1 - Explanation

Summary

The chapter begins with the observation that global history has an ambivalent attitude towards explanation. In many cases, the mere presentation of sources and voices from many different parts of the world seems sufficient to justify a global approach. The need for explanation is ignored or even denied. In other cases, global explanation is eagerly pursued, but often at the expense of more complex explanatory models that incorporate factors at different scales. In this perspective, global explanations are claimed to be inherently superior and a privileged way of explaining historical phenomena. After a cursory survey of current positions on causality and explanation in general methodology and ‘formal’ historical theory, the chapter proposes a brief typology of explanatory strategies. It goes on to discuss the peculiarities of explanation within a framework of connections across great distances and cultural boundaries. The much-exclaimed concept of narrative explanation is found to be of limited value, as it underestimates the difficulties of producing coherent narratives on a global scale. Concepts offered in the social science literature, such as the analysis of mechanisms and temporal sequences, could be helpful in refining purely narrative approaches to explanation.

Next to overcoming Eurocentrism – and perhaps any other form of value-laden centrism – the other big promise made by global history is to widen horizons, multiply the forms of human experience considered by the historian and increase the number of voices that are recovered from the past. Yet richness is not an end in itself. If the temptation of empirical overabundance is not resisted, world history turns into an ocean of the picturesque and the world historian into an old-fashioned polymath.

The obvious remedy is to employ concepts, patterns and strategies of emplotment with the purpose of giving shape to historical representation. In other words, sheer description tends to exhaust itself. In one of the most underrated contributions to historical theory, Siegfried Kracauer put it like this: ‘One might also say that the historian follows two tendencies – the realistic tendency which prompts him to get hold of all data of interest, and the formative tendency which requires him to explain the material in hand.’ ¹ It seemed to be a matter of course for Kracauer, writing in the 1960s, to equate the ‘formative tendency’ with ‘explanation’. As his subsequent discussion shows, Kracauer uses the term quite broadly – similar to how another great theorist, the Hungarian philosopher Ágnes Heller, was later to employ it when she wrote that ‘explanation’ was identical to ‘making something understood’.² In other words, to explain means ‘to make sense’ of what historians find in their sources. It also means to translate the past into the present and make it comprehensible while, ideally, not obliterating its strangeness.

Explanation builds upon description. It is the attempt to impart meaning to the evidence by distinguishing outcomes from causes and then tracing specific causes behind specific outcomes. Ágnes Heller adds an anthropological afterthought, which she does not really follow up: ‘“Why” is the elementary question, the first real question of a child. “How” is more sophisticated; it is a diffident “why”.’³ Explanation responds to a very basic human need; it is a naive expression of pristine amazement. Description – Kracauer’s ‘realistic tendency’ – already belongs to the answer. It requires care, even precision, a certain distance of the describing observer from the object under examination. It is by no means easy to give a good description, of historical events or of anything else. In a third step, to pick up Heller’s train of thought, explanation re-enters, fortified with method. Sense-making becomes systematic, follows certain conventional rules of logic and argumentation and sometimes aims at higher orders of abstraction.⁴

It should, therefore, be taken with a pinch of salt if historians deny any intention to explain, as was fashionable at the peak of the cultural turn. The more historians turned away from politics and economics, and the closer they drew to literary studies and certain tendencies in cultural anthropology, the lower fell the regard in which explanation was held. This was all the more true for the related concept of ‘causation’. In the 1980s, as R. Bin Wong aptly points out, ‘causation was no longer as central a concern of historians as it once was’.⁵ Neither was explanation.

Though it is impossible to speak about explanation without mentioning ‘causes’, I shall avoid the concept of ‘causation’. In analytical epistemology, ‘explanation’ and ‘causation’ or ‘causality’ are different if related topics.⁶ Among philosophers, causality presently seems to be the more exciting of the two. There is now consensus that (a) causation unfolds in processes, and (b) that it can be probabilistic. Other aspects of the topic are more controversial.⁷ The present chapter focuses on explanation, leaving aside causality.

I shall argue, empirically, that not all kinds of global history aim at explanation, but a lot do – sometimes explicitly, often in subcutaneous, implicit and hidden ways that should be brought to light. While I agree with Siegfried Kracauer that not everything in the social and political past is (rationally) explicable and that we must reckon with the existence of ‘irreducible entities’,⁸ I want to show that explanation ought to matter for the field of history, but even more so for that of global history, which is confronted with unusually rich and diverse evidence. Explanation is an important tool for reducing complexity in a ‘formative’ (Kracauer) way.

To be sure, there is a wide variety of explanatory approaches, none of them particular to global history. In actual practice, explanations are of differing quality, on a scale from brilliant to utterly unconvincing.⁹ It is an important task of scholarly critique to assess that quality in individual cases. The only general rule is a formal one that holds true for history as it does, more or less, for all scholarship: the imperative to avoid monocausality. Yet, under special circumstances, explanations have to be monocausal. Ancient Pompeii was destroyed in AD 79 by a volcanic eruption and by nothing else. Still, historians usually steer clear of monocausality and unilinear determinism. The decline and fall of the Western Roman Empire require a much more complex explanatory design than the end of Pompeii. At the same time, explanations should be elegant and parsimonious, stopping short of overcomplexity. Such overcomplexity may degenerate into long lists of factors that are suspected to be operative in a vaguely specified manner. In their practical work, historians are likely to look for graspable, intuitively plausible explanations; they prefer – or should prefer – controlled simplification to comprehensive fuzziness. Not everything is related to everything else – as a vulgar notion of globality tends to imply. The business of explanation consists, to a large extent, of taking decisions about what is relevant and what is less so in making sense of a particular historical constellation.

Since explanation has rarely been discussed in the theoretical literature on global history, this chapter begins with a brief overview of what historical theory and the methodology of the social sciences may have to offer global historians.

General Theories of Explanation

The word ‘explanation’ basically refers to two different things. Firstly, it can mean to give reasons – often moral justifications – for one’s own actions or those of other humans. As Charles Tilly has put it, human beings are ‘reason-giving animals’.¹⁰ For historians, this is an object of study. We look in the sources for attempts by historical actors to provide reasons and motives for their actions, and we do not expect such explanations to be ‘logical’ or ‘rational’. Secondly, to explain something can mean to account for states of affairs by identifying connections between ‘causes’ and ‘effects’ – in other words, by providing causal analysis. In this case, reasons and causes are established by the analyst, sidelining the self-expressions of the actors or taking a distancing view of them. The second meaning of ‘explanation’ transcends subjective intentionality and encompasses ‘structural’ considerations that often require hindsight and transgress the awareness of the historical actors. The question of ‘why people do what they do’ is not answered best by those people themselves.¹¹

The philosophical theory of explanation derives from Aristotle and, in modern times, from John Stuart Mill’s System of Logic (1843).¹² Its application to history basically begins with Carl Gustav Hempel’s 1942 theory of explanation that went through various modifications up to its final version developed in the 1960s – still a benchmark approach.¹³ Ever since the later Hempel, the theory rests on two assumptions: (a) it is nominalist or constructivist and does not require the assumption that causes ‘exist in reality’; (b) it presupposes that a specific effect and its specific cause, or causes, are connected by something more general: not necessarily an invariable and time–space-insensitive ‘natural law’ or ‘iron law of history’, but perhaps a more limited regularity.¹⁴ That regularity should be empirically rich as well as theoretically plausible, for instance when historians cite a well-established sociological theorem or an evidence-based insight from demography to help them account for a specific phenomenon in social history or the history of population.

General theories of explanation are nowadays worded much less rigorously than they used to be in the days of Carl G. Hempel.¹⁵ They often allow for ‘contexts’ of various kinds. Such ‘pragmatic’ theories¹⁶ have become acceptable, and even influential, as a result of the historicising and relativising turn in the philosophy and sociology of science inaugurated by Thomas S. Kuhn and others in the 1960s. Wesley C. Salmon, perhaps the most influential philosophical theorist of explanation in the generation after Hempel, has allowed for ‘causal networks’ and ‘etiological explanations’.¹⁷ In the last phase of his work, Hempel himself used ‘soft’ formulations such as asking what ‘made a difference’ or how ‘relevant’ causal factors were. He also envisaged ‘fine-grained mechanical explanations’¹⁸ that generate knowledge about ‘how things work’.¹⁹ Today’s major authority on explanation, James Woodward, even permits counterfactuals (i.e. sentences of the type ‘What if things had been different …?’).²⁰ In sum, the general theory of explanation is nowadays perhaps less parsimonious and elegant than in Carl Gustav Hempel’s foundational design of 1942, but much closer to the actual practice of scientists and less prescriptive than it used to be. Even so, philosophers still look for general criteria to assess the quality of particular explanations and to detect logical flaws in them.

Historians are busy people and unlikely to spend much time on the intricacies of the general theory of explanation. Still, denying its relevance would be an anachronistic relapse into crude dichotomies of science versus humanities, ‘nomothetic’ (law-based) versus ‘idiographic’ (case-based) disciplines or quantitative versus qualitative approaches. It would be a denial of the basic methodological unity of all the sciences.²¹ The boundaries between the famous ‘two cultures’ have become porous, not least through the rise of digital awareness in the humanities, including global history, where sometimes datasets of enormous volume and variety have to be processed.

In Defence of (Historical) Explanation

The heyday of debates on historical explanation was in the 1960s and 1970s. For our time, Paul A. Roth, one of the few remaining exponents of an ‘analytical’ theory of history in a loosely conceived Hempelian tradition, diagnoses ‘an almost total neglect of historical explanation within philosophy of science’.²² This is generally true for the ‘formal’ theory of history (in German: Historik), which has to be distinguished, following Ernst Troeltsch, from the ‘material’ philosophy of history that grapples with the big sweep of ‘real’ history.²³

The formal theory of history and the numerous programmatic self-reflections of historians tend to be almost silent on explanation. When historians ponder what they are actually doing, they rarely come to the conclusion that they elaborate explanations. The latest careful discussion of historical explanation, using numerous examples from the historiographical literature, dates from the previous century.²⁴ Achim Landwehr, a prolific German theorist, sees the historian’s task in the description of complexity and declines any further ambition; he does not even mention the issue of explanation.²⁵ Jörn Rüsen has downgraded the relative position of explanation within his comprehensive system of historical knowledge from version to version.²⁶ Reinhart Koselleck, who is enlightening on almost any question within the theory of history, was largely reticent on matters of explanation. Global historians, too, are diffident on the issue of explanation. Sebastian Conrad, today’s foremost theorist of global history, does not show much interest in it. Where he touches upon the matter, he apologises to the reader that his brief remarks might appear ‘rather technical and inconsequential’.²⁷ Diego Olstein has interesting things to say about contextualisation, comparison and connections, but next to nothing on explanation.²⁸

Why this white spot on the map of historical theory? There are at least three possible answers:

1. (a) Historians believe that explanation is something to be left to cliometricians, with their social-scientific minds, and to schematic historical sociologists, as it is no primary concern of the mainstream. They are reluctant to admit that, whether they are aware of it or not, they answer ‘why’ questions all the time. Few historians are likely to endorse the unequivocal assertion made by the Canadian philosopher Mario Bunge: ‘All the historical sciences have the same aim, namely, to discover what happened and why it happened: they seek truth and explanation, not just yarn.’²⁹ And not everyone would agree with Paul Veyne when he says that ‘to explain more is to narrate better’.³⁰

2. (b) A second explanation of non-explanation would be that this is not what the public expects from science in general, and from historical studies in particular. The public is said to be keen on ‘yarn’. This, too, is dubious. The pandemic year 2020 was a time when science – from virology to empirical social research – faced an unprecedented demand for discovering the causes of our multiple predicaments. Explanations were indispensable for finding remedies and practical solutions and for predicting the future. Historians were quite successful in explaining how we got to where we are – and, more specifically, how and why similar causes led to diverging outcomes.³¹ After Russia started its full-scale invasion of Ukraine in February 2022, historians of Eastern Europe were in great demand to give reasons for the Russian leadership’s motives, goals and conduct against the backdrop of the long-term history of the region.

3. (c) The third possible reason for the occlusion of explanation in historical theory comes closer to the mark: a powerful ‘narrativist turn’ since the 1970s, set in motion and sustained by cultural theorists and literary critics, gained intellectual hegemony at the expense of the analytical theory of history. It seemed to be closer to the activity of writing history than the abstract deliberations of the logicians in the Hempel tradition. Narrativist theorists believe that history is about constructing plots whose rootedness in evidential research, or the lack of it, is of subordinate importance to theory. While few working historians were (and are) persuaded that this approach offers an adequate description of what they are actually doing, narrativism conquered Anglo-American theory and came to lead a life of its own. The much more nuanced narrativism of the French philosopher Paul Ricœur was not as influential internationally as it should have been. Nor was Michel Foucault’s non-analytical concept of ‘genealogy’.³² Despite Foucault’s worldwide celebrity, few global historians have so far worked in such a genealogical mode.

In sum, explanation is a permanent concern and standard procedure of all historical sciences, including archaeology, palaeontology, historical demography and so on. Its current neglect in the formal theory of history does not mirror its real significance.

Analytical and Narrativist Theories

The analytical theory of history applies thought patterns from the logic of scientific research to the humanities and is primarily concerned with methodology. Reaching its high point in the influential works of Arthur C. Danto (1965) and Louis O. Mink (1987),³³ and represented today by Paul A. Roth and, with certain limitations, Aviezer Tucker,³⁴ it addresses the central theme of how historians establish the truth, or other forms of epistemic authority, of their verbal propositions. Analytical theorists have never shown much interest in analysing texts written by ordinary historians. They usually deal with brief and simple speech acts. Though this can hardly be otherwise for the sake of philosophical clarity, it limits the impact of analytical theory outside its own circles. Most historical explanations are complex argumentative constructions that cannot be reduced to atomistic events and isolated propositional sentences. Correspondingly, analytical theorists tend to have a reductive and old-fashioned understanding of real-life historiography, which they prefer to see as a linear chronicle of political events.

Paradoxically, the same is true for the arch opponent of the analytical school: narrativist theory. It is simply much easier to tell – and to analyse using the tools of narratology – a tale in the style of l’historie événementielle than to express multivariable causal arguments in narrative form.³⁵ Thus, both schools of theory suffer from an inbuilt bias against all kinds of structural history and also against cultural history of a more sophisticated bent.

Narrativist theory, to this day labouring under the shadow of Hayden White’s celebrated Metahistory: The Historical Imagination in Nineteenth-Century Europe (1973),³⁶ must not be taken too seriously in its far-reaching agnostic claims: that historians are unable to establish anything like the ‘truth’ about the past, that their utterances lack an extralingual referent and so on.³⁷ White himself has impressively analysed a handful of nineteenth-century historiographical classics that were written in co-evolution with the historical novel. His approach, and the more mundane and technical methods of the narratology of literary historians,³⁸ however, fail to do justice to research-based historical scholarship and its textual strategies, which are not primarily governed by literary techniques of spinning a tale. Moreover, form and rhetoric, though important for historical studies, are not essential for them. Whether the Gordian Knot can be cut by postulating something like ‘narrative explanation’ – in Jörn Rüsen’s view a ‘discursive practice’ that synthesises all aspects of historical writing³⁹ – remains controversial. A recent survey of the literature concludes ‘that it is hard to say what a narrative explanation precisely consists of’.⁴⁰ A new and promising approach to the connection between narration and argumentation suggested by a team of authors around the Spanish philosopher Paula Olmos has yet to reach the historiographical debate.⁴¹ So far, narrativism has difficulty offering criteria for assessing the quality of a specific explanation. To put it bluntly, any explanation seems to be acceptable as long as it is disguised as a good read.

A recent work by the Finnish theorist Jouni-Matti Kuukkanen fails to inspire more confidence than earlier narrativist theory. His Postnarrativist Philosophy of Historiography (2015) is the epitome of ultra-narrativism. The author, proud to represent ‘the dominant school’,⁴² is interested ‘not so much in the generation of historical knowledge and explanation as in the forms in which it is presented’,⁴³ and he dismisses Carl G. Hempel’s covering law theory, and with it analytical theory as a whole, as ‘(in)famous’,⁴⁴ unworthy of philosophical attention. Kuukkanen wants to liberate – in an age of interdisciplinarity – the humanities from ‘disciplinary externalism’, a term that appears to refer to the purported straightjacket of the natural sciences.⁴⁵ He sees no way of assessing the intrinsic worth of works of history. Rather, he subscribes to something like textual Darwinism: ‘The plausibility of a historical thesis depends on its impact within the argumentative field.’⁴⁶ The winner takes all.

More nuanced theoretical suggestions went down on the battlefield between the two tendencies, but deserve a new look. This applies to the philosopher Maurice Mandelbaum, with his urbane and learned attempt to bridge the chasm between the analytical and the narrativist schools.⁴⁷ Mandelbaum argues realistically that rather than spin linear plots, historians construct multilayered ‘sequences’ into which they incorporate explanatory elements.⁴⁸ The task of the historian is not so much to string together the pieces of a story as to clarify the relations between the various elements in a two-dimensional tissue. Mandelbaum also makes the important point that in analysing change historians should never forget ‘external’ factors.⁴⁹ It is always a promising working hypothesis, says Mandelbaum, that there is an ‘outside’ to one’s particular field of investigation – in other words, an external arena from where forces may impinge on what at first sight looks like a closed system: for instance, a nation-state. Though Mandelbaum is never quoted by global historians, his insights are much more pertinent for global history’s concerns than anything offered by current narrativism or analytical theory.

Another author worth (re)discovering is the Austrian sociologist, philosopher and historian of ideas Karl Acham, who began his career with an excellent summary and critique of the analytical school.⁵⁰ He has since reflected deeply on what Hempel already allowed for as ‘explanation sketches’: less rigorous than strictly universalist ‘nomological’ explanations and able to accommodate the plurality of factors and scales characteristic of the humanities.⁵¹ After many decades of struggling against the eviction of ‘meaning’ (Sinn) by a methodology of history subservient to the natural sciences, another veteran, the aforementioned Jörn Rüsen, has finally arrived at a sceptical verdict on narrativism to which he had always shown a close affinity. In Rüsen’s view, the triumph of that school has led to the consequence that ‘the problem of rationality was suppressed [verdrängt] rather than solved’, and, along with it, the question of the scientific nature (Wissenschaftlichkeit) of the work performed by historians.⁵² Following up on this, a slightly different answer to narrativism might be: the principal aim of historical studies is not to tell stories but to ask questions and provide the strongest possible rational justification for the answers given to those questions on the basis of the best available evidence.⁵³

In sum, both major schools within the formal philosophy of history - the disciples of Carl G. Hempel and the followers of Hayden White – offer only limited access to what Marc Bloch called le métier de l’historien, especially to the questions of how historians explain in actual practice and how their explanations might be improved. The analytical tendency, however, has a sense for the interplay between the general and the particular in historical reasoning and maintains the idea of an intersubjectively valid logic in the service of truth, whereas the narrativists lack respect for historical research and assimilate historical writing to the construction of fictional tales.

Sequences and Mechanisms: Explanation in the Social Sciences

The social sciences are close neighbours of historical studies. Both deal with individual and collective human behaviour; both study change over time; both differ from the natural sciences in that it is impossible (history) or difficult (social sciences) for them to observe reality directly or under laboratory conditions. A few remarks shall be offered about sociology, a discipline that since the days of Émile Durkheim and Max Weber has occupied a middle ground between Verstehen (hermeneutical understanding) and Erklären (explanation).⁵⁴

Unlike philosophers of history, sociologists are not interested in telling historians what to do. Thus, we have to reverse the perspective. Is there anything historians can learn from sociologists when it comes to explanation? Three points may be worth exploring further.

1. (a) Historical sociology has always been a decidedly explicatory discourse, comparison being its preferred method.⁵⁵ One of its favoured approaches is a dynamic comparison between developmental paths and trajectories. Historians tend to complain about the remoteness of historical sociologists from primary sources and of a certain formalism or schematism in their comparative thought experiments. Yet, in the best case, the chosen explanatory set-ups are complex as well as transparent, involving neatly defined factors and plausible hypotheses about the interplay between those factors over time.⁵⁶ The entire Great Divergence debate – to many observers quintessential global history – owes a lot to the social science methodology of comparison.⁵⁷

2. (b) A relatively new concept, explicitly conceived of as a way to facilitate causal explanations, is that of the mechanism.⁵⁸ Such an approach would either look at psychological mechanisms that make individual and collective behaviour more or less predictable,⁵⁹ or postulate medium-range and small-scale regularities between certain causes and certain effects in ‘processes involving large populations and interacting networks of organisations’.⁶⁰ Mechanisms as regularities of limited scope partly fulfil the requirements for ‘covering laws’ which are essential in Hempel’s nomological model of explanation. They also show a family resemblance with Reinhart Koselleck’s ‘patterns of repetition’, a fascinating though under-elaborated element of Koselleck’s mature, and rather sketchy, theory of history.⁶¹

3. (c) A recent innovation in theoretical sociology is the use of temporal sequences and ‘syntaxes’ for purposes of explanation.⁶² That concept involves the close study of temporal shifts and conjunctures and may help to better describe the concatenation of causal factors that historians like to employ in a less systematic fashion than sociologists. Sequencing works best when it is seen as preparing explanation rather than replacing it. Similarly, within the vast field of theories of time, sociological contributions stand out in their resolve to overcome the antagonism of experienced or subjective time against measured or objective time. They are particularly good at dissecting complex processes into their constituent elements and at postulating causal connections between those elements.

What these three sociological approaches have in common is that they reveal the bare bones of their explicatory arrangements in a way that can alert historians to their own strategies of reasoning. The explicatory or ‘configurational’ models ⁶³ used by (historical) sociologists tend to be much more intricate, and therefore better attuned to the practice of historians, than the often reductive and simplistic ideas about nomological, intentional or narrative explanation cherished by analytical and narrativist philosophers of history alike.⁶⁴

How Do (Global) Historians Explain?

In the study of historiographical texts, explanation has received much less attention than rhetoric and narrative emplotment. There are surprisingly few in-depth analyses of how historians actually practice explanation, even with regard to the great classics of the historiographical canon. Tim Rood’s Thucydides: Narrative and Explanation and Jonas Grethlein’s wide-ranging work on ancient historiography can serve as models for what is deplorably lacking for other authors and epochs.⁶⁵ Even less is known about the crafting of the routine research output in today’s discipline. How do normal historians handle explanation? Since actual practice remains obscure, firm foundations are lacking for normative assessments: what is a good explanation?

Nobody should expect a straightforward and universally valid answer. How we explain depends to a large extent on what we want to explain. Explanations in intellectual history are different from those in economic history. It is one thing to discover the reasons behind an individual political decision, quite another to account for a macro-process such as the outbreak of a multi-state war, the collapse of an empire, or a trajectory of economic development or implosion. A plausible guess is that what is difficult to analyse in most fields of historical study is even more difficult for global history. Though this should not be misunderstood as a claim to superiority, global history has to handle more factors and variables and a greater number of diverse actors and social configurations than is the case for most other fields of history. The ritualised assurance that global history is, or should be, multi-archival and multi-lingual is just a consequence of the fact that it is a rather complicated, disparate and sometimes messy affair. Since explanation is a way to reduce complexity, the burden that lies on explanation in global history is a particularly heavy one. It has to tame – to use a buzzword – ‘vibrant’ plurality.

Under these circumstances, critical interventions that probe the soundness (Woodward) of argumentation and explanation are highly welcome. I single out three of them.

Firstly, several distinguished historians, not known as sworn enemies of global history, have applied the emergency brake to a merry-go-round of high-sales publishing and launched a vehement attack against ‘fake global history’.⁶⁶ ‘Global history’, the critics declare, ‘has become an excuse for authors to make outlandish claims, based on the belief that they will not be subject to the usual scholarly scrutiny.’⁶⁷ Such ‘outlandish claims’ include sloppy explanations. The general justification of this charge derives from two related observations. On the one hand, there is a certain pressure in the ‘trade’ section of the book market even for respectable historians to exaggerate the colourfulness of their materials and the drama of their interpretations. On the other, wide-ranging works that cut across academic compartmentalisation are difficult to assess in terms of specialised scholarship. They easily slip through the net of responsible scrutiny and are applauded by overwhelmed reviewers for superficial virtues such as daring assertions, unparalleled comprehensiveness or the alleged uncovering of secrets.

Secondly, and closer to the issue of explanation, Princeton historian David A. Bell has undertaken an interesting thought experiment.⁶⁸ Bell, an expert on Europe around 1800, raises a helpful question: ‘What is gained from placing it [the French Revolution] in a global perspective, and what is lost?’⁶⁹ For ‘French Revolution’ one can easily substitute any major historical phenomenon or mega-event that has conventionally been considered within a non-global context and is now seen through the unaccustomed spectacles of global history. Bell supports his major question with a long lists of complaints about the over-ambitiousness of historiographical globalisers. While one does not have to follow him through all the twists and turns of his philippic, his focus on the French Revolution ensures that explanation occupies centre stage in his intervention. Very few episodes in history have been linked to more why-questions than the French Revolution. Why did a major upheaval occur in France and not elsewhere? Why in 1789? Why did the Ancien Régime succumb? Why did the Revolution go through a process of radicalisation? And so on. These questions have usually been raised and answered within a French or a European framework. A global approach, still under debate, suggests longer concatenations covering the entire North Atlantic space or even regarding French domestic developments as part of a general world crisis.⁷⁰ Bell offers a useful distinction of general applicability when he insists that ‘inward influences’ – the causal impact of external actors and events – and ‘outward influences’ – effects, often long-term and unspecific, reaching out into the world – follow different logics.⁷¹ This is generally true. Some local events have global ramifications, most others do not. (Not every assassination of the member of a royal house leads, as did the shots of 28 June 1914 at Sarajevo, to global war.) In reverse, the fact that an event acquired universal significance does not always mean that its origins were ‘global’; most ‘world religions’ have distinctly local roots.

From the vantage point of the historiographical practitioner, David Bell confirms a lesson also to be learned from the methodology of explanation: if additional – in this case, ‘global’ – factors are being added to an explanatory model, hypotheses are needed that specify precisely the possible causal connections between the new factors and the other elements of the model. Invoking an atmospheric ‘globality’ does not explain anything. Nor are fuzzy ‘waves’ that ‘sweep’ around the globe proper candidates for independent variables and fundamental causes. It is from theorists such as Maurice Mandelbaum and from the sociological analysis of temporal sequences that one should take away the imperative to disaggregate complex processes into their constituent parts and look for specific connections rather than for general connectivity.

Thirdly, Peer Vries, whose early work in the formal theory of history sharpened his later acuity as a global economic historian,⁷² sums up his experience as a pioneer of the field in a stern admonition: ‘More energy should be devoted to determining the exact extent and impact of the various kinds of “contacts” and “exchanges” of which global historians are so fond.’⁷³ Calling for ‘more methodological awareness’, Vries goes on to note that the seemingly avant garde label of global history camouflages a lot of scholarly practices that are ‘strikingly traditional’.⁷⁴ A new kind of history – as long as that ambition is kept alive – requires methodological adjustments and innovations. A global perspective, this is Vries’s persuasive argument, is always worth a try even though it cannot claim a priori superiority over conventional approaches.

Varieties of Explanation in Global History

Even if Peter Perdue, a distinguished historian of China, exaggerates when he says that the term ‘global history’ nowadays ‘can refer to almost anything’,⁷⁵ it remains true that many different kinds of history have comfortably settled under the umbrella of global history. That umbrella becomes even wider if one includes the more popularising trends within the global history discourse – in other words, those books that shape public impressions of what global history is about and why it is important. The discrepancy, for instance, between global microhistory and those macro-approaches that border on historical sociology and consider the ‘very long run’ is so enormous that a shared strategy of explanation is almost ruled out. Thus, there is no manner of explanation that is a distinctive feature of global history.

Still, a few basic ways of handling explanation can be discerned:

(1) Non-Explanation

A great deal of what goes under the label of global history was never meant to explain anything. I suggest calling this the ‘panoramic’ approach, to be distinguished from ‘analytical’ global history. Entirely legitimate, panoramic global history appears in the shape of various globalising genres. One such genre are the fashionable globalised histories of particular nation-states, patterned on L’histoire mondiale de la France.⁷⁶ They do not count as explanatory history simply because they pursue only modest analytical aims. These voluminous tomes have to be seen as synthetic statements intended for national education.

A second non-explicatory genre are general histories of the world. Nowadays they are wary of big questions and the corresponding big answers. William H. McNeill, writing in the halcyon days of the Pax Americana, put a major puzzle into the title of his work: The Rise of the West.⁷⁷ Yet his treatment remained safely in the descriptive mode and deserves to be remembered mainly for its imaginative periodisation and a few crisp chapter headings (‘Moslem Catalepsy, 1700–1840 AD’, etc.). McNeill told a story. He did not distinguish systematically between causes and effects and therefore did not offer an explanatory model, let alone a theory.

McNeill’s master of sorts, Arnold J. Toynbee, had been of a different cast of mind. Especially in his best decade, the 1930s, Toynbee was a dedicated explainer and anything but a spinner of epic tales. While Toynbee’s lack of interest in sociology and ethnology makes even his best works – the first six volumes of A Study of History⁷⁸ – look old-fashioned, his approach to explanation was rational and unpretentious. He did not believe in perennial ‘laws of history’ and, in a sense, anticipated the middle-range ‘causal mechanisms’ (Renate Mayntz) and ‘patterns of repetition’ (Reinhart Koselleck) mentioned earlier.⁷⁹

Today, the better one-volume world histories are playful philosophical reflections decked out with illustrative pluckings from the past.⁸⁰ When the burden of writing a history of the entire world is shouldered in scholarly earnestness, all sorts of rump explanations are attempted with hardly ever a sense of satisfaction. With everything remaining half-said, unintended monocausality can hardly be avoided.⁸¹ In a nutshell, much of published global history is never meant to serve as a stepping stone towards explanation. It is merely exhibitive: materials from all over the world are assembled and displayed, enriching people’s knowledge and strengthening their sense of diversity and their cosmopolitan outlook.

(2) Pan-Explanation and Explanation Through Comparison

The other end of the spectrum of explanatory intensity is marked by works where comparison is used to identify those variables that make a causal difference. These works are global – in a non-methodical way – if they straddle commonly respected cultural boundaries. ‘Non-methodical’ means that the logical strategies used in a comparison between Britain and France and in a comparison between France and Japan are basically the same. The only two differences, certainly requiring careful attention, are (a) a greater obtrusiveness of the ‘cultural’ dimension that cannot be disregarded or bracketed in a ceteris paribus way; and (b) a greater relevance of ‘emic’ (as distinct from ‘etic’) nomenclatures – in other words, ‘local’ or ‘indigenous’ terminologies. To illustrate the second point with examples from comparative social history: samurai in a Japanese, shenshi (scholar-officials) in a Chinese, gentry in an English and noblesse de robe in a French context are local categories that are almost impossible to translate and difficult to subsume under generic terms of higher abstraction and universality.

The apotheosis of comparativism in global history was attained in the Great Divergence Debate.⁸² Here everything revolves around explanations in response to one of the biggest why-questions ever asked: Why did ‘the West’ (or Europe, North-Western Europe, etc., respectively) achieve worldwide superiority in the modern era? What accounts for the increasing economic disparities between different parts of the world? The debate started almost a century before the publication of Kenneth Pomeranz’s famous book⁸³ with Max Weber’s titanic effort, undertaken at a time when the social and economic study of Asian societies had barely begun, to account for the emergence of rational capitalism in the Occident by contrasting it with supposedly dead-end trajectories in China, India and elsewhere. In the early twenty-first century, the debate has been the most important laboratory for macro-historical explanation through comparison. Regardless of innumerable disagreements among a vast array of authors, the participants share a few commonalities.

Though history books would be unreadable without narrative, no participant in the debate relies on narrative alone to produce explanations in the sense of the theorists’ ‘narrative explanation’. Mirroring developments in the real world, general attention in these explanations has shifted from probing the ‘rise’ of Europe to finding reasons for the delayed ‘rise’ of China. Thus, the explanandum remained more or less the same across more than a century – why did the ‘normal’ disparities in wealth, power and cultural creativity between major parts of the world result in one dramatic bifurcation, a great divergence among so many small divergences?⁸⁴ – whereas the candidates for the losing and winning positions changed several times. Concurrently, the basic parameters – or variables, in the language of quantification – of explanation have kept shifting, which complicates the debate considerably. What was it that diverged in the first place? Economic growth, capitalism, scientific ingenuity, power/imperialism or modernity at large? All those aspects are related but by no means identical.

In this teeming mass of sophisticated reasoning, monocausal explanations singling out culture, the environment or institution-building as the causal factor of last resort have not entirely disappeared. Yet there seems to be general consensus in favour of more complex models of explanation that meet the criteria of a multi-factorial design combined with parsimonious elegance. Pomeranz’s model shares these virtues with earlier contributions such as E. L. Jones’ pioneering contribution of 1981.⁸⁵ Subtle disagreements continue in regard to claims of explanatory ‘power’. Solutions to the big riddle of original bifurcation have moved from strong determination (i.e. a broad and powerful ‘Western tradition’ rooted in the Middle Ages or even Greek Antiquity⁸⁶) to weak determination through small differentials that engendered huge effects. This shift from necessity to contingency reflects a general transformation of – mainly Western – thinking from structuralism to postmodernism or poststructuralism, and also an evolution in theories of explanation as they incorporated probabilistic elements. Such a general intellectual stance, however, will not remain uncontested since those who see themselves as victors in historical struggles do not like to be told that they prevailed by mere chance. The current Chinese leadership and the scholars who happen to agree with it, for instance, insist on a deeply rooted (‘5,000 years’) path-dependency and thus on the unassailable necessity and legitimacy of the country’s ever-growing strength. The politics of explanation includes the question of how much explanatory weight one is projecting onto the past.

Much more remains to be said about the Great Divergence Debate. Addressing a major problem of world history, it is nevertheless conducted with intellectual tools pioneered by comparative sociologists from Max Weber to Charles Tilly, Theda Skocpol and Jack Goldstone. It is a truly transdisciplinary exercise. But is there anything specifically global about it? Surely the geographic scope of the cases considered is worldwide even if holistic ‘civilisations’ have been replaced by economic macro-regions as the preferred units of analysis. At the same time, the logical set-ups of explanation and the comparative procedures do not differ very much from those at closer range. Once ‘culture’ no longer counts as the major explicative variable,⁸⁷ the Great Divergence ceases to be a ‘transcultural’ issue. Thus, there is no longer a fundamental methodological gap between an intra-European comparison and one targeting different regions in China, India or Europe.

The Great Divergence Debate confirms the point that systematic comparison remains one of the most fruitful explanatory tools in the social sciences. Attempts to discredit comparison by playing it off against ‘relational’ history have failed as far as methodology is concerned. The basic compatibility of comparison and the analysis of transfer was already established two decades ago.⁸⁸ Remarkably, comparison has recently been rediscovered in the ‘soft’ humanities – for instance, in literary studies – even among those of a basically post-modernist persuasion.⁸⁹ However, outside of economic history, fields characterised by a preponderance of why-questions are relatively rare. Usually, explanation is set in wider descriptive frames. For example, it is an interesting problem of global cultural history why certain religions and some languages expand – and may even become ‘world religions’ and ‘world languages’ – while others stay local and do not ‘travel’.⁹⁰ Admittedly, these are not the central concerns of religious history and the history of languages, but they are very important issues for a history of cultural interaction that adopts a global perspective.

(3) Mixed Explanations

Between the extremes of non-explanation and pan-explanation, a wide middle ground opens up where the search for causes is mixed up with a host of other considerations. These cases are ‘global’ to the extent – one should remember David Bell’s discussion of explanations of the French Revolution – that external and ‘long-distance’ vectors are accorded special prominence in relation to internal ones. This also means that causal chains and sequences – what has been called ‘the transitivity of causation’ ⁹¹ – are usually longer than in internalist explanations. The factors impinging from the outside are often difficult to identify and trace to their origins. Characteristic, therefore, are forms of comparison that are incomplete, rudimentary or implicit and subcutaneous in a plurality of cases not strictly conforming to methodological requirements and standards. One could speak of ‘wild’ explanations, depending on the individual case, of quasi-explanations, crypto-explanations or proto-explanations, sometimes even of pseudo-explanations that qualify impolitely as ‘fake global history’. Since in the humanities the line between academic and popular forms of expression is much more difficult to draw than in the natural sciences, the rigour of explanation, comparison and other logic-bound methodical procedures can be softened in many grades and shades. James Woodward, the great philosophical authority on causal explanation, suggests some kind of ‘continuity’ between causal explanation in science and ‘causal knowledge of a more mundane, everyday sort’.⁹² In extreme cases, conspiracy ‘theorists’ concoct their own explanations of historical phenomena that can be perfectly consistent and formally rational, but based on substantially mistaken and irrational premises. Systems of delusion and closed worldviews derive their attractiveness to their true believers from a claim to be able to make sense of anything.

(4) Explanations as Counterfactual Thought Experiments

While philosophy takes counterfactuals very seriously,⁹³ manuals of historical method are likely to admonish us that they should be avoided, and no less an authority than Richard J. Evans has expressed well-considered reservations against the abuse of thought experiments for fanciful speculation about alternative pasts.⁹⁴ The genre of fictitious ‘alternate histories’ is an old one and is well-developed in contemporary popular culture. Laurent Binet’s novel Civilisations (2019),⁹⁵ in which Columbus fails and the Incas invade Europe, even won a prize from the Académie Française. Serious historians have peppered their books with speculations about China winning the Opium War and sending a punitive Armada to Britain.⁹⁶ Even so, few global historians are likely to risk their reputation with similar literary experiments. Still, one might pause and ponder whether we do not perform counterfactual thought experiments all the time. When we prepare a multi-factorial explanation of a complex phenomenon, is it not a normal, if pre-methodical, mental procedure to remove a factor – or to neutralise it to ceteris paribus status – and imagine the consequences of its deletion or disregard? Or to add another factor and see what happens? Perhaps it is worth considering Cass Sunstein’s advice to ‘dismiss counterfactual history when it is based on false historical claims’ and when it crosses the boundary between the plausible and the fantastic,⁹⁷ without rejecting it for experimental purposes: ‘any causal claim is an exercise of counterfactual history’.⁹⁸ Shouldn’t one simply add counterfactual speculation to the toolkit of historical heuristics?

(5) Explanation and Context

Almost anything can be placed within ‘a global context’ – in other words, a context of all contexts that encompasses the various national and regional contexts commonly handled in historiography. Global history could even be defined as an exercise in context maximisation. Bookshops are full of volumes on ‘X in global [world] history’. However, context as such is no virtue and no end in itself. It may be interesting to learn what happened elsewhere at the same time, or to draw parallels across the world, or to discover sources created by travellers and other eye-witnesses from afar whose existence had so far been overlooked by historians. Yet descriptive context as such does not explain anything. In each individual case, context has to be reduced to specific and traceable connections. To put it in more technical language: how does one select causally relevant contexts from among a huge repository of virtual contexts? How does one translate context into particular variables that correspond with classes of information found in the sources – in other words, variables that can be ‘tested’ empirically? How does one make claims about quantities – how much is ‘much’? – and proportions, about the relative power of impacts and the strength, stability and persistence of effects?

(6) Explaining Dynamics

While it may be correct to say that global historians tend to privilege the synchronic over the diachronic dimension – in other words, space over time – they are still keenly interested in dynamics. Global history is by no means a static discourse. The very centrality of mobility as a research topic speaks against such a suspicion. Dynamics enter the picture in two rather different shapes. On the one hand, the motive of long-term ‘change’ is being projected on the planet as a whole. Climate change and the shrinking of biodiversity are anthropogenic processes of worldwide scope. Does global history possess the intellectual tools necessary for making a significant contribution to explaining these processes? Probably, these kinds of macro-dynamics require micro-scaled and detailed analyses of their origins and consequences. On the other hand, global history is likely to be better equipped for understanding ‘diachronic’ dynamics – that is, processes that can best be observed as they move ‘horizontally’ from place to place. The frequently noted obsession of global studies with mobility and flows points in this direction. Processes of relocation and diffusion, of expansion and contraction, of the formation and metamorphosis of networks are rewarding objects of description. ‘Contagion’ has become a key term for global histories of disease and financial panics. But how to go beyond description? How to come up with accounts for motion that are neither unilinear and mechanical (A leads to B, B to C, etc.) nor tautological (mobility increases because the world is accelerating, and so on)? Would that not be a good opportunity to incorporate into explanatory models certain middle-range mechanisms and regularities of spreading and infectious connectivity drawn from epidemiology or financial market research?

That final question leads us back to the elementary options in the formal theory of history. Unfortunately, the squaring of the circle has not been accomplished: A concept of ‘analytical narratives’ was never elaborated adequately,⁹⁹ although it is intuitively obvious what such narratives might look like.¹⁰⁰ Global history – a wide umbrella covering very different approaches – cannot be content with producing narratives and, if explaining is intended at all, relying on the miracles wrought by a phantom called ‘narrative explanation’. Rather, explanation has to be made explicit as a logical procedure, with a little help from analytical theories of history, constrained as they largely have been by a fixation on a conventional history of political events. More promising are social science methodologies, especially middle-range theories, mechanisms and patterns of repetition. Such analytical devices can be incorporated into complex, though not overloaded explanatory sketches. In the event of success, global history is not just an exercise in diversity but makes a deprovincialised past speak to the future.