How long was the Twentieth Century? If you spend most of your time studying classical antiquity, that may sound like a trick question, but since Eric Hobsbawm published Age of Extremes: the short twentieth century in 1994, the idea that the 19th Century persisted until the shattering of the European political order in 1914 (not a new idea, of course; it’s found in Stefan Zweig’s Die Welt von Gestern, for a start) and the 21st Century began in 1989 with the collapse of Soviet communism has been widely recognised as a useful discussion point, if not as a definitive reading. There’s been a flurry of debate on this issue in the last week, with blogs on the topic from Brad DeLong and Branko Milanovic, plus multi-faceted exchange on the Twitter.*
DeLong starts with the question of whether the economic history of the 20th century is better organised around the ‘short’ (1914-1989) or ‘long’ (1870-2012) version, arguing for the latter on the basis of jumps in Total Factor Productivity and on the demographic transition; as he notes, the case for the short C20 is founded on (a) the idea that the crucial part of the Industrial Revolution had long since taken place and (b) the dominance of Marxism-Leninism and ideological struggle (which is of course Hobsbawm’s main focus). Milanovic’s response is essentially to re-assert the characterisation of the twentieth century in terms of the struggle between capitalism and an alternative political-economic philosophy, which naturally leads him back to Hobsbawm. As both note, what’s really driving this difference of opinion is a different choice of subject matter – in brief, if you want to write economic history in terms of productivity and demography, Hobsbawm’s framework makes very little sense because it’s about something else – but also with a certain hint that some approaches are more significant than others, at least when it comes to characterising the Twentieth Century as a period.
Querying periodisation is of course like shooting fish in a barrel for historians; drawing arbitrary lines in time, and then asserting that the period between two such lines is in some sense distinctive and homogeneous, is always silly. But it also seems to be inevitable. As DeLong noted on Twitter, we’re story-telling apes, making sense of the world through a rocess of selection, simplification, juxtaposition and connection of individual bits of data to create something that seems to convey understanding. What’s interesting, then, is not that we tell such stories about the past and impose some sort of external framework on events (while believing that this narrative structure is inherent to them), but what these stories and frameworks say about us.
Why centuries? I would speculate that part of the reason is the shift from understanding the past in terms of the regnal periods of monarchs or dynasties (‘the Elizabethan Age’; ‘the Victorian Era’), as a move to a more egalitarian (or at least less overtly Great Man orientated) view of history and a recognition that many important aspects of historical change are not driven by the choices of individual rulers. We can see this the way that Gibbon’s account of the (allegedly) most fortunate period in history, “the period which elapsed from the death of Domitian to the accession of Commodus”, has changed its label from ‘the Age of the Antonines’ to the Second Century. Gibbon had already indicated that this happy interlude needed to be understood in terms of fortuitous circumstances rather than attributed solely to the conscious virtue of rules, but later versions, even as they retain his overall perspective, place increasing emphasis on systematic explanations – with the stability of the Second Century doomed to be shattered not by the moral failings of Caracalla but by the Antonine Plague and the Third Century Crisis.
A focus on centuries as discrete chunks of historical development is of course entirely artificial and arbitrary – but that may be precisely its attraction. A century is longer than a normal lifespan, but a short enough period of time to be humanly comprehensible, as we’re conscious of events before our birth (our parents’ and grandparents’ histories) and can imaginatively project forward at least a decade or do beyond our own anticipated demise.The accepted chronological framework allows the coordination of all these different individual stories – though we may suspect that most people, other perhaps than professional historians, think of their own life and times more often in terms of decades rather than centuries, except around the turn of them.
Where this becomes trickier, from the point of view of historical understanding, is the subtle shift from centuries as a framework within which we locate events and developments and relate them to one another to centuries as a means of conceptualising and even explaining historical developments. Defining the 20th Century – long, short or conventional – in terms of, say, the struggle between capitalism and communism, privileges certain sorts of historical development. Further, it implies that they are the driving force behind events more generally, smuggling in interpretation in the guise of chronological arrangement, so that, when that process ends or is significantly transformed, everything else is assumed to change. A ‘pure’ century, 1900-2000, is too obviously a human construct for this to be terribly convincing; but if the period is shortened slightly or extended a little, it can appear more plausibly as the span of a ‘real’ development – that nevertheless just happens to gravitate towards the start and end points of a chronological century.
Unless you do assume that one strand of historical development – changes in productivity, or technology, or ideology – is determinative of all the others, then there’s no particular reason to assume that everything will change according to the same chronological pattern (indeed, a Braudelian approach would imply the precise opposite). The argument is not so much that Hobsbawm was wrong to think of the rise and fall of the communism-capitalism struggle as running from 1914ish to 1989ish, as that he’s wrong to pick on that as the defining theme of this entire chunk of time. If we focus on economic development, as DeLong does, then we need a different framework. If we focus on imperialism and decolonisation, as was raised in the Twitter, then we need a different framework. Ditto technology, or environmental change, or whatever. Obviously these all overlap and influence one another, in complex ways, and that’s the focus for historical investigation. What seems to be revealed in these debates is the wish for there to be a simple answer, for all the processes to line up neatly and hence, presumably, to make manifest their prime mover or determinative event.
In William Gibson’s Count Zero, the idea of When It Changed makes sense because there was such an event – but of course the point of the story was that its effects, let alone its cause, were neither immediate nor immediately obvious, and were in the first instance confined to cyberspace. The consequences of, say, 1989 took decades to become apparent – and it was equally striking how much less difference it made than had initially been anticipated. Our desire for narrative coherence wants moments like this to offer a definitive ending, or at least a new chapter; history goes on resisting such simplification.
[Update 5/12: Brad DeLong has now posted a response to part of this blog; roughly summarised, whereas I suggest above that there’s no particulat reason to suppose that different processes of historical change will neatly map onto one another (let along also map onto major chronological boundaries) unless you assume that there’s a single driving force, he argues that over the last 150 years “the pace of economic transformation has been so great as to force nearly every other aspect of history to respond according to the same chronological pattern”, i.e. there now is a single driving force, whereas previously this had been a mere fiction. Well, it’s a standard ‘tell’ of the advent of modernity (Polanyi’s Great Transformation), and as my stuff on conceptions of antiquity and modernity would imply, I’m inclined to stick to the argument that (a) we can always tell different stories about the past, even if the economic/technological one remains especially popular today, and (b) there are always multiple inter-locking processes following different rhythms, even if some are more prominent than others at particular times. But it’s a good summary of that opposing position, with some good comments below the line.]
*And isn’t this great? A combination of C18 pamphlets and the sort of erudite conversations one always imagined to take place at Oxbridge high tables and in fin-de-siècle Viennese coffee houses, in which anyone can join in. The key problem is now the temptation to stay up well beyond one’s bedtime to keep talking, while knowing that for US-based interlocutors the evening is just getting started; hence this blog instead.
In those faraway days when I taught Roman history, my method of teaching the Third-Century Crisis was to do a comparison between the essay questions set on it in University of London exam papers. They shifted from the ‘account for the TCC’ type towards the ‘why ever did anyone think there was a C in the TC?’ type. It certainly helped in showing that both crises and centuries are constructs!
Your point about centuries substituting for dynasties is intriguing. Braudel’s “Mediteranean” was about the “Long 16th Centuty”, but he felt compelled in the title to relate it to the reign of Philip II.
Absolutely – despite all his associations with the Annales school, which had led the charge against that sort of approach to history… Still in process of freeing himself from conventional thought? Or, keeping Philip in the title actually emphasised the radical nature of his new approach?
Another example, right from last night’s bedtime reading. In William McNeil’s “Venice”, the author demarcates the two centuries of Venice’s great power activity from 1282 (the Sicilian Vespers) to 1481 (death of Mehmet the Conqueror). We think like that.