Professional Documents
Culture Documents
1 - Waltz - Origins of War in NRM
1 - Waltz - Origins of War in NRM
1 - Waltz - Origins of War in NRM
Politics
KENNETI_I N. WALTZ
First published 2008
by Routledge
270 Madison Ave, New Yor*, NY I0016
Sirnultaneously published in d-re UK
by Routledge
2 Park Squarc, Milton Park, Abir-rgdon, Oxon OXI4 4RN
Routle dge is an imprint of the Taylor (t Francis Group,
an infonna business
Like most historians, rnany studcnts of inte rnrtional politics havc becn skepticirl
about thc possibiliq,of crcating a thcor).that rnight hclp onc to undcrstand and
cxplain thc intemational events that interest us. Thus Morgenthau, fbrcr.r-rost
among traditional rcalists, u,as fond olrepeating Blaise Pascal's remark that "the
histor,,, of the rvorld rvoulcl hlve been cliffbrent l'rird Cleopirtra's nose been a bit
shorter" trnd then irsking "Hou,cio you sl,stemize that)"1 His apprcciation of
the role of the accidental ar-rd the occurrcncc of thc unexpcctcd in politics
dampened his thcoretical anrbition.
The respclnse of ncorcalists is that, althougl-r difficultics abound, sor.nc of
the obstacles that seem most daunting lic in rnisapprcl-rcr.rsions irbont tl-rcor1,.
Tl'reory obviously cannot explain thc accidcr-rtal or account fbr nnexpectcc-l
e\rents; it deals in regularitics and rcpctitions and is possiblc or.rl1, if tl-rcsc can
be identifiecl. A firrtl-rer difficulq,is fbund in thc frrilr.rrc of rcalists to conceivc
of international politics as a distinct domain abor-rt u,hich tl-rcorics car-r bc f.rsl.r-
ioned. Morgenthlu, lbr exarnple , ir-rsistcd on "the autollomy of politics," but
hc fiilcd to appl)r the coucept to ilternational politics. A thcorl,is a depictiorr
of tl-re orgar-rizt.rtion of a dornair-r and of thc cor.rr.rcctiol'rs rntong its parts. A
theorl, indicatcs tl1i.1t some firctors arc r-norc ir-nportant than others and speci-
fies relatior-rs among then'r. In realih,, cvcrlrlfii1, is relatcd to everything clse,
and one don-rain cannot be separated from othcrs. 13ut the orl,isolates one reah-r'r
from ali others in order to dell u'ith it intcllcctualll,. S,1r 6[cfir]irrg the structllrc
of intcmational political svstcnls, r'reorealism cstablisl-rcs thc autonomv of
intcmtrtiontrl politics :rnc'l thus makes ir theorv about it possiblc.2
In developing a theorv olinternational politics, ncorcalism rctair-rs thc mr'Litr
tcrrcts of realpolitih, but me ans :urd encls arc viervcd diffcrcntl,v, as arc carlscs
ar-rd cffects. Morgcnthan, for example, thougl-rt of the "ratirxral" statesmln ls
ever striving to accurnnlate morc and morc po\\rer. IIe vieu,ed po\\'er irs lr-r cud
in itsclf. Altl-rough he:rcknorvledged that natiol-rs at timcs ilct or.rt of cor-rsidcr-
ations othcr than pou,cr, Morgcnthau ir-rsisted that, rvheu they do so, thcir
acti()l-rs arc not "ofla politicirl nature."3 In contrast, ncorcalism sccs po\\,cr as 11
possibly riscful means, \\.ith stites running rislis if thcy l.rave either too littlc or
too t-nuch of it. Excessivc u.cakncss r-r-rry invite an attack tlttt grerltcr strcrlgth
l'Hl'l ()ll.ICllNS ()l- WAIt IN N]rOltt.,\LIS I 'f rILOltY
lvould hlvc dissuirclccl rtn '.rclvcrsarv fi'or-n l:iunching. Exccssivc stl'cugtll nriry
promPt othcr strttcs to incrcasc thcir rrrurs irncl pool their ctlirrts rrgrrinst tl'rc
dor-nirtltrt state. Bccilusc po\\'er is rr possibll,nscfirl r-ncrrns, scnsible st.lresurelr
try to hi-rvc an r-tppropritrtc i.lmount ol it. In crLlcitl siturrtions, hori.cvcr, the
ultimate coltcenr of stiltcs is not for pou'cr bLrt tor sccuritr,. fl-ris rcvision is.in
ir-nportattt itt-tc.
An cvct-t lnore inrpol'tiurt rcvisiou is fixrnd in a shitt of crrusal rclations. -fhc
infir-ritc mittcrials ol rn1, rerrlnr can bc organizecl in cndlcsslv clilfcrent u,avs.
Realisr-r-r thinks olcruscs ls r-r-rcx'ir-rg ir-r on11,onc clircction, fronr thc intcractior-rs
of inciivicluals iurd states to the outcorncs that their rrcts irncl intcrrrctir)ns pro-
duce. Nlorgcr-rthau recoqnize d that, n'hcn the rc is cor-npctition firl sc.rrcc goocls
and no one to servc ls ubitcr, a stmgglc for pon'cr u,ill cnsr.rc ilnrong the con'r-
petitors trnd that consc(llrenth,thc strLlgglc fbr;ros'cr ciur bc c-rplrrincc.l u,ithout
refercnce to thc cvil born in mcn. The strugglc tirr pou,el lrises sin-rplv because
men \\,alrt tl.rir.rgs, not bccallsc ol the er.il in tl.rcir clcsires. He labclccl r.nrrn's
desire fbr scrrrcc goocls as one of thc tu'o roots of conflict, but, cve n n,hilc
discussing it, hc secn-tccl to pull tou,ard the "other root of cor-rf]ict irncl con-
comitant s1,il"-(c1he attinttts dominandi, thc dcsirc for pou,er." Hc oficn
considercc-l tltilt trrtn's clrir.e tbr pori,er is rnore basic than the chrrncc conclitirtns
under r.vhich strugglcs fbr pou,er occur. 'fhis attitr-rc'le is sccn in his statcr.ncnt
that "in a u,orlcl t,hcrc pou,er collnts) r'rrr rrltion pursuinu il rational policv hrrs
a choice betu,eeu rcr-rour-rcing r'urcl $,rrntins po\\,er; and, if it could,tltc lust ttrr
pou'er fbr tl're inclividr.rirl's sake rvould still confi'ont us u,ith its lcss spectrrculirr
yet no le ss pressine r-noral deftcts."r
Stuclcnts of intcrnational politics have rvpicrrlll, infcrrccl ourcorncs fl.ollt
salient attributcs ol the rrctors proc-lucing thenr. 'l'hr.rs Marxists, li[<e libcrals,
have linkcd thc outbrerrk of u,ar or thc previrlcncc ol pcilcc t() the intcmal
qualitics olstiltes. Got'entmentirl tbrr-ns, economic systcms, socirrl institutior-rs,
political idcologics-thesc irrc but a ten,e-rrrn-rples olu'hcre the c:iuscs olu,ar
have bect-t tbund. Yet, rrltl'rouul-r ciluscs ilre specific.rlll'assignec-I, u.c knou'tlrat
states lvith u'iclel1, clivcrgcnt cconor-r-ric ir-rstitutions, social custonrs, enrl polit-
ical ideologics hrrvc all tinght u,ars. More srrikins still, mrrnv clilfcre nt sorts of
organizations fight u,ars, u'licthcl thosc r>rqanizations be triLres,
1'rcttv princi-
palities, cmpires, niltions, ()r strcct gangs. If :rn iclcntificc'l conclition sccrlls to
have causecl ir givcn \\,ilr) ()ltc n'rust u,onclcr s,h1, 11,.',r, occur rcpeiltcrllt, cvcr-r
though thcir cattscs vur,. Vrrriations in the cl'r:rrirctcristics of tl-re strltcs r-rrc not
linked directh, to thc olltc()nres tl-rirt thcir bchrrviors proclucc, nor lrc vlri
atior-rs in thcir pilttcrlts ol interaction. Miir-ri, historirrns, tirr
cx;rntplc, hrrvc
claimed that Wcrrld \Yar I rvrs cr-use cl bi, tl-rc ir-rtcr:iction of
fir,o o;-rposccl rrncl
closely balar.rced coalitior.rs.
llut tl-rcn ,rr,,,r1, l-r,-,,,. clainrccl th:rr lYrrrlcl War II u,as
caused b1'thc firilurc
of sonrc stirtes t() cor-nbine fi;rccs in irn cflbrr to right rrn
imbalar-rce of pou,cr
crcatccl by irn cxistinq alli.uce .
rcalism. lly err-rpl-rasizing horv strulcturcs aflbct actions and outcomcs, ncorcal-
isr-n rejects the assumption that lran's innate lust for pou,cr constitrltcs a slrf-
ficient causc olrvar in thc absence ofan1,61he1. It rcconccivcs thc causal link
bet*,een interacting units and intemational outcomcs. According to the logic
of internatiorul politics, one must belicvc that some causes of international
olltcomcs are the result of intcractions at the ultit level, irnd, sincc variations in
presumeci causcs do rlot correspond verlr closcly, to variations in observed
olrtcomesl onc mllst also assumc that othcrs are located at the structural lcvc1.
Causcs at the level of ur.rits intcract rvitl.r those at thc lcvcl of structure, and,
because they do so, explanation at the unit lcvcl alone is bound to be n-rislcad-
ing. If trr-r approach allou's thc consideratior.r of both unit-lcvcl and structural-
lcvel ci.ruses, then it can cope u,itl-r both thc changes ar-rd the continuities that
occur it-t x s1r51grr.
Structural realism prcsents a systemic portrait of intcmational politics
depictir-rg component units according to the rlanncr of their arrangement. For
the purposc o[dcrcloping a tl]eor),, statcs trc crst as unilrry Jctors \\'autiug
at least t() sllrvivc, and are taken to bc the systelr's corlstittrent units. Tl're
essential stmctural quility of thc systcrn is anarchy-thc absence of a central
monopoly of legitimirte forcc. Cl-ranges of structurc and hence of system occur
u,ith variations in the numbcr of gretrt powers. Thc range of expected out-
comcs is inferred fron-r thc assumecl motivation of the units and thc structure
of thc svstem in u,l-rich tl-rcy act.
A systems theory of intcrnational politics dcals u,ith forces at tl-rc inter-
national, and not xt the national, level. With both s)rstems-level and unit-level
forces in plalr, horv cau ol-lc constrrlct a theor,v of internirtior-rrrl politics rvithout
simriltancoush, constmcting a theory of forcign policyf An ir-rternational-
political thcorv does not imply or require a thcory of foreign policl, ar-r1.morc
than a r.r-rarkct theory implics or requires a thcory of the firm. Syste n-rs thcories,
rvhether political or economic, are theories thirt cxplain horv thc organization
of ir realrn acts its x constraining and clisposing fbrce on thc ir-rtcractir-rg Lrnits
withir-r it. Sucl-r theories tcll us about the fbrccs to r.vhich the units are sub-
jected. Fror-n them, wc can dralv some infcrcnccs about the expectcd behavior
ancl fatc of the units: tramch', horv the1, u'ill havc to compete u'ith ar-rd adjust
to onc anothcr if thcy arc to sun'ive and flor-rrish. To tl-re cxtcnt that thc
dynamics of a sl,sten-r lin.rit the freedom of its i.u-rits, their bchar,ior and the
outcor-llcs of their behavior bccome predictablc. FIou, do \\,e expcct firms to
respond to diffcrently structrlrcd markets, and statcs to differently strlrctrlrcd
iutern:rtior.ral political s),stemsf Thcsc tl.reoretical qucstions require us to take
firms t-rs firms, and states as states, rvithout paying attention to diflerences
alTr()llg thcr-n. The questiolts are then ausu,cred by reference t() thc placemcltt
of the r.rnits in thcir s)/stern ancl not bv rcference t() thc internal qualities of
the units. Sl,stems theories cxplirin ri,h1, di16..,-rt ur.rits behavc similarly ar.rcl,
despite tl-rcir varirtions, producc oLltcomes that fall rvitl-rin cxpected rar-rgcs.
Convcrscll,, theorics xt thc rutit leve I te ll us rvhy ciif flbre nt ur-rits bchavc
F
.THE OITIGINS OF WAR IN NE,OREALIS'T THEOI{Y
negativc one: fb;rr of other states. Divergcllce colllcs u'hcn positive interests
are at issuc. L-r alliar-rccs amoltg ne.lr equtls, strategies arc irl\\'il)Is thc prclduct
of cor-npron-rise since the ir-rtercsts of rrllies irr-rd thcir notiolls of hou' to scclrrc
tlrcrn are ncr cr idcrtticrrl.
If con-rpctir-rg blocs arc sccn to bc closcll' b;rlancccl, ar.rd if cornpctitiou turns
or-r important matters) thcn to let onc's side clou'n risks one's ou'n destrucriot-t.
In a momcr-rt of crisis thc u'caker or the m()re advcuturous party is likely tcr
dctcrmine its sidc's policy. Its plrtncrs crln aflbrd ncither to lct the u'eakcr
mcmber be dcfcated nor to adve rtise tl-rcir disunitl, b1, fiiling to back a ventllrc
o,cn rvhile deploring its risks.
The preludc to World War I plovidcs striking cxantples olsuch a situetiotr.
Thc approxir.r-rate equality of prrrtncrs in trotl.r thc Triple Allilncc and Triplc
Entente t.uadc tl-rem closclf interclepentlcnt. This intcrdcpendencc) combinccl
rvith tl-re kccr-r competition bctu'eeu thc tu'o calttPs, mcallt that) althor.rgh anv
colrntr), could commit its rrssoci;ttcs) no ()nc coLtlttr)r on eithcr side coulcl
cxcrcise control. If Austria-Huugarl' t.uarched, Gcruranl' l'rad to fbllou'; thc
dissolution of thc Austro-Hungarian Er-npirc u'oulcl have left Gcmranl' aloue
in the middlc of Europc. If Frar.rcc t-t.tarchecl, Russia l.rad to follou'; a Gcrtt.ti.ttt
victory ovcr ts-rance rvoulcl bc l clefcat fbr Russia. Ar-rd so the vicious circlc
continucd' Becattse thc clcfci.rt or thc defcctio' of a nrirjor alh' *'otllcl h;rvc
shaken thc bahnce, cach stilte \\/r1s constrilil]ecl to adjttst its stratcglr ancl t[-re
use of its fbrce s to thc air-ns iurcl ttirs of its plrtucrs.
In alliar-rccs among cqtials, the clcfcction olonc nrcmber thrcate ns the sccnr-
it1. of the othcrs. In alliirnccs among unecluals, tlre contribtttions olthc lesscr
r-nembers afc at once \\'ilntccl irnd of rclativcly' srlall ir-nportattcc. Itt alli;rrlccs
alnong uncqnlls, alliancc lcaders ncccl rvorrr, littlc iibottt tl-rc fiitl-rfirh'ress of
their follou,crs, u,ho nsuirlll'l-ravc little ch<>icc ltrvn'rt1'. Coutrast thc sitttatiou in
l9l4 s,itl-r tl-rat of thc Unitccl State s and Rritain urd Frrrncc in I956. The Unitccl
States coLrld clissociatc itself liom thc Sucz rrdventurc olits nvo prir-rcip;rl allics
p
1'HL, oITIGINS OF WAlT IN NEOREALIST THEOI{Y
rcsist a chargc rhat anorhcr state tries to makc. As thc leadcrs ir-r a bipolar
systcm, tl-re United srates anc.l the Soviet Union are disposcd to clo
the rcsist_
ing, fbr in in.rporrant marters thc1, cannor hope that thcir allics rvill do
it fbr
them' Political action in the posnl,ar rvorlcl has reflectccl tl.ris colclitior.
comrnunist guerrillas operati.g i, Greece promptcd thc Trur.na, Doctrine.
Thc tightcni,g of soviet cortrol ovcr the srates oJ-Eastern Europc
led ro tire
Marshall Plar-r and the Atlantic l)efcr.rsc Treaq,, ar.rcl thcse ir-, t.r.,,
ga'e rise to
the cor-'ir.rtbrm and thc warsarv pact. Thc plar.r to crcate a welt
Ge.na.r
govcrnment produccd tl.re Berli, blockade. During the past
four decades, our
responses have bee, gcarcd to the Soviet union's actions,
and theirs to ours.
Miscalculatio, by some or arl of the great powcrs is a sourcc of
da.gcr i, a
multipolar r'r,orld; ovcrrcaction by either or both of tl-re great povucrs
is a solrrcc
of danger in t.r bipolar rvorld. Which is u,orsc: n.riscalculation or
overrcactio.f
Miscalculatio'r is the grearer e,il bccansc it is morc likely to permit
a, u.fold-
i,g of events that finall1'threatens thc status quo a,cr brings ihc powcrs
to \\rar.
o'erreaction is the lesscr c'il beca'sc ,t rr,.r.ri it costs only,roney
for unneces-
sary arms and possibly thc fighti,g of iimited rvars. The
clynarnics of a bipolar
system' morcover) provide a freasure of correction. L-r
a rvorlcl in r.vhich tg.o
states ,nited i, thcir m,tual antago,ism o'ershadou.
a,y others, thc be.efits
of a calculated rcsp.rsc stard out m.st crearly, ancl thc ,rrra,i.r,-r,
agai,st
irre sponsible behavior achicve their greatest force. Thus
two states, isolatio,ist
by traditio., unrurored in thc rvays of inter,ationar politics,
ar-rd fimed for
impulsivc behavior, havc shou,n themselves_not alrval,s
and cveryrvhere, but
alu'ilys in crucial cases-to be rvary, arert, cautious, flcxibre,
and forbcarir-rg.
Moreover, the cconomics of the great po\,vers in a bipolar rvorld
are ress
i,terdeper.rdent thirn those of the great pow..rs of a n-rultipolar
onc. The size of
grcat po$'ers tends to increasc as their numbers fhll, ancl
the largcr a state
is, thc grcirter the of its rcsources. states of continentar siic cro pro-
portior.rately less of'ariery
tl-reir busi'ess abroad than, for example, Britai., France,
and Germany dicl in thcir heydavs. Ncvcr befbre in modem
historl, have trre
grcat Powers dcpendcd so littlc on the outsicle u,orld,
and been so qlinvolr.ecl
in one a,other's economic affairs, as thc U'itcd States ar.rd the
So'ict Ur.rio,
havc bcen sincc thc rvtrr. Thc scparation of their intercsts
redtices the occasi.r-rs
for disp,te and pcrmits thcn'r, if thcy rvish, to lea'e cach other
alor-rc c'cr.r
thoueh each defi,cs its sccurity irtercsts rargely in terns ofthc
other.
of parties, diffusion or dangcrs, co,f,sio. of rcsponses:
-.Interdcpe'dcncc
These trre thc characteristics of great-powcr politi., ir.r
a multipolar rvorld.
Sclf:dcper-rdencc of parries, clarity of clangers, icrtirinry
about rvht has to facc
thcm: rhesc are the characteristics of great-po*,cr politics in
a bipolar,,vorld.
irigl-rer thc stalies ancl the closer a colrrltry comes to $'inning them, thc more
surely tl-rat colurrry invites rctaliation ar.rd risks its orvu destruction. The accu-
muitrtiorr ol significar-rt pou.cr through conqucst, even il only convcutional
\\,eapons arc uscd, is no lor.rger possiblc in thc tvorld of nuclcar porvers. Those
ir.rdividulls l,ho believc that the Soviet Union's lcaders .1re so bcllt otl rvorld
clot.uiuation that thcy nta)' bc u'illing to rlrn catastropl-ric risks for problenratic
gains firil to understand l-ron' govcrnmellts bchavc. Do tt'c cxpcct to losc one
ciq, or nvoi Ts,o cities or tcnf Wl.ren thcse are the pcrtinent qucstions, politi-
c;.rl lcadcrs stop tl-rir-rking lbout runuit'tg risks and start rvorrf ittg about ho\\' to
avoid thcm.
Deterrencc is more easily achieved than tt-tost militar.v strategists u'ould have
us bclicve. Lr tr convcntioual u'orld, A collntr\r can sensibly attack if it bclicvcs
that succcss is probablc. Itr a uuclcar u'orlc1, a couutrv callllot scnsibh' attack
unless it bclicvcs that succcss is assurecl. A natiot.r r.vill bc deterred liom attack-
ir-rg o'en if it bclicves that thcrc is onl1, a possibiliry that its adversary rvill
rettrliate. Uncertainq' of respousc, llot certainw, is reqr-rired fbr deterrencc
becausc, if retaliation occlrrs) one risks losing all. As Clauseu'itz wrote: If war
approtrchcs the absolutc, it becotnes imperative "not to take tl-rc first step
u,ithout tl.rir-rl<ir-rg u,hat ma1, be the last."u
Nuclear wcapons make thc inrplications cr,cn of victory too l-rorribie to con-
tcn-rplate. Thc problem that tl-rc t-tuclear pou,crs must soh'e is ho\\'to perPetuatc
pcace rvhen it is not possible to climinate :ril olthc causes oftvar. The structurc
of international politics has not been transfbnr-red; it remains anarchic in fbrm.
Nuclear stirtcs corltillue to compcte n-rilitarily. Witl"r each state striYing to ellsure
its ou,n securiw, u,ar remaius constrntl)r possible. In the anarchy of statcs,
impror.ir-rg thc means of defbrrsc and deterrencc rclative to the mcans of offerrsc
increascs thc chances of peacc. Wcapons and stratcgies that makc defbnse and
de terrcncc casicr, and offbr-rsivc strikcs l'rarder to moLlnt, decreasc the likelihood
of x.ar.7
Although the possibility of u'ar rcmains, the probability of a tvar involving
states u'ith nuclcar werpons has bcen drasticalll' rcduced. Over thc cettturics
great po\r'crs have fougl'rt ffrorc wafs than miuor states, alld the ficquencl' of
lvar has corrclatcd more closcll. u'ith tr structural characteristic-thcir inter-
national star-rdir-rg-thtrn u,ith unit-level attributes. Yet, becausc of a chauge
in militlrv tccllrology, a change at the unit lcvcl, r'vaging r,r'ar has increasingh'
becomc tl-rc privilege of poor aud u'e:rk statcs. Nuclear \\'eapons l-rave banisl-rcd
u.,ar fi-on-rthc ccnter of intcrr-ratiorul politics. A unit-level changc has dramatic-
r.rll1. rcduced a structural elfcct.
-fhc problbilitl, of mljor l'rrr xnrong states having nuclear \\'caPons
approachcs zcro. But the "rcal u'ar" trlay, as lirmes clain-red, lie in tl-rc prepar-
ations fbr u,agir.rg it. Thc logic of a dctcrrcnt strxtegv, il it is fbllou'cd, also
circurnscribes the ciruses of "rcal \\'alrs."3 In a couvetttional u'orld, the struc-
ture of internationll politics encouragcs statcs t() arm cot-npetitivcly. In a
nuclear u.orld, dcterrent stratcgies ofler the possibiliq. of dln'rpcr-rine thc
7
THI] ORIGINS OF WAR IN NEOITI.,AI,IS'I'1'HI.,ORY
Acknowledgment
Thc author thanks David Scl-rlcichcr, u'ho \\'i1s most hclpfirl in the cornplctior.r
of this article .
Notes
I Hans l.Morgcr-rthau, "Iuternation:rl Rclations: Quar-rtitative and Qualitatirc
Approacl-res," in Non-nan l). Pirln-rcr cd., A Dcsigtt for Internationnl Relatiorts
Research: Scope, Theory, Methods, and Relevnncr (Phitadclpl-ria,l970),p.78.
2 Morgerrthaq Politics among Nations (Ncu'York, 1973;Srh cd.), p. 11. Ludu'ig
Boltzmirn (trar-rs. Rudolf Weir-rgartncr), "Theories as Rcpre sentatious,"
cxcerpted ir.r Arthur Danto ancl Sidr.rey Morgcnbcsser eds.. Philonphy oJ'Science
(Cleveland, 1960), pp. 245-252. Ncorcalistn is sor.netitncs dubbcd strucftrrirl
realism. I usc thc tcrrns intcrchangcably, and, throughout this article, refbr tir
rrry o\\,n fbn-r-u-rlation ol neorealist tl-rcor1.. See Wirltz, Theory of Interno.tilnal
Politics (Rcadir-rg, MA, 1979); Robcrt I(eohrne ed., Neorenlistn and its Critics
(Nclr,York, 1986).
3 Morgenthau, Politics onxon! Nntions, y>. 27 .
4 Idem, Scientific Man vs. Pol,er Politics (Cl-ricago, l9a6), pp. I92, 200. Itrlics
trdclecl.
5 Sec lol-rn H. Herz, "Iclealist L-rtcrnationalisrr antl tl-rc Sccr-uiry Dllctnma," World
Politics, II (1950), pp. I57-180.
6 I(arl von Clauscu.itz, cd. Arratol Rap:rport; trans. J. l. Grirham, On War
(Hamn.rondsr'vorth, I 968 ), Y, p. 37 4.
7 Scc Malcolm !V. Hoag, "On Stability in l)eterrent Raccs," in Morton A. I(aplar.r
ed., The Revolu.tion in World Politics (Neu'York, 1962), pp. 388-410; Robcrt
)ervis, "Coopcration under the Securin,Dilcmma," World Politics,Xxx (I978),
pp.167-214.
8 William ]ames, "Thc Moral Equivirlent of !Var," iu Lcon Bramson and Gcorgc
W. Goetl.rals cds., Wnr: Studies frout Psl,chology, Sociology, and Antbropologl'
(Nov York, I9(r8; rer,. ed.), p. 23.
9 Cf'. Bernarcl Broclie, 'I'he Absohrtc Wrapltt: Atomic Pott,er nnd World Ot'der (Nctt'
York, 1946), pp.75-76.