Accidental Nuclear War.pdf

The New England Journal of Medicine

Special Report

ACCIDENTAL NUCLEAR WAR — A POST–COLD WAR ASSESSMENT

L ACHLAN F ORROW , M.D., B RUCE G. B LAIR , P H .D., I RA H ELFAND , M.D., G EORGE L EWIS , P H .D., T HEODORE P OSTOL , P H .D.,

V ICTOR S IDEL , M.D., B ARRY S. L EVY , M.D., H ERBERT A BRAMS , M.D., AND C HRISTINE C ASSEL , M.D.

A BSTRACT

Background In the 1980s, many medical organi-

zations identified the prevention of nuclear war as

one of the medical profession’s most important

goals. An assessment of the current danger is war-

ranted given the radically changed context of the

post–Cold War era.

Methods We reviewed the recent literature on the

status of nuclear arsenals and the risk of nuclear war.

We then estimated the likely medical effects of a sce-

nario identified by leading experts as posing a seri-

ous danger: an accidental launch of nuclear weap-

ons. We assessed possible measures to reduce the

risk of such an event.

Results U.S. and Russian nuclear-weapons sys-

tems remain on high alert. This fact, combined with

the aging of Russian technical systems, has recently

increased the risk of an accidental nuclear attack. As

a conservative estimate, an accidental intermediate-

sized launch of weapons from a single Russian sub-

marine would result in the deaths of 6,838,000 per-

sons from firestorms in eight U.S. cities. Millions of

other people would probably be exposed to poten-

tially lethal radiation from fallout. An agreement to

remove all nuclear missiles from high-level alert sta-

tus and eliminate the capability of a rapid launch

would put an end to this threat.

Conclusions The risk of an accidental nuclear at-

tack has increased in recent years, threatening a

public health disaster of unprecedented scale. Physi-

cians and medical organizations should work active-

ly to help build support for the policy changes that

would prevent such a disaster. (N Engl J Med 1998;

338:1326-31.)

the prevention of nuclear war should be one of the

medical profession’s most important goals. 5-9

CONTINUED DANGER OF A NUCLEAR

ATTACK

Although many people believe that the threat of a

nuclear attack largely disappeared with the end of

the Cold War, there is considerable evidence to the

contrary. 10 The United States and Russia no longer

confront the daily danger of a deliberate, massive

nuclear attack, but both nations continue to operate

nuclear forces as though this danger still existed.

Each side routinely maintains thousands of nuclear

warheads on high alert. Furthermore, to compensate

for its weakened conventional armed forces, Russia

has abandoned its “no first use” policy. 11

Even though both countries declared in 1994

that they would not aim strategic missiles at each

other, not even one second has been added to the

time required to launch a nuclear attack: providing

actual targeting (or retargeting) instructions is sim-

ply a component of normal launch procedures. 12-14

The default targets of U.S. land-based missiles are

now the oceans, but Russian missiles launched with-

out specific targeting commands automatically re-

vert to previously programmed military targets. 13

There have been numerous “broken arrows” (ma-

jor nuclear-weapons accidents) in the past, including

at least five instances of U.S. missiles that are capable

of carrying nuclear devices flying over or crashing in

or near the territories of other nations. 15,16 From

1975 to 1990, 66,000 military personnel involved in

the operational aspects of U.S. nuclear forces were re-

moved from their positions. Of these 66,000, 41 per-

URING the Cold War, physicians and

others described the potential medical

consequences of thermonuclear war and

concluded that health care personnel and

facilities would be unable to provide effective care to

the vast number of victims of a nuclear attack. 1-3 In

1987, a report by the World Health Organization

concluded, “The only approach to the treatment of

health effects of nuclear warfare is primary preven-

tion, that is, the prevention of nuclear war.” 4 Many

physicians and medical organizations have argued that

From the Division of General Medicine and Primary Care, Beth Israel

Deaconess Medical Center and Harvard Medical School, Boston (L.F.); the

Brookings Institution, Washington, D.C. (B.G.B.); Physicians for Social

Responsibility, Washington, D.C. (I.H.); Massachusetts Institute of Tech-

nology, Cambridge (G.L., T.P.); the Department of Epidemiology and So-

cial Medicine, Montefiore Medical Center and Albert Einstein College of

Medicine, New York (V.S.); Barry S. Levy Associates and Tufts University

School of Medicine, Boston (B.S.L.); the Department of Radiology and

the Center for International Security and Arms Control, Stanford Univer-

sity, Stanford, Calif. (H.A.); and Mount Sinai School of Medicine, New

York (C.C.). Address reprint requests to Dr. Forrow at the Division of Gen-

eral Medicine, Beth Israel Deaconess Medical Center, East Campus, 330

Brookline Ave., Boston, MA 02215.

1326

April 30, 1998

SPECIAL REPORT

cent were removed because of alcohol or other drug

abuse and 20 percent because of psychiatric prob-

lems. 17,18 General George Lee Butler, who as com-

mander of the U.S. Strategic Command from 1991

to 1994 was responsible for all U.S. strategic nuclear

forces, recently reported that he had “investigated a

dismaying array of accidents and incidents involving

strategic weapons and forces.” 19

Any nuclear arsenal is susceptible to accidental, in-

advertent, or unauthorized use. 20,21 This is true both

in countries declared to possess nuclear weapons

(the United States, Russia, France, the United King-

dom, and China) and in other countries widely be-

lieved to possess nuclear weapons (Israel, India, and

Pakistan). The combination of the massive size of

the Russian nuclear arsenal (almost 6000 strategic

warheads) and growing problems in Russian control

systems makes Russia the focus of greatest current

concern.

Since the end of the Cold War, Russia’s nuclear

command system has steadily deteriorated. Aging

nuclear communications and computer networks are

malfunctioning more frequently, and deficient early-

warning satellites and ground radar are more prone

to reporting false alarms. 10,22-24 The saga of the Mir

space station bears witness to the problems of aging

Russian technical systems. In addition, budget cuts

have reduced the training of nuclear commanders

and thus their proficiency in operating nuclear weap-

ons safely. Elite nuclear units suffer pay arrears and

housing and food shortages, which contribute to low

morale and disaffection. New offices have recently

been established at Strategic Rocket Forces bases to

address the problem of suicide 25 (and unpublished

data).

Safeguards against a nuclear attack will be further

degraded if the Russian government implements its

current plan to distribute both the unlock codes and

conditional launch authority down the chain of com-

mand. Indeed, a recent report by the Central In-

telligence Agency, which was leaked to the press,

warned that some Russian submarine crews may al-

ready be capable of authorizing a launch. 26 As then

Russian Defense Minister Igor Rodionov warned

last year, “No one today can guarantee the reliability

of our control systems....Russia might soon

reach the threshold beyond which its rockets and

nuclear systems cannot be controlled.” 24

A particular danger stems from the reliance by

both Russia and the United States on the strategy of

“launch on warning” — the launching of strategic

missiles after a missile attack by the enemy has been

detected but before the missiles actually arrive. Each

country’s procedures allow a total response time of

only 15 minutes: a few minutes for detecting an en-

emy attack, another several minutes for top-level de-

cision making, and a couple of minutes to dissemi-

nate the authorization to launch a response. 27,28

Possible scenarios of an accidental or otherwise

unauthorized nuclear attack range from the launch

of a single missile due to a technical malfunction to

the launch of a massive salvo due to a false warning.

A strictly mechanical or electrical event as the cause

of an accidental launch, such as a stray spark during

missile maintenance, ranks low on the scale of plau-

sibility. 29 Analysts also worry about whether com-

puter defects in the year 2000 may compromise the

control of strategic missiles in Russia, but the extent

of this danger is not known.

Several authorities consider a launch based on a

false warning to be the most plausible scenario of

an accidental attack. 20,29 This danger is not merely

theoretical. Serious false alarms occurred in the U.S.

system in 1979 and 1980, when human error and

computer-chip failures resulted in indications of a

massive Soviet missile strike. 10,30 On January 25,

1995, a warning related to a U.S. scientific rocket

launched from Norway led to the activation, for the

first time in the nuclear era, of the “nuclear suitcas-

es” carried by the top Russian leaders and initiated

an emergency nuclear-decision-making conference

involving the leaders and their top nuclear advisors.

It took about eight minutes to conclude that the

launch was not part of a surprise nuclear strike by

Western submarines — less than four minutes be-

fore the deadline for ordering a nuclear response

under standard Russian launch-on-warning proto-

cols. 10,24,27

A missile launch activated by false warning is thus

possible in both U.S. and Russian arsenals. For the

reasons noted above, an accidental Russian launch is

currently considered the greater risk. Several specific

scenarios have been considered by the Ballistic Mis-

sile Defense Organization of the Department of De-

fense. 31 We have chosen to analyze a scenario that

falls in the middle range of the danger posed by

an accidental attack: the launch against the United

States of the weapons on board a single Russian

Delta-IV ballistic-missile submarine, for two reasons.

First, the safeguards against the unauthorized launch

of Russian submarine-based missiles are weaker than

those against either silo-based or mobile land-based

rockets, because the Russian general staff cannot

continuously monitor the status of the crew and mis-

siles or use electronic links to override unauthorized

launches by the crews. Second, the Delta-IV is and

will remain the mainstay of the Russian strategic

submarine fleet. 27,32,33

Delta-IV submarines carry 16 missiles. Each mis-

sile is armed with four 100-kt warheads and has a

range of 8300 km, which is sufficient to reach al-

most any part of the continental United States from

typical launch stations in the Barents Sea. 34,35 These

missiles are believed to be aimed at “soft” targets,

usually in or near American cities, whereas the more

accurate silo-based missiles would attack U.S. mili-

Volume 338 Number 18

1327

The New England Journal of Medicine

tary installations. 36 Although a number of targeting

strategies are possible for any particular Delta-IV, it

is plausible that two of its missiles are assigned to at-

tack war-supporting targets in each of eight U.S. ur-

ban areas. If 4 of the 16 missiles failed to reach their

destinations because of malfunctions before or after

the launch, then 12 missiles carrying a total of 48

warheads would reach their targets.

hour (15 to 30 miles per hour), a 100-kt low-alti-

tude detonation would result in a radiation zone 30

to 60 km (20 to 40 miles) long and 3 to 5 km (2 to

3 miles) wide in which exposed and unprotected

persons would receive a lethal total dose of 600 rad

within six hours. 39 With radioactive contamination

of food and water supplies, the breakdown of refrig-

eration and sanitation systems, radiation-induced im-

mune suppression, and crowding in relief facilities,

epidemics of infectious diseases would be likely. 40

POTENTIAL CONSEQUENCES

OF A NUCLEAR ACCIDENT

We assume that eight U.S. urban areas are hit:

four with four warheads and four with eight war-

heads. We also assume that the targets have been se-

lected according to standard military priorities: in-

dustrial, financial, and transportation sites and other

components of the infrastructure that are essential

for supporting or recovering from war. Since low-

altitude bursts are required to ensure the destruc-

tion of structures such as docks, concrete runways,

steel-reinforced buildings, and underground facili-

ties, most if not all detonations will cause substan-

tial early fallout.

Deaths

Table 1 shows the estimates of early deaths for

each cluster of targets in or near the eight major ur-

ban areas, with a total of 6,838,000 initial deaths.

Given the many indeterminate variables (e.g., the al-

titude of each warhead’s detonation, the direction of

the wind, the population density in the fallout zone,

the effectiveness of evacuation procedures, and the

availability of shelter and relief supplies), a reliable

estimate of the total number of subsequent deaths

from fallout and other sequelae of the attack is not

possible. With 48 explosions probably resulting in

thousands of square miles of lethal fallout around

urban areas where there are thousands of persons

per square mile, it is plausible that these secondary

deaths would outnumber the immediate deaths

caused by the firestorms.

Physical Effects

Under our model, the numbers of immediate

deaths are determined primarily by the area of the

“superfires” that would result from a thermonuclear

explosion over a city. Fires would ignite across the

exposed area to roughly 10 or more calories of radi-

ant heat per square centimeter, coalescing into a gi-

ant firestorm with hurricane-force winds and average

air temperatures above the boiling point of water.

Within this area, the combined effects of superheat-

ed wind, toxic smoke, and combustion gases would

result in a death rate approaching 100 percent. 37

For each 100-kt warhead, the radius of the circle

of nearly 100 percent short-term lethality would be

4.3 km (2.7 miles), the range within which 10 cal

per square centimeter is delivered to the earth’s sur-

face from the hot fireball under weather conditions

in which the visibility is 8 km (5 miles), which is low

for almost all weather conditions. We used Census

CD to calculate the residential population within

these areas according to 1990 U.S. Census data, ad-

justing for areas where circles from different war-

heads overlapped. 38 In many urban areas, the day-

time population, and therefore the casualties, would

be much higher.

Medical Care in the Aftermath

Earlier assessments have documented in detail the

problems of caring for the injured survivors of a nu-

clear attack: the need for care would completely

overwhelm the available health care resources. 1-5,41

Most of the major medical centers in each urban

area lie within the zone of total destruction. The

number of patients with severe burns and other crit-

ical injuries would far exceed the available resources

of all critical care facilities nationwide, including the

country’s 1708 beds in burn-care units (most of

which are already occupied). 42 The danger of intense

radiation exposure would make it very difficult for

emergency personnel even to enter the affected ar-

eas. The nearly complete destruction of local and re-

gional transportation, communications, and energy

networks would make it almost impossible to trans-

port the severely injured to medical facilities outside

the affected area. After the 1995 earthquake in

Kobe, Japan, which resulted in a much lower num-

ber of casualties (6500 people died and 34,900 were

injured) and which had few of the complicating fac-

tors that would accompany a nuclear attack, there

were long delays before outside medical assistance

arrived. 43

Fallout

The cloud of radioactive dust produced by low-

altitude bursts would be deposited as fallout down-

wind of the target area. The exact areas of fallout

would not be predictable, because they would de-

pend on wind direction and speed, but there would

be large zones of potentially lethal radiation expo-

sure. With average wind speeds of 24 to 48 km per

FROM DANGER TO PREVENTION

Public health professionals now recognize that

many, if not most, injuries and deaths from violence

1328

April 30, 1998

SPECIAL REPORT

T ABLE 1. P REDICTED I MMEDIATE D EATHS

FROM F IRESTORMS AFTER N UCLEAR D ETONATIONS

IN E IGHT U.S. C ITIES .

and even its most optimistic advocates predict that

it cannot be fully protective. Furthermore, the esti-

mated costs would range from $4 billion to $13 bil-

lion for a single-site system to $31 billion to $60 bil-

lion for a multiple-site system. 46,47 In either case, the

system would not be operational for many years. 48

C ITY *

N O . OF

W ARHEADS

N O . OF

D EATHS

Atlanta

428,000

A Bilateral Agreement to Eliminate High-Level Alert Status

Since ballistic-missile defense offers no solution at

all in the short term and at best an expensive and in-

complete solution in the long term, what can the

United States as well as other nations do to reduce

the risk of an accidental nuclear attack substantially

and quickly? The United States should make it the

most urgent national public health priority to seek a

permanent, verified agreement with Russia to take

all nuclear missiles off high alert and remove the ca-

pability of a rapid launch. 49 This approach is much

less expensive and more reliable than ballistic-missile

defense and can be implemented in short order. In

various forms, such an agreement has been urged by

the National Academy of Sciences, 50 the Canberra

Commission, 51 General Butler and his military col-

leagues throughout the world, 52 and other experts,

such as Sam Nunn, former chairman of the U.S.

Senate Armed Services Committee, and Stansfield

Turner, former director of the Central Intelligence

Agency. 10,20,53 The Joint Chiefs of Staff and an inter-

agency working group are completing a detailed

study of de-alerting options that will be presented to

Defense Secretary William Cohen. 10

Major improvements in nuclear stability can be

achieved rapidly. In the wake of the 1991 attempted

coup in Moscow, Presidents George Bush and Mik-

hail Gorbachev moved quickly to enhance nuclear

safety and stability by taking thousands of strategic

weapons off high alert almost overnight. 27 Today,

there are specific steps that the United States can

take almost immediately, since they require only the

authority of a presidential directive. These steps in-

clude putting in storage the warheads of the MX

missiles, which will be retired under Strategic Arms

Reduction Treaty (START) II in any case, and the

warheads of the four Trident submarines that will be

retired under START III; placing the remaining U.S.

ballistic-missile submarines on low alert so that it

would take at least 24 hours to prepare them to

launch their missiles; disabling all Minuteman III

missiles by pinning their safety switches open (as was

done with the Minuteman II missiles under Presi-

dent Bush’s 1991 directive); and allowing Russia to

verify these actions with the on-site inspections al-

lowed under START I. Similar measures should be

taken by the Russians. 27,49 These steps — all readily

reversible if warranted by future developments or if

a permanent bilateral agreement is not reached —

would eliminate today’s dangerous launch-on-warn-

ing systems, making the U.S. and Russian popula-

Boston

609,000

Chicago

425,000

New York

3,193,000

Pittsburgh

375,000

San Francisco Bay area

739,000

Seattle

341,000

Washington, D.C.

728,000

Total

6,838,000

*The specific targets are as follows: Atlanta — Peachtree

Airport, Dobbins Air Force Base, Fort Gillem, Fort McPher-

son, Fulton County Airport, Georgia Institute of Technolo-

gy, Hartsfield Airport, and the state capitol; Boston —

Logan Airport, Commonwealth Pier, Massachusetts Insti-

tute of Technology, and Harvard University; Chicago —

Argonne National Laboratory, City Hall, Midway Airport,

and O’Hare Airport; New York — Columbia University, the

George Washington Bridge, Kennedy Airport, LaGuardia

Airport, the Merchant Marine Academy, Newark Airport,

the Queensboro Bridge, and Wall Street; Pittsburgh — Car-

negie Mellon University, Fort Duquesne Bridge, Fort Pitt

Bridge, Pittsburgh Airport, and the U.S. Steel plant; San

Francisco Bay area — Alameda Naval Air Station, the Bay

Bridge, Golden Gate Bridge, Moffet Field, Oakland Airport,

San Francisco Airport, San Jose Airport, and Stanford Uni-

versity; Seattle — Boeing Field, Seattle Center, Seattle–Taco-

ma Airport, and the University of Washington; and Washing-

ton, D.C. — the White House, the Capitol Building, the

Pentagon, Ronald Reagan National Airport, College Park

Airport, Andrews Air Force Base, the Defense Mapping

Agency, and Central Intelligence Agency headquarters.

and accidents result from a predictable series of

events that are, at least in principle, preventable. 44,45

The direct toll that would result from an accidental

nuclear attack of the type described above would

dwarf all prior accidents in history. Furthermore,

such an attack, even if accidental, might prompt a re-

taliatory response resulting in an all-out nuclear ex-

change. The World Health Organization has esti-

mated that this would result in billions of direct and

indirect casualties worldwide. 4

Limitations of Ballistic-Missile Defense

There are two broad categories of efforts to avert

the massive devastation that would follow the acci-

dental launch of nuclear weapons: interception of

the launched missile in a way that prevents detona-

tion over a populated area and prevention of the

launch itself. Intercepting a launched ballistic missile

might appear to be an attractive option, since it

could be implemented unilaterally by a country. To

this end, construction of a U.S. ballistic-missile de-

fense system has been suggested. Unfortunately, the

technology for ballistic-missile defense is unproved,

Volume 338 Number 18

1329

The New England Journal of Medicine

tions immediately safer. Both nations should then

energetically promote a universal norm against main-

taining nuclear weapons on high alert.

parties should cooperate to ensure that these meas-

ures are implemented rapidly.

CONCLUSIONS

The time, place, and circumstances of a specific

accident are no more predictable for nuclear weap-

ons than for other accidents. Nonetheless, as long as

there is a finite, nonzero, annual probability that an

accidental launch will occur, then given sufficient

time, the probability of such a launch approaches

certainty. Until the abolition of nuclear weapons re-

duces the annual probability to zero, our immediate

goal must be to reduce the probability of a nuclear

accident to as low a level as possible. Given the mas-

sive casualties that would result from such an acci-

dent, achieving this must be among the most urgent

of all global public health priorities.

The Role of Physicians

In awarding the 1985 Nobel Peace Prize to Inter-

national Physicians for the Prevention of Nuclear

War, the Nobel Committee underscored the “con-

siderable service to mankind” that physicians have

performed by “spreading authoritative information

and by creating an awareness of the catastrophic

consequences of atomic warfare. This in turn con-

tributes to an increase in the pressure of public op-

position to the proliferation of nuclear weapons and

to a redefining of priorities. . . .” 54 No group is as

well situated as physicians to help policy makers and

the public fully appreciate the magnitude of the dis-

aster that can ensue if changes in the alert status of

all nuclear weapons are not instituted. 5,6,8,9,55-57

The only way to make certain that an accidental

(or any other) nuclear attack never occurs is through

the elimination of all nuclear weapons and the air-

tight international control of all fissile materials that

can be used in nuclear weapons. In 1995, the World

Court stated that the abolition of nuclear weapons

is a binding legal obligation of the United States,

Russia, and all signatories to the Nuclear Nonprolif-

eration Treaty, under Article 6. 58 Preferring the term

“prohibition” to “abolition,” the Committee on In-

ternational Security and Arms Control of the U.S.

National Academy of Sciences concluded in its 1997

report, “The potential benefits of comprehensive

nuclear disarmament are so attractive relative to the

attendant risks — and the opportunities presented

by the end of the Cold War...are so compelling

— that...increased attention is now warranted

to studying and fostering the conditions that would

have to be met to make prohibition desirable and

feasible.” 59

Leading U.S. medical organizations, including the

American College of Physicians and the American

Public Health Association, have already joined Phy-

sicians for Social Responsibility, International Physi-

cians for the Prevention of Nuclear War, and over

1000 other nongovernmental organizations in 75

nations to support Abolition 2000, which calls for a

signed agreement by the year 2000 committing all

countries to the permanent elimination of nuclear

weapons within a specified time frame. 60-63 The

American Medical Association has recently endorsed

the abolition of nuclear weapons, 64 as have the Can-

berra Commission, 51 military leaders throughout the

world, 52 major religious organizations, 61,65 and over

100 current and recent heads of state and other sen-

ior political leaders. 66,67 Some supporters of the ab-

olition of nuclear weapons have specifically called for

immediate steps to eliminate the high-level alert sta-

tus of such weapons, as urgent interim measures. All

Supported by the Albert Schweitzer Fellowship and Physicians for Social

Responsibility.

We are indebted to Jose Berrocal, Gen. George Lee Butler, Thomas

Delbanco, Brian Forrow, Lawrence Gussman, and Bernard Lown.

REFERENCES

1. Ervin FR, Glazier JB, Arnow S, et al. Human and ecologic effects in

Massachusetts of an assumed thermonuclear attack on the United States.

N Engl J Med 1962;266:1127-37.

2. Sidel VW, Geiger HJ, Lown B. The physician’s role in the postattack

period. N Engl J Med 1962;266:1137-45.

3. Abrams HL. Medical resources after nuclear war: availability v need.

JAMA 1984;252:653-8.

4. Effects of nuclear war on health and health services. 2nd ed. Geneva:

World Health Organization, 1987.

5. Lown B, Chivian E, Muller J, Abrams H. The nuclear-arms race and the

physician. N Engl J Med 1981;304:726-9.

6. Relman AS. The physician’s role in preventing nuclear war. N Engl J

Med 1986;315:889-91.

7. Policies 520.999 and 520.997. Chicago: American Medical Association,

1981.

8. Cassel C, Jameton A. Medical responsibility and thermonuclear war.

Ann Intern Med 1982;97:426-32.

9. Lown B, Chazov EI. Physician responsibility in the nuclear age. JAMA

1995;274:416-9.

10. Hall B. Overkill is not dead. New York Times Magazine. March 15,

1998:42-9, 64, 76-7, 84-5.

11. National Academy of Sciences, Committee on International Security

and Arms Control. The future of U.S. nuclear weapons policy. Washington,

D.C.: National Academy Press, 1997:2.

12. Blair BG. Russian nuclear control: risks and solutions. Paper presented

to Foreign and Defense Ministry Officials of Finland, Helsinki, August 26,

1997.

13. Idem . Where would all the missiles go? Washington Post. October 15,

1996:A15.

14. Idem. Russian nuclear policy and the status of detargeting. Testimony

before the Subcommittee on Military Research and Development, House

Committee on National Security, Washington, D.C., March 13, 1997.

15. Leitenberg M. Accidents of nuclear weapons systems. In: SIPRI year-

book of world armaments and disarmament, 1977. Cambridge, Mass.:

M.I.T. Press, 1977:53.

16. Sagan SD. The limits of safety: organizations, accidents, and nuclear

weapons. Princeton, N.J.: Princeton University Press, 1993.

17. Abrams HL. Sources of human instability in the handling of nuclear

weapons. In: Solomon F, Marston RQ, eds. The medical implications of

nuclear war. Washington, D.C.: National Academy Press, 1986:490-528.

18. Idem . Human reliability and safety in the handling of nuclear weapons.

Science and Global Security 1991;2:325-49.

19. Butler GL. A call for nuclear abolition: remarks to the State of the

World Forum, San Francisco, October 3, 1996.

20. Nunn S, Blair BG. From nuclear deterrence to mutual safety. Wash-

ington Post. June 22, 1997:C1.

1330

April 30, 1998

Plik z chomika:

Inne pliki z tego folderu:

Inne foldery tego chomika: