I Hitler's squalid and ignominious death brings to an end this survey of the transformation of command across 2,000 years of Western history. Can we draw from it any general reflections on the nature of military power, the means by which it is exercised and the process by which its effects are invested with political value?
 

It is of overriding importance to recognize that military achievement is not an end in itself. Primitives may fight in blissful unconsciousness of performing any larger function than masculine self-expression. The professional warriors of advanced states may deny that they are anything more than simple soldiers doing their duty as they see it and dying when duty demands. Even their leaders may decry political purpose in strategy, claiming to be moved by military imperatives that stand at the furthest extreme from the dictates of diplomacy or the statesman's perception of national interest. 'A la guerre comme à la guerre,' say soldiers; by which they mean that war changes how warrior* look at the world, altering their priorities and submerging the preoccupations that animate peaceful society. Those - the profit motive, respect for property rights, obedience to the law, propitiation of the great, conciliation of minorities, performance of ritual and observance of custom and common courtesy - have no place, or only the very smallest, on the battlefield. There the race is indeed to the swift and devil take the hindmost. But, remote though the battlefield is from the marketplace and the court of law, its pre-existence, or the potentiality of recourse to it, underlie all assumptions citizens make about the order of things as they find them. Force, blind themselves to its sanction as the right-thinking may, provides the ultimate constraint by which all settled societies protect themselves against the enemies of order within and without; those with the knowledge and will to use it must necessarily stand close to or at the very centre of any society's power structure; contrarily, power-holders who lack such will or knowledge will find themselves driven from it.
 

There can, however, be nothing mechanistic about the exercise of power through force, whether naked or implicit long through the power-holding and power-hungry have sought such a secret. Force finds out those who lack the virtue to wield it. Such virtue in theocratic societies is deemed to descend from Cod or the gods and rulers by divine right may in consequence dispatch their subjects to the battlefield without thought or imputation of need to lead them there. Secular rulers enjoy no such moral exemption; in their worlds the virtues that attach to force are those by which it is resisted - resilience tenacity hardihood but above all, courage. They must therefore either go in person or else find the means of delegating the obligation without thereby invalidating their right to exercise authority outside the battlefield and in times of peace
 

In the preceding pages we have surveyed the practices by which four different societies dealt with the dilemma of command. By the heroic ethos of the Alexandrian world - an ethos that was widely to persist or later re-emerge elsewhere -command was simply subsumed within the art of government itself if the latter were not indeed subordinated to the former. No more distinction was made in Alexander's Macedonia between his roles as king and war-leader than between those of his leading subjects as electors and warriors. The legitimacy of all their roles was established and sustained by readiness to go to the battlefield and fight with courage once there; Alexander's function differed from that of his followers only in that he was expected to lead them to victory.
 

Not even defeat, if paid for by a kingly death could rob such a ruler of a hero's title. A heroic death, indeed, both glorified the victim and best legitimized his blood-heir's succession to his title. But it was in that cyclic rededication of the warrior ruler to legitimization by battle that the sterility of the heroic society lay. No development from it - political cultural, intellectual or economic - was possible as long as its elite's preoccupations were consumed by the repetitive and ultimately narcissistic activity of combat. All societies which achieved escape from the constrictions of heroism did so by separating the hero from the rest of society and accordingequal or superior prestige to functions more creative than his -those of the judge scholar diplomat politician and merchant.
 

Two routes which seemed to promise such escape proved in the long run to be dead ends: the mercenary and slave soldier systems By the second favoured in early Islam the function of warriordom was delegated to men who were the ruler's property. The logic of force however acting as it might have been expected to do worked over time to reverse the property relationship transforming those who exercised force into the possessors of power by feet if not title. The Mameluke kingdom that resulted was heroic by every test of that ethos proving to be incapable of civic development and so rooted in their traditional style of warmaking as to fail even militarily when confronted by the armies of societies which had achieved adaptation. The mercenary system on the other hand revealed its undesirability by contrary effect. Only those societies which had achieved a considerable degree of economic Development could afford to hire rather than breed soldiers; it was their own wealth which made the devolution of military duty so attractive to them and conversely to those who agreed to perform it on a commercial basis. The logic of force then working however to persuade the mercenary that he might take all that was available rather than the share he was offered states which had opted for hired defence tended to discover that they had sold their birthright. The social outcome proved to be either a reversion to heroic leadership or when enthroned mercenaries became softened by wealth and ease a resort to mercenarism all over again.
 

Successful escape from heroism was therefore to be by one of two other routes. The first epitomized by the society of which Wellington was a paragon lay through the creation of a military class compensated for its isolation from political power by an apparatus of established rewards anti privileges. Such classes emerged in few societies and at rare periods in history and the process by which they did so remains deeply mysterious. The Roman empire's class of professional soldiers is one example of the phenomenon; its evolution continues to engage the dedicated interest of historians of antiquity. Western Europe's regular armies are another. They are indeed an historical phenomenon in their own right hut the stages by which they detached themselves from the muddle of feudal levies royal retainers and hired freebooters who had served the mediaeval kingdoms are still shrouded in obscurity. All that can be said with confidence is that by the eighteenth century they existed in a finished form and that, by their liberation of their rulers' other subjects from the performance of military duty, they had released the energies of the rest of society for the tasks of creation - commercial, industrial, intellectual and artistic -which were to make Europe the master of the known world, and the conqueror of the globe's hidden parts, in their own time.
 

But even a professional military class, however stern the selfrestraint by which it lives, must in the last resort act to confine the scope for development of the society it serves. Military culture, central though it is and must be to the heroic ideal, can successfully adapt to the progressive separation of sovereign power from the person of the sovereign, even when the principle of the sovereign as hero on which it turns may have become a fiction. What it cannot accommodate is the formal transfer to the fact of sovereignty from ruler to ruled, that necessary process by which absolute states become democracies. Soldiers who have gone to the battlefield as the sovereign's surrogates and risked their lives in the name of the king instinctively recoil from the demand that they shed blood in the name of 'the people', a figment which can never be brought to represent the hero in any form. All peoples who have attempted an, rapid transition from monarchical to representational rule have, in consequence, encountered military opposition, the manifestation of which is called revolution.
 

By extraordinary ideological determination, as in the United States at its founding, or by subtle gradualism, as in nineteenth century Britain, a few have nevertheless succeeded in creating democratic constitutions to which soldiers could give their professional obedience. But the achievement of peaceful revolution does not dissolve the requirement for heroic leadership when a popular state calls on its people to die in battle. Then the eternal questions voice themselves again: 'Where is our Ieader? Is he to be seen? What does he say to us? Does he share our risks?' And the same questions in different form confront the leader himself: In front always, sometimes or never? is a dilemma that the elected statesman can ultimately no better escape than the heroic leader himself.
 

An elected leader who sticks to the rule 'never', perfectly proper though his decision may be by constitutional and practical judgement, will pay a terrible price if he inflicts on his people burdens heavier than they can or will bear: the disappearance of the French government of 1940 into one of history's oubliettes is a warning to that effect; the political extinction of President Lyndon Johnson at the height of the Vietnam war may be another. 'Never' may in the last resort stand even an unelected ruler with absolute power of repression at his disposal in no better stead. Hitler's suicide may be perceived as the due he had to pay the German people for leading them to defeat in 1945 and his foreknowledge of its inevitability appears in retrospect as a spectre with which he had long lived. The halfway house of 'sometimes' or 'I have shared such risks in my time' may not answer well either. Napoleon III's presence at the battle of Sedan could not rescue him from obloquy; Jefferson Davis who had been severely wounded in the Mexican War lost all hope of heroic epitaph when he cravenly fled from Richmond in 1865 at the appearance of Grant's army.
 

All such men of power may be judged to have met the fates they did and to deserve the reputations they enjoy from simple failure to understand the demands levied on them by the imperatives of command. Government is complex; its practice requires an endless and subtle manipulation of the skills of inducement persuasion
 

coercion compromise threat and bluff. Command by contrast is ultimately quite straightforward; its exercise turns on the recognition that those who are asked to die must not be left to feel that they die alone. But the relief of the warrior's ultimate loneliness is achieved by means quite as complex as those that attach to government. The successful leader - given that he is not doomed to fight an unwinnable war - is the person (women can lead as well as if not better than men) who has perceived command's imperatives and knows how to serve them. Those imperatives are few - but not all will necessarily yield to discovery even under assault by a mind as possessed by the urge to power as that of Hitler himself. How are they to be enumerated?

Command, the cliche has it, is a lonely task. But so it must be. Orders derive much of their force from the aura of mystery more or less strong with which the successful commander more or less deliberately, surrounds himself; the purpose of such mystification is to heighten the uncertainty which ought to attach to the consequences of disobeying him. The taskmaster who eschews mystification, who makes himself, his behaviour and his responses familiar to his subordinates, must then evoke compliance either by love or by fear. But love and fear, strong though the role of each hi in the masculine world of war, are emotions ultimately self-limiting in effect. True love is felt by two parties; it can rarely be simulated by either over the lifetime of a relationship. 'The commander who shows the love he feels when he gives orders must eventually cripple his will to expose his loved ones to danger. Fear, on the other hand, operates only if it is felt more keenly than the fear that it opposes. In the short-term, it can drive men to self-sacrifice ('Dogs,' Frederick the Great demanded of his grenadiers, 'would you live for ever?'). In the long-term it loses its power to compel by reciprocal mechanistic effect. Caught between two fears, the subordinate will eventually seek escape from both.
 

Mystification supplies the medium through which love and fear, neither ever precisely defined, cajole the subordinate to follow, often' to anticipate, the commander's will. But mystification is a function of distance, real or illusory, which the commander must impose or contrive. Hitler and the chateau genera's, on whose command style he modelled his own, created mystification by imposing distance, of fifty or so miles in their case, hundreds in his, between themselves and their subordinates. Alexander contrived a sense of distance by living within his aura of kingship, reinforced as it was by the priesthood whose offices he alone, as the Macedonians' sovereign, could perform. Wellington and Grant, in the very different societies to which they belonged contrived distance in appropriate ways: Wellington scion of a society dominated by gentlemen, created and maintained a gentleman's household of servants, hounds, horses and hunting companion wherever the vagaries of campaign took him, living a country - houslife in the heats of India or the snows of the Sierras; Crant, a small-town American, took the companionship of his own small town into the field, delimiting the distance necessary to his emotional comfort by setting a barrier of Main Street cronies between himself and the larger world of the army outside it.
 

Distance is, nevertheless, a negative dimension. The man who insists on it becomes a recluse and the reclusive commander achieves nothing. Distance must be penetrable by access either inward outward or both. Hitler allowed occasional inward access: Guderian for example had the self-confidence to insist on personal confrontations with the Fuhrer at Rastenburg when he felt that strategic crisis required it. Alexander thrived on outward access: he constantly moved among his subordinates showing himself lo his Macedonian subjects dramatizing his kingship and playing the hero to the ever ready audience his army provided. Wellington and Grant by contrast freely encouraged access both inward and outward. They were often seen by their subordinates in the field as they moved among them in an environment of shared danger- all too closely shared by Wellington they were also easy hosts Granl even more so than Wellington receiving guests from the body of the army in the small society of their headquarters making visitors feel at home and Ictting them go with the sense of having shared the vital intimacy of the commander chez sad.
 

The most important medium of penetrability however was supplied not by personal access but by the diaphragm of intimates and associates which surrounded the commander. Their selection and quality was crucial to the relationship that the general established with those to whom his orders were transmitted. Hitler isolated by real distance from his fighting and suffering armies needed his aloofness to be mediated by men with whom the common soldier could identify warriors who had also starved thirsted shivered sweated and bled with the man in the front line unlike the chateau generals of 1914-18. He signally failed to surround himself with anyone of that sort. Keitel his principal subordinate wobbled with the pounds of easy living and mindless sycophancy; Jodl his brainbox was marked by the stresses of the map table not the foxhole; Schmundt his chief army adjutant and so its principal representative at Fuhrcr headquarters babbled to his old comrades-in-arms when they met of the spell Hitler had cast over him, never of his chief's concern for the welfare or preoccupations of the men under his command. As a result it was only by the genius of Goebbels's propaganda efforts in representing the Fuhrer to the Wehrmacht as a front-fighter with the best of them that the force of his orders was sustained to the end.
 

Alexander's army was suffused by his personality from the outset of his anabasis to his death; the role of his intimates, who became the Diadochi, in interpreting and transmitting the nature of that personality is undeniable. But the limitation of his relationship with them is defined by their subsequent behaviour. The Diadochi were as much competitors in heroism with Alexander as mediators, and the posthumous fragmentation of his empire was the result of their desire to equal his achievement rather than propagate it. His essentially unstable system was held in equilibrium only by his day-to-day efforts; when his death disturbed the balance, both army and empire fell apart.
 

Wellington and Grant, representatives rather than embodiments of a system, used their circle of intimates to much more fruitful effect. Their intimates fulfilled the role on the one hand of remembrancers to the commander of his responsibility for the army's welfare, and on the other of witnesses to the army of the commander's concern for it. The extent of their success is borne out by the excellence of relations pertaining between headquarters and troops throughout all their campaigns, a success in the last resort attributable to the commanders' skill in selecting men who provided windows to both worlds.
 

Grant and Wellington both succeeded, in short, in creating a bond of kinship between themselves and their followers by surrounding themselves with men who posed no threat to their primacy yet were of sufficiently soldierly quality to command the army's respect. Alexander, on the other hand, was fated to be surrounded by men who, while their soldierly qualities were not in doubt, so powerful!! shared his ethic of heroic individuality that he could never truly rest at ease with them. Hitler went to the other extreme: his intimate circle was selected by the test of sycophancy, which made for perfect domestic ease at headquarters but denied him any bond of understanding with the fighting men at the front.

Understanding between commander and followers is not assured solely by the mechanisms of kinship. A commander must not only show what he feels for his soldiers by the quality of their representatives he chooses to keep at his side. He must also know how to speak directly to his men, raising their spirits in times of trouble, inspiring them at moments of crisis and thanking them in victory. The more directly heroic the nature of his leadership- and therefore in all likelihood the more extreme the predicament to which he exposes them - the stronger that imperative. Wellington and Grant, leaders of constitutional armies in inter-state wars, were bound comparatively lightly by that imperative and both were notably poor communicators. Hitler on the other hand - a demagogue fighting a demagogue's war - though he rarely spoke directly either to army or people during its course, controlled a propaganda machine of the highest sophistication and was acutely sensitive to its operation. And Alexander was, of course, a master orator, a brilliant stage-manager of his own speaking performances and a supreme psychologist in his choice of rhetorical devices - challenges, threats, cajolery, bribes, appeals to pride, evocations of past achievements, promises for the future. The means by which he brought the force of his personality and intellect to bear on his army remain obscure; no human voice, without artificial amplification, has the power to reach the whole of an army as large as he commanded. In consequence, he sometimes spoke only to his officers, and at others repeated his speech, or variations on it, to fractions of his army in turns. But it is quite possible that he occasionally paraded it in a natural amphitheatre where echo would make him heard simultaneously by all.
 

Whatever the means he employed to make himself understood, Alexander had grasped from the outset the imperative of prescription - the need of every commander to convey an impression of himself to his troops through words, to explain what he wants of them, to allay their fears, to arouse their hopes, and to bind their ambitions to his own. It is a mark of the depths to which the art of command fell in the era of chateau generalship that this need was served barely, if at all, by any of the generals of the First World War. Their armies were, by an ironic twist of social and constitutional development, the most literate and politically conscious mass forces ever to have taken the field. By an equally ironic twist, the Staff College culture which informed their leadership had, by a bogus scientism, so sanctified the importance of purely theoretical principles of warmaking, and consequently so depreciated the importance of human emotion, that the common soldiers were not thought worth the expenditure of their commanders' breath.
 

The lesson of that fatal misjudgement was to be widely drawn by the generals of the Second World War, many of whom were to become as adept at self-presentation and prescription as Alexander himself. Hitler may have scarcely ever been photographed among his soldiers; photographs abound of his subordinates -Guderian, Rundstedt, Dietl, Model, Student - among theirs. The dislike felt for Montgomery by his more blinkered contemporaries was largely provoked by his remarkable theatrical gifts, much appreciated by his audiences of ordinary soldiers. And the art of self-preservation became in the post-war years a positive cult in two armies committed to struggle against the odds, the lsraeli and the French. 'When I give difficult orders,' an Israeli general is remembered as saying, 'I like to do so in person, so that I can meet my soldiers' eyes.' 'Whatever else you may say about me,' General de Lattre de Tassigay assured the young officers of the army he was rescuing from the Indo-China disasters of 1950-1, 'you will not be able to say that you were not commanded.'
 

For all the importance of prescription, military literature is curiously deficient in discussion of how it should be done. What German classical scholars call the Felderrnrede - the general's speech before battle - was a well-known literary form in antiquity. In the modern world Raimondo Montecoccoli, the imperial general of the Thirty Years' War, is almost the only writer to have addressed the subject. His remarks are extraordinarily penetrating, many of them still closely relevant to the manipulation of soldierly emotion on the contemporary battlefield.
 

'Exhortation of the host' is how he describes the imperative of prescription, 'when the general speaks publicly to his soldiers in order to urge them to demonstrate virtu and to infuse them with courage.' He suggests four main ways by which those objects may be achieved.
 

The first is by 'arguments of use':
 

. . . captains can incite soldiers to fight wars by indicating the necessity of battle, which deprives men of all hope of saving themselves except through victory and which forces them either to conquer or die. The same result may also be achieved by depicting the justice of one's cause, by appealing to patriotism and love of the captain, and by evoking disdain for the enemy; by showing that the enemy is saying ignominious things about one's own troops; that he wants to take away their property, religion, liberty and lives; and that it is better to die generously than to languish under tyranny.

'Exploiting the fear of infamy' is the second:
 

... make soldiers see that they are in the presence of illustrious persons. In order that they may abhor cowardice and exalt valor and so that they will have witnesses to their actions, they should fight under the watching eyes of the general or the prince ... In order that the men will be prepared for the fray in a manner they can comprehend easily, the commander will declare that it is not the army of the fatherland but the fatherland itself that is endangered because it will have nothing left if the army is beaten.
 

'Exciting the desire for riches and prestige' is the third: 'It is also possible to make soldiers resolute by raising the hope of great rewards and prizes if they succeed, whereas they must be brought to dread severe punishment if they fail.' But it is Montecuccoli's fourth method which has the most convincing ring to modern ears, 'Developing confidence'. Let the captain, he says, show that
 

he himself is lighthearted and full of hope by means of his facial expression, his words and his dress. His visage should be severe, his eyes intrepid and luminous, and his clothing flamboyant. He should banter with his men, be clever and witty. They will then deduce that their general could not jest and enjoy himself like that if there were any real danger, if he did not think that he was much stronger or if he did not have good reason to scorn the enemy. The troops are bound to take confidence.
 

'The first quality of an officer,' wrote the future Marshal Lyautey in 1894, 'is gaiety,' independently echoing the point that Montecoccoli makes. Among the imperatives of command, that of speaking with all the arts of the actor and orator to the soldiers under his orders stands with the first.

It is self-deluding to expect, however, that men can be led to fight solely by encouragement, flattery or inspiration. Words supply an uncertain antidote to fear. Fear must be opposed by fear itself or by a material factor as strong or stronger, and the commander who shrinks from threatening his troops with punishment or who will not deign to bribe or reward them will make easy meat.
 

Grant, among our four commanders, had least recourse to either unction, the result of his access to very large reserves of manpower, from which the depredation of desertion could easily be made good, and also of his sensitivity to the populist ethos of contemporary America.
 

Outrage - rape or pillage - aroused his ire, as did treason or selfish profiteering, and he would punish peremptorily in such cases. But he did not regularly hang or imprison for cowardice or disobedience, because his citizen armies themselves tolerated such divergences from good military practice, recognizing them to be inseparable from their amateurism. For the same reason neither he nor ha soldiers placed any high value on decoration or exceptional payments; service freely undertaken for a cause (the North did not conscript until 1863) was held in itself to be a badge of honour, to which others were superfluous, if indeed not odious.
 

Wellington, on the other hand, commanding men brought into the army by want and serving in it without sense of public duty, punished ferociously and conceded reward, in the form of loot, as a necessity. His philosophy of sanction had been that of European armies since time immemorial and differed from that of mediaeval hosts or mercenary companies only by the stricter regularity with which it was enforced by military law and standing orders. In the aftermath of his wars, however, when military service was established throughout Europe on a footing of social obligation ether than hired enlistment, the basis on which both reward and punishment were administered was consonantly transformed. Punishment lost such barbaric features as flogging (a voter could scarcely be triced up at the triangles), though it retained the ultimate sanction of death for cowardice, desertion or mutiny. Reward, on the other hand, was enormously elaborated.
 

Napoleon, the first leader to command something approximating to a citizen army, had early grasped that the dignity of the citizen soldier required that he be rewarded for exceptional conduct not by the arbitrary prize of loot (falling though it naturally does to soldier foremost in the fight or breach) but by tokens of society's esteem. The Legion d'Honneur, instituted in 1802, was the first decoration for bravery to be created in any army for which all soldiers, irrespective of rank, were eligible. In a sense, it demonetarized reward in the field, and with such success that by the middle of the nineteenth century all Western armies had followed the French wit. The British Victoria Cross, the Prussian Iron Cross, the Ruesian Order of St George, the American Medal of Honor were all modelled on the Legion; their institution was followed by the creation of additional medals for lesser acts of bravery or devotion, so that by 1915, for example, a British general had at least six grades d decoration for which he could recommend soldiers under his command.
 

Decoration is a particularly potent tool in the management of a commander's direct subordinates, his staff officers and generals. Alexander had rewarded loyalty and success by marks of personal favour. Wellington and Grant, controlling armies formally structured by rank, arranged for their better subordinates to be promoted; a great deal of Grant's correspondence with Washington was devoted to that matter. Hitler, having the apparatus of both rank and decoration at his disposal, freely distributed promotion and rewards among his successful generals. Cunningly, and by a reversion to the conquering style of old, he also made so-called 'donations' to the favoured few, grants of land or money given privately and secretly to the very senior. It was a deliberately calculated means of compromising the integrity of the Ceneralitat, sowing disunity and disarming opposition.
 

Yet, until his outright breach with the army after July 1944, Hitler was curiously lenient with the unsuccessful, even with the contrary. Like any strong-minded generalissimo before him -Joffre, for example, in 1914- he dismissed on a large scale if combat efficiency required it; the mass purge of December 1941 showed how ruthless he could be if he chose. Yet, despite causing the Reichstag to accord him, in April 1942, absolute powers, he used such powers sparingly. Hoepner was deprived of his pension for his mishandling of his panzer group in 1942, von Wietersheim reduced to the ranks for incompetence, von Sponeck sentenced to be shot for abandoning the Kerch peninsula (the sentence was later commuted to imprisonment) and Falay and Stumma both dismissed outright as a result of breaches of documentary security in their commands. Until the Bomb Plot, however, Hitler's personnel policies were substantially lo harsher than those of Churchill's, and a good deal less draconian than Stalin's, who, having murdered half the senior officers of the IW Army in 1938, had no compunction about executing unsuccessful generals in the crisis of 1941; several anticipated their fate by committing suicide.
 

Speer, a civilian observer of proceedings at Hitler's headquarters, was indeed surprised by the apparent lack of awe in which the professional soldiers held their supreme commander. 'I had expected respectful silence during the situation conferences,' he wrote, and was therefore surprised that the officers who did not happen to be participating in a report talked together freely, though in low voices. Frequently the officers, showing no further consideration for Hitler's presence, would take seats in the group of chairs at the back of the room. The many marginal conversations created a constant murmur that would have made me nervous. But it disturbed Hitler only when the side conversations grew too excited or loud.
 
 

The treachery of the traditional military class in July 1944 put an end to the easy ways for good. Mistrust came to pervade all intercourse between Hitler and his generals and, as the tide of defeat engulfed the Reich, suffused the army at large. During the retreat from France, Hitler threatened Sippenhaft - punishment of family -against commanders who surrendered fortified places. And in the last days of the war all ordered discipline was thrown to the winds; 'flying' courts-martial summarily executed soldiers suspected of seeking to surrender and even those found separated from their units.
 

These were measures of desperation and, given the inevitability of impending defeat, anyhow quite fruitless. But the nakedness of the expedient nevertheless exposes in a peculiarly stark form the necessary ambiguity of the relationship by which leader and followers arc bound. Coercion is as essential a component of command as prescription or kinship. Ideally it should remain implicit, and when made explicit should manifest itself as rarely as possible as physical force, except in extreme emergency never falling arbitrarily or threatening the majority. Once a commander becomes as much an enemy to his followers as the enemy himself -and what else is a commander who breathes fire and sword against his own men? - the mystification of his role is destroyed and his power, essentially an artificial construct, dissipated beyond hope of recall.

Kinship, prescription, sanctions are all preconditions of command. They do not amount to command itself. There are, indeed, times when a commander must watch and wait, and then it will be by prescription and sanction that his authority is sustained. But in the last resort a commander must act. How should he do so?
 

Action without forethought or foreknowledge is foolhardy. Commanders must know a great deal before they act and see what they are about when they do. These prerequisites are defined in the military vocabulary as intelligence and control and form two of the major elements of what analysts of strategic affairs have recently come to call C{3}1; Command, Control, Communication and Intelligence. New definitions, however, do not change old realities. The essentials of action by the commander are knowing and seeing.
 

All four commanders whose methods we have surveyed grasped the central importance of knowing, both in general and in particular. Alexander's youthful obsession with the human geography of the Greek and Persian worlds - Who lived where? What did they grow? How did one travel from here to there? - was to be matched by Wellington's appetite for topographies and Grant's fascination with maps; even Hitler, indiscriminate as he was in choice of reading, the wordy frothings of racialist philosophers and the simple story-line of cowboy writers having an equal capacity to entertain him, tool' trouble to supply himself with exact military knowledge, if of a strictly limited usefulness. He certainly knew a great deal about the equipment of his armies and believed he knew all that was essential about soldiering; but he had an ignorance of climatic and terrain difficulties in the east, where he had never served, which was to prove fatal. Alexander, Wellington and Grant, on the other hand, knew their armies inside out, their theatres of campaign, and also a great deal about their enemies. Grant, of course, was privileged by special access to his opponents' minds; he had served with many of them, if he had not indeed known them as fellow cadets at West Point. Alexander's and Wellington's intimacy with the enemy was less complete. Both, however, understood a good deal about the forces they opposed, Alexander because the backbone of the Persian army was Greek, Wellington because he had been educated in France.
 

General knowledge is ultimately limited in its usefulness, however, precisely by its generality. Particular knowledge - of the enemy's whereabouts, strength, state, capabilities and intentions - is by contract the material on which effective command thrives. Its value is recognized by the simplest minded. The difficulty is to acquire it and, once acquired, to put it to use. Martin van Crefeld, in his study of staff systems, advances a reflection in this regard of the most acute insight: that in pre-industrial society, particular knowledge was generated in quantities small enough to be handled by an individual, but reached him at a speed not much faster than armies moved and so tended to be out of date when received - and was not therefore 'real time' intelligence, as communication experts now characterize the commodity; but once industrial technologies -of which the telegraph was the first - allowed intelligence to outpace the movement of armies, its volume at once increased to exceed the capacity of any one man to collect and digest it. The rise of general staffs - essentially collections of subordinates expert enough to process particular knowledge on the commander's behalf - almost exactly coincides with the appearance of the telegraph, thus bearing out the point that van Crefeld makes. But, as he goes on to emphasize, the delegation of information-processing to subordinates imposes a remove between the commander and his besetting realities, beyond those that already exist.
 

Chateau generalship- in some sense, an acceptance of the logic of circumstance - was one reaction to this development. But superior generals, of whom both Wellington and Grant were types, had always resisted the logic of circumstances, had been keenly alert to the danger of distancing themselves from reality that even the comparatively primitive technologies and staff systems with which they worked threatened. The antidote that they applied was an insistence on seeing. Grant, making allowance for the recently and very greatly heightened danger of moving exposed within the missile-zone on his battlefields, managed to see a great deal. Wellington, who gambled recklessly with the lesser but still acute dangers of the missile-zone in his time, saw as much as was possible for any individual horseman. Both acquired crucial 'real time' intelligence in large quantities, processed it instantly, gave necessary orders immediately and were able to monitor the effects almost as they watched.
 

Alexander, because of his direct involvement in hand-to-hand fighting, an inevitability of the heroic ethic, had been able to do no such thing. Nor, paradoxically, could Hitler. He, deluded by the apparent instantaneities of the radio, telex and telephone (though he disliked the latter instrument, which minimized his magnetism), believed that he saw with the immediacy of the men on the spot.. He was, however, wrong, and the workings of Fuhrer headquarters were afflicted by all that was and is worst about both the chateau generalship of his own youth and the elaborately mechanized and automated command centres of our day. Floods of information, collected and transmitted apparently in 'real time', arrived at his situation conferences with significant delay; precise and detailed orders, seemingly attuned to realities, returned from him to the point of action only after realities had moved on. The disjunction between intention and effect resolved itself in the undignified and impotent tirades to which the Fuhrer subjected his subordinates, both in headquarters and at the front, when events were revealed to have escaped his direction.
 

The problem of 'real time' intelligence probably defies solution. Armies are, in a sense , mechanisms designed to allow the will of an individual to bear directly on outcomes; that purpose is the justification for the hierarchy and discipline by which they are articulated. If the long experience of war demonstrates any one thing, however, it is that those moments when the scope of action and the size of armies lie in optimum relationship to each otherthose moments, that is, when the flow of information upwards and orders downwards will most nearly match the pace of events - are very, very few. The masters of gunpowder warfare, among whom Frederick the Great and Wellington were outstanding, operated at such optima; because the tactics and strategy then prevailing obeyed rules of almost mathematical constancy, the clever commander could use whatever privileged information came his way to predict, anticipate and influence outcomes with uncanny certainty. At almost all other times before or since, however, such disequilibrium has normally prevailed between the size of armies and the scope of action that outcomes have yielded no certainties at all. Armies have either been too small for a commander seized with a vision of outcome to achieve it; or too large for any commander, however elaborate his information-gathering means, to grasp where the opportunity for outcome lay. Strategic indecision - by far the most common end of all campaigns - results in the first case; painful and bloody attrition,
 

the all too frequent product of modern warmaking, in the second. The insolubility of the 'real time' intelligence dilemma accepted -the dilemma is as great today, allowance for relative velocities of force being made, 9a it ever was and far more crucial in importance - the actual issue of command may now be seen to confront us. In front _ always,, sometimes, never? is, I have suggested, the question which must lie at the heart of any commander's examination of conscience. Those, line Alexander, to whom 'always' was the instinctive response, solved the 'real time' intelligence dilemma by dismissing it; their response to the challenge of events was to determine outcomes by direct, personal intervention. Those, like Hitler, the chateau generals and the denizens of contemporary situation rooms, who choose to say 'never', do so because of their belief that the dilemma is solved by artificial vision - that supplied by telegraphic, telephonic and, today, televisual communication; their response to the challenge of events was and is to demand more information and to issue stronger orders. It is the third group, formed of those giving the answer 'sometimes', whose response to the dilemma is most fruitful. Wellington and Grant - but also Caesar among their predecessors, Guderian and Montgomery among their successors -accepted that neither knowing nor seeing alone return an answer to the challenge of events. Sometimes a commander's proper place will be in his headquarters and at his map table, where calm and seclusion accord him the opportunity to reflect on the information that intelligence brings him, to ponder possibilities and to order a range of responses in his mind. Other times, when crisis presents itself, his place is at the front where he can see for himself, make direct and immediate judgements, watch them taking effect and reconsider his options as events change under his hand.
 

The proof of the pudding is in the eating. The 'sometimes' generals, among those we have considered, achieved a notably more consistent record of success than the 'always" or 'nevers'. Alexander, for all the dramatic immediacy of his style, put the future of his army at risk whenever he tool' the field, since its survival depended upon his own, and he trifled with his survival as a matter of honour. Hitler exposed his whole army to constant risk of disintegration, once the tide turned against it, simply because he refused to contemplate the reality of its predicament, to which he insisted his own was superordinate. Grant and Wellington, on the other hand, by walking the narrow path between extreme and false heroism, succeeded in constricting the ambient risks both to themselves and their armies and thereby in 'leading'- if from the rear - their soldiers to victory.
 

But Wellington and Grant did more than obey the imperative of action - of selecting and performing, that is, the correct function for themselves in the context that the military circumstances of their time dictated. They also succeeded in obeying the best and greatest of imperatives - which Alexander had obeyed to the unsafe exclusion of all others - that of conspicuous participation in the dangers that confront the lowliest soldier most keenly; in short - the imperative of example.

The first and greatest imperative of command is to be present in person. Those who impose risk must be seen to share it, and expect that their orders will be obeyed only as long as command's lesser imperatives require that they shall. Presence may with limited and temporary success be simulated - by frequent visits to the danger zone at moments of quiescence or (what has been said about Jefferson Davis notwithstanding) by the invocation of a reputation for risk-taking in times past. Neither, however, guarantees that the seeming or one-time hero will thereby stimulate heroism in those he wishes to imbue with it. Legendary warriors like Churchill's Carton de Wiart, one-armed, one-eyed, seven times wounded on separate Sundays, or Franco's Millan d'Astray, founder of the Spanish Foreign Legion and also lacking an eye and an arm, may impel young soldiers to reckless deeds by the incontestable evidence of their own past contempt for danger; but few who have shown such contempt survive to infect others with their spirit. Old warriors who have survived risk intact seem to the young merely old; and would-be heroes not heroic at all. It is the spectacle of heroism, or its immediate report, that fires the blood.
 

Hence the collapse of so many armies whose commanders neglected to show themselves to their soldiers at the moment of danger. 'A rational army,' said Montesquieu, 'would run away.' And so, if we accept that self-preservation is the ultimate expression of rationality, we must agree it would. The thought is one that ought never to be far from any commander's mind. For the merest twitch of emotion stands between his exaltation and his descent to ignominy. At one moment he may, from his horse or headquarters, survey ten thousand, even a million men, ranked to heed his orders. At the next they may be streaming to the rear, obeying no order but 'sauve qui peut'. The transformation might sound over-dramatized; very large armies are as slow to disintegrate as they are to concentrate, since panique-terreur, the psychological state that eighteenth-century generals strove to create in the collective nervous systems of their opponents, can initially infect only those fractions of armies exposed to the enemy's main offensive effort. The rest will catch the infection indirectly, feeding their fears on rumour and sensation rather than the reality of rout at close hand, perhaps in consequence failing to find room on the roads to the rear, fighting rearguard actions willy-nilly or floundering in indecision until forced to offer their surrender when abandoned, encircled or marooned.
 

The sensation of defeat is, nevertheless, unmistakable and often uncontrollable. Few large modern armies have run with the instantaneity of Darius's at Issus or Gaugamela; parts of the Polish army preserved their integrity throughout the awful days of retreat from the frontier to Warsaw in September 1939, and the French defenders of Lille sustained such resistance in 1940 that their German opponents rendered them the honours of war when they eventually marched out to captivity. But when the germ of defeat takes a hold, even very large armies can fall apart with epidemic rapidity. Such was the fate of the Italian army at Caporetto in November 1917, of the bulk of the French army of the North-East in May 1940, of the German Arrny Group Centre in June 1944. The resulting humiliation of their commanders was pitiable. Cadorna, Georges, Busch had all been paladins; the first a general whose unapproachability struck fear into his subordinates, the second an Olympian of the generation of Foch, the third a victor of the French and Russian Blitzkriegs. Overnight they dwindled into despised nonentities. Cadorna was hurried into obscurity, Georges left weeping at his map table, Busch consigned to the pool of rejects unemployable even in the backwaters of Hitler's empire.
 

None wholly deserved his kite. The disorders which engulfed their armies were defeats that were waiting to happen, and perhaps no general could have averted them. But Cadorna and Georges had contrived to command in a fashion that ensured professional extinction would follow failure as night the day. Both were 'chateau generals' of the most extreme type, and though 'chateau generalship' was an understandable reaction to the recent appearance of long-range weapons, its effect on the relationship between leaders and led was so deadening that even the most arrogantly insensitive of generals should have taken steps to ameliorate it. By the time of Busch's disgrace in 1944 the more perceptive had already begun to do so. Cadorna and Georges appear never to have thought of attempting or even simulating heroic leadership themselves. To that extent they suffered their deserts.
 

Yet in their youth generals had shared rid' with their soldiers as a matter of course, just as leaders had done for a hundred generations. Why the submergence of heroic leadership by chateau generalship, which was its antithesis? The answer is in part cultural and intellectual - and to this we shall return - but in greater measure technical. The trend of weapon development had for several centuries been acting to drive commanders away from the forward edge of the battlefield, but they had nevertheless resisted it. What occurred at the end of the nineteenth century was a sudden acceptance by the generals of all advanced armies that the trend could no longer be gainsaid and that they must abandon the post of honour to their followers.
 

The option of command from the rear had, nevertheless, always been open. Alexander had chosen not to exercise it because the values by which he lived and reigned forbade his incurring any taint of cowardice. Within 200 years of his death, however, his own society had advanced to a recognition that a general's station need not be fixed at the point of maximum danger, that he might indeed serve the cause of victory better from a place where he could observe and encourage rather than fire others by his example. But that recognition was not to extinguish the power of the heroic ethic altogether. On the contrary, what resulted was the marriage of the two, giving birth in turn to a code of compromise. By its dictates the general would seek to set as striking an example of risk-sharing as he could, consonant with the need to keep a distance from danger sufficient to allow his controlling the battle as a whole.
 

It was by those dictates that such commanders of professional armies as Caesar and Wellington adjusted their response to threat and crisis. Caesar, articulating a weapon system technically no different from Alexander's though superior to it by the index of drill and discipline, was often impelled to its frontier of contact with the enemy, and both dressed and behaved accordingly. He affected a distinctive red battle cloak and had ready prepared a repertoire of battlefield oratory with which to inspire and instruct his subordinates. The death of the legions with that of the Roman empire brought back the heroic style. But with the return of regular armies, of which Wellington's was the most perfected type, the compromise between prudence and exposure re-asserted itself. Wellington's close encounters with death were never haphazard, but the result of a mathematical calculation of the ebb and flow of danger. On the open battlefields where he and his opponents chose to give action, it was a consistent possibility, given the known ranges at which weapons took effect, to anticipate the fine tolerances when this position or that would become untenable by the commander and to move accordingly. Wellington did not represent his style of command in terms of the judgement of 'fight' and 'flight' distance by which a lion tamer exerts his mastery over his charges- and a spell over his audience; but it was calculated in almost exactly the same way. If one dimension of command is the theatrical, one would say that, while Alexander's performance was relentlessly Grand Guignol, Wellington's was brilliant melodrama, a succession of perfectly timed exits and entrances, each advancing the plot to its triumphant conclusion by spectacular and risk-fraught effect.
 

It was a performance, nevertheless, that literally diced with death, as his tally of minor wounds and disabled mounts testifies. Just forty years after his last appearance on the stage of battle, the pattern of risk-taking he had run would have swiftly exhausted an imitator's invulnerability. The tide of probability had then begun to run against anyone foolish enough to keep to the saddle within 500 yards of the firing line - he had survived long exposure at 100 yards or less - and wise generals reacted accordingly. Grant, as we have seen, was very wise. Confident in the power of other means to legitimize his authority, he unashamedly held himself rearward of all but the incalculable odds - stray shells, ambush - while sending his soldiers forward without compunction to face the danger he had decided it was not his duty to share.
 

Yet Grant did not think it proper to exempt himself from the environment of risk altogether. Though leaving the heroic role to his subordinates, he kept a place for himself on the stage of battle as a sort of actor-manager, prompting the principal players at need and intervening from the wings when crisis threatened the development of the action. The actor-manager role he created - few contemporaries learnt to function as he did - was to prove, however, a transient one, intermediate between Wellington's style, rooted as it was in the heroic tradition, and that of the chateau generals to come. Some commanders of the Prussian wars of 1866-71 would ride the battlefield as if none but a silver bullet could touch them. But the majority kept to or near their headquarters, communicating with the front by messenger and surveying it, if they could at all, by telescope. Fifty years later, their descendants- Frcnch and German indiscriminately - were not to think of quitting their headquarters at any time. Berthelot, Joffre's operation officer at the Marne in 1914, would indeed spend the whole of the battle literally en pantoufles -shod in carpet slippers - and sitting at his desk from which only the summons to a meal (he might have doubled as the fat man in a circus) could shift him. The hazards of the preceding Great Retreat had obliged him to set up his office in a succession of town halls and schoolhouses. With the stabilization of the front in October, however, he would be solidly established in chateau comfort at Chantilly and his opposite numbers in the allied and opposing armies likewise, the Germans having chosen Spa, a health resort in Belgium, and the British Montreuil, a charming little walled town close to the English Channel. It was from those secluded places that the great slaughter of the trenches would be directed, totally out of sight and, unless for a trick of the mind, also out of sound of all the headquarters responsible for it.
 

One of the inhabitants of British headquarters, Charteris, Haig's chief of intelligence, has left us a picture of life at Montreuil in 1916:
 

Here at GHQ, in our own little town away back from the front line trenches [delicately put; Montreuil was fifty miles behind the lines], there are few visible signs of war. We might almost be in England . . . All the work in all the departments is now systematised into a routine. Most of it is done in office. One of the great difficulties of everyone at GHQ is to get away from the office often and long enough to get in close touch with the front. Few can ever get much further forward than the HQ of Armies . . . Forward of Army Headquarters, one is nearer the fighting, but even they are now mostly in towns or villages several miles behind the front line. Further forward still are Corps Headquarters, where there is generally plenty of evidence of war . . . but even Corps Headquarters are now pretty big organisations and are almost always in a village. In front of the Corpe Headquarters the Divisions are mostly in farmhouses, but well in the fighting line. One can almost always get one's car up to them. But that is about the limit, and visits forward of them consequently take up a good deal of time. We all manage, anyhow, to see something of a division headquarters, but it is only when there is some particular object, more than simply looking around, that one can give up the time to go beyond them. I have not even seen a Brigade Headquarters in the front line for the last month.
 

Since brigades stood a rank higher in the chain of command than battalions, which actually occupied the trenches, it may be seen that Charteris, whose duty was to form a picture of events at the front for transmission to his chief, did so at best largely second-hand. Haig himself, though his biographer, John Terraine, claims for him that he visited the trenches frequently, was rarely observed to do so by memoirists of the front line. Even at Montreuil he preserved an Olympian detachment from the work of the staff; one of them recalls that, as a special concession, staff officers were allowed to leave their desks to watch him ride in and out from his office provided they did not show themselves at the windows. Haig's residence was not even in Montreuil; he preferred to seclude himself from its relative hurry-burly at the chateau of Beaurepaire some ten miles away in the heart of the countryside.
 

The simulated absolute monarchy of chateau generalship ultimately provoked the military equivalent of revolution in almost all the armies on which it was imposed. In May 1917, after the failure of some particularly heartless offensive plans, nearly half the divisions of the French army downed tools, announcing their unwillingness to attack the Germans again until their grievances were redressed. In October of that year the Russian army, disillusioned by the pointlessness of its sufferings, simply 'voted for peace with its feet', as Lenin put it, allowing him to transform the power vacuum which resulted into political revolution. In November the Italian army effectively gave up the fight to which Cadorna had relentlessly driven it, with consequences that almost brought Italy to defeat. It was a crisis of morale in the German army in September 1918 that prompted Ludendorff to tell the German government it must treat for peace. And even the British army, in the aftermath of the March retreat of 1918, suffered a collapse of morale so acute that Haig wee impelled to subordinate his independence of command to the French, as the only means of securing reinforcements to shore up his shaken front.
 

At the root of all these spiritual crises lay a psychological revolt by the fighting soldiers against the demands of unshared risk. For two or three or, in the case of the German army in September 1918, four years, orders had emanated from an unseen source that demanded heroism of ordinary men while itself displaying heroism in no whit whatsoever. Far from it: the chateau generals had led the lives of country gentlemen, riding well-groomed horses between well appointed offices and residences, keeping regular hours and eating regular meals, sleeping between clean sheets every night of campaign and rising to don burnished leather and uniforms decorated with the high awards of allied sovereigns. Meanwhile those under their discipline, junior officers and soldiers alike, had circulated between draughty billets and dangerous trenches, clad in verminous clothes and fed on hard rations, burying their friends in field corners when spells from the front allowed and kicking a football about farmyards by way of relaxation. The implication of such disparities can be suppressed in the short term; modern armies are, indeed, mechanisms of such suppression. Their elaborate hierarchies - fourteen ranks interpose between a private and general - act as a system of screens to camouflage the altitude at which dangerous orders are generated. Since the subordinates most exposed to the consequences, ordinary fighting men, receive those orders from someone scarcely less exposed than themselves, or perhaps even more so - the platoon or company leader - resulting dissatisfactions are dissipated at that level if they are indeed felt or expressed. It takes much time for a bad or inconsiderate general's qualities to diffuse downwards through the barrier layers of rank, and even more time for that diffusion to type him for what he is. Even when so typed, he continues to be protected by a parallel mechanism of suppression, the code of military law. Unlike civil society, military society makes dissatisfaction with a superior, once expressed in any form, a criminal offence; even 'dumb insolence' attracts confinement, while fomenting dissent is mutiny, in time of war an act punishable by death.
 

Yet, as even bad generals know, hierarchy and discipline cannot suppress the implications of risk disparities for ever. Even while the First World War raged, some armies had begun to recognize the deficiencies of chateau generalship and taken steps to alleviate them. Pétain, appointed to rehabilitate the French army after the mutiny of May 1917, not only instituted enlightened measures of welfare, more generous. leave, better food, provision for entertainment - but also took care to design a series of limited operations against the Germans whose small scale ensured their success. By learning that their commanders could lead them to victory- and some French generals, like Marchand, had always been models of the exemplary style - the disheartened poilus were gradually weaned back to optimism.
 

That the commanders of citizen armies should have so gravely abused the reasonable expectations of their followers is evidence of how artificial and unreal was the general staff culture in which contemporary commanders had been raised. That culture was modern and its intensity a function precisely of its novelty. The perception by which it had been created was not false. The sudden heightening of danger on nineteenth-century battlefields quite properly required the commander to withdraw himself, and the consequent delay in the acquisition of 'real time' intelligence rightly demanded that subordinates should act for him at times and places when and where he could not be present. The cultural mistake lay in elevating those subordinates to the status of an elite and their function to superior expertise. General staff selection and training, based on fierce competitive examination, produced in the years 1870-1914 côteries of military specialists whose professional exclusivity was overweening. A social chasm was thereby opened between those who thought and those who fought; worse, thinking came to be deemed more important than fighting in the conduct of war, the emotions of ordinary soldiers subordinate to the perceptions of staff officers and the making of plans superordinate to their execution.
 

'Knowing', of a limited and theoretical sort, thus came to dominate 'seeing' in the system of military values, with results whose undesirability was to be concealed until the spiritual revolt of European armies in 1917-18 made them stand plain. The history of the emotional life of armies ever since has been one of a retreat from that disjunction. Staff officers who, even when general staff culture flourished at its most intense, had nominally been required to alternate between staff appointments and troop duty, were subsequently and with increasing strictness actually required to do so. Staff training, formerly restricted to a minority, teas progressively been extended to the majority of officers. The dynamics of combat -its stresses and psychological climate- have come to form an ever larger subject of consideration in that training. Those who undergo it have demonstrated the military society's change of heart by the enthusiasm with which they cultivate intimacy with the man in the ranks and the frequency with which they seek his company. Leadership, of a style sufficiently heroic to satisfy Alexandrian exigencies, is the command mode to which modern generals now aspire. Their armies perform accordingly. The Israeli army, animated by a code of which 'Follow me' is the central tenet, defeated its Arab enemies with a consistency that seemed routine until in 1973 the Egyptian army, its leadership transformed by an internal revolution inspired by the heroic ethic, very nearly succeeded in reversing the pattern. The Chinese and Vietnamese armies, outstanding among victors in the post-war years, both insist on the closest personal identification of leaders with led. The British army, once infected as badly as any by general staff culture, demonstrated how completely it had cured itself of the disease by its victory in the Falklands, a triumph of heroic leadership against odds. And the American army, trammelled by a theoretical approach to warmaking though it tends to be, has elevated the management of small groups to so high a place in its operational doctrine that its general staff culture may now be judged to persist only in a benign form.
 

And yet the cure to which so many armies have successfully subjected themselves may, with perspective, now come to appear irrelevant to command's current central problem. For armies have, by the nuclear revolution of 1945, been set aside from that central place in the defence of nations they have occupied since time immemorial.
 

'In order that the men will be prepared for the fray in a manner they can comprehend easily,' advised Raimondo Montecuccoli, the commander will declare that it is not the army of the fatherland but the fatherland itself that is endangered because it will have nothing left if the army is beaten; that it has entrusted all its resources and power to the soldiers; that they are the repository of all its hopes that they surely do not wish to be destroyed.
 

Montecuccoli's assumption that the army in war opitomized the state, so that its commander was therefore burdened with essentially sovereign responsibilities, is one which would have held good at virtually any moment of the twenty-four centuries which this book has surveyed. It holds good no longer. Armies are now but one means by which states of the first rank - those deploying nuclear weapons or belonging to an alliance which does - defend themselves, and not only that: they are a subordinate means. Truly critical command functions no longer belong to generals, but have emigrated to the centre of political power itself, have been returned into the hands of constitutionally sovereign authority itself and subject those who exercise them - president, prime minister, first secretary - to their burdens. Those burdens, always awesome, have been heightened by the dimensions of nuclear power, to the level of the almost unbearable. For it is not merely the 'resources and power' of the 'fatherland' - nation, rodina, patrie, call it what you will - that lie at risk should those exercising sovereign authority through nuclear weapons fail or miscalculate; it is the physical survival of the millions of human beings who have entrusted their wellbeing to him or her. Today the political leaders of the nuclear states have become Alexanders, the repositories of ultimate military as well as political responsibility in the polities they head, but with this unmanning- or unwomanning - difference: that those whose hands lie closest to the weapons by which society is defended are those who, in the eventuality of their use, would be placed furthest from the physical consequences of their impact. Nuclear war would expose every ordinary man, woman and child in every nuclear-armed nation to the risk of instantaneous disintegration or, failing that, to the inevitability of secondary irradiation. Presidents, prime ministers, first secretaries would, by contrast, belong to the only group - and that a tiny one - whose survival would in any way be assured against immediate or postponed nuclear extinction. The imperative of example would, in short, have been stood on its head; those least involved in the prosecution of war and least equipped to protect themselves against its consequences - suckling babes, nursing mothers, the sick, the lame, the very old - would stand in the front line; heads of government, by definition also nuclear force commanders, would be sheltered in deep headquarter bunkers or sequestered in airborne control posts. The weak would risk most, the 'strong' least of all. What are the implications of this extraordinary reversal of command ethic?

The sequestration of the commander from risk in nuclear-weapon states is, for all the paradox it entails, a perfectly proper procedural response to the dangers by which they are encompassed. The propriety flows from the anticipated nature of a nuclear war itself. For nuclear weapons may have three targets: the first is the civilian population and cities of an enemy state; the second are its weapons and weapon sites; the third its centres of command. Strategies dedicated to the destruction of each of them are called respectively counter-force, counter-value and decapitatory. The logic which underlies each may be characterized as follows: an attack on weapon systems (a 'first strike') would, if successful, win the war; but, weapons being numerous and well-protected, all opposed nuclear powers reserve their weapons- belief at any rate has it - for a 'second strike' against cities, the menace of which is held to deter the first. Nuclear weapons thus hold each in thrall by the logic of 'mutually assured destruction'. But the logic has a chink. If one side were able to outwit another's warning systems and destroy its command centres, it might thereby escape the retaliation of a 'second strike' -the authority to order which would have been paralysed - and proceed either to destroy the enemy's weapon systems or simply dictate peace under that or associated threats.
 

The spectre of this strategy of decapitation, long perceived and well understood by all nuclear states, explains and justifies the measures taken to protect their high commands from attack. There are many. One is that of direct defence, providing leaders with command shelters proof against nuclear strike. A second is escape, the provision of airborne command posts which would carry leaders away from points of nuclear impact at moments of danger. The third is alternative command, the empowering of nominated and instructed subordinates to exercise nuclear command authority in the case of death, disablement or isolation of the sovereign. The fourth, complementary to the other three, is redundancy, the multiple duplication of command centres and channels, so as to permit the free flow of orders even when the primary centres and channels have been interrupted.
 

The American nuclear command system, about which most has been revealed, is known to include all these features. A National Military Command Ccntre in the Pentagon and a hardened underground Alternate National Military Command Centre collect and collate the intelligence - chiefly from satellite surveillance and ground radar sources -- by which the danger of nuclear attack is monitored and transmitted to the President. He has a Situation Room in the White Houee to which he would go in the event of nuclear alert - two were caused by false alarm in 1979 and 1980 and there have been several deliberate alerts - and could, if time allowed and risk sufficiently threatened, transfer to a communications aircraft, the National Emergeney Airborne Command Post, kept permanently ready at Andrews Air Foree Base near Washington. Operational alert authority over nuclear forces is exercised by the North American Air Defense Command located inside Cheyenne Mountain, Colorado Springs (though it is not hardened to survive a major nuclear impact), while command of the strike forces themselves belongs to the Strategic Air Command, one of whose generals is permanently airborne in a 'Looking Glass' aircraft and which also deploys an Emergeney Rocket Communieation System, mounted in one or more Minuteman missiles, which (presumably) would broadcast attack orders if all other instruments of command had been destroyed. In the event of the President's death, disablement or isolation, however, his authority would devolve first on the Vice-President, followed by a succession of cabinet officers, in strictly specified constitutional order, and thence on commanders holding 'predelegated authority' whose identity is concealed (but are believed to number the six or seven exercising 'unified or specified' command). Serving all in the nuclear chain of command is a multibranched communication network, which uses the national telephone network as its medium and assures, in the event of anything except a stateswide catastrophe, that legitimate nuclear launch orders would reach launch centres if ever issued.
 

So comprehensive is the American nuclear command and control system that the role of the man at its centre,, the President, has been described as that not of implementing nuclear response (or attack) but of precisely the contrary: assuring that missiles will always remain in their tubes or silos, and aircraft within national airspace, unless he specifically orders otherwise. The President is, in short, like the wise elder of a pre-heroic society, an inhibitor of conflict, not its instigator, director or leader. The President's command centre, writes Paul Bracken, the major authority on the matter, has as its function 'not to act as a trigger to launch nuclear weapons but as a safety catch preventing other triggers from firing'. Between the pre-heroic inhibitor and moderator of conflict, however, and the nuclear-power President interposes a crucial difference of statue; the former acts by open, the latter by secret method. The tribal elder who urges restraint does so through his links of kinship with his people, by prescription, sanction, direct action and, if necessary, example. The President who exercises restraint on behalf of his society does so, necessarily, by mysterious means. 'Detailed information on the procedures for using [nuclear control measures],' writes Paul Bracleen, 'is one of the U.S. government's most closely held secrets. Information about which location the President would go to, which communication lines he would use, how much predelegated authority would be given to provincial commanders and which communication system would be selected for sending firing orders are all surrounded in much deeper secrecy than that surrounding the technical characteristics of the weapons themselves.'
 

'The reason for such secrecy,' he goes on, 'is not hard to fathom.' Indeed it is not. Nuclear command and control secrets are, more than any others of the strategic system, those that an adversary would most like to penetrate. For, if penetrated, an enemy would then be able to calculate if a decapitating strike was feasible and, if it were judged so (admittedly by no means a foregone conclusion), exactly how, when and where to target his missiles. Moreover such secrets are, in the last resort, the only ones that a nuclear power can realistically hope to deny to another. Everything else in the system missile sites, radar stations, command centres, airbases, satellites- is physically substantial. Even the minute-by-minute locations of ballistic submarines and nuclear-armed aircraft are, in the last resort, ascertainable by surveillance methods because submarines and aircraft, being physical objects, return sonar or radar signals and are therefore identifiable, even if with greater or lesser difficulty, in time and space. The one insubstantial, physically immaterial component of the system - not identifiable or penetrable by surveillance systemsis the procedure by which its physical elements would be activated and operated. True, the communication links by which procedures are initiated are vulnerable to direct attack, as, by cryptology, is the coded language in which communications are transmitted. But because of the high degree of 'redundancy' (which simply means-scale duplication) in the links, and because, as far as we know, even the most advanced cryptology cannot break current cyphers in 'real time', the communication system may be judged for the present secure. What ought always to remain beyond the reach of anything the enemy can deploy against it, except the efforts of traitors or 'agents in place', is the nuclear command protocol itself - authentification codes, launch orders and the Single Integrated Operation Plan or its Soviet equivalent.
 

The necessary secrecy that surrounds these inner secrets, however, brings with it, at least in democracies, a central contradiction; that the single most important process of government - for what else is that by which the survival of a people is ensured? - is itself kept secret from the electors themselves. The existence of this contradiction may not, at first consideration, have the power to shock. Confidentiality is, after all, an admitted right of government even in the most thoroughgoing democracies - cabinet discussions, for example, are kept secret, as are the internal processes by which ministries and departments arrive at their decisions of policy - and for confidentiality to embrace the heart of national security procedures might seem a quite proper extension of that principle. The difference, however, between cabinet confidentiality and national security confidentiality lies in the fact that the first concerns what are indeed discussions, unpredetermined in their form, while the second are indeed procedures, having - presumably -the same formality and sequentiality as constitutional practices for the enactment of law, the appointment of officers of state and the declaration of war.
 

The development of such other procedures in the history of democracies is one of the progressive rolling back of the curtains of secrecy by which they were originally surrounded; an excellent definition of democracy is that it is a system of government in which rule is conducted by rulers in open view of the ruled. In the United States, for example, not only must Congress discuss the passage of bills in open audience and the President accept public endorsement, or disallowal, of his appointments of cabinet officers, ambassadors and judges; Praident and Congress are both bound to conduct debate over the declaration of war in open session. Yet, by an unprecedented reversal of the historic trend of democratic development, we now live with a state of affairs which surrounds a matter far graver than traditional declarations of war- a decision to initiate and respond to nuclear attack - with a secrecy so complete that we do not, for example, even know whether the launch procedure has been computerized or is still amenable to human check and balance.
 

Consider what that obscurity implies: if launch procedures have already been computerized - if, that is, machines are now instructed to order the launch of missiles at some predetermined presentation of warning signals by the other machines of the surveillance system -then democratic government is already hollow at its centre, for the leader elected as the guardian of their security by the citizens of any democratic state that is also a nuclear power - the United States, Britain, France - is no longer empowered to exercise moderation, restraint or second thought in the matter which may determine whether it survives or not as a society. If the opportunity for moderation, restraint or second thought still pertains - if command procedures, that is, have not yet been computerized - democratic electorates may breathe again. But they are still left altogether uninformed - and therefore unable to express either their approval or disapproval - of the measures instituted by government to control and direct the weapons by which they expect to be defended.
 

There may be, given the intrinsic nature of the deterrent relationship which holds nuclear adversaries in mutual check, no way by which democracies can bring nuclear launch procedures within their system of accountability. Democracies may, in short, have to accept a permanent and unalterable diminution of their right to know, to criticize and to amend. But if that is the case, then the relationship by which the people of a democracy and their leader are bound together has not only been changed fundamentally and for good; the nature of that change requires that democratic leadership must in future partake of a style and a character altogether different from any that has prevailed before.
 

Let us briefly remind ourselves of the imperatives that have combined to define leadership in the past: they have comprised an element of kinship, by which the leader surrounded himself with intimates identifiable by his followers 88 common spirits with themselves, thus guaranteeing that their mutual humanity, in all its strength and weakness, will be constantly represented to each other; kinship has been bolstered by sanction, the reward - or punishment -of followers according to a jointly accepted value system; sanction has been reinforced by example, the demonstration of the personal acceptance of risk by the authority who requires others to bear it at his behest; example has been amplified by prescription, the explanation of the need for risk-taking by the leader, in direct speech, to his followers; and prescription has finally been made concrete- reified would be the technical term - by action, the translation of leadership into effect, of which victory was the desired result.
 

Power over nuclear weapons has undermined or invalidated all these imperatives. The exclusivity of the nuclear community, burdened by secrets it is legally forbidden to communicate, and physically isolated from the community it is charged to protect, has sundered all kinship between it and society at large; sanction has lost its force, since the proper management of a nuclear system will generate no occasion for either punishment or reward, or none at any rate that can be readily revealed; the opportunity for example is, as we have seen, denied by nuclear logic, which requires the leader to be at least risk among all members of his or her society; prescription, in consequence, is self-defeating, if not downright destructive of authority, since all exhortation to courage and fortitude invites the riposte, And What of You?; and action, the test by which leadership has always ultimately been validated, is, of course, denied by the necessity to avoid all outcomes in nuclear confrontation whatsoever.
 

The leaders of nuclear powers are therefore fixed in a dilemma: how to validate (legitimize, political scientists would say) their authority without recourse to any one of the heroic props always previously found necessary to that end? Autocracies, like the Soviet Union, confront this dilemma in a less acute form than democracies, since the autocrat does not shrink from using force to impose his will, up to the limit where force must rebound against his hold on power; but even an autocrat as extreme as Hitler took the precaution to employ reward, exhortation and a carefully contrived image of himself as hero as a means to palliate his dependence on direct coercion. Democracies, contrarily, have diluted the heroic appeal by adducing the principle of consent to justify disparities in risksharing. Thus Abraham Lincoln, with but a few days of bloodless campaigning against the redskins to his name, and Franklin Roosevelt, a man absolutely untouched by military experience, could both demand the ultimate sacrifice of their fellow citizens on the grounds that the voters had, by electing them president, willed them powers of war as well as peace, with all the consequences that flowed from that act.
 

Liberal democracy has never failed, none the less, to invoke the apparatus of heroic leadership when it could, its comparative brevity and localization as a form of government making consent alone too uncertain a means of legitimizing commands which regularly spared from their consequences those responsible for their issue. Thus Gladstone's anti-militarism could preserve its consistency because he was able to avoid leading Britain into blood-letting on any large scale. But the authority both of Asquith's government and of the Union sacrée of 1914-16 was undermined by the evident inconsistency of their pacific inclinations and their warmaking policies, to say nothing of the wholly civilian backgrounds of the ministers who composed them. Not only have the leaders of the democracies in the total war era subsequently taken trouble to publicize their individual military records, if they had any to claim -as Churchill, Kennedy and Eisenhower, for example, notably could and, more modestly, Truman, Nixon, Carter and even Reagan also; they have additionally - and the consent principle notwithstandingregularly mobilized the imperatives of kinship, prescription, sanction, action and, where possible, even example to heighten their military authority.
 

Those expedients, it has been demonstrated, no longer avail. What therefore is to take their place? No programme of national reassurance can do so. The early efforts of the United States and British governments to 'educate' their populations in techniques of nuclear survival foundered on evident and quite rational disbelief, and were indeed inevitable victims of the complementary but contrary half of the same strategic argument: that security in a nuclear world derived from the certainty of retaliation, otherwise known as Mutually Assured Destruction. That being so, the governments of democratic states which are also nuclear powersthose of nuclear autocracies should also take heed, but are under less compulsion to do so - must establish a new form of military command. It is best characterized as 'Post-Heroic Leadership'.
 

Post-heroic leadership will require that most difficult of all feats, in government, a transition from one system of appeal to human responses to another and quite different one. Traditional leadership in all its forms, even the most liberal and humanistic, has always had to delve deep into what is instinctual and emotive in the collective psyche to find the elements which will lend it force. Democracy, in its fundamental dimension, is a means of limiting the egotism and waywardness of those who exercise power by replacing them with others when their pretensions become intolerable. The alternation of declared political positions is a superficiality of the democratic system; it is the popular right to deprive one constituent group of the political class of authority and invest it in another that makes democracy morally superior to autocracy - call it monarchy, aristocracy or oligarchy - in any of its forms. But traditional democracy, fragile flower that it is, has never derived its force from moral argument alone. Morality is, in the last resort, founded upon reason, but mankind in the mass does not choose or unchoose by purely rational process. The most successful democratic leaders have known as much and acted accordingly, buttressing their reasoned arguments with a carefully calculated appeal to material interest and emotional response and, with an elaborate 'presentation of self', contrived to personify the image of leadership closest to that which a people, at one time or another, seeks for its own. The wastepaper basket of democracy is filled with the lives of would-be leaders whose highmindedness led them to reject such artifices and rest their approaches to the electorate on pure rationality.
 

The advent of nuclear weapons has put a term to the semi- and anti-rational style. Mankind, if it is to survive, must choose its leaders by the test of their intellectuality; and, contrarily, leadership must justify itself by its detachment, moderation and power of analysis. Hopes of transition to such a style of leadership need not be based on mere wish. The history of the world's first and only acute nuclear crisis lends substance to the belief that it may be achieved.
 

That episode was the Cuban Missile Crisis of October 14-27, 1962, which brought the United States and the Soviet Union to the brink of war. It retains the keenest significance for the modern world for three reasons: it is the only nuclear crisis of which we have a detailed account written by an insider (Thirteen Days by Robert Kennedy, brother of the President and US Attorney-General); it was presided over, on the American side, by a leader who had revived the heroic style in an extreme form; it was conducted, nevertheless, in a strictly post-heroic manner and resolved with rapid and complete success.
 

What were the bases of that outcome? The first was President John F. Kennedy's determination that the three competing velocities of the crisis - the velocity of the Russian initiative, the velocity of the necessary American response, and the velocity of assessment and decision - should be identified and separated. The second was that assessment should be entrusted to a group of men, the Executive Committee (Ex Comm), chosen for their expertise and sagacity, relieved temporarily of other responsibility and convened to meet outside his presence. The crises therefore developed, as far as was possible, in a way which ensured that the velocity of events did not accelerate the velocity of decision-making, with all the undesirable consequences of rushed and unconsidered judgement that might otherwise have ensued.
 

At the outset, the Ex Comm accepted the assessment that the deployment of Russian missiles to Cuba, the preliminaries to which had been discovered at an early stage, would take two weeks to complete while the appropriate American military dispositions required only forty-eight hours, thus leaving twelve days for rational consideration; the Ex Comm also quickly identified the three outcomes to which dispositions might lead - air blockade, bombardment and invasion of Cuba; and it finally and quickly agreed on how it should organize itself for the appropriate decision-making in the time it had decided was available.
 

The nature of the decision-making appears, in retrospect, the most impressive and significant feature of the crisis. The Ex Comm decided at the outset not to organize itself in a hierarchical way; it forswore 'leadership' from the start. 'We all spoke as equals,' Robert Kennedy recalled. 'There was no rank . . . we did not even have a Chairman . . . the conversations were completely unstructured and uninhibited. Everyone had equal opportunity to express himself or to be heard directly.' Some found the burden of equal responsibility too heavy to bear. Dean Rusk, the Secretary of State, underwent what Kennedy identified as a nervous breakdown early on and thereafter absented himself. McGeorge Bundy, the National Security Advisor, proved incapable of taking a consistent line. He was 'first for a strike, then a blockade, then for doing nothing'; finally for a strike again. But that at least demonstrates that the Ex Comm avoided 'groupthink'. In fact, and despite the insistence of its only military member, General Maxwell Taylor, on advocating military action, the Ex Comm took only three days to reach a majority decision for blockade. Another day was devoted to technical discussion outside the Ex Comm forum and a further two, October 21-2, to conferences with President Kennedy himself. By the sixth of the thirteen days available, therefore, a rational response to the Russian threat had been identified and endorsed. A week later the Russians also had accepted its logic and turned their missile-carrying ships away from Cuba to a homeward course.
 

The history of the Cuba crisis therefore offers both reassurance and hope that future nuclear crises may be resolved as rationally and harmlessly. But it must be remembered that the world has moved on singe 1962, and moved on apace. Pace is, indeed, the crux. Of the three velocities which drove that crisis - velocity of events, velocity of response and velocity d decision-making- the last has remained static, as it must; the human mind and the human tongue work no faster in 1987 than they did in 1962 or, to cast back to Alexander, than in 334 cc. But the reporting of events, which feeds the pace of crisis, has accelerated significantly, and the velocity of response - military disposition and alert - even more markedly. The Soviet Union in 1962 was seeking to dispose missiles into a gap in the American warning system by ponderous sea transport, taking days to complete. Today its missiles are disposed on submarines from which their flight-times to targets within the forty-eight contiguous United States are measured in minutes.
 

A velocity of unvarying pace - that at which human beings receive, assimilate and discuss information and decide what must be done in its light - therefore competes with velocities which are constantly quickening. Because they concern activities which are also swelling in volume- more long-range, short-flight-time weapons, more information - the human beings who are hampered by the unvarying velocity of their own thought processes seek to equalize the imbalance by reducing the flow of information to more manageable proportions and bringing the weapon array under ever more centralized control. The desired end of this trend is for all weapons to be made obedient to a single command, which in its turn will be determined by a single 'go' or 'no go' reading from all incoming information. The modern supreme commander- president, prime minister, first secretary- is, in short, seeking to return from the complexities of strategy to the simplicities of tactics; to a situation in which the warrior both sees his target and, by direct observation of its behaviour, launches or stays his weapon accordingly.
 

But desire and circumstance meet here, alas, in irresoluble conflict. To reduce a large volume of complex information quickly to a simple 'read out' can be done; but only by interposing a dense filter of machines and intermediary personnel between the decision-maker and reality. Machines, in the circumstances, can make of information only what they are told or programmed to do; while intermediary personnel, as they assess the information that passes before them, inevitably encroach upon the ultimate function of the supreme decision-maker. The result may be to persuade the strategist that he enjoys the tactician's direct vision and freedom of action; but the sensation will be an illusion.
 

Worse, it will be an illusion pregnant with disastrous consequences. Not only will it tempt the supreme commander into decisions which programmed and mediated information may, all desire to the contrary, have made for him. It will also tempt him to act the tactician and - therefore - the hero. Kennedy, as we saw, managed to resist that temptation. But Hitler assuredly did not. And though Hitler's personality was grossly aberrant, the means by and the environment in which he exercised command were not so at all. Indeed they directly resembled those that prevail in Washington, Moscow, London and Paris today. That being the case, the possibility that the supreme commander of a nuclear weapons state will at some time in the future yield to the temptation of false heroics and seek to play the tactician, just as Hitler did, cannot be ruled out of account.
 

The prospect is potentially catastrophic. How can it be forestalled?
 

Two methods suggest themselves. The first is to decelerate the two velocities- of events and appropriate response - that drive the critical velocity of decision-making. Easier said than done is the obvious response. But efforts to decelerate are nevertheless afoot through the vast American (and Soviet) scientific enterprise called the Strategic Defense Initiative. 'Star Wars' is both seen and represented as a system of protection; President Reagan's depiction of the eventual SDI product as a missile-proof 'astrodome' best conveys that aspiration. But even his warmest supporters concede that the dream of an astrodome is an illusion. Total missile-proofing probably lies beyond the capacity of any scientific community to achieve. That is not to say, however, that &Star Wars is without merit, political or military. On the contrary, it is an enormously hope-giving initiative, if it is seen, as it properly should be, as a mechanism to procure not total defence but relative delay. Deterrence, as we are currently reminded, derives its logic from its instantaneity, the certainty of instantaneous retaliation by #coed strike should a first strike fail. The outcome of that dialectic is called Mutually Assured Destruction. Because Star Wars threatens to dilute mutuality, assurance and destruction, it is seen as damaging to the deterrent principle. If, on the other hand, its influence is calculated not on the effect it might exert on outcomes but on the delay it could impose on decision-marring, its desirability switches from negative to positive. Nuclear Weapons strategy within a Star Wars system would, if a crisis boiled from menace to action, almost certainly result in some missiles reaching their target on one or both sides. But, horrible though such an experience would prove, the event would not only be bearable in a way that Mutually Assured Deatruction would not. It would also allow the contestants to think and calculate rationally during the course of the exchange - to act, that is, as strategists instead of tacticians- and to perhaps extricate themselves from deepening trouble rather than be driven further into it by the velocities of event and response.
 

Star Wars therefore offers hope; but only through the prospect it promises of reverting from the diplomacy of the hair-trigger to the more traditional rhythms which animated international relations before the coming of nuclear weapons. The outcome of a nuclear war, even one mediated by SDI mechanisms, would still be so much worse than of any ever known to the world that no strategic theorist may properly portray the Initiative as man's best and last resort. Mankind needs not new hardware but a change of heart. It needs an end to the ethic of heroism in its leadership for good and all. Heroism, as we have seen, is not a necessary constant in the way that societies work. Heroism is an irrational and emotional response to challenge and to threat. In a world of riches and poverty, better and worse land, full and empty spaces, good and bad gods, true and false creeds, the appeal of heroism was a natural temptation to those who felt that it would lend the decisive cutting edge to weapons otherwise inadequate for victory over the stronger, the more fortunate, the better favoured by history. It was also a splendid cloak for the bully, the tyrant, the ideologue and the lunatic, not least when the urge to tyrannize came to possess whole peoples, rather than those given to or taken by them as leaders.
 

For much of manta known past, the heroic ethic, in some guise or other, has characterized the style of government by which he has conducted his affairs in most quarters of the globe. A few people in a few places have found other means to legitimize the authority under which they have lived. The theocracies of China and the Middle East represent one alternative form. The liberal democracies of the nineteenth-century West represent another. 114th chose to preserve and cultivate the heroic ethic, none the less, in certain carefully isolated sectors of their societies, and to sustain the creed of struggle within their larger political philosophies. In the theocracies that creed belonged with the depiction of those 'outside' as barbarians or unbelievers. In the democracies the creed of struggle worked to energize politics from the inside, making 'heroes" of men and women simply through their advocacy of the opposed positions of right or left, red or white, us or them.
 

The concept of struggle, and its attendant ethic of heroism, broods over us all today. It lies at the heart of Marxism and hovers not far from the guiding belief of democracy in the values of human freedom and choice. Yet the spectre of risk, by confronting which the leader authenticated himself as hero, is no longer deflected from those who follow him by the singular role he takes for himself. On the contrary, it diffuses the whole arena of struggle, threatening everyone equally, if not indeed the led more directly than their leader. The traditional means by which the leader sought to validate his followers' sharing of the risk he led them to face - the cultivation of a sense of kinship, the use of sanction, the force of example, the power of prescription, the resort of action - now all fail. Indeed, what is asked first of a leader in the nuclear world is that he should not act, in any traditionally heroic sense, at all. An inactive leader, one who does nothing, sets no striking example, says nothing stirring, rewards no more than he punishes, insists above all in being different from the mass in his modesty, prudence and rationality, may sound no leader at all. But such, none the less, is the sort of leader the nuclear world needs, even if it does not know that it wants him. 'Post-heroic' is the title he might take for himself. For all is changed, utterly changed. Passing brave it may once have been to ride in triumph through Persepolis. Today the best must find conviction to play the hero no more.