CHAPTER 4 Possible Worlds

Counterfactuals, or what some historians term counterfactuals, come in many different guises, and it is important to disentangle them before reaching a conclusion as to their usefulness or otherwise. Politically motivated fantasies, however widely they are believed, involving, for instance, the survival of Nazis in secret underground bunkers, conspiring to build an authoritarian New World Order, are not really counterfactual because they lack any real interest in cause and effect. Nor are minutely detailed pseudohistorical investigations purporting to show that Hitler did not die in the Berlin bunker, or escaped to Argentina: here the interest is in the alleged fact itself, not in its possible consequences. A more sinister alteration of the past can be seen in Stalin’s attempt to airbrush his erstwhile rival Trotsky out of photographs taken during and immediately after the Bolshevik Revolution in Russia, and by Stalinist historians to write him out of the historical record—a practice chillingly represented in George Orwell’s novel 1984, where the hero Winston Smith’s job is to rewrite old newspapers to delete evidence inconvenient to his political bosses in the present. Retrospective historical falsification of this kind is not really counterfactual, however, because it is merely concerned with rewriting the entire factual record, not with positing an alternative to it deriving from one small change in it. Nathaniel Hawthorne’s story “P’s Correspondence” falls into a similar, related category: he imagines a situation in which people like Napoleon, Shelley, or Byron have survived into the time at which he is writing—1845—instead of dying earlier, as they actually had.1 This is simply imagining another reality, rather like Jonathan Swift’s Gulliver’s Travels or Thomas More’s Utopia and other fictions of its kind. The survival of Hawthorne’s subjects does not actually alter the larger situation of the world in 1845, it is simply interesting in itself, just as satirical alternative world like Brobdingnag or Lilliput do not actually change the real world their author lived in, they just hold up an ironic mirror to it.

Historical novels might seem to fall into the category of the counterfactual, and in his introduction to Sir John Squire’s collection If It Had Happened Otherwise …, Sir John Wheeler-Bennett discusses what he called “imaginary history,” including Anthony Trollope’s and Benjamin Disraeli’s parliamentary and political fictions. These are set in a recognizably contemporary or recent historical world—in their case, Victorian England—with recognizable institutions such as the House of Parliament, and recognizable figures, such as prime ministers and bishops, but narrate fictional events that never happened and depict fictional characters who never existed. The same might be said of many other novels in the genre begun by Sir Walter Scott’s influential medieval fictions. But these are not counterfactual novels, because they do not posit any causal influence that created the parallel world they depict, as Wheeler-Bennett correctly noted. Moreover, neither they nor the much broader genre of historical novels make no significant changes to the basic historical context; they invent dialogue, character traits, and people, but not the big events, structures, or institutions of history; indeed, historical novelists usually go to considerable pains to “get it right” by reading and consulting standard historical works on their chosen period; Hilary Mantel’s extraordinarily successful novels about the Tudor statesman and politician Thomas Cromwell, for example, have garnered numerous prizes and a vast readership not just because of the brilliance of their style, characterization, and structure, but also because of the patina of historical authenticity with which they decorate the action.2

Nor, for the same reason, can one classify as counterfactual the many novels written in England before the First World War warning of the consequences of continuing British complacency about the German arms buildup: novels in which Britain is invaded by German armies and ends up groaning under the iron heel of the Kaiser’s jackboot, or similar fictions written in the 1930s, or the Eurosceptic novels of the 1990s like Andrew Roberts’s The Aachen Memorandum, which envisions a future in which British domination by the EU in the twenty-first century is not dissimilar to what British domination by the Germans might have been like in the twentieth. Roberts’s imaginary dystopian future is clearly based on a real dystopian past, but its conditional is based in the future and does not require any alteration to be made to what has happened already.3 Similarly, novels or essays that reverse reality to provide the basis for a satirical commentary on the present, as, for example, with the imagined discovery and conquest of Europe by the Mayas or the Incas, the subject of fantasies by Spanish writers like Unamuno or Fuentes, do not depend on diverting the stream of time, but simply on turning reality upside down.⁴ True counterfactual scenarios, whether historical or fictional, always involve drawing historical consequences, often far-reaching in nature, from altered historical causes.

A great deal of the time, what this produces is banal in the extreme. Jeremy Black’s book, for example, is wholly devoted to accounts of how things might have turned out differently from the way they did, which, he claims repeatedly, all goes to show that contemporaries didn’t know what was going to happen next, and therefore that they had considerable freedom of choice. Indeed, “rulers and ministers,” he claims, had “the free will to enable them to defy the normative character of the policies that should derive from an understanding of national interests,”⁵ or perhaps better said, their view of national interests was not necessarily the same as other people’s. But the point here is that by bracketing out the factors that constrained their freedom of choice, we present a wholly false picture of that choice as having been unfettered. Rulers like Catherine the Great of Russia did indeed indulge in abrupt changes of policy, but always within the limits of what was acceptable; a good number of Russian czars were murdered because they transgressed those limits. To take a different example, namely, Ferguson’s analysis of the outbreak of World War I, Aviezer Tucker notes that “Ferguson constructs decontextualized historical agents, isolated from larger cultural and economic contexts that precluded the kind of decision-making he would have undertaken” had he been a member of Asquith’s government. Here again, to make the counterfactual work, individual decision makers have to be presented, implausibly, as free-floating agents. The larger contexts in which Sir Edward Grey actually made the decision to go to war in 1914 made it likely that even had the liberal government fallen as a result of the resignation of ministers who were against going to war with Germany, the British would have intervened in the war sooner or later anyway.6

Black’s version of the “counterfactual,” which amounts to no more than the possibility that things might have turned out differently, in no way helps us explain how or why they turned out as they did, not least because he prefers “counterfactuals that offer complexity and indeterminacy,” rather than those that offer “answers.” Ultimately, his counterfactuals are not really counterfactuals at all, or rather, they are not counterfactuals in the sense used by most other exponents of the genre. If the only thing that can be said in their favor is that they “make us much more aware of the role of the contingent,” that is not very much, and once more confuses contingency with counterfactualism.⁷ Other contributions to counterfactualist collections go in the opposite direction and eschew speculation altogether; thus, for example, Tucker points out that Michael Burleigh’s archivally grounded essay in Ferguson’s collection on Nazi plans for Europe is not counterfactual because it does not say what would have had to be different in the history of the war for them to have been put into effect.⁸ A much larger category of writing speculates on too limited a level to be worth discussing as counterfactualism, namely, the very numerous essays of military history that discuss how a battle that went one way might have gone the other. Undoubtedly the most comprehensive of these is the volume published in 1997 by Dennis E. Showalter and Harold C. Deutsch, If the Allies Had Fallen. In no fewer than sixty separate essays, military historians consider counterfactuals ranging from what might have happened if the British had not managed to decode German radio traffic to “what if Stalin had adopted the supposed Shaposhnikov proposal for concentrating defenses along the Stalin line?” Many of these are unrealistic (e.g., what would have happened had Hitler let his generals get on with the job they were paid to do instead of interfering all the time). Essentially the essays, many of them very brief, are devoted to refighting the strategy and tactics of World War II with a view to rectifying the mistakes, or occasionally underlining the correct decisions, made by the participants.

Rather different in style and intent is another subcategory of counterfactual, extremely popular in Britain but also enjoying some vogue in the United States, namely, speculations about what would have happened had one or other politician become prime minister or president instead of the individual who actually did. The enterprising Conservative publisher Politico has made something of a minor industry out of such imaginings. Duncan Brack’s 2006 collection President Gore … and Other Things That Never Happened, contains nineteen essays on individual politicians, mostly of the wishful thinking variety: Al Gore becomes U.S. president in November 2000 and avoids the Iraq War; the liberal statesman Gustav Stresemann does not die, as he did, in 1929 but lives on to save the Weimar Republic; Gavrilo Princip’s bullet in Sarajevo misses the archduke Franz Ferdinand and the twentieth century turns out to be far less disastrous than it was in reality. More narrowly focused on British politicians, Duncan Brack and Iain Dale’s 2003 collection Prime Minister Portillo … and Other Things That Never Happened contains a familiar mixture of wishful thinking and its reverse, peppered with not a few essays that conclude that things would have been much the same anyway had, for example, Ted Heath won the 1974 election or Margaret Thatcher lost the 1979 election. Much the same can be said of the same editors’ Prime Minister Boris … and Other Things That Never Happened (London, 2011). All these collections are indeed counterfactuals, but they focus exclusively on individual personalities, taking what is known about them, perhaps spicing it up a bit, and then, acting in the belief that personality is all, they postulate (with a few exceptions) major changes resulting from the alteration in their personal fortunes. Slightly different is Francis Beckett’s The Prime Ministers Who Never Were (London, 2011), which takes individual British politicians and puts them into Number 10 Downing Street. The usual suspects make their appearance, including Oswald Mosley, who creates the European Union, and Lord Halifax, who confounds his posthumous critics by deciding to carry on fighting Hitler. Reading through the often witty and irreverent pieces in all these collections, it is hard to avoid the conclusion that they are intended as a lighthearted entertainment for the chattering classes, perhaps also as an implicit encouragement to aspiring politicians.

For all their emphasis on the seriousness of their intentions, the new counterfactualists are not—fortunately for their readers—averse to mixing argument with entertainment or larding their speculations with whimsy and humor. Reviewers of the Roberts collection noted that all twelve essays were “good” and “entertaining” and praised their “playful” nature, calling the book “a hymn to the accidental and the erratic.”9 Humor creeps into even the most straight-faced counterfactuals, as for example when Holger Herwig has a collaborationist British government under the rule of the Germans appoint William Joyce (“Lord Haw-Haw”) as director-general of the BBC.¹⁰ Dominic Sandbrook’s series of forty counterfactual historical essays written for the New Statesman in 2010–11 is clearly intended as an entertainment (the only really serious one, the last in the series, and some four times the length of any of the others, is predictive, focusing on the performance of David Cameron’s Conservative-led coalition government in its first five years of office, from 2010 to 2015, even if it is written as a pseudohistorical article from the perspective of the end of this period). Sandbrook ranges across the whole of recorded history, beginning by having Egypt replace Rome after Octavian loses the Battle of Actium (“in the long term the rise of Egyptian power was inevitable”) and Africans, convinced of Europeans’ “stupidity, indolence and general inferiority,” launching a “Scramble for Europe” in the fullness of time. William the Conqueror loses the Battle of Hastings, and the Anglo-Saxons continue in power up to the administration of “Chief Ealdorman Aedgifu Thatcher,” maintaining a sturdy British independence from the Continent (a bit of Eurosceptic wishful thinking here, perhaps). Henry V of England lives into a ripe old age and conquers France (more wishful thinking), with Joan of Arc as his mistress (a likely story), but the united Anglo-French kingdom becomes too remote from the people, and the monarchy is overthrown when the Tower of London is stormed in 1789 by a republican mob led by Charles James Fox. Catholicism triumphs in England in two of Sandbrook’s fantasies, to be led in the present by “Cardinal Dawkins,” but the nation fails to industrialize and slips into a “nightmare: the civil war of the 1930s, the leftist massacres of priests and nuns, and the reactionary backlash under the long, ultra-clerical rule of President Muggeridge” (this is the very opposite of the wishful thinking usually characterized by those who imagine the failure of Protestantism in early modern England).11

This is all great fun, and Sandbrook’s determination to keep it so is evidenced by his rigorous avoidance of topics that might get too serious, like a possible Nazi occupation of Britain. As he nears the present, his amusing counterfactuals focus more heavily on British politics, often achieving their effect through witty reversals of what we know happened, dexterous parallels with the present, and unexpected twists and turns in the imaginary story. One should not burden these brilliantly clever brief essays with a ponderous excess of analytical zeal. Still, it is noticeable that most of the time, Sandbrook offers not a counterfactual history, in the sense of one change in the chain of events leading on to a series of other changes that seem with at least some degree of plausibility to follow from it, but a parallel history in which, for example, Oliver Cromwell’s son Richard (“Farmer Dick”) does not retire to the country but becomes the Merry Monarch, to be succeeded in due course by “the dissolute George Cromwell in the 1820s” (i.e., a parallel George IV), a “tipsy, philandering Herbert Henry Cromwell” in the early 1900s (a parallel H. H. Asquith, the actual prime minister of the day), and even, in the primaries for the presidency founded in the seventeenth century by Oliver, “two Cromwells, Praise-God and Ed” (the Miliband brothers, who fought the Labour Party leadership in 2010, needless to say).12

In this lengthy series of short essays, Sandbrook is following the procedure set by Niall Ferguson in the afterword to his Virtual History volume, which posits an alternative history from 1646 to 1996, the date at which the book was completed. Ferguson has a great deal of fun attacking “determinists” who see the alternative course of history he describes as having been inevitable, presenting in his afterword what we know actually happened as a series of “counterfactuals” and thus repeatedly emphasizing the contingency of events. The narrative starts with the Royalists winning the English Civil War, and goes on to a future in which the Stuarts still become constitutional monarchs whose political flexibility and military skill allow the American colonies to be retained and the French Revolution to be avoided by financial reform; in addition, industrialization brings rising living standards that appease working-class discontent, Marx becomes a “millenarian Jewish prophet” and Lenin an Orthodox priest (later executed as a German spy), and the Holy Roman Empire is reformed through an Austro-Prussian alliance and keeps going as a decentralized federation through the nineteenth century, winning the First World War in 1915 while Britain stays neutral. The result is a “European Union” that respects the integrity of the British Empire, but still falls under the influence of the Nazis, who transform it into a “leader-state,” conquer France and invade Britain, forcing it to join the new “German-European Union.” In eastern Europe, the Germans defeat the Russians, but Hitler’s death from the bomb planted by Stauffenberg destroys German morale, the Russians, led by the Patriarch Djugashvili (Stalin) counterattack, the Japanese bring the Americans into the war, but D-Day fails, and the Russians end up in control of Europe. In Britain, Prime Minister Thatcher loses the Falklands War and is succeeded by Michael Foot, whose calamitous premiership helps open the way to the economic and political collapse of the West and the breakup not only of the transatlantic confederation but also of the United Kingdom into its constituent parts. The way is open for the future dominance of the East.13

Cleverly weaving together the threads spun by the preceding essays in the book, this is a highly entertaining narrative that repeatedly brings a smile to the lips. Yet what is offered here runs counter to the general tenor of the Virtual History collection in a number of ways. First of all, it is patently intended as a jeu d’esprit in the style of Sir John Squire’s collection (and as an entertainment it is a good deal superior). Ferguson’s intention, as announced in his preface, is clearly to get away from this kind of thing and establish counterfactual history as a serious tool of scholarship.¹⁴ Secondly, what the afterword offers is in fact not a counterfactual history, an altered event or circumstance leading through an apparently logical chain of consequences on to an altered state of affairs some time in the future, but, like Sandbrook’s essays, a parallel history that shadows the history that actually happened but reverses it at every step. Causation is thrown out the window, and there is no attempt to consider how one change in the pattern of events might have affected others. It is simply asserted, for example, that a victory by Charles I in the seventeenth century would have led in time to a constitutional monarchy in Britain, though no reasons are given for supposing this would have happened. Ferguson has to do this so as to get history roughly back on course, as he has to have a First World War and a Second World War; otherwise the narrative would drift too far away from the actual course of events and so cease to mirror them. But this procedure is entirely arbitrary and ignores both the possible chains of causation and the possible intervention of contingency.

In Ferguson’s “virtual history,” Argentina wins the Falklands War in 1982, for example. But the point of course is that had Charles I been victorious over the Parliamentarians in the seventeenth century, it is probable that the subsequent course of events over the next three centuries, though unpredictable, would have been such that there would have been no Falklands War, because an event such as the Falklands War is the product of a specific set of historical circumstances that in practice would not have occurred without a whole series of previous historical circumstances; a lengthy chain of causation, in other words. Indeed if Charles I had been victorious over the Parliamentarians, it is unlikely that Mrs. Thatcher would have been prime minister in Britain 340 years later, since the conditions that would have enabled women to achieve the vote, stand for Parliament, and lead political parties, might never have materialized. Altering one part of the kaleidoscope of history shakes up all the others in ways that are quite unpredictable. In this narrative, Ferguson jettisons all the rules and cautions he has so carefully elaborated earlier in the book and engages in wild historical fantasy.

Nor is this fantasy in any way value-neutral. A counterfactual narrative that includes the survival of the British Empire (including North America and ruled by the Stuarts), along with the Holy Roman Empire, into the late twentieth century, the absence of the French and Russian Revolutions, and more besides, is conservative wishful thinking with a vengeance. Yet although Aviezer Tucker claims that Ferguson develops these ideas in fulfillment of “some kind of personal utopia,”15 there are negative elements too, notably the loss of the Falklands War, the disastrous premiership of Michael Foot, and the ominous rise of the East, serving as a warning for the future. Conservative Euroscepticism appears yet again in the portrait of the European Union as an expanded German Reich, though wishful thinking then has it destroyed in the end. Such imaginings go far beyond anything that is even remotely plausible; nor, it is clear, is it intended to be. Just to take a handful of examples: the victory of the Stuarts in the seventeenth century would have been unlikely to have led to a constitutional monarchy unless there were profounder social and economic forces at work. An absolutist Stuart England was hardly likely to have been the site of the industrial, political, scientific, technological, and indeed sociopolitical achievements that underpinned the imperial domination achieved by Britain in the nineteenth century.

A Europe dominated by the German military, as I have already suggested, would have been unlikely in 1915 to have allowed the British Empire to continue in its previous form, as indeed the much weaker Germany of the Kaiser was already contemplating in the decade and a half before his putative victory in the First World War. The historical record shows that Michael Foot invested just as much personal and political capital in the Falklands War as did Mrs. Thatcher, so that a defeat would have damaged him too (though one could also speculate that it would have won her a sympathy vote in the following election). Moreover, Ferguson’s parallel narrative contradicts other counterfactual hypotheses he has developed elsewhere. In his own essay in the volume, for instance, he hypothesizes, as we have seen, that a German victory in the First World War would have prevented the spread of the bitter resentments and economic catastrophes that destroyed the Weimar Republic and brought Hitler to power, whereas in the afterword he has Hitler coming to power anyway. Even if we do not take the afterword too seriously, this is a contradiction that suggests a strong degree of arbitrariness in such speculations.

The suspicion of arbitrariness is underlined by counterfactual essays that offer alternative courses following the initial change of circumstance rather than trying to build up a coherent case for one particular chain of consequences. Thus, for example, in Roberts’s collection the Catholic historian Antonia Fraser starts by presenting an “optimistic” scenario in which a successful Gunpowder Plot, having blown up king and Parliament in 1605, enthrones a new monarch, Elizabeth II (known to history as “the Winter Queen” through her short-lived occupancy of the Bohemian throne at the beginning of the Thirty Years’ War in 1618 as wife of the quickly deposed Protestant Elector Palatine), reconciles Catholics and Protestants, establishes religious toleration, and cements a close alliance with France, all arguments of the “wishful thinking” variety. A “pessimistic” scenario, she says, however, conceding “an inexorable beat to the march of history,” sees religious strife continuing.16

More strikingly still, quite a few counterfactual speculations end up by concluding that things would have taken the course they did anyway. In Ferguson’s volume, John Adamson describes a victory for Charles I making little real difference in the end, given the strength of the forces working in favor of an increase in parliamentary power, thus implying that the Civil War and the execution of Charles I were historically unnecessary. Like Fraser’s essay this embodies a strong element of wishful thinking but also admits that chance events only have a limited effect on the course of history, thus rather going against the editor’s proclamation of the supremacy of contingency and chance in history.¹⁷ Anne Somerset sees Philip II’s conquest of England after a successful invasion by the Spanish Armada in 1588 as making little difference in the end: a reconversion to Catholicism does little harm, English autonomy is preserved, Parliament meets as before, and Shakespeare writes his great plays. Simon Sebag Montefiore has Stalin panicking when the Germans invaded in June 1941 and reached the gates of Moscow six months later. He flees the Soviet capital (in reality, after a great deal of hesitation, he decided to stay, though it was a close-run thing), and the Soviet forces then abandon the city. But Foreign Minister Molotov and the Politburo arrest and shoot Stalin, who is replaced by Molotov, the great general Marshal Zhukov mounts a successful counterattack, and the Red Army wins the war anyway. Molotov stays in power until his death in 1986, when Mikhail Gorbachev takes over and implements reforms that bring about the end of the Soviet Union. In other words, Stalin’s abandonment of Moscow in December 1941 makes no difference in the long run. Similarly, in the scenario painted by Conrad Black, where the Japanese do not attack Pearl Harbor, the Americans still enter the war against the Axis powers; while Jonathan Haslam concludes that even had there not been an ideological confrontation between the Soviets and the West, the Cold War would have happened for geopolitical reasons in any case.18

The limited support many of these essays give in practice to the cause of counterfactuals as a vehicle for overcoming “determinism,” in the sense of the prioritization of larger historical forces over smaller, personal, chance, and contingent events and circumstances, is compounded by the fact that despite Ferguson’s criticisms of wishful thinking, other contributions to his volume, and a number of the contributions to Roberts’s collection, and still others again, fall headlong into the trap of imagining things would have been better had they been different. In his contribution to Roberts’s collection, the right-wing Anglo-Polish historian Adam Zamoyski envisions a world after Russia’s defeat by Napoleon in 1812 in which there is no socialism and no nationalism to breed wars, and Russia is driven back to the margins of Europe and therefore not in a position to oppress Poland, as it did in reality for the whole of the nineteenth century and much of the twentieth; but Zamoyski is evenhanded enough to posit a future in which Europe, while prosperous and united in something that looks remarkably like the European Union, is hamstrung by a bureaucratic system that crushes economic and cultural initiative. The Eurosceptic wishful thinking so evident in Zamoyski’s contribution comes in again as Norman Stone speculates that the putative failure of the assassination attempt on the archduke Franz Ferdinand at Sarajevo in 1914 would have been followed after a time by a violent implosion of the Ottoman Empire and its partition between Britain and Russia (thus involving the withdrawal of Russia from Europe and the strengthening of the British Empire). The Germans and French, who had missed the feast, join together in an economic federation looking, once more, like the present-day European Union, while Britain stands aside and prospers as a global power—another example of Eurosceptic counterfactual historical fantasy. Andrew Roberts has Lenin assassinated in 1917, the Bolsheviks taking a more moderate course than they did, the liberal Alexander Kerensky concluding peace, and a constitutional Russia helping defeat the Nazis in 1938, leading to the overthrow of Hitler and his Third Reich.19 In all of this, wishful thinking trumps historical sense, and the extrapolation of consequences over the decades is too far-reaching to be convincing.

In Roberts’s collection, the right-wing journalist Simon Heffer’s essay on the consequences of the bomb set off at the Conservative Party conference in Brighton in 1984 has Prime Minister Thatcher killed in the blast (in fact, of course, she survived) and being succeeded by Michael Heseltine, a charismatic Tory but committed European. The essay falls into the mold of “there-but-for-the-grace-of God” as Heseltine divides his party with his pro-European policies and opens the door for Labour to win the 1992 election. But the damage would already have been done, and “Britain would have been set as a highly taxed, inefficient country like modern France, Germany or Japan.” In similar vein, in a knockabout piece, David Frum imagines a U.S. president Gore hamstrung by environmentalist scruples, political correctness, and respect for international institutions when faced with the destruction of the Twin Towers in New York by Al-Qaeda in 2001, so that he is unable to mount an adequate response.²⁰ The possible objections to the politically driven imaginings here are legion. The evidence is that it was Mrs. Thatcher, with her late-onset Euroscepticism, who divided the Conservative Party, while Heffer’s negative image of France, Germany, and Japan would not be recognized by many even in the 1990s. Similarly, the real victor of the U.S. presidential election of 2000, George W. Bush, turned out by common consent to be one of the most incompetent American presidents of the modern era, and his illegal and poorly thought-out invasions of Afghanistan and Iraq achieved very few concrete results in the end.

A large number of these essays ignore Ferguson’s requirement that only alternatives actually consciously considered by contemporaries may be taken into account when developing a counterfactual. In practice, they throw a deus ex machina into the workings of history, whether it is a bomb that kills Margaret Thatcher, or an election that brings Al Gore to power, or a military victory by a general who was actually defeated, or some other reversal of historical circumstance that actually had little or nothing to do with alternatives confronting a decision maker. Grey in 1914, or Churchill and Halifax in 1940, may fulfill Ferguson’s requirement, but very few of the historians who spin out their counterfactual scenarios follow suit, preferring to focus on other pieces of the historical mosaic. This is, of course, because the requirement to take a decision as the starting point proposes an impossibly narrow area in which counterfactuals can operate. But in practice that area is narrow enough anyway. Counterfactuals have almost always been hypothesized in the history of politics and policies, diplomacy, war, and government, and in these fields only in explanations of events.

Some advocates and practitioners of “counterfactual history” share this view, from Robert Cowley, seeking to use it to restore belief in great men, to Jeremy Black, who similarly concedes that the “undermining of any sense of necessity has tended to appeal largely to historians working on those aspects of history where contingent events and human actions are most intuitively central, such as political and military history.”21 It is striking how often the same topics turn up in the counterfactual world, from Charles Martel and the Spanish Armada to the Battle of Waterloo and the Second World War. Of the twelve essays in Andrew Roberts’s collection, nine deal with wars, while the other three, respectively, ask what would have happened had the Gunpowder Plot succeeded, had Margaret Thatcher been killed by the Brighton bomb, or had Al Gore and not George W. Bush won the presidential election in 2002. Cowley’s first collection is exclusively devoted to military history; and rare indeed are the contributions to the genre that stray beyond the realm of high politics and warfare, like William H. McNeill’s essay on what would have happened had the potato not reached Europe, or Joel Mokyr and Kenneth Pomeranz’s essays in Unmaking the West on alternative scenarios of economic history.22

For the vast majority of counterfactualists, as Tristram Hunt has pointed out, “‘what-if’ versions of the past posit the powerful individual at the heart of their histories: it is a story of what generals, presidents and revolutionaries did or did not do. The contribution of bureaucracies, ideas or social class is nothing to the personal fickleness of Josef Stalin or the constitution of Franz Ferdinand.”²³ Yet, paradoxically, once the counterfactual narrative is launched, the personality at the center of the fantasy takes a back seat. The attention is all focused on the general, contextual factors that he or she affects by being killed (or not) or by taking a decision at odds with the one he or she actually took, or by winning (or losing) a battle or some similar action. Lubomir Dolezel, indeed, suggests that “historians focus on modulating significant (global) social, political, economic, or military historical conditions” in the counterfactual worlds they create, not on modulating the individual human being. “For counterfactual history, the individual is interesting only as a ‘leader’ of social, historically relevant actions or as an occasional agent of historically significant events.”²⁴ When Napoleon wins the Battle of Waterloo, the world political order is changed; when the Gunpowder Plot succeeds, England becomes Catholic; Franz Ferdinand is not assassinated, so World War I would not have happened and the whole subsequent history of Europe and America would have been changed; when Britain stays out of the First World War, the European Union is created.

As these and many other examples suggest, counterfactualism focuses almost exclusively on traditional, old-fashioned political, military, and diplomatic history of the sort that used to be dominant in the 1950s. Even in Unmaking the West most of the contributions are on wars and revolutions. And indeed it more or less has to be so. Large-scale counterfactuals are implausible by their very nature. The leap, for example, from the absence of the Black Death in fourteenth-century France to a fall in fertility in the eighteenth century that led to an acceleration of economic growth, as proposed in Geoffrey Hawthorn’s volume, is unsupported either by the extremely unlikely starting point, or by convincing historical linkages that posit a consequence over three hundred years later in the development of industry through a combination of labor shortage and high consumer demand such as happened in Britain.25 Similarly, Joel Mokyr, in asking how science and technology might have developed if there had been no industrial revolution in the West, is compelled to admit that “at the end of the day, it is hard to know precisely whether oriental science, had it been left alone long enough by the West, would not have developed into something so radically different from what we are use to that we cannot even imagine it.”²⁶ So this in itself is one major reason to be sceptical of the more far-reaching claims of counterfactual history: it not only assumes but also implicitly preaches a history where politics and warfare are the most important subjects to be studied; in other words, it advocates a narrow, traditional approach to the past that most historians have long since moved beyond, an area, however, where counterfactuals are almost impossible to deploy. Nowadays, the world’s most innovative historians focus overwhelmingly on social, economic, and cultural history, on global and transnational representations of the past, not on political or diplomatic history.

When it comes to painting a larger picture that goes beyond the political or the military, indeed, Ferguson himself falls back into what looks very much like determinism. Thus in his book Civilization, Ferguson explains the fact that Europe dominated the world between 1815 and 1914 by arguing that European economies were based on competition, European science was superior to that of China and other civilizations, European law respected property rights and gave rise to stable forms of government, European medicine improved health and life expectancy, European society was based on consumer culture, and Europeans worked harder than anyone else. As he goes through these factors chapter by chapter, the unavoidable cumulative impression is one of inevitability. These were all long-term developments, beginning at the end of the Middle Ages if not before. By 1800 these factors were all in place, making European supremacy after 1815 inevitable. Alternative hypotheses—counterfactuals—do not come into this history, though according to Ferguson elsewhere they would undermine the determinism of this narrative.27

If even Ferguson implicitly concedes that counterfactuals are inappropriate when we are dealing with large-scale historical change, then how possible is it to devise procedures that avoid the many pitfalls that threaten the unwary practitioner who tries to apply them on a narrower front, from politically driven wishful thinking to implausible chains of consequences? Geoffrey Parker and Philip Tetlock have issued some basic rules of counterfactual history to the contributors to their volume of counterfactual essays, Unmaking the West. There has to be a “minimal-rewrite” rule, they say, along the lines suggested by Ferguson, to avoid arbitrariness. The ceteris paribus rule is another, or in other words, the rule that a counterfactual must only make one change in the causal chain and leave everything else the same as it was in reality. Thus, for example, as Aviezer Tucker points out, the counterfactual speculations of Jonathan Clark in Ferguson’s volume—had there been no “Glorious Revolution” in 1688, America would have remained part of the British Empire—are meaningless because the change in the initial condition ignores the ceteris paribus rule; that is, it is too large to make the speculation historically sustainable, for England would have needed to have been a completely different society and polity for there to have been no “Glorious Revolution” overthrowing the Catholic absolutist James II and bringing in the Protestant Dutch monarch William III.28

A self-denying ordinance whereby counterfactual scenarios do not extrapolate themselves too far into the future is the second principle suggested by Parker and Tetlock, and in doing so they add to Clark’s destructive point about the implausibility of very long-term scenarios by noting that “the deeper authors try to see into the futures of their counterfactual worlds, the frailer their connecting principles become.”²⁹ Finally, in order to get round Carr’s charge, exemplified by so many contributions to other collections of counterfactual essays, that such speculations are hopelessly self-serving, Parker and Tetlock asked their contributors to be explicit about their perspective and allow themselves to be self-critical, as all historians should be. The acid test of essays written in this frame of mind is surely whether they contribute anything to historical knowledge or understanding. Surveying the contributions to their volume at the end, the editors make the point that “close-call” counterfactuals that show how a crucial decision or event might easily have gone another way can contribute usefully to a sense of the possibilities open to contemporaries. But this is not quite the point. The real interest of close-call counterfactuals is in pointing up the limited nature of such possibilities and the constraints within which they operated.

In writing about the Nazi seizure of power, for example, I argued in my book The Coming of the Third Reich that the choices facing Germany in 1933 were essentially between an authoritarian military regime and a Nazi dictatorship. It is certainly true, as Parker and Tetlock say, taking their cue from empirical studies by Henry Turner, that the Nazis were in decline, in voting support, financial resources, and internal cohesion, at the end of 1932 and the beginning of 1933, and that if you follow the many twists and turns of the complex political negotiations that brought Hitler into the chancellorship on 30 January 1933, chance played a role. But we do not need a counterfactual to tell us that, and indeed Parker and Tetlock do not deploy one. Moreover, to focus exclusively on the small number of decision makers around President Hindenburg, whose maneuverings made Hitler chancellor, while inevitably underlining the impact of chance factors such as “personal affinities and aversions, injured feelings, soured friendships, and desire for revenge,”30 is to bracket out the larger picture of rapidly growing, mass, murderous Nazi violence on the streets, combined with the complete ungovernability of the Reichstag, which dissolved in chaos, with Communists and Nazis chanting slogans at each other across the chamber and joining together only to vote down any measure proposed by the government. This was an intolerable situation that could not be sustained, least of all in a massive economic crisis that left well over a third of the population unemployed and bankruptcies and bank failures across the land. The political crisis could only be resolved by bringing the Nazis into government in one form or another. The key factor was not so much the political decline of the Nazis from the elections of July 1932 to the elections of November 1932, though this, fatally, helped persuade the conservatives around Papen and Hindenburg that they could control them, but the Nazi storm troopers’ escalating use of violence, which, the army leader General Schleicher feared, was pushing the country into civil war.³¹

The whole history of Germany from the fall of the last democratic government in 1930 had narrowed the options down and eliminated the possibility of the return of democracy. This at least is what key contemporaries thought. One might speculate that more skillful maneuvering by men like General Schleicher might have brought a representative of the army to power in Germany instead of the Nazis, though the Nazis would either have had to be put down by force—a difficult proposition, given the fact that the storm troopers were far more numerous than the army’s own soldiers—or still brought into some kind of coalition. Even had this happened, it is reasonable to suppose that, given the policies of the army officer corps and its commitment to reversing the Treaty of Versailles, rearming the nation, remilitarizing the Rhineland, invading Austria and Czechoslovakia, and in general preparing to reverse the defeat of 1918, would have at least brought a general European war much closer, especially since they would have needed to keep the Nazis onside. In practice, indeed, a coalition government was formed, bringing together the Nazis, the conservatives, and the army. Hitler comprehensively outmaneuvered his coalition partners and established a one-party dictatorship in a few months. The dynamism of the Nazis, their violence on the streets, the ruthless ambition of their leader—all these speak in favor of such a hypothesis.32

The way in which these scenarios contribute to historical understanding is by making it clear that the immediate restoration of Weimar democracy and the maintenance of the international status quo in Europe was not an option in 1932–33. The point is to argue not for the freedom of maneuver that German politicians had in 1933, but for the lack of it. Similarly, the experience of the uprising of the Austrian Socialists against the dictatorship of Dollfuss in February 1934, gunned down by the army in a few days, can show that even had the Communists and Social Democrats been united in Germany, they would not have been able to withstand the violence of the German armed forces, or the Nazis, let alone both. The aim here is to undermine any kind of wishful thinking on these counts, or, in other words, to reinforce the depiction of actual historical reality in the narrative of events in Germany in 1932–33, not primarily to emphasize the openness of the future at this crucial point; it was open only in the most limited sense. Such speculations fit in well enough with Parker and Tetlock’s criteria of “minimal rewrite” or of “close-call” or short-term speculation on possible alternatives, though not with the purposes for which they think they should be undertaken, but in the end, they scarcely fit the description of “counterfactual” in the way it is usually applied.

In another area, Parker and Tetlock criticize the claim that Kaiser Wilhelm II was unaware of the precariousness of Bismarck’s achievement of German unity and saw the process as historically preordained. “The only way to argue [t]his case,” they claim, “is by making a host of counterfactual assumptions about how easy or difficult it is to reroute European history in the mid-nineteenth century.” But this is not true at all; in fact the case can easily enough be made by quoting Bismarck, the Kaiser, and their contemporaries on the topic. The fact is that Bismarck thought that Germany’s position in Europe and the world was precarious, the Kaiser did not, and both modeled their behavior as statesmen on these beliefs. A simple factual narrative of German unification, so long as it doesn’t fall victim to the inevitabilism of the Borussian school of historians, indeed a simple factual narrative of the Battle of Sadowa, should be enough without any explicit counterfactual speculations to demonstrate the role of chance in the process, but then of course the Kaiser did indeed think victory at Sadowa was preordained, as it was in Sedan. The crucial point here is what the two men thought; the historian does not need to decide whether they were right in order to understand their actions, though the Kaiser did indeed in the end risk everything Bismarck had achieved and brought his empire to ruin in the First World War. To reach such a conclusion is not to claim one knows better than contemporaries what they should have done; it’s simply a matter of historical observation.

Finally, Parker and Tetlock argue that what they call reversionary counterfactualism, emphasizing not short-term chance events but longer-term large-scale processes, can be helpful in explaining, for instance, how and why things such as industrialization or changes in the balance of power between nations and states happened the way they did; that is, the counterfactual speculation leads back in the end to the conclusion that things would have been the same even had events and processes occurred rather differently. It is possible to imagine for instance, Hitler winning the war against the Soviet Union only to be destroyed by the American atom bomb, so Germany lost the war in the end anyway. The argument here is that such thought-experiments help weigh up the importance of different, impersonal factors in historical outcomes. But the problem here is that one can just as easily do this without such thought-experiments. They also seem to violate Parker and Tetlock’s principle of avoiding long-term rewrites of history. And anyway, Nazi Germany was defeated without the intervention of the atom bomb, so what is the point of imagining that it might have needed the bomb to achieve this?

Parker and Tetlock’s distinction between short-term and long-term counterfactualism is very similar to that made by Allan Megill between what he calls restrained counterfactual history, which “involves an explicit canvassing of alternative possibilities that existed in a real past,” and exuberant counterfactual history, or “virtual history,” which “deals in past historical outcomes that never in fact came to be.”33 Restrained counterfactual history is restrained because it starts out from an actual event and looks back in time, moving “from observed effect to hypothesized cause.” Thus, for example, John Adamson’s essay “England without Cromwell: What If Charles I Had Avoided the Civil War?,” in Ferguson’s Virtual History, looks at a range of possible alternatives to what actually happened (namely, that Charles I fought and lost a civil war between 1640 and 1649 and was executed at the end, being replaced by Oliver Cromwell, the leading general on the opposing side, as “Lord Protector”), and devotes the bulk of his essay to explaining why these alternatives did not come to pass. This kind of counterfactual history is epistemologically relatively tenable, according to Megill, because it starts with known evidence and uses counterfactual speculation to argue about why things ended as they did and not differently. Indeed, Megill argues that all causal explanations in history must be counterfactual in this sense, because all causal explanations involve not only explaining why things turned out the way we know they did, but also why they did not turn out some other way.³⁴ For example, in explaining why Hitler came to power in 1933, we are also explaining why the German army did not come to power, why the Left failed to resist the Nazis, and why democracy was not restored.

Megill argues that historians need counterfactuals because they cannot, as a matter of principle, adduce regularities or constant conjunctions of repeated structures or larger influences as causes. We might say for example that imperialism caused the First World War, but we cannot say that imperialism causes all wars. In practice, saying that imperialism caused the First World War is a perfectly reasonable statement, even if it operates at a very high level of generalization (it does not of course say why the war broke out when it did, or why some countries fought on one side and some on the other). And of course it has an implied counterfactual, or, in this sense, an alternative vision of what might have happened: if there had been no imperialism, there would have been no First World War. But this is not the main point of the argument: it is not necessary in any way to discuss what would have happened in 1914 had there been no imperialism, so there is no need for counterfactual speculation. Historians, then, only need counterfactuals at a much lower level of generality, following Parker and Tetlock’s principle of the minimal rewrite. Exuberant counterfactual history, argues Megill, operates not from an actual event back to a hypothesized cause, but from an invisible or hypothesized cause to an event or sequence of events that never actually took place. In other words, “speculations concerning virtual history are far more deeply permeated by under-supported assumptions about the real nature of the world than is the case when the normal canons of historical method operate.” These speculations amount in effect to a theory that drives the counterfactual speculations, because they lack any direct evidence to support them.35 Thus the argument here becomes metaphysical rather than historical.

Aviezer Tucker has put these same points in a slightly different way. Tucker points out that “every counterfactual has a ceteris paribus clause: the historian assumes that the historical reality remained constant, except for the examined factors.”³⁶ So, for example, if we thought about what might have happened had Hitler been killed in the First World War (a far from unlikely contingency), we would assume everything else remained the same, in order to make the speculation meaningful: Germany would lose the war, and the Nazi Party would still be created, only with a different leader; then we imagine what difference Hitler’s absence might have made to its policies, its electoral prospects, and so on. Counterfactuals in other words have to be consistent with other things we know about the subject of the speculation: we change one thing but leave everything else the same. It would make little sense to speculate on the course events might have taken if the Nazi Party had been philosemitic because we know that far-right parties in Germany immediately after the First World War were antisemitic. The implication here is that to be meaningful at all, counterfactuals have to posit small changes not large ones. If we shake up the kaleidoscope of history by moving one piece then we can think creatively about the effect this might have on all the other pieces; but if we shake them all up, we can’t make any generalizations at all. Nevertheless, who is to say that had Hitler been killed in the war, the Nazi Party would ever have come into existence after it, as opposed to the very ordinary fringe movement, the German Workers’ Party, from which it emerged? The ceteris paribus rule leads to unconvincing decisions to ignore the possible effects of a changed starting point on later developments that the counterfactualist does not want to consider. Other things won’t necessarily remain the same, in other words, and even this principle still leaves out the possibility of unforeseen chances.37

As Johannes Bulhof points out, many if not most historical investigations contain what he calls modal claims, by which he means that if a had not happened, then b would not have happened either. Counterfactuals in this sense permeate historiography because historians are concerned to explain why things happened, and this necessarily involves explaining why other things, alternatives, did not happen. But this is a truism. The crucial step that turns an absent but plausible alternative into a counterfactual is the extrapolation from it of further unrealized but plausible consequences. And in practice these are not central to the task of explaining why things happened as when and why they did. Thus, for example, Daniel Goldhagen’s book Hitler’s Willing Executioners argues that non-Germans such as the Ukrainian auxiliaries who manned the death camps of the Reinhard Action “were not essential to the perpetration of the Holocaust,” and backs this with a counterfactual: “To be sure, had the Germans not found European (especially East European) helpers, then the Holocaust would have unfolded somewhat differently, and the Germans would likely not have succeeded in killing as many Jews.” But this speculation—expressed in vague terms, as a possibility or at most a probability—is not actually necessary to the explanation, which is contained in Goldhagen’s evidence-based statement that non-Germans were not essential to the Holocaust because “they did not supply the drive and initiative that pushed it forward.” Naturally this supposes another counterfactual, namely, that if the Germans had not acted as they did, there would have been no Holocaust. But this is completely redundant as a statement, because we know from an enormous mass of evidence that the Germans did indeed originate, initiate, and implement the Holocaust and that Ukrainians and others were only auxiliaries.38 All causal statements may indeed have implied alternatives, but it is not necessary to consider these alternatives and their implications in order to drive the causal explanation forward.

Bunzl argues that “the problem of the basis for deeming counterfactual judgments to be reliable is a function of the reliability of our claims on which we base the counterfactual judgments. If we base a counterfactual judgment on a causal claim, then the question is, how reliable is that causal claim?”³⁹ Thus, for example, it seems at first sight to be a reasonable statement that had partisans in Nazi-occupied Eastern Europe had nuclear weapons at their disposal they would have defeated Hitler’s army, but the background condition here is implausible because if they had had nuclear weapons then other, richer, better resourced organizations would surely have got there before them, including the Nazis, but above all the Americans, so the entire background situation would have been entirely different. Thus, as Bunzl concludes, “a counterfactual inference is only as good as the assumptions that one makes about the background conditions.”40 Counterfactuals, however, can involve not only extrapolating forward in time from an alternative event to the one that actually happened, but also extrapolating back in time to see whether the same event would have occurred under different conditions or with changes in the antecedents. Counterfactuals in this sense are compatible with determinism because “some of the most interesting historical questions have to do with probing the stability of an outcome under a range of variations from the way things were.”⁴¹

But this is only true, as many critics of “exuberant” or “long-range” counterfactualism have pointed out, if we stick as closely as possible to what was known, and use counterfactual speculation to understand why things turned out as they did. Understanding what the options and possibilities were helps us penetrate to the core of why one and only one became reality. For example, it helps understand the motives and purposes of the men who tried to kill Hitler on 20 July 1944 if we think about what might have happened had they succeeded. In practice, they enjoyed only very limited support among the armed forces, and the death of Hitler would probably have unleashed civil war as the SS and other fanatical Nazi troops fought to put them down and exact revenge. The plotters more or less knew this, and their final letters indicate they had moved on from thinking they could take over Germany and make peace with the Allies (unrealistic in the extreme, given the Allied policy of unconditional surrender) to rescuing at least an element of German honor by an act they knew to be in all probability self-sacrificial. Given their hostility to parliamentarism and their belief in social and political inequality, they were not likely to follow Hitler’s death by trying to create a new, democratic Germany. Speculations such as these help identify what their real options were in July 1944, not least because they are based closely on evidence of what the plotters thought these options were.

It is altogether a different matter to extrapolate from this, as C. J. Sansom does, to a scenario in which the war on the Eastern Front in Europe continues as a result of a German victory over Britain, and a military plot to seize power succeeds in 1952 because an unwinnable war has continued for eight more years with massive loss of life, and—crucially—Hitler has died, freeing the military from their oath of allegiance to him.42 There are plenty of risky conditionals here. Given the extent to which the Third Reich was outnumbered in human resources by the Soviet Union and outpaced by its production of arms, ammunition, and equipment by 1944, it is not at all certain that it could have continued fighting for a further eight years, and improbable that—as others have speculated—that the two powers were sufficiently evenly matched to fight each other to a standstill or a stalemate.⁴³ Still, the death or incapacity of Hitler would indeed, as many authors, from Philip K. Dick onward, have suggested, surely have unleashed a power struggle within the Nazi elite: and such was the centrality of Hitler to the whole Nazi system that it would have been unable to continue for much longer anyway, whether or not the military staged another revolt. German war graves commonly honored the dead as “fallen for Führer and Fatherland,” and no other Nazi leader, neither Goering, nor Goebbels, wielded the same charisma as Hitler.

Historical explanation generally involves a concept of historical necessity—in other words, necessary causes—in which converging causal chains lead to a certain type of result without determining the particular form it takes. Thus, for example, we may say that the initial condition of Hitler’s intention to wage a European war to establish German hegemony in Europe provided a necessary cause of the Second World War, but does not in itself explain why that war broke out in September 1939. For that to happen, other causal chains had to be set in motion—for example, the appeasement policy of Britain and France in the mid-1930s and its reversal in 1939, the successes of German rearmament, and so on; and there were also contingent events that influenced the outcome, such as the illness that made Hitler accelerate the drive to war because he was afraid he might die before being able to put into effect his original intention of starting a war in 1942. In this kind of analysis, according to Tucker, counterfactuals might be useful in understanding just how contingent a given event might have been.44 Thus we cannot and should not imagine that Hitler had no intention at all of waging a general European war—that was the mistake the appeasers made—nor does it make sense to imagine Hitler would not have marched into Prague in March 1939, which was the event that brought appeasement to an end. But we can imagine what might have happened had the generals succeeded in their tentative plot to overthrow Hitler in 1938 because they did not think Germany was ready for war, or had Neville Chamberlain not secured the Munich agreement that made them abandon their plans, or had Hitler not decided to accelerate his external and internal aggression in 1937–38, and indeed in examining these events, we implicitly do engage in such speculations because these were contingencies that might easily have ended in different outcomes.

In the counterfactual scenarios that I have examined in detail, what is most striking perhaps are the radical divergences of opinion between the counterfactualists on the same topic, depending partly on their political motives, partly on the contemporary context within which they are writing. If Britain had signed a separate peace with Nazi Germany in 1940 or 1941, would Hitler have beaten the Soviet Union in the war on the Eastern Front, or would the two totalitarian powers have fought each other to a stalemate? Or, as Andrew Roberts speculated in 2001, contradicting his earlier views on the topic, would Stalin have conquered the whole of Europe, with horrendous consequences?⁴⁵ Would British institutions including the empire have been preserved, or would Hitler have imposed his will gradually on the British, Nazifying their institutions, forcing them to surrender British Jews to be taken to Auschwitz, and dismantling their imperial possessions one by one? If Nazi Germany had occupied Britain would there have been mass collaboration or mass resistance? Would the Duke of Windsor have become a pro-Nazi puppet king or not? Would a collaborationist Number 10 Downing Street have been occupied by Lloyd George, Sir Oswald Mosley, Lord Halifax, Sir Samuel Hoare, or Rab Butler (Holger Herwig hedges his bets by having a collaborationist government under Edward VIII led by Halifax, Hoare, and Lloyd George, all three together)?46 What, in any case, do these speculations tell us about the state of British and European politics in the 1930s and 1940s that we did not know already?

Such considerations apply to many if not most other counterfactuals as well. Surveying the contributions to their own collection of counterfactual essays, Parker and Tetlock point out that for Jack Goldstone, discussing English history in the late seventeenth century, “the exact moment of William III’s death possesses great significance because he believes it would have unleashed a totally different and irreversible course for British, European, and even global history,” while another of them, criticizing Goldstone’s long-range extrapolation of this singular event, “believes it scarcely matters because William would have been succeeded first, by his wife Mary and, if she had predeceased William, by her sister Anne (as happened in 1702).”⁴⁷ The issue at hand here is one of the most oft-addressed in history counterfactuals, covered already in very different ways by Chesterton, Fraser, Russell, and many others: What would have happened had England been a Catholic country rather than a Protestant one? While Fraser and Chesterton, as Catholic historians, both think things would have been better (though as we have seen, Fraser concedes that they might not), Goldstone takes a far gloomier view. If William of Orange’s invasion of England in 1688 had failed, he says, there would have been no scientific revolution, no parliamentary constitution, no British Empire, no modern world. As Carla Gardina Pestana has pointed out, this would have required James II, whom William in reality ousted in 1688, to have reached “a level of political sagacity rarely reached by any mere mortal, much less any Stuart king.”48 The power of James II to repress opposition in the country, which was massive by that time, would have had to be far greater than it actually was, while Protestantism would have needed to be far weaker. The English would not have accepted the status of a satellite of Louis XIV’s France. The civil wars of a few decades before might well have begun all over again.

In terms of methodology, Pestana is “uncomfortable with the way in which Goldstone privileges certain kinds of causal explanations in his counterfactual,” namely, the actions, beliefs, and character of great men. Similarly, Goldstone has James II destroying the scientific revolution by getting rid of Isaac Newton; yet of course we know that scientific breakthroughs in the seventeenth century involved far wider circles of men and a far more widespread field of knowledge. Goldstone might adhere in terms of events to the minimal-rewrite rule, which, as Pestana comments, is often put into effect “by finding someone to kill,”⁴⁹ but then has to ignore the ceteris paribus rule (by making James II an effective monarch, for example) in order to get his counterfactual to work. Moreover, one might point out in any case that some European countries were quite content to have a monarch of one religious denomination ruling over a population with another, and that there was no necessary opposition between Catholic belief and scientific enterprise, as the case of the Emperor Rudolf II and his patronage of scientists and astronomers such as Tycho Brahe and Johannes Kepler showed.⁵⁰

All of this goes to show how counterfactual speculations frequently, perhaps even generally, tread on thin evidential ice, too often selecting the conditions of their starting point with insufficient care, and failing to distinguish between different levels of causation. They frequently tackle historical topics of enormous complexity, cutting the Gordian knot of interpretation by simply asserting the power of the individual actor to change things. Moreover, every counterfactual is locked into a particular historical interpretation that is almost bound in itself to be hugely controversial even before the initial change in the timeline is effected. Thus, for example, there have been many counterfactual essays about the outbreak of World War I—characteristically for the genre, they have focused either on the non-assassination of Franz Ferdinand or on the decision of the British foreign secretary not to go to war, rather than, say, on alternative outcomes of the decision-making process in the Russian, Habsburg, Serb, or other governments. Thus any counterfactual based on the outbreak of the war has to take into account the fact that there were multiple lines of causation intersecting in a whole variety of non-predictable ways. They had already almost joined up in the winter of 1912–13 over war in the Balkans; it is likely they would have joined up in a different way at a different time had Franz Ferdinand not been shot, though one can never be quite sure. The variables are just too many for it to be plausible to isolate one and reduce the whole complex mess of causation to the effects of a single change in the causative chain.51

I have argued in this book that long-range counterfactual speculations are unconvincing and unnecessary for the historian because they elide too many links in the proposed causative chain after the initial altered event. As Eric Hobsbawm puts it, all that you can say if a condition is altered in a timeline, such as, for example, Lenin being stuck in Switzerland in 1917 instead of being able to make his way to Russia, is that “‘things could have been very different’ or ‘not very different.’ And you can’t get any further, except into fiction.”⁵² Of course, counterfactualists do habitually say more, and this quickly leads them into trouble. Too often they fall prey to wishful thinking. Counterfactualists who propose sensible limitations to their thought experiments too often forget them altogether in the heady excitement of speculative imagining. Ferguson lays down some sane and workable rules about counterfactuals then breaks them in a whole variety of ways, whether by taking his speculations too far up the stream of time, or by descending into a morass of politically motivated wishful thinking. Parker and his colleagues insist on short-term counterfactuals based on minimal rewriting of the factual situation from which they start out, and then include long-term and large-scale discussions of vast imaginary social and economic changes extending over several centuries. Roberts wants to free his contributors from the tyranny of hindsight and open up history to contingency and chance but ends up including in his collection several essays that conclude that what happened was more or less inevitable anyway. Everyone in recent years has insisted on the seriousness of the counterfactual enterprise, but this doesn’t stop many of them penning witty and whimsical essays designed as much to entertain as to inform, if not more so.

Wrestling with problems such as these, the German-born American historian Holger Herwig confessed that in considering what might have happened had Hitler won the war in the east, “a bewildering choice of possible counterfactuals confused me. Which to choose?”53 In practice, the choice is the outcome of the historian’s intention, political orientation, factual knowledge, and contemporary context. It also reflects to a degree the aesthetic purposes of the author, striving to produce the most satisfying, the most coherent and, often, the most entertaining counterfactual scenario. Based on a minimal rewrite, and confined to the short run, a counterfactual can illuminate the choices that confronted individual politicians and statesmen, and the limitations that the historical context imposed on those choices. But the further away it gets from the starting point the less useful it becomes, and the more it enters the worlds of alternative reality to which increasing numbers of people are turning in search of spaces for their imaginations to roam free, unfettered by facts. Frustration at the complexities and uncertainties of modern life leads them to inhabit the Middle Earth of Tolkien’s Lord of the Rings rather than the Middle Ages of real historical time, or the rational world of Sherlock Holmes’s Victorian London rather than the conditionally complex world of the late Victorian city. Such fantasy worlds are particularly appealing in times of political and cultural anxiety, uncertainty, crisis, or disappointment.54 Far from demonstrating “how central counterfactual framing should be to serious historical research,” as Parker and Tetlock have claimed,⁵⁵ I hope I have shown that it is not central at all, but marginal. It can be useful under certain strictly limited conditions and with strictly limited purposes, but surveying what is by now a very voluminous literature with hundreds of case studies in print, the conclusion surely has to be that it is most useful, and most interesting, as a phenomenon in itself, as a part of modern and contemporary intellectual and political history, worthy of study in its own right, but of little real use in the serious study of the past. As Max Weber observed: “In every line of every historical book, judgments of possibilities are hidden and must be hidden, if the publication is to have any intellectual value.”⁵⁶

Such a conclusion has been reached by a number of students of counterfactualism from the historian Gavriel Rosenfeld, who views counterfactualism’s motives as “fundamentally presentist,”⁵⁷ to the literary scholar Benjamin Wurgaft, who points out that “the effort to ask such questions often serves as an excellent guide to the prejudices and interests of the historian asking them.”⁵⁸ This is not least because, as Nietzsche remarked, the question “what would have happened if … ?” “turns everything into a matter of irony.”⁵⁹ Counterfactuals are ironical because, ultimately, they always cast more light on the present than on the past. In the end, perhaps, the last word should be left to Walther Rathenau, himself, as we have seen, the subject of one of the more elaborate of counterfactual novels, by his Italian admirer Guido Morselli. Looking back in 1918 over the events of the First World War, and looking forward to the new world he hoped to make when peace came, Rathenau remarked: “History does not conjugate in conditionals, it speaks of what is and what was, not what would be and what would have been.”⁶⁰