Military Culture and Inadvertent Escalation in World War II
Author(s): Jeffrey W. Legro
Source: International Security, Vol. 18, No. 4 (Spring, 1994), pp. 108-142
Published by: The MIT Press
Stable URL: http://www.jstor.org/stable/2539179 .
Accessed: 16/05/2011 13:50
Your use of the JSTOR archive indicates your acceptance of JSTOR's Terms and Conditions of Use, available at .
http://www.jstor.org/page/info/about/policies/terms.jsp. JSTOR's Terms and Conditions of Use provides, in part, that unless
you have obtained prior permission, you may not download an entire issue of a journal or multiple copies of articles, and you
may use content in the JSTOR archive only for your personal, non-commercial use.
Please contact the publisher regarding any further use of this work. Publisher contact information may be obtained at .
http://www.jstor.org/action/showPublisher?publisherCode=mitpress. .
Each copy of any part of a JSTOR transmission must contain the same copyright notice that appears on the screen or printed
page of such transmission.
JSTOR is a not-for-profit 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 [email protected].
The MIT Press is collaborating with JSTOR to digitize, preserve and extend access to International Security.
http://www.jstor.org
MilitaryCultureand
Inadvertent
Escalation
in WorldWarU
Jeffrey
W. Legro
H
ow can the use of
"unthinkable" means of warfarebe avoided? How can states successfully
observe mutually desired limitationson "taboo" formsof combat?1These
questions are importantbecause of concern that nuclear,chemical,and biological weapons and terrorismwill spread and be used. The growingnumber
of states-e.g., Israel, Iraq, Libya, North Korea, Ukraine-that have such
means of inflictingharm increases the likelihoodthatany futureconflictwill
involve a desire for restrictions.Countries may pursue restraintbecause
popular opinion vilifiescertainweapons;2 because leaders calculate that escalationwould damage theirdomesticand internationalpoliticalsupport;3or
because states fear retaliatoryattacks. Unfortunately,even when nations
agree thatlimitationsare desirable, restraintdoes not always endure. A key
source of this disparitycan be found in accidentsand inadvertentescalation.
In contemporaryaffairs among major powers, the apparent absence of
grounds forintentionalaggression, against a backdrop of change and instability,makes the unintended expansion of conflicta centralconcern.4States
may not seek a spiral of hostilitybut stillcan stumbleinto escalation. Why?
Jeffrey
W. Legrois AssistantProfessor
ofPoliticalScienceat theUniversity
ofMinnesota.
I am gratefulto RobertJervis,Ido Oren, ArthurStein, RichardRosecrance,and two anonymous
reviewersfortheircommentson the ideas presented here.
1. The term"unthinkable"comes fromHerman Kahn's studyon nuclearwarfare,Thinking
About
theUnthinkable
(New York:Horizon Press 1962). Here the terms"unthinkable"or "taboo" indicate
the stigma attached by public opinion and the internationalcommunityto the use of certain
instrumentsof warfare.It does not mean thatstates,and especially theirmilitaryorganizations,
do not thinkabout, and plan for,theiruse in war.
2. Often thereseems to be littlelogic to such a stigma.For example, in World War II it became
acceptable to the allies to roast Japanese soldiers alive in caves with flame throwers,while the
use of gas forthe same task was considered by some to be illegitimate.There is a large literature
on the "just" use of forcefroma moral or legal perspective.See e.g., Michael Walzer, Justand
UnjustWars:A MoralArgument
withHistoricalIllustrations
(New York:Basic Books, 1977); Geoffrey
Best, Humanityin Warfare
(New York: Columbia UniversityPress, 1980); WilliamV. O'Brien, The
Conductofa Justand LitnitedWar (New York: Praeger,1981).
3. In the 1990-91 Gulf War bomber and missile attacks against targetsin Iraq, the U.S.-led
coalitionmade concertedeffortsto avoid civiliancasualties which mightshow up on television
and promote opposition to the war.
4. Accidentsand inadvertencehave receivedconsiderableattentionrecently.See ScottD. Sagan,
TheLimitsofSafety:Organizations,
anidNuclearWeapons(Princeton:PrincetonUniversity
Accidents,
International
Security,Vol. 18, No. 4 (Spring 1994), pp. 108-142
(C 1994 by the Presidentand Fellows of Harvard College and the Massachusetts Instituteof Technology.
108
MilitaryCultureand Inadvertent
Escalation1109
An intriguingset of cases fromthe Second World War offersnew insights
and leverage on the problem of taboo warfare and inadvertence.5In the
interwar years, three means of warfare were especially singled out and
denigratedas inhumane and illegitimate:submarineattacksagainstmerchant
ships, the aerial bombing of non-military
targets,and the use of poison gas.
At the outset of World War II, countriesexplicitlyrecognized and desired a
distinctlimitor "firebreak"between restraintand escalation in each of the
three means of warfare,despite the fact that all were considered militarily
significant.Shortly after fightingbroke out, however, submarine warfare
escalated beyond restrictions.Strategicbombing,restrainedat first,was later
employed extensively.In contrastchemical weapons, despite expectations
and preparations,were never used. Accidents and inadvertence,while not
always the main factor,were involved in each of these cases.6 How were
these incidentsallowed to occur? Why did some unintended events lead to
escalation, while otherswere brushed aside allowing restraintto endure?
I argue that the main existingtheoriesof inadvertence-Clausewitz's notionoffriction,the securitydilemma,and organizationtheory7-providepoor
predictionsforthe events of World War JJ.8I develop and test an alternative
approach, organizational
culture,that provides a betterexplanation. Military
cultures-beliefs and norms about the optimal means to fightwars-are
importantbecause they have a pervasive impact on the preferencesand
actions of both armies and states. While traditionalorganizational theory
Press, 1993); BarryPosen, Inadvertent
Escalation:Conventional
WarandNuclearRisks(Ithaca: Cornell
UniversityPress, 1992); Bruce G. Blair, The LogicofAccidentalNuclearWar (Washington,D.C.:
Brookings,1993).
5. Much of the literatureis limitedby its focus on the bipolar U.S.-Soviet nuclear competition.
The emergingmulti-polarnuclear world, however, makes the historicalexperienceof countries
other than the two superpowers increasinglyrelevant, especially if, as I argue, restraintis
influenced by national traits. The narrow concern with nuclear warfare precludes historical
analysis of variationsin use in this formof combat because of the welcome absence of instances
of inadvertentnuclear escalation. One study that takes a historicalcomparativeperspectiveon
escalation is Richard Smoke, War: Controlling
Escalation(Cambridge: Harvard UniversityPress,
1977).
6. My broader study CooperationUnder Fire (Ithaca: Cornell UniversityPress, forthcoming)
examines the overall patternof cooperation and conflictimplicitin the restraintand escalation
of World War II. This articlefocuses on the accidental and inadvertentelements of escalation
ratherthan the intentionaldecisions of states to violate limitations.
7. These threeare highlightedby Posen, Inadvertent
Escalation,pp. 12-23.
8. A fourthimage mightbe based on a realist"strategicrationality."However, the focus of this
essay is inadvertencewhere states do not desire escalation. From the perspectiveof a unitary
state, a strategicrationale for accidental escalation makes little sense. Nonetheless, strategic
rationalityis discussed at differentpoints below in order to betterdelineate the organizational
cultureapproach.
International
Security18:4 | 110
emphasizes the importanceof formalstructurein causing uniformmilitary
behavior, a culturalapproach contends that differencesin beliefcan lead to
dissimilaractions.9Militaryorganizationalculturesnot only influencewhat
types of accidents mightoccur,but more importantly,
what the implications
of those incidentswill be forescalation. Where specificmeans of warfareare
cultureofa country'skey military
compatiblewiththe dominantwar-fighting
services, that nation is likelyto take actions that contributeto escalation. In
such situations,the militarywill emphasize the antagonisticrole the other
side played, encourage propagandisticuse of the incident,and highlightthe
advantages in escalation. Yet when a type of warfareis antitheticalto one
side's militaryculture, that state will support restrainteven in the face of
provocative enemy incidents. It will suppress informationthat might encourage escalation, accept accidents as such regardless of evidence, make
effortsto communicategood will to the opposing side, and rejectany internal
proposals to seize propaganda advantages. In short,organizationalculture
leads to dynamicsin use and restraintthatare not predictedby the randomness of friction,the securitydilemma, or traditionalorganizationtheory.
This argumenthas implicationsforboth theoryand policy. Much of the
work in the securitystudies literaturehas emphasized the internationaldeterminantsof the use of forcesuch as the militarybalance, the strugglefor
security,or the prevailingnorms or laws of war.10This emphasis, however,
has tended to ignore importantdomestic determinantsof escalation. The
historicalcases below indicate the powerfulinfluenceof organizationaldy9. Whereas a cultural approach predicts that militariescan eitherfosteror inhibitescalation,
workin the traditionalschool such as Graham T. Allison, EssenceofDecision:ExplainingtheCuban
Missile Crisis (Boston: Little, Brown and Company, 1971); BarryPosen, The SourcesofMilitary
Doctrine:France,Britain,and Germanybetweenthe WorldWars (Ithaca: Cornell UniversityPress,
1984); Posen, Inadvertent
Escalation;and Richard Betts, Soldiers,Statesmen,and Cold War Crises
(Cambridge: Harvard UniversityPress, 1977), anticipates that all militarieswill favor offense
and tend to provoke escalation in war. Stephen Peter Rosen, WinningtheNext War:Innovation
and theModernMilitary(Ithaca: Cornell UniversityPress, 1991) challenges the monolithicview
of organizationsin his study of innovation,which emphasizes the politicsof promotion,measures of strategiceffectiveness,and the managementofinnovation.I argue thatcultureis critical
to each of these threefactors.
10. Sidney Verba, "Assumptions of Rationalityand Non-rationalityin Models of the International System," WorldPolitics,Vol. 14, No. 1 (October 1961), p. 115; Arnold Wolfers,Discordand
Collaboration
(Baltimore:JohnsHopkins UniversityPress, 1962), pp. 13-16; Theodore Lowi, The
End of Liberalism:Ideology,Policy,and the Crisis of Public Authority(New York: Norton, 1969),
pp. 158-160; Posen, SourcesofMilitaryDoctrine,pp. 59-79, 228-236; Charles Lipson, "International Cooperation in Economic and SecurityAffairs,"WorldPolitics,Vol. 37, No. 1 (October
1984), pp. 1-23; Benjamin Miller, "Explaining Great Power Cooperation in ConflictManagement," WorldPolitics,Vol. 45, No. 1 (October 1992), pp. 17-26.
EscalationI 111
MilitaryCultureand Inadvertent
namics withinnation-states.My contentionis not thatmilitariesalways tend
to fosterescalation, as traditionalorganizationtheorywould suggest. Rather,
the point is that the armed forces, depending on their culture, can either
reinforcerestraintor instigateescalation. In the past, countrieshave made
considerable effortsboth to reach diplomaticagreements(e.g., the Hotline,
the Incidents at Sea accords) and to develop technical procedures (e.g.,
permissive action links [PALS] on nuclear weapons) intended to control
unwanted events. My thesis suggests that leaders concerned with avoiding
undesired escalation in futureconflictsmust also pay attentionto and manage, not only the militaryorganizationsthatimplementthe use of force,but
also the beliefs and norms that characterizethose organizations.This is no
small task and it may demand a new systemof civil-military
relations,a topic
taken up in the conclusion.
I develop this argument by firstexploring the logic of three traditional
approaches to inadvertentescalation-friction,the securitydilemma, traditional organizationtheory-and the new one, organizationalculture. These
ideas, particularlythe organizationalculture perspective,are then assessed
in comparativecases involvinginadvertencein submarine,aerial, and chemical warfare between Britain and Germany in World War II. Finally, the
concluding sections draw togetherthe empirical evidence and outline the
importof the findingsfortheoryand policy on escalation and restraint.
ImagesofInadvertence
It mightseem contradictoryto speak of explanations of accidents and inadvertentescalation-phenomena that seem inherentlyunpredictable-but explanations do in factexist. There are two elements demanding explanation.
The firstis to account for the originsof accidents-those unintended and
unexpected events. The second issue, central here, is how to explain the
consequencesof accidents: why some lead to the widespread crossing of a
recognizedlimiton war, while othersare ignoredand restraintendures. Four
approaches offeranswers: Clausewitz's notion of friction,the securitydilemma, traditionalorganizationtheory,and organizationalculture.
FRICTION
The most widely accepted theory on accidents is Clausewitz's thesis presented in On War. He posits that accidents are unpredictable.His termfor
this is "friction"or "the fog of war." This concept asserts that a varietyof
International
Security18:4 | 112
factorscan impose themselves unexpectedlybetween plans and actual outcomes. For example, communicationand control of forces are difficultin
war; amidstthe chaos of combat,intelligenceis oftenuncertainor misleading;
soldiers get scared or tired and make mistakes. Clausewitz writes, "This
tremendous frictionwhich cannot, as in mechanics, be reduced to a few
points, is everywherein contact with chance, and bringsabout effectsthat
cannot be measured, just because they are largelydue to chance."11
His theme is that accidents are random and unavoidable under the demands of battle. Clausewitz's ideas make intuitiveand empiricalsense. By
theirverydefinition,accidentshave multipleand oftenunforeseeablecauses.
This thesis is somewhat blunt and only gives the most generalnotion of why
and how theymightcome about. Nonetheless, the logic ofClausewitz's ideas
suggests that unintended escalation is particularlylikelywhen the employment of forceis complex, when the battleis intense, and when information
(on what one should do, or what the enemy is doing) is uncertain.12
The mishaps of the 1980 U.S. raid to freethe hostages held in Iran exemplifythe complexity,intensity,and "fog" problems that can beset military
operations. The rescue effortwas a high-stakes,high-riskmulti-servicemission in unfamiliarenemy territory.Although the weather forecastwas for
clear skies, the helicoptersran into giant dust clouds. At the "uninhabited"
meetingspot in the middle of the Iraniandesert,therescue team encountered
a bus-load of Iranian travellers.Three helicopters sufferedproblems that
grounded them and resulted in the collapse of the overall operation. One
helicopterwas forced down by a crew member's simple mistake of leaving
a flakjacket or duffelbag over an engine cooling unit.13 No one could have
foreseen these particular difficulties,but as Clausewitz warns, given the
complexityand pressures of the operation, some undesired incidentswere
unavoidable.
11. Carl von Clausewitz, On War,ed. and transl.by Michael Howard and PeterParet (Princeton:
PrincetonUniversityPress, 1984), p. 120; also see pp. 113-122.
12. Modern day accident theoristsecho Clausewitz's thesisby positingthatunwanted incidents
will be more likelyin areas of warfarewhere thereis technicaland organizationalcomplexity.
Charles Perrow,NormalAccidents:LivingwithHighRiskTechnology
(New York:Basic Books, 1984),
pp. 3-4, 330-335. Perrow's argument is that complex organizations that deal with high-risk
technologieswill inevitablyincur accidents. An application of this thesis in the militaryrealm
is Chris C. Demchak, MilitaryOrganizations,ComplexMachines:Modernization
in theU.S. Armed
Services(Ithaca: Cornell UniversityPress, 1991). For a test of this thesis vis-a-vis the nuclear
weapons accidents of the United States, see Sagan, TheLimitsofSafety.
13. Paul B. Ryan, The Iranian Rescue Mission: WhyIt Failed (Annapolis, Md.: Naval Institute
Press, 1985), esp. chapters4 and 5.
Escalation| 113
MilitaryCultureand Inadvertent
THE SECURITY
DILEMMA
What RobertJervisand others have called the "securitydilemma" links the
structureand technologythat characterizea conflictsituation to the possibilityof inadvertence.14This dilemma is produced by the anarchicnature of
internationalrelationsthat causes distrustamong states: one states's defensive effortscan make others less secure. The resultinginsecurityleads to
action and reaction,producing spirals of hostilityand escalation, ending in
an arms race or the actual use of force.
It is easy to grasp how the insecurityand uncertaintycharacteristicof
ongoing armed conflictwould produce fearon each side thatthe othermight
abrogate a pledge of non-use in order to gain an advantage. The security
dilemma is magnifiedand escalation made likelywhen defensivecapabilities
cannot be distinguishedfromoffensiveones and offensehas the advantage.
This structuralcircumstanceseems particularlylikelyto lead to firstuse in
two ways. First,when a state's doctrinesor weapons depend on surprisefor
effectiveness,that countryhas an incentiveto undertakea firststrike.Recognizing this incentive,an opponent is likelyto be especially nervous about
its vulnerabilitiesand will keep a tightfingeron its trigger.15Leaders may
face a "use-'em-or-lose-'em"dilemma, in which theyperceive thattheirown
securityis endangered if they do not act first;this invites preemptiveescalation. Thus, according to securitydilemma logic, in war inadvertentescalationis likelywhen first-strike
or use-'em-or-lose-'emincentivesare present.
One example of the dangers of this dilemma comes fromthe fable of King
Arthurat the battleof Camlan; a negotiationbetween two suspicious armies
erupted into unpremeditatedslaughterwhen one soldier drew his sword to
kill a snake and others, thinkingthat a battle had begun, followed suit to
defend themselves.16Similarlyin 1890, duringa U.S. Armysearch of a Sioux
village, both soldiers and inhabitantshad a tense grip on theirrifles.When
one rifleaccidentally went off, a storm of unintended and unanticipated
gunfirewas unleashed; thisbecame known as the Battleof Wounded Knee.17
14. Robert Jervis,"Cooperation Under the SecurityDilemma," WorldPolitics,Vol. 30, No. 2
(January1978), pp. 167-214. This dilemma is also capturedby the notion of the "reciprocalfear
of surpriseattack"in Thomas C. Schelling,TheStrategy
ofConflict
(Cambridge:Harvard UniversityPress, 1980), pp. 207-208.
15. Jervis,"Cooperation Under the SecurityDilemma," pp. 187-205.
16. This comes fromHerman Kahn, On Thermonuclear
War,2nd ed. (Princeton:PrincetonUniversityPress, 1961), p. 525, citedin Stephen Van Evera, "Causes of War,"Ph.D. diss. (University
of Californiaat Berkeley,1984), pp. 40-41, where otherexamples of inadvertencecan be found.
17. Sagan, TheLimitsofSafety,p. 263, describes thisand otherexamples of accidentalescalation.
International
Security18:4 | 114
The securitydilemma notion captures how insecurityin certain situations
can lead to unintended outcomes.
ORGANIZATIONS:
TRADITIONAL
APPROACH
A third perspective on accidents and escalation derives fromorganization
theoryapplied to professionalmilitarybureaucracies.There are two variants:
one that is quite common in the literatureI will call "traditional,"while a
second relies on a culturalperspective. The traditionalview of militariesis
that,like all organizations,they seek to maximizeautonomyand size and to
reduce uncertainty.18
In the armed forces,these tendencies are expected to
produce certain common characteristics,such as that militarieswill prefer
offensivestrategiesand that they will resist civilian interventionin operaThe underlyingpremise of the traditional planning and implementation.19
tionalperspectiveis thatsimilarlystructuredorganizationswith similarfunctions should have similarinterestsand behavior.20
In a general sense, the traditionalperspectiveexpects escalation because
restraintcontradictsthe very nature of autonomy-seeking,offense-oriented,
war-winningmilitaryorganizations. Research has indicated that soldiers do
not necessarilydesire war, but that afterthe war is under way, professional
warriorsdo seek operational autonomy and are inclined to use all means at
theirdisposal. Gradualism and restraintcan cost lives and are inconsistent
with such hallowed militaryprinciplesas concentrationof forceand the goal
of totalvictory.21
From a traditionalorganizationalperspective,thereis little
reason to expect any dampening of escalation based on organizationalinflu18. Allison, EssenceofDecision,pp. 67-100.
19. Building on Allison's work, this is the interpretationgiven by Posen, SourcesofMilitary
Doctrine,pp. 41-59; Jack Snyder, The Ideologyof the Offensive:
and the
MilitaryDecision-Making
Disastersof1914 (Ithaca: Cornell UniversityPress, 1984), pp. 24-25; Stephen Van Evera, "Why
Cooperation Failed in 1914," in Kenneth Oye, ed., Cooperation
UnderAnarchy(Princeton:Princeton UniversityPress, 1986), p. 97; Van Evera, "Causes of War," esp. chapter 7; Leon Sigal,
Fightingto a Finish: The Politicsof War Termination
in the UnitedStatesand Japan,1945 (Ithaca:
Cornell UniversityPress, 1988), pp. 19-25. Snyder and Van Evera, while emphasizing the
structuraltendencies toward similaritiesamong organizations (e.g., militariesare offense-oriented), seem to allow for the possibilityof a defensive policy, depending on organizational
essence. The latterview is more compatiblewith the notion of organizationalculturedeveloped
below. On organizationalessence see Morton H. Halperin, Bureaucratic
Politicsand ForeignPolicy
(Washington,D.C.: Brookings,1974), p. 28.
20. Posen, TheSourcesofMilitaryDoctrine,p. 37 statesthismostexplicitly:"[Organizationtheory]
predictssimilarbehavior of units in the contextof similarstructures."
21. Betts,Soldiers,Statesmen,
and Cold WarCrises.
Escalation| 115
MilitaryCultureand Inadvertent
ence.22This perspective, however, would anticipate that accidents and inadvertentescalation are particularlylikelyin ways that are compatiblewith
the standard operatingprocedures of the armed forces-i.e., where military
organizationshave developed routinesforuse of a particularmeans of warfare.23The proposition that follows fromtraditionallogic is thatinadvertent
escalationis more likelyin any means ofwarfarewhere militaryorganizations
have developed routinesforuse of thatmeans.24
ORGANIZATIONS:
CULTURAL
APPROACH
An organizationalcultureview of inadvertentescalation predictsa different
dynamic. In the last decade, culture has emerged as a central concept in
organizationalresearch,primarilyin the fieldofbusiness management.25The
reason forthis development was dissatisfactionwith existingstructuraland
functionalorganizational studies such as those found in the traditionalapproach discussed above. More specifically,analysts were puzzled why Japanese firmsperformedso differently
(i.e., better)when theirformalstructures
were so similarto those of Westerncompanies. Many have contended that
the answer is organizationalculture:the patternof assumptions, ideas, and
beliefsthat prescribehow a group should adapt to its externalenvironment
and manage its internalstructure.26
22. Attestingto this expectationare the discussion and examples on organizationtheorygiven
in Posen, Inadvertent
Escalation,pp. 16-19.
23. For example, during the Cuban Missile Crisis, a routineU-2 mission went ahead as usual,
but the aircraftstrayed off-courseover the Soviet Union, contributingto tensions that nearly
led to a violent U.S.-Soviet clash. This and other examples are given in Allison, Essenceof
Decision,p. 141.
24. Van Evera, "Causes of War," chapter 7, presents a detailed case on how the organizational
dynamics of militariesfavorescalation.
25. For an overview of the early evolution of the concept of organizationalculture,see William
G. Ouchi and Alan L. Wilkins, "Organizational Culture," Annual Reviewof Sociology,Vol. 11
(1985), pp. 457-483; more recentlysee PeterJ.Frost,et al., eds., Reframing
Organizational
Culture
(Newbury Park, Calif.: Sage Publications, 1991). In securitystudies, the concept of culturehas
been applied in differentways. See e.g., Ken Booth, Strategyand Ethnocentrism
(New York:
Holmes and Meier Publishers,Inc., 1979); JackL. Snyder,TheSovietStrategic
Culture,R-2154-AF
(Santa Monica, Calif.: The RAND Corporation,1977); Colin Gray,NuclearStrategyand National
Style(Lanham, Md.: Hamilton Press 1986); Carl H. Builder,TheMasksofWarAnalysis(Baltimore:
JohnsHopkins UniversityPress, 1989). Three notable recentstudies are Elizabeth Kier, Culture,
Politics,and MilitaryDoctrine:Franceand BritainBetweenthe Wars (manuscript,January1993);
Thomas U. Berger,"America's ReluctantAllies: The Genesis of the PoliticalMilitaryCultures of
Japan and West Germany," Ph.D. diss. (MIT, 1992); Alastair I. Johnston,"An Inquiry into
StrategicCulture: Chinese StrategicThought, The Para Bellum Paradigm, and Grand Strategic
Choice in Ming China," Ph.D. diss. (Universityof Michigan, 1993).
26. The definitiongiven here is loosely based on Edward Schein, OrganizationalCultureand
Leadership(San Francisco: Jossey-BassPublishers, 1985), p. 9. Large organizations are rarely
International
Security18:4 1116
Organizational culture deserves attentionbecause it has an autonomous
influenceon the preferencesof militaryservices and ultimatelyon those of
states. Cultures are not simply reducible to the desires of individuals who
guide organizations,nor to the environmentalcircumstancesin which they
exist. Organizational cultures are more than the individuals that run the
organizations,in two senses. First,cultureis a collectivelyheld phenomenon.
It is not generally the reflectionof a single leader nor is it some simple
mathematicalaggregation of many individual beliefs. Second, instead of
individuals changing cultures, the reverse is usually the case: people are
socialized by the beliefs that dominate the organizationsof which they are
part.27Those who heed the prevailingnorms are rewarded and promoted.
Those who do not are given littleauthorityor are fired.28
Cultures are also not mere weathervanes to environmentalforces or to
The number of large companies that have failed to
"strategicrationality."29
adapt to marketchanges are legion.30Organizationalbeliefsoftendetermine
which externalcircumstancesget attentionand how costs and benefitsare
weighed. Cultures act as a heuristicfor organizationaldevelopment, much
the same way a theoreticalparadigm can shape intellectualthought. They
provide a limitinglens forinterpreting
and selectingwhat is importantamidst
uncertainty.31
Environmentaldata and factswhich contradictculturewill be
characterizedby one culture,but have several. Often,however,especiallywhen the organization
is hierarchicallyordered, a dominant cultureprovides the main creed.
27. One theoristhas noted that one "does not live formonths or years in a particularposition
in an organization,exposed to some streamsof communication,shielded fromothers,without
the mostprofoundeffectsupon what he knows, believes, attendsto, hopes, wishes, emphasizes,
fears,and proposes." HerbertA. Simon, Administrative
3rd ed. (New York: Free Press,
Behavior,
1976), p. xvi.
28. For many years, failureto wear a white shirtand dark suit at IBM had hazardous consequences foran employee's career. The company has no writtenpolicy on attire.But according
to one formerexecutive, there was "an unwrittendress code that's as effectiveas if it were
engraved in steel-or as if it had a loaded gun behind it." See F.G. "Buck" Rodgers, The IBM
Way(New York: Harper and Row, 1986).
29. The issue of where culturecomes fromand how it changes is a broad topic thatis generally
outside the scope of this paper. I do, however, provide some evidence below that military
culturesare not simply a product of the strategiccircumstancesthese organizationsconfront.
For a more extensive discussion of organizationalinnovation see Barbara Levittand James G.
March, "Organizational Learning," AnnualReviewofSociology,Vol. 14 (1988), pp. 319-340.
30. Many tryto change theirculturesby large-scalepersonnel changes at the top. See Thomas
C. Hayes, "FalteringCompanies Seek Outsiders," New YorkTimes,January18, 1993,pp. Cl and
C4.
31. Mary Douglas and Aaron Wildavsky, Cultureand Risk (Berkeley:Universityof California
Press, 1982), argue that the risks people face and the ways risks are assessed are a product of
an earlierculturalchoice.
Escalation| 117
MilitaryCultureand Inadvertent
discounted as deviant and will be discarded. Likewise, resources are channeled to methods suited to culture,which consequentlyappear more feasible
than those deprived of fundingand attention.32
Finally,culturespersist for
utilitarianreasons: it can be difficultand expensive to reorientoperational
philosophy,especially in a large and complex organization.33
The organizational culture view posits that the pattern of assumptions,
ideas, and beliefs that prescribehow a militarybureaucracyshould conduct
battlewill influencestate preferencesand actions on the use of thatmeans.34
Each service, repeatedlyfaced with tough decisions about how, where, and
when to employ violence, develops a culturethatsets prioritiesand allocates
resources.35Where the traditionalapproach would expect all organizational
activitiesto be equally likelyto resultin inadvertentescalation, the cultural
variant contends that accidents and escalation are likely in those means
compatible with beliefs about the "rightway" to fightwars. These "paradigms" provide maps foraction thateitheradvocate or ignore specificmeans
of warfare.Those means compatiblewith the dominantwar-fighting
culture
will be developed and advocated by the military.In such areas doctrine,
preparations, and intelligence will be geared towards use, not restraint.
Furthermore,those accidents that do occur will be seized on as proof of
intentionalenemy use or of an unavoidable intensificationof the war that
must be met in kind or bettered. In those types of warfarethat are incompatible with the dominantculture,therewill be littleplanning and advocacy
fortheiruse.36More attentionwill be given to avoiding accidents; those that
32. This is the "competencytrap"where experiencewith,and sunk costs in, a certaintechnology
or means make it seem bettereven if alternativesare actually superior. See Levittand March,
"Organizational Learning," p. 322.
33. David Kreps asserts thatcorporateculture,even if some costs are involved, has a beneficial
functionalrole of facilitatingcommunicationand coordination. David M. Kreps, "Corporate
Cultures and Economic Theory,"in JamesE. Alt and Kenneth A. Schepsle, eds., Perspectives
on
PositivePoliticalEconomy(New York: Cambridge UniversityPress, 1990), pp. 90-143.
34. The degree of impact a particularorganizationwill have on executive decisions seems to
vary along at least three dimensions: 1) the extentto which it has monopoly power on issue
expertise;2) the complexityof its responsibilities;and 3) the time period available for action.
Militariesare particularlyinfluentialin wartimebecause they generallyhave monopoly control
over expertiseon the use of force,militaryoperations are complex and not easily understood
by non-specialists,and time periods for alteringpre-arrangedplans are limited. For a more
detailed discussion of this topic, see Legro, Cooperation
UnderFire.
35. Organizational cultures can be discerned in a varietyof sources including interviews,the
memoirsof participants,doctrinaldevelopment, organizationalcorrespondence,planning documents, and internalexercises.
36. For example, the traditionalviewpoint has difficulty
explainingthe defensiveorientationof
some militaries(e.g., the French Army)in the interwarperiod. For an excellentanalysis of this
issue, see Kier, Culture,Politics,and MilitaryDoctrine.
Security18:4 | 118
International
do occur will be accepted as such and managed so as to avert an escalation
spiral. In short, the organizational culture perspective predicts a directalbeit subtle-link between militarybeliefs and customs and the likelihood
of inadvertentescalation.
WorldWarII: Submarines,
Bombing,and ChemicalWeapons
Each of these perspectiveson inadvertentescalation tells a differentstoryof
why states unintentionallycross fire-breaksin the midst of an on-going
conflict.How the four approaches account for the actions of the two main
antagonists of World War 11-Britain and Germany-is examined below in
more detail. The cases deal with three areas of combat-submarine attacks
on non-combatants,the bombing of civilians, and chemical warfare-that
were taboo means of combat during the interwarperiod. These cases all
occurredwithin the same conflict,and thereforedifferencesin what was at
stake cannot explain the outcomes. They include both situationswhere inadvertentescalation took place as well as when it did not.
These cases involve two aspects of accidents: theiroriginsand theirconsequences. In the submarine warfaresection, the main focus is on origins
and covers three situations that had differentoutcomes: an accident that
broke a taboo, an accident did not involve taboos, and an "accident waiting
to happen" thatwas avoided. Withregardto strategicbombingand chemical
warfare,the main interestis the consequences of the accidents: why did
some lead to escalation while othersdid not? I pay more theoreticalattention
to organizationalculturebecause it is a newcomerand untested;nonetheless,
evidence for competing explanations is considered, particularlywhen they
offera compellingcontrastingpredictionof the event.
SUBMARINE
WARFARE:
ACCIDENT
VS. NON-EVENT
The submarinein World War II became known as the "viper of the sea" and
its use against merchantand passenger ships was reviled. At the London
Naval Conference of 1930, rules that had been formulatedand generally
approved at the 1922 Washington Naval Conference were accepted into
internationallaw. These rules prohibitedsubmarinesfromsinkingmerchant
or passenger ships withoutprovidingforthe safetyof the crews and passen-
EscalationI 119
MilitaryCultureand Inadvertent
gers.37This hampered the submarinesomewhat because it mightbe vulnerable to aircraftand otherships when heeding the required search and safety
provisions. Nonetheless, all the major powers supported the treaty.Despite
the collapse of other negotiationsand agreements,in 1936 the major powers-including Britain and Germany-reaffirmedtheir commitmentto the
rules in the London Protocol.
A close inspection of three situationsof submarinewarfareinvolvingGermanyand Britainsuggests how organizationalculturecan generateaccidents.
The most infamous accident of World War II submarine combat was the
sinking of the passenger liner Atheniaby a German U-boat at the start of
hostilities.In the second case, one Britishsubmarine sank another,and the
thirdwas a historicalnon-event: a Britishsubmarine adhered to the rules
despite strong incentives to torpedo a German liner. The four approaches
offerdifferentpredictionson these cases. Clausewitz's frictionthesis would
anticipate an equal likelihood of escalation, given that the situations were
comparable in intensity,complexity,and uncertainty(e.g., clarityof battle
instructionsand abilityto identifyfriendor foe). Similarly,traditionalorganization theorywould also predictescalation in all cases because both navies
had organizational routines torpedoing ships and were eager to join the
fight.Conversely, the securitydilemma image would predictno accidental
escalation: the submarinesfaced neitheruse-'em-or-lose-'emnor preemption
incentives.That is, theywere not put at riskthemselvesby not attackingbut
instead waiting to confirmthe identityof the enemy as eithercombatantor
civilian,as directedby internationallaw.
Organizational culture provides the best explanation in that it correctly
predicts that there would be escalation in Germanywhere unrestrictedattacks were central to naval warfarethinking,but not in Britainwhere the
Royal Navy belittledthe value and threatof anti-tradesubmarine warfare.
Here is what happened.
GERMANY S SUBMARINE
ACCIDENT.
When GermanystartedWorld War II by
attackingPoland, strictorders had been issued to the U-boats that the war
was to be conducted with meticulous restrainttowards merchantand passenger ships in accordance with internationalagreements. Hitler hoped to
37. Richard Dean Burns, "Regulating Submarine Warfare,1921-1941: A Case Study in Arms
Control and Limited War," MilitanyAffairs,Vol. 35 (April 1971), pp. 56-62; Janet Manson,
DiplomaticRamifications
of Unre'stricted
SubmarineWarfare,1939-1941 (New York: Greenwood
Press, 1990), pp. 33-52.
Security18:4 | 120
International
avoid conflictwithBritainand America.38In WorldWar I, unrestrictedattacks
by German underwaterboats had increased animositieswith the formerand
led to hostilitieswith the latter,a developmentthatsignificantly
contributed
to Germany's defeat.39Hitler hoped to avert a repetitionof history and
thereforerequired restrictionson submarinewarfare.
At the startof hostilities,eighteen of Germany's twentysix ocean-going
U-boatshad already takenup action stationsaround the BritishIsles.40One of
the submarines was Joseph Lemp's U-30 on patrol 250 miles northwestof
Ireland. On the evening of September3, 1939, the day Britaindeclared war,
Lemp located a potential targetand identifiedit as enemy. The vessel was
indeed British,but it was the passenger liner Athenia.The crew aboard the
Atheniaknew that war had been declared, but they did not worry about
U-boat attacksbecause they believed the ship was protectedby the London
Protocol.41 Lemp's U-30, however, spit out two torpedoes that burst the
vessel. 1088 passengers took to lifeboats,112 went to the bottom.The sinking
of the Atheniacontributedto the onset of unrestrictedsubmarinewarfare.42
How can this incident be explained? It is probable that the "fog of war"
had somethingto do with it. The young captain Lemp was probably tense
in the face of possible enemy contactand not thinkingas clearlyas he might
have been. It is unlikelythat he purposely blasted a passenger liner. Since
accidents are inherentlyundesired and are more likely in times of tension
and confusion,it is difficultto "falsify"Clausewitz's frictionthesis. Maybe
Lemp just made a mistake.But simplyto blame "operatorerror"is to conflate
human presence with causation.43We must ask if there was evidence to
suggest thatfactorsotherthan pure chance under complexitywere at work,
38. F.H. Hinsley, Hitler's Strategy(Cambridge, U.K.: Cambridge UniversityPress, 1951), pp.
4-9.
39. Ernest R. May, The WorldWar and AmericanIsolationism,1914-1917 (Cambridge: Harvard
UniversityPress, 1959), pp. 113-136.
40. Oberkommando des Kriegsmarine,"Chronikdes Seekriegs:Heft1 (1939 and 1940)," Berlin,
PG 32610B, Roll 4078, National Archives,Washington,D.C. (hereafterNA).
41. Edwin P. Hoyt, Death oftheU-Boats(New York: McGraw-HillCo, 1988), pp. 15-16.
42. Britainclaimed that the incidentwas intentionaland implementedits own set of defensive
measures, some of which were violations of the spiritif not the letterof the London Protocol.
German submarinersused the Britishmeasures to argue to theirsuperiors that they must be
allowed to attackmerchantvessels withoutrestrictions.Germanywas largelydoing so against
Britainby October of 1939, while Britaindid not do so against Germanyuntilthe springof 1940.
43. When accidents happen thereis a tendencyto focus on human erroras a cause, ratherthan
on other underlyingcauses that would tend to produce incidentsregardless of the individual
involved. See Perrow,NormalAccidents,pp. 9, 23-30, 330-331, 339.
Escalation| 121
MilitaryCultureand Inadvertent
and whether certainpredilectionstended to push randomness and human
in a particulardirection.
fallibility
The environmentof combat that evening was not particularlysuited to
"friction"in the formof Lemp's misidentification.
Although Clausewitz had
much more in mind than the weather, the analogy of "fog" does not fit.144
The eveningwas clear,makingmisidentification
unlikely.Nonetheless, Lemp
claimed that he thoughtthe Atheniawas an auxiliarycruiserbecause it was
zig-zaggingand showing no lights,neitherof which was true. Furthermore,
Lemp's U-boat closed to such a short distance that it could not easily have
missed the outline of the safetyboats and the lack of guns that marked the
ship as a passenger liner.45Hitler's demand for restraintwas also clearly
communicatedto the U-boat commanders. The submarineswent to sea with
orders to heed the submarinerules, and this orderhad been repeated at the
veryoutset of World War 11.46Finally,therewere no strongsecuritydilemma
incentives to escalate: the Nazi submarines were not at risk by avoiding
attacks on non-combatantsand would stillbe able to impair Britishtrade if
they decided to do so at a later time.47In Lemp's particularcase thereis no
evidence thathe thoughtthe Atheniahad detectedthathis U-boat was in the
area. Thus the source of the U-30's action was not a perceived "us or them"
dilemma as if, forexample, the Atheniahad been a cruiserthat spotted the
U-boat, forcingLemp to choose to kill or be killed. Neither frictionnor the
securitydilemma accuratelycapture the dynamicsof the Atheniaaccident.
44. Clausewitz, On War,p. 120 writes, "One [source of friction],forexample, is the weather.
Fog can preventthe enemy frombeing seen in time,a gun fromfiringwhen it should, a report
fromreachingthe commandingofficer.Rain can preventa battalionfromarriving,make another
late by keeping it not three,but eighthours on the march,ruin a cavalrychargeby bogging the
horses down in mud, etc."
45. Manson, DiplomaticRamifications,
pp. 64-66 and note 44; Peter Padfield, Donitz: The Last
Fuhrer(London: Golanz Ltd., 1984), p. 191.
46. The orders were laid out to the commanders as they put to sea in August and Donitz
reminded his captains of that order by wire on September3, 1939. See "Opbefehl Nr. 2 furUBoote 'AlarmiubungNordsee' (U27, U30), Kiel 21.8.39," PG32012-NID, NA; Padfield, Donitz,
p. 191.
47. They were more vulnerableby followingthe rules than by not doing so; however, restraint
was centralto Hitler's grand strategy.With immediateunrestrictedwarfare,the U-boats could
have scored a few easier kills rightat the beginningbeforeBritaincould organize its convoys
and defensivemeasures, but no significantstrategicadvantage was expected fromsuch action.
Germany only had some 26 ocean-going U-boats at the startof war, just one-thirdof which
could normallybe on stationat a time. Donitz argued 300 would be needed to get the job done.
See "Gedanken uiberden Aufbau der U-Bootswaffe,"Memo by Donitz, September3, 1939, RM
7/891,Bundesarchiv-Militararchiv,
Freiburg,Germany(hereafter,BA-MA).
Security18:4 | 122
International
From an organizationalcultureperspective,however, the incidentis less
puzzling. The cultureof the U-boat corps ofwhich Lemp was a partis central
to his subsequent "mistake."48 To a degree not seen in any other country,
Germany had a vibrantpro-submarineculturewithin its maritimeforces.49
Even though this culture was suppressed for a good part of the interwar
period due to internationaltreatiesand internalpolitics,the submarineranks
served as a magnet for some of the most ambitious and talented officersin
the Navy. Karl Donitz, commanderof the U-boats, sought to infusehis men
withan offensive,anything-is-possible
spirit.He led the revivalof the World
War I U-boat doctrinebased on an anti-tradeoffensive,one that seemed at
odds with the pledge to adhere to the submarinerules.50
In light of this culture, the Atheniaincident is not surprising.Lemp was
zealously implementinghis training.Germany's"sea wolves" were firstand
foremosttaughtto be aggressive,and not to miss opportunities.51 The notion
of differentiatingattacks was not ingrained in the training of the
U-boat mariners,which is why they had to be given instructionson the eve
of war that explained the procedures forheeding the submarinerules.52But
Lemp violated the restrictionsby sinking the Atheniain a situation where
there was every reason, except organizational predisposition,to show restraint.53
48. In thiscase, as in otherswhere the military'spreferredway of war favorsuse and escalation,
the distinctionsbetween culturaland traditionalorganizationtheoryargumentsare not discernible and thereforeI do not address the latter.
49. Germanyalso had a strongbattleshipculture.But the submarinehad a respected tradition
and valued role. This contrastssharply with the interwarexperience of countriessuch as the
United States and Britain,which ignored the underwaterboat, despite its potentialvalue in a
war with Japan.
50. TerrenceRobertson,NightRaideroftheAtlantic(New York:Dutton, 1956), p. 16; Befehlshaber
der Unterseeboat,Kriegstagebuch,September15, 1939, RM 87/3,BA-MA; Karl Donitz, Memoirs:
Ten Yearsand TwentyDays, trans. R.H. Stevens (New York: World, 1959), pp. 12-13. Padfield,
Donitz,pp. 158-160.
51. Padfield, Donitz,p. 196.
52. See testimonyof FregattenkapitanHessler (Donitz's son-in-law)at Nuremburg. Trialof the
Major War CriminalsBeforethe International
MilitaryTribunal,Nuremburg:Vol. XIII (New York:
AMS Press, 1971), p. 528.
53. One naval staffmember advocated giving submarines permission to sink darkened ships
withoutany warning.Due to the "politicalsituation,"the possibilitythatanotherincidentwould
generate internationalopposition, this could not be completelyapproved. The suggestion was
made thatthe Navy leadership give its "silentapproval" to attackdarkened ships in areas where
only Britishvessels operated. The one conditionwas that the submarineshad to claim in their
war diaries that any sunken merchantvessels had been mistaken forwarships. This is what
Lemp claimed and what Padfield,Donitz,p. 193, says was officialNavy policyin such situations.
See "Forderungen des B.d.U. und militarischeMoglichkeitender Durchfiihrung,"September
Escalation| 123
MilitaryCultureand Inadvertent
When news of the Athenia'sdestructionreached Germanyon the day that
Britaindeclared war, Hitler and Raeder, the head of the Navy, were convinced that it could not have been caused by a German U-boat, given the
aims of the Third Reich and the explicitinstructionsissued. Only Donitz,
the trainerof the U-boat captains and crews, thoughtthatit mightverywell
have been one of his captains, despite instructions,who sank the passenger
liner.54
The U-boats' war-fighting
dogma meantthatsuch an "accident" could
be expected.
BRITAIN S SUBMARINE
ACCIDENT
AND NON-EVENT.
Britain'sbehavior in submarinewarfarecontrastssharplywith thatof Germany.The Royal Navy had
dramaticallydifferentbeliefs about the utilityof the submarine. In Britain,
the battleship was considered the "final arbiter" of naval combat and it
dominated the war-fightingcultureof the BritishNavy.55This culture,embodied in doctrine and plans, belittled submarine warfare, particularly
against commerce. The underwaterboat was regarded as the tool of weaker
powers, not of mightyBritain.56Sight unseen, the submarine could strike
without warning at undefended merchantvessels and even at the proud
warships. Not only was it a threatto the war fleet,but it also required that
warships be engaged in less heady tasks, such as accompanying convoys,
when theyotherwisewould be seeking battle. The Royal Navy did not train
submarinersnor develop boats to attackcommerce.Submarineswere meant
to be used primarilyforintelligenceand, when they were lucky enough to
get the opportunity,occasional attackson enemy warships.57
22, 1939, RM 7/844,BA-MA. Donitz and Wagner testifiedat Nuremburgthat this memo was
writtenby a staffofficerand that the Navy never forwardedsuch an order.
54. AnthonyMartienssen,Hitlerand His Admirals(New York: Dutton, 1949), p. 23. The leaders
did not find out what actually happened until Lemp returnedfrompatrol at the end of the
month.
55. "Final Report of the Post-WarQuestions Committee,"March 27, 1920, ADM 1/8586,Public
Records Office,Kew, UK (PRO) as cited in Stephen W. Roskill,Naval PolicyBetweentheWars,
Vol. I: The PeriodofAnglo-American
Antagonism,
1919-1929 (London: Collins, 1968), p. 115.
56. The originof thisbias was not simplygeo-strategicadvantage; i.e. thatBritainwas an island
power dependent on sea tradeconfronting
Germany,a continentalstatethatwas less vulnerable
to sea interdiction.Afterall, the Royal Navy's main expected opponent untilthe late 1930s was
Japan, a countryvery vulnerable to an anti-commercecampaign. Nonetheless, despite the fact
that the UK did not have and could not affordthe battleshipsto take on Japan in the Pacific,
the Navy did not seriouslyconsider using cheaper submarinesto blockade the island.
57. The Britishbelieved that the submarine was not a threatto battleshipsbecause U-boats
would be detected with the sonar device called "ASDIC." But because of the cultural bias
stressingthe inferiority
of submarines, ASDIC did not receive adequate criticaltestingand its
flaws were not appreciated. David Henry,"'BritishSubmarine Development and Policy, 19181939," Ph.D. War Studies (King~sCollege, Universityof London, 1976); on ASDIC see pp. 320321.
Security18:4 1124
International
The influenceof cultureon unintended Britishactions is especially apparent when contrastedwith the German navy's actions. As with Germanyand
the Atheniaincident,the firstsubmarine attackby Britainwas also an accident. On September 10, 1939, the Tritonmistakenlytorpedoed another En-
glishsubmarine,Oxely.58
was actingin accordance
Like the U-30,theTriton
with its organizational culture. Each inadvertentlydestroyed a ship, but it
was the type of vessel each had been trained to attack. Britishsubmarines
had a legacy of successful anti-submarinewarfare(ASW) in World War 1.59
In comparison, German strategywas based on an anti-merchantoffensive.
This differenceappears to have been reflectedeven in the physical structure
of the submarines. The acoustical detection array in German U-boats was
designed for an anti-shippingrole, while in Britishsubmarines it was for
attackingotherunderwaterboats.60Froman organizationalcultureviewpoint
it is not surprisingthat the first"accident" of the war fora Britishship was
the destructionof a friendlysubmarine,while Lemp's errorwas an "out of
bounds" passenger liner.
The differencebetween Britain and Germany is just as evident in the
accidents that did not happen. Britishsubmarines incurredno Atheniaincident despite opportunities.61For example, when the British submarine
Salmon sighted the German passenger liner Bremen2000 yards away on
December 12, 1939, a replay of the incidentinvolvingthe Athenia(which had
been only 800 yards fromthe U-boat) seemed likely.Yet the Salmonsurfaced
and ordered the ship to stop forthe search and seizure proceduresthatwere
mandated by internationallaw. Unexpectedly,however, a Luftwaffeplane
appeared and chased the submarine off.62Ironically,the Bremenwas being
used as a troopshipat the time and thereforewas a legitimatetarget,which
58. The storyof thisincidentis told in A.S. Evans, BeneaththeWaves:A HistoryofH.M. Submarine
Losses(London: William Kinder, 1986), pp. 195-199.
59. The day beforethe Oxelywas destroyed,AdmiralWatson, the commanderof the submarine
force,requested thathis boats be used more in an ASW role. "The Use of Submarinesin Defence
of our Trade in the Atlantic,"From RA(S) to Secretaryof the Admiralty,September 9, 1939,
ADM 199/1920,PRO.
60. Rear Admiral G.W.G. Simpson, PeriscopeView (London: Macmillan London Ltd., 1972),
p. 61.
61. The Britishsurface fleet did clear much of the German commerce from the seas at the
beginningof the conflict.
62. The captain of the Salmonnoted, "I had no special instructionswithreferenceto intercepting
Bremenand considered myselfbound by internationallaw, a rigidadherence to which had been
specificallystressed to submarine commanding officersat the beginning of war." See "HMS
SalmonPatrol Report December 2-16, 1939," Memo fromCommanding OfficerHMS Salmonto
Captain (S) Third Submarine Flotilla,ADM 199/288,PRO.
MilitaryCultureand Inadvertent
Escalation| 125
the Salmoncould have easily torpedoed fromthe safetyof the ocean depths.63
The differencebetween Britain's and Germany's accidents is inexplicable
froma traditionalorganizationalperspective. Both navies had adopted the
London Protocol as part of theirorganizationalrules, but beliefs on how to
fighta war-and the typesof accidentsthatoccurred-diverged. As a cultural
approach expects,Britain'sbehaviorreinforcedits restraint,while Germany's
provoked an accident thatviolated taboos.
STRATEGIC
BOMBING:
VIOLATIONS
IGNORED
OR RECIPROCATED
From20,000 feetoffthe ground, the familiarsurroundingsof life-buildings,
homes, cars-appear unnaturallysmall. Perhaps this perspectiveeased the
task of the young aviators of World War II who were ordered to flattenthe
homes and habitationsof enemy civilianshundreds of miles behind the front
lines. Althoughthis practicebecame commonplaceduringWorldWar II (and
in many conflictssince), in the 1920s and 1930s it was considered barbaric
and potentiallyavoidable. Statesmen made considerable effortsboth to reduce air armamentsand to find ways to regulateair attacksby agreeing on
rules and restrictions.64
The main distinctiontheyhoped to enforcewas that
between civilians and combatants. No officialtreatieswere concluded, but
Britainand Germanywere able to reach accord at the beginningof the war
to avoid bombing undefended civiliansand cities.65That pact retainedlegitimacy for the firstnine months of conflict,but as we will see, it did not
survive the war.
In the strategicbombing cases, organizationalculture played the pivotal
part, not necessarilyby causing accidents as in the submarinecases above,
but instead by affectingthe responses of states to the incidents-i.e., either
to use accidentsas justificationforescalationor,by ignoringthem,to reaffirm
restraint.Consider, forexample, the differencein the responses of Germany
and of Britainto the other side's transgressionof a major limitationin the
air war. In one case, Britain'sbombing of the German homeland broughtno
63. Nigel John Gilbert,"BritishSubmarine Operations in WWII," UnitedStatesNaval Institute
Proceedings,
Vol. 89 (March 1963), p. 73.
64. These were serious efforts.Britaineven considered giving up her most effectivemeans of
sea warfare,the sea blockade, in exchange forrestrictionson air warfare.See especially C.I.D.
Limitation of Arms Sub-Committee, 2nd Meeting, July 18, 1938, in "Humanization of Air
Warfare,"AIR 9/202,PRO.
65. Those participatingdirectlyin the war effortwere generallyseen as legitimatetargetsof air
power. J.M. Spaight, Air Powe&and War Rights,3rd ed. (London: Longmans, Green, 1947),
pp. 43, 259.
Security18:4 1126
International
comparable response; Germany virtuallyignored the event. Yet when Germany accidentallytransgresseda second limitationon the bombingof capital
citieslaterin the war, Britainresponded immediately.How can these different outcomes be understood?
Organizational culture predicts correctlythat Britainand Germany had
differentways of warfarethat accorded differentimportanceto a strategic
offensiveaimed at civilians. That contrasthad a considerable impact on
decision making and events. The otherperspectivesare less precise in their
predictions. The frictionhypothesis would anticipate escalation fromboth
participants,particularlyin lightof the technicalcomplexitiesof navigation
and precision bombing in the early war period. Likewise, traditionalorganization theorywould also expect both to escalate since both the RAF and
Luftwaffehad organizational capabilities and plans for strategicbombing.
The securitydilemma argumentexpectssimilarbehaviorfromthe opponents,
but in the opposite direction,toward restraint:there was not a strongsurprise-attackor preemptive incentive to initiatestrategicbombing since no
single attack could cause a devastatingamount of damage, and there were
few penalties involved in continuingrestraintto ascertainif the enemy had
actually-even if accidentally-violated restrictions.Yet while Germany
showed restraint,Britainescalated.
On the nightof May 11GERMANY S RESPONSE TO BRITAIN S FIRST ATTACKS.
12, 1940, RAF bombers undertook, possibly accidentally,the firststrategic
raid of the war in an attack on Monchen-Gladbach in Germany. Another
RAF assault on Aachen and Monchen-Gladbach took place on May 14-15.66
The officialgo-ahead forthe RAF strategicoffensivewas given on May 15.
What is of interesthere is the German response: therewas none. The Germans undertook no immediate retaliation,in kind or otherwise, against
Britain,neitherforthe earlier,possibly accidental,raids, nor in response to
the approved offensiveafterMay 15. The Britishhistoryof the Luftwaffe
pointsout thatin the entire1940WesterncampaignthroughthefallofFrance,
66. Because the Cabinet was debating on a daily basis whetherto undertakethese raids, H.W.
Koch, "The StrategicAir OffensiveAgainst Germany:The EarlyPhase, May-September1940,"
TheHistoricalJournal,Vol. 34, No. 1 (1991), p. 127, has speculated thatBomberCommand acted
on its own authority.The evidence, however,is not conclusive. The factthatlittleirritationwas
expressed at the Cabinet meetingon May 15 when the Secretaryof State forAir mentionedthe
May 14-15 raids suggests that,at least by thatdate, some sortoflimitedoperationwas approved.
See W.M. (40) 123, Conclusions, Minute 2, May 15, 1940, CAB 65/13,PRO; MartinMiddlebrook
and Chris Everitt,The BomberCommandWar Diaries: An OperationalReference
Book 1939-1945,
p. 42.
MilitaryCultureand Inadvertent
Escalation| 127
German bombers were not used strategically,except during one four-day
period in Junewhen they attacked the French aircraftindustryin Paris and
fuel dumps in Marseilles.67The Nazis were, of course, aware of the RAF's
attacks as was noted in reportsby the securityservice of the SS.68 Hitler
himselfallegedlydismissed the raids at one pointas inadvertent.He assumed
that the attacks on German territory
were the result of someone losing his
head due to the pressure of the Battleof France, or that the RAF had acted
on its own. He saw no need to retaliatein kind.69Despite the factthatit had
been made clear that its restraintwas contingenton reciprocity,Germany
simply ignored the Britishactions. Why did reciprocalescalation not occur
immediately?
A pivotal factorin this restraintwas the organizational culture of the
Luftwaffe.Unlike the RAF, Luftwaffefaithin strategicbombing-particularly
in a civilian-targetedmorale bombing campaign-never took hold. Germany's strategicculture, of which the Luftwaffewas a part, was land-oriented and heavily influencedby a traditionalarmy outlook. Like the RAF,
the Luftwaffewas an independent service in the sense that it was organizationally separate fromthe Army and Navy and not subordinate to their
orders. Nonetheless, cultures oftenrun deeper than formalstructuresand
the Luftwaffewas constrainedby Germany'scontinentalorientationto combat. The German Air Force did not prepare equipment or plans to wage the
type of large-scale air assault required forbombing and particularlyforan
unrestrictedcampaign. Instead, doctrine was oriented more towards supportingthe land battle.70Even the German heavy bombers best suited for
67. BritishAir Ministry,The Rise and Fall of the GermanAir Force,1933-1945 (New York: St.
Martin's Press, 1983 ed.), p. 72. There is, however, a debate on whether the prior German
attackson Warsaw and Rotterdamwere against "defended" citiesand thus permissibleor were
instead simplyillegitimateterrorraids. E.g., see Olaf Groehler,"The StrategicAir War and its
Impact on the German Civilian Population," pp. 282-283; cf. Horst Boog, "The Luftwaffeand
IndiscriminateBombing up to 1942," p. 386, in Horst Boog, ed., The Conductof theAir War in
theSecondWorldWar:An International
Comparison(New York: Berg PublishersLtd., 1992).
68. See Koch, "The StrategicAir OffensiveAgainst Germany,"p. 127.
69. As reportedby General Warlimontto WalterAnsel, HitlerConfronts
England(Durham, N.C.:
Duke UniversityPress, 1960), p. 113.
70. WilliamsonMurray,TheLuftwaffe
(Baltimore:Nautical and AviationPublishingCo. of America, 1985), esp. pp. 1-23, argues thatGermanydid have a strategicbomber emphasis. My view
is thatwhile a strategicmission-orientedsub-culturecertainlyremaineda part of the Luftwaffe,
it did not take rootand dominate the organizationas in Britain.Germanplanningand operations
consistentlylisted the hierarchyof Luftwaffe'saims as: 1) the destructionof enemy airpower;
then 2) support of the Armyand Navy; and finally3) tasks which mightbe considered strategic
bombing. See "Instructionsof the Commander of the Air Force forthe Conduct of Operations
in the InitialPeriod of War," November 18, 1935, as reprintedin Karl-Heinz V6lker,Dokumente
Security18:4 | 128
International
strategicbombing were required to have dive-bombingcapabilities,a specificationthat hindered their strategicbombing effectiveness.71 Thus when
Britain'sbombs fell on the Ruhr in the spring of 1940, Germany ignored
them. The Luftwaffewas absorbed fulfillingits established role of helping
the ground forces advance in assault on France. Even when France was
defeated and the Luftwaffewas no longer occupied with its part in that
victory,German restraintin strategicbombing endured.72
BRITAIN S RESPONSE TO GERMANY S ACCIDENTAL
RAID ON LONDON.
Compare
the German reactionto escalation with the Britishresponse to an accidental
German raid that appeared to breach the restraintstillin effecton bombing
capital cities. On the night of August 24, 1940, twelve German bombers
overshottheirintended targetsconsistingof aircraftfactoriesand oil refineries located at Rochester and Thameshaven, twentymiles east of London.
Instead they dropped theirloads on London, settingoffa chain of reprisals
that ended any hope forrestraintin strategicbombing in the Second World
War.73
Britaindid not ignorethe event as Germanyhad done in Britain'sbombing,
but instead seized on it as an act that required response in kind. The next
day Churchillcalled forretaliationand some 100 bombers were dispatched
explains that
against Berlin. Websterand Frankland's StrategicAir Offensive
escalation was motivatedby the prime minister'sdesire that "the Germans
get as good as they were giving."74 Perhaps in the heat of battle (with its
und Dokumnentarfotos
zur Geschichte
der DeutschenLuftwaffe
(Stuttgart:Deutsche Verlags Anstalt,
1968), esp. p. 478; "Aufmarsch-und Kampfanweisungen der Luftwaffe:Weisungen fur den
Einsatz gegen Osten," May 1939, RL 2 II/21, BA-MA; FuhrerWar DirectiveNo. 16, "Preparations
fora Landing Operation Against England," July16, 1940.
71. Richard J. Overy, "From 'Uralbomber' to 'Amerikabomber':The Luftwaffeand Strategic
Bombing," Journalof StrategicStudies,Vol. 1, No. 2 (September 1978), pp. 168-169; Edward L.
The ReichAir Ministryand theGermanAircraft
Homze, ArmingtheLuftwaffe:
Industry,1919-1939
(Lincoln: Universityof Nebraska Press, 1976), pp. 63-68.
72. Strategicrationalitypartiallyaccounts for this decision: Hitler sought a peace with Britain
so that he could turn his forcestoward the East. But this does not explain why, even afterit
became clear in Julyof 1940 thatBritainwould not yield and Germanywould have to fight,the
Luftwaffe'sassignment was firstto defeat the RAF, then to assist the army and navy in an
invasion. Terrorbombing was to be used only in retaliation.
73. Koch, "The StrategicAir OffensiveAgainst Germany,"p. 137; Boog, "The Luftwaffeand
IndiscriminateBombing to 1942," p. 389. The Fuhrer was enraged that his orders had been
disregarded, even if it was an accident. As punishment,the bomber crews responsible were
sent to the infantry.This punishment contrastssharply with the minimal rebuke that Lemp
received forviolatingthe submarinerules by torpedoingthe Athenia.Cajus Bekker,TheLuftwaffe
WarDiaries(New York: Doubleday, 1968), p. 172.
74. Charles Websterand Noble Frankland, The StrategicAir Offensive,
1939-1945 (London: Her
Majesty's StationeryOffice[HMSO], 1961), p. 152.
MilitaryCultureand Inadvertent
Escalation1129
attendant"fog"), Churchillsimplydid not understandthatthe Germanaction
was inadvertent.There is evidence to suggest, however, that Churchillwas
aware that this was an accident and was looking foran excuse to startcity
bombing. Britain may have known frominterceptedmessages that Hitler
had forbidden the Luftwaffeto bomb London.75 Yet even without such
intelligence,there was good reason to suspect an accident. The unintended
foray involved twelve planes which caused light damage and only four
fatalities,hardly the type of decisive operation to be expected froma purposeful breach of this importantlimitation.76
Churchillhimselfhad earlier
downplayed the gravityof the German raids, noting that very few people
were affectedby any one attack.77Moreover,in July,well beforethe German
assault, Churchillhad already shown an interestin bombing Berlin. At that
time he expressed interestin being able to respond to German attacks on
London. But he also gave a planning date of September 1, suggesting that
his intentionswere not necessarilydependent on German actions.78
Why did Britaindecide to escalate, breakingthe last taboo against strategic
bombing? This is a question of considerable historicalcontroversyand not
one easily answered with the evidence available. While there are many
argumentsabout what motivated Churchill(not the least of which was his
personal predilectionto seize the initiative),what is clear is that his outlook
and options were influencedby RAF culture.79From the end of World War
75. David Irving,Churchill'sWar:TheStruggleforPower(Australia:VeritasPublishingCo. 1987),
p. 365, especially note 30. This informationis based on an interviewwith R.V. Jonesthatis not
corroborated.However, it is not unthinkablethatsuch knowledge was gleaned fromintercepts
in theSecondWorld
of the signals trafficof the Luftwaffe.See F.H. Hinsley, BritishIntelligence
War, VolumeI: Its Influenceon Strategyand Operations(New York: Cambridge UniversityPress,
1979), pp. 179-182.
76. Harvey B. Tress, BritishStrategicBombingPolicy Through1940: Politics,Attitudes,and the
Formation
ofa LastingPattern(Lewiston, N.Y.: Edwin Mellen Press, 1988), p. 68.
77. Martin Gilbert,WinstonS. Churchill,Vol. VI: FinestHour, 1939-1941 (London: Heinemann,
1983), pp. 602-603.
78. Minute to Secretaryof State forAir and CAS fromPrime Minister,July20, 1940, AIR 19/
458, PRO; Minute fromDirectorHome Officeand CAS, July21, 1940,AIR 19/458,PRO. Churchill
also noted the desirabilityof waiting,in case of the need to targetBerlin,forlonger nightsand
the arrivalof the new Stirlingbombers. Gilbert,WinstonS. Churchill,Vol. VI, p. 673. Churchill
invitedPortal, the head of Bomber Command, to his countryhome to discuss the idea on July
20 and August 17. See Irving,Churchill'sWar,pp. 371, 403.
79. George Quester, Deterrence
Before
Hiroshima:TheAirpowerBackground
ofModernStrategy
(New
York: Wiley and Sons, Inc., 1966), pp. 117-118, suggests a strategicrationaleforescalation. At
the time,FighterCommand was under pressure due to Luftwaffeattacks.Churchillrecognized
that command of the air was the key to Britain'sdefense: if FighterCommand failed, Britain
was lost. Thus to buy Fighter'Command breathingroom, it is argued, Churchill purposely
attacked Berlin in order to draw the Luftwaffe'sattacks on the Britishcapital and away from
Security18:4 j 130
International
I, Britain's Air Force had promoted and institutionalizeda philosophy of
strategicbombing, whose central tenets were that the best way either to
preventor to win a war would be to threatenor launch a massive assault on
the enemy's sources of power. This included both deprivingthe enemy of
the physical means to fightand breakingits morale to continue the battle.
Hugh Trenchard,the Chiefof the Air Staff(CAS), had no doubt about which
was more important:"The moral effectof bombing stands undoubtedly to
the materialeffectin a proportionof 20 to 1."8? This philosophy,effectively
A
cultivatedin the organization,was well-suitedto unrestrictedwarfare.
The RAF's preferencesmade themselves feltin a number of ways. Air
Force officialshad been directlylobbyingforescalation since the invasion of
France, arguing that the battle had to be taken to the German homeland.
RAF Intelligenceboldly asserted that large "moral effects"were resulting
fromits bombing operations, a conclusion that seems to have been driven
more by wishful thinkingthan objective evidence. Influencedby these arguments,the Chiefsof Staff(COS) concluded on May 25, 1940,thatGermany
mightbe beaten by economic pressure, the bombing of economic and psychological targets,and the instigationof popular revoltin German-occupied
territories.By late June, Churchillhad picked up on this thinking,arguing
thatairpowerwould cause Hitler"possibly decisive difficulties"
in Germany
and other areas he had to feed and defend.82On July8 he asserted: "There
is one thing that will bring him back and bring him down and that is an
absolutely devastating, exterminatingattack by very heavy bombers from
the RAF. This thesis seems plausible with hindsightbecause it reflectswhat actually resulted.
Yet directevidence to support it is sparse. In addition, as FrederickM. Sallagar, The Road to
TotalWar(New York: Van Nostrand Reinhold Co., 1969), pp. 181-182, has noted, the decision
was made at a time, August 24, when FighterCommand was not in terribleshape as it would
be two weeks later.
80. Websterand Frankland, The Strategic
Air Offensive,
pp. 46, 55; Neville Jones,The Beginnings
ofStrategic
Air Power:A HistoryoftheBritishBomberForce,1923-1939(London: FrankCass, 1987),
p. 34.
81. One way the culturespread was throughthe selectionof personnel. Trenchardwas known
to have kept on only those officerswho agreed withhim. Robin Higham, TheMilitaryIntellectuals
in Britain:1918-1939 (New Brunswick,N.J.: Rutgers UniversityPress, 1966), p. 200. Perhaps
more important,institutionswere founded that would propagate the bomber offensivephilosophy, including an air force staffcollege, a cadet college, technicaltrainingschools and other
facilities.
Air Offensive,
82. Websterand Frankland,TheStrategic
pp. 145-146; Gilbert,WinstonS. Churchill,
Vol. VI, p. 603. On the general issue of RAF bias in intelligencesee Harold L. Wilensky,
and Policyin Government
and Industry
Organizational
Intelligence:
Knowledge
(New York:Basic Books,
1967), pp. 24-28.
Escalationj 131
MilitaryCultureand Inadvertent
this countryupon the Nazi homeland. We must be able to overwhelmthem
by this means, withoutwhich I do not see a way through."83
The organizationalcultureview suggeststhatChurchill'sviews and choices
were a consequence, not a cause, of organizationalplanning. His opinion of
the utilityof bombing had evolved froma pre-warhistoryof opposition to,
and lack of faithin, the independent strategicair offensiveaimed at morale.84
In May when he became prime minister,Churchillwas particularlyexposed
to the organizationallobbyingon the few air plans available. Perhaps more
important,the situation and RAF culture constrainedthe options available
to Churchill: Britainwas most prepared to strikeout at Germany through
strategicbombing.85Given the RAF's militaryadvice, the interpretationof
events, and the limited capabilities available (strategic area bombing),
Churchill's choice of strategicbombing is largely explained by a cultural
perspective. Without the RAF's bomber culture,Churchillmightwell have
shared Hitler's dispositionagainst unrestrictedair operations.The difference
in the compatibilityof each side's air forceculture with strategicbombing
explains why Britainand Germanyhad opposite responses to incidentsthat
violated restrictionson the use of aerial force.
CHEMICAL
WARFARE:
BARKING
DOGS
SILENCED
Like submarine warfareand strategicbombing, chemical warfare(CW) was
a forbiddentool of conflictduring the interwaryears. The limitationof CW
83. Gilbert,WinstonS. Churchill,Vol. VI, pp. 655-656. At about the same time (July17), Portal,
the Commander-in-Chief(CINC) of Bomber Command, personally advocated unleashing the
bomber offensive.RAF Narrative,TheRAF in theBombingOffensive
AgainstGermany:Volume II,
Restricted
BombingSeptember1939 to May 1941 (Air History Branch, Air Ministry),AIR 41/40,
p. 117, AIR 41/40,PRO.
84. At the end of World War I, Churchill doubted that victorycould be had by terrorizing
civilians. Webster and Frankland, The StrategicAir Offensive,
p. 47. In the 1930s, Churchill
advocated air defenses as a means of mitigatingair attacks. He did not believe the "bomber
would always get through." See Winston S. Churchill,The Gathering
Storm(Boston: Houghton
MifflinCo., 1948), pp. 147-152; Tress, BritishStrategicBombing,pp. 69-70. In September 1939,
citingthe results of the Spanish Civil War, Churchilldoubted that "the essential elements of
war" would be changed by the air arm. On May 7, 1940, Churchillhad opposed unrestricted
bombingbecause of Britain'sperceived inferiority
to Germanyin air power. See W.M. (40) 114,
Conclusions, Minute 1, May 7, 1940, CAB 65/13,PRO. AfterSeptember 1940, Churchillvaried
between doubt and support for the air offensive.See Maxwell Philip Schoenfeld, The War
MinistryofWinstonChurchill(Ames, Iowa: Iowa State UniversityPress, 1972), pp. 92-101.
85. But Britain'sair power resources could also have been used to attack Germany's invasion
effortor help with the battle at sea. See Tress, BritishStrategicBombingPolicy,pp. 215-220. On
the uses of air power in the Battle of the Atlantic,see Williamson Murray,"The Influenceof
Pre-WarAnglo-AmericanDoctrine on the Air Campaigns of the Second World War," in Boog,
ed., TheConductoftheAir War,pp. 245-246.
Security18:4 1132
International
was discussed at many of the negotiationsof thatera. But only one treatythe Geneva Protocol of 1925 prohibitingfirstuse of chemicalweapons-was
signed. At the beginning of World War II, Britainand Germanyexchanged
pledges reaffirmingtheir no-first-usecommitments.Despite these agreements, nations expected CW use and went to considerableeffortspreparing
for such combat. CW was widely recognized by all participantsin World
War II to have significantmilitaryutility.86However, throughoutthe war,
even though a range of CW accidents occurred,theynever led to escalation.
The friction,traditionalorganization, and securitydilemma perspectives
all would predicta spiral of use fromrestrictedweapons even ifby accident.
Intense conflict,complex operations, and great uncertaintycharacterized
many of the situationswhere CW incidentsoccurred. These sources of friction helped cause accidents, but did not cause escalation. All of the military
organizationsinvolved also had plans, troops,and weapons forCW use but,
contraryto traditionalorganization theory,theirroutines did not generate
escalation. According to the securitydilemma argument,both sides should
have feltinsecure due to the surprise-attackadvantage of CW, should have
been poised to strikeback to minimizedisadvantage, and should have been
leery of any type of trustin enemy restraint.It was widely acknowledged
thatgas was most effectively
used in a surpriseattack. For example, General
Ochsner, the head of German CW in World War II, argued that where
attackerand defenderare equally well prepared, the attackerhas the advantage because of abilityto completepreparations,achieve surprise,and choose
86. This is evident in the analysis and calculations of individual countriesand outside experts.
For example, in Britaina 1939 review of gas requirementsconcluded that "with added and
improved weapons chemical troops will be used in a futurewar more than they were in the
last." See "Gases for Use in the Field and the Quantity of Each Required," prepared by the
Directorof MilitaryTrainingand Directorof StaffStudies by request of the Intra-serviceCommitteeon Anglo-FrenchChemical WarfareConversations,July7, 1939,War Office193/740,PRO.
From December 1939 tests, the Britishconcluded that, "we have at our disposal a potential
weapon of great value." See "Chemical Warfare-High Spray Trials," fromMO1 to DDMO,
January30, 1940, WO 193/726,PRO. Militaryintellectualssuch as B.H. Liddell Hart and J.F.C.
Fuller were also proponents of CW.
87. Gas was not, however, a use-it-or-lose-it
weapon, nor was CW just an "offensive"weapon.
It could be used forboth offenseand defense. Generallyitwas seen as benefitingthose interested
in inhibitingfast moving operations afterit had been introduced to the battlefield.But initial
use could facilitatean offensiveby opening wide gaps in enemy lines, as was the case with
Germany's firstuse of CW in World War I. Britishplanners in World War II argued that CW
might be effectivein breaking through enemy lines if the offensivetowards Germany were
stalled in Italy or France. See JointPlanning StaffMemorandum, "MilitaryConsiderations
Affectingthe Initiationof Chemical and Other Special Forms of Warfare,"July27, 1944, PREM
3/89,PRO.
Escalation1133
MilitaryCultureand Inadvertent
the time, place and scale of attack. The defenderwould be hard pressed to
keep defenses ready and would have to fightunder the most unfavorable
conditions.88Inadvertent escalation would seem likely in these circumstances. However, the historyof incidents involving chemical warfare in
WorldWar II-that is, its non-use-challenges thisnotion.The organizational
cultureapproach argues that both Germanyand Britainhad "ways of war"
thatworked to suppress inadvertentescalation. Indeed, thatis what we find.
BRITAIN AND THE ABSENCE OF CW INADVERTENCE.
Of the armed forcesof
Britainaffiliatedwith chemical warfare,none found it a desirable weapon,
and organizationalroutines forits use were not accorded the attentionand
fundingotherareas received. Given theirprevailingorthodoxyof war-fighting, this is not surprising.In the Army,chemicalwarfaredevelopment was
relativelyignored forthreereasons. First,the legacy of gas use in WorldWar
I had alienated mainstreamofficersto this formof warfare.They did not like
the interferenceof civilian chemists, the special privilegesaccorded to the
companies that were tasked to wage CW, and the way thatgas complicated
the traditionalbattlefield.Second, the Army's conservativeapproach to innovation worked against the acceptance of chemical weapons. The Army
was but a loose collectionof traditionalregiments,like sportsor social clubs,
thatsoldiersrelishedas a refugefromsocial and technologicalchange. Within
this system,the technicalor mechanical officerwas looked down upon. Gas
was a technical weapon. The Director of Artillerywas left in charge of
chemicalwarfare,yet artillerymenseemed more concernedwith theirhorses
than with their technical equipment. Finally,the Army had few resources
and no central war scenario. This inhibited development of weapons like
CW that were affectedby the specificgeographical and climatologicalconditions of the area in which theywould be used.89
88. Lt. General Herman Ochsner, HistoryofGermanChemicalWarfare
in WorldWarII, PartI: The
MilitaryAspect,P-004a (HistoricalOfficeof the Chief of the [U.S.] Chemical Corps, 1949), p. 4,
NA.
89. See L.F. Haber, ThePoisonousCloud:ChemicalWarfare
in theFirstWorldWar(Oxford:Clarendon
Press, 1986), pp. 269, 273; J.B.S. Haldane, Callinicus:A Defenseof ChemicalWarfare(New York:
Dutton, 1925), pp. 34 and 37; Brian Bond, BritishMilitaryPolicyBetweenthe Two WorldWars
(Oxford: Clarendon Press, 1980), pp. 35-71, 132; Shelford Bidwell and Dominick Graham,
Firepower:
BritishArmyWeaponsand Theories
ofWar1904-1945(London: George Allen and Unwin,
1982); chapter9, and p. 180; and M.M. Postan, D. Hay, and J.D. Scott, Designand Development
of Weapons(London: HMSO, 1964) pp. 238-240, 253. It was recognized that the effectof CW
depended on local weather arid geographic conditions. See CID, "Chemical WarfarePolicy,"
November 1924, p. 15, WO 188/144,PRO.
International
Security18:4 1134
RAF thinkingwas more compatiblewith gas use yetCW developmentwas
in the hands of the Army and it received littleattentionor advocacy in the
Air Ministry.The RAF recognized a potential role for CW, but when it
appeared that CW development would reduce fundingformore preferred
tools, such as bombing with high explosives and incendiaries,the RAF was
willing to forgo the option.90The individuals that did advocate CW were
censured by higherauthorities.9'In wartime,and specificallyin the decisions
of 1944, the COS, led by the RAF, did not want theirhigh-explosiveand
incendiarybombing loads cut in favorof gas. The former,theyargued, were
well-triedand known to be effective.Although the bomber offensivewas
not workingas intended, the blinders of RAF cultureinhibitedBritainfrom
seeing this.92
This culturalaversion to CW was reflectedin the Britishmilitary'sstrong
interestin avoiding its use. This is a dynamic unanticipatedby traditional
organizational theory,which expects that militariesin war tend to foster
escalation.93Yet in several instances, incidents that could have led to CW
escalation were ignored or purposefullysuppressed. The firstwas in late
1940 when the War Officereceived reportsfromthe Middle East that Italy
was preparing to use gas in Ethiopia. The Commander-in-Chiefof Middle
East Forces suggested that a threatof retaliationin kind be made to deterit.
The War Officequashed the suggestion,fearingthatgivingattentionto cases
of possible use withoutactuallyretaliating(therewas doubt Britaincould or
would) mightindicate to Germany that the UK feared a gas war, and thus
encourage the Nazis to use it. It was decided thatshould the Italians employ
90. Also see CID, "The Manufactureof Toxic Gas forUse in War," Memorandum by the War
Office and Air Staff,July 26, 1938, CAB 4/28, PRO. Paul Harris, "British Preparations for
OffensiveChemical Warfare,1935-1939," Royal UnitedServicesInstituteJournal,Vol. 125, No. 2
(June 1980), p. 61. For a similarassessment during the war, see CAS to the COS, "Chemical
Warfare,"November 14, 1941, WO 193/711,PRO.
91. In 1942, Hugh Dowding, CINC of FighterCommand, wrote in a draftof an article that
"mustard gas should be used in an air attackon Germany."But this view was not approved by
the COS and he had to delete it. This was noted in a letterfromDowding to Basil Liddell-Hart.
See B.H. Reid, "Gas Warfare:The Perils of Prediction,"in David Carlton and Carlo Schaerf,
eds., ReassessingArmsControl(London: Macmillan, 1984), p. 153.
92. See F.H. Hinsley, BritishIntelligencein the Second WorldWar, Vol. III, Part I (New York:
Cambridge UniversityPress, 1988); pp. 298-307; Wilensky,Organizational
Intelligence,
pp. 24-28.
93. This is particularlytrue in formsof warfarethatalready have well-developed routinesand
thatare non-innovative.CW fitboth of these categories.Britainhad prepared to use it and had
already done so extensivelyin World War I; thus its use was not innovative.In fact,of the three
means of warfare,strategicbombing was the most radical change fromconflictin World War I.
MilitaryCultureand Inadvertent
Escalation| 135
gas, the whole matterwould be swept under the rug: "Publicityshould not
(repeat not) be given to the fact."94
The same phenomenon happened in 1942 with respectto reportsfromthe
Far East that leftlittledoubt that Japan was using gas in China. The Army
thoughtthatit was best to ignore this use since any Britishretaliationmight
lead to unrestrictedJapanese CW in India, an area that was considered
vulnerableto gas attacks.95This was a strategiccalculation,but one thatwas
definedby a culturehostileto CW use. For even when strategiccircumstances
changed, when India was not at riskin 1944, the COS stillrefusedto accept
the evidence that the Japanese had used gas.96
In another example, when the Soviets became worried in the winter of
1942 that the Germans were readying to unleash a gas war, Stalin asked
Churchillforhelp. This was given in the formof a pledge: If the Germans
used CW against the Russians, the Allies would use CW against the Nazis.
This idea threw the Britishmilitaryinto frenzy.The COS feltthe promise
mightlead to immediate chemical warfare.Many were especially upset because the United Kingdom had no means to verifywhetherSoviet claims of
German use were actually true.97
A final incident occurred during the German bombing of Bari Harbor in
Italyin December 1943. One of the Luftwaffebombs hit a U.S. supply ship,
the S.S. JohnHarvey,that was carrying2,000 100-lb. mustard bombs to be
used in case CW escalated. The gas was released into the harborwhere many
sailors ended up in the water during the raid. Clouds of the toxic agent
driftedover the town. Some 1,000 civilians, as well as soldiers, were killed
at Bari, many fromthe contaminatedwater and air. In contrastto Britain's
reactionto the accidental bombing of London, the Allies did not propagandize the event or use it as an excuse forretaliation,but instead covered it
up. Medical reportsof wounds were allowed to describe chemical weapon
injuries only in general terms, and strictcensorship was institutedat all
militarybases. When it was clear that the accident could not be kept secret,
the Combined Chiefs of Staffprepared a statementwhich reiteratedthat
"Allied policy is not (repeat not) to use gas unless or until the enemy does
94. Telegram fromthe War Officeto the CINC Middle East, December 16, 1940, WO 193/721,
PRO; "Chemical Warfare:Use of Gas by Italians and Policy forRetaliation,"WO 193/725,PRO.
95. "Japanese Gas Warfarein China," July14, 1942, WO 193/723,PRO.
96. Ismay (forCOS) to PM, June28, 1944,.PREM 3/89,PRO.
97. 11thMeeting of the Defence Committee,April 17, 1942, WO 193/711,PRO.
International
Security18:4 1136
so firstbut that we are fully prepared to retaliate and do not deny the
accident,which was a calculated risk."98
CW. German military
GERMANY AND THE LACK OF INADVERTENT
culturewas
no more favorable to gas than was Britain'sand the effectof unintended
incidents on escalation was the same: restraintendured. Dating back to
Moltke, German militarythought focused on fast, decisive maneuver and
encirclementas a means of victory.In World War I, a skepticalmilitarywas
convinced to trygas based on the argumentthatit would break the stalemate
of trenchwarfare.99It did not. In fact it turned out to be a poor fitwith
Germany's desired operations in World War I because it inhibitedmobility
and impeded the aggressiveness of soldiers. In addition it led to civilian
interferencein militaryaffairs,particularlyby officiallysanctioned chemists.'00
German doctrineon the eve of WorldWar II was decidedly offensive:fastmoving and long-rangearmored and motorized units would spearhead the
attack,break throughthe enemy's front,penetratequicklyand deeply into
the rear, counter enemy effortsto block encirclementor escape, and sever
communications,supply, and command lines. Infantrydivisionswould then
move in for annihilationbattle fromthe front.Chemical weapons, which
were cumbersome, could play only a limited role in such a strategy.CW
equipment and munitionswould jam supply lines, and chemical casualties
were difficultto handle: they did not die easily and needed intensivecare.
This was particularlytrue since the use of gas in offensiveoperationswould
have demanded centralized control which clashed with the decentralized
German auftragstaktik
system.'0'
For reasons that differfromBritain's-Germany's later defensive stance,
the threatto its survival, and the mercurialnature of Hitler,who ignored
many norms of state behavior-it may seem surprisingthat the Reich did
98. RobertHarris and JeremyPaxman, A HigherFormofKilling:The SecretStoryofChemicaland
BiologicalWarfare
(New York: Hill and Wang, 1982), pp. 119-123.
99. Robert M. Citino, The Evolutionof BlitzkriegTactics:GermanyDefendsItselfAgainstPoland,
1918-1933 (Westport,Conn.: Greenwood Press, 1987), p. 81; Matthew Cooper, TheGermanArmy
1933-1945 (New York: Stein and Day, 1978) pp. 139-140; FredericJ. Brown, ChemicalWarfare:
A
Studyin Restraints(Princeton,N.J.: PrincetonUniversityPress, 1968), p. 5.
100. Haber, The PoisonousCloud, p. 269, 273. Some of the reasons for the poor fitbetween
cultureand CW in Germany,as in Britain,are related to factorsthat traditionalorganizational
theoristswould stress (forexample, a bias towards offenseand autonomy),but the outcomerestraint-does not fitthe predictionsof the traditionalschool.
101. Ochsner, TheHistoryofGermanCW, p. 5; StockholmInternationalPeace Research Institute
[SIPRI], The Problemof Chemicaland BiologicalWarfare,VolumeI: The Rise of CB Weapons(New
York: Humanities Press, 1971), p. 307.
Escalation1137
MilitaryCultureand Inadvertent
not reflexivelylapse into CW use. Coupled with securitydilemma considerations,such as surprise-useadvantages, it is even more curious why there
was no accidental escalation. Yet froma militarycultureperspective,the lack
of escalation is not a puzzle.
During Germany'sinvasion of Poland, mustardgas was used in the Polish
defenseoftheJaslobridge,resultingin several Germancasualties and deaths.
Instead of responding in kind, however, the Third Reich's militaryassumed
that the Polish Supreme Command had not ordered the use of gas. They
were right,but the choice seemed more a product of hopeful expectation
than shrewd analysis.'02The benign assumption that the Germans made in
this situation,that the gas use was not intentional,contrastssharply with
what Britain concluded about the Luftwafferaid on London, or Lemp's
judgment on the status of the Athenia.When the German Foreign Ministry
wanted to use the incident for propaganda, General Halder was quick to
squelch the idea. It appears thathe, like his Britishcounterparts,was afraid
it mightlead to the initiationof CW.103
A second incidentin July1941 testifiesto the unusual effortssome states
went to in hopes of avoiding escalation in certainareas. The Soviet Union
claimed that Germany was gettingready to use chemical weapons. This
accusation was based on the capture of a German manual on the offensive
use of gas. In response, Germanywas quick to announce throughits official
news agency that the manual was merelya trainingguide, allowed by the
Geneva Protocol,and not an imminentplan.'04That same summer,German
militaryleaders had received fivereportsfromthe fieldthatthe Soviets had
used chemical weapons. One involved a bomber,two were artilleryattacks
and two were armored vehicle assaults. The Germans decided that not
enough "objective" evidence existed thatthe attackshad occurred.But since
twelve German soldiers had mustard gas wounds, it was conceded that
perhaps a single gas bomb had been dropped. Otherwise, however, the
incidentwas ignored.'05
102. Ochsner states that this findingwas "a great reliefto us." Ochsner, TheHistoryofGerman
CW, p. 16.
103. "Pressepropaganda Gelbkreuzgasverwendungdurch die polnischenTruppen," September
23, 1939, RW 5/v.346,BA-MA.
104. But, "if the Soviets use the discovery of German instructionsabout gas as an excuse to
begin gas warfare,Germanywill answer appropriately.""Abschrift.Auszug aus der Times vom
26 July1941," RW 5/v.346, BA-MA.
105. See Armeeoberkommandd11 an Oberkommando des Heeres, July1, 1941, RW 5/v.346,
BA-MA, for a list and descriptionof the injured soldiers. For the analysis of the incidents see
International
Security18:4 1138
Another provocativeaccident occurredduring the Allied invasion of Italy
at Anzio in 1943. A German shell struckan Allied weapons depot containing
chemical munitions. The explosion released a cloud of gas that driftedtowards the German lines. The Allied commander was quick to notifyhis
German counterpartthat this release of gas was strictlyinadvertent.The
German officeraccepted the explanation despite his disadvantage had the
Allied officerbeen lying.'06
Even at the end of the war, when the Germans faced imminentpolitical
extinctionand the Allies feareddesperate escalation, therewas no last resort
to CW in the confusing, threatening,frenzied disintegrationof the Third
Reich. In fact, Germany became particularlycautious about unauthorized
use. Supplies were ordered moved, not destroyed,so as to avoid any event
that might give the enemy a pretextfor CW use.'07 Chemical stocks and
factorieswere given top priorityin the allocation of scarce transportspace.
Despite precautions, on April 18, 1945, an accident at a chemical depot in
centralGermany led to the contaminationof the surroundingtwentykilometers. The Wehrmachtanticipated that the enemy mightpoint to such an
incidentas an excuse forinitiatingdeliberateuse of CW and it recommended
haltingthe riskytransferof chemical stocksand givingthe Allies the location
of the sites. Hitlervetoed this order.'08Nonetheless, therewas no escalation.
ORGANIZATIONAL
CULTURE
AND
INADVERTENCE
Variationsin inadvertentescalation in World War II are explained betterby
organizational culture than other prominentimages of escalation. The significantinfluence of culture is summarized in Table 1. There are two key
facetsof the link between organizationalcultureand inadvertence.The first
im Ausland Nr. 10," August 12, 1941, RH 11 IV/v.
"Mitteilungenuber Gaskriegsvorbereitungen
17, BA-MA.
106. This incident is related by Lord Ritchie-Calder,who was Directorof Political Warfarein
the Foreign Office during the Second World War. See Steven Rose, ed., CBW: Chemicaland
BiologicalWarfare(Boston: Beacon Press, 1968), p. 14.
107. Oberkommando der Wehrmacht,"Gaskriegsvorbereitungen,"February4, 1945, RW 4/v.
720, BA-MA.
108. Rolf-DieterMueller, "World Power Status Through the Use of Gas? German Preparations
and GermanRearmament
(London:
forChemical Warfare,"in Wilhelm Diest, ed., The Wehrmacht
Macmillan, 1981), pp. 200-201; Stephen L. McFarland, "PreparingforWhat Never Came: Chemical and Biological Warfarein World War II," DefenseAnalysis,Vol. 2, No. 2 (1986), p. 114; F.H.
in theSecondWorldWar: Its Influenceon Strategyand Operations,Vol.
Hinsley, BritishIntelligence
III, PartII (New York: Cambridge UniversityPress, 1988), pp. 577 and 929-930. Brown, Chemical
Warfare,p. 237, suggests that Hitler may have ordered gas attacks at the end of the war but
officersand officialsdid not carryout his command.
Escalation| 139
MilitaryCultureand Inadvertent
Table 1. A Summary of the Cases. Does Theory Predict Inadvertence?
Security
Friction Dilemma
SUBMARINE WARFARE
U.K. Oxely accident
U.K. Bremen non-accident
Germany
STRATEGIC BOMBING
U.K.
Germany
CHEMICAL WARFARE
U.K.
Germany
Traditional
Organization Organizational Actual
Theory
Culture
Outcome
?
No
No
?
No
No
Yes
Yes
Yes
Yes
No
Yes
Yes
No
Yes
Yes
Yes
No
No
Yes
Yes
Yes
No
Yes
No
Yes
Yes
Yes
Yes
Yes
Yes
No
No
No
No
connectsorganizationalpredilectionsto the types of accidents thatare likely.
Clausewitz tells us that accidents are unpredictableand unavoidable in the
complexityof war. The incidentsreviewed above do seem to have multiple
causes and do not fitneatly under any one theoryexcept the vastly generalized and residual one of friction.Nonetheless, some accidentsmay fitmore
of a patternthan would be suggested by a random-walkthesis. Some degree
of regularitymatches organizations to accidents, as Graham Allison suggested in EssenceofDecision.'09But the thrustofAllison's argument-and that
of otherswho have employed traditionalorganization-theory
logic-assumes
that,in gross terms,similarorganizationsact in similarways. According to
thislogic we should expect militarieswith the same structuresand functions
to incurthe same types of incidents.But this emphasis on structureneglects
the importanceofbeliefsand norms.As was demonstratedby the differences
between the accidents of German and Britishsubmarines,similarstructures
and functionsdid not produce the same type ofresults.The submarineforces
of both navies were prepared to targetand destroy enemy ships. But the
BritishNavy saw only a role for hittingwarships, whereas the German
109. E.g., Allison, EssenceofDecision,pp. 139-140. Where frictionwas comparable-for example
in the Britishand German submarine situations-an accident occurred in one, but not in the
other. And even where frictionis expected to be most decisive-where warfareis complex,
fightingintense, and informationuncertain-accidents did not cause escalation. This was apparentin the restraintthatenduted in CW duringthe fiercebattleson the continentin the latter
stages of WWII.
Security18:4 1140
International
Navy considered anti-tradeattacksas a centralformof warfare.A focus on
this divergence in the hierarchyof beliefs in militaryorganizations more
accuratelyaccounts for what occurred. The role of culture is more important-and the influenceof the armed forcesmore varied in thattheycan also
inhibitaccidents, than is recognized by friction,the securitydilemma, or a
traditionalorganizationviewpoint.
The second, and more important,tie between organizationalcultureand
accidents concerns the impact of the unintended incidents that do occur.
Avoiding accidents altogethermay be an impossible task. But as we have
seen, some accidents lead to escalation, while othersdo not. Contraryto the
expectationsof the traditionalorganizationalschool, militariesdo not always
push events towards escalation,even afterhostilitieshave begun. Depending
on culture, militaryorganizations may act as inhibitorsor as advocates of
escalation. The armed forces play a centralrole in war throughthe development of capabilities, planning, informationprocessing, operational response, and judging militaryutility,and theirculturestherebyhave a decisive
impact on national choices. Where specificmeans of warfareare compatible
with one side's organizational culture,accidental use of a taboo means of
warfareby the enemy oftenleads to escalation. Militariesand states in such
situationsare likelyto emphasize the antagonisticrole the otherside played,
encourage propagandisticuse of the incident,and stressthe militaryadvantage in theirown escalation. Such tendencies increase the likelihood of inadvertentescalation, as was evident in German submarinewarfareand Britain's bombing of Berlin.
In contrast,when a type of warfareis antitheticalto organizationalculture,
restraintendures in the face of provocative enemy incidents. Information
encouragingescalation is suppressed, enemy actions are taken on faithto be
accidents, effortsare made to communicategood will to the opposing side,
and internalproposals to seize propaganda advantages are rejected. These
dynamics were evident in German strategicbombing, and Britishand German decisions related to chemical warfare. As we have seen, escalation
windows are inevitablythrown open in the midst of conflictby the unintended and often unpredictable incidents that occur in "unthinkable" restrictedmeans of warfare.But whetherstates jump throughthose windows
seems to be importantlyaffectedby militaryculture.
Escalation1141
MilitaryCultureand Inadvertent
Inadvertence
Contemporary
Some of the taboos of the inter-warperiod remain intactto varyingdegrees
today. Chemical and biological warfarecontinue to be stigmatized,and so
too is strategicbombing,as seen in the allied efforts
to avoid civiliancasualties
in the Gulf War. The most strikingdistinctionbetween the period examined
above and the modern age is the immensely destructiveforce of nuclear
weapons."0I Some believe thatany serious armed clash between major power
antagonists would be likely to result in nuclear use. Some assume that an
accidental event involving a nuclear weapon, particularlyin the midst of
war, would lead to a widespread exchange.111
What the organizationalculture
approach suggests, however, is that this need not be the case. Depending
on organizationalpredilection,some incidentsare likelyto lead to escalation
while others will not. Thus fornational leaders, the abilityto controlescalation involves understandingand managing bureaucraticculture. This is a
differentsortof enterprisethan those suggested in otherrecentstudies. It is
not about the technicalspecificationsand procedures of command and control systems.112 Nor is it about the formalstructuraltraitsof organizations
themselves.113 Furthermore,the primaryfocus is not on the explicitnature of
civil-military
relations and the problem of gettingsoldiers to adhere to the
orders and aims of the higher militaryor civilianleadership, although that
is certainlya concern.114
Rather, the policy task is first,to understand the norms and beliefs of
militaryservices on war-fightingthat permeate the plans, capabilities, and
110. It is unlikelythat these powerful tools of violence have neutralized the dynamics of the
images of inadvertencediscussed here. Friction,the securitydilemma, and organizationaldynamics have all figuredprominentlyin research on the nuclear age; e.g., BarryPosen uses all
threein his study Inadvertent
Escalation.
111. This is the predominantthrustof traditionalorganizationtheory,"friction,"and "normal
accident" theorists,along with most who study nuclear accidents. See Bracken, "Accidental
Nuclear War," in Graham T. Allison, Albert Carnesale, and Joseph S. Nye, Jr.,eds., Hawks,
Doves, and Owls: An AgendaforAvoidingNuclearWar (New York: W.W. Norton, 1985), pp. 3749; and Sagan, The Limitsof Safety,pp. 250-51, 259-264. For an argument that rejects this
thinking,see JohnMueller,Retreat
fromDoomsday:TheObsolescence
ofMajor War(New York:Basic
Books, 1989), pp. 237-238.
112. E.g., Blair,The LogicofAccidentalNuclearWar.
113. Although they acknowledge the role of culture,this is the focus of the "high reliability"
theoristsdiscussed in Sagan, The LimitsofSafety,pp. 14-28.
114. On control of U.S. nuclear weapons, this issue is covered thoroughlyby Peter Douglas
Feaver, Guardingthe Guardians:CivilianControlof NuclearWeaponsin the UnitedStates (Ithaca:
Cornell UniversityPress, 1992).
International
Security18:4 1142
skills available to keep the peace. Secondly, it is to shape that culture, if
necessary, so that it is compatible with national objectives. These tasks are
far from simple. Professional militarycultures formin a society set apart
fromthe broadernation theyserve, and such organizationshave no domestic
competitorsto ensure that the consumer (i.e., national interest) is wellserved. Even in the United States, where there is a traditionof civilian
oversightand "independent" bodies to contributeto militarythinking,outsiders (especially those who spend fouryears in Washingtonand then move
on) sometimes have problems gaining obedience to explicitdirectives,let
alone the abilityto affectorganizational culture.115Nonetheless, there is a
need foralternativeparties-probably civilianexpertsauthorizedby the highest levels of government-to review operational plans.116 This is not about
who has the final say in using force:that is clearlythe politicalleadership.
Instead, the issue is how choice is subtly,but powerfully,circumscribedby
the pre-existingorganizationalmind-set,withits attendantcapabilities,skills,
and intelligence,that dominates operationalthinking.Thus, the centraltask
is to explicate and assess the assumptions and beliefs that shape the way
that militariesthinkabout practicalwar-fighting
in termsboth of theirown
efficacyand of politicalobjectives.
The aim of such an effortwould not be to de-professionalizeor to politicize
America's competent armed forces.117Nor is it to blame soldiers fordoing
an inadequate job. Militaries cannot be considered the cause of war, nor
should they be pictured as an unwavering source of escalation. The armed
forcescan act as a friendof restraintas well as a foe (and eitherrole might
serve a national purpose). Nonetheless, the pervasive influenceof military
cultureon inadvertencesuggests the need to improve understandingof the
beliefsand norms that characterizethese organizations.It is certainlyin the
nationalinterest,and to the benefitofinternationalsecurity,thatwar-fighting
culturesbe compatiblewith higherlevel politicalstrategyand goals.
115. A range of anecdotes and analysis on this topic are found in JanneE. Nolan, Guardiansof
theArsenal:The PoliticsofNuclearStrategy(New York: Basic Books, 1989), pp. 5-6, 31-32, 248285; and Feaver, GuardingtheGuardians,pp. 227-229, 232-234, 242-244.
116. Posen, Inadvertent
Escalation,p. 217, offerssome good proposals along these lines.
117. The dangers of doing so have been articulatedin the classic workby Samuel P. Huntington,
The Soldierand the State: The Theoryand Politicsof Civil-Military
Relations(Cambridge, Mass.:
Harvard UniversityPress, 1957).