Air power is an unusually seductive form of military strength, in part because, like modern courtship, it appears to offer gratification without commitment. Francis Bacon wrote of command of the sea that he who has it "is at great liberty, and may take as much and as little of the Warre as he will," and a similar belief accounts for air power's attractiveness to those who favor modest uses of force overseas. Statesmen may think that they can use air attacks to engage in hostilities by increments, something ground combat does not permit. Furthermore, it appears that the imminent arrival of so-called nonlethal or disabling technologies may offer an even more appealing prospect: war without casualties.

This rise in air power's stock comes from its success in the Persian Gulf War. In the view of some, that conflict represented the opening shot of a fundamental transformation in the nature of warfare, a "military-technical revolution" as the Russians have termed it for more than a decade. Thus the Russian military sadly read the outcome of a war that vindicated their predictions even as it sealed their profound sense of inferiority vis-à-vis the United States. Secretary of Defense Richard Cheney agreed: "This war demonstrated dramatically the new possibilities of what has been called the 'military-technological revolution in warfare.'" Others, outside the Bush administration, expressed this view no less enthusiastically. William Perry, now Deputy Secretary of Defense, wrote in Foreign Affairs that "a new class of military systems...gave American forces a revolutionary advance in military capability."1

The lopsided struggle with Iraq has already affected the way Americans understand modern war, inducing the ornithological miracle of doves becoming hawks. More than one distinguished commentator who had reservations about aerial bombardment in the Persian Gulf expressed a newfound belief in its utility as a tool of American foreign policy in the Balkans. Thus, Anthony Lewis of The New York Times wrote during the Persian Gulf War in disgust at the ruin wrought by aerial bombardment, "We should never again tolerate anyone who talks about 'surgical strikes.'" Since then he has developed a keener appreciation of air power, asserting that "a few air strikes in Dubrovnik" would have stopped the Yugoslav horrors in 1991. There is a "straightforward way to apply force" in Bosnia that involves "minimum risk" and provides a course that is not merely right but "clear and doable"--precision air attacks.2

Many of these individuals came away from the Persian Gulf War with a far healthier respect for air power, believing it had made all the difference. Indeed it had.3 Some 52,000 air-to-surface sorties delivered approximately 210,000 unguided bombs, 9,300 guided bombs, 5,400 guided air-to-surface missiles and 2,000 anti-radar missiles; American forces also hurled more than 300 cruise missiles at the enemy. To what effect? Of its 700 aircraft, the Iraqi air force lost 33 in the air; approximately 140 perished in hardened aircraft shelters and more than 120 were flown to Iran. The Iraqi air defense system succumbed within days--really, hours--to an extremely sophisticated attack, and it managed to shoot down only a tiny fraction of the attacking aircraft. The Iraqi electrical grid, oil refineries and most of the telephone and communications system stopped functioning. The Iraqi ground forces in the Kuwait theater attracted the most attention from coalition air forces. Before the ground war began on February 24, the Republican Guard located in the northern part of the theater had lost nearly a quarter of its armor to air attacks, and frontline units had suffered even heavier losses. Moreover, air power had completely disrupted Iraqi logistics and immobilized the Iraqi army. Aircraft operating around the clock crushed the one attempt by the Iraqis to launch a large-scale operation--the two-division thrust southward that barely got over the Saudi border at the town of Khafji. Although ground action necessarily consummated the final victory for coalition forces, air power had made the final assault as effortless as a wartime operation can be.

The Persian Gulf War has an importance that goes beyond its immediate (and considerable) effects on politics there. If the claims of air power advocates are correct, the United States has acquired a military edge over conventional opponents comparable to that exercised in 1898 by the soldiers of Lord Kitchener against the sword-wielding dervishes of the Sudan. The way would lie open to a reorientation of the defense budget toward an air-dominated force structure.4 Indeed, if air power performed as spectacularly as first appeared in the Persian Gulf, the way to an American-policed world order might look remarkably smooth. Was, then, the Persian Gulf War a major departure in the history of warfare, and does it point the way to unshakable American military preeminence?


The war saw the first extensive use of some new technologies. The F-117 stealth fighter penetrated the Iraqi radar system safely and secretly. The Joint Surveillance Target Attack Radar System, still undergoing testing, detected Iraqi trucks and tanks along highways in Kuwait, and an array of satellites provided unparalleled support to military commanders for intelligence gathering, map-making, communication, navigation, meteorology and missile-launch detection. Still, the bulk of the work came from much older systems and mundane technologies such as air refueling (which was required for three fifths of all combat missions). Most of the weapons used in the Persian Gulf dated back two decades or more. The airframes of Air Force B-52 and Navy A-6 bombers had seen at least 20 years of active service, although they employed newer electronics. Guided bombs first appeared in the Vietnam War, when American aircraft dropped 4,000 or so on the Vietnamese communists. Even modern, first-line aircraft such as the F-16 have been in use for almost a decade. The "military-technical revolution" sparkled in the new systems, but it drew as much on considerably more mature technologies.

The most profound change in military technology, however, was the vast increase in usable and communicable information. This war saw, for example, the first combat use of the Global Positioning System, which allows units to locate their position in three dimensions by a mere press of a button. (The Gulf War was the last war in which only one side will have this knowledge. The technology for tapping into the GPS is widespread, easy to use and relatively low cost. And even if American operators were to attempt limiting access by degrading the GPS satellite signals, a clever enemy could largely circumvent such spoofing.) Space-based information gathering systems also churn out vast quantities of information, and even a Third World country, for example, can tap into international weather satellites or buy militarily useful commercial imagery from the French satellite imaging system, commonly known as SPOT.

Conventional warfare depends increasingly on the skillful manipulation of electronically transmitted information. The advantage goes overwhelmingly to combatants who can bring together information from many sources, updating old databases (e.g., removing from the target list a radar station already destroyed) and acting on perishable information. Countries such as the United States or, to a lesser extent, Israel have enormous and growing advantages in these areas. In the future, the struggle for information may take the place that the contest for geographic position took in the past. But the information explosion does not mean, as a casual observer might think, that war will become more transparent to those who conduct it. Clausewitz's "fog of war" may now descend on the battlefield not so much from a paucity of good information as from a plethora of half-knowledge. A fog it remains, however, and it lay heavily over the Persian Gulf War.

Any editor knows that the advent of the personal computer and facsimile machine has prompted authors to fiddle with articles constantly and to submit them at the very last moment. Similarly, the flood of combat information prompts commanders to change targets or tactics at the last moment. In the Gulf War commanders changed one fifth of all missions during the few hours between the time staffs printed the centralized air tasking order and the time aircraft took off. They made many more changes before the ATO was officially issued and still more after aircraft had left their bases. Sometimes these decisions made sense; other times they did not. In all cases they created great uncertainty among the pilots flying the missions.

Despite the wisdom of proverbs, pictures sometimes lie, or at least deceive. Coalition air planners in Riyadh tried to do their own bomb damage assessment by looking at videotape footage of laser-guided bomb strikes. Lacking time (and, in some cases, experience in photo interpretation), they sometimes misinterpreted what they saw--mistaking an exploding fuel truck or decoy for a mobile missile launcher, for example, or thinking that a bomb bursting on a concrete roof meant that the contents of the building had been destroyed. The short decision times created by modern weapons can also force quick decisions on the basis of electronically gathered information whose ambiguity may not be readily apparent. The shooting down of an Iranian airliner in 1988 by an American naval cruiser is a case in point. Time pressure created by an abundance of data breeds longer-term problems as well. The constant pressure of the data stream, together with the growth of nighttime operations, means that leaders try to keep on top of events at the cost of sleep and acuity.

Combat information increasingly takes the form of abstract representations of reality compiled from multiple sources. It becomes harder to discriminate between different types of information when a distant, anonymous expert or even a machine has done the sifting. Since the Vietnam War, American generals have decried civilian micro-management of military operations, an indictment only partly warranted but accepted uncritically by politicians as well as soldiers. Today, however, the danger of military micro-management looms much larger. A general in Washington, an admiral in a command ship or a theater commander in rear headquarters may have access to almost the same information as a forward commander, and in some cases more. Those distant commanders will often succumb to the temptation to manipulate individual units in combat accordingly.

Dependence on vast quantities of electronic information, of course, poses certain risks. During the war, pilots complained of having to fly missions without the kind of target graphics they had used in training. In the future, soldiers may become overly dependent on detailed, well-presented and accurate information that simply may not exist in wartime. And as the verisimilitude of computer simulators and war games increases, future warriors may paradoxically find themselves all the more at a loss when the real world differs sharply from a familiar cyberworld.

Furthermore, the more sophisticated and expensive the information gathering system, the greater the premium opponents will put on disabling it with anything from electronic attack to homing missiles. The payoff for shooting down a state-of-the-art radar surveillance aircraft, for example, will surely attract intense efforts to do so. The Persian Gulf War also demonstrated a trend toward what one might call the speciation--that is, the evolution of distinct families--of munitions. The new munitions can, in theory and often in practice, achieve effects unimaginable with the conventional, high-explosive bomb. Anti-radiation missiles, for example, not only destroyed Iraqi radars, but intimidated air defense crews from turning them on in the first place. Specially tipped laser-guided bombs punctured hardened aircraft shelters immune to regular high explosives.

But as air-delivered munitions have become increasingly specialized in their effects, they have become susceptible to unintentional misuse. Air campaign planners expected extremely high kill rates against Iraqi armor, for example, based on the use of a scatterable mine, CBU-89. But the projected lethality of CBU-89 relied on calculations made for a war in Europe, fought against Soviet tank armies on the move. Against static, dug-in Iraqi tanks, CBU-89 had much less to offer. The speciation of munitions brings unusual capabilities, but it also poses the risk of creating forces so specialized that they lack flexibility, and weapons so expensive that commanders will have only slender inventories to use when a war starts. Moreover, "dumb" (or at least relatively unintelligent) weapons will keep a place. Massive raids by B-52s raining down conventional bombs helped crush the morale of Iraqi soldiers and smash the large military-industrial facilities that figured so prominently in Saddam Hussein's aspirations for power.


The successes of the air campaign in the gulf rested almost as much on organizational innovations as on technology. To speak of a revolution in warfare as a purely technological affair is to miss half the significance of the war. In the Gulf War, for example, Lieutenant General Charles Horner, the commander of CENTAF (the Air Force component of Central Command), also controlled in some measure the airplanes of all the services, as well as helicopters flying above 500 feet and Navy Tomahawk cruise missiles. In this respect he embodied a doctrine dear to airmen for half a century: "Control of available air power must be centralized and command must be exercised through the Air Force commander if this inherent flexibility and ability to deliver a decisive blow are to be fully exploited."5 "Centralized planning, decentralized execution" remains a catchphrase of Air Force doctrine, much as "don't divide the fleet" preoccupied American naval strategists in earlier times.

In practice, though, Horner's authority had its limits. The Navy controlled maritime air operations, the Marines determined the assignments of their short take-off and landing aircraft plus at least half of their fighter-bombers, while the allies exercised discretion regarding which targets they would attack. Special operations forces--which in effect constitute a fifth armed service--continued to struggle for control of their own air operations.

Horner, directing 1,800 combat aircraft, had a staggering fleet at his command. Nonetheless, even his gently wielded centralized control elicited suspicion and hostility from officers in other services who feared an Air Force attempt to dominate all aerial warfare. Grudgingly conceding the necessity of a single command center, they argued that in theory it could dilute the synergy of, for example, Marine air and ground forces, and that in practice it proved cumbersome and slow.

The centralized control of air power made for a much more coherent campaign than would otherwise have occurred. But, as officers from the other three services bitterly observed, the centralized control rested overwhelmingly in the hands of Air Force officers. Although the core planning staff, the so-called "Black Hole," included representatives from the Navy, Army and Marine Corps, as well as the British and Saudi air forces, its membership came predominantly from the Air Force. In theory a joint targeting board should have selected targets; in practice it did very little. Furthermore, much of the inspiration for the Black Hole's targeting decisions came from an Air Force staff in the Pentagon, an organization known as "Checkmate" led by Colonel John Warden, a fervent believer in air power.

Thus, an Air Force staff (nominally under Joint Staff auspices) dominated the flow of targeting information and proposals to the theater. This certainly violated the intent if not the letter of a command system that in theory excluded the service staffs from operational activity. But in fact this arrangement proved fortuitous because the Air Staff could tap a far wider range of expertise than could Gen. Norman Schwarzkopf's understaffed and overworked Air Force staff in Riyadh, which had to cope first with the task of deploying a vast force to the Persian Gulf and then with the myriad chores of daily flight activity and defensive planning. Furthermore, Air Force dominance in both Washington and Riyadh meant that the plan was conceptually consistent. Normal procedures, which give each service an equal voice in the name of cooperative participation, would not have done nearly so well.

The procedurally orthodox abhor the notion of Washington service staffs feeding operational suggestions to a theater staff. Such a practice subverts the idea--enshrined in the Goldwater-Nichols Act of 1986--of theater commanders as semi-autonomous warlords who take only the broadest strategic direction from Washington. But American defense planners should look at what happened and ask whether these improvisations do not point the way to greater effectiveness. After several decades of insisting that the word "service" means "parochial," military reformers might ponder the individual merits of the services, each of which can pool a great deal of operational expertise along with a common world view and an esprit de corps difficult to find among a mélange of officers.

The abundance of reliable and secure voice, data and facsimile communications to (and within) the theater transformed command and control. Communications technology subverted hierarchies and rendered abundant exchanges between the theater and the United States both inevitable and desirable. During the Persian Gulf War staffs in Colorado relayed warnings of Scud launches to Riyadh and Tel Aviv, and the now-defunct Tactical Air Command near Norfolk, Virginia, managed CENTAF logistical accounts. Watching CNN and using other sources of instantaneous news, the chairman of the Joint Chiefs of Staff (and, to only a slightly smaller degree, his civilian superiors) monitored the day-to-day activities of U.S. Central Command's (CENTCOM) forces. The new technologies threaten age-old principles such as unity of command and delegation of authority. Those pieces of military folk wisdom have so much authority that they will persist in peacetime even if they must disappear in war. A new concept of high command, one that acknowledges that technology inevitably diffuses authority, will have to take root.

Information gathering and processing technology can help but not solve the problem of bomb damage assessment. In many cases, commanders sent out sorties uncertain about the degree of damage a target had already received. Part of the problem stemmed from excessive reliance on intelligence derived from satellites rather than locally controlled reconnaissance aircraft. The theater intelligence staff was small in number and had little experience in tasking the array of satellites at its disposal. There remained, moreover, the sheer difficulty of knowing what damage has been done. From an overhead photograph, for example, it may prove difficult to figure out whether a small black hole on top of a hardened aircraft shelter indicates a hit by a dud bomb, an explosion in the thick, rubble-filled space between the shelter's inner and outer walls, an explosion within the shelter or an artful paint job by Iraqi camouflage experts. And unless reconnaissance units can keep targets under near-constant surveillance from many angles and with different kinds of sensors, intelligence analysts may not know which projectile did what kind of damage. Finally, functional damage may differ sharply from physical damage. Air Force planners desired the first, hoping, for example, that a few hits on a command post would discourage the Iraqis from using it, even if according to normal measures of damage (which depend mainly on engineering criteria) the facility had not received a mortal blow. Not surprisingly, they became increasingly frustrated with intelligence reports that paid more attention to physical than functional damage.

The problem of bomb-damage assessment means that the fog of war will persist, although intelligence services will work to develop ever more sophisticated means of interpreting imagery and cross-checking damage through different sources of information. These individual uncertainties can add up to much higher aggregate levels of confusion. Hard as it may be to figure out what a particular target consists of and how to strike it, figuring out a target system--for example, how destruction of one microwave relay affects an entire telephone network--is even more difficult. It is depressing to recall that the United States began the war with two known Iraqi nuclear facilities on its target list, that it added six during the conflict and that another dozen were uncovered by U.N. inspectors only after the war. Furthermore, it appeared that the Iraqis had vitiated the effects of bombing by stripping these buildings of equipment and materials used in the nuclear program. This too was not known until after the war.

Although some aircraft, such as the F-117, had excellent videotape recorders, most could not take pictures of where their bombs or missiles had landed. Nor did the coalition have nearly enough unmanned aircraft or manned reconnaissance aircraft to do a proper job, given the magnitude of the effort. These shortfalls revealed an institutional failure before the war to accept the notion that knowing what a bomb has done is almost as important as delivering it. In future conflicts, where commanders might have less time or much smaller forces, an inability to track damage to an enemy could prove crippling. Even in retrospect it has proven extremely difficult to decipher the air war's effects. There was no comprehensive survey of the battlefield conducted in the wake of advancing coalition forces, in part because of deficient prewar planning and in part because of CENTCOM's antipathy to visiting study teams. The American armed forces could have done a far more thorough job of recording and analyzing battles as they unfolded, and certainly after they were over.


Reliance on air power has set the American way of war apart from all others for well over half a century. Other countries might field doughty infantry, canny submariners or scientific artillerists comparable in skill and numbers to America's. Only the United States, however, has engaged in a single-minded and successful quest for air superiority in every conflict it has fought since World War I. Air warfare remains distinctively American--high-tech, cheap in lives and (at least in theory) quick. To America's enemies--past, current and potential--it is the distinctively American form of military intimidation.

Air warfare plays to the machine-mindedness of American civilization. Aircraft can direct massive and accurate destructive force at key points without having to maneuver cumbersome organizations on land or sea. Air power can indeed overawe opponents, who know quite well that they cannot hope to match or directly counter American strength. On the other hand, these enemies will find indirect responses. The Saddam Husseins of the world have surely learned that they need not take American children hostage to deter bombardment; they can take their own citizens' young with no less effect. When F-117s struck the so-called al-Firdos bunker--a perfectly appropriate military target--on February 13, they apparently killed the wives and children of Iraqi leaders using the facility as a shelter. For the next four days all air operations against Baghdad ceased, and when they resumed, politically motivated controls reduced the number of targets to the barest handful. Mobility, when abetted by camouflage and tight communications security, can also shield a potential opponent from harm. Publicly available evidence does not suggest that air attacks destroyed any Scud missile launchers, for example.

The soldier or marine will surely say to the air power enthusiast that nothing can substitute for the man with the bayonet. True, but some politically desired effects (elimination of electrical power in Baghdad, for example) required no use of ground forces. And in some cases the United States has proved unwilling to use ground forces to achieve its objective (for example, the overthrow of Saddam). Air power may not decide all conflicts or achieve all of a country's political objectives, but neither can land power.

All forms of military power seem likely to benefit from the imminent arrival of "nonlethal" or "disabling" technologies, which offer the prospect of war without casualties. But here, perhaps, lies the most dangerous legacy of the Persian Gulf War: the fantasy of near-bloodless uses of force. Set aside, for the moment, the question of so-called nonlethal weapons. No military technology (indeed, no technology at all) works all the time. Inevitably, even the best-aimed laser-guided bomb will lose its fix on a target because of a passing cloud or a steering mechanism failure, and hurtle into an orphanage or hospital. As one wise engineer puts it, "[T]he truly fail-proof design is chimerical."6

In many cases, so-called nonlethal weapons will prove just the reverse. The occupants of a helicopter crashing to earth after its flight controls have fallen prey to a high-power microwave weapon would take little solace from the knowledge that a nonlethal weapon had sealed their doom. Some of these weapons (blinding lasers for example) may not kill, but have exceedingly nasty consequences for their victims. And in the end a disabling weapon works only if it leaves an opponent vulnerable to full-scale, deadly force.


The simple and brutal fact remains that force works by destroying and killing. In the Gulf War the commanding generals ostentatiously, indeed obsessively, abjured Vietnam-style body counts, but that did not diminish the importance of terrifying enemy soldiers through the fear of violent death from tons of ordnance raining down on them. And fear of violent death only comes from the imminent possibility of the real thing. True, in the Gulf War relatively small numbers of Iraqis (perhaps 2,300 civilians and up to 10,000 soldiers) died before the ground war, although others suffered indirectly from the combined effects of air attack and the coalition embargo. That so few died reflects, among other things, the potential of the new technologies and the scrupulous regard for civilian life shown by coalition planners. But the essential ingredient of fear remained constant.

Sometimes fear does not suffice. The objectives of conflicts such as the war with Iraq will frequently mandate killing. The destruction of some 50 percent of the Republican Guard's armor (in roughly equal proportions by air and ground action) made little difference outside the Kuwait theater. The Republican Guard remained at war's end an organized force and, after drawing on ample stocks of weapons in Iraq proper, put down the Kurdish and Shi'ite uprisings. To stop that and to undermine Saddam's regime (which the Bush administration certainly wished to do) would have required killing or wounding the men who constituted the bulwark of the regime. This uncomfortable fact, long known to the Israelis, who have had few scruples about killing German rocket scientists in the past or rogue super-gun designers in more recent years, sits poorly with Americans. When General Michael Dugan, Chief of Staff of the United States Air Force, hinted to journalists in September 1990 that the most effective use of air power might consist of attacks on Saddam Hussein, his intimate associates and key members of the Iraqi general staff and Ba'ath Party, he only pointed to the truth, impolitic as his outraged superiors found it.

It appears likely that civilian populations or large portions of them will continue to be the objects of terror. General William Tecumseh Sherman described the grim purpose of his 1864 march through Georgia and South Carolina thus: "My aim then was, to whip the rebels, to humble their pride, to follow them to their inmost recesses, and make them fear and dread us. 'Fear of the Lord is the beginning of wisdom.'" Sherman's troops did not massacre the inhabitants of the South: they merely ruined their private and public possessions, attacking (as a contemporary strategic analyst might antiseptically observe) the region's "economic infrastructure." In many cases today, war means bringing power, particularly air power, to bear against civil society. Those who hope for too much from air power desire to return to a mode of warfare reminiscent of the mid-eighteenth century in western Europe--war waged by mercenary armies isolated from society; war with (by modern standards, at any rate) remarkable efforts to insulate civilians from its effects. Sherman, reflecting the character of armed struggle in his century as well as ours, believed that in modern conditions civil society must inevitably become a target.

As leaders attempt to use their civilians as hostages against American air power this will become ever more true, whether we like it or not. Moreover, throughout the nineteenth and twentieth centuries military power has become increasingly intertwined with civil society. The electric generators that keep a defense ministry's computers running and its radars sweeping the skies also provide energy for hospitals and water purification plants. The bridges indispensable to the movement of military forces support the traffic in food, medicine and all other elements of modern life for large civilian populations. Sherman faced a similar situation when he besieged Atlanta. "You cannot qualify war in harsher terms than I will," he told the hapless leading citizens of that city. "War is cruelty, and you cannot refine it."


American air power dominated the Persian Gulf War as no other conflict since World War II. Special circumstances helped account for this achievement, but in the end airmen were probably correct in their belief that this war marked a departure. No other nation on earth has comparable power, nor will any country accumulate anything like it, or even the means to neutralize it, for at least a decade and probably much longer. American air power has a mystique that it is in the American interest to retain. When presidents use it, they should either hurl it with devastating lethality against a few targets (say, a full-scale meeting of an enemy war cabinet or senior-level military staff) or extensively enough to cause sharp and lasting pain to a military and a society. But both uses of force pose problems. The first type represents, in effect, the use of air power for assassination, a procedure not without precedent (American pilots stalked and slew Japan's Admiral Isoroku Yamomoto in 1943.) But it sets troubling precedents and invites primitive but nonetheless effective forms of revenge. The second involves the use of air power in ways bound to offend many, no matter what pains commanders take to avoid the direct loss of human life. To strike hard, if indirectly, at societies by smashing communication or power networks will invite the kind of wrenching television attention that modern journalists excel at providing.

Still, to use air power in penny packets is to disregard the importance of a menacing and even mysterious military reputation. "The reputation of power is power," Hobbes wrote, and that applies to military power as well as to other kinds. The sprinkling of air strikes over an enemy will harden him without hurting him and deprive the United States of an intangible strategic asset. American leaders at the end of this century indeed have been vouchsafed with a military instrument of a potency rarely known in the history of war. But glib talk of revolutionary change obscures the organizational impediments to truly radical change in the conduct of war and, worse, its inherent messiness and brutality. In the end, students of air power will serve the country well by putting the Persian Gulf War in a larger context, one in which the gloomy wisdom of Sherman tempers the brisk enthusiasm of those who see air power as a shining sword, effortlessly wielded, that can create and preserve a just and peaceful world order.


1 Department of Defense, The Conduct of the Persian Gulf War, Washington: Government Printing Office, 1992, p. 164; William Perry, "Desert Storm and Deterrence," Foreign Affairs, Fall 1991, p. 66.

2 See his columns in The New York Times, February 8, 1991; June 14, August 3 and December 7, 1992; and April 12, 1993.

3 See Thomas A. Keaney and Eliot A. Cohen, Gulf War Air Power Summary Report, Washington: Government Printing Office, 1993.

4 Such a possibility seems implicit in Christopher Bowie et. al., The New Calculus: Analyzing Air Power's Changing Role in Joint Theater Campaigns, Santa Monica: RAND, 1993.

6 Henry Petroski, To Engineer is Human: The Role of Failure in Successful Design, New York: Vintage, 1992, p. 217.

You are reading a free article.

Subscribe to Foreign Affairs to get unlimited access.

  • Paywall-free reading of new articles and a century of archives
  • Unlock access to iOS/Android apps to save editions for offline reading
  • Six issues a year in print, online, and audio editions
Subscribe Now
  • Eliot A. Cohen is Professor of Strategic Studies at the Paul H. Nitze School of Advanced International Studies, The Johns Hopkins University. From 1991 to 1993 he was Director of the Gulf War Air Power Survey, an independent study commissioned by the Secretary of the Air Force. This article represents the author's opinion solely, and not the views of the U.S. Air Force or any other government agency.
  • More By Eliot A. Cohen