Thinking about Nuclear Terrorism Author(s): Thomas C. Schelling Source: International Security, Vol. 6, No. 4 (Spring, 1982), pp.

61-77 Published by: The MIT Press Stable URL: http://www.jstor.org/stable/2538678 . Accessed: 21/06/2011 14:22
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 support@jstor.org.

The MIT Press is collaborating with JSTOR to digitize, preserve and extend access to International Security.

http://www.jstor.org

Thinking about Nuclear Terrorism

ThomasC. Schelling

Sometime

in

the

1980s an organizationthat is not a national governmentmay acquire a few nuclear weapons. If not in the 1980s then in the 1990s. The likelihood will grow as more and more national governmentsacquire fissionable material fromtheirown weapon programs,theirresearchprograms,theirreactor-fuel programs,or fromthe waste products of theirelectricpower reactors. By "organization" I mean a political movement, a governmentin exile, a separatist or secessionist party, a militaryrebellion, adventurers from the undergroundor the underworld,or even some group of people merelybent just a sample of on showing that it can be done. My list is not a definition, of the possibilities.Two decades of concernabout the proliferation weapons list have generated a familiar of national governmentsthatmay have motive and opportunityto possess weapons-grade fissionable material, and some ideas about how they mightbehave if they had it and what they mightuse it for.But thereis also a possibilitythat somebody otherthan a government may possess the stuff. Who they mightbe and how they mightacquire it are related questions. While not impossible, it is unlikely that an entitynot subject to nationalgovernmentregulationcould independentlyobtain and enrich uranium for use in explosives or could produce plutoniumas a reactorproductand refine it for weapons use. There are undoubtedly corporations technicallyand able to do it, but not many with both the motive and the opporfinancially tunityto do it withoutbeing apprehended by an adversely interestedparty. Access to weapons or a weapon program, or to an authorized nuclear fuel cycle, or to an officialresearch establishmentlicensed and authorized by the the some national governmentis currently only way to do it. Identifying and the access to those opportunitiesgenerates some answers opportunities to the question "Who?" Theftof weapons is an obvious possibility.As far as we know, it hasn't happened. Despite the thousands in existence, including the thousands on
This essay was originallypresented at the Conference "War and Politics," held in November 1979 at the University California,Los Angeles and sponsored by the Center forInternational of and Strategic Affairs. will also appear as a chapterin NationalSecurity International It and Stability, edited by BernardBrodie, Michael D. Intriligator, and Roman Kolkowitz (forthcoming). ThomasC. Schelling Professor PoliticalEconomy theJohnF. Kennedy is of at Schoolof Government, HarvardUniversity. is theauthorof Strategyof Conflict(HarvardUniversity He Press,1960; Oxford University Press,1963), Arms and Influence (Yale University Press, 1966), and Micromotivesand Macrobehavior(W. W. Norton,1978). International Security, Spring 1982(Vol. 6, No. 4) 0162-2889/82/040061-17 $02.50/0 and Institute Technology. of ? 1982bythePresident FellowsofHarvard Collegeand oftheMassachusetts

61

Security| 62 International

foreignsoil, and the large numbersof people who participatein the custody, of maintenance, and transport nuclear weapons, and despite earlierreports of by the General AccountingOfficethatproper care in the transport weapons has sometimes been lacking, I am not aware of any hint that thefthas has the occurredin any country.And iftherehas been a theft, thiefcertainly not made a public announcement. negative can be said about theftof separated plutonium Nothing so flatly of between theft the material or enriched uranium. An importantdifference and theftof a weapon is that nobody is likely to remove a weapon or a warhead in small pieces. A weapon, if stolen, is likely to be taken whole. Materials fromsome sources, on the other hand, would have to be secreted cumulativelyover a protractedperiod. organization Giftis a possibility.There may be things a non-government can accomplish that a national governmentwould prefernot or dare not to or try.Surreptitious anonymous activitiesby agents of a governmentwe can problem; but weapons entrustedto consider part of the national proliferation independent or uncontrolledparties who have at least some autonomy in what to do with them should be counted as part of the non-nationrisk. The giftmay be extorted;blackmailagainst a governmentpossessing weapons or weapons material is one way of obtaining a "gift." Extortionwould be if especially pertinent the recipientsknew of a clandestineweapon program, and iftheyhad a capabilityto hurtthe governmentor the people concerned, but would target the nuclear capability elsewhere. In principle we should add the possibilityof purchase; but in mattersof corruption,"briberyand extortion"are so often togetherthat gift,blackmail, and purchase can be motivated by various inducethought of as unilateral intentionaltransfers ments. forces, or officials, units of national military Defectionof civilianor military is an obvious possibilityonce a national governmenthas weapons, especially available for military use. Had nuclear weapons been in weapons officially the hands of French forces in Algeria in 1958, the paratroopersand Secret Army Organization that challenged the Paris governmentmighthave made use of such weapons or arranged theirdisappearance foruse on threatening some later occasion. When Batista went into exile from Cuba, or Somoza holding of nuclear weapfromNicaragua, or the Shah fromIran, any official ons or the materialsfromwhich they can be made could have accompanied him, or could have leftthe country,or gone into hiding with other officials. The civilwar thathas ravaged Lebanon, the war thatseparated the two parts

| NuclearTerrorism 63

of Pakistan, the overthrowof Sukarno in Indonesia, and the turmoilin Iran fromthe first confrontations between officers loyal to the Shah and officers disobeying suggest the circumstancesin which somethingcould have happened to an official arsenal of nuclear weapons. Three otherNATO countries besides France-Portugal, Greece, and Turkey-have undergone changes in govemment by violence or the threat of violence in such a way that any weapons could have reached "unauthorized hands," eitherat the time or on a later occasion. It is worth speculating on what mighthave happened to Iranian nuclear weapons had the turmoilthatbegan in 1978 broken out a dozen years later, when the Shah's originally planned nuclear power programwould have been generatingspent nuclear fuel fromwhich plutoniumcould be reprocessed or weapons fabricated.The Shah or someone loyal to him might have taken weapons or material out of the country(or sent it out in advance) for the purpose of staginga comeback or defendingwealth and personal securityin exile. They mightalready have been entrustedforsafekeepingto somebody abroad. Loyal troops without access to weapons might have taken steps to obtain them to keep them out of dangerous hands; disloyal troops might have sought to capture weapons forpolitical use or, also, to keep them out of even more dangerous hands. The disintegratedregime presided over by the Ayatollah might have found it awkward and controversialto have nuclear-weapons materialin the possession of some impermanentofficial-the President, say, who disappeared with his executioners in pursuit in July 1981. Or whoever managed to possess the material might have claimed authorityunder the Ayatollah to withhold it fromthe president or prime minister, withoutany countermandfromthe Ayatollah. Indeed, it isn't clear to whom it might have been entrustedby whoever might have been in a position to do the entrustingduring the near-anarchy of 1979-80. Other interestedparties would undoubtedly have been willing to consider commando tactics to preempt the stuff, eitherbecause of its value in the right hands or because of its value in the wrong hands. Israel, the Soviet Union, and the United States of America come to mind, as does the P.L.O. What would have come out of the scrambleis a guess thathas many answers. And as with the jewelry,the cash, or the negotiableassets thatthe Shah or others thatfled may have taken with them, it mightnever have been known how much nuclear materialtherehad been and how much was missing.A prudent royalfamilywould have exercised the same caution with nuclear materialas with gold or Panamanian bank accounts.

International Security| 64

These different routes by which weapons or material might get out of officialhands tend to compound themselves. Whatever the likelihood that, say, the governmentof Iran would have sold nuclear weapons or given them away, or yielded them to blackmailby the P.L.O., weapons in the hands of militantsor of disaffected naval or air forcesmightbe subject to less inertia and less inhibition,or more subject to theftor capture or plain loss in the confusionof escape and evacuation. Whoever had such weapons or material would be unlikelyto hold it or transportit in labeled containers.Justwhat do you do with a hundred pounds of plutoniumwhen the building is under siege and you are in a hurryto reach the airport,or you have to leave town in disguise, and you don't exactlytrustthe "authorized" official whom it to is supposed to be surrenderedunder standard operatingprocedures? For now this is all just rehearsal. In ten or fifteen years it may be a live performance.

It Getting All Together

If weapons materialis obtained in any of these ways froma nation that has a small or clandestine weapons program or especially fromone that has no weapons programat all, the materialis not likelyto be obtained in the form of completelyassembled bombs or warheads. For reasons of custodial security,the weapons might be unassembled. If weapon development is a continuingprocess, a governmentmightnot commitscarce materialto permanentlyassembled weapons of obsolescing design. A governmentunwilling to entrustweapons to its own armed forcesmightfind it expedient, in to dealing with senior military officials, keep its arsenal of finishedweapons in the future.A governmentpreferring keep a weapons option, but not to to declare or leak a nuclear militarycapability,might reduce the lead time between weapon decision and completed weapon without marryingthe fissionablematerialand the othercomponents,exceptin laboratory rehearsal. So an organizationthat obtains the materialmay still face the task of cona structing portable explosive. That fact makes strong demands on the organization. Highly qualified scientistsand engineers are required. Some years ago, Hans Bethe publicly calculated a minimumof six well trained people representingjust the right specialties.Unless theyalready possessed weapon designs, theywould need access to computersand a library. They would need the disciplineand loyalty to work in secrecyand in trust.And it would take time. Recruiting the team

NuclearTerrorism 65 |

would be difficult; even clandestineadvertisingcould give the whole secretly thingaway. Mercenaryengineersmightfeel uneasy about an enterprisethat required them to leave no trace behind, an enterprisepremised on willingness for large-scale violence and not likely to give much thought to the welfare of mercenaryengineers if secrecy were at stake. If the key people were sympathizersratherthan mercenaries,they would want to participate in the planningof what was to be done with the weapons. In sum, it appears to requirea group of significant size, high professionalquality,and excellent organizationand discipline to convertunauthorized or illicitly obtained materialsinto a useable weapon. A consequence is thattherewould be timeand opportunity make plans, to to assure thatthe weapons or the opportunity were not wasted, and to work out the technicaland tacticalproblems of exploitingthe weapons once produced. If the materialwere obtained foran existingemergencytheremight be urgency in completing a weapon promptly and exploiting it at once; otherwise,a major considerationwould be that exploitingthe weapon early would eliminate secrecy, stimulate countermeasures, close the source of materialif it remained open, and use up the unique occasion of "firstrevelation." I conclude that people capable of all this will be able to do some pretty sophisticated planning. There will be time and motivationto think about how to use this unique capability.They may wait forthe rightoccasion; they createthe rightoccasion. They may be in no hurry.They could may patiently hope not to need to actually use theirweapon. And they are likely,having thoughtlong and explored alternativesin detail before embarkingon their than theiradversaries. plan, to be betterprepared intellectually If theydecide to claim possession of a nuclear bomb and, as at Hiroshima, do not wish to deplete their supply by an innocuous demonstrationor to riskfailurein an open test, theywill have to be prepared to prove thatthey have what they claim. But they will have had the time, and probably the right people, to have considered how to authenticate their claim-what are technicaldemonstrations feasible,even which ambassadors to appoint in presentingtheirclaim. If they need to pre-position a few weapons and can best accomplish it before their victims have any inkling of what is up, there will be further reason to postpone revelationand more time to plan the campaign. And a campaign what they will plan, not an episode. Unlike most "teris rorist" acts of recent years, the activityinitiatedby the announcement or

International Security| 66

demonstrationof a nuclear weapon is not likely to culminate in a decisive outcome that terminatesthe episode. If a weapon were used to coerce, neitherthe success nor the failureof coercion would necessarilylead to the surrenderor capture of the weapon. Even the explosion of a weapon might with any confidence,as the exhaustion of the arsenal and not be interpreted, the end of the threat.Acquisitionof a nuclear-weapons capabilitymay confer permanentstatus on the organizationand create a permanentsituation,not just initiatea finiteevent.

WillThey Terrorists? Be
The question is sometimes posed, will terrorists come to acquire nuclear weapons? That question startsa series of thoughts:thinkof all the different terrorist organizations,fromQuebec to Uruguay, Ireland to the Basque country,Rhodesia to Israel. Who among them might acquire weapons material during the coming decade or two? Who would know how to produce a weapon, or even to build an organization technicallyand institutionally Who among them would capable of bringingsuch an enterpriseto fruition? actually trusteach otherenough to want such a thing? Those questions probablyfocus on the wrong issue. The proper question will is not whetheran organizationof the kind thatwe thinkof as "terrorist" get nuclear weapons in pursuit of theirgoals. It is whetherany organization thatacquires nuclear weapons can be anything terrorist the use of such in but weapons. Does possession of a nuclear weapon, or a few weapons, necessarily make an organization "terrorist"?Are the weapons themselves so in terrorist, any use thatcan be made ofthem,thattheymake theirpossessors whatever else theymay or may not be? supreme terrorists, Except foracquiring a weapon just to prove it can be done-a motive that I don't doubt appeals to some, but one that I doubt adequate to accomplish the feat-I find it hard to thinkof any use that would not be "terrorist."I also findit hard to thinkof any exploitationof nuclear weapons by a national other than one with a sufficient arsenal forbattlefield government, use, that would not be terrorist. Even the language of "mass destruction"in categorizing these weapons suggests intimidation and reprisal ratherthan battlefield effectiveness. The concept of "massive retaliation" is terrorist. My dictionarydefines as terrorism ". . . the use of terror, violence, and intimidation achieve an to end." And to terrorize ".'. . to coerce by intimidation fear." The passive is or

| NuclearTerrorism 67

known as "deterrence"(using the root of the word "terror"),need not form, but connote bloodthirstiness, thereis a 30-yeartradition thatthe appropriate nuclear targetsfornuclear forcesare cities or populations, and that strategic forces induce caution and moderation in an adversary by threateningthe destructionof the enemy society. I imply nothingderogatoryor demeaning nuclear forcesby emphasizing the traditional about strategic expectationthat their primaryuse is to deter or to intimidate, and thereby to influence behavior, throughthe threatof enormous civilian damage. It is worth rememberingthat on the only occasion of the hostile use of nuclear weapons, they were used in a fashion that has to be considered "terrorist."There was a nation that had a very small capabilityto produce nuclear bombs. The need was sufficiently urgent that it was decided to go ahead with "revelation" when only two were in hand. The hope was to stun the enemy into surrender,or to create such a tremorthat the government itselfwould change into one disposed to surrender. The possibility of a harmlessdetonationin an unpopulated place was considered but rejectedon grounds that the demonstrationmightfail (possibly throughincomprehension by the witnesses) and in any event would deplete the stockpileby half. The weapons couldn't be wasted on a remote battlefield; and even military destructionin the Japanese homeland would be incidental compared with the shock of an anti-populationattack. With a modest pretense at militaryindustrialtargeting, industrialcityof Hiroshima was chosen. No warning the was given thatmighthave allowed interference with the demonstration. And so much was at stake-the possibilityof continued warfareand the prospective loss of more millionsof lives, mostlyJapanese-that the demonstration's casualties mightbe justifiedas the price of persuasively communicatingthe threatand precludingany militarist refusalto take the threatseriously.When the response to the firstbomb was not prompt enough, the second was dropped on Nagasaki, depletingthe arsenal altogether.(Whethera thirdwas ready fordelivery,the victimgovernmentcould not know.) The bombs on Hiroshima and Nagasaki remind us that,fromthe point of view of those who use the weapons, use favorsthe rightside. It may even be hoped that use will minimize the ultimateviolence to the enemy. To make the comparisonmore contemporary, can contemplatethe kind we of use that mightbe made by a national governmentthat had a few nuclear weapons, "few" meaning single numbers or teens, few not only by comparison with so-called nuclear powers, but too few for decisive use on the battlefield.It is generally expected that population threats,verbal or com-

International Security 68 |

municated by detonation,would be the obvious role forsuch weapons, and that even nominally"battlefield"use would be intended mainly forintimiwith only dation ratherthan local targetdestruction.A national government mode in which to a few such weapons would not have any but a "terrorist" a use them. There are many ways to differentiate national governmentfrom quality in any use of nuclear a non-governmental entity,but the "terrorist" weapons does not seem the crucial differentiation. Ifa governmentthatexploitedany genuine or pretendednuclear capability organizationby doing would appear to "descend" to the level of a terrorist thatpossessed or could so, an organizationotherthan a national government crediblyclaim to possess nuclear weapons converselymight"ascend" to the status of a government.It mightseek its own permanence as a nuclear miniOr or state, even if lacking territory. it mightclaim a territory seek a homein claimantto legitimate authority some itselfas the rightful land, identifying existingstate. Considering the status and "prestige" thatare supposed to go with the fearsomeaccomplishmentof producing or otherwiseacquiring nuclear bombs, and recognizingthat somethingas vaguely defined as a Palestine Liberation Organization can achieve diplomatic recognition,observers should not have to tax their imaginations to suppose that an organization with the ability to acquire the wherewithal to produce nuclear weapons mightproceed to set up its own foreignofficeand dispatch its own ambassadors to the governmentswith which it proposed to do business. This idea is supported by the considerationthatan organizationthatcould have one bomb could well have more than one. So even violent use of a "successful" exploiweapon mightnot exhaust the campaign. Furthermore, entityas fora national governtationof a weapon, fora non-governmental ment,would likelyachieve itspurpose withoutexplosion, withoutdestroying the weapon itself.So the organizationneed not contemplateits own demise, in as a nuclear mini-state, the event of initialsuccess. Modes of Utilization It is hard to foretellwhat mode of exploitationmightbe adopted by some organizationthat comes into possession of nuclear weapons. It is especially hard because the people who would decide the strategycould have had weeks or months or longer to think and argue intenselyabout what their "nuclear strategy"should be. Most of us have not spent weeks or months our thatexercise. And because we have not identified to attempting anticipate

NuclearTerrorism 69 |

specific candidate, we have no vivid image of either its objectives or its limitations help sharpen our imaginationor to screen out the implausible. to We can, however, generate a menu of some possibilities with which to acquire some notion of the range of surprises to which we may be subject someday. way that a small nuclear-weapon Not the most dramaticbut an effective capabilitymightbe exploited would be to announce it. Announce it nonand do nothing else. The announcement might require idenbelligerently, it was that had it, unless the organizationor its affiliation was who tifying the could be allowed to leak; already well known. Alternatively information and if there was already conjecture that the organization was acquiring a nuclear capability,the "announcement" mighttake the formof merelynot responding negatively(or convincingly)to inquiries. The factof possession as could be interpreted an implicitthreat.If the object were to gain attention and status,a believable announcementought to do the trick.The announcement could of course be secretand limitedto a few addressees, perhaps the target governments toward whom the weapons represented an implicit threat,perhaps a friendly governmentthatmightor mightnot welcome the news of such support. Some effort authenticatethe claim would accompany the announceto ment. We have fictionaland journalistic accounts of how this might be or done-revealing details of theft capture thatonly thieves or captors could know, invitingor kidnapping witnesses fora "show and tell," deliveryof a weapon facsimileor a sample of weapons-grade material,open declaration by the distinguishedscientistswho had designed and assembled the weapons, or of course a "test" detonation. Depending on the riskof interception, the lattercould be by surprise, at an unannounced time and place, or by "invitation"or advance tip-off. An alternativepassive, but more dramatic, mode might be to eschew announcementand proceed to demonstration.If the object were to strikea littleterror,an anonymous blast might do; and although minor terrorist groups mightclaim authorship in the usual style,the authenticperpetrator should have no difficulty proving his claim. In case it seems thatmere announcement,unaccompanied by overtthreats and demands, would lack the climactic qualityone expects to findassociated that announcement a non-nation nuclear capability,it helps to reflect with leaks and undenied conjectures, are what one usually exor, alternatively, pects of a national government.Except for the initial American revelation,

International Security| 70

anothergovernment the announcing) do mere announcement(or even letting has been the custom. Those who believe that Israel has nuclear weapons, whether in readiness or ready for quick assembly, must believe that "announcement by denial" is a believable tactic. (Whether they believe that Israel also made some explicitbut discreetrevelations,I don't know.) In any case, thereis nothingabout "mere" announcement thatprecludes becoming more explicitlater. The way it is done would depend, too, on the manner in which some national governments might, between now and then, have way to conducted their own revelations or tests or firstuses. An effective announce possession if the most likely target government had just announced that it possessed nuclear bombs would be to counter,"we do too." In a more active mode, the organizationcould approach the targetgovernment, i.e., the governmentwhose behavior it hoped to influenceby formulation of demands. It could do this secretlyor openly. There are at least four combinations:open announcementcoupled with secretdemands, secretdisplay or announcement associated with open demands, or both secret, or both open. There mightbe more than one targetgovernment;the "victim" from the "target," the overt threat made against one might be different another. A Libya that nominally governmentwith the intentof intimidating threatenedIsrael mightbe aiming its terror Americans. at The target might be a populace, not a government. Panic, evacuation, or politicalpressure on a government, mass civil disobedience or obstruction could be the response either to mere announcement or to a threat aimed obliquely or directlyat the population. Winston Churchill was reportedly thatthe German bombing of London mightcause concerned, at least briefly, disorderlyevacuation, and those were prettysmall bombs. would be to detonate a weapon on The most active mode of intimidation a live target. Hiroshima represented intimidation:the object was not to eliminate that city; the true targetwas the emperor's palace in Tokyo. A weapon could disable a nation by destroyingthe center of the capital city; installation and a weapon or several mightbe used by surpriseon a military But even on a military or troop concentration. target,nuclear weapons might a a be intended more to terrorize populace or to intimidate governmentthan might be just to cause local blast and radiation damage. (A nuclear facility judged an appropriate targetby an organization that wanted the drama of militaryuse but wanted to target hearts and minds rather than bodies.) Detonation against some specifictargetis evidentlymore likelyif there is a war on.

NuclearTerrorism 71 |

Not out of the question is the possibilitythat a weapon would be offered forsale. Whethera dedicated group of scientistsand engineers would elect to turna profit ratherthan perform diplomaticmiraclesstrainsmy credulity, but only somewhat; a weapon obtained opportunisticallyfrom a military arsenal seems more compatiblewith mercenary utilization.Puttinga weapon up to public auction would be a clever bit of mischief.It mightbe purchased by someone desiringthe weapon, or ransomed by a governmentthat could financepreclusivepurchase. There mightbe the usual difficulty with kidnap ransom, namely, of showing up to collectthe money or giving one's mailing address; but the fatherof Patty Hearst was ordered to deliver ransom to a if thirdparty-food forthe poor-and that strategy, the owner isn't afterthe money itself,takes care of half the problem. The other half is crediblysurrenderingthe weapon. An outside possibilityis that an organizationacquiring a nuclear weapon would surrenderit with ceremony,not against ransom or on any conditions or concessions. It could be an effective anti-nucleardemonstration.Such a consummation seems incompatiblewith the deadly serious and dangerous a task of designing and constructing weapon; but a less professional organization that obtained a militaryweapon by theftor hijacking might be temptedto dramatizedisarmamentin thatfashion. (Or theymightannounce a contest:now that they have it, what should they do with it?) Use Strategies Terrorist for Nuclear weapons are enormous, discrete,unrenewable, and scarce. They are too valuable to waste, or to entrustto any but the most reliable operatives. demand in They are out of proportionto most of the things that terrorists response to some finitethreat,like something or somebody held hostage. terrorist has a single finite threat-the Americans Only recently opportunistic captured with the embassy in Teheran-been parlayed into a quasi-permanent source of intimidation.Assassination of a head of state is an act to compare to a nuclear threator nuclear test, but assassination doesn't lend itselfto a protractedstrategyof coercion. Few terrorist incidents of recent contain interestingsuggestions for terrorist decades therefore activityon a nuclear scale. A characteristic nuclear weapons, unlike live hostages, occupied buildof ings, hijacked ships or aircraft, stolen precious objects, is that thereis no or inherentlimitationon how long a nuclear threatcan last and no necessity

International Security| 72

forsurrenderof the weapon at the end of a successfulnegotiation.In ancient times hostages were taken in large numbers as securityagainst the good behavior of a tribeor a town or a nation. In recenttimes,with the ambiguous exception of Iran, terrorist hostages have been dynamicallyunstable assets, in and safetyeitherforcaptives not capable of being held indefinitely comfort or captors. The ultimate mobilityand securityof the captors depends on release fromas well as release of the hostages. We have to look back to the occupied countries of World War II, or eastward to more recent military occupations, to find instances of "steady-state" ratherthan "episodic" hostages. An important use distinction the terrorist of nuclear weapons, whether for by a small nation or organizationor by great nations like France, the United States, and the Soviet Union, is between deterrence and compellence. (I apologize for the word "compellence," which I introduced some years ago which has been understood since forthis purpose, but not forthe distinction, historicaltimes.) By "deterrence" I mean, of course, inducing an adversary or a victimnot to do something,to continue not doing something.The word takes the preposition"from." Accordingto my dictionary, discourage or "to keep (a person) fromdoing somethingthroughfear,anxiety,doubt, etc." It is the more passive kind of coercion. By "compellence" I mean inducing a person to do something through fear, anxiety,doubt, etc. "Compel" takes the preposition"to." My dictionary containsno word specialized toward this more active kind of coercion, so I coined the word. Deterrence is simpler. The command to do somethingrequires a date or deadline; to keep on not doing something is timeless. Acquiescence to a compellent threatis visiblyresponsive; doing nothingin face of a deterrent threatis not so obvious. Acquiescence to a compellentthreatinvitesanother demand; complyingwith a deterrent leaves things unchanged and leads to no sequel. Compellent threatshave to be spelled out: "go back" needs an indicationhow far,"give help" an indicationof what and how much; while "don't" usually takes its definitionfromwhat exists. And if a compellent threat is met by inaction, no event or initiativetriggersor mandates the threat,unless softlyand gradually execution, while violation of a deterrent done, initiatesthe action. In militaryaffairsthe deterrentthreat can often target the same activitythat constitutesthe violation, while a compellent threat must often find a target disconnected from the desired response: "don't advance or I'll shoot" makes the connection,while "send money or ..." cannot usually targetthe money but must findsome linkage elsewhere.

NuclearTerrorism 73 |

I propose that terrorist nuclear threatshave a comparativeadvantage toward deterrence.The more familiarterrorist actions of recentyears had the dynamics and disconnectedness of compellence. Organizations making nuclear threats, like nuclear nations, have a credibilityproblem; deterrent threatsare more credible. One of the great advantages of deterrentthreats is thattheyoftendo not need to be articulated.They are typically addressed to some obvious overt act. If you draw a gun on somebody approaching he may stop whether or not you tell him to; if you want him to turn around, take the keys fromhis pocket, and throw them to you over his shoulder, you have to say something. Not only does the deterrentthreateconomize communicationand minimize ambiguity,it permits one to create a threat without doing anything belligerent, without acknowledging, even while denying,thatone is threatening. It was easy to draw up a list of things that the U.S. governmentmight have done but didn't do forfear of jeopardizing the hostages in Teheran. It is not so easy to draw up a convincinglist of affirmative U.S. actions that were coerced throughthe hostages, except forthe finalransom. A second distinction terrorist is and for nuclear strategy between military civilian targets.A nuclear threat,and especially a detonation,will frighten, terrorize,and intimidate,whatever the targetindicated in the threator attacked by a bomb. But if the unofficialmini-stateaspires to international diplomacy, and identifiesits role as taking part in internationalconflict, it especially military conflict, may want to legitimizeits possession of weapons, its campaign of intimidation,even its use of nuclear explosives, by nominally confiningits attention to militarytargets. It may not want to discreditthe side that it favors. If the organizationconsiders itselfimmune to reprisalbecause it has no homeland, and its location is unknown, it may a not wish to make its intended beneficiary targetforreprisal.As in a conflict between India and Pakistan, Argentinaand Brazil, Israel and Syria, Cambodia and Vietnam, or South Africaand any black nation of the continent, and terrorist organizationsmay not want to appear inhumane and terrorist, may select military targetsto attackor to threaten. Similarly,the actions deterredmay be military. Justas the United States would not consider nuclear threats except against some military"aggression," and all nations that make referenceto nuclear weapons of theirown do so in referenceto self-defenseor commitmentsin defense of allies, the terrorist organization may adopt the custom and confine its threatsto the deterrenceof military actions.

International Security174

than France or the United In doing so it need not look more terrorizing States. By eschewing "massive retaliation"against homeland populations, and avoiding threatsof destroying "enemy societiesas such," it may legitimize its nuclear role and appear less inhumane or destructive than the greater the nuclear powers. Whateverit achieves forthem, striking posture may not cost them much. In choosing the actions to deter and the targetsto threaten,nuclear terroristsmay enjoy the advantage of multiple victims. When an action they want to detercan be obstructedby any one of several deterrableactors,they may select their victims and targets for likely compliance. Let me use a historical instance as an example, one that does not have to be wholly invented. During the October War of 1973 the United States was denied by ammumany European countriesthe privilegeof refuelingaircraft carrying nition and equipment forIsrael. No nuclear threatwas needed: it was megatons of energyto be withheld ratherthan delivered that deterredcooperation with the United States. But Portugal,justifying two decades of pretense that that countryhad anythingto contributein NATO, allowed use of airfields in the Azores. A pro-Arab terrorist organization,acting on its own or frontingfor a government that had a few nuclear weapons, might have declared the resupplyof Israel an aggressiveintrusioninto a military conflict, one thattheywould resistwith theirmodest nuclearforce.They could simply threatento explode theirweapon near a base on which U.S. planes were ceased. Assuming they could prove possession, refuelingunless the airlift theirabilityto do what theythreatenedcould be credible. The United States the Portugese governmentmight be demightbe deterred. Alternatively, terred.Or third,civilianemployees and othersnear the threatenedsite could be motivatedto strike,to obstruct, evacuate. If the threatwere not heeded, to and a weapon were detonated in the vicinity, perpetrators the would at least a have caused minimalciviliandamage in attacking military operation. Whatever the political and diplomatic consequences of such a stunning eventand on thatI am not even going to conjecture-it at least does not appear a wholly self-defeating threat,or even a self-defeating action, fromthe point to of view of an organizationwholly sympathetic the Syrian-Egyptiancause. Notice that if we go back and replay the scenario, allowing some Israeli that might entail eqully imaginative targeting,we nuclear counter-threat have an indeterminatesituation, one in which the fact that the nuclear organization is not a recognized nation does not make much difference.I have heard comparable scenarios conjecturinghow a governmentlike that of Libya, if it had a few nuclear weapons, might operate through a front

| NuclearTerrorism 75

organization,or mightissue a threatanonymously,or mighteven detonate a weapon without acknowledging its role. I have heard scenarios in which some small nation's nuclear weapons are entrustedto the air forceor navy and escape the control(or appear or are alleged to have escaped the control) of the national government.Whetherthe nuclear-armedentityin these stories is a nation or not does not make much difference. Like governmentsin these scenarios an organization not a government would not necessarilycondition its nuclear participationon a request from the governmentfor whose benefit it appears to be intervening.The real purpose may not even be to help the "beneficiary." Governmentsare of course more susceptible to nuclear reprisal,or almost any other kind of reprisal,than an unattached or anonymous organization. Countries can be denied exports, blockaded, diplomaticallyostracized, or even subjected to non-nuclearattackiftheyappear about to create a nuclear infernoin a population center or to rupture a moratoriumon the use of an nuclearweapons thathas survivedsince V-J Day. In contrast, organization thatneeds only a small boat to dock in a metropolitan harbor,with a nuclear weapon on board and someplace to operate a two-way radio, can hardlybe starvedinto second thoughtsby denial of soybeans, military spare parts, or air traffic; and it evidently cannot be invaded or captured or we wouldn't have the problemin the first place. The difference undeniably important,but need not be decisive. Up to is the point of actuallyexploding a weapon, a national governmentmay be in counteractionthan an organization no more danger of diplomaticor military thatis not a government.Even a nation that actually used nuclear weapons against enemy troops or enemy population would not necessarily be subjected to nuclear reprisal. Fear of escalation, abhorrence of anti-population attacks, and an interestin enhancing rather than abandoning nuclear restraint,could inhibitsome of the potential nuclear retaliators,especially if other modes of reprisal or disabling attack were available. And if the nongovernmentorganization uses its nuclear status in militarysupport of a particularcountry,the countryso "helped" mightnot escape all blame. So the difference, is though important, not decisive. Safeguards AgainstNon-National Weapons Eventuallywe may need a domain of strategyfor coping with these lesser nuclear threats,coming from either national governmentsor non-governmentalorganizations.It is likelyto be different fromthe principlesand ideas

International Security 76 |

that have been developed during the past three decades for the SovietAmericanor NATO-Warsaw Pact confrontation, at least as complicated. and It will be especially complicatedby the utterlack of symmetry between the United States and any such nuclear adversaries or clients as we have been talkingabout. I do not propose to outline any such strategy here. But in preventingthe acquisition of nuclear capabilities by non-national entities,thereis at least one principlethatI thinkis undeniable: the best way to keep weapons and weapons-material out of the hands of non-governmental entitiesis to keep them out of the hands of national governments. Saying that doesn't get us very far, but it does remind us that this is the problem we have already worried about, or a large part of it. International war-war between nations that are enemies of each other-is only one relevant occasion when possession of nuclear weapons mightmake a terrible difference. The main military activity military of forcesaround the world is overthrowing their own government or fightingother militaryforces in the same country-air forceagainst navy, officers against enlisted men, East against West Pakistan, Moslem against Christian,royalistagainst anti-royalist. The risk that nuclear weapons, or the materialfromwhich such weapons can be assembled, would actuallyplay a role in hostilitiesseems to be enhanced by the fact that militaryaction in much of the world is so characterized by internaldisorder and loss of national controlover military forces. There is just a chance, maybe only a prayer, that some governmentsor heads of governmentmay appreciate the dangers, to theirown countryand to themselves, of acquiring a weapons capability.Aside frombecoming a targetfor nuclear blackmail or nuclear attackby joining the ranks of declared nuclear nations, some governmentsmight possibly recognize how divisive a small nuclear capabilitycould be. What should a wise head of governmentrespond if offeredimmediate delivery of a few nuclear weapons? I think,"Not yet-let me think about where to put them." A prudent citizen mightbe appalled if a truckarrived unexpectedlyand conspicuously delivered objects of greatvalue thathe had won in some lottery, inadvertently fearingthat robberswould arrivebefore safeguardscould be erected and he would be leftin worse shape than before the goods arrived. Ownership of nuclear weapons poses embarrassingquestions for a head of state. Does he trusthis senior officers to sufficiently put his weapons in hands? Can he dare to display his distrustby keeping them from military

NuclearTerrorism 77 |

the military?Would he have to provide them to the competing military services or can he elevate one service as the sole nuclear force?Could he, as SecretaryMcNamara could not in the early 1960s, get enthusiasticmilitary acquiescence in electronically safeguarded presidential control? And could he in some futurewar-a contingencythat has to be considered possible, otherwisewhy nuclearweapons?-let an armybe surrounded,immobilized, even captured, as suggested by the finalstage of the October War, without authorizing some use of these supposedly awesome weapons? And what use mightthatbe? And ifhe withheldthe weapons, wouldn't he then regret having possessed them? There is littlethat the United States can crediblydo to increase the alarm any heads of non-nucleargovernmentsmightfeel at the prospect of getting close to having the weapons. We have too evident an interestof our own. It must furthermore exceedinglydifficult get a president,defense minbe to ister, ministerfornuclear power, and chief of armed forcesto sit together around a table and acknowledge that theymay shortlybe on opposite sides of a coup or civil war, or evacuating in disorder, and that they would be therefore wise not to encumberthemselveswith as competitivea prize as a nuclear arsenal. I wish there were evidence that some heads of governmentshared my of misgivings.I have never heard representatives Pakistan or Iraq or Argentina explain that a positive interestin nuclear weapons should never be imputed to thembecause the weapons are farmore dangerous than theyare worth. We mightnot believe them if we heard them say it, but it would be good to know thatit had occurred to them.