of C
ivil S trife," in H
ugh D avis G
raham and T
ed R
obert G urr, eds.,
V iolence in A
m erica: H
istorical and C om
parative P erspectives (W
ashington, D .C
.: N
ational C om
m ission on the C
auses and P revention of V
iolence, 1969). 3 A
ccording to data collected by L ew
is F . R
ichardson, Statistics o f D
eadly Q uar-
rels (P ittsburgh: T
h e B
oxw ood P
ress, 1960), 32-43, to w hich events after 1948 are
added here. T h
e conflic;ts m ost destructive of life w
ere W orld W
ars 1 and Il. O f the
other eleven al! b u
t one w ere prim
arily interna! w ars and al! caused m
ore than 300,000 deaths each; the T
ai-P 'ing R
ebellion, 1851-64; the A m
erican C ivil W
ar; the G
reat W ar in L
a P lata, 1865-70; the post-revolutionary C
ivil W ar in R
ussia, 1918-20;
the first
and second C hinese C
ivil W ars, 1927-36 and 1945-49; the
S panish C
ivil W ar; the com
m unal riots in India and P
akistan, 1946-48; the V iet-
nam W
ar, 1961-present; the prívate w ar betw
een Indonesian C om
m unists and their
opponents, 1964-66 (casualty
figures are
problem atic);
and the N igerian C
ivil W
ar, 1 9
6 7
-.
3 t i
r . J t; t ! 1 ' 1
W H
Y M
E N
R E
B E
L
w ithin a political com
m unity against th
e political regim e, its actors
- including com
peting political groups as w ell as in
cu rn
b en
ts- or its policies. T
h e co
n cep
t represents a set of events, a com m
on property of w
hich is the actual or th reaten
ed use of violence, b
u t
the explanation is n o
t lim ited
to th at property. T
h e co
n cep
t sub- sum
es revolution, ordinarily defined as fundam ental sociopoliti-
cal change accom plished through violence. lt also includes guer-
rilla w ars, coups d
' état, rebellions, an d
riots. P olitical violence is
in tu rn
su b
su m
ed u
n d
er "force," th e use or threat of violence b
y
any party or institution to attain en d
s w ithin or outside the political
order. T h
e definition is n o
t b ased
on a prejudgm ent th
at political violence is undesirable. L
ike th e uses of violence q
u a force by th
e state, specific acts of political violence can b
e good, bad, or neutral according to th
e view point of th
e observer. P articipants in
political violence m
ay value it as a m eans 'of expressing political dem
ands or opposing u
n d
esirab le policies. L
im ited violence also can b
e useful for rulers an
d for a political system
generally, especially as an expression of social m
alaise w h
en other m
eans for m aking
dem ands are inadequate. E
thical judgm ents are h
eld in abeyance
in this study to avoid dictating its conclusions. B
ut it does not re- quire an ethical ju
d g
m en
t to observe that in ten
se violence is de- structive: even if sorne political violence is v
alu ed
by b o
th citizens
an d
rulers, the greater its m agnitude the less efficiently a political
system fulfills
its o
th er functions. V
iolence generally consum es
m en
and goods, it seldom enhances them
. 4
D esp
ite the frequency an d
social ím pact of political violence, it
is not now a conventional category of social analysis. Y
et sorne com -
m on properties of political violence encourage attention to it rather
than m ore general or m
ore specific concepts. T heoretically, all
such acts pose a th reat to th
e political system in
tw o senses: th
ey
challenge the m onopoly of force im
p u
ted to th
e state in political theory; and, in
functional term s, th
ey are likely to interfere w
ith and, if severe, to destroy norm
al political processes. E m
pirical justification for selecting political violence as a universe for analy- sis is provided by statistical ev
id en
ce that political violence com -
prises events distinct from o
th er m
easured characteristics of na- tions, an
d hom
ogeneous en o
u g
h to justify analysis of th
eir com m
on characteristics
an d
causes. F
o r exam
ple, countries experiencing 4 It is possible th
at political violence can increase the sum total of satisfactions
of society's m em
bers. T his can b
e true if violence and its im m
ediate effects are intrinsically valued m
ore than th e m
aterial and h u
m an
resources it consum es, or
if violence serves a popularly approved regulatory function, as it d id
for th e A
m eri-
can vigilante m ovem
ents. __ A hypothetical relationship of this type is show
n sche- m
atically in fig. A
. lt is likely th at high m
agnitudes of violence destroy m ore than
they create, at least in th
e short run. W h
en the tim
e dim ension is taken into ac-
4
E X
PL A
N A
T IO
N S O
F P
O L
IT IC
A L
V IO
L E
N C
E
extensive poli ti cal violence of one k in
d -
w h
eth er riots, terrorism
, coups d'etat, or guerrilla w
a r-a
re rather likely to experience other
kinds of political violence, b u
t are n eith
er m ore or less likely to
b e en
g ag
ed in
foreign conflict. 5 T h
e properties an d
processes that distinguish
a riot from
a
revolution are substantively an d
the- oretically interesting, and are exam
ined at len g
th in
this study, b u
t at a general leve! of analysis th
ey seem
to b e differences of degree,
not kind. 6 T h
e search for general causes an d
processes of política! violence is further en
co u
rag ed
by th e convergence of recen
t case, com
parative, an d
theoretical studies. O n
e striking feature of these studies is th
e sim ilarity of m
any of th e causal factors and proposi-
tions th ey
identify, w h
eth er they deal w
ith revolution, urban riot- ing, or o
th er forros of political violence. T
his sim ilarity suggests
th at sorne of th
eir findings can b e synthesized in
a m ore efficient
set of testable generalizations. H
o w
ev er good th
e prospects seem for a general analysis of poli ti-
cal violence, research o n
it has b een
quite u n
ev en
, both in sub-
count, how ever, in
ten se political violence, though it destroys m
uch in th e short
run, m ay have th
e long-run payoffs eith er of stim
ulating rulers to increase outputs or of restructuring society in
such a w ay that total satisfactions are substantially
increased. T his kind of relationship is sketched in fig. B
.
G ain
O verall
V alue
P osition
M agnitudes"'-.of
L oss
1 violence
F igure A
high
F igure A
. H
ypothetical effects o f vio-
len ce o
n satisfactions in
a society in w
hich violence is valued, at tim e X
.
G ain
O verall
V olue
P osition
L oss
Y eor
F igure B
F igure B
. H
ypothetical effects of in- tense
violence on
satisfactions in
a society in w
hich violence leads to re- forros, over tim
e. 5 R
udolph R
um m
el, "D
im ensions
of C onflict
B ehavior W
ithin an
d
B etw
een N
ations," G
eneral System
s Y
earbook, V
III (1963),
1-50; an
d
R aym
ond T anter,
"D im
ensions of
C onflict
B ehavior W
ithin and
B etw
een
N ations,
1958-1960," ]ournal o
f C onfiict R
esolution, X (M
arch 1966), 41-64. T h
ere are statistically sig- nificant relations am
ong sorne form s of interna! and externa! conflict, b
u t they are
relatively w eak. F
o r factor-analytic evidence that types of political violence are
in d
ep en
d en
t of m ost m
easu red
characteristics of nations, see R udolph R
um m
el, "D
im ensionality of N
ations P roject; O
rthogonally R otated F
actor T ables for 236
V ariables," D
ep artm
en t of P
olitical S cience, Y
ale U niversity (N
ew H
aven, July 1964), m
im eographed.
8 B illington points out, for exam
ple, that the R ussian R
evolution com prised an
''insurgency" from b
elo w
-actu ally
extensive rioting an d
m u
tin ies-fo
llo w
ed by a
coup d 'état a
n d
a protracted civil w
ar. Jam es H
. B illington, "S
ix V iew
s of the R us-
sian R evolution," W
orld P olitics, X
V III (A
pril 1966), 452ff.
5
'1 Ir 1 11
W IIY
M E
N llE
B E
L
stance an d
in disciplinary approach. T
h ere is considerable E
uro- p
ean historical scholarship on segm
ents of th e subject, notably th
e peasant rebellions of th
e tw elfth through n
in eteen
th centuries an
d
the great revolutions o fE
n g
lan d
, F rance, an
d R
ussia. A m
erican an d
E
u ro
p ean
scholars, m ost of th
em also historians, have in
recen t
years contributed a m o
d est case-study literature. A
m erican policy
scientists have w ritten a sm
all flood of treatises on the causes an d
prophylaxis of subversive w
arfare, m ost of w
hich seem to have h
ad
n eith
er academ ic nor policy im
pact. T h
e lapses of attention are striking b
y com
parison. O f all the riotous m
obs that have clam oured
through th e streets of history, only the revolutionary crow
ds of eighteenth-
and nineteenth-century E u
ro p
e an d
th e ghetto rioters
of tw entieth-century A
m erica have attracted m
uch scholarly atten- tion.7 T
h ere are relatively few
case studies of political violence in the non-W
estern w orld, an
d few
er system atic com
parative studies or
attem pts
at em
pírica! theory.
E xperim
ental studies
dealing w
ith socia1-psycho1ogical m echanism
s o f collective violence can
b e counted on one hand. 8 A
m ong social scientists the historians
have b een
by far th e m
ost active; A m
erican política! scientists have until
n eg
lected th
e subject. 9 O f 2,828 articles that ap-
peared in the A
m erican P
olítica[ Science R eview
from its estab-
7 A useful survey o
f current know ledge of urban an
d rural tunnoil in eighteenth-
and nineteenth-century E u
ro p
e is G eorge R
udé, T he C
row d in H
istory: A Study
o f P
opular D isturbances in F
rance and E ngland, 1730-1848 (N
ew Y
ork: W iley,
1964). A representative survey of cu
rren t know
ledge of urban violence in m
odern societies is L
ouis H . M
asotti an d
D o
n R
. B ow
en, eds., R iots and R
ebellion: C ivil
V iolence in the U
rban C o
m m
u n
ity (B everly H
ills: S age, 1968).
8 T h
e studies,
using a
variety of experim
ental techrtiques,
include N
onnan P
olansky, H onald L
ippitt, am i F
ritz R edl, ''A
n Investigation of B ehavioral C
on- tagian in G
roups," H um
an R elations, ll1
(N o. 3, 1950), 319-348; G
. E . S
w anson,
"A P
relim inary L
aboratory S tudy of th
e A cting C
row d," A
m erican Sociological
R eview
, X V
III (O ctober 1953), 522-533; N
orm an C
. M eier, G
. H . M
ennenga, and H
. J. S toltz, "A
n E xperim
ental A pproach to th
e S tudy o
f M ob B
ehavior," jo u
m a
l o
f A bnorm
al and Social P sychology, X
X X
V I (O
ctober 1941), 506-524; K urt L
ew in,
R onald L
ippitt, an d
R alph K. W
hite, "P atterns of A
ggressive B ehavior in E
xperi- m
entally C reated S
ocial C lim
ates," jo u
m a
l o f Social P
sychology, X (M
ay 1939), 271-299;
an d
D avid S
chw artz, "P
olitical A lienation: A
P relim
inary E x
p erim
en t
on the P sychology of R
evolution's F irst S
tage," p ap
er read at th e A
nnual M eeting
of the A
m erican P
olitical S
cience A ssociation, 1967.
M any other experim
ental studies are relevant to collective violence; these deal explicitly w
ith social-psy- chological factors facilitating violence.
• T h
e renaissance of system atic theoretical an
d em
pirical w ork on poli ti cal vio-
len ce by poli ti cal scientists can b
e d ated
from th
e publication of three articles that appeared in
1962 and 1963: Jam es C
. D avies, "T
o w
ard a T
heory o f H
evolution," A
m erican
Sociological R
eview ,
X X
V II
(F ebruary
1962), 5-19;
H arry
E ckstein,
"Interna! W ar: T
h e P
roblem -
of A nticipation," in
Ithiel de S ola P
ool et al., Social R
esearch and N ational Security: A
R eport P
repared b y the R
esearch G roup in P
sy- chology
and the
Social Sciences
(W ashington,
D .C
.: S
m ithsonian
Institution, M
arch 5, 1963), p u
b lish
ed in revised f01m
as "O n
th e E
tiology of Interna! W ars,"
H istory and T
heory, IV (N
o. 2, 1965), 133-163; an d
R um
m el, "D
im ensions of C
on- flict B
ehavior."
6
E X
P L
A N
A T
IO N
S O
F P
O L
IT IC
A L
V IO
L E
N C
E
lish m
en t in 1906 through 1968, only tw
enty-nine ap p
ear from their
titles to b e concerned w
ith political disorder or violence. M ore-
over tw elve of th
e tw enty-nine w
ere concerned specifically w ith
revolution, an d
fifteen ap p
eared after 1961. 1
0
P olítica! scientists m
ight b e ex
p ected
to have a greater concern w
ith political violence th an
others. A uthoritative coercían in the
service of the state is a crucial co n
cep t in
política! theory an d
an issue of continuing d
isp u
teY S
orne h av
e identified th e distinctive
characteristic of the state as its m onopoly of physical coercion.
M ax W
eber, for exam ple, w
rote that violence is a "m ean
s specific" to th
e state an d
that "th e right of physical violence is assigned to
all o th
er associations or individuals only to th e extent perm
itted b
y th
e state; it is su p
p o
sed to b
e th e exclusive source of the 'right'
to use violence." 1 2
T hom
as H o
b b
es, dism ayed b
y the brutish an-
archy of m en
living outside the restraint of com m
onw ealths, con-
ceiv ed
the sovereign's control of coercían to b e tl1e foundation of
th e state an
d th
e social co n
d itio
n P
S chattschneider sees conflict,
w hich
subsum es
violence, as
the central
concept of political
science. 14
N ieberg
em phasizes
the positive
functions of non-
authoritative violence an d
its th reaten
ed use as an instrum
ent of social ch
an g
eY F
ro m
an y
o f th
ese perspectives the occurrence of collective, nonauthoritative violence appears to pose fw
o funda- m
ental questions for política! science: F rom
w hat sources and by
w h
at processes does it arise, an d
how does it affect the política!
an d
social order?
W h
at Is to B e E
xp lain
ed ?
T his study propases sorne general answ
ers to th ree basic ques-
tions about our occasional
disposition to
disrupt violently th e
ard er w
e otherw ise w
ork so h ard
to m aintain: W
hat are the psy- 10 B
y contrast, 111 a1ticles deal specifically w
ith G reat B
ritain, sorne 140 w ith
th e U
.S. C ongress, an
d about 250 w
ith constitutions or constitutional issues. T h
ese counts w
ere facilitated by th e use of the K
ey-W ord-in-C
ontext index o f article titles
in K
enneth Janda, ed., C um
ulative Index to the A m
erican P olitical Science R
e- view
: V olum
es 1-57: 1906-1963 (E vanston: N
orthw estern U
niversity P ress, 1964).
11 A useful analysis o
f alternative conceptions o f authoritative violence is E
. V .
W alter, "P
o w
er an d
V iolence," A
m erican P
olitical Science R eview
, L V
III (June 1964), 350-360. A
sürvey of th e origins, functions, an
d form
s o f legitim
ate and il- legitim
ate violence, w ithin an
d am
ong states, is C h
ap ter 8, "C
ivil C onflict and W
ar," A
lfred d e G
razia, P olitics and G
overnm ent, V
ol. 1: P olitical B
ehavior, rev. edn. (N
ew Y
ork: C ollier, 19q2, 1962), 265-301.
12 M ax W
eber, "P olitik als B
eruf," G esam
m elte P
olitische Schriften (T ubingen:
J. C . B
. M ohr, 1958), 494, translated in
E . V
. W alter, 359.
13 T hom
as H obbes, L
eviathan (O xford: B
asil B lackw
ell, 1946). 14 E
. E . S
chattschneider, "Intensity, V isibility, D
irection an d
S cope," A
m erican
P olítica! Science R
eview , L
I (D ecem
b er 1957), 933-942.
1 5
H . L
. N
ieburg, "T h
e T h
reat o f V
iolence an d
S ocial C
hange," A m
erican Po- lítica! Science R
eview , L
V I (D
ecem b
er 1962), 865-873, an d
P olítica! V
iolence: T he
B ehavioral P
rocess (N ew
Y ork: St. M
artin's P ress, 1969).
7 ,
I t r. g e e o n vi at tu h; m
pe se di: R
t Ft' e o se' " (P
r gra
t T
ar Int• R
os
J C
ro 271
8
r \' i i, 1:: t i l l " t h:. r E' ' t i ¡ 1
W H
Y M
E N
R E
B E
L
chological and social sources of th e potential for collective vio-
lence? W hat
determ ines
the extent to
w hich
that potential is
focused on the poli ti cal system ? A
nd w h
at societal conditions af- fect the m
agnitude an d
form , an
d h
en ce the consequences, of vio-
lence? T h
e study has four prim ary objects of analysis. T
w o are
intervening variables: th e potential for collective violence an
d th
e potential for political violence. P
ropositionally, potential for col- lective violence is a function of th
e extent an d
intensity of sh ared
discontents am
ong m em
bers of a society; the potential for political violence is a function of the d
eg ree to w
hich such discontents are b
lam ed
on the political system an
d its agents. T
h e rem
aining ob- jects of analysis are d
ep en
d en
t variables: the m agnitude o
f political violence an
d the form
s o f political violence, both of w
hich are dis- cussed below
. T
heories of revolution are usually concerned w ith specifying a
relationship b etw
een sorne set of preconditions an
d th
e occur- rence of revolution.
P olitical violence, how
ever, is a pervasive
phenom enon, as w
as suggested above: few contem
porary or his- torical societies have b
een free of it for long. It m
ay b e useful for
m icroanalysis to specify w
h eth
er política! violence is likely in a
given society a ta
particular point in tim
e. F o
r m acroanalysis, how
- ever, the m
ore interesting questions are th e determ
inants of the extent of violence an
d of th
e forros in w
hich it is m anifested. If
one's interest is the effects o f political violence on the political
system , the questions of its m
agnitude an d
kind are b o
th highly
relevant. A nd if o
n e is co
n cern
ed in
an ethical w ay w
ith political violence,
th en
alm
ost certainly one w ants to assess
its h u
m an
and m
aterial costs, an d
co n
seq u
en tly
th e determ
inants of its m agni-
tude. V arious m
easures of th e relative extent o
f política! violence have b
een used in recen
t com parative studies. S
orokin co m
b in
ed
m easures of the proportion of a nation affected (social area), pro-
portian of population actively involved, duration, intensity, an d
severity of effects of violence in
assessing the m agnitude of interna!
disturbances. T illy an
d R
ule m ake use of m
an-days of participation. R
um m
el and T an
ter have u sed
counts of num bers o
f events. T h
e F
eierab en
d s have d
ev elo
p ed
a scaling procedure that takes ac- count of both n
u m
b er of events an
d a pr,iori judgm
ents about th e
severity of events of various types. 16 S orne researchers have u
sed
16 Sorokin, op. cit.; C harles T
illy and Jam es R
ule, M easuring P
olitical U pheaval
(Princeton: C enter of lnternational S
tudies, P rinceton U
niversity, R esearch M
ono- graph N
o. 19, 1965); R um
m el, "D
im ensions of C
onflict B ehavior .
. . "; R
aym ond
T anter, "D
im ensions of C
onflict B ehavior W
ithin N ations, 1955-60: T
urm oil and
Interna! W ar," P
eace R esearch Societies P
apers, III (1965), 159-164; and Ivo K. and R
osalind L .
F eierabend, "A
ggressive B
ehaviors W ithin
P olities, 1948-1962; A
C
ross N ational S
tudy," ]ournal o f C
onflict R esolutíon, X
(S eptem
ber 1966), 249- 271. 8
l.
¡,·
E X
P L
A N
A T
IO N
S O
F P
O L
IT IC
A L
V IO
L E
N C
E
the grisly calculus of n u
m b
er of deaths resulting from v
io len
ceY
T h
e p ro
p o
sed relation b
etw een
p erceiv
ed deprivation an
d the
frustration concept in frustration-anger-aggression theory, to b
e dis- cussed in
ch ap
ter 2, provides a rationale for a m ore general defini-
tion of m agnitude of violence an
d a m
ore precise specification of w
hat it com prises. T
h e basic frustration-aggression proposition is
that the greater th e frustration, th
e greater th e quantity of aggres-
sion against th e source of frustration. T
his postulate provides the m
otivational base for an initial proposition about poli ti cal violence: the greater th
e intensity of deprivation, the greater th e m
agnitude of violence. (O
ther perceptural an d
m otivational factors are also
relevant to política! violence, b u
t m any of th
em can b
e su b
su m
ed
by th e deprivation cóncept, as is suggested in ch
ap ter 2.) In
ten se
frustration can m otívate m
en eith
er to intense, short-term attacks
o rto
m ore prolonged, less severe attacks on th
eir frustrators. W hich
tactic is chosen is probably a function of anticipated gain, oppor- tunity,
an d
fear of retribution,
w h
ich
in política! violence are
situationally determ ined. H
en ce th
e severity of deprivation affects both th
e intensity of violence, i.e. in th
e extent of h u
m an
and phys- ical dam
age incurred, an d
its duration. M
oreover there are evi- dently individual d
ifferen ces-presum
ably norm ally d
istrib u
ted -
in th e intensity of frustration n
eed ed
to precipitate overt aggres- sion. E
xtension o f this principie to th
e deprivation-violence rela- tionship suggests th
at th e proportion o
f a population th at partici-
pates in violence ought to vary w
ith th e average intensity of per-
ceiv ed
deprivation.
M ild deprivation w
ill m
otívate few to vio-
lence, m oderate deprivation w
ill p u
sh m
ore across th e threshold,
very in ten
se deprivation is likely to galvanize large segm ents of a
political com m
unity into action. T
his arg
u m
en t suggests
th at m
agnitude of political
violence has th
ree com ponent variables that ought to b
e taken into account in
system atic analysis: th
e extent of participation w ithin the politi-
cal u n
it b ein
g stu
d ied
(scope), th e destructiveness of action (in-
tensity), an
d
th e
len g
th
o f tim
e violence
persists (duration).
S orokin's em
pírica! w ork takes all three aspects into account; so
does m ine. 18
T h
e intensity an d
scope of relative deprivation an d
m agnitude of
17 A m
ong them R
ichardson, especially chap. 2 and 4; R um
m el, "D
im ensions of
C onflict B
ehavior"; T anter; and B
roce M .
R ussett, "Inequality and Instability:
T he R
elation of L and T
enure to P olitics," W
orld P olitics, X
V I (A
pril 1964), 442- 454.
16 T ed
G urr w
ith C harles R
uttenberg, T he C
onditions o f C
ivil V iolence: F
irst Tests o
f a C ausal M
odel (P rinceton: C
enter of Intem ational S
tudies, P rinceton
U niversity, R
esearch M onogm
ph N o. 28, 1967); T
ed G urr, "A
C ausal M
odel of C
ivil Strife: A C
om parative A
nalysis U sing N
ew lndices," A
m erican P
olitical Sci- ence R
eview , L
X II (D
ecem ber 1968), ll04-1124.
9
fjj ::¡· " ¡, ii l' J
W H
Y M
E N
R E
B E
L
violence are
unidim ensional
variables. T
heoretically, an
d em
- pirically, one can conceive of degrees or quantities o
f each in any
polity. T h
e form s of violence, how
ever, are attributes that do not form
a sim ple dim
ension. A society m
ay experience riots b u
t not revolution, revolution b
u t n
o t coups d
' état, coups d ' état b
u t n
o t
riots. H ypotheses about form
s of violence as d ep
en d
en t variables
thus are
necessarily different from those about deprivation an
d
m agnitude of violence. T
h ey
are expressed in term
s of probabilities (the greater x, th
e m ore likely y) rather than strict concom
itance. T
h e question is h
o w
m any form
s of political violence ought to b e
accounted for in a
general theory. T h
e principie of parsim ony,
w hich should apply to d
ep en
d en
t as w ell as in
d ep
en d
en t variables,
suggests using a typology w ith a sm
all n u
m b
er of categories, events in each of w
hich are fairly num erous.
C onventional
taxonom ies,
o f w
hich there
are m
any, provide
little help. S orne; like that of L
assw ell an
d K
aplan, provide sim ple
typologies for revolutions b u
t n o
t for política! violence generally. 19 E
ckstein propases a eom posite typology com
prising unorganized, spontaneous violence (riots), intraelite conflicts (coups), tw
o varie- ties of revolution, an
d w
ars of in d
ep en
d en
ce. 20 P erhaps th
e m ost
com plex typology is R
u m
m el' s list of tw
enty-five types of dom es tic
conflict, the analysis o f w
h ich
provides an em pírica! solution to th
e problem
of a parsim onious typology. In
R um
m el's analysis, a
n d
in a
num ber of su
b seq
u en
t studies, data on th e incidence an
d character-
istics of various types of poli ti cal violence w ere collected an
d tabu-
lated by country an d
th e "co
u n
try scores" (num
ber of riots, assassi- nations, coups, m
utinies, guerrilla w ars, an
d so on, in
a given tim e
period) w ere factor analyzed. W
hatever th e typology em
ployed, the period of reference, or th
e set of countries, essentially the sam e
results w ere reported. A
strong turm oil dim
ension is characterized by largely spontaneous strife such as riots an
d dem
onstrations. It is quite distinct both statistically an
d substantively from
w hat can
b e called a revolutionary dim
ension, characterized b y
m ore or-
ganized an d
intense strife. T his revolutionary dim
ension has tw o
com ponents that appear in
sorne analyses as separate dim ensions:
interna[ w
ar, typically
in clu
d in
g
civil w
ar, guerrilla w
ar, an
d
sorne coups; an d
conspiracy, typically in clu
d in
g plots, m
utinies, an
d m
ost coups. 21
19 H arold L
assw ell and A
braham K
aplan, P ow
er and Society: A F
ram ew
ork for P
olitical lnquiry (N ew
H aven: Y
aie U niversity P
ress, 1950), 261-268. T h
e types are palace revolution, political revolution, an
d social revolution.
2 0
"O n
the E tiology of Interna! W
ars," 135-136. 21 R
um m
el's typology
appears in
"D im
ensions of C
onflict B ehavior," 25-26.
T w
o sum m
ary articles o n
the factor analyses are R udolph J. R
um m
el, "A F
ield
lO
¡ ¡,
E X
P L
A N
A T
IO N
S O
F P
O L
IT IC
A L
V IO
L E
N C
E
T h
ese types are not absolutely distinct. T h
e analyses m en
tio n
ed
on pp. 4-5 indicate that, a ta
m ore g
en erallev
el of analysis, política! violence is a relatively hom
ogenous uní verse. W ithin that u
n í verse,
how ever, sorne kinds of violence ·ten
d to occur together, an
d the
occurrence of sorne types tends to p reclu
d e th
e occurrence of other types. T
h e principal distinction b
etw een
turm oil and revolution
is th e degree of organization an
d focus of violence, a distinction
also m ade by E
ckstein in his com
posite typology. A m
ajor differ- en
ce b etw
een the interna! w
ar an d
conspiracy com ponents of the
revolutionary dim ension is
one of scale. G
eneral definitions of th
e three form s of political violence exam
ined in this analysis are
as follow s:
T urm
oil: R
elatively spontaneous, unorganized politi.cal vio- . lence
w ith
substantial
popular participation,
in- cluding
violent political
strikes, riots,
political clashes, an
d localized rebellions.
C onspiracy:
H ighly
organized política! violence w ith lim
ited participation, including organized political assassi- nations, sm
all-scale terrorism , sm
all-scale guerrilla w
ars, coups d'état, an d
m utinies.
Interna! w ar:
H ighly
organized political
violence w
ith w
ide- sp
read
popular participation,
d esig
n ed
to
over- throw
th e regiine or dissolve the state an
d accom
pa- n
ied by extensive violence, including large-scale
terrorism an
d guerrilla w
ars, civil w ars, an
d revo-
lutions.22
In sum
m ary, this study is an attem
pt to analyze, an d
develop testable general hypotheses about, th
ree aspects of political vio- lence: its sources, m
agnitude, an d
form s. T
h e processes b
y w
hich the potential for violence develops an
d th
e kinds of conditions an
d events that ch
an n
el its outcom e are exam
ined as p art of this
T heory of S
ocial A ction W
ith A pplication to C
onflict W ithin N
ations," Yearbook o
f the Society for G eneral System
s R esearch, X
(1965), 183-204, and T anter, "D
i- m
ensions of C onflict B
ehavior." 2
2 E
ckstein, in "O
n th
e E tiology of Interna! W
ars," reintroduced the term "in-
terna! w ar," b
u t defined it considerably m
ore broadly as "an y
resort to violence w
ithin a political order to change its constitution, rulers, or policies," 133. In or-
dinary language, how ever, th
e phrase connotes a degree of participation an d
or- ganization of conflict that is not characteristic o
f th e events 1 h
a ve separately cate- gorized u
n d
er turm oil an
d conspiracy. T
h e three categories differ from
th e cate-
gories w ith the sam
e labels in G urr, "A
C om
parative S tudy of C
ivil S trife," an
d G
u u
, "A
C ausal M
odel of C ivil S
trife," only in their exclusion of nonviolent strife an d
of violent conflict b
etw een
apolitical groups.
ll
y r •.¡ 1 j 1 r t f' J M li ,:, i!i ll! i!+ )!' ¡,: J'r 1' l
.j 'j
W H
Y M
E N
R E
D E
L
analysis. T w
o tapies often exam ined in theories of revolution are
exam ined h
ere only in passing: the im
m ediate precipitants of vio-
lence, about w hich m
ost generalizations appear trivial; and th e
long-run outcom es of various
kinds of political violence, about
w hich there is
little em
pirical evidence or d etailed
theoretical speculation.
T o
w a rd
a n
ln te
g ra
te d
T h
e o
ry o
f P o
litical V io
len ce
T h
e basic m odel of the conditions leading to political violence
used in this study incorporates both psychological an
d societal
variables. T h
e initial stages of analysis are actor-oriented in th e
sense that m any of the hypotheses about the potential for collective
action are related to, an d
in sorne instan ces d ed
u ced
from , inform
a- tion about the dynam
ics of h u
m an
m otivation. T
h e approach is
not w holly or prim
arily psycholÓ gical, how
ever, an d
it w o
u ld
b e
a m isinterpretation o
f th e argum
ents an d
evidence p resen
ted h
ere to categorize it so. M
ost of the relationships an d
evidence exam ined
in su b
seq u
en t stages of analysis are those that are proposed or
observed to hold b etw
een societal conditions an
d political vio-
lence. T h
e psychological m aterials are used to h
elp provide causal
linkages b etw
een an
d am
ong societal variables an d
the d ep
en d
en t
variables specified above: th e potential for collective an
d political
violence; th e m
agnitude of political violence; an d
th e likelihood
that political violence w ill take th
e forro of turm oil, conspiracy, or
interna! w ar.
U se of psychological ev
id en
ce in this w
ay m akes
certain kinds of social
uniform ities
m ore
clearly apparent an d
com
prehensible, an d
contributes to the sim plification of theory.
A t the sam
e tim e th
e analysis of societal relationships is crucial for identifying the sources of sorne com
m on psychological properties
of violence-prone roen an d
for generalizing about the m any facets
of political violence that have no parallels in psychological dy- nam
ics. T h
e goal of this analysis, at b est only partly realized, w
as proposed b
y Inkeles in
th e context of a discussion of social struc-
ture an d
personality: "W
h at is
req u
ired ... is an integration or
coordination of tw
o basic sets
of data in
a
larger explanatory sch
em e-
not a reduction of eith er m
o de of analysis to th e allegedly
m ore fundam
ental m ode of th
e other." 23 T
h e outlines of th
e theory can now b
e sketched briefly. T h
e pri- m
ary causal seq u
en ce in political violence is first the d
ev elo
p m
en t
of discontent, second the
politicization of that discontent, an
d
23 A lex lnkeles, "P
ersonality and S ocial S
tructure," Sociology T oday (N
ew Y
ork: B
asic B ooks, 1959), 272, quoted in M
arvin. E . W
olfgang and F ranco F
erracuti, The Sub culture o
f V iolence: T
ow ards an Integrated Theory in C
rim inology (L
on- don: S
ocial S cience P
aperbacks, 1967), 8.
12
E X
P L
A N
A T
IO N
S O
F P
O L
IT IC
A L
V IO
L E
N C
E
finally its actualization in violent action against political objects
an d
actors. D iscontent arising from
th e perception of relative depri-
vation is th e basic, instigating condition for participants in
collec- tive violence. T
h e lin
k ed
concepts of discontent an d
deprivation com
prise m ost of th
e psychological states im plicit or explicit in
such theoretical notions about the causes of violence as frustration, alienation,
drive an
d
goal conflicts,
exigency, an
d strain
(dis- cu
ssed in
ch ap
ter 2). R
elative deprivation is defined as a p erceiv
ed discrepancy be-
tw een
m en
's value expectations an d
th eir value capabilities. V
alue expectations are the goods an
d conditions of life to w
hich p eo
p le
b eliev
e th
ey are
rightfully entitled. V
alue capabilities
are th
e goods an
d conditions th
ey think th
ey are capable of attaining or
m aintaining, given th
e social m eans available to them
. S ocietal
conditions that increase the average level or intensity of expecta- tions w
ithout increasing capabilities increase the intensity of dis- content. A
m ong the general conditions th
at have such effects are the value gains of o
th er groups an
d th
e prom ise of n
ew opportuni-
ties (chapter 4). S ocietal conditions th
at decrease m en
's average value position w
ithout decreasing th eir value expectations sim
ilarly increase deprivation, h
en ce the intensity of discontent. T
h e in-
flexibility of value stocks in a society, short-term
deterioration in a
group's conditions of life, an d
lim itations of its structural oppor-
tunities have such effects (chapter 5). D
eprivation-induced discontent
is a
general sp
u r
to action.
P sychological theory an
d group conflict theory b
o th
suggest th at
th e greater th
e intensity of discontent, the m ore likely is violence.
T h
e specificity of this im pulse to action is d
eterm in
ed b
y m
en's beliefs ab
o u
t the sources of deprivation, an d
about th e norm
ative an
d utilitarian justifiability of violent action directed at th
e agents responsible for it.
S ocietal variables th
at affect th e focusing of discontent on politi-
cal objects in clu
d e th
e extent of cultural an d
subcultural sanctions for overt aggression, th
e extent an d
degree of success of past poli ti- cal violence, th
e articulation an d
dissem ination of sym
bolic appeals justifying violence, th
e legitim acy of th
e political system , an
d the
kinds of responses it m akes an
d has m
ade to relative deprivation (chapters 6 and 7). T
h e b
elief that violence has utility in obtaining
scarce values can b
e an in d
ep en
d en
t source of political violence, b
u t w
ith in
political com m
unities it is
m ost likely to provide a
secondary, rationalizing, rather th an
prim ary, m
otivation. W ide-
spread discontent provides a
general ím petus to collective vio-
lence. H ow
ever, th e great m
ajority of acts of collective violence in recen
t decades have h ad
at least sorne political objects, an d
13
:,i ¡:: \:\ ¡.¡ '¡:': ¡, ¡;.; ¡,: i 1' i!'i ¡, !:
W H
Y M
E N
R E
B E
L
th e m
ore in ten
se those violent acts are, th e m
ore likely th ey
are to b
e focused prim arily or exclusively on th
e political system . In-
ten se
discontent is
q u
ite likely to
b e
politicized; th
e prim ary
effect of norm ative an
d utilitarian attitudes tow
ard violence is to focus th
at potential. T
h e m
agnitude of political violence in a system
, an d
th e form
s it
takes, are
partly d
eten n
in ed
by
the scope
an d
intensity of politicized discontent. P
oliticized discontent is a necessary con- dition for th
e resort to violence in politics. B ut h
o w
ev er in
ten se
an d
focused th e ím
petus to violence is, its actualization is strongly influenced b
y the
patterns of coercive control an
d institutional
su p
p m
t in the politicál com
m unity. P
olitical violence is of greatest m
agnitude, an d
m ost likely to take th
e form of interna! w
ar, if regim
es an d
those w lw
oppose tl)em exercise approxim
ately eq u
al degrees of coercive control, an
d co
m m
an d
sim ilar an
d relatively
high degrees of institutional su p
p o
rt in th
e society. T h
e coercive capacities of a re gim
e '1.nd th e uses to w
hich th ey
are p u
t are crucial variables, affecting th
e form s an
d extent of political violence in
b o
th
th e short an
d long run. T
h ere is m
u ch
ev id
en ce, sorne of it sum
m ar-
ized in
chapters 8 an d
10, that sorne patterns of re gim e coercive con-
trol increase rather th an
th
e intensity of discontent, an d
can facilitate th
e transform ation of turm
oil into full-scale revolution- ary m
ovem ents. D
issidents, b y
contrast, u se w
hatever degree of co- ercí ve capacities th
ey acquire principally for group defense an
d for
assaults on th e re gim
e. T h
e degree of institutional su p
p o
rt for dis- sidents an
d for regim
es is a function of th e relative proportions
of a nation's
population th eir organizations m
obilize, th e com
- plexity an
d cohesiveness o
f those organizations, th eir resources,
an d
th e extent to w
h ich
th ey
provide regularized procedures for value
attainm ent,
conflict resolution,
an d
ch
an n
elin g
hostility
(chapter 9). T h
e grow th o
f d issid
en t organization m
ay in th
e short ru
n facilitate política! violence, b
u t it also is likely to provide th
e d
isco n
ten ted
w ith
m an
y of th
e m eans to alleviate deprivation in
the long run, thus m
inim izing violence.
T h
e p reced
in g
th ree paragraphs are an outline of th
e fram ew
ork in
w h
ich the hypotheses an
d definitions of this study are devel-
o p
ed , a
n d
a sum
m ary of sorne of its generalizations. T
h e hypotheses
an d
th eir interrelationships are sum
m arized m
ore fully an d
sys- tem
atically in ch
ap ter 10. T
h e A
ppendix lists all hypotheses de- veloped, categorized according to th
eir d ep
en d
en t variables, an
d
the chapters in w
h ich
th ey
are proposed. T
h e th
ree stages in th
e process o f political v
io len
ce-th o
se in
w h
ich discontent is generated, politicized, an
d actualized in
políti- ca! v
io len
ce-are each d ep
en d
en t on the p
reced in
g one, as th
e 1
4
(, ..
E X
P L
A N
A T
IO N
S O
F P
O L
IT IC
A L
V IO
L E
N C
E
outline indicates.
lt is likely b
u t n
o t necessarily the case that
th ere is a tem
poral relationship am ong the three stages, w
hereby a sharp increase in
th e intensity of discontent precedes the articula-
tion of doctrines th at justify politically violent action, w
ith shifts in
th e balances of coercive control an
d institutional adherence oc-
curing subsequerttly. T h
e conditions can b e sim
ultaneously opera- tive, how
ever, as th e outbreak of th
e V en
d ée counterrevolution in
1793 dem
onstrates: im
plem entation
of procedures
for m
ilitary conscription intensified th
e discontent of w orkers an
d peasants
already sharply hostile to th e bourgeoisie an
d the governm
ent it ruled. M
ass action against th e bourgeoisie b
eg an
in a m atter of
days; the social context for dissident action w as provided in
part by preexisting com
m unal an
d political organization, action that w
as facilitated by the co
n cu
rren t w
eakness of governm ent forces and
institutions in the region. 2
4 T
h e point is that m
any of the attitudes an
d societal conditions that facilitate political violence m
ay be p
resen t an
d relatively unchanging in
a society over a long period; th
ey becom
e relevant to or operative in th
e genesis of violence only w
h en
relative deprivation increases in scope an
d intensity.
Intent;e politicized discontent also can b
e w id
esp read
and per- sistent over a long p
erio d
w ithout overt m
anifestation because a regim
e m
onopolizes coercive
control an
d
institutional support. A
w eakening of regim
e control or th e d
ev elo
p m
en t of dissident
organization in such situations is highly likely to lead
to m assive
violence, as it d id
in H
ungary in 1956 an
d C
h in
a in 1966-68, and
as is likely at sorne future date in S
outh A frica.
T h
e concepts, hypotheses, an d
m odels of causes an
d processes
d ev
elo p
ed in
th e follow
ing chapters are not in ten
d ed
as ends in them
selves. lntellectually pleasing filters through w hich to view
an
d categorize the p
h en
o m
en a of a disorderly w
orld are not know l-
edge. S ystem
atic. know ledge requires us to propose an
d test and
reform ulate an
d
statem ents about h
o w
an d
w hy things hap-
pen. W e know
enough, an d
know it w
ell enough, only w h
en w
e can say w
ith sorne certitude not ju
st w hy things h
ap p
en ed
yesterday, b
u t how
o u
r actions today w ill affect w
h at happens tom
orrow ,
som ething w
e can alw ays h
o p
e to know better, though never per-
fectly. T his analysis m
ay dem onstrate that too little is know
n about the violence m
en do one another, an
d th
at it is know n too w
eakly an
d im
precisely. lt is d esig
n ed
to facilitate th e processes by w
hich that know
ledge can b e increased.
24 C harles
T illy,
The V
endée (C
am bridge:
H arvard
U niversity
Press, 1964),
passim .
15
r o o p lt li al ir 01
p: is el 01 0
(
ti« re is fo
an th " Th, C
n P
et of t Sh€
16
1 f !\_.
), 5
W H
Y M
E N
R E
D E
L
T h
e N atu
re o f S
o cial T
h e o
ry
(1 n this section 1 w
ish to com m
ent on so m e criteria for social
theory and their application to political violence. T hese criteria
are not essential for understanding the substance o f the book, b
u t
they help explain its approach to analysis and form o
f presentation. R
eaders not concerned w ith
these questions m ay turn directly to
chapter 2.) M
any descriptive generalizations an d
propositions can b e m
ade about the origins, processes, an
d effects of political violence gen-
erally, an d
of its severa} form s. V
ariables com m
only exam ined in-
• elude th
e kinds of societal an d
political structures m ost susceptible
to violence, revolutionary ideologies, th e developm
ental seq u
en ce
of revolutionary m ovem
ents, th e effects of force on the outcom
e of violence, the nature of revolutionary organization, an
d m
u ch
else. S
orne less variables for analysis are the nonideological
m otivations of revolutionaries, th
e effects of institutional patterns on the form
s an d
objects o f violence, th
e efficacy of different kinds of governm
ental response, an d
the long-range consequences of
political violence. T
w o paradigm
atic approaches to "th eo
ry " about political vio-
lence can b e identified. O
n e is to analyze a related p
h en
o m
en o
n ,
like revolutionary organization or ideology, in order to generalize about its origins, characteristics, or effects. A
n in terest in
explain- ing the occurrence or outcom
e of political violence is often im plicit
or explicit in such an analysis, b
u t typically th
e analysis does not p
reten d
to b e com
plete; the relevance of other variables ordinarily is recognized. T
h e second is to choose a particular violent ev
en t or
class of events as an object of explanation an d
to specify sorne m ore
or less exhaustive set of conditions or variables that determ ines the
occurrence, extent, or outcom e of th
e events. T heories of revolu-
tion, like those of E d
w ard
s, B rinton, P
ettee, an d
T im
asheff, are representative of this second approach. 25 A
variant of this approach is to select a com
m on, quantifiable property of a variety of events
for analysis an d
explanation (see pp. 8-9). "T
h eo
ry " is a term
loosely u sed
in the social sciences generally
an d
in the study of political violence specifically. T h
e k in
d o
f theory· aim
ed at in this study is an interrelated set of general, falsi-
20 T h
e three best-know n A
m erican theories of revolution are L
yford P. E dw
ards, The N
atural H istory o
f R evolutions (C
hicago: U niversitY
. o f C
hicago P ress, 1927);
C rane B
rinton, T he A
natom y o
f R evolution (N
ew Y
ork: N orton, 1938); an
d G
eorge P
ettee, The P rocess o
f R evolution (N
ew Y
ork: H arper, 1938). A
m ore recen
t theory of this type is proposed by N
icholas S. T im
asheff, W ar and R
evolution (N ew
Y ork:
S h
eed an
d W
ard, 1965).
1 6
E X
P L
A N
A T
IO N
S O
F P
O L
IT IC
A L
V IO
L E
N C
E
fiable hypotheses that specify causal or concom itant relationships
b etw
een
in d
ep en
d en
t an
d d
ep en
d en
t variables.
T h
e nature of such relationships
can b e argued inductively from
observations ab
o u
t th e relevant p
h en
o m
en a at various levels of analysis, or de-
d u
ced from
o
th er
statem ents
com prising th
e theory, or both. 26
S uch "explanatory" theory is distinct from
w h
at E ckstein calls
pretheoretical exercises, su ch
as th e form
ulation of classificatory schem
es and conceptual fram ew
orks, w hich are som
etim es called
theory. 27 N or are descriptive generalizations about relationships
am ong variables
theory in the sense in
ten d
ed here, unless th
e sources an
d nature of such relationships are stated explicitly an
d
th e generalizations them
selves interrelated. M
ost hypotheses d ev
elo p
ed in
this study are d eriv
ed from
th e
juxtaposition an
d
generalization of
relationships observed
in
studies of political violence an d
of th e individual behaviors m
ani- fest in
it. T h
e approach reflects an assum ption that social science
theory ought to b u
ild on w
hat is already know n about th
e subject b
ein g
theorized about, an d
that it ought to b e consistent w
ith, or at least not directly contradictory to, w
hat is m ore generally know
n about th
e nature an d
processes of individual an d
aggregate hum an
behavior. T his is not to say that hypotheses m
u st b
e consistent w ith
w
hat L evy calls "going com
m on sen
se," b u
t rather that it is non- sensical an
d inefficient to invest
effort in testing hypoth-
eses th at contradict w
hat is precisely know n, unless there are com
- p
ellin g
logical or em pirical grounds for questioning w
hat is as- su
m ed
to b e "precisely know
n." T
h e central scientific criterion for theory is that it b
e subject to em
pirical assessm ent. F
o u
r attributes of theory th at facilitate
its assessm ent are its falsifiability, definitional clarity, identifica-
tion of relevant variables at various levels of analysis, an
d ap-
plicability to a universe of events for analysis. T
h e first tw
o are
necessary conditions
for assessm
ent, th
e others
desirable. A
fundam ental lim
itation of m ost older theories and conceptions
26 F o
r a critique o f w
hat passes for theory in th
e social sciences an d
a proposed set o
f criteria for good theory, see M arion J. L
evy, Jr., "'D o
es It M atter if H
e's N
aked?' B aw
led th e C
hild," in K
laus K norr and Jam
es R osenau, eds., C
ontending A
pproaches to
Intem ational
P olitics
(P rinceton:
P rinceton
U niversity
P ress,
1969). F o
r a statem ent o
f criteria for theory in /a field related
to th e subject o
f this book, a
n d
a survey o
f extant "theory," see W olfgang an
d F
erracuti, chap. 2. 27 H
arry E
ckstein, "Introduction:
T ow
ard th
e T
heoretical S
tudy o
f Interna! W
ar," Interna! W ar: P
roblem s and A
pproaches (N ew
Y ork: T
h e F
ree P ress, 1964),
7-29, classifies pretheoretical operations as delim itation (statem
ent of th e bounda-
ries of the subject); classification an d
analysis (m eaning by "analysis" the dissec-
tion o f a subject into its com
ponents); and problem ation (the form
ulation o f specific
problem s for theory construction).
M ost of these "operations" w
ith respect to political violence are attem
pted in this chapter.
17
··t ti ,¡ l¡ Ji '¡ !·; li 11 li,
W H
Y M
E N
R E
U E
L
of revolution is the
difficulty of deriving falsifiable hypotheses
from them
. 28 F
ew if any w
ere form ulated w
ith reference to ap- plicable em
pirical m ethods. T
h e fact that very few
case or com para-
ti ve studies m ake use of th
em is further evidence of th
eir lim ited
usefulness ev en
for taxonom ic or conceptual purposes. T
w o
gen- eral conventions regarding falsifiability are u
sed in
social science. O
ne is to state that an in d
ep en
d en
t (causal) variable is a necessary and/or sufficient condition for the d
ep en
d en
t variable one w ants
to explain. T his usually m
eans that both variables are defined in
dichotom ous
te rm
s- "d
iseq u
ilib riu
m " does
or does n o
t o c c u
r- ev
en if one or both are continuous variables, an
d typically leads
to propositions that art trivial or falsifiable b y
a single deviant case. T
h e second is to rely on tests of statistical significance of relation-
ships b
etw een
variables,
w hich
has led
to th e proliferation of
w eakly su
p p
o rted
b u
t not-quite-fálsified hypotheses w hich have
yet to b e integrated into a m
ore general an d
parsim onious theory.
A third convention is u
sed in this study: a m
ínim um
strength of relationship
is postu1ated
for each
hypothesis in
correlational
term s; if the relationship found in
an em pirical test is w
eaker than stipulated, an
d no serious sam
ple or instrum ent error is likely,
the hypothesis
is rejected.
F o
r exam ple,
one basic hypothesis
is that the greater the intensity an d
scope of relative deprivation in a population, the greater th
e potential for collective violence. T
h e proposed relationship is a strong one: if interval-order m
eas- u
rem en
t of b o
th variables for a substantial n
u m
b er of cases gives
a product-m om
ent correlation coefficient less than .45 (less than 20 p
ercen t of the variance explained), th
e hypothesis is rejected. 29
28 F or exam
ple, E dw
ards, B rinton, and P
ettee. A sam
pling of th e older theoreti-
cal speculation about revolution and political violence, w hich seldom
finds its w
ay even into the footnotes of current w ritings, w
ould include the w ork of G
us- tave L
e B on, especially T
he P sychology o
f R evolutíon, trans. B
ernard M iaU
(L on-
don: U nw
in, 1913); B rooks A
dam , T
he T heory o
f Social R evolutions (N
ew Y
ork: M
acm illan, 1913); E
verett D ean
M artin, T
he B ehavíor o
f C row
ds: A P
sychologícal Study
(N ew
Y ork:
H arper,
1920), chap.
7, "T
h e
P sychology of R
evolutionary C
row ds"; C
harles A .
E llw
ood, T he P
sychology o f H
um an Society: A
n Introduc-
tíon to Socíologícal T heory (N
ew Y
ork: A ppleton, 1925), chap. 8, "C
hanges W ithin
the G roup: A
bnorm al"; R
obert H unter, R
evolution: W hy, H
ow , W
hen? (N ew
Y ork:
H arper, 1940); and M
ark A .
M ay, A
Social P
sychology o f W
ar and P eace (N
ew
H aven:
Y ale
U niversity P
ress, 1943), chap. 7, "A ggressive S
ocial M ovem
ents." T
h e flavor of the literature to 1925 can b
e sam pled in D
ale Y oder, "C
u rren
t D efi-
nitions of R evolution," A
m erican ]ournal o
f Sociology, X
X X
II (N ovem
ber 1926), 433-441. S
orne w ritings of this period are curious indeed, for exam
ple, S idney A
. R
eeve, T he N
atural L aw
s o f Social C
onvulsíons (N ew
Y ork: D
utton, 1933). 29 T
h e
strength of relationship is req u
ired of m
ultiple regression coefficients (R
) for studies in w hich m
ultiple m e asures of a particular in
d ep
en d
en t variable are
used. I have reported evidence to the effect that the .45 threshold is by no m eans
too stringent for the hypotheses m entioned in G
urr, ''A C
ausal M odel of C
ivil S
trife." T h
ree com posite m
easures of deprivation related to m agnitude of civil
strife for 114 polities give an R of .60, despite the presence of considerable m
eas- urem
ent error.
1 8
l.
E X
P L
A N
A T
IO N
S O
F P
O L
IT IC
A L
V IO
L E
N C
E
O th
er hypotheses
specify relationships
of m
oderate strength,
requiring a m
ínim um
r of .30
if th ey
are not to b
e discarded. H
ypotheses about in d
ep en
d en
t variables that do not m eet these
criteria are not necessarily false in th
e conventional sense, b u
t they denote relationships that are too w
eak to con tribute to parsim onious
theory. 30 In
d ep
en d
en t an
d d
ep en
d en
t variables m
ust b e defined w
ith sufficient precision so th
at a researcher can determ ine w
hat par- ticular conditions or set of events constitute "X
" for purposes of em
pirical assessm ent. T
h e definitional inadequacy of m
any the- ories, n
ew an
d old, contributes to the difficulty of assessing them
. F
o r exam
ple, a great m any conditions have b
een said to "cau
se" or to constitute a basic potential for collective violence, am
orig th
em general properties of sociopolitical system
s lab eled
"cram p,"
"d iseq
u ilib
riu m
," an
d
"strain"; m
otivational characteristics
of violent m
en su
ch as "frustration" an
d "discontent"; an
d particu-
lar institutional patterns su ch
as repressive governm ent an
d inade-
quate circulation of, or divisions w ithin·: an elite. 31 A
difficulty of explanations citing such general properties is that th
e conditions as
defined, if in d
eed th
ey are defined, can
usually b e found in
m ost societies an
d am
ong m ost m
en, w h
eth er or not revolutionary;
an d
usually th ey
are not form ulated precisely en
o u
g h
to p erm
it one to assess th
e effects of th eir various elem
ents an d
degrees. T
h e em
pirical referents of sorne concepts are m ore easily isolated
than those of others. T
h e co
n cep
t "participation in clandestine
organization" is m ore readily m
ade operational than "revolution- ary m
obilization," an d
revolutionary m obilization is a m
ore m al-
leab le concept th
an "p
o w
er deflation." T his is not to argue that
m ore general concepts b
e avoided, b u
t that the m ore general they
are th e m
ore necessary it is to define th em
fully an d
to catalog th eir
m anifestations so that th
ey can b
e evaluated em pirically.
It is desirable that theoretical statem ents b
e testable at various levels of generalization. F
eld m
an , for exam
ple, attributes revolu- tionary potential to the increasing salience o
f goal conflicts b etw
een
th e increasingly nurnerous subsystem
s that are said to result from
30 T h
ese tw o criteria im
ply an absolute m axim
um of five
strong variables or eleven m
oderately strong variables for any theoretical system . H
ypotheses that ·are not supported at the .30 leve! b
u t that nonetheless m
eet statistical tests of significance
should lead to postulation of m
ore general hypotheses that m
ight account for a set of such w
eak relationships. 31 S
ee P ettee, passim
; C halm
ers Johnson, R evolutionary C
hange (B oston: L
ittle, B
row n, 1966), especially chap. 4; N
eil J. S m
elser, T heory o
f C ollective B
ehavíor (N
ew Y
ork: T h
e F ree P
ress, 1963), passim ; D
avies, passim ; and R
onald C . R
idker, "D
iscontent and E conorhic G
row th," E
conom ic D
evelopm ent and C
ultural C hange,
X I (O
ctober 1962), 1-15. A categorization of specific and general causal factors cited
in th e literature is given in E
ckstein, "O n
the E tiology of Interna! W
ars," 143- 144.
19 ,.
' l . W
H Y
M E
N R
E B
E L
structural differentiation in a society. 32 Id eally
it should b e pos-
sible to observe m anifestations of such a process at b
o th
th e m
acro- level of an en
tire social system , for exam
ple eighteenth-century F
ran ce, and at th
e level of th e com
m unity an
d am
ong sm all, face-
to-face groups. 33 S m
all-group p h
en o
m en
a are not likely to resem b
le sim
ply a scaled-dow n m
odel of m acrophenom
ena, b u
t it should b
e possible to identify th e sm
all-group interactions an d
individual behaviors
that are com prised b
y m
acroevents. If this translata-
bility cannot b e ach
iev ed
directly, th en
at least the m acrophe-
nom ena postulated in
theory sh o
u ld
b e com
patible w ith w
hat is know
n about m icrophenom
ena. T
his is not a reductionist argu- m
en t that analysis of social system
s or collective behavior can or should be red
u ced
to analysis of com ponent individual behaviors.
T h
e point is that analysis on one level can an d
should inform th
e other,
an d
that hypotheses
w hose relationships
are m anifested
at, and subject to exam ination at, different levels of analysis are
usually m ore interesting an
d fruitful
th an
those w hich refer to
one level o f analysis only. 34
T h
e fourth attribute o f theory that facilitates its assessm
ent con- cerns th
e types an d
num bers of cases or settings in
w hich th
e pro- posed relationships can b
e exam ined. T
h e preference is for theory
that can b
e subject to test both in case studies an
d in
large-sam ple
com parative
studies using statistical m ethods.
C ase studies
are useful for elucidating the fine structure o
f revolutionary events and for providing a sen
se of understanding of h o
w general vari-
ables act an d
interact. T h
ey also can b
e u sed
to test theoretical statem
ents expressed in dichotom
ous term s: o
n e can define "ac-
celerators of dysfunction"
an d
"revolution" in d
ep en
d en
tly an
d
in sufficiently rigorous fashion that a single case stu
d y
m ay b
e suf- ficent to falsify the proposition that acceb
rato rs o
f dysfunction are a
necessary precondition of revolution. 35 B ut m
any variables of interest to social theory can b
e dichotom ized only at great loss o
f inform
ation, and scholars undertaking case studies often find it difficult
to distinguish
b etw
een
general relationships
and th
e 32 A
rnold F eldm
an, "V iolence and V
olatility: T
h e L
ikelihood of R evolution,"
in Internal W ar, 111-129.
33 S idney Y
erba, Sm all G
roups and P olitical B
ehavior: A Study o
f L eadership
(P rinceton: P
rinceton U niversity P
ress, 1961), dem onstrates the usefulness of re-
lating concepts and flndings at one leve! of analysis to those at another. T illy's
case study of an interna! w ar, T
he V endée, show
s how analysis can proceed from
m
acrolevei concepts
such as
urbanization and social change to the individual
leve!, and back again. 34 O
n the relative m erits of using dilferent levels of analysis in social inquiry
see D avid S
inger, "T h
e L eve! of A
nalysis P roblem
in International R elations,"
W orld P
olitics, X IV
(O ctober 1961), 77-92.
•• T his proposition is adapted from
Johnson, op. cit.
2 0
1 ¡ 1
E X
PL A
N A
T IO
N S O
F P
O L
IT IC
A L
V IO
L E
N C
E
u n
iq u
e historical
an d
cultural circum stances of each case.
H y-
potheses that
specify system
atic relations
am ong
continuous variables, or b
etw een
a continuous an d
a dichotom ous variable,
are subject to scientific assessm ent only if substantial num
bers of cases
are exam
ined. T
his provides
further justification for the
choice of political violence rather than revolution as the subject of this study: th
e form er is far m
ore com m
on than the latter. In
this
study hypothetical
relationships are
form ally
stated and given alphanum
eric descriptors. T h
e first term in
each hypoth- esis is its d
ep en
d en
t variable. D efinitions ordinarily are proposed
in th e paragraphs im
m ediately follow
ing th e first appearance of
a theoreticai term in
a hypothesis. U n
d erlin
in g
of a term in
this context indicates that th
e sen ten
ce in w hich it appears is a form
al definition. S
ynonym s are u
sed for severa! of the frequently cited
theoretical concepts to avoid th
e m in
d -d
ead en
in g
effects of repe- tition. "R
elative deprivation," defined as a p erceiv
ed discrepancy
b etw
een m
en 's value expectations an
d value capabilities, is rep-
resen ted
by its initials, R D
, an d
th e synonym
s deprivation, dis- crepancy, and, loosely, frustration. "D
isco n
ten t," the psychologi-
cal state said to b e cau
sed by R
D , has as its synonym
s anger, rage, an
d dissatisfaction. S
ynonym s for "v
alu e expectations" are w
ants, aspirations, an
d expectations. "V
alues" are the goods an d
condi- tions o
f life w hich m
en seek; th
e phrase "goods an d
conditions of life" is
u sed
synonym ously w
ith values. "V alue opportunities,"
th e courses of action p
eo p
le have available to th em
for attaining or m
aintaining th eir d
esired values, are m
ore sim ply referred to as
"m ean
s." T h
ese synonym s are u
sed only w
h en
there is no am -
biguity about th eir referent; th
e defined term
s are u sed
in the d
ev elo
p m
en t of th
e basic theoretical argum ents.
21
2 . R
ela tiv
e D ep
riv a
tio n
a n
d th
e Im
p etu
s to V
io len
ce Our desires an
d pleasures
spring from
society; w
e m easure them
, therefore, b y
society
an d
not
b y
th
e objects
w hich
serve for th eir satisfaction. B
ecause they are of a social nature, they are of a rela- tive nature.
·
K arl M
arx an d
F riedrich E
ngels, W
age L abor and C
apital
B E
N E
A T
H th
e com plexity of h
u m
an m
otivation neurophysiologists have identified
great "ap p
etitiv e system
s" th at provide th
e m
otivating feelings against everything that h
ap p
en s to us
is m easu
red an
d judged. S
tim ulation of o
n e of these system
s pro- vides our feelings
of elation, satisfaction, an d
love. S tim
ulation of th
e other leads to sensations of anxiety, terror, depression, an d
rage. T
h ese feelings color o
u r perceptions of th
e w orld an
d ener-
gize our actions. L earning is b
ased on th
ese appetitive system s,.
first directly, th en
indirectly: w e learn to do an
d to seek
o u
t those things that b
rin g
satisfaction, an d
to avoid those that have noxious effects. 1
M eri.'s
circum stances
change, how
ever, an
d
w hat th
ey
have learn
ed does not alw
ays prove suitable for deriving satisfactions from
ch
an g
ed
circum stances.
"W e
b eco
m e frustrated,"
C antril
w rites, "w
h en
w e sen
se a conflict b etw
een th
e significances w e
bring to a situation an d
w hich have w
orked in th
e past b u
t seem to
have no correspondence . . . to th
e em erging situation w
e face. •
• : · 2 T
his conflict or tension is fundam entally unpleasant: to b
e avoided or overcom
e if possible; to b e released in
expressive, "nonrealistic" w
ays if not. lt is th e fundam
ental source of b o
th
innovation and destruction in h
u m
an affairs. W
hy inriovative be- havior should occur in
response to tension is clear enough: th e
socialization process teaches
m en
to learn to avoid u n
p leasan
t stim
uli, and only severe n ew
conflicts are likely to paralyze the adaptive capacities m
en acquire in
that process. D estructive be-
havior m ay
b e
explained by reference
to another fundam
ental property of the h
u m
an organism
: if m en
are exposed to noxious
1 A b
rief introduction to the neurophysiological literature, an d
an interpretation of its im
plications for m otivation generally, are provided by H
adley C antril, .. Sentía,
ergo sum : 'M
otivation' R econsidered," ]ournal o
f P sychology, L
X V
(January 1967), 91-107. T
h e appetitive system
s w ere located by neurophysiologists in
th e m
id- 1950s an
d have b
een identified an
d studied in both m
an and other m
am m
als. 2 C
antril, p. 99.
22
R E
L A
T IV
E D
E P
R IV
A T
IO N
A N
D T
H E
IM P
E T
U S
T O
V IO
L E
N C
E
stim uli that th
ey cannot avoid or overcom
e, they have an innate disposition
to out at th
eir sources. S
triking out m ay or
m ay n
o t reduce the frustration, b
u t it seem
s to b e an inherently
satisfying response to th e tension b
u ilt u
p through frustration.
3
T h
e desire to release tension is not th e only source of aggression,
how ever.
Innovative responses
to tensions m
ay them selves in-
elude th e resort to violence. M
ost im portant, th
e choice of tactical or "realistic" violence as an innovative response to tension is rein- forced by the innate disposition to aggression created
by the ten- sion.
D istinctions
b etw
een
"realistic" an
d
"nonrealistic" con-
flict or aggression thus m ay b
e analytically useful, b u
t the physio- logical an
d psychological ev
id en
ce suggests that elem ents of t#
e latter are alm
ost alw ays present. 4 It is likely to b
e absent only am
ong those w ho are coerced into participation in collective con-
flict. T
h ese principies operate in a w
ide range of individual behavior, including th
e actions of those in rebellion against th
eir political com
m unity. W
e n
eed concepts an
d hypotheses b
etter su ited
to . analyzing th
e social an d
psychological transactions th at provide
the ím petus to political violence am
ong m em
bers of a collectivity. "R
elative deprivation" (R D
) is th
e term u
sed in
the p reced
in g
chapter to denote the tension that develops from
a discrepancy b
etw een
the "o u
g h
t" an d
th e "is" of collective value satisfaction,
and that disposes m en
to violence. T h
e term 's definition is dis-
tinct from its conventional sociological usage, b
u t not so difterent
as to w arrant using a neo!ogism
like "cram p
" or "exigency." T his
chapter exam ines th
e R D
concept an d
its subordinate concepts: values, value classes, value expectations, value capabilities, an
d
value opportunities. T h
e frustration-aggression relationship pro- vides
th e
psychological dynam
ic for the proposed relationship
b etw
een intensity of deprivation an
d the potential for collective
violence; consequently it is exam ined in
sorne detail. O th
er con- ceptual
interpretations of th
e ím petus to political violence are
3 T h
e drive properties of frustration-induced aggression are exam ined an
d docu-
m ented by N
orm an R
. F . M
aier, F rustration: T
he S tu
d y o
f B ehavior W
ithout a G oal
(N ew
Y ork: M
cG raw
-H ill, 1949), passim
, and L eonard B
erkow itz, .. T
h e C
oncept of A
ggressive D rive: S
orne A dditional C
onsiderations," in B
erkow itz, ed.,A
dvances in E
xperim ental
P sychology, V
ol. u
(N ew
Y ork:
A cadem
ic P ress, 1965), 307-322,
am ong others. L
ike th e appetitive system
s, it appears to b e characteristic of m
an and of higher-order anim
als general! y. 4 C
onsiderable significance has b een
attached to the distinction b etw
een realistic
and nonrealistic conH ict b
y L
ew is C
oser, T he F
unctions o f Social C
onfiict (N ew
Y
ork: T h
e F ree P
ress, 1956), 4 8
-5 5
an d
passim . T
h e distinction is com
m only m
ade in conH
ict theory, as pointed o u
t by R aym
ond W . M
ack an d
R ichard C
. S nyder,
..T he A
nalysis of S ocial C
onH ict: T
ow ard an O
verview an
d S
ynthesis," ]ournal o f
¡;. C
onfiict R esolution, 1 (June 1957), 221-248.
( ! 23
fi, j
!!!'
' ' ) (,e W
H Y
M E
N R
E B
E L
related to th
e relative deprivation m
odel, including notions of dissonance,
anom ie,
an d
social
conflict. F
inally th ree
patterns of disequilibrium
b etw
een value expectations an
d value capabili-
ties are proposed to facilitate dynam ic analysis.
R elativ
e D ep
riv atio
n D
efin ed
H ypothesis
V .l:
T h
e potential
for collective
violence varíes
strongly w ith" th
e intensity an d
scope of relative deprivation am
ong m em
bers of a collectivity.
R elative deprivation
(R D
) is defined as
actors' p ercep
tio n
o f
discrepancy b etw
een th
eir value expectations an d
th eir value capa-
bilitjes. V alue expectations are th
e goods and conditions of life to w
hich p eo
p le b
eliev e th
ey are rightfully entitled. V
alue capa- bilities are th
e goods an d
conditions th ey
think th ey
are capable of getting an
d keeping. (T
hese concepts are m ore precise! y defined
below .) T
h e em
phasis o f th
e hypothesis is on th e perception o
f deprivation;
m ay b
e subjectively d ep
riv ed
w ith reference
to th
eir expectations ev
en though an objective observer m
ight not ju
d g
e th em
to b e in
w ant. S
im ilarly, th
e existence o f w
h at
the observer ju d
g es to b
e abject poverty or "absolute deprivation" is not necessarily th
o u
g h
t to b e unjust or irrem
ediable by those w
ho experience it. A s R
uncim an puts it, "if p
eo p
le h a ve no reason
to expect or h o
p e for m
ore than th ey
can achieve, th ey
w ill b
e less discontented w
ith w
h at th
ey have, or ·even grateful sim
ply to b e
able to hold on to it." 5 T h
e concept of R D
w as first u
sed system
ati- cally in
th e 1940s b
y th
e authors of T he A
m erican Soldier to denote
th e feelings of an
individual w ho lacks sorne status or conditions
that h e thinks h
e sh o
u ld
have, his standards of w h
at h e sh
o u
ld
have generally b ein
g d
eterm in
ed by reference to w
hat sorne o th
er person or groups has. 6 T
h e co
n cep
t is w idely u
sed in
sociological research, w
h ere it is usually assu
m ed
for operational purposes th at
value are set b
y reference to sorne group or status
w hich an individual does or is thought to identify_7 It is m
ore gen-
5 W . G
. R uncim
an, R elative D
eprivation and Social ]ustice (B erkeley: U
niversity of C
alifornia P ress, 1966), 9.
6 S ee note 27, chap. l.
7 S ee for instance R
uncim an, ll ff; D
avid F . A
berle, "A N
ote on R elative D
epriva- tion T
heory," in S ylvia L. T
hrupp, ed., M illennial D
ream s in A
ction: E ssays in
C om
parative Study (T he H
ague: M outon, 1962),
G ordon R
ose, "A nom
ie and D
eviation: A C
onceptual F ram
ew ork for E
m pirical S
tudies," B ritish ]ournal
o f Sociology, X
V II (M
arch 1966), P
eter T ow
nsend, "T h
e M eanings of Pov-
erty," B ritish ]ournal o
f Sociology, X III (S
eptem ber 1962),
and the status- inconsistency literature beginning w
ith G erhard L
enski, "S tatus C
rystallization: A
N on-V
ertical D im
ension of S ocial S
tatus," A m
erican Sociological R eview
, X IX
(A
ugust 1954),
2 4
p r
R E
L A
T IV
E D
E P
R IV
A T
IO N
A N
D T
H E
IM P
E T
U S
T O
V IO
L E
N C
E
erally recognized, how ever, th
at value standards can have other sources. A
n individual's point of reference m ay b
e his ow n past
condition, an abstract ideal, or the. standards articulated by a lead er
as w
ell as a "referen ce group." T
h e definition u
sed h
ere m akes
no assum
ptions about
th e
sources of value
expectations; it is
sim ilar to A
berle's definition of R D
as "a negative discrepancy
b etw
een legitim
ate expectations an d
actuality." 8 V
alues are th e d
esired events, objects, an
d conditions for w
hich m
en strive. 9 T h
e values m ost relevant to a theory of política! vio-
lence are th e general categories of conditions v
alu ed
by m any m
en, not those idiosyncratically sought b
y particular individuals. In
psy- chological term
s, values are th e goal objects of h
u m
an m
otivation, presum
ably attributable to or d eriv
ed from
basic "n eed
s" or "in- stincts." T
h ere have b
een in
n u
m erab
le attem pts to identify and
categorize "n
eed s,"
"goals," or
"v alu
es" for
purposes of psy-
chological, sociological, an d
political analysis. F reu
d postulated a
single basic n eed
, E ros; H
en ry
M urray listed 12 "viscerogenic"
and 28 "p sy
ch o
g en
ic" needs. 10 T
h ree influential an
d reasonably
parsim onious lists are sum
m arized in table l
an d
related to one another. A
three-fold categorization th at includes w
elfare values, pow
er values, an d
interpersonal values is u sed
here. T h
ere is no n
eed for originality in
su ch
a schem e; it is a com
posite typology, representing
values com
m on to o
th er schem
es an d
relevant to the genesis o
f collective R D
. W
elfare values
are those
th at contribute directly to physical
w ell-being an
d self-realization. T
h ey
in clu
d e th
e physical goods of life-fo
o d
, shelter, h
ealth
services, an d
physical
co m
fo rts-
and th e d
ev elo
p m
en t an
d u
se of physical an d
m ental abilities.
T hese
tw o
classes of w
elfare values
are referred to
b elo
w as
econom ic an
d self-actualization values. S
elf-actualization values m
ay b e instrum
ental to th e attainm
ent of o th
er w elfare values an
d
vice versa. A side from
this, how ever, M
aslow an
d D
avies have argued persuasively th
at "self-actualization" is an en
d in
itself for m
an y
m en: w
e take intrinsic satisfaction in exercising o
u r in-
tellects an d
o u
r h an
d sY
P ow
er values are those th at d
eterm in
e the extent to w
hich inen can influence the actions o f others and
avoid u n
w an
ted interference b
y others in
th eir ow
n actions. P o
w er
values especially salient for política! violence include th e desire
8 A berle, 209.
9 F ollow
ing the usage of H arold L
assw ell and A
braham K
a plan, P ow
er and So- ciety: A
F ram
ew orkfor P
olitical Inquiry (N ew
H aven: Y
ale U niversity P
ress, 1950), 10 S
um m
arily listed and discussed in Jam es C
. D avies, H
um an N
ature in P olitics
(N ew
Y ork: W
iley, 1963). 11 M
aslow , passim
; D avies,
2 5
1 ' H
1! •:¡ ::¡ _111
i ' :q .¡. :. i' r !•111: 1''1 !f ,, ,,¡1 ¡ Hl 1 '8 iil:. .. ¡··¡···¡· ;: ,.,,. i ' ·, ';¡ ;"j '1 ' ,¡¡; : .. :11' :¡, .• :¡·fl f 1 ; ... .. t.l·· 1¡f' il J! 'ra !if:',¡ ik 1:
1 .. .
'\1 '
1
1,!!
W H
Y M
E N
H E
H E
L
M aslow
's N eed
H ierarchy"
P hysical
S elf-actual- ization
S afety, order
L ove, belong- ingness
S elf-esteem
..
T A
B L
E 1
F o
u r L
ists of V alue C
ategories
L assw
ell and K
aplan's V
aluesb
R uncim
an's D
im ensions o
f Social
I nequality e
W E
L F
A R
E V
A L
U E
S
W ell-being,
E conom
ic class w
ealth S
kill, enlight- en
m en
t
D E
F E
R E
N C
E
V A
L U
E S
P o
w er
P ow
er A
tfection -
R espect
S tatus
R ectitude
-
C om
posite T
ypology
} voluo'
P o
w er values
l ln
terp erso
n al
values
• A .
H .
M aslow
, "A T
heory of H u
m an
M otivation," P
sychological R eview
, L
(1943), 370-39& , sum
m arized and discussed in
Jam es C
. D avies, H
um an N
ature in P
olitics: The. D ynam
ics o f P
olitical B ehavior (N
ew Y
ork: W iley, 1963), 8-63.
M aslow
postulates a hierarchy am ong needs: safety an
d order needs w
ill not em erge
until physical .needs are satisfied, !ove needs em erge only after safety n
eed s are
satisfied, etc. T h
e needs are listed h ere in
M aslow
's proposed order w ith th
e ex- ception of self-actualization,
w hich h
e suggests
em erges
after love n
eed s
are satisfied.
b P ow
er and Society, 55-56. e R
elative D eprivation,
chap. 3.
R uncim
an does
n o
t treat these explicitly as values or needs. b
u t as conditions that groups have in varying degrees, an
d w
ith respect to w
hich p eo
p le ju
d g
e their relative satisfaction or deprivation.
to in
collective d
ecisio n
-m ak
in g
-to
vote, to
take part in
political com petition, to b
eco m
e a m em
b er of th
e political e lite
- and th
e related desires for self-determ ination an
d security,
for exam ple freedom
from oppressive political regulation or from
disorder. T
h ese tw
o classes of p o
w er values are referred to b
elo w
as participation an
d security values. Interpersonal values are the
psychological satisfactions w
e seek
in nonauthoritative interac-
tion w ith other individuals an
d groups. T
h ese values in
clu d
e th e
desire for status, i.e., occupancy of a generally recognized role b y
virtue
of w hich
w e
are granted sorne
m easure
of prestige b
y
those w ith w
hom w
e interact; the related n eed
to participate in
stable, supportive
g ro
u p
s-fam ily,
com m
unity, asso
ciatio n
s- th
at provide com panionship an
d affection; an
d th
e sense of cer- tainty
that derives from
sh ared
adherence to beh.efs about th e
nature of society an d
one' s place in it, an d
to norm s governing
social íhteraction.
T h
ese th
ree classes
of interpersonal values
are lab eled
status, com m
unality, an d
ideational coherence.
2 6
;. b. t:: ,. [ i
R E
L A
T IV
E D
E P
R IV
A T
IO N
A N
D T
H E
IM P
E T
U S
T O
V IO
L E
N C
E
T h
e value expectatíons of a collectivity are the average value positions to
w hich its
m em
bers b eliev
e th ey
are justifiably en- titled. V
alue position is th e am
ount or level of a val u e actuall y attained. V
alue expectations refer to both p resen
t an d
future con- ditions. M
en ordinarily expect to keep w
h at they have; th
ey also
generally have a set of expectations an d
d em
an d
s about w hat they
should have in th
e future, w hich is usually as m
uch or m ore than
w hat they have at present. It is im
portant to note that value ex- pectations are defined w
ith reference tojustifiable value positions, m
eaning w h
at m en
b eliev
e th ey
are en titled
to g et or m
aintain, not m
erely w h
at they faintly h o
p e to attain. H
oselitz an d
W illner
m ake
a precisely
com parable
distinction b
etw een
expectation
and aspiration:
E xpectations are a rnanifestation of th
e prevailing norm s set b
y
th e
irnm ediate
social an
d cultural environm
ent. W h
eth er ex-
p ressed
in econom
ic or social term s, the basis upon w
hich th e
individual forros his expectations is th e sense of w
hat is right- fully o
w ed
to him . T
h e source o
f that sen se of rightness m ay b
e w
hat his ancestors have enjoyed, w hat h
e has h ad
in the past,
w hat tradition ascribes to him
, an d
his position in relation to
that of others in th
e society. A spirations, on th
e other hand, rep
resen t th
at w hich h
e w o
u ld
like to have b u
t has not neces- sarily h
ad or co
n sid
ered his d
u e .
. . P
T he value capabilíties of a collectivity are the average value
positions its
m em
bers perceive them
selves capable of attaining or m
aintaining. V
alue capabilities
also have
both p
resen t an
d
future connotations.
In the present, value capabilities are
rep- resen
ted b
y w
hat m en
have actually b een
able to attain or have b
een p
ro v
id ed
by th eir environm
ent: th eir value position. In
the future, value capabilities are w
h at m
en b
eliev e th
eir skills, th eir
fellow s, an
d th
eir rulers w ill, in th
e course o f tim
e, p erm
it th em
to keep or attain: th
eir value potential. It is possible to distinguish b
etw een
p erceiv
ed an
d actual value potential: m
en 's capacities
for attaining th eir value expectations m
ay b e substantially greater
or less th an
they believe th em
to be. H ow
ever, it is perceived value potentiai that determ
ines p resen
t behavior. It is also likely that
p erceiv
ed
value potential
is considerably
m ore
im portant
than p
resen t value
position in determ ining how
p eo
p le assess
their capabilities. T h
e attained value positions of a group m ay
1' B ert H
oselitz an d
A nn W
illner, "E conom
ic D evelopm
ent, P olítica! S
trategies, and A
m erican A
id," in M orton A
. K
aplan, ed., T he R
evolution in W orld P
olitics (N
ew Y
ork: W iley, 1962), 363.
2 7
L . '
Y 1
-
' ' ¡: ' l ¡ W
H Y
M E
N R
E B
E L
b e quite low
w
ith respect to value expectations, b u
t p erceiv
ed
deprivation an d
m anifestations of discontent w
ill ten d
to b e low
to the extent that potential is perceived to b
e high. T h
e obverse relationship
characterizes sorne
prerevolutionary societies:
at- tain
ed value positions appear relatively high w
ith respect to value expectations, b
u t th
e potential for increasing or ev en
m aintaining
value positions is p erceiv
ed to b
e declining. T h
ese assertions are d
o cu
m en
ted in
th e follow
ing chapters. T
h e courses of action p
eo p
le have available to th em
for attain- ing or m
aintaining th eir d
esired value positions are th
eir value opportuníties, th
ree types o f w
hich can b e distinguished: personal,
societal, an d
política!. P ersonal opportunitíes are individuals' in-
h erited
an d
acquired capacities for value-enhancing action.
In- h
erited capacities are norm
ally distributed in m
ost collectivities an
d thus have little relevance to a theory of collective violence.
T h
e technical
skills an
d
general know
ledge acquired
through education, how
ever, can greatly increase m en
's sense of personal com
petence, particularly in im
proving th eir m
aterial value posi- tions.
Societal opportunitíes
are th
e norm
al courses
of action available to m
em bers of a collectivity for direct value-enhancing
action. S
ocietal opportunities for econom ic value attainm
ent in- elude th
e range an d
n u
m b
er of rem unerative occupations, th
e ease of access to those occupations, an
d th
e econom ic resources avail-
able to com pensate those engaged. P
articipation values can b e
attained through
routinized channels for
political participation
an d
recruitm ent to th
e political elite; the attainm ent of security
values is largely a function of th e capacity of th
e political system
for sim
ultaneously m
inim izing
detailed regulation
of h
u m
an
activity an d
m aintaining interna! arder. Interpersonal values are
en h
an ced
to th e extent that fam
ilia! an
d com
m unal life is free
from externa! disruption, an
d to th
e extent that there are generally accepted norm
s on the basis of w hich status an
d resp
ect are ac- corded in
interpersonal relations. P olítical opportunítíes are th
e norm
al courses of action available to m em
bers of a collectivity for
inducing others
to provide
th em
w
ith value
satisfactions. P
olitical opportunities refer to political actions as m
eans rather than ends; opportunities for political participation as an en
d in
itself are com
prised u n
d er societal value opportunities. T
h e sam
e procedures
an d
institutions th at provide. th
e latter usually also provide th
e m eans
by w h
ich collectivities can d
em an
d w
elfare an
d p
o w
er benefits from a governm
ent. rh e re
are o th
er kinds of opportunities
that are
"political" in
th
e sen
se inti:m
ded here,
including collective bargaining procedures by w hich w
orkers can d
em an
d greater w
elfare benefits from
th eir em
ployers, an d
as-
2 8
t 1 · .. :
R E
L A
T IV
E D
E P
R IV
A T
IO N
A N
D T
H E
IM P
E T
U S
T O
V IO
L E
N C
E
sociational activity
by subcultural
groups designed to increase
their m em
b ers' status in dealing w
ith m em
bers of other groups. T
h e scope of R
D is its prevalence w
ith respect to each class of values am
ong th e m
em bers o
f a ·collectivity. S orne deprivations
are characteristic o f sorne m
em bers of all groups. D
eprivation is relevant
to th
e disposition to
collective violence to the extent that m
any p eo
p le feel discontented about the sam
e things. U nex-
p ected
personal deprivations such as failure to obtain an expected prom
otion or th e infidelity of a spouse ordinarily affect few
people at any given tim
e an d
are therefore narrow in
scope. E vents an
d
patterns of conditions
like th e
suppression of a political party,
a drastic inflation, or th e decline of a group's status relative to its
reference groüp are likely to precipitate feelings of R
D am
o.ng w
hole groups or categories of p eo
p le an
d are w
ide in scope. A
berle dichotom
izes w hat is h
ere caB ed scope into tw
o general classes of deprivations, those th
at are personal an d
those that are group ex- periences.13 S
cope is b etter regarded as a continuum
: it should b e
possible to identify, for exam ple by survey techniques, th
e propor- tion of p
eo p
le in any collectivity that feels d
ep riv
ed w
ith respect to any specified class of
. T
h e intensity of R
D is th
e extent o f negative affect that is asso-
ciated w ith its perception, or in
o th
er w ords th
e sharpness of dis- content or anger to w
hich it gives rise. R uncim
an sim ilarly speaks
of the "d eg
ree" of deprivation, defined as "th e intensity w
ith w hich
it is felt." 14
Intensity, like scope, is subject to direct em
pirical assessm
ent: one can infer th e intensity of m
en 's feelings about R
D
using interview
, projective,
an d
content
analytic techniques,
am ong others. 15
M oreover it is possible to specify a n
u m
b er of
properties of value expectations an d
value capabilities that increase or decrease the scope an
d intensity o
f deprivation, an d
th at can b
e exam
ined w ithout necessarily relying on survey techniques. S
orne determ
inants of th e scope an
d intensity of R
D are exam
ined in the follow
ing chapter. P
otential for collective violence, th e d
ep en
d en
t variable of the hypothesis stated at th
e outset of this section, is defined as th e
scope an d
intensity of th e disposition am
ong m em
bers of a col- lectivity to take violent action against others. F
o r m
any research ·purposes this potential m
ay b e treated
as a hypothetical construct, a disposition to act inferred to exist in
the m inds o
f m any m
em bers
13 A berle, 210.
14 R uncim
an, 10. 15 O
n e appropriate interview
ing technique, the self-anchoring scale, is u sed
in H
adley C antril, T
he P attern o
fH u
m a
n C
oncerns (N ew
B runsw
ick: R utgers U
niver- sity P
ress, 1965).
2 9
•
W H
Y M
E N
R E
D E
L
of a collectivity b u
t m easu
red only in term
s of its antecedents, the intensity an
d scope of R
D , or in term
s of its consequences, th
e m agnitude of collective violence. If it could not conceivably
b e assessed m
ore directly there w o
u ld
b e no p
o in
t in stating hy-
potheses about it; th e only testable hypothesis w
o u
ld b
e that th e
greater th e intensity an
d scope of relative deprivation, th
e greater the m
agnitude of collective violence. In principie, how
ever, th e
potential for collective violence can b e in
d ep
en d
en tly
assessed. O
n e m
eans is
th e
u se of interview
techniques that specifically ask p
eo p
le w h
eth er th
ey are p
rep ared
to participate in a riot, or
that allow th
em to project violent sentim
ents in response to am
- bivalent
stim uli.
T h
ese tech
n iq
u es
can b e
u sed
in
structured or laboratory situations. T
h ey
also can b e an
d have b
een em
p lo
y ed
in
natural
populations. L
ouis H
arris, for
exam ple,
has p
o lled
black A
m ericans ab
o u
t th eir
to riot. 16 lt also is pos- sible
to construct sim
ulation studies
of prerevolutionary situa-
tions an d
assess th e responses of players, an approach b
ein g
de- v
elo p
ed
by S
chw artz.
17 T
his diversity
of approaches seem
s to
justify treating potential for collective violence as a crucial in-
tervening variable b etw
een deprivation-induced discontent an
d
political violence, rather than as a m erely hypothetical an
d su-
perfl.uous construct.
T h
e S
o u
rces o f A
g g
ressio u
18
P sychological theories about the origins of h
u m
an aggression
provide an
explicit m
otivational explanation
for the
p ro
p o
sed
causal link b etw
een relative deprivation an
d collective violence.
T h
ere is a variety of theoretical w ritings on this question, sorne
of it speculative, sorne of it b ased
on em pirical research. S
orne psychological "th
eo ries" ab
o u
t th e sources of aggressive behavior
can b e disregarded at th
e outset. T h
ere is little support for pseudo- psychological assertions
th at m
ost or all revolutionaries or con- spirators are deviants, fools, or th
e m aladjusted.
19 P sychodynam
ic explanations of the "revolutionary personality" m
ay b e useful for
16 W illiam
B rink .and L
ouis H
arris, B lack and W
hite: A Study o
f U .S.
R acial
A ttitudes T
oday (N ew
Y ork: S
im on an
d S
chuster, 1967), 266. 17 D
av id
S chw
artz, "P olitical A
lienation: A P
relim inary E
x p
erim en
t o n
th e P
sy- chology of R
evolution's F
irst S tage," p
ap er read
at th e A
nnual M eeting of th
e A
m erican P
olitical S cience A
ssociation (1967). 18 P
ortions of this section first ap p
eared in T
ed G
urr, "P sychological F
actors in
C ivil V
iolence," W orld P
olitics, X X
(January 1968), 247-251. 19 S
ee for exam ple K
urt R iezler, "O
n the P
sychology of th e M
o d
em R
evolution," Social R
esearch, X (S
eptem ber 1943), 320-336; portions of E
ric H offer's generally
useful T he True B
eliever: T houghts on the N
ature o f M
ass M ovem
ents (N ew
Y ork:
H arper, 1951); an
d D
o n
ald J. G
oodspeed, T he C
onspirators: A Study o
f the C oup
d'E tat (N
ew Y
ork: T h
e V iking P
ress, 1962).
30
llE L
A T
IV E
U E
P IU
V A
T IO
N A
N U
T H
E IM
P E
T U
S T
O V
IO L
E N
C E
m icroanalysis of particular events, b
u t contribute relatively little
to general theories of collective action. 2
0 A
ggression-prone victim s
of m aladaptive socialization processes are found in
every society, and
am ong
th e
actors in
m ost outbreaks
of political violence,
b u
t th ey
are m u
ch m
ore likely to b e m
obilized b y
strife than to constitute it in
its entirety. N or can a general theory of political
strife b
e b ased
solely on culturally specific
theories of m
odal personality traits, though it sh
o u
ld take account of their effects
(discussed in chapter 6). T
h e m
ost generally relevant psychologi- cal theories are those that deal w
ith the sources an d
characteristics of aggression in
all m en, regardless of culture. S
uch psychological theory
provides a
m otivational
base for
theory about political
violence an d
prov:i:des a m eans for identifying an
d specifyipg the
operation of sorne explanatory variables. T
h ere are three distinguishable psychological assum
ptions about the generic sources of h
u m
an aggression: that aggression is solely
instinctive, th at it is solely learn
ed , or th
at it is an innate response activitated by frustration.
21 O n
e or another of these is im plicit in
m ost theoretical approaches to
civil strife that have no explicit m
otivational base. T h
e instinct theories of aggression, rep resen
ted
am ong others by F
reu d
's qualified attribution of th e im
pulse to destructiveness to a death instinct an
d by L
orenz's view of aggres-
sion as a survival-enhancing instinct, assum e that m
ost or all m en
have w
ith in
th em
an autonornous source of aggressive im pulses,
a drive to aggress that, in L
orenz's w ords, exhibits "irresistible
outbreaks w
h ich
recur w
ith
rhythm ical
regularity." 22 A lthough
there is no definitive su p
p o
rt for this assum ption, its advocates,
including F reu
d an
d L
orenz, have often ap p
lied it to the explana-
tion of collective as w ell as individual aggression. 23 T
h e assum
p-
2 0
A recen
t study of this type is E . V
íctor W olfenstein, T
he R evolutionary Per-
sonality: L enin, T
rotsky, G andhi (P
rinceton: P rinceton U
niversity P ress, 1967).
21 A threat-aggression seq
u en
ce is discussed below .
22 K onard L
orenz, O n A
ggression (N ew
Y ork: H
arcourt, B race, an
d W
orld, 1966), chap. 4, quotation from
xii. T h
e aggressive instinct in anim
als an d
m an is said ordi-
narily to b e triggered b
y th
e p resen
ce or approach of an o
th er creature. In
th e absence
of such an activator, how ever, aggression w
ill occur spontaneously. S uch assertions
are su p
p o
rted b
y som
ew hat idiosyncratic observational reports on anim
al behavior. 23 S
ee for exam ple S
igm und F
reu d
, C ivilization and Its D
iscontents, trans. Joan R
iviere (L
ondon: T h
e H
ogarth P
ress, 1930); L
orenz, chaps. 13, 14; an d
F ranz
A lexander, "T
h e P
sychiatric A spects of W
ar an d
P eace," A
m erican ]ournal o
f So- ciology,
L X
V I (1941), 504-520. F
reu d
's instinctual interpretation of aggression is advanced in
his later w orks; his early view
w as th
at aggression is a response to frus- tration o
f pleasure-seeking behavior. F o
r review s an
d critiques of o
th er instinct
') theories o
f aggression see L eo
n ard
B erkow
itz, A ggression: A
Social P sychological
A nalysis (N
ew Y
ork: M
cG raw
-H ill, 1962), chap.
1, an d
R alph L
. H
ollow ay, Jr.,
fe "H
u m
an A
ggression: T h
e N
eed for a
S pecies-S
pecific F ram
ew ork,"
in M
orton F
ried an
d others, eds., W
ar: T he A
nthropology o f A
rm ed C
onflict and A ggression
- i'' · (C arden C
ity, N ew
Y ork: N
atural H istory P
ress, 1968), 29-48.
31 .
-
1""1
¡ J t ! 1
T
W H
Y M
E N
R E
B E
L
tion is ev id
en t in
H o
b b
es' characterization of m an in
th e state of
nature, an d
perhaps im plicit in
N ieburg's recen
t concern for "th e
people's capability for outraged, uncontrolled, bitter, an d
bloody violence," 24 b
u t plays no significant role in
contem porary theories
of civil strife. Just th
e opposite assum ption, that aggressive behavior is solely
or prim arily learned, characterizes th
e w ork of sorne child an
d so-
cial psychologists, w hose evidence indicates that sorne aggressive
behaviors are learn ed
an d
u sed
strategically in th
e service of par- ticular g
o als-
aggression by children an d
adolescents to secure at- tention, by adults to express dom
inance strivings, by groups in com -
petition for scarce values, by m ilitary personnel in
th e service of
national policy. 25 T his assum
ption, that violence is a learn ed
re- sponse, rationalistically chosen an
d dispassionately em
ployed, is com
m on to a n
u m
b er of recen
t theoretical approaches to collective conflict. A
m ong theorists of revolution, J ohnson repeatedly, though
not consistently, speaks of civil violence as "purposive," as "form s
of behavior intended to disorient th e behavior of others, th
ereb y
bringing about the dem
ise of a h ated
social system ." 26 T
im asheff
regards revolution as a "resid u
al" event, an ex p
ed ien
t "reso rted
to w
h en
o th
er w ays of overcom
ing tensions have failed." 27 M or-
rison attributes rural discontent an d
strife in developing nations
to "relative deprivation," defined as it is here, b u
t h e explicitly
assum es rationality in
th e behavior of the d
ep riv
ed w
h en
h e hypoth-
esizes that "all attem pts to red
u ce discontent are selected
on th e
basis of th e actor' s perception of the probability of th
e attem pt' s re-
ducing th e discontent." 28 P
arsons attem pts to fit political violence
into th e fram
ew ork. of social interaction theory, treating th
e resort to force as a w
ay of acting chosen b y
the actor(s) for purposes of deterrence, punishm
ent, or sym bolic dem
onstration of capacity to act. 29 S
chelling represents those conflict theorists w ho explicitly
24 H . L
. N ieburg, "T
h e T
hreat ofV iolence and S
ocial C hange," A
m erican P
olitical Science R
eview , L
V I (D
ecem ber 1962), 870.
25 A characteristic study is A
lbert B andura and R
ichard H . W
alters, Social L earning
and P ersonality D
evelopm ent (N
ew Y
ork: H
olt, R inehart and W
inston, 1963).
2 6
C halm
ers Johnson, R evolutionary C
hange (B oston: L
ittle, B row
n, 1966), 12, 13, italics added.
27 N icholas S. T
im asheff, W
ar and R evolution (N
ew Y
ork: S heed and W
ard, 1965), 154.
2 8
D en
ton E . M
orrison, "R elative D
eprivation and R ural D
iscontent in D evelop-
ing C ountries: A
T heoretical P
roposal," paper read at the A nnual M
eeting of the A
m erican A
ssociation for the A dvancem
ent of S cience (1966), 6.
29 T alcott P
arsons, "S orne R
eflections on the P lace of F
orce in Social P rocess,"
in H arry E
ckstein, ed., Interna! W ar: P
roblem s and A
pproaches (N ew
Y ork: T
h e
F ree P
ress, 1964), 34-35.
32
¡.
e t r t
R E
L A
T IV
E D
E P
R IV
A T
IO N
A N
D T
H E
IM P
E T
U S
T O
V IO
L E
N C
E
assum e rational behavior an
d in
terd ep
en d
en ce of adversaries' de-
cisions in all types of conflict. 30
T h
e third psychological assum ption is that m
uch aggression oc- curs as a response to frustratiori. "F
ru stratio
n " is an interference
w ith goal-directed behavior; "aggression" is behavior designed to
injure, physically or otherw ise, those tow
ard w hom
it is directed. T
h e disposition to resp
o n
d aggressively w
h en
frustrated is part of m
an' s biological m akeup; th
ere is a biologically in h
eren t tend-
ency, in m
en an
d anim
als, to attack th e frustrating agent. T
his is n
o t necessarily incom
patible w ith the p
reced in
g tw
o assum p-
tions. F rustration-aggression theory is m
ore system atically devel-
oped, how
ever, an
d
has substantially
m ore
em pirical
support than theories th
at assum e eith
er that all m en
have a free-flow ing
source of destructive en erg
y
th at all aggression is im
itative an d
instrum
ental. T
h e m
ost influential form ulation of frustration-aggression theory
w as proposed b
y D
ollard an d
his colleagues at Y ale in
1939. T h
e basic postulate is "th
at th e occurrence of aggressive behavior al-
w ays presupposes the existence of frustration and, contrariw
ise, that th
e existence of frustration alw ays leads to sorne form
of ag- gression." It is clear from
the rem ain
d er of th
e study that the sec- ond p
art of th e postulate w
as n o
t in ten
d ed
to suggest eith er that
aggression w as th
e only possible response to frustration, or that there
w as
no difference b etw
een the
instigation to aggression,
su b
seq u
en tly
called "an
g er,"
an d
the
actual occurrence of ag- gression.31 M
iller later offered a clarification: frustration produces instigations
to various
responses, o
n e
of w hich
is aggression.
If the non-aggressive
responses do
n o
t relieve the frustration,
"th e greater is th
e probability that the instigation to aggression eventually w
ill b eco
m e dom
inant so that sorne response of ag- gression
w ill
occur." 32 E
m pirical
studies identify fundam
ental responses to frustration o
th er than aggression. H
im m
elw eit sum
-
30 T hom
as C . S
chelling, T he Strategy ofC
onjlict (C am
bridge, M ass.: H
arvard U ni-
versity P ress, 1960), 4.
31 John D ollard and others, F
rustration a n
d A
ggression (N ew
H aven: Y
ale U ni-
versity P ress, 1939), quotation from
p. l. M ajor sum
m aries of the experim
ental and theoretical literature include H
ilde T . H
im m
elw eit, "F
rustration and A ggression:
A R
eview of R
ecent E xperim
ental W ork," in P
sychological F actors o
f P eace and
W ar, ed. T
. H . P
ear (L ondon: H
utchinson, 1950), 161-91; E lton D
. M cN
eil, "P sy-
chology and A ggression," ]ournal o
f C onjlict R
esolution, U I (}une 1959), 195-294;
A rnold H
. B
uss, T he P
sychology o f A
ggression (N ew
Y ork: John W
iley, 1961); A
ubrey J. Y ates, F
rustration a n
d C
onjlict (N ew
Y ork: John W
iley, 1962), especially chaps. 2-4; and B
erkow itz, A
ggression. 32 N
eal E . M
iller and others, "T h
e F rustration-A
ggression H ypothesis," P
sycho- logical R
eview , X
L V
IT I (July 1941), quotation from
339.
33
:¡ ¡ ,¡ 1 r-1 l 1,1 w 'lj w \r .. r[ :1 '1' \\ : '1 w 1+ 1¡: 'J¡ 1 ' ''''1 .:'!l
1 .
,.
W H
Y M
E N
R E
B E
L
m arizes
experim ental
evidence that frustration in children can
lead to regression in th e fonn of low
ering intellectual perform ance,
and to evasion. 33 F o
u r response patterns w
ere found in an exam
ina- tion of frustration-induced behaviors in a N
ew G
u in
ea tribe: sub- m
ission, d
ep en
d en
ce, avoidance,
and aggression. T
h e
prim ary
em otional response to frustrations am
ong children is rage or anger, m
odified by later learning experiences. 34 P rolonged frustration,
in th e forro of continuous unem
ploym ent, has b
een observed to
result in apathy. 35
-
T h
ese findings an d
observations are qualifications of th e basic
frustration-aggressÚ m
thesis, not refutations of it. T
h e basic ex-
planatory that frustration-aggression theory contributes
to the understanding of h
u m
an conflict, and specifically to th
e analysis of political violence, is th
e principie that anger functions as a drive. In
th e recen
t reform ulation of th
e theory b y
B erkow
itz, the perception of frustration is said to arouse anger. A
ggressive responses ten
d to occur only w
h en
they are evoked by an externa! cue, that is, w
h en
th e angered person sees an
attackable object or person that h
e associates w ith th
e source of frustration. T his argu-
m ent, and th
e experim ental evidence that supports it, suggests
that an angered person is not likely to strike o u
t at any object in
his environm ent, b
u t only at th
e targets h e thinks are respon-
sible. T h
e crucial point is that occurrence of such an attack is an inherently
s·atisfying response to
anger; if th
e attacker has
done sorne harm to his frustrator, his anger is reduced, w
h eth
er or not he succeeds in
reducing the level of frustration p er se. 36 If
frustration continues, aggression is likely to recur. If it is red u
ced
as a result of the attack, th e ten
d en
cy to attack is reinforced, an
d
th e onset of anger in th
e future is increasingly likely to b e accom
- p
an ied
by aggression. M
aier has u n
d ertak
en m
any studies w hich support the thesis
that innate frustration-induced behaviors becom e ends in them
- selves for the actors, u
n related
to further goals, and qualitatively different from
goal-directed behavior. H e suggests that there are
four frustration-induced responses, including regression, fixation, and resignation as
w ell as aggression. F
rustration-instigated be- havior is distinguished from
goal-directed behavior by a n u
m b
er of
33 H im
m elw
eit, 172. 34 J.
M . V
. W hiting, '"T
he F rustration C
om plex in K
w om
a S ociety," M
an, X
L IV
(N ovem
ber-D ecem
ber 1944), 140-144. 35 M
arie L azarsfeld and H
ans Z eisal, '"D
ie A rbeitslosen von M
arienthal," P sy-
chologische M onographen, v (1933), sum
m arized in
H irnrnelw
eit, 172. 36 O
n the drive properties of anger see B
erkow itz, "T
h e C
oncept of A ggressive
D rive," and S. F
eshbach, "'T he F
unction of A ggression and the R
egulation of A g-
gressive D rive," P
sychological R eview
, L X
X I (July 1964), 257-272.
3 4
r r ' '"1'
llE L
A T
IV E
D E
P R
IV A
T IO
N A
N D
T H
E IM
P E
T U
S T
O V
IO L
E N
C E
characteristics: it tends to b e fixed an
d com
pulsive; it is not neces- sarily d
eterred by punishm
ent, w hich m
ay instead increase the degree
of frustration; it takes
the forro m
ost readily available, little infiuenced by anticipated consequences; an
d it is satisfying
in itself. 37 F
urtherm ore the original goal w
hich suffered frustra- tion m
ay becom e largely irrelevant to behavior. "A
ggression then becom
es a function of the frustration, th e previously existing goal
response having b een
replaced b y
behavior w hich is controlled
by an entirely different process." 38
T h
e threat-aggression sequence is another behavioral m
echa- nism
that a n u
m b
er of psychologists h a ve argued is as fundam
ental, if not as com
m on, as th
e frustration-aggression relationship. C lín-
ica! and observational evidence suggests that th e greater the per-
ceived threat to life, th e greater the violent response. A
ccording to W
edge, "W h
en the .value directly at stake is life, violent response
occurs as reaction to fear rather than expression of anger." 39 S ur-
veys of th e effects of bom
bing on J a panes e, G erm
an, and E nglish
civilian populations during W orld W
ar II show that heavy bom
b- in
g s-in
clu d
in g
those of H iroshim
a an d
N ag
asak i-first produced
acute fear, not anger, b u
t also generally led to increased hostility
tow ard both th
e enem y and th
e g o
v em
m en
t that failed to p rev
en t
the bom bings. 40 E
xperim ents w
ith anim als provide substantiating
evidence: events that im m
ediately an d
actively threaten the con- tinued existence of th
e organism trigger avoidance-survival m
ech- anism
s, w hich can include extraordinarily violent behavior. T
h e
threat-aggression sequence can b e in
terp reted
as a special case of the
frustration-aggression relationship,
as B
erkow itz
does. A
. threat to life is an anÚ
cipated frustration; as the degree o f threat
increases, fear and anger rise sim ultaneously, and th
e extent to w
hich fear predom inates m
ay b e "a function of th
e individual's perceived pow
er to control or h u
rt his frustrater relative to the frustrater's p
o w
er to control or harm him
." 41 lt nonetheless seem s
likely tl1at people have a fundam ental disposition to respond ag-
37 M aier, 92-115, 159-161; also see Y
ates, 24-30, 36-56. 38 N
orm an R
. F . M
aier, "T h
e R ole of F
rustration in S ocial M
overnents," P sycho-
logical R eview
, L X
IX (N
ovem ber 1942), 587.
39 B ryant W
edge, '"T he C
ase S tudy of S
tudent P olitical V
iolence: B razil, 1964,
and the D om
inican R epublic, 1965," W
orld P olitics, X
X I (January 1969), 195-196.
A lso see Jerom
e F rank, Sanity and Survival: P
sychological A spects o
f W ar and
Peace (N ew
Y ork: V
intage B ooks, 1968), 75.
4 0
l. L . Janis, A
ir W ar and E
m otional Stress; P
sychological Studies o f B
om bing
and C ivilian D
efense (N ew
Y ork: M
cG raw
-H ill, 1951), 4-152. C
om parable reactions
have b een
observed in rnany disaster studies. S
ee for exam ple G
eorge W . B
aker and D
w ight W
. C haprnan, eds., M
an and Society in D isaster (N
ew Y
ork: B asic
B ooks, 1962). 41 B
erkow itz, A
ggression, 42-46, quotation from 45.
35 •. , f i:. ; l r f t 1 :{' l. 1 f, ¡;-r \ ,, ¡; S ¡,: ['! r f i·
l \ ·,;.; i{ l. ¡ ·-'
W H
Y M
E N
R E
D E
L
gressively to extrem e fear itself; if so, the response can reinforce
an d
b e reinforced b
y frustration-induced anger. T
h e relationship
appears especially
relevant in
evaluating
th e
effect of police
and m
ilitary actions in
precipitating
an d
prolonging collective
violence. T
h e
frustration-aggression an
d
the related
threat-aggression m
echanism s provide th
e basic m otivational link b
etw een
R D
an d
th
e potential for collective violence. T h
ey are n
o t inconsistent
w ith the p
resen ce of learn
ed an
d purposive elem
ents in acts of in-
dividual an
d
collective violence,
how ever.
M en
feel
d ep
riv ed
w
ith respect to w h
at th ey
have learn ed
to value an d
to w hat th
ey
have to do. T
h e beliefs an
d sym
bols th at d
eterm in
e the tim
ing, form s, an
d objects of violence are learned. If th
eir anger is pow
erful an d
persistent, m en
can em ploy m
uch reason an d
inven- tiveness in
devising w ays to givé it violent expression. S
orne such m
en m
ay learn to value violence for its o w
n sake. B
u t m
uch of this learning takes place after anger has already b
een aroused; indi-
viduals w
ho are
dispassionately violent
often are
using tech- niques th
at p ro
v ed
useful an d
satisfying in response to p
ast frus- trations.
T h
ere also is an ev id
en t sense of p
u rp
o se am
ong m any of th
e participants in
m ost outbreaks of collective violence, in
th e sen
se that th
ey expect violent action to en
h an
ce th eir value position.
R evolutionary leaders p
u t th
eir follow ers' anger to th
eir purpose of seizing pow
er; rioters take advantage o f disorder to loot stores
for food an d
furniture; dem onstrators h
o p
e to p ersu
ad e th
eir rulers to take rem
edia} action. T h
e nature an d
strength of th ese purposes
are m ajor determ
inants of th e form
an d
tactics o f collective vio-
lence. B ut in
m ost instances th
ey ap
p ear to reinforce or ch
an n
el th e
ím petus to violence, an
d are infrequently an autonom
ous m otive
for violence. T his assertion is m
ad e w
ithout an attem pt to su
p p
o rt
it, b u
t it is suceptible to em pirical test. It is true, an
d th
e frustra- tion-aggression relationship is significant for política} violence; to th
e extent that actors in political violence m anifest or adm
it to sorne degree of anger. 42
In sum
m ary, the prim
ary source of th e h
u m
an capacity for vio-
len ce appears to b
e the frustration-aggression m echanism
. F rus-
tration does
not necessarily lead to violence, an
d violence for
42 A nger is quite often accom
panied by other affective elem ents, notably excite-
m ent; th
e test proposed here is w h
eth er anger is present in
substantial degree. E
xciternent does not seern to b e an in
d ep
en d
en t affective state b
u t a general em
o- tional
arousal that precedes
an d
accom panies
goal-consurnrnatory activity; th
e angry m
an w ho anticipates a riot is excited for m
uch the sam e reasons that a starv-
ing m an is excited at th
e sm ell of food cooking.
3 6
1
·!'
R E
L A
T IV
E D
E P
R IV
A T
IO N
A N
D T
H E
IM P
E T
U S
T O
V IO
L E
N C
E
sorne m en
is m otivated b
y expectations of gain. T
h e anger in
d u
ced
by frustration, how ever, is a m
otivating force that disposes m en
to aggression, irrespective of its instrum
entalities. If frustrations are sufficiently prolonged or sharply felt, aggression is quite likely, if not certain, to occur. T
o conclude that th
e relationship is not rele- vant to individual or collective violence is akin to the assertion that the law
o f gravitation is irrelevant to the theory o
f flight because not everything th
at goes u p
falls back to earth in accord w ith the
basic gravitational principie. T h
e frustration-aggression m echanism
is in
this sense analogous to th e law
of gravity: m en
w ho are frus-
trated have an innate disposition to do violence to its source in
proportion to th e intensity of th
eir frustration, ju st as objects are
attracted to one another in direct proportion to th
eir relative m asses
an d
inverse proportion to th eir distance. A
n u
m b
er of o th
er vari- ables influence th
e behavior of m en, an
d of objects, in
such cir- cum
stances: fo rm
en , th
eir beliefs, inhibitions, an d
social environ- m
ent; for objects in a gravitational field, th
eir energies, configura- tion, an
d the properties o
f th e m
ed iu
m in
w hich th
ey are situated.
B ut it seem
s ev en
less feasible to account for political violence w
ithout reference to th e properties o
f that dispose th
em to
violence th an
it is to construct a theory of flight w ithout refer-
en ce to the law
of gravitation. O n
earth, gravity can b e assum
ed a constant; am
ong m en, levels o
f frustration vary greatly.
R elativ
e D ep
riv atio
n a
n d
A n
alo g
o u
s C au
ses o f
P o
litical V io
len ce
R elative
deprivation, defined
as p
erceiv ed
discrepancy
be- tw
een
value expectations
an d
value capabilities, is
sufficiently general to com
prise o rb
e related to m
ost of the general "precondi- tions o
f revolution" identified in o
th er theoretical analyses. S
orne of th
ese conceptual relationships are exam ined here, n
o t to dem
on- strate th
at R D
is so m
eh o
w a "correct" concept an
d that others are
not, b u
t to show that in
addition to its relatively clear definition it can synthesize diverse o
th er notions.
F o
r A ristotle the principal cause of revolution is th
e aspiration for econom
ic or political eq u
ality on th
e p art of the com
m on p
eo p
le w
ho lack it, an d
th e aspiration o
f oligarchs for greater inequality . than they have, i.e. a discrepancy in both instances b
etw een
w hat
people have o f political an
d econom
ic goods relative to w hat they
think is justly theirs. 43 E d
w ard
s, w riting sorne tw
enty-three cen- turies later, asserts th
at all revolutions are d u
e to "repression of
43 T he P
olitics o f A
ristotle, trans. J. E . C
. W elldon (N
ew Y
ork: M acm
illan, 1883, 1905), 338-342.
3 7
)\ ' •j 1 ( •
¡_¡_'_:¡: ,,
·i¡ :¡,J
i.¡¡!
1 1.· ·jí : :¡ ! L ¡ .ir, ·¡·i¡¡ í• > i! 1 n f:,¡ lil' 1 d:t! :!FJ! '' t! '·' 1_···· r __ ¡,l•J_ 1 ,.¡. 'il!fl' 1· 1 : !1 ¡¡ ¡ '' 1 f:M ,¡¡ _-'tm .··. ¡[!:& ·¡··l! ·'-'1 ·:1 ·l.íjri' 1 .¡1_:_* . ·¡·_·,_1: '!, ·-w: h: ill;' F''' ''""i ¡"¡ r.;¡l li'''·! '3:! 1 {: ' 111 i :f\i' . •_·:! f_! i ¡.-•,¡¡ ii! ji fi !! ,¡j,l!' :·YtJP.
! '
1 1 >
W H
Y M
E N
R E
D E
L
elem ental w
ish es" an
d th
at th e violence of any revolution is pro-
portional to the d eg
ree of such repression. T h
e sense of repression, or "b
alk ed
disposition," develops w h
en "p
eo p
le com e to feel that
their legitim ate aspirations and ideas are being rep
ressed or per-
verted, that their entirely proper desires an d
am bitions are b
ein g
h
in d
ered
and th
w arted
. . . . " 44
P ettee's
concept of
"cram p"
sim ilarly resem
bles R D
.. P eople feel "cram
p ed
" w h
en th
ey find
that satisfaction of their basic needs for liberty an d
security is interfered w
ith, an d
m oreover regard this repression as unneces-
sary and avoidable, h en
ce unjustified. "A revolution takes place
w h
en th
e great m ajority of th
e society feel cram ped b
ey o
n d
toler- ance." 45
A nalogous'concepts are u
sed b
y contem
porary theorists. L ass-
w ell and K
aplan attribute political instability to the discrepancy b
etw een
·expectations an d
the "d:'egree o f .
. . realization of value
for th
e m
ass . . . . lt
is a
low
d eg
ree of realizatio
n -disparity
b etw
een
value position
an d
value
d em
an d
ed
and ex
p ected
- w
hich is
m ost
directly effective." 46
Z ollschan
argues that
all activity, including revolutionary activity, begins w
ith "exigency," defined as "a discrepancy (for a person) b
etw een
a consciously or unconsciously d
esired or expected state o
f affairs an d
an actual situation." 47
B oth
of th
ese concepts
m ake
assum ptions
about states of m
in d
o f revolutionary actors, as do th
e concepts m en-
tioned in the p reced
in g
paragraph. J ohnson m akes no such assum
p- tions in identifying a "d
iseq u
ilib rated
social system " as a neces-
sary precondition
for revolution,
w hich
is explicitly
a m
acro- analytic
concept. Its
m anifestations
at the
individual level
of analysis m
ay b e T
eadily in terp
reted in term
s o f R
D , how
ever: it constitutes a discrepancy b
etw een
m eu's value expectations (col-
lectively, their "value structures") au d
their m eans for attaining
those values (collectively, the social system ' s "pattern of adapta-
tion to the environm
ent" an
d its capacity to "fulfill functional
requisites"). 48 S
orne theorists explicitly use th e term
s "frustration" or "depriva- tion" to represent th
e ím petus to collective violence. D
avies at- tributes
revolutionary outbreaks to th e frustration w
hich results
•• L yford P. E
dw ards, T
he N atural H
isto1·y o f R
evolution (C hicago: U
niversity of C
hicago P ress, 1927), 3-4, 30-33.
45 G eorge S. P
ettee, T he P
rocess o f R
evolution (N ew
Y ork: H
arper, 1938), chap. 2, quotation from
33. 46 L
assw ell and K
aplan, 264. 47 G
eorge K. Z ollschan and W
alter H irsch, eds., E
xplorations in Social C hange
(B oston: H
oughton M iffiin, 1964), x"xv, 89.
•• Johnson, chap. 5. F or a sim
ilar analysis see N eil J. S
m elser, T
heory ofC ollective
B ehavior (N
ew Y
ork: T h
e F ree P
ress, 1963), chaps. 2, 3.
3 8
R E
L A
T IV
E D
E P
R IV
A T
IO N
A N
D T
H E
IM P
E T
U S
T O
V IO
L E
N C
E
from a short-term
decline in achievem ent follow
ing a long-term
increase that generated expectations about continuing increase.49 L
ern er sim
ilarly describes th e gap b
etw een
w hat people w
ant and w
hat they get as "frustrating" an d
. suggests revolutionary conse- quences. "T
h e spread of frustration in areas developing less rapidly
than their p eo
p le w
ish can b e seen
as th e outcom
e o f a d
eep im
- balance b
etw een
achievem ent an
d aspiration . . . aspiration out-
runs achievem ent so far that m
any people, e ven if they are m aking
sorne progress tow ard th
eir goal, are dissatisfied because they get so m
u eh less than th
ey w
ant." 5° C rozier says that th
e one elem en
t com
m on to all rebels is frustration, defined as "th
e inability to do som
ething one badly w ants to do, through circum
stances beyond one's control." 51
T h
e F eierab
en d
s associate political instability w
ith aggressive behavior, w hich is said to vary w
ith the extent o f
"system ic frustration." T
h e extent of system
ic fm stration is the
ratio of social w ant satisfaction to social w
ant form ation, or, in
R D
term
s, the discrepancy b etw
een p
resen t value position an
d value
expectations. 52
R elative deprivation is related to frustration by C
oser, and ap- p
lied to th
e explanafion of suicide rates. 53 H
oselitz an d
W illner
extending their distinction b etw
een expectations an
d aspirations:
link deprivation w ith the potential for revolution.
U nrealized
aspirations produce
feelings o
f disappointm
ent, b
u t unrealized
expectations result in feelings of deprivation.
D isappointm
ent is generally tolerable; deprivation is often in- tolerable. T
h e d
ep riv
ed individual feels im
pelled to rem edy,
b y
w hatever m
eans are available, th e m
aterial an d
psychic frus- trations p
ro d
u ced
in him . W
hereas disappointm ent m
ay b reed
th
e seeds of incipient revolution, deprivation serves as a catalyst for revolutionary action. 54
4 9
!am es C
. D
avies, "T ow
ard a T heory of R
evolution," A m
erican Sociological R
em ew
, X X
V II (F
ebruary 1962), 5-19. T h
e D avies thesis is also used by R
aym ond
and M anus M
idlarsky, "A T
heory o f R
evolution," ]ournal o f C
onfiict R eso-
lutw n, X
I (S eptem
ber 1967), 264-280, w ho characterize the discrepancy as a "rev-
olutionary gap." 50 D
aniel L em
er, "T ow
ard a C om
m unication T
heory of M odernization: A
S et of
C onsiderations," in L
ucían W . P
ye, ed., C om
m unications and P
olitical D evelop-
m ent (P
rinceton: P rinceton U
niversity P ress
1963) 327-350 quotations from 330-
335. ,
, ,
51 B rian C
rozier, T he R
ebels: A Study o
f P ost-W
ar lnsurrections (L ondon: C
hatto and W
indus, 1960), 15-16. 52 Ivo K.
and R osalind L
.. F eierabend, "A
ggressive B ehaviors W
ithin P olities
1948-1962; A C
ross-N ational S
tudy," ]oum al ofC
onfiict R esolution x
1966), 250-251. ,
53 L ew
is A . C
oser, C ontinuities in the Study o
f Social C onfiíct (N
ew Y
ork: T he
F ree P
ress, 1967), 56-62. 54 H
oselitz and W illner, 363.
3 9
r . ' í :, .:¡ j -
i)
'l _i
W H
Y M
E N
R E
B E
L
T h
e anthropological literature
on A m
erican In d
ian response to
w hite co
n q
u est also m
akes use of th e deprivation concept. N
ash, for
exam ple,
show s
how deprivation m
ay occur eith er through
acceptance or
rejection by
Indians of w
hite values an
d skills,
an d
proposes th at th
e aggressive com ponents in
In d
ian revivalism
are a response to that deprivation.
55 G esch
w en
d er attributes th
e A
m erican N
egro revolt of the 1960s to "relative deprivation," de- fined in its conventional sociological sense of status discrepancy vis-a-vis a reference group. 56 G
altung, although h e does not u
se the concept of relative deprivation p
er se, attributes aggression w
ithin an d
am ong societies to status discrepancy, or "ran
k dis-
equilibrium ," in
w hat is essentially a generalized rephrasing of
the A ristotelian thesis w
ith w
h ich
this catalog of concepts began. If m
en or groups are high on o
n e dim
ension of a stratification sys- tem
, b u
t low on another, e.g., if th
ey have high p
o w
er or education b
u t low
incom e, th
ey are said to b
e disposed to use aggression to attain a high or eq
u ilib
rated position on all dim
ensions. 57
T his catalog could b
e ex ten
d ed
at length, b u
t only at the risk of belaboring th
e A
lm ost all theories-
th at p
u rp
o rt to ex-
plain violent collective b eh
av io
r assign a central place to a variable or co
n cep
t that generally an d
often specifically resem bles R
D as it
is defined here. S orne salient characteristics of th
e R D
variable are
not necessarily incorporated in th
ese o th
er concepts, how -
ever. S orne of them
, particularly those m aking use of "w
an t/g
et" form
ulations, m ake no referen
ce to th e justifiability or intensity o
f m
en ' s
val u e expectations,
nor to th e theoretical desirability of
taking into account b o
th actual an
d anticipated discrepancies b
e- tw
een
goals an
d
attainm ents.
M oreover,
w h
ile m
any of th
em
specify by illustration th e kinds of societal and political condi-
tions that constitute the variable or increase its m agnitude, few
include
specific propositions
ab o
u t its
determ inants,
an d
only sorne suggest categories for classifying the variable's m
anifesta- tions.
F inally, m
any theories do not provide a m
otivational ra- tionale for th
e causal connection th ey
propase b etw
een th
e vari- ables an
d the violent events tow
ard w hich it is su
p p
o sed
to dis- pose m
en.
55 P hi !leo N
ash "T h
e P lace of R
eligious R evivalism
in th
e F orm
ation of the ln ter-
cultural on K
lam ath R
eservation," Social A nthropology o
fN o
rth A
m eri-
can T ribes, ed. F
red E
ggan (C hicago: U
niversity of C hicago P
ress, 1937), •• Jam
es A . G
eschw ender, "S
ocial S tructure and the N
egro R evolt: A
n E xam
m a-
tion of S orne H
ypotheses," Social F orces, X
L IIT
(D ecem
ber 1964), 248-256. 57 Johan G
altung, "A S
tructural T h
eo ry
of A ggression," ]ournal ofP
eaceR esearch,
N o. 2, 1964, 95-119. G
altung also posits severa! variables that intervene b etw
een
rank disequilibrium and aggression, including lack of alternative m
eans an d
th e
extent o f cultural experience in
aggression.
4 0
R E
L A
T IV
E D
E P
R IV
A T
IO N
A N
D T
H E
IM P
E T
U S
T O
V IO
L E
N C
E
T h
ere are
th ree
o th
er concepts frequently
em p
lo y
ed in
the analysis
of disruptive collective
b eh
av io
r th at are
not directly analogous to R
D b
u t that appear .to b
e alternatives to it: dissonance, anom
ie, an d
conflict. W ithout passing ju
d g
m en
t on th eir analytic
usefulness in o
th er contexts, it can b
e su g
g ested
th at th
ey relate
to conditions eith er m
ore or less specific than R D
. D
issonance is a co n
cep t w
idely u sed
in individual psychology.
In F
estin g
er's form ulation of cognitive dissonance theory, the term
refers to inconsistency b
etw een
tw o cognitive elem
ents or clus- ters of elem
ents. C ognitive elem
en ts, "th
e things a person know s
about him self, ab
o u
t his behavior, an d
about his surroundings," are dissonant if "th
e obverse of one elem en
t w ould follow
from
the other." 58 F o
r exam ple, a citizen w
ho b eliev
ed th
at his govern- m
en t w
ould b eg
in a w
ar only if attacked an d
th en
learn ed
through new
s m ed
ia th at it h
ad initiated a w
ar w ithout provocation is said
to experience dissonance. T h
e assum ption is th
at p eo
p le sensing
dissonance are
m otivated to red
u ce or elim
inate it, w hich th
ey
can do b y
changing th eir behavior or beliefs, b
y changing th
e corresponding situation, or by seeking n
ew inform
ation to red u
ce dissonance
w hile avoiding inform
ation that w o
u ld
increase dis- sonance.59
D ep
en d
in g
on th e m
agnitude of dissonance, i.e. th e
im portance or value of th
e dissonant elem en
ts to him , th
e citizen m
ay change his view s about th
e governm ent, attem
p t to change th
e personnel or policies of th
e governm ent through political action,
or, m ost likely in
this instance, d en
y th
e validity of the evidence that th
e w ar w
as b eg
u n
w ithout provocation an
d b
e receptive to any ev
id en
ce of provocation, how ever flim
sy. M
any hypotheses have b
een proposed an
d considerable em
pirical research done on effects of dissonance in
decision-m aking, com
pliance, receptivity to inform
ation, social su p
p o
rt for m ass p
h en
o m
en a, an
d the in-
fluence process in sm
all groups. 60 T
h e
dissonance an
d R
D
concepts n eith
er com prise nor sub-
sum e o
n e an
o th
er b u
t overlap. R D
is p erceiv
ed w
ith reference to individuals' w
elfare, pow er, an
d interpersonal value expectations;
dissonance can obtain am ong any set of cognitive elem
ents, not only those
that relate
to v
alu ed
goods an
d conditions
of life. M
oreover only sorne perceptions of deprivation entail dissonance, in
its original sen se of contradiction am
ong cognitive elem ents.
58 L eon F
estinger, A
T heory o
f C ognitive D
issonance (S tanford: S
tanford U ni-
versity P ress, 1957), quotations from
4, 13. T h
e perception of contradiction can arise on logical, experiential, cultural, o
r other grounds, according to F estinger.
59 F estinger, 18-24.
6° F estinger, passim
; Jack W . B
rehm and A
rthur R . C
ohen, E xplorations in C
ogni- tive D
issonance (N ew
Y ork: W
iley, 1962).
41
\1 11 \i :¡¡ '1 \¡ t¡
W H
Y M
E N
R E
B E
L
M ost black A
m ericans know
very w ell th
at th eir expectations of
econom ic an
d social equality w
ill b e only grudgingly an
d gradually
satisfied, if at all; th e failure of the political system
to take m assive
rem edial action is "d
isso n
an t" only for those N
egroes w ho at o
n e
tim e thought th
at it w ould. F
o r the great m
ajority th e lack of ac-
tion m erely confirm
s prior cognitions of the system an
d intensifies
discontent. T h
e dissonance co n
cep t an
d hypotheses seem
directly applicable to only o
n e particular aspect of the processes b
y w
h ich
th
e p ercep
tio n
of R D
develops: w h
en an individual first perceives
inadequacies in value opportunities h
e h ad
thought w ere appro-
priate an d
sufficient for attainm ent of his value expectations, h
e can b
e said to éxperience dissonance. T h
e onset of R D
through increasing value expectations, or the existence of a discrepancy b
etw een
expectations an d
capabilities p er se, does n
o t itself con-
stitute nór as
F estin
g er p
o in
ted . o
u t does
dissonance ex ist·w
h en
ev er an individual encounters res-tstance or
frustration in trying to achieve an objective. 61
A nom
ie, in th
e sense D u
rk h
eim u
sed it in
Suicide, is a situation ...
in w
hich eith er en
d s (value expectations) outstrip m
en 's m
eans, or ends rem
ain constant w hile m
eans are severely restricted, an d
corresponds quite closely to th
e R D
concept. T h
e m ore general-
ized sense that D u
rk h
eim gave the term
in T
he D ivisíon o
f L abor
in Society, one w h
ich M
erton popularized in his essay o n
"S ocial
S tructure an
d A
nom ie," is th
at anom ie is a b
reak d
o w
n of social
standards governing social behavior, or norm lessness. 1t is specif-
ically a sociological concept: "th e degree of anom
ie in a social
system is in
d icated
b y
th e extent to w
h ich
there is a lack of con- sensus on norm
s ju d
g ed
to b e legitim
ate, w ith its atten
d an
t uncer- tainty and insecurity in
social relations." 62 M anifestations of nolm
- lessness
w ithin
individuals are
characterized as
anom ia.
H ose,
sum m
arizing the
literature, identifies
three kinds
of anom
ie: w
eakness of norm s p
er se, th e existence of several strong b
u t con-
flicting norm s, an
d ignorance of norm
s. A ll three situations lead
to the pervasive sen
se of uncertainty to w hich is attributed m
u ch
deviant behavior, in
clu d
in g
crim inality, suicide an
d drug addic-
tion, an d
gang behavior. M erton originally suggested th
at anom ie
co u
ld lead
to w id
esp read
d ev
ian t behavior an
d th
e establishm ent
of alternative norm s, w
h ich
constitutes "rebellion." "W h
en rebel-
61 F estinger, 278.
62 R obert K. M
erton, "S ocial S
tructure and A nom
ie: C ontinuities," Social T
heory and Social Structure, rev. ed. (N
ew Y
ork: T h
e F ree P
ress, 1957), quotation from
266-267. T his discussion of the anom
ie concept m akes substantial use of G
ordon R
ose "A
nom ie and D
eviation," 29-45. A m
ore com prehensíve sum
m ation of w
ork use of the concept is
M arshall B
. C
línard, ed., A nom
ie and D eviant B
e- havior: A
D iscussion and C
ritique (N ew
Y ork: T
h e F
ree P ress, 1964).
4 2
R E
L A
T IV
E D
E P
R IV
A T
IO N
A N
D T
H E
IM P
E T
U S
T O
V IO
L E
N C
E
lion becom es en
d em
ic in a substantial p
art of th e society, it pro-
vides a potential for revolution, w h
ich reshapes b
o th
th e norm
a- tive an
d th
e social structure." T h
ere seem to h
a ve b e en
no sub- stantial later attem
pts to relate ahom ie to th
e occurrence of collec- tive violence o
th er th
an gang behavior, how
ever. 63 A
nom ie can b
e related to th e R
D concept, as defined here, in
tw
o w ays. If group norrns are w
eak or in conflict about how
m em
- bers can satisfy value expectations, value opportunities are thereby lim
ited. T his is particularly th
e case w ith respect to personal an
d
societal opportunities;
the less
certain p eo
p le
are about w
hat w
ays of acting are appropriate for attaining th
eir goals, i.e. the greater th
eir anom ia, the low
er th eir value capabilities are likely
to b e an
d h
en ce th
e greater is R D
. It sh o
u ld
b e n
o ted
that m en
can experience norm
lessness or norm conflict in
any social role an d
outside social roles, for exam
ple as subordinates in an authority
relationship or as passengers
on a jet flight. T
h u
s anornie con- stitutes or increases R
D only w
h en
it relates to norrns of value- m
aintaining or vahie-enhancing activities. It also can b
e argued, partly on th e basis of cognitive dissonance
theory, th at sets of internally consistent norrns are intrinsically
valued an d
th at th
e breakdow n of system
s of norm s therefore con-
stitutes R D
w ith
resp ect to "co
h eren
ce" values. T h
e argum ent is
not th at everyone w
ho suffers from anornia is therefore deprived;
it is th at R
D is ex
p erien
ced b
y those persons w
ho at one tim e ac-
cep ted
an internally consistent set of norrns as valid guides to
action b u
t h a ve su
b seq
u en
tly felt those norm
s to b e seriously chal-
lenged, w ithout b
ein g
replaced by another internally consistent set o
f norm s. D
eprivation in this instance results from
value loss, i.e. a declining value position w
ith resp
ect to the ideational co- h
eren ce category of interpersonal values. In
an anom ic society,
how ever, there are likely to b
e rnany p eo
p le w
ho have n ev
er h eld
one u
n q
u estio
n ed
, consistent set of nonns. N ever having h
ad a
consistent set of norrns, th ey
are less likely to b e discontented over
its lack than those w ho experience a threat to or loss of strongly
h eld
norm s.
T h
e m ost p
o ten
t effect of anom ie o
n R
D probably is its im
pact on value
opportunities. W
h eth
er norm
ative systerns
are intrin- sically
v alu
ed in
the · sense th at econornic goods, security, an
d
63 M erton, 191. O
ne application to political violence is E lw
in H . P
ow ell, "R
eform ,
R evolution, and R
eaction as A daptations to A
nom ie," R
eview o
f M exican Sociology,
X X
V (1963), 331-355. D
avid C . S
chw artz has developed a process m
odel of revolu- tionary behavior w
hich uses the related concept of alienation and specifies the círcum
stances in w hich política! alienation occurs, in "A
T heory of R
evolutionary B
ehavior," in Jam es C
. D avies, ed., W
hen M en R
evolt and W h
y (N ew
Y ork: T
he F
ree P ress, 1970).
4 3
1
l t
W H
Y M
E N
R E
B E
L
status are valued is an em pirical question m
ost directly an sw
ered
by ascertaining the extent to w hich threats to norm
system s an
g er
people. G enerally, the anom
ie and anom ia concepts are related
to R D
in a w ay com
parable to dissonance: they overlap R D
, b u
t they cannot b
e su b
su m
ed by it nor do th
ey incorporate all or m
ost aspects of it.
"C onflict" in
its collective sense is som etim
es defined as a con- dition, som
etim es as a process, an
d som
etim es as an event. G
altung defines it as a condition: "A
n action-system is said to be in con-
flict if the system has tw
o or m ore incom
patible goal-states." 64 C
oser initially defines it as a process, "a struggle over values an d
claim
s to scarce status, p o
w er an
d resources in
w hich the aim
s of the opponents are to neutralize, in
ju re or elM
ninate th eir rivals." 65
In conventional usage conflict
an event, a violent or nonviolent clash
tw o groups. B
y· eith er of th
e first tw o
conflict resem bles th
e R D
co n
cep t m
ore closely than eith er dis-
sonance or anom ie. 66 O
n e difference is that R
D refers both to in-
dividual states of m ind an
d th
eir collective w
hereas social conflict is
generally treated as a property o f collectivities
w ithout reference to its
m anifestations in
th e m
inds of th e in-
dividuals involved. G altung identifies a m
ore consequential dif- ference:
C onflict sh
o u
ld . . . b
e distinguished from frustration, w
hich is th
e m ore general case w
h ere goals are not achieved (needs
are not satisfied, gratification n o
t obtained, values not fulfilled, etc.) for sorne
reason. A
very sim ple case is
that of scarcity. . . . A
nother sim ple case is
w h
en som
ething is· blocking th e
access to
the source
o f gratification .
. . . B
ut th
e m
ost im
- portant special case .
. . is
th e case o
f conflict w here efforts
by o n
eself or others to obtain sorne value can b e seen
as th e
source of frustration. 6
7
C onflict defined as a condition is essentially a special case of R
D
in w
hich the so urce of the discrepancy b etw
een val u e expectations
an d
capabilities is another group com peting for th
e sam e values.
C onflict defined as
a process
refers to
the interaction b etw
een
groups in th eir respective attem
pts to alleviate R D
. 64 Johan G
altung, "Institutionalized C onH
ict R esolution: A
T heoretical P
aradigm ,"
]ournal o f P
eace R esearch, N
o. 4, 1965, 348. •• C
oser, T he F
unctions o f Social C
onfiict, 8. 6
6 C
onH ict al so has a psychological usage not exam
ined here, nam ely the presence
in an individual of tw o (or m
ore) com peting and incom
patible m otives. F
or a re- view
of this concept and the literature see Y ates, esp. chap. 5.
67 G altung, "Institutionalized C
onflict R esolution," 349.
4 4
R E
L A
T IV
E D
E P
R IV
A T
IO N
A N
D T
H E
IM P
E T
U S
T O
V IO
L E
N C
E
T h
e conflict concept is less appropriate than R D
for the analysis of political violence for severa! reasons. O
n e is sim
ply that in its
connotations o f process or class. o
f events it includes by definition sorne of the variables th
at this analysis propases to explain. 1 hope to account for th
e form s and extent of violent política! conflict by
proposing sorne of th eir general determ
inants, not to "explain" them
b y
defining th em
as conflict. M oreover, conflict defined as
a condition in w
hich group X has w
hat group Y w
ants refers to a particular k
in d
of R D
, b u
t b y
no m eans exhausts the R
D concept.
R D
refers to any collectively felt discrepancy b etw
een a sought
an d
an attainable value position, w
h eth
er or not sorne other group has th
e value sought an d
w h
eth er or not group Y
tries to seize it from
group X . T
h ere are psychological grounds, suggested in the
p reced
in g
section, for expecting collective violence to result from
any R D
that is of w ide scope an
d intensity. It also seem
s ev id
en t
that m an
y acts of political violence, revolutionary m
ovem ents in
particular, do n
o t involve a struggle for values so m
uch as a d em
an d
that system
s b e resh
ap ed
so th at they can create n
ew values. In
this connection, one difficulty that arises if political violence is treated as a category o
f conflict is that política! violence is th ereb
y
assum ed to entail a struggle for scarce values, an assum
ption that forecloses exam
ination of o th
er causal factors. O
n e last lim
itation o f conflict theory for o
u r purposes is th
e distinction com
m only m
ade b y
conflict theorists b etw
een w
h at is
called "realistic" an d
"nonrealistic" conflict (C oser), or "rational"
an d
"nonrational" conflict (S chelling), or "destructive behavior"
and "conflict b eh
av io
r" (G altung). T
h e essence o
f the distinction is
b etw
een
actions instrum
ental in securing the values
sought and actions destructive for th
eir ow n sake. T
h e analytic usefulness
of th e distinction is
n o
t in question; w hat ís questionable is at-
tem pting to account for política! violence using theoretical ap-
proaches that assum e th
at only the instrum ental m
anifestations of violence are relevant or subject to analysis. C
o ser an
d G
altung, am
ong o th
er conflict theorists, recognize th at b
o th
elem ents are
p resen
t in
m ost
conflict. C
o ser
criticizes others
for failure
to realize th
at "conflict m ay be m
otivated b y
tw o distinct yet inter-
m in
g led
facto rs-
a realistic conflict situation an d
th e affective in-
vestm ent
in
it . . . . "
G altung
sim ilarly
m akes
the theoretical
point that underlies this en tire discussion: "conflict behavior tends
to becom e destructive behavior (because of th
e frustration-aggres- sion cycle) an
d destructive b
eh av
io r tends to b
e self-reinforcing." 68 D
esp ite this recognition, m
ost o f the concepts an
d hypotheses
•• C oser, 59; G
altung, 349. A lso see M
ack and S nyder, 219, 222-223.
4 5
i[ ¡:( ii ![l 'f¡ ¡' ,f 1'1 ,1 1 J ,, ,,
j··
l !-
: .. ¡
¡\· ¡j ,.¡ 1' ;¡· 1 ·!· ,1::
W H
Y M
E N
H E
B E
L
of conflict theory
are co
n cern
ed
w ith
th e
instrum entalities
of strife. T
his analysis gives equal w eight to its nonrational origins
an d
m anifestations.
P a tte
rn s o
f R elativ
e D ep
riv atio
n
In static term
s, R D
is a discrepancy b etw
een value expectations
an d
value capabilities, its intensity an
d scope
d eterm
in ab
le in
any accessible population b y
th e use of survey an
d o
th er tech-
niques. D ynam
ic analysis requires conceptual tools th at take into
account patterns of changes in value expectations an
d value capa-
bilities over tim e. O
n e can b
eg in
w ith th
e assum ption that, be-
cause R D
is a psychically uncom fortable condition, m
en ten
d over
th e long ru
n
to adjust
th eir value
expectations to
th eir
value capabilities.
S ocietal conditions
in w h
ich ·sought an
d attainable
value positions are in approxiU
:ate eq u
ilib riu
m consequently can
b e reg
ard ed
as "norm
al," h o
w ev
er uncom m
on th ey
m ay b
e in
the contem porary w
orld, an d
provide a base-line from w
hich to evaluate patterns of change. T
h ree distinct patterns of disequi-
librium can
b e
specified: decrem ental deprivation, in
w hich a
group' s value expectations rem
ain relatively constant b u
t value capabilities
p erceiv
ed to
decline; aspirational
deprivation, in
w h
ich capabilities rem
ain relatively static w hile expectations
increase or
intensify; 69 an
d
progressive deprivation,
in w
hich there is substantial an
d sim
ultaneous increase in expectations an
d
decrease in
capabilities. A ll th
ree patterns have b een
cited as
causal or predisposing factors for political violence.
D E
C R
E M
E N
T A
L D
E P
R IV
A T
IO N
T h
e m o
d el sh
o w
n graphically in
figure 1 represents settings
in w
hich group consensus about justifiable value
positions has varied little
tim e, b
u t in
w h
ich the average attainable value
position or P<!tential is p erceiv
ed to decline substantially. M
en
in th
ese circum stances are an
g ered
over th e loss of w
h at th
ey once
h ad
or thought they co u
ld have; th
ey experience R
D by reference
to th eir o
w n
p ast condition. T
h e value position of an entire so-
ciety m ay fall b
ecau se of declining production of m
aterial goods, declining capacities of th
e political elite to provide order or re- solve crises, im
position of foreign rule, or loss of faith in th
e so- ciety' s
in te grating
structure of beliefs
an d
attendant n
o n
n s
of action. V
alue capabilities also m ay fall
one or m ore seg-
m ents o
f society b ecau
se its m em
bers lose out in absolute tenns
in conflict w
ith o th
er groups over scarce values.
E xam
ples in-
69 T hese patterns are labeled and described brieH
y by M orrison, 5.
4 6
R E
L A
T IV
E D
E P
I'IV A
'
T IO
N A
N D
T H
E IM
P E
T U
S T
O V
IO L
E N
C E
high
C ollective
val u e position low
value expectations ----------------------------------------
value capabilities
F igure l. D
ecrem enta! deprivation.
elude of progressive taxation on the w
ealthy an d
of taxation on th
e poor; th e loss of political influence by
an d
oppositional groups
n ew
ly
b arred
from political ac-
tlvlty; an
d the
decline in status
an d
influence felt by m iddle-
class groups
the status
of w
orking-class groups
increases. T
h e val_ue p
o sltw
n or potential of a particular group also m
ay seem
to d
eclm e
not because of any
dim inution
or redistribution of total values available in
a society b u
t because of the declinin or
appropriateness o
f opportunities,
for exam
ple th
! shnnk1_ng
of em
p lo
y m
en t
opportunities for
unskilled labor
m
h1ghly industrialized societies an
d th
e lack o f stable
c?m m
unal relationships o p
en to
m igrants
new ly arrived in
the clty
rural villages. T h
e system ic sources of declining value
capab1hbes are exam ined in
m ore detail in chapter 5.
n u
m b
er o f theorists have attributed political violence w
holly or m
p
art to decrem enta! deprivation. T
h e revolutions that A
ris- h
eld to b
e characteristic of dem ocracies an
d oligarchies w
ere sa1d to
result partly from
deprivations of this type. "T h
e cause of revolutions in
D em
ocracies is th e intem
perate conduct of the
dem agogues
w ho force
th e propertied class
to com
bine partly b
y instituting m
alicious prosecutions against individuals ... ?Y inciting th
e m asses against th
em as a body," w
hereas 111 ohgarch1es o
n e of th
e tw o general causes of revolution is seen
as the oppression of th
e m asses b
y th
e oligarchs. 70 In dem
ocracies the relatively high, stable value position of th
e oligarchs is threat-
do 70 A
ristoH e, The
B ook v, chaps. v and V
l. Q uotation from
J. E . C
. W ell-
n, trans., The
P olttics o
f A ristotle
(N ew
Y ork:
M acm
illan, 1883, 1905), 355.
4 7
f : " d: \i
¡ --------------------------------·
,l '
W H
Y M
E N
R E
B E
L
en ed
, in oligarchies
it is the
low , stable value
position of th e
m asses that suffers interference. In
1925 S orokin advanced a gen-
eral thesis about the significance of "repression" in th
e genesis of revolution that influenced a n
u m
b er of later w
riters. T h
e im m
e- diate cause of revolution, h
e w rote, "is alw
ays the grow th of 're-
pression' of th e m
ain instincts of the m ajority of society, and th
e im
possibility of obtaining for those instincts the necessary m íni-
m um
of satisfaction." If th e desire for food of a substantial n
u m
b er
is "rep
ressed " b
y fam
ine, riots
result. If the "reflexes
of indi- vidual self-preservation" are "rep
ressed " by arbitrary executions,
total w ar, or terror, th
e result is the sam e. O
th er rep
ressed in-
stincts w
hich dispose roen tow ard violence are said to include
collective self-preservation,
w ant
of housing, th
e "instincts
of ow
nership" an d
of an
d so forth.
71 T h
e M arxian
view in its original dispensation .is sim
ilar. M arx an
d .E
ngels ar- g
u ed
the inevitable grow th of profound dissatisfactions in
th e pro-
letariat as a co n
seq u
en ce of absolute deprivations or oppressions:
th e destruction· of th
e w
orker' s pride through his subjection to
th e m
achine and the
m arket;
econom ic deprivation b
ecau se of
m inim
al w
ages an d
job insecurity, th e latter a
co n
seq u
en ce o
f crises
in th
e econom ic system
; an
d repressive m
easures of th e
bourgeois state. D
ecrem enta! deprivation is probably m
ost com m
on in "tradi-
tional" societies
an d
in traditional segm
ents of transitional so- cieties. N
atural disasters in traditional societies often gave rise to
collective violence,
as N
orm an
C o
h n
observes in
his study of violent m
illenarianism in
m edieval E
urope:
A gain an
d again o
n e llnds that a particular outbreak of revo-
lutionary chiliasm took place against a background of disaster:
th e plagues that p
relu d
ed th
e F irst C
rusade an d
the flagellant m
ovem ents
of 1260,
1348-1349, 1391,
an d
1400 .
. . . T
h e
greatest w ave of chiliastic excitem
ent, one w hich sw
ep t through
the w hole of society, w
as precipitated b
y th
e m ost universal
natural disaster of th e M
id d
le A ges, th
e B lack D
eath . . . . 72
H obsbaw
m says
that social bariditry w as m
ost pervasive in th e
precapitalist peasant societies of S outhern E
u ro
p e "w
h en
th eir
traditional equilibrium
[ w as] upset; during an
d after periods of
- -
abnorm al hardship, such as fam
ines an d
w ars, or at th
e m om
ents
n P itirim
A . Sorokin, The Sociology ofR
evolution (P hiladelphia: J. B
. L ippincott,
1925), 367-369. 12 N
orm an R
. C . C
ohn, T he P
ursuit o f the M
illenium , 2nd edn. rev. (N
ew Y
ork: H
arper, 1957, 1961), 315. A lso see S
m elser, chap. 6.
4 8
R E
L A
T IV
E D
E P
R IV
A T
IO N
A N
D T
H E
IM P
E T
U S
T O
V IO
L E
N C
E
w h
en the jaw
s of the dynam ic m
o d
ern w
orld seized th
ese static com
m unities in
o rd
er to destroy an d
transform th
em ." 73 S
uch ir- resistible
events inflicted decrem
enta! deprivations on an
d led
to banditry b y
peasants w ho h
ad no effective m
ethods of social agitation,
i.e., no
political value
opportunities. T
h e im
position of alien
authority on non-W
estern p
eo p
les h
ad sim
ilar effects. C
olonialism to varying degrees disrupted interpersonal relation-
ships an
d
u n
d erm
in ed
tribal
an d
o
th er
traditional authority.
Interpersonal an
d
p o
w er
deprivations w
ere inflicted on
m any
new ly co
n q
u ered
p eo
p le at the sam
e tim e that traditional value
opportunities, especially political
ones, w
ere disrupted.
T h
ese conditions
often g
en erated
violent traditionalist
resistance th at
su b
seq u
en tly
w as transform
ed into nationalist rebellion. 74
D ecrem
enta! deprivation m ay b
e less com m
on th an
o th
er form s
of R D
in societies undergoing socioeconom
ic transform ation
b u
t it is
n o
t uncom m
on an d
it can have virulent effects. T h
e hum
an an d
m aterial sacrifices of th
e R ussian p
eo p
le during th e
first W orld W
ar p ro
v id
ed th
e basic potential for th e first R
ussian R
evolution; th e refusal o
f th e K
erensky regim e to term
ínate R us-
sia's involvem ent in
th e w
ar led directly to th
e B olshevik seizure
of p o
w er an
d indirectly to th
e civil w ar th
at follow ed. A
lm ost all
analyses of th e post-w
ar fascist m ovem
ents em phasize th
e abso- lute
deprivations th
at m
otivated m
ost of th eir adherents.
In a
com parative stu
d y
o f fascist m
ovem ents in
riine countries C arsten
concludes that
certain social
groups resp
o n
d ed
m
u ch
m
ore strongly to
th e
F ascist appeal
than others. T his
is particularly tru e of those
w ho
w ere
uprooted an
d
th reaten
ed
b y
social an
d
econom ic
change, w hose position in society w
as b ein
g u
n d
erm in
ed , w
ho h
ad
lost th
eir traditional
place, an
d
w ere frightened
o f the
future. T h
ese w ere above all [segm
ents of] th e lo
w er m
id d
le classes .
. . . P
erhaps ev en
m ore im
portant in th
e early stages w
ere th e form
er officers and non-com m
issioned officers of the first w
orld w ar for w
hom no jobs w
ere w aiting, w
ho h ad
got accustom
ed to th e use of violence, an
d felt them
selves d ep
riv ed
of th
eir "leg itim
ate" rew ards. 75
73 E . J. H
obsbaw m
, Social B andits and P
rim itive R
ebels: Studies inA rchaic F
orm s
o f Social M
ovem ent in the 19th and20th C
enturies (G lencoe: T
he F ree P
ress 1959) 23-24.
• •
74 S ee, for exam
ple, W illiam
K ornhauser, "R
ebellion and P olitical D
evelopm ent,"
in H arry E
ckstein, ed., Intem al W
ar: P
roblem s and A
pproaches (N ew
Y ork: T
he F
ree P ress, 1964), 145.
75 F . L. C
arsten, T he R
ise o f F
ascism (B
erkeley: U niversity of C
alifornia P ress,
1967}, 232.
4 9
'1
W H
Y M
E N
R E
B E
L
C antril's social psychological analysis of th
e roots of N azism
is sim
ilar. N azism
w as capable of succeeding as a m
o v
em en
t b ecau
se "o
ld norm
s, o ld
cultural standards, w ere no longer able to pro-
vide th e fram
ew ork necessary for a
adjustm ent of th
e individuals w
ho com posed the culture." D
ecrem enta! deprivation
w as
intense for a variety of groups on m any values; there w
as econom
ic distress, m any people h
ad suffered status reductions,
an d
m ost felt a profound sense of personal insecurity b
ecau se of
th e
disintegration of b
o th
social
structures an
d norm
ative sys- tem
s.76 A
ny absolute decline in th
e value position or value potential of a social group constitutes decrem
enta! deprivation, an d
m any
additional cases of political violence attributable to this pattern of R
D could b
e cited. S orne othyrs are exam
ined in chapter 5. D
e- spite the considerable em
phasis given in contem
porary social anal- ysis to "revolutions of rising expectations," exam
ined below , over
the long run of h u
m an
history decrem enta! R
D has probably b
een
a m ore com
m on source of collective violence th'an auy other p
attern
of R D
. A nd o
n e can speculate th
at decrem enta! R
D of a given de-
gree probably
instigates roen to
greater in ten
sities o_f violence th
an an eq
u iv
alen t level of aspirational R
D . M
en are h
k ely
to b e
m ore intensely an
g ered
w h
en they lose w
hat th ey
h a ve than w
h en
th
ey lose h
o p
e of attaining w hat th
ey do n
o t yet have.
A S
P IR
A T
IO N
A L
D E
P R
IV A
T IO
N
T h
e aspirational R D
m odel, sk
etch ed
in figure 2, is character-
ized b
y an increase in
m en
's value expectations w ithout a con-
com itant change in
value position or potential. T hose w
ho experi- en
ce aspirational R D
do n o
t anticípate or experience significant loss of w
hat th ey
have; th ey
are an g
ered b
ecau se th
ey feel th
ey
have no m eans for attaining n
ew or intensified expectations. A
n "in
crease" in value expectations m
ay reflect dem ands for a greater
am ount of a value already h
eld in sorne degree, for exam
ple for m
ore m aterial goods an
d a greater degree of political order an
d
justice. It m ay b
e a d em
an d
for n ew
values n ev
er previously h eld
, such as political participation for i!O
lonial peoples an d
personal equality for m
em bers of lo
w er cla.ss an
d caste groups. T
h ird
, it
'" H adley C
antril, The P sychology o
f Social M ovem
ents (N ew
Y ork:
1941), 228-269, quotation from
266. A
lso see F rederick L
. S chum
an, T he Na;ra D
Jctator- ship: A
Study in Social P athology and the P
olitics o f F
ascism (N
ew Y
ork: K n_opf,
1935). Individual docum entation of th
e nature and severity of these frustrations is provided by the autobiographical essays of N
azi P arty m
em bers w
hich are ana- lyzed by T
heodore A bel, The N
azi M ovem
ent: W hy H
itler C arne to P
ow er (N
ew
Y ork: A
therton P ress, 1938, 1966).
50
R E
L A
T IV
E D
E P
R IV
A T
IO N
A N
D T
H E
IM P
E T
U S
T O
V IO
L E
N C
E
high
C ollective
value position
/ /
/
/
/ /
/
value ex¡;ectations /
/ .,..--
/ /
_ ..... ...- value capabilities
low
T im
e -+
F igure 2. A
spirational deprivation.
m ay rep
resen t intensification of th
e com m
itm ent to (salience of)
a value position that earlier w as w
eakly sought, for exam ple in-
tensifying dem ands for w
elfare goods am ong those w
ho experience breakdow
n of com m
unal life during th e early stages of m
oderni- zation, an
d intensified dem
ands for access to political elite posi- tions
am ong th
e upw
ardly m obile
bourgeoisie of seventeenth-
an d
eighteenth-century E urope. D
egrees an d
saliences of value expectations q
u ite often increase sim
ultaneously, b u
t no neces- sary relation holds b
etw een
th e tw
o. F o
r exam ple, th
e contem po-
rary dem
ands of black A m
ericans for social equality com
pared w
ith d em
an d
s m ade in
th e 1940s appear to reflect an increase in
the salience of status values am
ong N egroes, a shift from
faintly h
eld aspirations for eq
u ality
to an intensely h eld
b elief that equal-
ity is
d eserv
ed now
, rather than an increase in
th e degree of
equality sought. M
any sources of increasing value expectations are identified in
studies of R
D
an d
of political instability. F
o r sorne traditional
peoples m ere exposure to, or know
ledge of, a b etter m
aterial w ay
of life is assum
ed to raise expectations. In m
edieval an d
early R
enaissance E u
ro p
e th e grow
th of industrial an d
com m
ercial cen- ters dem
onstrated n ew
possibilities b ey
o n
d any that life h
ad to
.offer th e peasant. T
h e n
ew w
ays attracted the surplus population in particular, b
u t also those w
ho w ere in
sorne w ay dissatisfied
w ith
m anor life.
"A s
social an
d econom
ic horizons
expanded, hardship an
d poverty an
d d
ep en
d en
ce ceased to appear th
e in- eluctable fate of com
m on folk." 77 A
special case of th e dem
onstra- 77 C
ohn, 27-28. A lso see N
eil J. S m
elser, Social C hange in the Industrial R
evolu- tion (C
hicago: U niversity of C
hicago P ress, 1959).
51
W H
Y M
E N
R E
B E
L
tion effect is
"relative deprivation" in its narrow
sense, th at is,
setting one' s value expectations by reference to th e h
ig h
er value position of sorne o
th er individual or group. In
particular, expec- tation levels are often accelerated b
y the dem
onstration effect of other groups
th at are
im proving w
hile one's ow n group is not.
B rogan observed that the n
ew w
ays an d
n ew
w ealth of the Indus-
trial R
evolution im
p elled
m
any intellectuals
to revolutionary
fervor. M
en like F ried
rich E
ngles "w ere struck b
y the paradox
that th e m
eans of w ealth w
ere vastly increasing an d
that the b en
e- fits
of that w ealth seem
ed to b
e m ore and m
ore narrow ly be-
stow ed." 78 D
iscrepancy b etw
een an individual's or group's rela-
tive share
of w elfare, .PO
W er
an d
interpersonal
also is
specifically related to political violence b y
a n u
m b
er of theorists, am
ong th em
A ristotle, w
ho w rote th
at the source of disposition to revolution
is the
aspiration after equality w hich provokes
the com m
ons to sedition w
h en
th ey
suppose th at th
ey have a
sm all share
. . . although they are th
e equals of the privileged F ew
, an d
it is the aspiration after inequality or in
o th
er w ords after su-
periority w h
ich provokes th
e O ligarchs to sedition, w
h en
th ey
im
agine th at d
esp ite th
eir inequality th eir share is n
o t greater
than that of others b u
t is equal or ev en
sm aller. 79
T h
ese and o th
er conditions that increase expectation levels be- yond the capacities of m
en to satisfy them
, an d
h en
ce dispose m
en to collective violence, are exam
ined m ore closely in chapter
5. S orne com
parable effects of n ew
beliefs an d
ideologies are con- sidered there an
d in
ch ap
ter 7.
P R
O G
R E
S S
IV E
D E
P R
IV A
T IO
N
T h
e third pattern o f R
D , sk
etch ed
in figure 3, is a generalized
version of a m odel proposed by D
avies, w ho refers to it as the
"J-curve" hypothesis: "revolutions are m ost likely to occur w
hen a prolonged period of objective econom
ic an d
social developm ent
is follow ed b
y a short p
erio d
of sharp reversal." 80 It can b e re-
garded as a special case of aspirational R
D , one in
w hich long-
ron, m ore-or-less-steady im
p ro
v em
en t in
peoples' value position generates
expectations ab
o u
t continued im
provem ent. If value
capabilities stabilize or decline after such a period of im prove-
. ·
78 D enis W
. B rogan, T
he P rice o
fR evo
lu tio
n (L
ondon: H am
ish H am
ilton, 1951), 79 T
he P
olitics, W elldon trans., 343-344. F
or a restatem ent of the
position see F red
K ort, "T
h e Q
uantification of A ristotle's T
heory of H evolutí•on,'
A m
erican P olitical Science R
eview , L
X V
I (June 1952), 487. 80 D
avies, "T ow
ard a T heory of R
evolution," 6.
52
high
C ollective
value position
low
R E
L A
T IV
E D
E P
R IV
A T
IO N
A N
D T
H E
IM P
E T
U S
T O
V IO
L E
N C
E
/ /
/ /
/
value exp ectation
s /,.,...,..
/ /
/
/ /
/
------ /
/ value capabilities
/ /
/ /
T im
e -+
·F igure 3. P
rogressive deprivation.
m ent, progressive R
D is th
e resu lt S
uch a pattern is m ost com
m on
in societies
undergoing sim ultaneous
ideological an
d system
ic change, E
conom ic depression in
a grow ing econom
y can have this effect. S
o can the articulation o f an ideology of m
odernization in a society that has
structural inflexibilities .that p rev
en t expansion
of value o u
tp u
t b ey
o n
d a certain point. T
h e m
odel also can b e
used to
su b
su m
e sorne
"social ch an
g e" theories
of revolution w
hich in th
eir general form postulate that political violence is
co n
seq u
en ce
o f decreasing responsiveness
of social structures
beliefs, norm s, or all three to objective change.
' " D
avies _em phasizes that th
e revolutionary state o f m
in d
requires the
even habitual b
u t dynam
ic expectation of greater opportum
ty to
satisfy basic n eed
s," by
w hich
h e
includes all
types o f values, physical, social, an
d political. In
addition, w hat
is n eed
ed
is a persistent, u
n relen
tin g
threat to th
e satisfaction of these needs: n
o t a th
reat w hich actually returns p
eo p
le to a state o f
sh eer survival b
u t w
hich puts th em
in th
e m ental state w
here they b
eliev e th
ey w
ill n o
t b e able to satisfy one or m
ore basic n
eed s .
. . . T
h e crucial factor is the vague or specific fear that
ground gained over a long p erio
d o
f tim e w
ill b e quickly los t.
political system is p
ercep tu
ally related
to th ese fears; th
ey
generated "w h
en th
e existing g o
v ern
m en
t suppresses or is for su
p p
ressin g
such opportunity." 81 In su
p p
o rt of the the-
53
1 • 'f ':1 '
' '
'1¡ 1: .¡J
W H
Y M
E N
H .E
B E
L
sis D
avies identifies a }-curve p
attern of progress follow
ed b y
decline in case studies of the preconditions of a n
u m
b er
of revolutions an d
rebellions, including the F ren
ch , R
ussian, an d
N
azi R evolution
th e A
m erican C
ivil W ar, an
d th
e E gyptian R
evo- lution of 1952. F
o r exam
ple, h e show
s th at D
orr's R ebellion in
R
hode Islan d
in 1842 occurred after a forty-year p
erio d
of im prov-
ing econom ic conditions
an d
expanding political nom
ic depression in 1835-40 an
d the rejection by an ohgarch1c
state governm ent of d
em an
d s for further extension of suffrage
to th
e drafting
of a P
eo p
le' s C
onstitution, an attem pt to se1ze
state buildings, an d
sporadic violence. T h
ere also is ev id
en ce _of a
progressi,),¡e deprivation p attern
in the econom
ic statu_s of N
egroes relat_ive to w hites in
th e tw
o decades th
e reb
ellio n
" of the 1960s. T h
e incom e of N
egroes relatlve to w httes
of com parable education
rapidly tow ards equality be-
tw een
1940 an d
the early 1950s b u
t th en
b eg
an to decline, so that
by 1960 half the relative gains of th e earlier period
In this study's term
inology, b o
th cases w
ere b
y n
s- ing value expectations, set in
m otion b
y
?f im
proving value positions. D
im inishing capabilthes, ev
td e':t m
politicians' reluctance to ex
ten d
political rights and_ irrem
ediable econom
ic decline, p ro
v id
ed th
e background condttions necessary for th
e outbreak of violence. L
e V ine's explanation of m
ob violence by A frican colonial pea-
pies agai'ust th eir rulers
resem bles th
e }-curve S
u ch
turm
oil is said to b e a co
n seq
u en
ce of psychologtcal confliC t gen-
erated w
h en
colonial
officials encouraged expectations
of rule by th
eir stated policies an
d practices, b
u t th
en acco
m p
am ed
or follow
ed those policies w ith others w
hich, A fricans thought,
interfered th
e attaim
nent of those expectations.
L e
V ine
describes sev
en cases
th at appear to
support th e hypothesiS
. 83
T h
e }-curve hypothesis is also im
plicit in D
eu tsch
's tion
of th e
relationship b
etw een
governm
ental capabthty
an d
poli ti cal stability in
societies in th
e early an d
m _iddle stages of m
od- ernization. ln
creasin g
capabilities,
as an obJect of governm
ent policy, require increasing m
obilization of for
in
th e
m arket econom
y an
d
political life.
S uch m
obthzatw n,
•• C ase studies of th
e R ussian an
d E
gyptian R evolutions and D
orr's R ebellion
appear ibid. T h
e other studies are in clu
d ed
in Jam
es C . D
avies, "T h
e !-C u
rv e of
R ising an
d D
eclining S atisfactions as a C
ause of sorne G reat R
evolutw ns. an
d a
C o
n tain
ed R
ebellion," in H
u g
h D
avis G raham
an d
T ed
R obert G
urr, eds., V1ol_ence in A
m erica: H
istorical and C om
parative P erspectives (W
ashington, D .C
.: N atw
nal C
om m
ission o n
the C auses an
d P
revention of V iolence, 1969), 547-576.
sa R obert A
. L e V
ine, "A nti-E
uropean V iolence in
A frica: A
C om
parative A naly·
sis," ]o u
m a
l o f C
onjlict R esolution, lll (D
ecem b
er 1959), 420-429.
5 4
H .E
LA T
JV E
D E
P R
IV A
T IO
N A
N D
T H
E IM
P E
T U
S T
O V
IO LE
N C
E
ho":ever, is usually accom panied by rising needs an
d expectations,
w hiC
h m ust not b
e frustrated if stability is to b e preserved." T
o preserve stab
ility -that is, to m
inim ize progressive R
D an
d the
co n
seq u
en t ím
petus to
v io
lén ce-outputs
m u
st continue to in- crease. D
eu tsch
cites only tw o types of value outputs, increasing
per capita incom e an
d expanding governm
ental activity, b u
t his argum
ent is readily ex ten
d ed
to o th
er value classes.B 4
P rogressive R
D is a com
m on th
em e in m
any old and sorne n ew
theories
th at attribute
revolutionary potential
to general social
change. S im
ple versions of th ese theories em
phasize structural in- flexibility, i.e. the inability of social an
d political institutions to
adapt th eir value outputs rapidly en
o u
g h
to changing conditions. Y
oder propases, for exam ple, that change in so
ciallife is and th
at the organized group m ust an
d ordinarily does co
n stan
tl; readjust to changing situations, resulting from
inventions, discov- eries, culture contact, an
d so forth. In
sorne cases, how ever, groups
w ithin th
e society desire to preserve the old o rd
e r-
"th e traditional
institutions, th e tim
e-tried folkw ays, m
ores, conventions an
d cus-·
to m
s"- ev
en though irrelevant to p
resen t situations. T
h e en
d of
adaptation is the
b eg
in n
in g
of revolution.
85 Johnson's "social-
dysfunction" theory of the origins of political violence is of the sam
e genre, though its vocabulary has a m ore contem
porary ring. O
ne necessary condition of revolution is said to b e a disequilibrated
social system , i.e. a discrepancy b
etw een
the b elief structure of a
society an d
its division of labor, w hich m
ay result from any com
bi- nation of interna! or externa! changes in
values or technology. T
he seco n
d necessary cause is elite refusal ("intransigence") to
take action d esig
n ed
to relieve the disequilibrium
. T h
e conse- quence is a loss of authority by the elite an
d its reliance u p
o n
force to m
aintain its position. T h
e sufficient cause of revolution in such
a situation is an "accelerator o f dysfunction," any condition that
decreases the
ability of the elite to control its
arm ed
forces. 86
T h
ese an d
o th
er social change theories refer to system s that w
ere able to adapt value outputs to changing environm
ental require-
• 84 K arl D
eutsch, "T o
w ard
an Inventory o f B
asic T rends an
d P
atterns in C
om para-
ti ve an d
International P olitics," A
m erican P
olítica! Science R eview
, X
IV (M
arch 1960), 39 .
.., D ale Y
oder, "C u
rren t D
efinitions o f R
evolution," A m
erican ]ournal o f Sociol-
ogy, X X
X II (N
ovem ber 1926), 440-442. C
om parable explanations are proposed by
A . E
llw ood, T
he P sychology o
f H um
an Society: A n
Introduction to Socio- l?glcal T
heory (N ew
Y ork: A
ppleton, 1925), chap. 8; R ex H
opper, "T h
e R evolu-
hon_ary P rocess: A
F ram
e of R eference for th
e S tudy o
f R evolutionary M
ovem ents,"
F orces, X
X V
III (_M arch 1950), 270-27_1; and P
ettee, passim .
Johnson, R evolutw
nary C hange, passtm
. A n earlier an
d briefer statem
ent o f
t?e theory is his R evolution and the Social System
(S tanford: T
h e H
oover Institu- tion on W
ar, R evolution an
d P
eace, S tanford U
niversity, 1964).
5 5
W H
Y M
E N
R E
B E
L
m ents and to m
en 's changing value expectations u
p to sorne p
o in
t in
tim e, b
u t gradually or abruptly lost th
eir adaptive capacities. societies that have ex
p erien
ced positive social change, th
e condi- tions
described in th
ese theories fit
th e }-curve or progressive
deprivation m odel; in
static societies, they m ore closely resem
b le
th e absolute deprivation m
odel. T h
e point is that th e R
D m
odels correspond generally to m
any of the abstract theories of th e pre-
conditions of political violence.
T h
e R D
M o
d els: S
orne Q u
alificatio n
s an d
U ses
N o one of the R
D m
odels is necessarily characteristic of a par- ticular type of society, although absolute R
D seem
s m ore li.kely to
b e found in
static societies, w h
ile aspirational an d
progress1ve R D
ap
p ear m
ost often in societies, undergoing substantial socioeco-
nom ic change. In
any given society at any given tim e, how
ever, sorne groups are likely to experience R
D o
f each type. M oreover,
sorne groups can ex p
erien ce different patterns of R
D w
ith respect to different classes of values.· F
o r exam
ple, large segm ents of th
e w
orking classes of the d efeated
E u
ro p
ean countries
W ar I sim
ultaneously ex p
erien ced
decrem enta! d
ep n
v atw
n w
lth respect to security an
d participation values, an
d progressive dep-
rivation on w elfare values. A
com plete profile of R
D in
a society requires specification of th
e extent an d
patterns o f R
.D w
ith spect to each class of values for every consequential socw
econom 1c
group. . .
O n
e operational u se of th
e R D
m odels is th
at th ey
tem
atic inferences ab o
u t the effects of changes in
vanous m diC
es of econom
ic an d
política! perform ance. Indicators of econom
ic per- form
ance an
d of governm
ental fiscal activity provide exam
ples. R
ather than resorting to
inferences about econom
ic discontent b
ased on such m
easures as p er capita incom
e in country X
coro- p
ared w
ith country Y ,
or th eir relative rates of grow
th, one .can infer it m
ore accurately from such patterns as short-term
declm es
in
productivity follow ing a
period of stable production m
ental R D
) and short-term changes in inflation rates, co
m m
?d ü
y
prices, or total productivity relative to rates in th
e m ore d1stant
past (decrem enta! or progressive R
D ). T
im e-series w
age an d
salary data for various occupational groups are available for m
any m odern
an d
sorne m odernizing societies, from
w hich w
elfare an d
in sorne
instances status R D
can b e inferred. O
n e also can search system
- atically for ev
id en
ce of conditions associated w ith rising expecta-
tions such as rates of increase in
schooling and literacy, announce-, m
en ; of program
s of reform , articulation b
y political
of idéologies of m
odernization, an d
m obilization of previously non-
5 6
¡ 1 1 ¡ i 1 ¡ i ¡ 1 r t r l.
R E
L A
T IV
E D
E P
R IV
A T
IO N
A N
D T
H E
IM P
E T
U S
T O
V IO
L E
N C
E
participant citizens for .purposes of política] an d
associational ac- tivity. T
o assess the effects o f governm
ental action one can exam ine
the changing
balance over
tim e
b etw
een
value-depriving and value-satisfying decisions, giving special attention to relative in- creases in
th e form
er. Interpersonal R D
of various patterns al so can be inferred from
changes over tim e in
m easures of condÍtions as di-
verse as interna! m igration, religious affiliation, changing size and
com position
of occupational groups,
an d
social origins of elite groups. P
rocedures for m aking system
atic, com parative inferences
of these kinds are not exam ined h
e re; exam ples o
f sorne o f them
are reported in o
th er p
u b
lish ed
studies. 87 r
th ree
o f R
D proposed do n
o t exhaust all the logically
valu_e expectations and value capa- bdities.
D eclm
m g
expectations m 1ght b
e found in sorne groups, for exam
ple am ong m
em bers o
f a w ell-disciplined colonial elite
w ho expect gradually to lose th
eir political authority to indigenous rulers, or am
ong m em
bers of a m illenarian religious group w
ho ex- pect the social o
rd er to deteriorate and collapse as a p
relu d
e to th e
com ing o
f a n ew
order. S om
ew hat m
ore com m
on m ay b
e a "boom
and b u
st" set o f expectations about value satisfactions, for exam
ple am
ong sorne traditional societies, u n
d er sorne totalitarian regim
es, and p
erh ap
s am ong sorne groups an
d isolated individuals in m
od- ernized societies. R
esig n
ed anticipation o
f a "tim
e o f troubles"
follow ing the death o
f a king w as as com
m on in
feudal kingdom s of
tropical A frica as it w
as in m
edieval E u
ro p
e. T hose w
hose liveli- hood d
ep en
d s on the land ten
d to develop a sim
ilar tolerance for the vagaries o
f w eather: tw
o years o f p
len ty
an d
one o f scarcity
is the w ay o
f th e seasons. T
h e B
em ba o
f S outh C
entral A frica, like
m any o
th er p
eo p
le living n ear the m
argin o f subsistence, b
u ild
their w
ay of life around a cycle of n in
e m onths o
f abundance fol- low
ed_ by a h u
n g
er season o f three m
onths. 89 B u
t in all such cycles
there IS an elem en
t of predictability, an d
if this predictability van- ishes R
D is th
e likely result. If, for exam ple, conditions in
such a 87 S e e T
ed G
urr, N ew
E rror-C
om pensated M
e asures for C om
pari ng N ations: S o m
e C
or:elat.es o
f C ivil Strife, C
enter of International S tudies (P
rinceton: P rinceton
U m
versity, R esearch M
onograph N o. 25, 1966); T
ed G
urr w ith C
harles R uttenberg
C onditio'!s o
f C _ivil V
iolence: F irst T
ests o f a C
ausal M odel, C
enter of n_ahonal
Stud1es (P nnc;;ton: P
rinceton U niversity, R
esearch M onograph N
o. 28, and T
ed .
A C
:ausal M
odel of C ivil Strife: A
C om
parative A nalysis
U sm
g N ew
Indices, A
m en
ca n
P olítica[ Science R
eview
L X
I (D ecem
ber 1968) 1104-1124.
, '
88 A m
ore c?m prehensive set of m
odels of social change, including variations on three basiC
m odels proposed here, appears in Ivo K. F
eierabend, R osdind L
. and
A . N
esvold, "S ocial C
hange and P olitical V
iolence: C ross-
N !!IO
nal P atterns,
G raham
and G urr, eds., 497-542.
S ee
R IC
hards, L and, L
abour, and D iet inN
orthern R hodesia (L
ondon: O
xford U m
versity P ress, 1939), especially chaps. 1-3.
57
W H
Y M
EN R
E B
E L
.
b d
th
su lt is decrem
enta! R D
; a: a'
;:v e l substantially b
ey o
n d
an
d ¡f va u
e post w ns s
te clical decline, any su
b seq
u en
t de- the expected o
n set o{ th
e nex th
e potential for violent protest. cline
t bet of th
e m odels it w
as suggested ru
n m
en 's value
ten d
th eir value
m en
pectations or d
eclm m
g
1 .
g th eir value capabilities to
collectively eith er .su
ccee. m
prove im pervious
m eet th
eir expectatw ns or, ¡f
c u
F
. tensely felt expecta-
1 1
er th eir expectatw
ns. or m
d
to e Iange,
ow "
ll " b
e m easu
red in
years, d eca es,
. h
er ev
en tu
a y
m ay
b h
tw ns,
ow ev
, . h. h are likely to b
e characterized b y
ot
or ev en
generahons, w
. l
S orne of th
e variables that constructive
eical m odes of response to it
affect the persistence of R
h typ
If any single sen ten
ce can .
d · b
seq u
en t e
ap ers. are
m
su advanced in
this chapter, it is th at m
en
su m
m an
ze the argum ents
. . 1
s an d
quick to anger are quick to aspire
tldletqr to accept their lim
i- w
h en
those m eans prove n
m e
' tations.
5 8
3 . T
h e In
ten sity
a n
d S
eo p
e o f
R ela
tiv e D
ep riv
a tio
n .
P eo
p le foreign to our land an
d race m
ade u
p th
e governm ent; th
e m iddle class w
as ru
in ed
through th e scarcity of food an
d
the depreciation
o f m
oney; scoundrels
an d
parasites cheated and robbed us, an d
in an incredibly brief tim
e ru in
ed under-
takings it h ad
taken a w hole people cen-
turies to build.
P eople lacked th
e very essentials
of living .... H aving felt th
e results
of th e econom
ic collapse on m y
ow n pulse, 1 w
as only too happy to take m
y place in th
e van o f th
e m ovem
ent.
A N
azi P arty m
em b
er,1 9
3 3
1
T H
E IN
TEN SITY
of our anger at its onset is a function of four psycho- cultural variables. T
h e greater the discrepancy w
e see b etw
een
our expectations an d
capabilities, th e greater is our discontent.
T he greater th
e im porta.rice w
e attach to the values affected, an d
the
few er
th e
o th
er satisfactions w
e have to fall back on, th
e greater is o
u r discontent. If w
e have m any alternative w
ays of ti·y- ing to satisfy our expectations, w
e are likely to defer discontent over o
u r failures; if w
e h av
e few alternatives w
e are likely to feel the an
g er of desperation. A
fifth d eterm
in an
t is tim e: if our anger
is d en
ied expression in the short ru
n it intensifies before it sub-
sides. T h
e greater its intensity, th e longer it persists; m
any m en
carry the b
u rd
en of profound grievances throughout th
eir lives an d
pass th
em o
n to their children. T
h e first four variables resem
ble . but are not m
erely collective m anifestations of the three variables
that D ollard said d
eterm in
e the strength of instigation to aggression (anger) in
th e frustration-aggression sequence: the strength of in-
stigation to th e frustrated response (salience), the degree of inter-
ference w ith th
e frustrated response (degree), an d
th e n
u m
b er of
frustrated response-sequences (alternative values an d
opportuni- ties).2 R
uncim an specifies tw
o com parable variables as sources o
f variance in
R D
. "T h
e m agnitude o
f a relative deprivation is th e
extent of th e difference b
etw een
the desired situation an d
that of the ·person desiring it (as h
e sees it) .... T h
e degree of a relative
1 F rom
an autobiogniphical essay w ritten in 1933, cited
in T heodore A
bel, T he
N azi
M ovem
ent: W hy H
itler C arne to P
ow er (N
ew Y
ork: A therton P
ress, 1938, 1966), 125-126.
2 John D ollard and others, F
rustration and A ggression (N
ew H
aven: Y ale U
ni- versity P
ress, 1939), 33.
5 9
. ' 1