Niall Ferguson Criticism - Essay

Ulrich Wengenroth (review date June 1996)

(Contemporary Literary Criticism)

SOURCE: A review of Paper and Iron, in Journal of Economic History, Vol. 56, No. 2, June, 1996, pp. 506-7.

[In the following review, Wengenroth offers a tempered assessment of Paper and Iron, concluding that final judgement of the book depends upon a reader's “political taste.”]

After a plethora of books examining the German inflation after World War I via the documents of the Reichsbank, the Berlin government, and those of the magnates of the Ruhr’s heavy industry, Niall Ferguson’s Paper & Iron represents a most welcome shift of perspective in making Hamburg and its business elite the point of departure. At the center of this elite...

(The entire section is 785 words.)

Ben Lieberman (review date Fall 1996)

(Contemporary Literary Criticism)

SOURCE: A review of Paper and Iron, in The Historian, Vol. 59, No. 1, Fall, 1996, pp. 186-7.

[In the following review, Lieberman provides a generally positive assessment of Paper and Iron.]

Entering into a field of history crowded with recent studies, including the authoritative analysis of Gerald D. Feldman, Niall Ferguson makes a valuable contribution to the historiography of German inflation by recasting the history of the inflation of the early 1920s as a study of long-term structural weaknesses of the German state together with the interplay between local business élites and the national government. In contrast to most studies of inflation, which...

(The entire section is 492 words.)

J. Adam Tooze (review date October 1996)

(Contemporary Literary Criticism)

SOURCE: A review of Paper and Iron, in Social History, Vol. 21, No. 3, October, 1996, pp. 386-7.

[In the following review, Tooze analyzes the methodology of Paper and Iron, which he describes as “unconvincing and distasteful,” though the book's themes are “of considerable interest.”]

In recent years the Oxford historian, Niall Ferguson, has made a name for himself as a right-wing pundit in the pages of the British press. Paper and Iron is Ferguson’s first scholarly monograph and, true to form, it is a highly ambitious, politically charged and intellectually provocative work.

Ferguson’s topic is the history of inflation...

(The entire section is 778 words.)

David Caute (review date 26 April 1997)

(Contemporary Literary Criticism)

SOURCE: “Some Revised Versions,” in The Spectator, April 26, 1997, pp. 34-5.

[In the following review, Caute offers a negative evaluation of Virtual History.]

This is another novelty item in the post-modernist pavilion of ‘what if?’ history. Virtual History begins with Niall Ferguson’s long, heavyweight and impressively erudite bombing raid on ‘determinist’ theories of history. He may be justified in putting down previous volumes in this genre, including If I Had Been … Ten Historical Fantasies (1979) and For Want of a Horse (1984), but the majority of the contributors to his own book are still writing orthodox history spiced by...

(The entire section is 1409 words.)

Conrad Russell (review date 6 June 1997)

(Contemporary Literary Criticism)

SOURCE: “Castles in the Air,” in New Statesman, June 6, 1997, p. 45.

[In the following review, Russell offers a tempered analysis of Virtual History.]

Historians are less impressive at handling conceptual questions than at assembling empirical data. Like bad builders, we erect edifices of factual bricks on soft, sandy soil. When, sapped by a generation of further research, they fall down, another historian rejects some bricks, makes a few new ones and erects an edifice with a different design, but on an equally sandy soil.

Perhaps this is not a particular weakness of historians. The conceptual eccentricities of Plato or Marx, to name only two,...

(The entire section is 780 words.)

The Economist (review date 14 November 1998)

(Contemporary Literary Criticism)

The Economist (review date 14 November 1998)

SOURCE: “Avoiding Hitler,” in The Economist, November 14, 1998, p. 5.

[In the following review, the critic discusses Ferguson's historical argument in The Pity of War.]

For Europe at least, the first world war was by many accounts the defining event of this century. But the very magnitude of the catastrophe it represented for liberal civilisation has tended to paralyse historical imagination, fixing the study of it in patriotic apology or moral lament—still two of the dominant modes of writing about the Great War. At one massive stroke, Niall Ferguson has transformed this dismal intellectual...

(The entire section is 853 words.)

Natasha Fairweather (review date 20 November 1998)

(Contemporary Literary Criticism)

SOURCE: “A Gilded Brand,” in New Statesman, November 20, 1998, pp. 46-7.

[In the following review, Fairweather offers a positive assessment of The World's Banker.]

Superlatives attach themselves to the Rothschild family like burrs. They are the greatest, richest, most enduringly successful banking dynasty the world has ever seen. From their five European power bases, the family was able to influence world politics, during the 19th century in particular, to a degree unrivalled since David was king of the Jews. Their palaces, gardens, art and antique collections, vineyards and stud farms are the stuff of legends. And the family name, gilded with as much mythology as...

(The entire section is 1141 words.)

Aviezer Tucker (review date 1999)

(Contemporary Literary Criticism)

SOURCE: “Historiographical Counterfactuals and Historical Contingency,” in History and Theory, Vol. 38, No. 2, 1999, pp. 264-76.

[In the following review, Tucker examines the methodology and merits of “alternative history” and offers an analysis of Virtual History.]

“Alternative history” is an established genre of literature that is sometimes classified as science fiction. Hundreds of books explore what would have happened had, for example, Hitler never been born, or won the war. Recently, grassroots non-academic interest in this genre has surged. A yearly “Sidewise Award” is given for the best alternative history book and short story. An internet site, “Uchrania: The Alternative History List,” is devoted to a comprehensive annotated bibliography of works of alternative history in all genres in all languages. Thus far, most academic historians have pretended to ignore this genre. Virtual History is significant in contributing to breaking the academic taboo on explicit discussions of historiographical counterfactuals.1 Still, what are historiographical counterfactuals good for, beyond an entertaining exercise of our imaginative faculties? A closely related question is: which criteria are relevant for evaluating the quality of historiographical counterfactuals?

In literature, the genre of alternative history has been used for a wide spectrum of purposes. P. K. Dick used an alternative world, in which the Axis Powers won the Second World War and divided the USA between Germany and Japan, in order to explore the nature of human decency (associated with the Americanophile Japanese occupation in San Francisco) and fanatic totalitarianism (associated with the German occupation of the East coast). Dick used this alternative historical reality to emphasize through his protagonists the banality of good and the chaotic nature of human life, dominated by greater unfathomable forces symbolized by the I-Ching.2 Michael Morecook constructed alternative radical political utopias by introducing alternative revisions of history at the turn of the twentieth century to create histories in which imperialism, racism, and European domination were successfully challenged by indigenous Asian or African radical political forces. Morecook then followed the development of these utopias as they turned into distopias. The purpose of Morecook’s alternative historiographies is the examination of radical political utopias from the perspective of a well-wishing, naive, fin-de-siècle British romantic officer turned pilot of airships.3 Robert Harris imagined how Europe would have been during the 1960s had the Nazis won the War, emptied it of its Jews, enslaved the Slavs, and settled Germans in an ethnically-cleansed Eastern Europe. He speculated that the Nazi regime would have moderated by the 1960s, and attempted a detente with the United States. Against this fascinating background, Harris developed a detective plot: a murder investigation leads a Berlin police detective to discover an elaborate conspiracy by the Nazi regime to hide the existence of the Holocaust by silencing its functionaries who participated in the Wansee Conference that plotted the “Final Solution.”4 Some of the appeal of Harris’s bestseller, as well as many other novels and short stories that describe a world where the Nazis won the war, may be ascribed to an aesthetic fascination with apocalyptic landscapes, with consistent realistic depictions of a horrendous alternative universe, like a Bosch painting. The theme of a Europe/world dominated by Germans by reversing the results of one of the World Wars is perhaps the most popular among writers in the genre of alternative history, followed by a victory of the Confederacy in the American Civil War. A German imperial- or Nazi-dominated world opens endless opportunities for all the literary tropes with the exception of comedy.

Yitzhak Laor wrote an alternative history novel that Israeli critics hailed as one of the best in the Hebrew language, offering a penetrating insight into the Israeli experience and character by posing a counterfactual in which an army major named Kishon (a parody on the name of the famous author of feuilletons) dies during his siesta under a eucalyptus tree, thereby causing a chain of events in the desert supply camp he commands that prevents the outbreak of the 1967 Six Days War.5

All these uses of the genre of alternative history are legitimate and fascinating. They are also distant from the kind of issues that concern historians, social scientists, and philosophers of historiography and the social sciences in the normal course of their scholarly work. This is not to say that history and social science eschew counterfactuals entirely. Historians and social scientists use counterfactuals regularly though implicitly when they assign necessary causes, and sometimes in assigning degrees of importance to causes. The assignment of necessary causes assumes that had the causes not occurred, neither would the effects. One possible way to support assigning degrees of importance to causes is by posing counterfactuals that isolate the effects of each cause separately, and then comparing them with the actual historical result. The greater the difference, the more important the cause. Still, the vast literature on this topic indicates that though historians assign causes very often, only rarely do they claim to have discovered the necessary causes of an event. Though the use of counterfactuals is one of the ways by which historians may support assigning degrees of significance to causes, the use of comparison situations and criteria of relevance is more common. More ambitious neo-positivist attempts to explicate causation and natural laws by using counterfactuals have failed.6

In the introduction to the book under review [Virtual History], Niall Ferguson suggests that historical counterfactuals are useful in refuting what he calls “historical determinism.” Ferguson rejects as “too elastic” (53) the standard use of “determinism” as the metaphysical doctrine that holds that all events have causes without which they would not have occurred. The “anti-determinism” Ferguson upholds does not relate to the indeterminism of quantum physics, where either all the causes are known but more than one kind of outcome can follow, or events have no apparent causes. Much like Popper’s use of “historicism,” Ferguson uses “historical determinism” as a conceptual blanket that covers several otherwise independent historiographical doctrines and methods of which he disapproves:

1. Teleological schemes of history as leading necessarily to certain ends according to a plan.

2. Scientistic theories about history that claim to have discovered linear and deterministic laws of historical development.

3. Historiographical theories that privilege certain large impersonal factors as determining all other historical processes (such as Braudel’s emphasis on geographical factors, or Marx’s on the relations of production, or game theory’s on rationality).

4. “Narrative determinism,” the literary interpretation of history as a text implies in Ferguson’s opinion its predictability, as the end of a trope narrative, such as a comedy or tragedy, is predictable.

Ferguson conflates these four types of doctrines, theories, and methods under the title “historical determinism.” The first doctrine is metaphysical: it claims that history (past and present) is predetermined. This metaphysical doctrine does not imply that predetermined history is knowable, though the monotheistic religions and some substantive philosophers of history like Vico, Hegel, and Marx claimed indeed that the predetermined course of universal history is knowable at least in its broad outlines through faith or through their various methods. Unless the advocates of a teleological plan of history set a precise date in the near future for the end of the world, there is nothing that historians or social scientists can do to refute their metaphysics. Any historical development can be incorporated into a loose teleological plan of history. For example, some rabbis have been able to integrate the Holocaust into their teleological historical scheme, holding that God’s exact plans are beyond our grasp though each step brings us nearer to the coming of the Messiah; the Holocaust was followed by the establishment of the state of Israel, a step in the correct messianic direction. Similarly, orthodox Marxists are able to integrate the fall of the Soviet Empire into their teleological scheme, claiming that we are still witnessing only the early stages of global capitalism, and the revolution is yet to come. Counterfactuals may participate in casting doubt on this metaphysical doctrine only on its own metaphysical level; this requires adherence to modal realism which holds that numerous possible worlds or histories exist,7 in which some of them do not fulfill any particular teleological scheme of history. Though this metaphysical approach is developed masterfully in Gregory Benford’s excellent novel Timescape,8 historians do not care for it, and the contributors to Virtual History do not address it.

The second and third meanings of Ferguson’s “historical determinism” are types of historiographical theories. It is hardly necessary in this day and age to flog again the dead horse of scientistic linear theories of historiography, nor is it clear how counterfactuals can do a better job of refuting these theories than plain historical facts. Historiographical theories that privilege certain factors do not necessarily deny the historical significance of factors that are not represented in their theories. They must claim merely that some variables tend to be more independent than others in history. Mandelbaum noted that a deterministic science is possible only in relatively closed systems. Unlike astronomical systems, for example, historical processes are not sufficiently isolated for deterministic description. Factors external to processes studied by historians may intervene and affect the outcome. These historical accidents do not imply metaphysical indeterminism, merely the intersection of two or more causal chains, each of which may be deterministic. Accident is relative to a system, not indeterminist.9 It would take fairly extreme and self-destructive Marxist or Braudelian historians to claim that their historiographical theory is sufficiently comprehensive to account for the whole of history. Quite the reverse: faithful Marxist and Braudelian historians would rely on coincidental intervening variables and auxiliary theories to explain away the obvious deviation of history from what their favored theory would lead us to expect.

Though historiographical and social theories that privilege certain independent variables often outline a more or less linear relation between independent and dependent variables (in the sense that had their relations been modeled on a graph between two axes, known as Cartesian space, that graph would have been approximately a straight line), they are not necessarily so. For example, a theory may claim that history is dominated by the interactions between great heroes; a revised Marxism may claim that the history of class struggle is determined by the effects of chaotic processes of technological innovation. Historiographical counterfactuals as such are not a challenge to privileging theories as such, though particular counterfactuals may cast doubt on particular theories. For example, Hawthorn was successful in questioning Braudel’s favoring of the long durée by demonstrating the role of political agency in the etiology of demographic events such as the Black Plague and the pre-modern French birth rate.10

Finally, if historians choose to present their conclusions in a narrative form, it does not imply what Ferguson calls “narrative determinism.” Their study of the events usually precedes their choice of narrative trope. If a historian presents, say, the history of Czechoslovakia as tragic, the historian does not predetermine this history, but merely represents the information gathered from the primary sources in a trope that seems fitting. The historiographical use of what Danto called “narrative sentences” does not precede the historical events but conceptualizes and follows after them. Most significantly, Ferguson ignores a basic rule of literary criticism: a predictable narrative is bad literature.

In a recent very useful article, Yemima Ben-Menahem has suggested understanding historical necessity and contingency in terms of degrees of sensitivity to initial conditions. Ben-Menahem claims that “historical necessity” describes circumstances in which several independent causal chains may lead to a certain type of result. In this respect, whether or not specific initial conditions cause a chain of events that lead to a certain result, that type of result will come about by a different chain of events. There can be degrees of necessity according to how sensitive the type of result is to initial conditions. For example, human death is necessary within certain time limits because some chain of events will bring it about irrespective of any particular initial conditions. When events are extremely sensitive to initial conditions, they are contingent. The degree of their contingency is measured by their sensitivity to initial conditions. The most contingent events are chaotic. Chaotic events may be governed by deterministic laws, but their complexity usually prohibits prediction. Chaos theory describes stochastic behavior in deterministic systems that are sensitive to minor changes in initial conditions.11

If we accept Ben-Menahem’s understanding of historical contingency, the study of historiographical counterfactuals may be useful in ascertaining how contingent historical events were. If there is a sufficiently large number of such studies, it may become possible to make some tentative generalizations about contingency in history. Ferguson approaches this conclusion as well, recognizing that counterfactuals are important for understanding the significance of complexity and chaos theory for historiography; chaos theory reconciles causation with contingency by linking causally unpredictable outcomes to initial conditions (76–79). Still, Ferguson does not express interest in less extreme cases of historical contingency. I think that this is the one undeniable significance of discussing historiographical counterfactuals for historians.

Ben-Menahem speculates that “where large ensembles of systems are concerned, we might find a large degree of overall necessity due to the law of large numbers, but low degree of necessity at the level of individual events. Evidently, one has to consider many more possibilities than the simple paradigmatic ones.”12 Specifically, in relation to historical events, she does not commit herself: “The sensitivity of historical events to initial conditions may or may not be as radical as that of chaotic phenomena, but it can certainly be very significant. …”13

The evaluation of the degree of contingency of historical events—how complex and chaotic historical processes are—is a controversial issue. The intuitive evaluation of the complexity and contingency of a system may lead easily to mistakes. For example, John Stuart Mill proposed a theory of the history of science according to which the pace of development of the sciences is commensurable with the degree of complexity of their subject matter. Mill thought that since cosmology is the simplest system, Newtonian physics was the first deterministic science. The weather system and the system of tides appeared to Mill more complex and therefore they were the subjects of statistical sciences that account only for the “larger causes” but not for the “perturbations.” Since human nature and society were the most complex in Mill’s opinion, he predicted that sociology would be the last science.14 Oddly enough, though Mill spent most of his life in the British Isles, he grossly underestimated the complexity of the weather. Edward Lorenz developed chaos theory in the early 1960s as an attempt to model meteorology mathematically.

If Mill is right, historiography should prove to be an even more chaotic science than meteorology. Still, there have been radical disagreements both on how contingent or chaotic history is and on which historical subsystems are more contingent than others. The old arguments about the significance of the individual or the leader in history foreshadowed the current argument about contingency and chaos in history because in the former the issue revolved around how linear the effect is of some individuals in history.

Jon Elster has suggested that a theoretical background participates in the construction of historiographical counterfactuals both by connecting the hypothetical antecedent with the consequent, and by deciding on the legitimacy of the antecedent in the given historical context in the first place.15 Every counterfactual has a ceteris paribus clause: the historian assumes that the historical reality remained constant, except for the examined factor. For example, we may ask what the effect would have been of the death of Hitler in the First World War, assuming that the historical context of the First World War remained the same. Then we would need to study the early stages of the formation of the Nazi party and use a theoretical background to deduce who would have led the party in Hitler’s absence, and how different its electoral attraction, policies, and so on would have been. But Elster noted that not all counterfactual antecedents are consistent with the ceteris paribus historical context; in those cases the counterfactual is meaningless, filtered out by a theory. For example, if a historian asks what the effects would have been of a Nazi racial ideology that assumed that the Jews are the superior race, our background theories would tell us that to have a history where the Nazis liked Jews is inconsistent with most of the other relevant historical conditions in Germany, so the counterfactual is internally inconsistent (unlike, say, a history where Hitler died before he became politically active).

Elster thought that this dual role of theoretical background in constructing counterfactuals leads to an epistemic zero sum game: the better our theoretical background, the greater support we have for connecting counterfactual antecedent with consequent, but there will be fewer counterfactuals consistent with the ceteris paribus conditions. Correct though Elster probably is about most cases, there is one group of antecedents that seem to defy this epistemic zero sum game: contingent factors, especially chaotic ones. Tiny counterfactual variations in actual history, such as Hitler’s death in World War I, are not inconsistent with their contemporary historical context according to any social theory, but still would have had enormous effects on history. Thus, it seems that historical counterfactuals whose antecedents are contingent are privileged in being able both to be backed by a strong theoretical background and still be consistent with their ceteris paribus conditions. But then the theoretical background would also have to assist in identifying which counterfactuals are contingent by examining whether other chains of events would not have led to a result similar to the actual one. The extent and location of contingency in history are theoretically contested. For example, Marxists and Braudelians are generally hostile to historical contingency, upholding, for example, the insignificance of individuals. Others, like Hegel, thought that the influence of certain individuals in history is nonlinear. Other disputes revolve around which parts of society are more chaotic than others. Braudelians hold that political history (“history of events,” in their terminology) is chaotic while “history of the environment” is not. This is as hotly debated an issue as has appeared in the pages of History and Theory.16

The participants in the History and Theory debate about chaos in history had to admit that they do not really know how chaotic history is and where exactly chaos reigns, though they had strong intuitions. Reisch wrote:

Of course I cannot demonstrate that history is chaotic in the same way that we decided our input-output machine was. … The subjects of histories do not constitute isolated systems, and even if they did their histories are not repeatable. But I will stand by the claim [that history is chaotic] nonetheless, for I am only applying a technical term to highlight what is an obvious feature of history. … 17

Reisch acknowledged that non-chaotic social and economic forces help forge history, though individuals may render the result chaotic. Still, he provided no empirical evidence for evaluating which historical processes are chaotic.

McCloskey also acknowledged that it is impossible to know whether many systems are linear or not. Still McCloskey thought that history may well be chaotic and suggested a few anecdotal examples from the history of the American Civil War. But though it seems likely that the exact course of the American Civil War depended at least partially on chaotic conditions, McCloskey did not prove that something similar would not have been brought about by slightly different initial conditions.

Roth and Ryckman questioned the plausibility of Reisch and McCloskey’s premise that history is chaotic. Believers in chaotic history must assume counterfactual situations in which the initial conditions were slightly different but the outcome was radically other. In Roth and Ryckman’s opinion, since it is impossible to know which causes were necessary for particular historical events such as the Great Depression or the Civil War, it is impossible to substantiate counterfactuals. They argue that since there is no concrete proof that history is chaotic, and since there are no quantifiable chaotic models of history that use chaotic mathematical calculations, chaos is just a scientistic analogy or metaphor of little value.

In his reply to Roth and Ryckman, Reisch conceded that indeed he cannot rerun history to examine empirically how sensitive it is to initial conditions. Though there are no precise quantifiable non-linear historiographical models as in the natural sciences, Reisch believed that if such models were knowable, they would be chaotic. A completely linear history would be inconsistent with too many anecdotes. The open question is how much of history is nonlinear, where and when?18

Shermer attempted to satisfy the requirement for concrete examples of chaos in history, claiming that historians have been discussing contingency and necessity for centuries, though their use of these terms was ambiguous (as is Shermer’s which confuses them at times with planned versus unforeseen from the perspective of historical agents). He claimed that:

There is here a rich matrix of interactions between early pervasive contingencies and later local necessities, varying over time, in what I shall call the model of contingent-necessity which states that in the development of any historical sequence the role of contingencies in the construction of necessities is accentuated in the early stages and attenuated in the later. … a historical sequence is what the historian says it is.19

Shermer improved on his predecessors by suggesting where and when history is likely to be chaotic, rather than just making a sweeping generalization. Shermer’s examples of chaotic elements in history include inventions, discoveries, ideas, and paradigm shifts. The sensitivity of historical sequences to leaders and exceptional people is dependent on the stability of the social circumstances. “The straw that broke the camel’s back” is in the nature of earthquakes, avalanches, economic depressions, ecological disasters, and possibly wars, revolutions, paradigm shifts, and so on. Shermer analyzed concretely cases of mass hysteria such as witch hunting, “repressed memory syndrome,” and “alien abduction” as chaotic.

In science fiction, the evaluation of the degree of contingency in history has been related to attempts to change or “engineer” history. Asimov invented the ultimate chaotic narrative in which the replacement of a single letter in the forename of a scientist prevents nuclear war.20 Jakes imagined a future world where time travel is used by opposing groups of radical African-Americans and white supremacists to attempt to change history through assassinations, respectively, to create African domination, or to maintain slavery.21 The radical African-Americans attempt to assassinate the prophet Muhammad before he developed Islam, to prevent the defeat of the African Ashanti Empire by a Muslim army. They also attempt to assassinate Booker T. Washington before he advocated the concept of “separate but equal.” The white supremacists attempt to assassinate Benjamin Franklin to prevent the development of the abolitionist movement; and Abraham Lincoln before he declared the emancipation of the slaves. Jakes recognized only a single chaotic element in history that can be manipulated by a determined team of time travelers: individual leaders. Jakes assumed that the disappearance of a leader eliminates the movement he leads. For example, once the radical African-Americans succeed in killing Muhammad, the narrative proceeds to an America that was discovered and settled by the Ashanti who then imported European slaves. As an ironic twist of narrative, this is not bad, but historiographically it is hardly convincing. At least sometimes, leaders may be replaceable.

In my opinion, evaluating how contingent, complex, and chaotic history is can only be done empirically: it is the work of historians and social scientists. There is no philosophical superhighway to bypass the careful historiographical study of concrete historical processes and events. Instead of speculation, the only way to examine the contingency of history is to study it empirically, and attempt with the help of theoretically based counterfactuals to find out how sensitive particular historical outcomes were to initial conditions. I think that evaluating how contingent history is, and in which parts or aspects, are some of the most seminal questions that face the historiographical profession. Some counterfactuals may be obvious, if the facts are undisputed, and the theoretical background is trivial or even implicit. Then, the theories involved may even be local and probabilistic. Historians may disagree on the degree of contingency of other events, either because they assume different theories used to warrant certain counterfactuals,22 or because they examine alternative causal chains different from the one that actually took place, so the necessity that some discover in several alternative causal chains may elude other researchers who have not noticed it. At least in the second case, agreement may follow close empirical examination.23

The bivalent choice between chaos and linearity is false because between these two poles there can be various levels of contingency. The absence of mathematically precise historiographical models of chaos does not exclude the possibility of discussing historical chaos because there are hardly any alternative mathematically precise linear models of history either, though obviously there are useful “linear” theoretical models that are employed by historians and social scientists. If we accept that there are theories in the social sciences and historiography (and geology and biology) without quantification, there can be chaos there as well. The philosophical evaluation of the quality of historiographical counterfactuals should also be empirical. Instead of commencing with some metaphysical criteria for evaluating the justification of counterfactuals, philosophers of historiography should examine how historians actually attempt to justify counterfactuals, and then and only then criticize and evaluate.

From this perspective of looking for empirical studies of historical contingency, the eight historical counterfactual studies that compose Ferguson’s anthology have varying levels of significance. The most interesting studies demonstrate where, when, and how history was necessary or contingent. Other studies are no better or worse than conventional historiography, while still others are irrelevant for the kind of problems in which historians and philosophers of history are interested.

Alvin Jackson argues convincingly that had the political constellation in Westminster been slightly different in 1912, the Irish Home Rule Bill would have passed and the subsequent history of Ireland would have been different. Diane Kunz makes a strong argument for necessity by examining how different US history would have been had John F. Kennedy not been murdered. Her conclusion is that the circumstantial constraints that forced Lyndon B. Johnson to maintain and escalate the war in Vietnam would have forced Kennedy to do the same. She further claims that Kennedy would not have pushed the liberal agenda of civil rights and welfare beyond L.B.J., and most probably the support of the rich kid from Harvard for these issues would have been more lukewarm than that of the poor man from Texas. Andrew Roberts denies any chaotic element in the process that led to the outbreak of World War II. Though it is easy to analyze in hindsight what the democratic countries could have done better to curb Nazi expansionism, at the time these reasons were unknown, and there is no nonlinear gap in the causal chains that led to appeasement. Roberts notes, as many historians have before, that had Hitler not decided to avoid obliterating the British expedition force in France before it was evacuated from Dunkirk, the British Isles would have been left with little ground defense and the Germans could have attempted to launch an invasion of England almost immediately. It is difficult to assess the results of such an invasion because the German fleet would have had to control the channel for about twelve hours facing the Royal Navy and the possible use of huge amounts of mustard gas by the RAF against the invading force. Still, this was Germany’s greatest chance to win the war in Europe, and it missed it due to a chaotic decision of a single person. In another study, Jonathan Haslam examines several possible causes of the Cold War, and proves, using counterfactuals, that none of them was necessary because other causes were sufficient.24

Other counterfactual studies in this anthology have doubtful relevance for the discussion of historical contingency. Michael Burleigh attempts to sketch how Nazi Europe would have looked, drawing on the plans in the Nazi archives. There is nothing very contingent or historiographically unusual here (though the article does evoke our fascination with horror). John Adamson demonstrates in his contribution that had Charles I had better intelligence about the actual strength of the Scottish Covenanter army, he could have decided to fight and defeat it. Adamson further speculates that a Scottish defeat would have prevented the Puritan Revolution. The first conclusion appears more convincing than the second. Would not the deep rifts within English society have caused something similar to the Puritan Revolution, even had the Scots been defeated?

Jonathan Clark suggests that had there been no Glorious Revolution, America would have remained part of the British Empire. Mark Almond holds that had the Russian nomenklatura elite headed by Gorbachev not conceptualized the crisis of Communism as it did, their empire would not have collapsed. Here changes in the initial conditions (the Glorious Revolution; the worldview of the Russian Communist elite) are of sufficient magnitude to exclude a contingent interpretation. Further, they are inconsistent with their ceteris paribus conditions: for England to have no Glorious Revolution or something very similar, it would have had to be a different country. Almond’s claim that had the nomenklatura had different beliefs it would have behaved differently and Communism would have survived is logically equivalent to saying that the Soviet Empire collapsed rather than was destroyed by external factors.25 For the nomenklatura to view the world differently, late Communism would have had to be completely different, so Almond’s counterfactual is also inconsistent with its context.

When historical counterfactuals are stretched across too many causal links, they become too difficult to substantiate. If history displays some characteristics of chaotic systems, the longer a counterfactual is stretched, the more assumptions must be added, and it becomes impossible to substantiate. It is one thing to examine whether the fall of the Stuarts was chaotic. It is quite another to attempt to outline the next 300 hundred years of Stuart rule. Attempting to extrapolate historiographical counterfactuals across many links on intertwining causal chains resembles attempting to predict the weather exactly one year from today. Ferguson appears to develop some kind of personal utopia by combining most of the counterfactuals in this anthology into a single narrative: the British Empire, preferably including North America, under the absolute rule of the Stuarts, and a continental Europe incorporated into a German Empire that created, as Ferguson puts it, “a European Union.” Ferguson bemoans the loss of British world domination following the two World Wars, the ascendancy of the USA, and the economic dominance of Germany in Europe. A neutral but strong Britain ruling its Empire and forsaking Europe to the Prussians would have been more to his liking. It is tempting to challenge this eccentric utopia: it is unlikely that a British absolute monarchy would have been conducive to the kind of economic, social, political, scientific, and cultural achievements that we associate with the extraordinary success of Britain and its former colonies. It is inconceivable that Europe under the Kaiser, Krupps, economic corporatism, Prussian junkers, and the imperial officers’ corps would have resembled anything like the current European Community, but rather would have been an authoritarian, semi-feudal, antimodern, and antiliberal sort of miserable place. It is unlikely that after consolidating its power in Europe, the Prussian militaristic Empire would not set its goals on the next military prize, the British Empire, and start a naval build-up. Generally, it appears that Ferguson’s day-to-day analysis of the decisions of the European political elites and their logical possibilities precludes him from considering the deeper social, ideological, and economic interests of the historical agents that make events more necessary and less contingent. Undoubtedly, had Ferguson been in Asquith’s government he would have delayed sending the British expeditionary force to France, and would have allowed the latter to fall. But the historical reality is that most members of the Whig government and all the Tories sensed compelling reasons to go to war over Belgium and France. In the name of anti-presentism, Ferguson constructs decontextualized historical agents, isolated from larger cultural and economic contexts that precluded the kind of decision-making he would have undertaken. Ferguson suggests that had the dissenting ministers in Asquith’s government brought down his liberal government, by the time a new government had been established the defeat of France would have been a fait accompli and there would have been no reason for Britain to continue fighting alone. Yet, as he himself notes, the dissenting ministers consented to postpone their public resignation to prevent sending a message of disunity to Germany that would have facilitated the occupation of Belgium. Accordingly, Churchill communicated with the opposition Conservatives precisely to prevent such a possibility. It is likely that had the liberal government fallen, the British intervention in the War would not have been affected.

My own strong intuition is that history is composed of processes of varying levels of contingency and necessity. In other words, neither Marx nor Ferguson and Clark’s revisionism are absolutely right. The interesting question, though, is the extent and location of contingency and chaos in history. The only way to approach an answer to this question is empirically, by having more studies of virtual history as “calculations about the relative probability of plausible outcomes in a chaotic world” (85). The repercussions on our understanding of history and society cannot be overestimated.26

Since Pascal, Cleopatra’s nose has become the paradigmatic example of historical chaos. Monica’s dress may be a paradigmatic case of the difficulties in assessing the degree of contingency in history. On the face of it, it may appear that the events surrounding the definite proof of the president’s mishap were contingent. But it may also be argued that Clinton’s troubles were necessary: perhaps his politically self-destructive approach to the opposite sex would have landed him in trouble irrespective of the sensitive initial conditions that together with deterministic laws initiated a chain of events that left their mark on US presidential history.


  1. Exceptions to the rule followed Fogel’s use of counterfactuals in cliometrics to support his thesis about the limited influence of railways on US history; see, for example, Jon Elster, Logic and Society: Contradictions and Possible Worlds (Chichester, 1978), 175–221. In the social sciences, with the notable exception of Weber, counterfactuals have been ignored as well. An exception is Counterfactual Thought Experiments in World Politics: Logical, Methodological, and Psychological Perspectives, ed. Philip E. Tetlock and Aaron Belkin (Princeton, 1996). I found this anthology disappointing because though the authors discuss at length concepts such as theory, cause, and explanation, they largely ignore the great advances in our understanding of these concepts that have been made since the late 1960s in the philosophy of science. At most, they refer to the outdated accounts of Hempel from the 1940s and Goodman from the 1950s. This ignorance of the relevant philosophical literature cost these articles much of what could have been their methodological rigor, sophistication, and relevance. In philosophy, especially in philosophical logic, debates about the correct understanding and formal representation of counterfactuals are common; see, for example, Igal Kvart, A Theory of Counterfactuals (Indianapolis, 1986). These debates in philosophical logic are not directly relevant for evaluating historical counterfactuals.

  2. Philip K. Dick, The Man in the High Castle (London, 1965).

  3. Michael Morecook, A Nomad of the Time Streams: A Scientific Romance (London, 1993).

  4. Robert Harris, Fatherland (London, 1992).

  5. Yitzhak Laor, The People, Food for Kings (Tel Aviv, 1993) [in Hebrew].

  6. Morton White, Foundations of Historical Knowledge (New York, 1965), chapters 4–5; William Dray, Perspectives on History (London, 1980), chapter 4; Michael Hammond, “Weighing Causes in Historical Explanation,” Theoria 43 (1977), 103–128.

    Geoffrey Hawthorn, Plausible Worlds: Possibility and Understanding in History and the Social Sciences (Cambridge, Eng., 1991), also establishes the relation between counterfactuals and the assignment of necessary causes in three solid case studies, though the author seems to have attempted to prove much more in two semi-philosophical chapters that precede and succeed the case studies. These rather unclear and confused chapters are disappointing because Hawthorn cites there most of the recent updated relevant works in philosophy of science and epistemology that could have led him to far more interesting and clear conclusions. Had Hawthorn fully comprehended the works by van Fraassen and Quine that he cites, he could not have reiterated a hackneyed distinction between historiographical understanding and scientific knowledge in his conclusion, after reaffirming van Fraassen’s universal model of explanation and Quine’s equally universal underdetermination of theories. Cf. the justified criticism by Haskell Fain in his review of this book in History and Theory 32 (1993), 83–90.

  7. For a defense of modal realism, see David Lewis, Counterfactuals (Oxford, 1973).

  8. Gregory Benford, Timescape [1980] (London, 1996). Benford imagines an alternative universe that split from ours during the 1960s when minor mistakes in the chemical industry created a chain of events that led to irreversible environmental disaster that threatened the survival of the human race. In that universe, on the brink of destruction, scientists discover a method for sending messages to the past, using particles faster than light. A young scientist in the 1960s deciphers the messages and prevents the chain of events that led to the environmental disaster. But then, instead of preventing the alternative universe, he causes the splitting of the universes. Though the universe with the environmental disaster continues on its destructive course, the messages it sends to the past cannot reach the universe where the disaster was averted because it is no longer the past of the first universe. The reader is thus left comparing three universes, our own, and the two universes that split from us during the 1960s. David Lewis would approve.

  9. Maurice Mandelbaum, The Anatomy of Historical Knowledge (Baltimore, 1977), 105–108ff.

  10. Hawthorn, Plausible Worlds, 39–80.

  11. Yemima Ben-Menahem, “Historical Contingency,” Ratio 10 (1997), 99–107. The weakest part of Ben-Menahem’s explication of contingency is the absence of an explication of how different though similar results of different causal chains can be classified as of the same type. Such an explication is likely to be context- dependent.

  12. Ibid, 102.

  13. Ibid.

  14. Aviezer Tucker, “J. S. Mill’s Philosophy of the History of Science,” Explorations in Knowledge 13, no. 2 (1996), 21–31.

  15. Elster, Logic and Society.

  16. George Reisch, “Chaos, History, and Narrative,” History and Theory 30 (1991), 1–20 and Donald N. McCloskey, “History, Differential Equations, and the Problem of Narration,” History and Theory 30 (1991), 21–36 supposed that history is largely chaotic. Paul A. Roth and Thomas A. Ryckman, “Chaos, Clio, and Scientistic Illusions of Understanding,” History and Theory 34 (1995), 30–44 responded, denying the relevance of chaos theory to historiography. Rejoinders by Reisch and Michael Shermer followed in the same issue.

  17. Reisch, “Chaos, History, and Narrative,” 9. Reisch then discussed an alleged correspondence between chaotic subject matters and narrative explanation as an alternative to Hempel’s covering-law model of explanation. Since the bivalent choice between narrative and covering law is obsolete (there are many models of explanation that are debated in contemporary philosophy of science, and the covering-law model was falsified a generation ago), this discussion fails to become relevant for contemporary discussions of explanation.

  18. George Reisch, “Scientism without Tears: A Reply to Roth and Ryckman,” History and Theory 34 (1995), 45–58.

  19. Michael Shermer, “Exorcising Laplace’s Demon: Chaos and Antichaos, History and Metahistory,” History and Theory 34 (1995), 70–71.

  20. Isaac Asimov, “Spell My Name with an S” [1958], in The Complete Short Stories (London, 1994), I, 397–414.

  21. John Jakes, Black in Time (New York, 1970).

  22. Historiography is notoriously fragmented theoretically: Aviezer Tucker, “A Theory of Historiography as a Pre-Science,” Studies in History and Philosophy of Science 24 (1993), 633–667.

  23. I have argued elsewhere that the present institutional arrangements of historiographical research are a hindrance for such research: Aviezer Tucker, “Scientific Historiography Revisited: An Essay in the Metaphysics and Epistemology of History,” Dialogue: Canadian Philosophical Quarterly 37 (1998), 235–270.

  24. On page 363 in this article there is an inconsistency concerning the year of Litvinov’s commission. First it is 1943 and then 1944.

  25. Almond’s factual assumption may be disputed as well: he assumes that the nomenklatura had an interest in maintaining political power and their empire. The prevailing sentiment in Eastern Europe now, when the results of the collapse of political Communism are apparent, is that the nomenklatura stole the state and gained rather than lost from the collapse of political communism.

  26. Oddly enough, despite Ferguson’s anti-Marxist Tory rhetoric, he does not mention the obvious relation between Marx’s linear theory of economic development and central planning on the one hand, and Hayek’s chaotic theory of economic development and a free enterprise system on the other.

Fritz Stern (review date 8 February 1999)

(Contemporary Literary Criticism)

SOURCE: “Family Values,” in The New Republic, February 8, 1999, pp. 34-7.

[In the following review, Stern offers a positive evaluation of the first volume of The House of Rothschild.]

The house of Rothschild had many mansions, but it was one house—and this gave it unique power and allure. The dynasty began toward the end of the eighteenth century in the dank Judengasse of Frankfurt, where Mayer Amschel Rothschild and his wife Gutle Schnapper begat nineteen children, of whom ten lived, five sons and five daughters. The founder, as he came to be called, began humbly, burdened by all manner of disabilities leveled against the Jews, as a dealer in antiquities and...

(The entire section is 3102 words.)

Arthur Hertzberg (review date 21 February 1999)

(Contemporary Literary Criticism)

SOURCE: “All in the Family,” in Los Angeles Times Book Review, February 21, 1999, p. 5.

[In the following review, Hertzberg offers praise for the first volume of The House of Rothschild.]

Some years ago, a British journalist surveyed the descendants of the Jews who had been granted hereditary peerages since the middle of the 19th century. He found that the overwhelming majority of these titles belonged to grand-children and great-grandchildren who had converted to Christianity. The Rothschilds were the important and striking exception. To be sure, a number had left the faith. Many of the women had married Gentiles, and they had followed their husbands into the...

(The entire section is 1393 words.)

Robert S. Boynton (essay date 12 April 1999)

(Contemporary Literary Criticism)

SOURCE: “Thinking the Unthinkable,” in The New Yorker, April 12, 1999, pp. 43-6, 48-50.

[In the following essay, Boynton analyzes the critical reaction to The Pity of War and provides an overview of Ferguson's academic career and historical writings.]

In 1918, on the eleventh hour of the eleventh day of the eleventh month, the “war to end all wars” came to an end. Every November since 1919, the people of Great Britain have observed two minutes of silence to mark the moment when the guns on the Western Front fell quiet. For the eightieth anniversary of the Armistice, last year, the Royal Family, political leaders, and the Bishop of London gathered on...

(The entire section is 5245 words.)

Benjamin Schwartz (review date May 1999)

(Contemporary Literary Criticism)

SOURCE: “Was the Great War Necessary?,” in Atlantic Monthly, Vol. 283, No. 5, May, 1999, pp. 118-20, 122-8.

[In the following review, Schwartz provides an extended analysis of The Pity of War, citing weaknesses in Ferguson's “impassioned and distorted argument.”]

Americans scarcely marked the eightieth Armistice Day, this past November 11. But standing with stricken faces before the Cenotaph at Whitehall and the Ossuaire at Verdun, and tolling bells in the gloomy villages of Lancashire and the Pas-de-Calais, the British and the French, our erstwhile co-belligerents, mourned as if freshly wounded. For them the Great War is not yet merely history.


(The entire section is 4703 words.)

Andrew J. Bacevich (review date Summer 1999)

(Contemporary Literary Criticism)

SOURCE: A review of The Pity of War, in The Wilson Quarterly, Vol. 23, No. 3, Summer, 1999, pp. 131-3.

[In the following review, Bacevich concludes that The Pity of War is an “important contribution” to the study of World War I but finds Ferguson's analytical approach inadequate and reductive.]

Soldiers, statesmen, and scholars have long shared a common conceit: that, given sufficient effort and the right analytical tools, they might one day fully decipher the nature of war. As to where that understanding would lead, though, these groups part company. The soldiers and statesmen imagine bending war to their will and employing military power more...

(The entire section is 1105 words.)

Charles S. Maier (review date 28 June 1999)

(Contemporary Literary Criticism)

SOURCE: “Slaughterhouse Jive,” in The New Republic, June 28, 1999, pp. 51-4.

[In the following excerpt, Maier examines Ferguson's historical arguments and unconventional conclusions in The Pity of War.]

The British have taken to most of their wars. Slow learners who are sometimes handicapped at first by their own dismissive prejudices about the enemy, they usually recoup their losses in the decisive final periods, in part because of stolid battlefield virtues, in part because they know how to sustain coalitions. And they write about their wars very well, whether as memoir, poetry, novel, or history. World War I remains perhaps the preeminent object of this...

(The entire section is 3444 words.)

David Gress (review date 12 July 1999)

(Contemporary Literary Criticism)

SOURCE: “The Second Fall,” in National Review, July 12, 1999, pp. 50-1.

[In the following excerpt, Gress offers a positive assessment of The Pity of War but notes shortcomings in Ferguson's counterfactual approach.]

The late political philosopher Sidney Hook, though a staunch atheist, referred to the outbreak of World War I as “the second fall of man.” The phrasing reflected the profound sense, held by nearly all true democrats who witnessed the effects of that war, that the conflict marked a vast, ominous, and tragic diversion of the course of human history. Before 1914, the liberal principles of free trade, expanding suffrage, and increasing prosperity...

(The entire section is 1102 words.)

Leonard W. Boasberg (review date 9 August 1999)

(Contemporary Literary Criticism)

SOURCE: “Scholar Challenges Theories on the Origins of World War I,” in Chicago Tribune Books, August 9, 1999, p. 3.

[In the following review, Boasberg offers a positive assessment of The Pity of War.]

World War I was a stupid war, stupidly fought. The carnage was horrendous. In the four years and three months that the war dragged on, from August 1914 to November 1918, the Allied Powers—France, Britain, the British Empire, Belgium, the United States (entering in 1917), Russia (leaving, defeated, in 1917), Italy, Romania, Serbia, Greece and Portugal—lost 5,421,000 killed in action, more than 7 million wounded. The Central Powers—Germany, Austria-Hungary,...

(The entire section is 525 words.)

Hans Koning (review date 9-16 August 1999)

(Contemporary Literary Criticism)

SOURCE: “Still Not Over Over There?,” in The Nation, August 9-16, 1999, pp. 34, 36-8.

[In the following review, Koning disputes Ferguson's historical arguments in The Pity of War.]

The estimates of the number of books written about World War I are in the hundreds of thousands. By my estimate, Yale University Library holds 34,000 titles published before 1977 and more than 5,000 since. (The second category is on its computer, which counts up to 5,000 only.) The bibliography of Niall Ferguson’s The Pity of War lists about a thousand titles. But the author, no shrinking violet, advertises his book with the subtitle Explaining World War I.


(The entire section is 2613 words.)

Paul Kennedy (review date 12 August 1999)

(Contemporary Literary Criticism)

SOURCE: “In the Shadow of the Great War,” in The New York Review of Books, August 12, 1999, pp. 36-9.

[In the following excerpt, Kennedy examines Ferguson's controversial arguments and conclusions in The Pity of War.]

It is interesting but perhaps not surprising that, as this conflict-torn century nears its end, the shadows cast over it by the Great War of 1914–1918 seem in some ways longer, darker, and more daunting than ever before. For what that struggle meant and did changed the course of history more than any other in modern times, including its great successor war of 1939–1945. Consider only a few of the consequences of the Great War, offered here in no...

(The entire section is 4034 words.)

Donald Kagan (essay date October 1999)

(Contemporary Literary Criticism)

SOURCE: “Lessons of the Great War,” in Commentary, Vol. 108, No. 3, October, 1999, pp. 48-52.

[In the following essay, Kagan responds to Ferguson's arguments in The Pity of War concerning the length and resolution of World War I and Ferguson's counterfactual assertions.]

In August 1914, the Central Powers—Germany, Austria-Hungary, and their lesser allies—went to war against Russia, France, Great Britain, and their lesser allies. The Great War, as contemporaries called it, World War I to those who lived through its horrible successor a few decades later, raged for more than four years, doing awful damage. Battle casualties alone mounted to 4...

(The entire section is 4222 words.)

John Shaw (review date October-December 1999)

(Contemporary Literary Criticism)

SOURCE: “Five Arrows Make History,” in Europe Business Review, Vol. 2, No. 9, October-December, 1999, p. 40.

[In the following review, Shaw provides an overview of the Rothschild banking dynasty and offers a positive assessment of The World Banker.]

Many European and Australian bankers have recently been adding a bulky volume to their carry-on luggage.

The 1,300-page, $75 book is about banking, but unlike a Treasury report or a High Court judgment, it is readable enough to help pass the time on a long flight.

The World Banker by Niall Ferguson (published by Weidenfeld & Nicholson, London) is a revealing history of...

(The entire section is 724 words.)

Brian Bond (review date November 1999)

(Contemporary Literary Criticism)

SOURCE: A review of The Pity of War, in The English Historical Review, Vol. 114, No. 459, November, 1999, p. 1360.

[In the following review, Bond offers a favorable assessment of The Pity of War, but finds Ferguson's counterfactual approach problematic.]

At the outset of his provocative and immensely readable study, The Pity of War, Niall Ferguson poses ten bold, revisionist questions. These include: was the war (of 1914–18) inevitable; why did Germany’s leaders gamble on war; why did Britain’s leaders decide to intervene on the Continent; why did not the huge economic superiority of the British Empire cause the defeat of the Central Powers...

(The entire section is 514 words.)

Frederic Morton (review date 21 November 1999)

(Contemporary Literary Criticism)

SOURCE: “Dynasty,” in Los Angeles Times Book Review, November 21, 1999, p. 4.

[In the following review, Morton praises the unprecedented inside perspective offered in the second volume of The House of Rothschild but complains of it's excessive detail and documentation.]

In the annals of both high finance at its most feral and of high society at its most opulent, no entry has quite the glow of “Rothschild.” “That family,” the editor of the French edition of Vogue once sighed, “is the true successor to the Bourbons.” In England, the Rothschilds probably matched the fortune, and certainly transcended the flair, of the Windsors. In Austria, they...

(The entire section is 1425 words.)

Charles Paul Freund (review date December 1999)

(Contemporary Literary Criticism)

SOURCE: “Split History,” in Reason, Vol. 31, No. 7, December, 1999, p. 50.

[In the following review of Virtual History, Freund examines the function and application of counterfactuals in contemporary political discourse and historical scholarship.]

After the historical, comes the conditional: That’s how Robert E. Lee lost a battle this year in Virginia, where things had otherwise gone so well for the general since the unpleasantness in Appomattox that he’d become a rare American example of honor traduced by fate, of the peculiar fulfillments of the tragic. Yet in June, just as officials in Richmond were placing a Lee mural as a tribute along a new James...

(The entire section is 4060 words.)