Sage Publications, Ltd. is collaborating with JSTOR to digitize, preserve and extend access to Journal of Peace Research.
http://www.jstor.org
Violence, Peace, and Peace Research
Author(s): Johan Galtung
Source: Journal of Peace Research, Vol. 6, No. 3 (1969), pp. 167-191
Published by: Sage Publications, Ltd.
Stable URL: http://www.jstor.org/stable/422690
Accessed: 07-04-2015 01:57 UTC
Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at
http://www.jstor.org/page/info/about/policies/terms.jsp
JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content
in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship.
For more information about JSTOR, please contact support@jstor.org.
This content downloaded from 133.30.212.88 on Tue, 07 Apr 2015 01:57:19 UTC
All use subject to JSTOR Terms and Conditions
VIOLENCE, PEACE, AND PEACE RESEARCH*
By
JOHAN GALTUNG
International Peace Research Institute, Oslo
1. Introduction
In the presentpaper we shall be using the
word 'peace'very many times. Few words
are so often used and abused- perhaps,it
seems, because 'peace' serves as a means
of obtainingverbalconsensus-it ishardto
be all-out against peace.l Thus, when effortsaremade
to plead almost any kindof
policy- say technicalassistance,increased
trade,tourism,new formsof education, irrigation,
industrialization, etc.- then it is
often assertedthat that policy, in addition
to other merits,will also servethecauseof
peace. This is done regardless of how
tenuous the relation has been in the past
or how dubious the theory justifying this
as a reasonableexpectation for the future.
Such difficultiesare avoided by excluding
any referenceto data from the past or to
theories about the future.
This practice is not necessarily harmful.
The use of the term 'peace' may
in itself be peace-productive, producing
a common basis, a feeling of communality
in purpose that may pave the
ground for deeper ties later on. The use
of more precise terms drawn from the
vocabulary of one conflict group, and
excluded from the vocabulary of the
opponent group, may in itself cause
dissent and lead to manifest conflict
precisely because the term is so clearly
understood. By projecting an image of
harmony of interests the term 'peace'
may also help bring about such a harmony.
It provides opponents with a
one-word language in which to express
values of concern and togetherness because
peace is on anybody's agenda.2
One may object that frequent use of
the word 'peace' gives an unrealistic
image of the world. Expressions like
'violence', 'strife', 'exploitation' or at
least 'conflict', 'revolution' and war
should gain much higher frequency to
mirror semantically a basically non-harmonious
world. But leaving this major
argument aside for the moment, it is obvious
that some level of precision is necessary
for the term to serve as a cognitive
tool. At this point, of course, nobody has
any monopoly on defining 'peace'. But
those who use the term frequently in a
research context, as peace researchers
(will do) do, will at least havegained some
experience when it comes to definitions
that should be avoidedfor one reason or
another.
To discuss the idea of peace we shall
start from three simple principles:
1. The term 'peace' shall be used for social
goals at least verballyagreedto by many, if
not necessarilyby most.
2. Thesesocial goals maybe complexand difficult,
but not impossible, to attain.
3. The statementpeace is absence of violence
shall be retainedas valid.
The third principle is not a definition,
since it is a clear case of obscurumper
obscurius. What we intend is only that
This content downloaded from 133.30.212.88 on Tue, 07 Apr 2015 01:57:19 UTC
All use subject to JSTOR Terms and Conditions
168 Johan Galtung
the terms 'peace' and 'violence' be linked
to each other such that 'peace' can be
regarded as 'absence of violence'. The
reasons at this early point in our semantical
excursion, are twofold: the statement
is simple and in agreement with
common usage, and defines a peaceful
social ordernot as a point but as region as
the vast region of social orders from
which violence is absent. Within this region
a tremendousamount of variation is
still possible, making an orientation in
favor of peace compatible with a number
of ideologies outlining other aspects of
social orders.
Everything now hinges on making a
definition of 'violence'. This is a highly
unenviable task, and the suggestions will
hardly be satisfactory to many readers.
However, it is not so important to arrive
at anything like the definition, or the
typology - for there are obviously many
typesof violence. More importantis to indicate
theoreticallysignificantdimensions
of violence that can lead thinking, research
and, potentially, action, towards
the most important problems. If peace
action is to be regardedhighly because it
is action against violence, then the concept
of violence must be broad enough to
include the most significantvarieties, yet
specific enough to serve as a basis for
concrete action.
Thus, the definition of 'peace'becomes
a major part of a scientific strategy.
It may depart from common usage by
not being agreed to 'by most' (consensus
not required),yet should not be entirely
subjectivistic ('agreed to by many'). It
should depict a state of affairsthe realization
of which is not utopian ('not impossdile
to obtain'), yet not on the immediate
political agenda ('complex and
bifficult'). And it should immediately
steer one's attention towards problems
that are on the political, intellectual, and
scientific agenda of today, and to-
morrow.2
2. On the definition and dimensions of
'violence'
As a point of departure,let us say that
violenceis present whenhumanbeingsare
being influencedso that theiractualsomatic
and mental realizations are below
theirpotentialrealizations.This statement
may lead to moreproblemsthanit solves.
However, it will soon be clear why we
are rejecting the narrow concept of
violence - according to which violence is
somatic incapacitation, or deprivation of
health, alone (with killing as the extreme
form), at the hands of an actor who intends
this to be the consequence. If this
were all violence is about, and peace is
seen as its negation, then too little is
rejected when peace is held up as an
ideal. Highly unacceptable social orders
would still be compatible with peace.
Hence, an extended concept of violence
is indispensablebut that concept should
be a logical extension, not merely a list
of undesirables.
The definition points to at least six important
dimensions of violence. But first
some remarks about the use of the key
words above, 'actual' and 'potential'.
Violence is here definedas the cause of
the difference between the potential and
the actual, between what could have been
and what is. Violence is that which increases
the distancebetween the potential
and the actual, and that which impedes
the decrease of this distance. Thus, if a
person died from tuberculosis in the
eighteenth century it would be hard to
conceive of this as violence since it might
have been quite unavoidable, but if he
dies from it today, despite all the medical
resources in the world, then violence is
present according to our definition. Correspondingly,
the case of people dying
from earthquakes today would not warrant
an analysis in terms of violence,3but
the day after tomorrow, when earthquakes
may become avoidable, such
deaths may be seen as the result of voiThis
content downloaded from 133.30.212.88 on Tue, 07 Apr 2015 01:57:19 UTC
All use subject to JSTOR Terms and Conditions
Violence, Peace, and Peace Research 169
lence. In other words, when the potential
is higher than the actual is by definition
avoidableand when it is avoidable, then
violence is present.
When the actual is unavoidable, then
violence is not present even if the actual
is at a very low level. A life expectancy of
thirty years only, during the neolithic
period, was not an expression of violence,
but the same life-expectancy today
(whether due to wars, or social injustice,
or both) would be seen as violence according
to our definition.
Thus,the potentiallevel of realizationis that
which is possible with a given level of insight
and resources.If insight and/or resources are
monopolized by a group or class or are used for
otherpurposes,thenthe actual level falls below
thepotentiallevel, andviolenceis presentin the
system.In additionto thesetypesof indirectviolence
there is also the direct violence where
means of realization are not withheld, but directly
destroyed. Thus, when a war is fought
thereis directviolencesincekillingor hurtinga
personcertainlyputs his 'actual somatic realization'
below his 'potential somatic realization'.
But there is also indirectviolence insofar
as insight and resources are channelled away
from constructive efforts to bring the actual
closer to the potential.4
The meaning of 'potential realizations'
is highly problematic, especially when we
move from somatic aspects of human
life, where consensus is more readily
obtained5, to mental aspects. Our guide
here would probably often have to be
whether the value to be realized is fairly
consensual or not, although this is by no
means satisfactory. For example, literacy
is held in high regardalmost everywhere,
whereas the value of being Christian is
highly controversial. Hence, we would
talk about violence if the level of literacy
is lower than what it could have been,
not if the level of Christianity is lower
than what it could have been. We shall
not try to explore this difficult point further
in this context, but turn to the dimensions
of violence.
To discuss them, it is useful to conceive
of violence in terms of influence, as indicated
in the statement we used as a
point of departure above. A complete
influence relation presupposes an influencer,
an influencee, and a mode of influencing.6
In the case of persons, we can
put it very simply: a subject, an object,
and an action. But this conception of
violence in terms of a complete interpersonal
influence relation will lead us
astray by focussing on a very special
type of violence only; also truncated
versions where either subject or object
or both are absent are highly significant.
To approach this we shall start
with two dimensions characterizing the
violent action itself, or the mode of in-
fluence.
The first distinction to be made is
between physical and psychological violence.
The distinction is trite but important
mainly because the narrow concept
of violence mentioned above concentrates
on physical violence only. Under
physical violence human beings are hurt
somatically, to the point of killing. It
is useful to distinguish further between
'biological violence', which reduces somatic
capability (below what is potentially
possible), and 'physical violence as
such', which increases the constraint on
human movements7- as when a person is
imprisonedorputin chains,butalso when
access to transportation is very unevenly
distributed, keeping large segments of a
population at the same place with mobility
a monopoly of the selected few. But
that distinction is less important than the
basic distinction between violence that
works on the body, and violence that
works on the soul; wherethe latterwould
include lies, brainwashing,indoctrination
of various kinds, threats, etc. that serve
to decrease mental potentialities. (Incidentally,
it is interesting that such English
words as 'hurt' and 'hit' can be used
to express psychological as well as physThis
content downloaded from 133.30.212.88 on Tue, 07 Apr 2015 01:57:19 UTC
All use subject to JSTOR Terms and Conditions
170 Johan Galtung
ical violence: this doubleness is already
built into the language.)
The second distinction is between the
negative and positive approach to influence.8
Thus, a person can be influenced
not only by punishing him when he does
what the influencerconsiders wrong, but
also by rewarding him when he does
what the influencer considers right. Instead
of increasing the constraints on his
movements the constraints may be decreasedinstead
of increased,and somatic
capabilities extended instead of reduced.
This may be readily agreed to, but does
it have anythingto do with violence?Yes,
because the net result may still be that
human beings are effectively prevented
from realizing their potentialities. Thus,
many contemporary thinkers9emphasize
that the consumer's society rewards amply
he who goes in for consumption,
while not positively punishing him who
does not. The system is reward-oriented,
based on promises of euphoria, but in so
being also narrows down the ranges of
action. It may be disputed whether this
is better or worse than a system that
limits the range of action because of the
dysphoric consequences of staying outside
the permitted range. It is perhaps
better in terms of giving pleasure rather
than pain, worse in terms of being more
manipulatory, less overt. But the important
point is, the awareness of the concept
of violence can be extended in this
direction, since it yields a much richer
basis for discussion.
The third distinctionto be made is on
the object side: whether or not there is
an object that is hurt. Can we talk about
violence when no physical or biological
object is hurt? This would be a case of
what is referred to above as truncated
violence, but nevertheless highly meaningful.
When a person, a group, a nation
is displaying the means of physical violence,
whetherthrowing stones around or
testingnucleararms,theremaynot beviolence
in the sense that anyone is hit or
hurt, but there is nevertheless the threat
of physical violenceand indirect threat of
mental violence that may even be characterized
as some type of psychological
violence sinceit constrainshuman action.
Indeed, this is also the intention: the
famous balance of power doctrine is
based on efforts to obtain precisely this
effect. And correspondinglywith psychological
violence that does not reach any
object: a lie does not become more of a
truth because nobody believes in the lie.
Untruthfulness is violence according to
this kind of thinkingunderany condition,
which does not mean that it cannot be
the least evil under some widely discussed
circumstances.
Is destruction of things violence?
Again, it would not be violence according
to the complete definition above, but
possibly some 'degenerate' form. But in
at least two senses it can be seen as psychological
violence: the destruction of
things as a foreboding or threat of possible
destruction of persons,10and the destruction
of things as destruction of
something very dear to persons referred
to as consumers or owners.11
The fourth distinctionto be made and
the most important one is on the subject
side: whether or not there is a subject
(person) who acts. Again it may be
asked: can we talk about violence when
nobody is committing direct violence, is
acting? This would also be a case of what
is referredto above as truncatedviolence,
but again highly meaningful. We shall
refer to the type of violence where there
is an actor that commits the violence as
personal or direct,and to violence where
there is no such actor as structural or
indirect.12In both cases individuals may
be killed or mutilated, hit or hurt in both
senses of these words, and manipulated
by means of stick or carrot strategies.
But whereas in the first case these consequences
can be tracedback to concrete
This content downloaded from 133.30.212.88 on Tue, 07 Apr 2015 01:57:19 UTC
All use subject to JSTOR Terms and Conditions
Violence, Peace, and Peace Research 171
persons as actors, in the second case this
is no longer meaningful. There may not
be any personwho directlyharmsanother
person in the structure. The violence
is built into the structure and shows up
as unequal powerandconsequently as unequal
life chances.'3
Resources are unevenly distributed, as
when income distributions are heavily
skewed, literacy/education unevenly distributed,medical
servicesexistentin some
districtsand for some groups only, and so
on.'4 Above all the power to decide over
the distributionof resources is unevenly
distributed.l5The situation is aggravated
furtherif the persons low on income are
also low in education, low on health, and
low on power - as is frequently the case
because these rank dimensions tend to be
heavily correlated due to the way they
are tied together in the social structure.'6
Marxist criticism of capitalist society
emphasizeshow the power to decide over
the surplus from the production process
is reserved for the owners of the means
of production, who then can buy themselves
into top positions on all other rank
dimensions because money is highly convertible
in a capitalist society - if you
have money to convert, that is. Liberal
criticism of socialist society similarly emphasizes
how power to decide is monopolized
by a small group who convert
power in one field into power in another
field simply because the opposition cannot
reach the stage of effective articula-
tion.
The importantpoint hereis that if people
are starving when this is objectively
avoidable, then violence is committed,
regardless of whether there is a clear
subject-action-object relation, as during
a siege yesterday or no such clear relation,
as in the way world economic relations
are organized today.17 We have
baptized the distinction in two different
ways, using the word-pairs personalstructuraland
direct-indirectrespectively.
Violence with a clear subject-objectrelation
is manifest because it is visible as
action. It corresponds to our ideas of
what dramais, and it is personal because
there are persons committing the violence.
It is easily captured and expressed
verbally since it has the same structure
as elementary sentences in (at least IndoEuropean)
languages: subject-verb-object,
with both subject and object being
persons. Violence without this relation
is structural, built into structure. Thus,
when one husband beats his wife there
is a clear case of personal violence, but
when one million husbands keep one
million wives in ignorance there is structural
violence. Correspondingly, in a
society where life expectancy is twice as
high in the upper as in the lower classes,
violence is exercised even if there are no
concrete actors one can point to directly
attacking others, as when one person kills
another.
In order not to overwork the word
violence we shall sometimes refer to the
condition of structuralviolence as social
injustice.18 The term 'exploitation' will
not be used, for several reasons. First, it
belongs to a political vocabulary, and
has so many political and emotional
overtones that the use of this term will
hardly facilitate communication. Second,
the term lends itself too easily to expressions
involving the verb exploit, which in
turn may lead attention away from the
structural as opposed to the personal
nature of this phenomenon - and even
lead to often unfounded accusations
about intended structuralviolence.19
The fifth distinction to be made is
between violence that is intended or
unintended.This distinction is important
when guilt is to be decided, since the
concept of guilt has been tied more to
intention,both in Judaeo-Christianethics
and in Roman jurisprudence, than to
consequence (whereas the present definition
of violence is entirely located on
This content downloaded from 133.30.212.88 on Tue, 07 Apr 2015 01:57:19 UTC
All use subject to JSTOR Terms and Conditions
172 Johan Galtung
the consequence side). This connection
is important because it brings into focus
a bias present in so much thinking about
violence, peace, and related concepts:
ethical systems directed against intended
violence will easily fail to capture structural
violence in their nets - and may
hence be catching the small fry and letting
the big fish loose. From this fallacy
it does not follow, in our mind, that the
opposite fallacy of directing all attention
against structural violence is elevated
into wisdom. If the concernis with peace,
and peace is absence of violence, then
action should be directedagainstpersonal
as well as structural violence; a point
to be developed below.
Sixth, there is the traditional distinction
between two levels of violence,
the manifest and the latent.20Manifest
violence, whether personal or structural,
is observable; although not directly since
the theoretical entity of 'potential realization'
also enters the picture. Latent
violence is something which is not there,
yet might easily come about. Since violence
by definition is the cause of the
difference (or of maintaining the nondecrease)
between actual and potential
realization, increasedviolence may come
about by increasesin the potential as well
as by decreases in the actual levels.
However, we shall limit ourselves to the
latter and say that there is latent violence
when the situation is so unstable that the
actual realization level 'easily' decreases.
For personal violence this would mean
a situation where a little challenge would
trigger considerable killing and atrocity,
as is often the case in connection with
racial fights. In such cases we need a way
of expressing that the personal violence
is also there the day, hour, minute, second
before the firstbomb, shot, fist-fight,
cry - and this is what the concept of
latent, personal violence does for us. It
indicates a situation of unstable equilibrium,
where the level of actual realization
is not sufficiently protected against
deterioriationby upholding mechanisms.
Similarly with structural violence: we
could imagine a relatively egalitarian
structure insufficiently protected against
sudden feudalization, against crystallization
into a much more stable, even
petrified, hierarchical structure. A revolution
brought about by means of a
highly hierarchical military organization
may after a brilliant period of egaliatarianism,
and after major challenge,
revert to a hierarchical structure. One
way of avoiding this, of course, is to
avoid hierarchicalgroup struggleorganizations
in the first run, and use nonviolent
nonhierarchical guerrilla organizations
in the fight so as to let the means
be a preview of the egalitariangoal.21
That concludes our list of dimensions
of violence, although many more could
be included. One question that immediately
arises is whether any combinations
from these six dichotomies can be
ruled out a priori, but there seems to be
no such case. Structural violence without
objectsis also meaningful;truncation
of the complete violence relation can go
so far as to eliminate both subjects and
objects. Personal violence is meaningful
as a threat, a demonstration even when
nobody is hit, and structuralviolence is
also meaningful as a blueprint, as an
abstract form without social life, used to
threaten people into subordination: if
you do not behave, we shall have to
reintroduceall the disagreeablestructures
we had before.
Disregardingthe negative-positive distinction
as less important in this context,
we end up, essentially, with the typology
illustrated in Figure 1.
If peace now is regardedas absence of
violence, then thinking about peace (and
consequently peace research and peace
action) will be structured the same way
as thinking about violence. And the
violence cake can evidently be cut a
This content downloaded from 133.30.212.88 on Tue, 07 Apr 2015 01:57:19 UTC
All use subject to JSTOR Terms and Conditions
Violence, Peace, and Peace Research 173
Figure 1. A Typologyof Violence
intended . i. manifest
VTIOLENCFr
4,
structural
4,
without
objects
not intended
physical - *
- personal
psychological *--es"ol
J Y?
without with
objects objects
number of ways. Tradition has been to
think about violence as personal violence
only, with one important subdivision in
terms of 'violence vs. the threat of violence',
another in terms of 'physical vs.
psychological war', still another (important
in ethical and legal thinking)
about 'intended vs. unintended', and so
on. The choice is here to make the distinction
between personal and structural
violence the basic one; justification has
been presented (1) in terms of a unifying
perspective (the cause of the difference
between potential and actual realization)
and (2) by indicating that there is no
reason to assume that structuralviolence
amounts to less suffering than personal
violence.
On the other hand, it is not strange
that attention has been focussed more
on personal than on structural violence.
Personal violence shows.22The object of
personal violence perceives the violence,
usually, and may complain - the object
of structural violence may be persuaded
not to perceive this at all. Personal violence
representschange and dynamism not
only ripples on waves, but waves on
otherwise tranquilwaters. Structuralviolence
is silent, it does not show - it is
essentially static, it is the tranquilwaters.
In a static society, personal violence will
be registered,whereas structuralviolence
may be seen as about as natural as the
air around us. Conversely: in a highly
dynamic society, personal violence may
be seen as wrong and harmful but still
somehow congruent with the order of
' latent
_ physical
^ ^~> ~psychological
with
objects
things, whereas structural violence becomes
apparentbecause it stands out like
an enormous rock in a creek, impeding
the free flow, creating all kinds of eddies
and turbulences. Thus, perhaps it is not
so strange that the thinking about personal
violence (in the Judaeo-ChristianRoman
tradition) took on much of its
present form in what we today would
regard as essentially static social orders,
whereas thinking about structural violence
(in the Marxist tradition) was formulated
in highly dynamic northwestEuropean
societies.
In other words, we conceive of structural
violence as something that shows a
certain stability, whereas personal violence
(e.g. as measuredby the tolls caused
by group conflict in general and war in
particular) shows tremendous fluctuations
over time. This is illustrated in Figure
2.
violence ,
level - ? I > time
Figure2. TimeandtheTwoTypesof Violence
This is to a large extent tautological. A
type of violence built into the social
structure should exhibit a certain stability:
social structuresmay perhaps sometimes
be changed over night, but they
may not very often be changed that
quickly. Personal violence, which to a
larger extent is seen as subject to the
whims and wishes of individuals, should
show less stability. Hence personal vioIY
- -
-44
This content downloaded from 133.30.212.88 on Tue, 07 Apr 2015 01:57:19 UTC
All use subject to JSTOR Terms and Conditions
174 Johan Galtung
lence may more easily be noticed, even
though the 'tranquilwaters' of structural
violence may contain much more violence.
For this reason we would expect a
focus or personal violence in after-war
periods lest they should become betweenwar
periods; and if the periods protracts
sufficiently for the major outburst of
personal violence to be partly forgotten,
we would expect a concentration on
structuralviolence, provided the societies
are dynamicenough to makeanystability
standout as somehow unnatural.23
3. The means of personal and structural
violence
To make this distinction less abstract, let
us now explore how personal and structural
violence, are, in fact, carried out.
Starting with personal violence, concentration
on 'actual somatic realization':
how can it be reduced or kept low at the
hands of somebody else? The question is
simple, as are the answers since they
suggest an instrumental approach to the
problem of violence. There is a wellspecified
task to be done, that of doing
bodily harm unto others, and there are
persons available to do it. But this is a
production relation, suggesting a 'development'
much like in the economic sector
of society, with the introduction of
increasinglyrefinedtools and differentiated
social organization - only that the
tools in this case are referredto as weapons
or arms, and the organization is not
called a workshop or a factory, but a
gang or an army.
A typology of personal, physical violence
can now be developed focussing on
the tools used, starting with the human
body itself (in the elementary forms of
fist fights and the more advanced forms,
such as Karate and Aikido), proceeding
towards all kinds of arms culminating,
so far, with ABC weapons. Another
approach would use the form of organization,
starting with the lone individual,
proceeding via mobs and crowds ending
up with the organizations of modem
guerrilla or army warfare. These two
approaches are related:just as in economic
organizationsthe means and mode of
production (here direct bodily violence)
depend on each other, and if one is lagging
behind a conflict will arise.Together
these two approaches would yield the
history of military warfare as a special
case, since much bodily violence is not
military. The approach would be cumulative
for a weapon or technique, and a
form of organizationonce developedmay
become obsolete but not erased; hence
this typology would not be systematic,
but always open to record new devel-
opments.
A more systematic approach can be
obtained by looking at the target; the
human being. He is relativelyknown anatomically
(structurally)and physiologicTable
1. A Typologyof PersonalSomatic Violence
Focussed on the anatomy Focussed on the physiology
1. crushing(fist fight, catapults) 1. denialof air (choking, strangulation)
2. tearing(hanging,stretching,cutting) 2. denialof water(dehydration)
3. piercing(knives,spears,bullets) 3. denialof food (starvationdue to siege, em-
4. burning(arson, flame, thrower) bargo)
5. poisoning(in waterand food, in gases) 4. denialof movement
6. evaporation(as in nuclearexplosion) a. by body constraint(chains, gas)
b. by space constraint (prison, detention,
exile)
c. by brain control (nerve gases, 'brain-
washing')
This content downloaded from 133.30.212.88 on Tue, 07 Apr 2015 01:57:19 UTC
All use subject to JSTOR Terms and Conditions
Violence, Peace, and Peace Research 175
ally (functionally), so typologies can be
developed on that basis. One primitive
typology might be as shown in Table 1.
The basic distinction is not water-tight,
but neverthelessuseful: for one thing is to
try to destroy the machine (the human
body) itself, another to try to prevent the
machine from functioning. The latter can
be done in two ways: denial of input
(sources of energy in general, air, water,
and food in the case of the body), and
denial of output(movement). The human
output can be somatic, recorded by the
outside as movement (with standstill as
a limiting case) or mental not recorded
directly from the outside (only by indicators
in the form of movements, including
movements of vocal chords). The
borderline between physical and psychological
personalviolence is not veryclear,
since it is possible to influence physical
movements by means of psychological
techniques, and vice versa: physical constraints
certainly have mental implica-
tions.
In Table 1 some of the techniques have
been indicated in parenthesis. A note
should be added here about explosions. In
principle they are of two kinds: to propel
some missile, and to work directly on
human bodies. Explosions are much used
for the latter purpose because they combine
the anatomical methods: a standard
bomb would combine 1 and 2; add some
shrapnel and 3 is also taken care of; add
some simple chemicals so as to make it
a fire bomb and 4 is taken into account;
some gases would include 5 and if in addition
the contraption is made nuclear
the crowning achievement, 6, is there presumablyfor
ever, at least in principle,
since it is difficult systematically to unmake
an invention, it can only be suppressed.
New weapons can always be invented,
based on one or any combination
of the principles in the Table. But there
is also room for the more basic innovation:
the introduction of a new principle.
Is it now possible to construct a corresponding
typology for structural violence?
If we accept that the general formula
behind structuralviolence is inequality,
above all in the distributionof power,
then this can be measured;and inequality
seems to have a high survival capacity
despite tremendous changes elsewhere.24
But if inequality persists, then we may
ask: which factors, apart from personal
violence and the threat of personal violence,
tend to uphold inequality? Obviously,
just as military science and related
subjects would be indispensable for the
understanding of personal violence, so
is the science of social structure, and
particularly of stratification, indispensable
for the understanding of structural
violence.
This is not the occasion to develop
general theories of social structure, but
some ideas are necessary to arrive at
some of the mechanisms. Most fundamental
are the ideas of actor, system,
structure, rank and level. Actors seek
goals, and are organized in systems in
the sense that they interact with each
other. But two actors, e.g. two nations,
can usually be seen as interacting in
more than one system; they not only
cooperate politically, e.g. by tradingvotes
in the UN, but also economically by
trading goods, and culturally by trading
ideas. The set of all such systems of
interaction, for a given set of actors, can
then be referred to as a structure. And
in a structure an actor may have high
rank in one system, low in the next, and
then high in the third one; or actors may
have either consistently high ranks or
consistently low ranks.
However, if we look more closely at
an actor, e.g. a nation, we shallvery often
be able to see it as a structurein its own
right, but an integrated structure since
it is able to appear as an actor. This
'Chinese boxes' view of actors is very
important, and leads to the concept of
This content downloaded from 133.30.212.88 on Tue, 07 Apr 2015 01:57:19 UTC
All use subject to JSTOR Terms and Conditions
176 Johan Galtung
level of actors. There are three major
interpretations:25
- in terms of territories: a nation can be seen
as a set of districts, in turn seen as a set of
municipalities,and these are then seen as a
set of individuals;
- in terms of organizations: a factory can often
be seen as an assemblylinewithsub-factories
feedinginto the assembly-linewiththeirproducts,
finally coming down to the individual
worker.
- in terms of associations: they can often be
seen as consisting of local chapters,ending
up with individualmembers.
Thus, the image of the social orderor disorder
can be presented as in Figure 3.
structure,level n
highrankO
Oactor
low rank( )
systems
structure
level n-l
0
0
0
systems
{_4
aytm
Figure 3. An Imageof the Social Order
In all these systems there is interaction,
and where there is interaction, value is
somehow exchanged. It then makes very
much sense to study what the value-distribution
is after the system has been operating
for some time, and the gross
distinction has been made between egalitarian
and inegalitarian distributions.
We can now mention six factors that
serve to maintain inegalitarian distributions,
and consequently can be seen as
mechanisms of structuralviolence:
1. Linearrankingorder- the rankingis complete,
leaving no doubt as to who is higher
in any pair of actors;
2. Acyclicalinteractionpattern- all actors are
connected,but only one way - thereis only
one 'correct'path of interaction;
3. Correlationbetweenrankandcentrality- the
higher the rank of the actor in the system,
the more centralhis position in the interaction
network;
4. Congruencebetweenthe systems- the interaction
networksarestructurallysimilar.
5. Concordancebetweentheranks- if an actor
is highin one systemthenhe also tendsto be
highin anothersystemwherehe participates
and
6. High rank couplingbetweenlevels- so that
the actoratIevel n-l arerepresentedat level
n throughthe highest rankingactor at level
n-1.
The factors can best be understood by
examining to some extent their negation,
starting with the last one.
Thus, imagine that a nation is dominated
by an economic and cultural capital,
but has a much smaller political
capital through which most political interaction
in the international system is
carriedout. This would tend to distribute
the power at the level of cities since the
coupling is not at the highest point.
Similarly,we could imagine that the major
road from the capital to a districtdid
not connect directly with the district
point of gravitybut with some peripheral
point; as when a governmentis represented
abroad not by the president or prime
minister but by the foreign minister or
a sub-factory not by the manager
but by his deputy. But very often the top
actor at level n-1 is made the representative
at level n - with a number of im-
plications.26
Similarly, imagine there is considerable
rank discordance, even to the point
where the summated rankings of the
actors tend to be relativelyequal. In that
case, patterns of inequality would be less
consistent and less reinforcing, and the
amount of disequilibrium in the system
"i/
This content downloaded from 133.30.212.88 on Tue, 07 Apr 2015 01:57:19 UTC
All use subject to JSTOR Terms and Conditions
Violence, Peace, and Peace Research 177
would also tend to upset any stability.
Moreover,if thesystemsarenot congruent
but differ in structure,actors will not so
easily generalize interaction patterns but
be more flexible,less frozen into one way
of acting (for instance servility). And if
the actor with highest rank did not necessarily
have the most central position
in the network this would diminish his
power, which would also be diminished
if actorswith lower rankswereto a larger
extent permitted direct interaction
(not only interaction mediated through
the actors with high rank). Finally: nonlinear,
pyramidal (also known as partial)
ranking orderpermitsmoreleeway, more
flexibility in the system.27
Many propositions can now be developed
about this, a basic one being that
social systems will have a tendency to
develop all six mechanisms unless deliberately
and persistently prevented from
doing so. Thus the pattern is set for an
aggravation of inequality, in some structures
so much so that the lowest-ranking
actors are deprived not only relative to
the potential, but indeed below subsistence
minimum. Inequality then shows
up in differentialmorbidityand mortality
rates, between individuals in a district,
between districts in a nation, and between
nations in the international system
- in a chain of interlocking feudal
relationships. They are deprived because
the structure deprives them of chances
to organizeand bringtheir power to bear
against the topdogs, as voting power,
bargaining power, striking power, violent
power - partly because they are
atomized and disintegrated, partly because
they are overawed by all the
authority the topdogs present.
Thus, the net result may be bodily
harm in both cases, but structural viol-
encewillprobablyjustasoftenberecorded
as psychological violence. Hence, highly
differentmeans may lead to highlysimilar
results- a conclusion to be exploredlater.
4. The relation between personal and
structuralviolence
In this section some comments will be
offered on this relationship, following
this outline:
1.Is therereally a distinctionbetweenpersonal
and structuralviolence at all?
2. If thereis, does not one type of violencepresuppose
the manifestpresenceof the other?
3. If pure types exist, could it not nevertheless
be said that they have a pre-history of the
other type?
4. If this is not generallythe case, could it not
be that one type of violencepresupposesthe
latent presenceof the other?
5. If this is not the case, could it not be that
one is the price we have to pay for the absence
of the other?
6. If this is not generallythe case, could it not
be that one type is much more importantin
its consequencesthan the other?
Let us startwith the firstquestion.
It may be argued that this distinction
is not clear at all: it disregards slights
of the structuralelement in personal violence
and the personal element in structural
violence. These important perspectives
are regained if a person is seen as
making his decision to act violently not
only on the basis of individual deliberations
but (also) on the basis of expectations
impinging on him as norms
contained in roles contained in statuses
through which he enacts his social self;
and, if one sees a violent structure as
something that is a mere abstraction
unless upheld by the actions, expected
from the social environment or not, of
individuals. But then: does not this mean
that there is no real distinction at all?
Cannot a person engaging in personal
violence always use expectations from
the structure as an excuse, and does not
a person upholding an exploitative social
structure have responsibility for this?
This content downloaded from 133.30.212.88 on Tue, 07 Apr 2015 01:57:19 UTC
All use subject to JSTOR Terms and Conditions
178 Johan Galtung
The distinction that nevertheless remains
is between violence that hits human
beings as a directresult of Figure 4
type actions of others, and violence that
hits them indirectly because repressive
structures (as analyzed in preceding section)
are upheld by the summated and
concerted action of human beings. The
qualitative difference between these actions
is the answer. The question of guilt
is certainly not a metaphysical question;
guilt is as real as any other feeling, but a
less interesting one. The question is
rather whether violence is structured in
such a way that it constitutes a direct,
personal link between a subject and an
object, or an indirect structuralone, not
how this link is perceived by the persons
at eitherend of the violence channel. The
objective consequences, not the subjective
intentions are the primaryconcern.
But are personal and structural violence
empirically, not only logically, independent
of each other? Granted that
theremay be a corrrelationso that structures
richly endowed with structuralviolence
often may also display above average
incidence of personal violence, it is
possible to have them in pure forms, to
have one without the other? are there
structureswhereviolence is person-invariant
in the sense that structural violence
persistsregardlessof changes in persons?
And conversely, are there persons where
violenceis structure-invariantin the sense
that personal violence persists regardless
of changes in structuralcontext?
The answer seems to be yes in either
case. The typical feudal structure, with
a succession of incapsulating hierarchies
of metropole-satellite relationships is
clearly structurally violent regardless of
who staffs it and regardless of the level
of awareness of the participants: the
violence is built into the structures. No
personal violence or threat of personal
violence are needed. And there are persons
who seem to be violent in (almost)
any setting- often referredto as 'bullies'.
Characteristic of them is precisely that
they carry their violent propensity with
them far outside any structural context
deemed reasonable by society at large,
for which reason they will often be institutionalized
(in prison or mental hospital,
dependingon which basic norms they
infract first and most clearly). Hence, we
may conclude that the two forms of
violence are empirically independent:
the one does not presuppose the other.
But from this alone it cannot be concluded
that there is no necessary (not
only sufficient) causal relationship between
the two types of violence, or that
the even stronger condition of one-way
reductionism is not fulfilled. One may
argue that all cases of structuralviolence
can, by closer scrutiny,be traced back to
personal violence in their pre-history.
An exploitative caste system or race
society would be seen as the consequence
of a large-scale invasion leaving a thin,
but powerful top layer of the victorious
group after the noise of fighting is over.
A bully would be seen as the inevitable
product of socialization into a violent
structure: he is the rebel, systematically
untrained in other ways of coping with
his conflicts and frustrations because the
structureleaves him with no alternatives.
That structural violence often breeds
structural violence, and personal violence
often breeds personal violence nobody
would dispute - but the point here
would be the cross-breeding between
the two. In other words: pure cases are
only pure as long as the pre-history of
the case or even the structuralcontext are
conveniently forgotten.
Far from denying that these may be
fruitful perspectives both for research
into the past and the etiology of violence
as well as for search into the future and
therapy for violence we would tend to
reject the position that violence presupposes
a pre-history of violence of the
This content downloaded from 133.30.212.88 on Tue, 07 Apr 2015 01:57:19 UTC
All use subject to JSTOR Terms and Conditions
Violence, Peace, and Peace Research 179
same or opposite kinds. This view is a
breeding theory, and like all breeding
theories it fails to answer two questions:
how did the process come into being at
all? and is spontaneous generation of
violence impossible, or are all cases of
violence the legitimate offspring of other
cases of violence - handed down through
some kind of apostolic succession, the
content being more like 'original sin'
though?
Take the case of structural violence
first. Here it may be arguedwe will never
get the perfect test-case. Imagine we
based our thinking on something like
this: people, when left to themselves in
isolation (in a discussion group, stranded
on an isolated island, etc.) will tend to
form systems where rank, or differential
evaluation of relatively stable interaction
patternsreferredto as status, will emerge;
high rankstend to clusteron persons who
already have some high ranks, and interaction
tends to flow in their direction hence
the net result is sooner or later a
feudal structure. One might then object:
yes, because these persons are already
socialized into such structures, and all
they do is to projecttheirexperiencesand
their habits so as to give life to an embryonic
structure. And there is no way
around it: human beings, to be human,
have to be rated by humans, hence there
will always be an element of succession.
Maybe, but, we also suspect that the
reasoning above holds true even under
tabula rasa conditions because it probably
is connected with the fact (1) that
individualsaredifferentand (2) that these
differences somehow are relevant for
their interaction behavior. Hence, special
measures are needed to prevent the
formation of feudal structures: structural
violence seems to be more 'natural'
than structuralpeace. And similarlywith
personalviolence: it is difficultto see how
even the most egalitarianstructurewould
be sufficient to prevent cases of violence,
whether they result from conflicts or not.
Personal violence is perhaps more 'natural'
than personal peace. It could also
be argued that an inegalitarian structure
is a built-in mechanism of conflict control,
precisely because it is hierarchical,
and that an egalitarian structure would
bring out in the open many new conflicts
that are kept latent in a feudal structure.
One could now proceed by saying
that even if one type of violence does not
presuppose the manifest presence of the
other, neither synchronically, nor diachronically,
there is nevertheless the
possibility that manifest structural violencepresupposeslatent
personalviolence.
When the structure is threatened, those
who benefit from structural violence,
above all those who are at the top, will
try to preserve the status quo so well
geared to protect their interests. By observing
the activities of various groups
and persons when a structureis threatened,
and more particularly by noticing
who comes to the rescue of the structure,
an operational test is introduced that can
be used to rank the members of the
structure in terms of their interest in
maintaining the structure. The involvement
that does not show clearly in times
of unimpeded persistence is brought up
to the surface when there is turbulence.
But one has to observe carefully, for
those most interestedin the maintenance
of status quo may not come openly to
the defence of the structure: they may
push their mercenariesin front of them.28
In other words, they may mobilize the
police, the army, the thugs, the general
social underbrush against the sources
of the disturbance, and remain themselves
in more discrete, remote seclusion
from the turmoil of personal violence.
And they can do this as an extrapolation
of the structural violence: the violence
committed by the police is personal by
our definition, yet they are called into
action by expectations deeply rooted in
This content downloaded from 133.30.212.88 on Tue, 07 Apr 2015 01:57:19 UTC
All use subject to JSTOR Terms and Conditions
180 Johan Galtung
the structure- thereis no need to assume
an intervening variable of intention.
They simply do theirjob.
This view is probably generally very
valid, even if it may underestimate the
significance of a number of factors:
1. the extent to whichthe'tools of oppression'
may have internalizedthe repressivestructure
so that theirpersonalviolence is an expressionof
internalized,not onlyinstitution-
alizednorms;
2. the extent to which those who benefitfrom
the structuralviolencemay thenlselveshave
severe and sincere doubts about that structure
and prefer to see it changed, even at
theirown expense;
3. the extent to which the 'challenge of the
structure'may be a personalconfrontation
with the police etc. more than with the
structure,and reveal more about the dynamics
of interpersonalrelations than about
the structure.29
4. the extent to whichall membersin a violent
structure,not only the topdogs, contribute
to its operationandhenceareall responsible
as they can all shake it through their non-
cooperation.
But these are minor points; social affairs
always refuse to be capturedin simplistic
formulations. More importantis whether
one can also turn the proposition around
and gain some insight by saying that
manifest personal violence presupposes
latent structural violence - which is not
the same as saying that it presupposes
manifest structural violence. The idea
would be that of an egalitarian structure
maintained by means of personal violence,
so that when this pattern of violence
is challenged to the point of abolition
there will be an emergence of structural
violence.
The proposition is interesting because
it may open for some possible insights in
structuresyet unknown to us. It does not
seem a priori unreasonable to state that
if the absence of personal violence is
combined with a pattern of structural
violence, then personal violence is nevertheless
around the corner - and correspondingly
that if absence of structural
violence is combined with personal violence,
then structural violence is also
aroundthecomer. Allwe aresayingisonly
that the sum of violence is constant, only
that onehas to takeintoaccount thelatent
variety of the type of violence 'abolished'
to see more clearly how that type is in a
standby position, ready to step in once
the other type crumbles. Absence of one
type of violence is bought at the expense
of the threat of the other.
But, however insight-stimulating this
may be in certain situations we refuse to
accept this pessimistic view for two reasons.
First, the two propositions seem
simply not to be true. It is not at all
difficult to imagine a structureso purely
structural in its violence that all means
of personal violence have been abolished,
so that when the structureis threatened
there is no second trench defense by
mobilizing latent personal violence. Similarly,
a structure may be completely
unpreparedfor freezing the released forces
stemming from a reduction of personal
violence into a hierarchical order.
Empirically such cases may be rare, but
yet significant.
Second, the assumption would be that
human beings somehow need violence to
be kept in line; if not of the personal
type, then of the structural variety. The
argument would be that if there is no
personal violence or threat of personal
violence then a very strong hierarchical
order is needed to maintain order and
to control conflict; and if there is no
structuralviolence or threat of structural
violence, then personal violence will
easily serve as a substitute. But even if
this may be a reasonabletheory to explain
possible empirical regularities, that in
itself is not sufficient argument for reifying
a regularity into a principle supposedly
eternally valid. On the contrary,
this would be a highly pessimistic view
This content downloaded from 133.30.212.88 on Tue, 07 Apr 2015 01:57:19 UTC
All use subject to JSTOR Terms and Conditions
Violence, Peace, and Peace Research 181
of the human condition, and to accept
it fully would even be a capitulationist
view.
From the problem of whether one type
of violence is necessary to obtain or
sustain the other type, whether at the
manifest or the latent levels, it is not far
to the opposite problem: is one type of
violence necessary or sufficient to abolish
the other type? The question, which
actually splits into four questions, brings
us directly into the center of contemporary
political debate. Let us examine
briefly some of the arguments.
1. Structural violence is sufficient to abolish
personalviolence.This thesisseems to have a
certainlimitedand short-termvalidity.If all
the methodsmentionedabovefor sustaining
structural violence are implemented, then
it seemsquitepossiblethatpersonalviolence
betweenthe groupssegregatedby the structure
is abolished. The underdogs are too
isolated and too awed by the topdogs, the
topdogs have nothing to fear. But this only
holds between those groups; within the
groupsthe feudal structureis not practised.
And although the structure probably is
among the most stable social structures
imaginable, it is not stable in perpetuity.
There are many ways in which it may be
upset,and resultin tremendousoutburstsof
personalviolence. Hence, it may perhapsbe
saidto be a structurethatservesto compartmentalizepersonalviolence
in time, leading
to successionsof periodsof absenceand presence
of personalviolence.
2. Structuralviolence is necessary to abolish
personalviolence.This is obviouslynot true,
since personal violence will cease the moment
the decision not to practiseit is taken.
But this is of course begging the question:
underwhat condition is that decision made
and really sustained? That structuralviolence
representsan alternativein the sense
that much of the 'order' obtained by means
of (the threatof) personalviolence can also
be obtained by (the threat of) structural
violence is clear enough. But to state a relation
of necessity is to go far outside our
limitedempiricalexperience.
3. Personal violence is sufficient to abolish
structuralviolence. Again, this thesis seems
to have a certainlimitedshort-termvalidity.
Personal violence directedagainst the topdogs
in a feudal structure incapacitating
them bodily by means of the techniquesin
Table 1, used singly or combined. Whenthe
topdogs are no longer there to exercise
their roles the feudal structurecan clearly
no longer function. Hence, just as under 1
above between-group structural violence
may be abolished by this process. But to
abolish the topdogsin a violent structureis
one thing, to abolish the violent structure
quite another, and it is this fallacy of misplaced
concretenessthat is one of the strongest
argumentsagainstthe proposition. The
new power group may immediatelyfill the
vacancies, retaining the structure, only
changing the names of the incumbentsand
possibly the rationalizationof the structure,
in which case the structuralviolence is not
evenabolishedfor a shortterm.Orthestructuremayre-emergeaftersome
time, because
of internaldynamismor becauseit has after
all been firmly imprinted on the minds of
the new power-holders and has thus been
presentall the time in latentform.
4. Personal violence is necessary to abolish
structuralviolence.This is, of course, a famous
revolutionaryproposition with a certain
currency.One may argue against it on
three grounds: empirically, theoretically
and axiologically. Empirically one would
point to all the cases of structuralchange
decreasingstructuralviolence that seem to
take place without personal violence. The
counter-argumentwill be thattherewerecaseswith
no basicchangeof thestructure,for
if therehadbeena fundamentalthreatto the
power-holdersthentheywouldhaveresorted
to personalviolence.Theoreticallyone would
point to the qualitative differencebetween
the means of personal and structural violenceand
ask: even if personalviolence may
lead to the abolition of structuralviolence,
is it not likely that some, and possibly also
moreeffectivemeansof changinga structure
would be structural,for instancesystematic
changes of interactionnetworks, rank profiles
etc.? In other words, the belief in the
indispensabilityof personalviolencecould be
said, on theoreticalgrounds,to be a case of
fetishizationof personal violence. And then
there is the axiological argument: even if
personalviolencecould be seen as indispensable
up till today, on empirical and/or
theoreticalgrounds,this would be one more
good reason for a systematicsearchfor the
conditions under which this indispensability
would disappear.
This content downloaded from 133.30.212.88 on Tue, 07 Apr 2015 01:57:19 UTC
All use subject to JSTOR Terms and Conditions
182 Johan Galtung
Again our search seems to fail to
uncover any absolutes. It is hard to
sustain a belief in sufficiency or necessity
one way or the other. The two types of
violence simply do not seem to be more
tightly connected empirically than logically
- and as to the latter, the whole
exerciseis an effortto show that they may
be seen as logically independent even
though they are continuous with each
other: one shades into the other.
But even if one now rejectsreductionism
one way or the other there would
still be good reason for focussing research
attention more on one kind of violence
than on the other: it may always be argued
than one is much more important
in its consequences than the other. Thus,
imagine we were able to calculate the
losses incurred by the two forms of violence,
or the gains that would accrue to
mankind if they could be eliminated. In
principlethis should not be quiteimpossible,
at least not for the simpler physical
forms of violence that show up in terms
of mortality,and possibly also in termsof
morbidity. Mortality and morbidity rates
under the condition of absence of war
can usually be calculated relatively well
by extrapolation from pre-warand postwar
data. It is more difficult for the case
of absence of exploitation, but not impossible:
we could calculate the levels attained
if all available resourceswere used
for the purpose of extending and improving
the biological life-span and in
addition were distributed in an egalitarian
fashion in social space. The costs
incurred by violence of one form or the
otherwould then appearas the difference
between the potential and the actual, as
the definition requires, and the costs can
then be compared. One could also imagine
calculations of the costs of the joint
operation of the two forms of violence.
One significant feature of such calculations,
that definitelyshould have a high
priorityon the researchprogramof peace
researchinstitutes, is that the door would
be opened for answers to questions such
as whether the costs in terms of personal
violence were higher or lower than the
gains in reduction of structuralviolence
in, say, the Cuban revolution. The present
author would say that they were
definitely lower, using comparable Latin
American countries as a basis for evaluating
the costs of the structuralviolence
under Batista, but in the equation one
would of course also have to include the
personal violence under Batista and the
structuralviolence under Castro, e.g. in
the form of almost complete alienation of
the formerbourgeoisie, not only as status
holders, but as persons. Such statements
are impressionistic however, they should
be backed up empirically.
But however attractive such calculations
may be - for reasons of intellectual
curiosity about the dynamics of violence,
structuraland personal, even to develop
much higher levels of theoretical insightsinthese
phenomenathanwe possess
today - this is not the same as accepting
cost-benefit analysis in this field as a
basis for political action. The point here
is not so much that one may have objections
to projectingthe mathematical 'one
human life-year = one human life-year',
regardlesshow it is lost or gained, on to
the stage of political action, but rather
that this type of analysis leads to much
too modest goals for political action.
Imagine that the general norm were formulated
'you shall act politically so as
to decreaseviolence, taking into account
both before and after levels of personal
and structuralviolence'. A norm of that
kind would be blind to possible differences
in structural and personal violence
when it comes to their potential for getting
more violence in the future. But it
would also condone action as long as
there is any decrease,and only steerpolitical
action downwardson the violence
surface, not lead to a systematic search
This content downloaded from 133.30.212.88 on Tue, 07 Apr 2015 01:57:19 UTC
All use subject to JSTOR Terms and Conditions
Violence, Peace, and Peace Research 183
for the steepest gradient possible, even
for a descent route hitherto unknown
to man.
But equally important is to recall that
it is hardly possible to arrive at any
general judgment, independent of time
and space, as to which type of violence
is more important. In space, today, it
may certainly be argued that research
in the Americas should focus on structural
violence, between nations as well
as between individuals, and that peace
researchin Europe should have a similar
focus on personal violence. Latent personal
violence in Europe may erupt into
nuclear war, but the manifest structural
violence in the Americas (and not only
there) already causes an annual toll of
nuclear magnitudes. In saying this, we
are of course not neglecting the structural
components of the European situation,
(such as the big power dominance and
the traditional exploitation of Eastern
Europe by Western Europe) nor are we
forgetful of the high level of personal
violence in the Americas even though it
does not take the form of international
warfare (but sometimes the form of interventionist
aggression).
5. On the definition of 'peace' and
'peace research'
With the distinction between personal
and structuralviolence as basic, violence
becomes two-sided, and so does peace
conceived of as the absence of violence.
An extended concept of violence leads to
an extended concept of peace. Just as a
coin has two sides, one side alone being
only one aspect of the coin, not the complete
coin, peace also has two sides:
absence of personal violence, and absence
of structural violence.30 We shall refer to
them as negativepeace and positivepeace
respectively.31
For brevity the formulations 'absence
of violence' and 'social justice' may
perhaps be preferred,using one negative
and one positive formulation. The reason
for the use of the terms 'negative' and
'positive' is easily seen: the absence of
personal violence does not lead to a
positively defined condition, whereas the
absence of structuralviolence is what we
have referred to as social justice, which
is a positively defined condition (egalitarian
distribution of power and resources).
Thus, peace conceived this way
is not only a matter of control and reduction
of the overt use of violence, but of
VIOLENCE
Personal
(direct)
absenceof
personalviolence
or
Negative
peace
Structural (alsoreferredto
(indirect) as *socialinjustice)
Iabsenceof
structuralviolence
or
Positive (alsoreferredto
peace as usocialjustice)
PEACE
Figure4. The ExtendedConceptsof Violence
and Peace
what we have elsewhere referred to as
'vertical development'.32And this means
that peace theory is intimately connected
not only with conflict theory, but equally
with development theory. And peace
research, defined as research into the
conditions - past, present and future of
realizing peace, will be equally intimately
connected with conflict research
and development research; the former
often more relevant for negative peace
and the latter more relevant for positive
peace, but with highly important over-
laps.
To justify this way of looking at peace
and peace research, let us see where the
many efforts to conceive of peace in
terms of only one of these 'sides' or
aspects leads us. Such efforts are likely
to bring into focus, in theory and indeed
2 Journal of Peace Research
This content downloaded from 133.30.212.88 on Tue, 07 Apr 2015 01:57:19 UTC
All use subject to JSTOR Terms and Conditions
184 Johan Galtung
in practice, the onesidedness on which
they are based and to highlight the need
for richer concepts of peace. Here only
a very sketchy outline of this type of
analysis will be presented, particularly
since relations between personal and
structural violence were to some extent
explored in the preceding section.
Thus, a research emphasis on the reduction
of personal violence at the expense
of a tacit or open neglect of research
on structuralviolence leads, very
easily, to acceptance of'law and order'
societies.33Personal violence is built into
the system as work is built into a compressed
spring in a mattress: it only
shows when the mattress is disintegrating.
And on the other hand there may
be a researchemphasis on righting social
wrongs on obtaining social justice at the
expense of a tacit or open acceptanceand
use of personal violence. The short-term
costs of personalviolence appearas small
relative to the costs of continued structural
violence. But personal violence
tends to breedmanifestphysical violence,
not only from the opponent but also inside
one's own group - and the aftermath
of violent revolutions generally seems to
testify to this.
We may summarizeby saying that too
much researchemphasis on one aspect of
peace tends to rationalize extremism to
the right or extremism to the left, depending
on whether onesided emphasis
is put on 'absence of personal violence'
or on 'socialjustice'. And these two types
of extremism are of course not only
formally, but also socially closely related
and in a dialectic manner: one is often a
reaction to the other. When put into
practice both may easily develop into
well-known social orders where neither
of the two aspects of peace are realized:
gross social injustice is maintained by
means of highly manifest personal violence.
The regime usually tries to maintain
a status quo, whetherit means forceful
maintenance of traditional social injustice
that may have lasted for generations,
ortheforcefulmaintenanceof some
new type of injusticebroughtin by an attempt
to overthrow the old system.
If 'peace' now is to be interpreted as
an effort to play on both, one may ask:
does this not simply mean some kind of
'moderate' course, some effort to appear
'objective' by steering carefully between
the two types of extremism outlined
above? There is no doubt a danger in
this direction. Efforts to avoid both
personal and structural violence may
easily lead to accept one of them, or even
both. Thus, if the choice is between
righting a social wrong by means of
personal violence or doing nothing, the
latter may in fact mean that one supports
the forces behind social injustice. And
conversely: the use of personal violence
may easily mean that one gets neither
long-term absence of violence norjustice.
Or, we can put the argument in a
slightly different framework. If we are
interested in e.g. social justice but also
in the avoidance of personal violence,
does this not constrain our choice of
means so much that it becomes meaningful
only in certainsocieties?And particularlyin
societiesthathave alreadyrealized
many social-liberal values, so that there
is considerable freedom of speech and
assembly, and organizations for effective
articulation of political interests?Whereas
we are literally immobilized in highly
repressive societies, or 'more openly repressive
societies' as modem critics of
liberalismmight say? Thus, if our choice
of means in the fight against structural
violence is so limited by the non-use of
personal violence that we are left without
anything to do in highly repressivesocieties,
whether the repression is latent or
manifest, then how valuable is this recipe
for peace?
To this we may answer along many
lines.
This content downloaded from 133.30.212.88 on Tue, 07 Apr 2015 01:57:19 UTC
All use subject to JSTOR Terms and Conditions
Violence, Peace, and Peace Research 185
One answer would be to reject the definition
given above of peace, because
we want 'peace' to refer to something
attainable and also in fact attained, not
to something as utopian as both absence
of personal violence and social justice.
We may then slant the definition of
'peace' in the direction of absence of
personal violence, or absence of structural
violence, depending on where our
priorities are. In our definition above we
have suggested that the two enter in a
completely symmetrial manner: there is
no temporal, logical or evaluative preference
given to one or the other. Social
justice is not seen as an adornment to
peace as absence of personal violence,
nor is absence of personal violence seen
as an adornment to peace as social justice.
Unfortunately, on the printed page,
one has to appear before the other or
above the other, and this is often interpreted
as priority (compare the recent
debate on whether a certain group's political
slogan should be 'peace and freedom'
or 'freedom and peace'). Actually,
somebody should invent some way of
printing so that absolutely no connotation
of priority is implied.
This approach presupposes that we
somehow are attracted by the term
'peace' and would like to let that word
express our goal rather than some other
word. But another answer would be to
give up the word 'peace' and simply
state our interest in one or both of the
two values and then try to do our best
along both dimensions, so to speak. This
appears less satisfactory, because of the
generally widespread use of the term
'peace' - so widespread and so generally
acknowledged that it possibly presents
some kind of substitute in this secular
age for feelings of devotion and community
that in former ages were invoked by
reference to religious concepts. In fact,
'peace' has indeed religious overtones
to many, and that this probably contributes
to the use of the word 'peace' as a
carrier of feelings of universal love and
brotherhood in our days. Hence, in
spite of the many possibilities for semantic
confusion, we would argue in favor
of retaining the term 'peace'.
A third answer would be to combine
the first two approaches, to talk little or
at least not very loudly about peace - for
fear of blushing, among other reasons and
to give up one of the two goals, absence
of violence and social justice. This
attitude, found today in several circles,
may be commended for its honesty and
lack of hypocrisy. Neither the 'law and
order' racist or primitive capitalist
society, nor the openly repressive postrevolutionary
society is presentedas realizations
of 'peace', but as social orders
where one made a choice between two
evils, direct violence or social injustice,
using what was seen as the lesser evil to
drive out the greaterevil (possibly ending
up with both).
And then there is a fourth approach
which will be preferred in this context.
It may be expressed as follows:
Both values, both goals are significant,
and it is probably a disservice to man to
try, in any abstract way, to say that one
is more important than the other. As
mentioned, it is difficult to compare the
amountofsufferingandharmthathasbeen
caused by personal or structuralviolence;
they are both of such an order of magnitude
that comparisons appear meaningless.
Moreover, they seem often to be
coupled in such a way that it is very
difficult to get rid of both evils; more
likely the Devil is driven out with Beelzebub.
In view of this difficulty, so amply
testified through human history, we
should be very careful in passing moral
judgements too readily on those who fail
to realize both goals. To realize one of
them isno meanachievementeither,particularlyif
weconsiderthenumberof social
orders and regimes that realize neither.
This content downloaded from 133.30.212.88 on Tue, 07 Apr 2015 01:57:19 UTC
All use subject to JSTOR Terms and Conditions
186 Johan Galtung
But the view that one cannot meaningfully
workfor both absence of personal
violence andfor socialjustice can also be
seen as essentially pessimistic, as some
sort of intellectualandmoralcapitulationism.34First
of all, there are many forms
of social action available today that
combine both in a highlymeaningfulway.
We are thinking of the tremendously
rapid growth in the field of nonviolent
action, both in dissociative nonviolence
that serves to keep parties apart so that
the weaker part can establish autonomy
and identity of its own, and associative
nonviolence that can serve to bring them
together when a basis for equal nonexploitative
partnership exists.35We are
thinking of all that is known about the
theories of symmetric,egalitarian organization
in general.36We are thinking of
the expanding theory of vertical development,
of participation, decentralization,
codecision. And we are thinking of the
various approaches to arms control and
disarmament issues, although they are
perhaps of more marginal significance.37
This is not the place to develop these
themes; that will be done in other contexts.But
secondly,once the double goal
has been stated - that peace research is
concerned with the conditions for promoting
both aspects of peace - there is
no reason to believe that the future will
not bring us richer concepts and more
forms of social action that combine
absence of personal violence with fight
against social injustice once sufficient
activityis put into researchandpractice.38
There are more than enough people
willing to sacrificeone for the other - it
is by aiming for both that peace research
can make a real contribution.
NOTES
* The present article (PRIO-publicationNo. 23-9 - is a revisedversionof talksoriginallypresented
by the author at the Oslo Conferenceon the plan for a peacemaker'sacademy,organized
jointly by the Peacemakers'AcademyCommittee,Vermontand the InternationalPeaceResearch
Institute,Oslo, 14-17 November 1968;at the peace researchseminarorganizedby the Gandhian
Instituteof Studies,Varanasi,8-9 March1969;at the meetingof theJapanPeaceResearchGroup
Tokyo, 27 March 1969; at a seminarorganizedby the Seminarfor Peace and ConflictResearch,
Lund, 26 April 1969,and at the internationalseminarorganizedby the CentroStudie Iniziative,
Partinico, 3-4 May 1969. I am indebted to the organizersof these meetings, Randolph Major,
Sugata Dasgupta, Hisako Ukita, HaIkanWibergand Danilo Dolci and to manyparticipantsfor
highly stimulatingcommentsand criticism.But special gratitudeshould be expressedto Herman
Schmid,LundUniversity,Sweden,for his lucidand importantcriticismof some conceptsof peace
research,in Journalof Peace Research,1968,pp. 217-232. Although I agreeneitherwith his critiquenor
withhis proposals,andfeel that his way of presentingmy own viewsis misleading,there
arecertainlyfew personswho have stimulateddiscussionand rethinkingin this fundamentalfield
so much. However, the presentarticle L not a systematicanswerto his arguments,but ratheran
effort,partlystimulatedby him, to indicatewhat to the presentauthorseems to be a fruitfulway
of thinkingabout violence, peace and peaceresearch.
1 This point is elaboratedfurtherin Theoriesof Peace (forthcoming),Chapter1.1.
2 This, of course, is not strictlytrue.It wasnot on Fascistornaziagendas,noris it on the agenda
of contemporaryrevolutionarythinking.However,evenfor thesecasesviolenceis not an end, but
rathera meansto overcomeobstaclesimpedingthe realizationof a futureorder,the millennium,
the communistsociety, etc; thesefutureordersdo not seem to containviolence. But this is hardly
a universalhumaninvariant.TheVikingparadiselooks violent,andwarliketribes/societieslike the
Pathanswould probablyput complete absenceof violence last on the agenda,if at all.
3 But whatif a social orderis suchthatsomepeoplelivewell in solid,concretehouses andothers
in shacks that crumbleunderthe first quake, killing the inhabitants?In other words, even if the
natural disaster is unavoidable, differentialsocial impact may have been avoidable. This may
certainlyjustify the use of the term 'structuralviolence'for such differentialhousing standards,
not only because of differentialexposure to earthquakes(as in the earthquakezone in Western
Sicily), but becauseof implicationsfor differentialhealthstandardsin general,educationalpossiThis
content downloaded from 133.30.212.88 on Tue, 07 Apr 2015 01:57:19 UTC
All use subject to JSTOR Terms and Conditions
Violence, Peace, and Peace Research 187
bilities,and so on and so forth. Whetheritjustifiesthe use of such epithets as 'violent' or 'assassin'
to the people sustainingsuch social structures,or (whichis not quite the same) to the people
on top of such social structures,is anothermatter.
4 Since the potential level dependsnot only on the use and distributionof availableresources,
but also on insight, a crucialperson in this pictureis the scientistor anyone who opens for new
insights into how old, or new, resources may be utilized. In other words, anyone who makes
possible what was formerlynot feasible raisesthe level of potentiality.But the level may also be
lowered,perhapsnot so often because insight is forgotten (although history is full of such cases
too) as because resourcesbecome more scarce - for instance due to pollution, hoarding, overutilization,
etc. In short, we make no assumption about the shape of the potential realization
curve through time, nor do we make any assumptionabout the correspondingcurve for actual
realization.In particularwe rejectthe optimisticassumptionaccordingto which both curvesare
monotonously increasingand witha decreasinggap so that thereis asymptoticconvergenceof the
actual to the potential, 'until the potentialities of man are fully realized'. This is an ideology,
usuallyin the form of an underlyingassumption,not a descriptionor reality.As BertrandRussell
writes(Autobiography,Vol, III, p. 221): 'WhenI was young, Victorianoptimism was taken for
granted.It wasthoughtthatfreedomandprosperitywould spreadgraduallythroughoutthe world
by an orderly process, and it was hoped that cruelty, tyranny and injustice would continually
diminish. Hardly anyone was haunted by the fear of great wars. Hardly anyone thought of the
nineteenthcenturyas a brief interludebetween past and future barbarism-.' In short, let us
makeno assumptions,but focus on the causesfor a discrepancybetweenthe curves,admittingfor
a lag in the application and distributionof new insights; whetherthey are called technological
or social.
5 However,it is by no means obvious how potentiallife-spanshould be defined.Onecannot use
the age at deathof the oldestpersondyingtoday or this year;this maybe too low becausehe does
not benefitfrom possibleadvancesin hygieneetc. madetoo late to have an impacton him, or not
yet made, and it may be too high becausehe is speciallyadvantagedgenetically.But the average
of the p%of the social orderbenefitingfully from insightand resourcesalreadyavailableshould
at least yield a basis for an estimateof what is possible today.
6 In an article'On the Meaning of Nonviolence', Journalof PeaceResearch,1965,pp. 228-257
the concept of influenceis basic in an effortto analyze the differencebetweenviolence and nonviolence,
and positive and negativeversions of the latter. In the presentarticle the focus is on a
typology of violence, not on a typology of non-violence.
7 Ibid., pp. 230-234.
8 Loc. cit.
9 This is a recurrenttheme in Herbert Marcsue,One-dimensionalMan (Boston Press, 1968),
especiallyPart I, 'One-dimensionalSociety'.
10 This is a recurrenttheme in much of the analysis of violence in the US. Violence against
property is seen as training,the firstwindow-pane crushed to pieces is also a blow against the
bourgeoisin oneself, a liberationfrom formerconstraints,an act of communicationsignallingto
eithercamp a new belongingnessand above all a rejectionof tacit rules of the game. 'If they can
do that to property,what can they do to persons-'
11 It was pointed out by Herman Kahn (at a seminar at PRIO, May 1969) that middle class
students and lower class police may have highly differentrelations to property: as something
highly replaceablefor the middle class student in an affluent society, as something difficult to
attainfor a lowerclass Irishcop. Whatto one is a relativelyunproblematicact of communication
may to the other be sacriligeous,particularlysince studentsprobablyaspireto mobility and freedom
unfetteredby propertyties.
12 The term 'institutionalviolence' is often sometimes used, but we have preferred'structural'
sinceit is often of a moreabstractnatureand not anythingthatcan be traceddown to a particular
institution.Thus,if the police are highlybiased the terminstitutionalizedviolence may be appropriate,
but this is a highlyconcretecase. Theremaybe violence built into a structurewithoutany
police institutionat all, as will be developedin the next section.
13 Thisis clearlyexpressedby StokeleyCarmichaelin 'BlackPower'(TheDialecticsof Liberation,
David Cooper ed., London Penguin,p. 151, 1968):
'It is important to this discussion of racism to make a distinction between the two types:
individualracism and institutionalracism. The first type consists of overt acts by individuals,
with usually immediate results of the death of victims, or the traumatic and violent
This content downloaded from 133.30.212.88 on Tue, 07 Apr 2015 01:57:19 UTC
All use subject to JSTOR Terms and Conditions
188 Johan Galtung
destructionof property.Thistypecan berecordedon TVcamerasand canfrequentlybe observedin
the processof commission.
The secondtype is less overt,farmoresubtle,less identifiablein termsof specificindividuals
committingthe acts, but is no less destructiveof humanlife. The second type is more
the overalloperationof establishedandrespectedforcesin thesociety andthusdoes not receive
the condemnationthat the firsttype receives.
His distinctionindividual/institutionalis the same as our personal/structural.But we preferthe
term'personal'becausethe person sometimesacts on behalf of groups, whereas'individual'may
be interpretedas the opposite of 'group'. But particularlyin the context Carmichaeldiscusses
groupviolenceis immenselyimportant- themob lynchingas opposedto the individualmurdererbut
that does not make the violence institutional. It still satisfies all the other criteria,e.g. it
consists of 'overt acts by individuals','can be recordedon TV-cameras'(as in a war), etc.
14 The difficultyhere,as often pointed out, is that internationalstatisticsusuallyreflectaverages
and not dispersions,rankingnations in orderof averageachievement,not in terms of degreeof
equalityachievedin distribution.Onereasonis of course that such dataarenot readilyavailable,
but that is only beggingthe questionwhytheyarenot available.Onereasonfor thatagainmaybe
thatit upsetsrankingordersandrevealslesspositiveaspectsof social ordersused to definethemselves
as world leaders,but that is hardly a sufficientexplanation.Anotherreasonmightbe that
the problem is simply not sufficientlyclearly defined,nor is it regardedas sufficientlyfeasible or
indeed desirableto decreasedispersions.When this becomes sufficientlycrystallizedit will also
find expressions in internationalstatistics.
15 The remarkin the precedingnote holds a fortiori here: not only is it difficultto presentany
measureof dispersionof power,it is difficultenoughto measurepowerat all, exceptin the purely
formalsense of voting rights.He who comes up with a reallymeaningfulmeasurein this fieldwill
contributegreatlyto crystallizationof political fightingas well as administrativeendeavors.
16 Again the same: the publications of these correlationswould contribute significantlyto increasedawareness,since
the currentideology is preciselythat correlationsbetweenachievedand
ascribedranksshould be as low as possible, preferablyzero.
17 Economic sanctions occupy interestingmiddleposition here. They are clearlyviolent in their
ultimate consequences,which are starvation etc., but the hope is of course that they are slow
enough to permitcapitulationmuchbeforethat. At the same time they areclearlyalso built into
thestructure,for the mostvulnerablecountriesarealso thecountriesthattend to be at the bottom
of the internationalstratificationin general: high in dependenceon trade, low in commodity
dispersionand low on trade partnerdispersion.See Johan Galtung, 'On the Effects of International
Economic Sanctions, With Examples from the Case of Rhodesia', WorldPolitics, 1967,
pp. 387-416.
18 One expressionof what is meant by social justice is found in declarationsof human rights,
wherea numberof normsaboutequalityarestated.However,theyveryoften sufferfrom the deficiency
that they are personalmore than structural.They referto what individualscan do or can
have,not to who or whatdecideswhattheycan do or have; theyreferto distributionof resources,
not to poweroverthe distributionof resources.In otherwords,humanrightsas usuallyconceived
of arequite compatiblewith paternalismwherebypower-holdersdistributeanythingbut ultimate
power over the distributions,so that equalizationwithout any change in the power structureis
obtained.It is almost painful to see how few seem to realizethat muchof the currentanti-establishment
anti-authorityrevolt is precisely about this: concessions are not enough, not even
equality is enough, it is the way in which decisions about distributionare arrivedat and implementedthat
is basic. But thereis littlereasonto believethatthiswillnot alsoin duetimecrystallize
into some kind of humanrightandbe addedto that list of philosophicalandpoliticalbattlefields.
19 Exploitationalsohasanambiguitywhichweactuallyhaveexploitedin thissection.Thereseems
to be a liberalinterpretationin termsof distributionand inequality,and a Marxistinterpretation
in terms of power, particularlyover the use of the surplus produced by others (in a capitalist
economy). Clearlyone can have one type of exploitationwithout the other.
20 I am indebtedto Hans Riegerand otherparticipantsin the seminarat the GandhianInstitute
of Studiesfor pointing out the possibility of using the manifest-latentdistinction in connection
with both personaland structuralviolence.
21 This is a point whereGandhiand Mao Tse-Tung would agreein theory,althoughin practice
they areboth so dominantin theirorganizationsthat it probablywasnot too meaningfulto speak
cf realegalitarianism.
This content downloaded from 133.30.212.88 on Tue, 07 Apr 2015 01:57:19 UTC
All use subject to JSTOR Terms and Conditions
Violence, Peace, and Peace Research 189
22 See Note 13 for Carmichael'sanalysis. The basic point in our communicationstructureis of
course that personal violence much more easily 'can be recordedon TV cameras',although this
is not correctstrictlyspeaking.Thereis no intrinsicreason why structuralviolence should not be
registeredon TV cameras; in fact, really good cameramendelight in doing exactly this. But the
concept of newsis againstits prominentdisplay;that concept is in itself gearedto personalrather
than structuralviolence. For an analysis, see Johan Galtung and Mari Holmboe Ruge, 'The
Structureof Foreign News', Journalof Peace Research,1965, pp. 64-91, especially on personvs
Structure-orientednews.
23 Herman Schmid seems to be very correct when he points out (op. cit., p. 217) that peace
researchgrew out of a certain historicalcondition and the basic concepts were colored by that
condition. No doubt this explainssome of the emphasison symmetricconflict,and we would add,
on personalviolenceboth becauseof warmemoriesandwarthreats.However,thethreatsof a major
war in the North Atlantic areafailed to materialize,economic growthcontinued, but exploitation
remainedconstant or increased.So, towardsthe end of the 'sixties the focus changes;'for
some persons to a completely new focus (as when Schmid and others would argue in favor of
conflictcreationresearch,of polarizationand revolutionresearch),for others (as the presentauthor)
to an extension of focus, as arguedin the present article.
24 Thus, it is almost unbelievablehow little the gap between rich and poor seems to be affected
by the generalincrease,withinnations and betweennations.
25 This is the generaltheme in Johan Galtung, 'A structuralTheory of Integration',Journalof
Peace Research,1968, pp. 375-395.
26 One of these implicationsis of course that it enhanceshis power: he monopolizes information
from the level above and can convert this into power at his own level. Another implication is
that he is very often untrainedfor or unfit for the task to be performedat the higherlevel since
his frame of referenceall the time has been level n-1. The managerof a certaintype of products
suddenlyfindshimselfon the boardof a big businesscorporationdoing quite differentthings; the
teadingnation in a regionalalliancesuddenlyfindsitself responsiblefor world affairsand forced
to think within a completelynew frame of reference,and so on.
27 We have not discussedthe possibilityof denyingrankdifferencescompletelyby makingeverybody
equal, since there seem always to be some differencesthat elude equalizationattemptsand
thesedifferencestend to becomesignificant.Makeeverybodycitizenswithequalvotingrights,and
differencesin style of life become overwhelming,abolish class differenceson trainsand the upper
classes go by plane, and so on.
28 Few haveexpressedthis imageas well as EldridgeCleaverin SoulonIce (London: Cape, 1969,
p. 92):
'Both police and the armedforces follow orders. Orders.Ordersflow from the top down.
Up there, behind closed doors, in antechambers,in conferencerooms, gavels bang on the
tables, the tinkling of silverdecanterscan be heardas ice wateris pouredby well fed, conservatively
dressed men in horn-rimmedglasses, fashionably dressed American widows
with rejuvenatedfaces and tinted hair, the air permeatedwith the square humor of Bob
Hope jokes. Here all the talking is done, all the thinking, all the deciding.
Grayrabbitsof men scurryforth from the conferenceroom to spreaddecisions throughout
the city, as News. Carryingout ordersis a job, a way of meetingthe paymentson the
house, a way of providingfor one's kiddies.In the armedforcesit is also a duty,patriotism'
Not to do so is treason.'
29 See Note 11 for Kahn's analysis, where he added that fighting with fists would be about as
naturalfor the Irishcops as it is unnaturalfor the uppermiddle class student, and fightingwith
words as naturalfor that studentas it is unnaturalfor the cop. Hence, when the studentdestroys
propertyand heaps abuse on the police he challengesthe police muchbeyond the tolerancelevel,
and the police respond with the reaction they know, violence; a reactionfor which the students
are untrained.One does not need structuralexplanationsto account for an outburstof violence
in suchcases.But one could ask whysuchpeoplearein the police department,and one explanation
can supplementratherthan supersedeanother.
30 This coin metaphor, of course, is not to suggest that one side excludes the other. Indeed, as
pointed out so many timesin the precedingsection: a givensocial ordermayexhibitboth, one or
(perhaps)neither of them. The metaphor applies to the conceptualizationof peace, not to the
empiricalworld.
31 Of course, I am very much aware of changes in my own presentationof these concepts, just
This content downloaded from 133.30.212.88 on Tue, 07 Apr 2015 01:57:19 UTC
All use subject to JSTOR Terms and Conditions
190 JohanGaltung
as I am confidentthat new formulationswill follow in the wake of those presentedhere. Whereas
'negativepeace' remainsfairlyconstant, meaning'absenceof violence', I think it gains from the
precision given to 'violence' in that context, a 'personal violence'. But 'positive peace' is constantly
changing (as is 'positive health' in medicalscience). I used to see it in terms of integration
and cooperation ('An Editorial', JPR, 1964, pp. 1-4), but now agree fully with Herman
Schmid that this expressesa much too integratedand symmetricview of conflict groups, and
probably reflectsthe East-West conflict or a certain ideology in connection with that conflict.
I would now identify'positive peace' mainly with 'social justice', the latter taken in the double
sense of this article- but I thinkone could also be open to othercandidatesfor inclusionsince the
definitiongivenof violenceis broadenoughalso to point in otherdirections.Thisis to some extent
attemptedin section 1.3 of Theoriesof Peace. Moreover,I think Schmidis basicallyright(op. cit.
p. 221) in saying that thereis a tendencyto focus on negative peace becauseconsensus is more
easily obtained - but I share his rejection of that tendency. To reveal and unmask the subtle
mechanismsof structuralviolence and explore the conditions for theirremovalor neutralization
is at least as important,although comparisonsof the two types of violence in termsof priorities
seemsa little bit like discussingwhethermedicalresearchshouldfocus on canceror heartdiseases.
And to this shouldbe added,emphatically,thata disciplinefully satisfiedwithits ownfoundations
and definitionis probablya dead discipline. Fundamentaldebate and debate over fundamentals
are the signs of health,not of disease.These issuesare difficult,and we shall makeprogressonly
throughmore practicein analyzingthem and more praxisin workingwith them.
32 In Theoriesof Development,forthcoming.
33 Thus,thereis littledoubt that in generalpeaceresearch(Schmid,op. cit., p. 222)in this decade
that has passed since it was launchedhas met with more approvalfrom the north-westernestablishmentin
the world than from other quarters,but so has cancerresearch.From this it does not
follow thatpeaceresearchis meaninglessto the thirdworldand to revolutionaryforces. Thesame
skewed distributioncan be found almost anywhere, due to the skewed distribution of world
resourcesand the generallyfeudal structureof the world. But Schmidis certainlyrightin setting
peace researchin a social setting: 'who will pay for it', and 'who will be able to implementadvice
from peace researcher'are basic questions. I only fail to see that there should be any implicit
reasonwhypeaceresearchshouldfallinto thearmsof theestablishmentmorethaninto otherarms
not to mention be able to retainconsiderableautonomy in its pursuits.This presupposesan academic
structurethat does not steerall researchinto the arms of the power-holders,left or right,
but leavesthe road open for pursuitsof insightsinto the mechanismsbehindany kind of violence,
any kind of obstacleto humanself realization.
34 Thus, peace researchis seen here as an effort to promote the realizationof values.To what
extent these values coincide or not with the interestsof certaingroupsis another matter.Hence,
peace researchcould not be identifiedwith the ideology of a group unless that group professed
the same values.Itik also an open questionwhethergroupidentificationwith these valueswill in
fact serve to promote these values.
35 Some of this is exploredin 'On the Meaning of Nonviolence', and infinitelymuch more can
be done in this direction.However,the importantthing seems to be thatthereis no reason whatsoever
why peace researchshould be tied to study of symmetricconflictonly, and to integrative,
or as we preferto say, 'associative'(integrativebeing too strong a term)approaches.Any effort
to explorestructuralviolencewill lead to awarenessof asymmetricconflict,betweenpartieshighly
unequalin capabilities- and I think it is unfairto state that this is neglectedin the type of peace
researchcarriedout at the InternationalPeace ResearchInstitute in Oslo. The terms 'topdogunderdog'may
be unfamiliarand even be resentedby those who preferto do this researchin a
Marxisttraditionandjargon,but it is neverthelessan effort.Moreprecisely,the efforthas beento
understandbetter the structureof structuralviolence, one little indication of which is given in
section 3 of this article.And thereis no implicitreason why the remedyshould be in associative
policies only. On the contrary,I tend to feel in generalthat associativepoliciesarefor equals,i.e.
for symmetricconflict, whereaspolarizationand dissociativepolicies are much betterstrategies
for exploitedgroups.This is also reflectedin the doublenessof non-violentstrategies,all themesto
be morefully developedin Theoriesof Conflict(forthcoming).WhenSchmidsays (op. cit., p. 219)
that peaceresearch'should explain . .how latent conflictsare manifested- /and/how the present
international ystem is seriouslychallengedor even broken down' he seems to betraythe same
type of onesidednessthat he accuses peace researchof - interestin controlling manifestconflicts
only, in bringingabout integration,in formulatingproblemsin termsmeaningfulto international
This content downloaded from 133.30.212.88 on Tue, 07 Apr 2015 01:57:19 UTC
All use subject to JSTOR Terms and Conditions
Violence, Peace, and Peace Research 191
and supranationalinstitutions.But this onesidednesswill almost inevitablyresultif researchshall
be gearedto servethe interestsof specificgroups,high or low, insteadof the promotion of values.
It is as hardto believethat disintegration,polarization,dissociationis alwaysthe beststrategyas it
is to believethe opposite.
But this seems to be closely relatedto Schmid'sconflictology (op. cit., pp. 224-228), where he
seems to believethat I have a subjectivisticconception of conflict.If there is anything the conflict
trianglepurportsto achieveit is exactlythe opposite: the definitionof conflict independently
of attitudes and behavior, and also independentlyof perceptions of the situation held by the
parties(as differentfrom theirattitudesto each other).To me, conflictis incompatibilityof goals,
but how these goals are establishedis a quitedifferentmatter.To ask the partiesfor theirperception
of what they pursueand what, if anything,stands in the way is one, but only one approach.
I have nothing against definitions in terms of 'interests' the concept of 'goal' is wide enough to
encompass. The difficultyis, as Schmid readily and frankly admits (op. cit. p. 227) to 'decide
what the interestsare' and I sharewith him the idea that 'this is a challengeratherthan a reason
to abandon the idea of an interestdefinitionof conflict'.But I feel these interestshave to be postulated,as
I think Marxto a largeextent did, and then one has to explorethe implications.I also
thinktheycan be seenas expressionsof values,but not necessarilyheld by the actor,nor necessarily
held by the investigator,just as postulated values.Thus, if one feels it is contraryto the interests
of children,as autonomous humanbeings, to acceptthe tie as the childrenof theirbiological
parents,then there is certainlyan incompatibilityin the presentfamily system: parentshave interestsas
owners incompatiblewith the children'sinterestsas self-owners.The only differencebetween
this example and Schmid'smaster-slaveexample is that he gives a paradigmfor a conflict
of the past, I a paradigmfor a conflict of the future, and moreoverfor a conflictI thinkwill be
manifestedfairly soon, in line with the generalwave towardsdefeudalizationof the social order.
And I certainlyagreewith Schmidthat polarizationwill herebe a partof the solution.
36 For an effortin this direction,see Johan Galtung, CooperationinEurope(Strasbourg:Council
of Europe, 1968).
37 An effort to give some reasons why are found in 'Two Approaches to Disarmament: The
Legalist and the Structuralist',Journalof Peace Research1967,pp.161-195.
38 And it is of course not necessarythat all or most or much of this sails underthe flag of 'peace
research'or any other flag for that matter- only the slightly totalitarian minded would be inclined
to feel so. What is importantis that it is done, and that thereis contact between different
approachesso that they and others can benefitfrom ideological and institutionalpluralism.
This content downloaded from 133.30.212.88 on Tue, 07 Apr 2015 01:57:19 UTC
All use subject to JSTOR Terms and Conditions