Vol. 14 No. 3 · 13 February 1992

Had Lenin’s train collided with the Berlin-Basle express

Charles Maier

3299 words
Plausible Worlds: Possibility and Understanding in History and the Social Sciences 
by Geoffrey Hawthorn.
Cambridge, 192 pp., £27.50, September 1991, 0 521 40359 6
Show More
New Philosophy of Social Science: Problems of Indeterminacy 
by James Bohman.
Polity, 273 pp., £35, October 1991, 0 7456 0632 6
Show More
Show More

Certainly not the saddest for historians, according to Geoffrey Hawthorn’s wonderfully playful and intelligent book: rather, the most instructive. Hawthorn is intrigued by the philosophical standing of counter-factuals – hypothetical ‘other worlds’ – and their usefulness for historians and social scientists. Some historians resist the legitimacy of invoking counter-factual stories. They stringently insist that we can research and speculate only about what we believe actually occurred; anything else is merely fanciful. Call them factualists.

Hawthorn believes, as I do, that the factualists are mistaken both about what historians should do and what they actually do. If I read my colleagues correctly, they are continually, and properly, testing stories about what might have happened in order to figure out what did happen. To argue why the Han empire collapsed, why European populations contracted in the 14th century, why World War One erupted, why the United States decided to ‘contain’ Soviet Communism, usually requires demonstrating either why alternative outcomes did not occur or why other factors did not bring about the one that did. History as a story of wie es eigentlich gewesen always rests on at least an implicit wie es anders gewesen waere, how it might have been otherwise.

Probably what separates factualists from counter-factualists is a matter of temperament. For the factualist there is only one psychologically plausible world: the one that ‘is’. For counter-factualists the world as it is often rests on very thin ontological ice: for them, other worlds possess at least the same intensity of psychological presence as imaginary numbers offer the mathematician. Some of these other worlds, however, are more conceptually useful than others. Just because the historian departs from the actual world, he or she cannot wander like Saint-Exupéry’s Little Prince unrestrained among infinite possibilities. As Hawthorn emphasises, counter-factuals that are too inclusive do not help test causal hypotheses. Not all worlds that are conceivable in what we might call hypo-space have equally privileged subjunctive status. The counter-factual evoked must be a plausible world, not a totally different one.

The most powerful objection to counter-factuals, Hawthorn maintains, is not the factualist’s stubborn conviction that there is only one actual world to explain. The weightier argument is that counter-factuals are not controllable as thought experiments; they cannot be fitted to the real world one variable at a time. The real world is so holistically interconnected that to change one element must be to change them all: ‘the sheer thickness of circumstance that constitutes any past precludes any alternative at all except a wholly different world unwinding itself from the start of all worlds.’ This view is most closely associated with Leibniz, who maintained that every substance or monad had determinate relations with every other – past and future as well as contemporary – in any given possible world. Any counter-factual speculation would thus require opening up all history and recalculating all the switching-points since the Creation, even if the historian sought to speculate, say, only about why the atomic bomb was dropped on Hiroshima.

The historian or social scientist, however, must be able to think through imaginary changes one element at a time – the analogue in thought to a laboratory experiment. Mill outlined the procedure, which he called the ‘method of difference’, in his Logic, but he also seemed to caution that it was not really workable since societal factors, unlike those in the laboratory, could not be altered in isolation. (Mill fudged his rejection, however. Counter-factual states, he explained, could not be altered one element at a time when they pertained to definitional properties, though they might be selectively tested when they referred to empirical causal factors. For Leibniz there was no distinction between an entity’s definitional properties and its causal embeddedness in the world.)

Are counter-factuals, then, a seductive but hopeless strategy for the historical imagination? David Lewis argued in Counterfactuals (1973) that warranted counter-factuals not only had to obey causal logic, they had to imply a real alternative world, of which it seemed there must be an infinite number. Hawthorn along with earlier critics rejects this pluriverse, which for most common-sense analysts would entail rejecting all counter-factuals. More usefully, Jon Elster proposed that a valid counter-factual had to develop consistently out of a single historical branching-point: ‘the alternative state,’ he argued, had to ‘be capable of insertion into the real past’ (Logic and Society, 1978). Hawthorn endorses this criterion, which eliminates many badly-framed counter-factuals that pose alternatives before the historical preconditions for them had come into existence. As Hawthorn explains, in pondering why the Black Death abated in Europe, it is unacceptable to assume that the aetiology of the plague might have been understood earlier than it was.

Hawthorn, however, claims to differ from Elster in an important respect. Elster argued that counter-factuals were warranted only when they incorporated an alternative theory or general explanation of events. Hawthorn argues that this requirement entails applying to human behaviour the notions of law and generalisation that pertain in the natural sciences. The endeavour to frame such laws – what James Bohman characterises as the ‘strong programme’ or ‘old logic of social sciences’ – has been under attack for decades. And Hawthorn is not just pressing the familiar, if not banal arguments for hermeneutics or restating Stanley Fish’s claim that interpretation is the only game in town. He is also proposing that by its resistance to theoretical explanation, all history remains histoire événementielle. What Braudel and the Annalistes offer is just a diachronic version of Leibniz. As Hawthorn argues, a historical structure ‘is not an unchangeable state of affairs, but one that just happens not to have been much changed.’

Really? The notion of underlying structures and qualitative changes, or ‘crises’, has certainly remained a murky fetish among historians. Nonetheless, so many other disciplines, from paleontology to mathematics, are currently wrestling with issues of large-scale and discontinuous change that historians would do well not to abandon the concept of structure for the warm and friendly rhetoric of Post-Modern contingency. Agreed: plausible counter-factuals have to be debated almost jurisprudentially, case by case. Nonetheless, a valid counter-factual must ipso facto incorporate some alternative model of how history works. Theory, Hawthorn rightly emphasises, cannot reveal which counter-factual is warranted. But any plausible world will suggest its own theory.

Valid counter-factuals – in this respect Hawthorn does second Elster – impose a sort of trade-off whose specific terms are hard to stipulate in advance. They must leave open enough other possibilities to test the one that emerged, but they must simultaneously remain close enough to the actual scenario so as not to seem fanciful. Hence warrantable counter-factuals cannot stipulate that human agents might have had a psychological make-up at odds with what we know about them: to imagine Richelieu less concerned with state power, or, say, Hitler less obsessed by Jews, would involve positing actors strange even to themselves, thus implausible. Hence it is apparently not warranted to ask what if Hitler had undergone a shrapnel lobotomy on the Western Front and emerged a kinder and gentler veteran. And if a head wound is excluded, a fortiori historians should not be allowed to posit the early death of a known historical actor. Hawthorn implies that such speculation is truly idle. Historians must work with the individual lives and psyches that history did generate; we can posit different choices but not different choosers.

If I am reading Hawthorn correctly, he believes that a counter-factual based on mere chance, such as the sudden death of a major agent, is unwarranted and produces no interesting insight. At the same time, in conjuring up plausible alternative histories, he eloquently defends notions of contingency. At the macro level contingency apparently makes sense and is viewed as past possibility, whereas at the micro level it is disdained as a fanciful speculation and treated as mere chance. This may be too stringent an approach, however. Historians do seek to resolve issues about the respective roles of individuals and milieus, and no matter how naive, they try to think through complex situations altered by the absence of key leaders. If Lincoln or Franz Ferdinand had not been assassinated, or Roosevelt had lived ... Are all such procedures invalid?

The counter-factual represents a road not taken as a potential crossroads of history. As Hawthorn argues, too far in the past and the counter-factual remains too speculative to carry conviction. But too close to the event and the counter-factual hardly establishes alternative causal outcomes and thus inadequately tests the chain of causation that did prevail. Most historians are prepared to debate whether or not the Bolshevik Revolution might have been averted had Lenin’s apocryphally sealed train collided with the Berlin-Basle express or Kerensky initiated armistice talks. But it makes no sense to discuss whether or not it might have occurred had Peter the Great rather than Charles XII lost at Poltava. Too many events have intervened to establish any plausible chains of causation. On the other hand, the price of the counter-factual closer to the event is that it will fail to illuminate the broad spectrum of causative factors and put too much stress on the weight of micro-contingent factors. Histoire événementielle does not mean histoire par hasard.

The admirable quality that distinguishes Hawthorn’s book from previous discussions of historical explanation is his actual engagement with historical reflection through four extended case-studies. Hawthorn is not content to rework the classic instances that historians have used to invoke counter-factuals (the outbreak of war, the profitability of slavery). He applies counterfactual analysis to a broader range of historical phenomena: demography, political choices and the development of artistic styles. Some of the background circumstances are detailed beyond the needs of the argument, and Hawthorn has clearly developed an autonomous historical interest in each case and has done significant research. The presentation is dense but intellectually callisthenic; if you enjoy mathematical or logical puzzles you will delight in these expositions.

The first test case is the plague. Hawthorn proposes that counter-factuals can help us decide whether the plague abated because the bacterial strains lost virulence, the host rodents changed, or men learned how to enforce quarantine. A decline in virulence seems unproven, but so, too, does any significant medical learning. The best defence was keeping bearers away and segregating victims. Only a few cities had the determination and power to impose such a quarantine. Ultimately the counter-factual scrutiny suggests to Hawthorn that institutional capacity proved to be the major force in halting the devastation of the disease.

The argument about the plague reveals the implicit agenda of this book, which is to demonstrate that what historians often depict as macro or large structural developments – demographic changes, disease, styles in art – frequently hinge on political phenomena; and conversely, that political alternatives may be more intractable than those usually taken to be long-term and structural. The second case involves comparisons of British and French fertility from the late 17th to the late 18th century. Put simply, why did the French birth rate within marriage remain higher than the British until near the end of the 18th century? The reason, Hawthorn suggests, is that, as in Bangladesh, large families hedge against risk. The French had more children so long (and only so long) as they faced the risk of falling below the margin of subsistence. Sociopolitical institutions in England protected against destitution. No such safety net existed in France. Moreover, Hawthorn proposes, the wars of the French state under Richelieu and his successors required such a heavy tax burden that they placed French peasants close to the subsistence margin.

Ingenious though it is, this discussion hardly fulfils Hawthorn’s own criterion of plausibility. There is no measure of tax burden across the 17th and 18th centuries. Do the fiscal exactions of the state constitute an increase of risk? Even if they placed peasants closer to the margin of subsistence, expected economic burdens usually tend to discourage fertility. Does it really make sense to argue that when Olivares went to war he sparked the Catalan revolt, but that when Richelieu went to war his peasant subjects abandoned coitus interruptus?

The Korean argument is less strained. Hawthorn sensibly concludes that the United States could have decided not to occupy the southern half of Korea in 1945, but once there ‘it could not easily and without loss have decided to do anything very different from what it did.’ Before 1945, there was no Korean issue for which Americans had to produce a policy; after 1945, the options for withdrawal become increasingly implausible. The United States was a great power: ‘Powers expand to pre-empt; and great powers expand greatly.’ Politics are no less subject to refractory causes than economics or demography. This cannot resolve the issue of whether statesmen were right or wrong; their alternatives were at best possibilities. The possibilities illuminate how they weighed alternatives but only at the cost of increasing uncertainty with knowledge. ‘To understand more about any politics is thus to be less certain about it.’

Art history presents an additional challenge, according to Hawthorn. Just to describe the artifact imposes a theoretical vocabulary, a subjective evaluation of goal and technique. Why did Duccio in his great altarpiece for the duomo of Siena not achieve the breakthrough to perspective that Cimabue and Giotto did? Eliminating the alternatives – lack of knowledge or ability – Hawthorn, following other authorities, argues that Duccio’s concept of spirituality and the requirements imposed by his commissioners made perspective less desirable than it was for the Florentines. Christian art had devoted a millennium to suppressing pagan naturalism and establishing the pictorial constraints appropriate for depicting events simultaneously divine and human. Why trivialise these epiphanies, so Duccio might well have asked, by gimmicks with oblique angles? It is thus possible to divest Western art history of its teleology of naturalism.

Hawthorn’s treatment is an exemplary specimen of recent social science ‘cool’. It is chastened in its claims, distrustful of theory, tentative about interpretation, disarmingly modest in its post-Wittgensteinian sensitivity to communities of language. The ‘old’ social scientist believed his own knowledge was power – progressive and benevolent power. The new one has long since learned to distrust the benevolence. He or she employs a certain dewey-eyed and Deweyite stylistic trope, aspires to knowledge as play, or even, in Hawthorn’s case, renounces knowledge. ‘Success in history and the social sciences, as perhaps in life itself, consists in understanding more and knowing less.’

The aphorism rightly suggests that Hawthorn’s moral and political agenda is complex. At first glance, he deploys the counter-factual mode to reinforce the Post-Modern critique of scientific programmes descended from the Enlightenment. He joins the offensive against what Lyotard calls meta-narrative: the great teleological theories of social development which, so the Post-Modernists suggest, are close-minded, metaphysical and disdainful of individuality and agency. In the spirit of this book we might ask whether the revolutions of 1989 in Eastern Europe didn’t drive the nail into the coffin of critical theory as well as that of really-existing socialism. Did not 1989 free us from 1789 as well as fulfilling its promise?

The counter-factual temperament should also, Geoffrey Hawthorn believes, preclude our imposing abstract schemes on reality to see if it measures up to some ethical standard. The important distinction, he concludes, is not between social sciences and history, or theory and fact, but between actual and possible.

That is a more quietistic conclusion than the book perhaps warrants. Why not between possible and impossible? Or should such a question imply trop de zèle, why not between plausible and implausible? Counter-factuals preclude what Popper called historicism, the assumption of great deterministic schemes of development. On the other hand, Hawthorn concedes, they also preclude mere deconstruction, which undercuts the human agency Hawthorn feels remains critical in history. Like the end of a James Bond film, Hawthorn’s conclusions rescue us from some looming, and by now utterly familiar, implosion of theory. None the less, he is still uncomfortable with theory; I think excessively so. No plausible world, actual or potential, can be specified by theory alone: but none exists without forcing us to generate theory.

Presumably James Bohman would argue in the same spirit. Bohman has written a far drier, but learned and acute survey of the ‘new logic of social science’. The ‘old logic of social science’ yielded reductionist models of behaviour and tautological pseudo-explanations. Canvassing some of the most important recent debates in economics, sociology, anthropology and social philosophy, Bohman demonstrates how the champions of the interpretative social sciences have superseded these approaches. But he also criticises false starts in the new programme: the tautology inherent in theories relying on revealed preference or expected utility modelling; the collapse into empathy among participant observers in ethnomethodology.

As over a century of debate on this issue has made clear, the challenge of explanation is to understand and communicate the meaning of social activity. Some of the most influential practitioners of the new social science have argued that it is impossible to escape the ‘hermeneutic circle’ of subjectivism. The activity observed becomes a language game or community in-joke whose point is often lost in the description. The scientific activity of the observers becomes a sort of agonistic but co-operative competition according to other fetishised procedure. Ritual play described by ritual murderers.

Bohman insists, however, that explanation is possible within the framework of the interpretative social sciences. Habermas’s ‘critical theory’, with its continuous questioning of norms and reflection on the conditions that surround interpretation, keeps knowledge from becoming totally subjective. Geertz’s anthropology – despite the strictures of Post-Modernist critics who find it implies a project of Western cultural imperialism – endeavours both to penetrate a community’s activities and to keep a scientific distance. Bohman applauds these defences against Post-Modern or deconstructionist relativism. We can, he insists, accept indeterminacy, open-endedness, the fact that we may not unambiguously be able to describe social activity, far less to predict it, but still provide trustworthy explanations.

It is interesting to speculate how Bohman would react to Hawthorn’s trade-off of knowledge and understanding. He might find Hawthorn too willing to yield to a fashionable Post-Modern scepticism, too tentative about the possibilities of knowledge, and too modest in his claims for what history might teach. Bohman believes that we can rescue a concept of explanation. Nevertheless, the indeterminacy inherent in Hawthorn’s counter-factuals aligns his agenda with Bohman’s. While Hawthorn remains more sceptical of what he characterises as Habermas’s metaphysics, he, too, finally rejects the subjectivist implication of deconstruction and Post-Modern pyrrhonism. Hawthorn and Bohman agree that a ‘strong programme’, or a natural-science model of theory, can serve us no longer. But some practical or critical understanding of historical and social action remains possible, indeed necessary. Both converge – rightly I believe – on social-scientific understanding as a necessary and therefore feasible recourse.

In this last decade of the 20th century the issues these books explore are not just ‘theoretical’. If, as Hawthorn argues, we achieve understanding by envisaging possibilities, knowledge and politics overlap. Counter-factuals summon to reconstruction as well as reflection. Plausible alternative worlds suggest the implausibility that clings to the one we have made our own. Both the dismantling of Communism in Eastern Europe and the intellectual disarray of the Left in Western Europe and the United States have reinforced an often unreflective faith that public politics can be adjourned in favour of private transactions. It is hard to generate enthusiasm for any reconstruction of public norms. Hawthorn’s stance, as I have said, is decidedly cool. But cheerful coping may have its limits. From the perspective of the present, all political futures are possible worlds. More than one is ‘plausible’. Which world will achieve the ontological hegemony of actually coming to pass is for the moment indeterminate. But once one does, those that got away may not provide a very consoling spectacle.

Send Letters To:

The Editor
London Review of Books,
28 Little Russell Street
London, WC1A 2HN


Please include name, address, and a telephone number.


Vol. 14 No. 5 · 12 March 1992

Counterfactual history repels many historians because of the difficulty of knowing enough about all aspects of a period for the counter-factual proposal to be, as your reviewer Charles Maier says, ‘capable of insertion into the real past’. What we know about the past is constantly changing as research invalidates older views. Maier’s review of 13 February reveals that for certain periods either the reviewer or the authors of the books reviewed – it is not clear which – have not been in close touch with certain aspects of economic and social history over the last fifteen years. On plague, the reviewer mentions the importance of quarantine, but only at the local level. Recent work has revealed how Western Europe as a whole benefited from the quarantine and espionage system set up by the Austrian Empire to regulate contacts with the Turkish Empire. Then there is the comparison of British and French fertility from the late 17th century. Presumably for ‘British’ is meant English, for we have no trustworthy figures for Scottish 17th-century demography, and little also is known about Wales. The argument used depends on French families being under a much heavier tax burden than English ones. But for the 18th century Mathias and O’Brien showed in 1976 that the opposite was the case. Family limitation is suggested as in use in France in the 17th century, yet any serious study of comparative birth rates should surely recognise the achievement of Wrigley and Schofield in stressing the importance of variation in the age and likelihood of marriage as a major influence on the birth rate.

It is stated that ‘sociopolitical institutions’ prevented destitution in England and that ‘no such safety net existed in France.’ If it is poor relief that this phraseology refers to, the review, though rightly praising the efficacy of the English poor law, ignores the work of Hufton, Schwartz, Colin Jones and many others who have explored the French system.

Rosalind Mitchison
University of Edinburgh

Vol. 14 No. 7 · 9 April 1992

Rosalind Mitchison (Letters, 12 March) shoots each of her feet. I wrote the book she scorns in order to show the risk in the attitude she so nicely reveals: to try to know too much is to risk understanding less. How does a historian know when she knows enough? By considering the counterfactual implications of what she thinks she does know. What’s involved in doing that? Testing her explanations against the alternatives they suggest, for some of which there’ll be evidential support from elsewhere. And if Professor Mitchison had looked at the chapter in question, she’d have seen that most of the relevant evidence which she mentions is – as it happens – discussed.

Geoffrey Hawthorn
Cambridge University

send letters to

The Editor
London Review of Books
28 Little Russell Street
London, WC1A 2HN


Please include name, address and a telephone number

Read anywhere with the London Review of Books app, available now from the App Store for Apple devices, Google Play for Android devices and Amazon for your Kindle Fire.

Sign up to our newsletter

For highlights from the latest issue, our archive and the blog, as well as news, events and exclusive promotions.

Newsletter Preferences