La longue purée

In The History Manifesto, two historians, Jo Guldi of Brown and David Armitage of Harvard, urge their peers to turn away from microhistory and go back to doing Big History in the longue durée tradition of Fernand Braudel. The book also doubles as a rant against the public influence of the natural and social sciences, particularly economics, arguing that historians can do “causal analysis” better. But the contents of the book cast doubt on the authors’ understanding of complex social and scientific issues. 

The modellers’ putsch

Whilst historians slept, Jo Guldi’s ideological enemies took over the world and she doesn’t like it. Historians once spoke truth to power. But starting in the 1960s they retreated from the longue durée into the “Short Past”. As they became absorbed in the inner lives of witches, the testimonies of slaves, and the day-dreams of landless peasants, their role as counsellers to the public on the great questions of contemporary life were usurped by economists, climate scientists, ecologists, evolutionary biologists, psychologists, and other ahistorical peddlers of grand universal models.

These modellers, hopelessly naïve about the institutional bias inherent in their “impoverished array of historical evidence”, concoct thin, static, outmoded, “essentialist”, deterministic, and patently false narratives. Their reductionist mythologies serve the interests of the powerful by monopolising the public debate on inequality, climate change, and world governance, and by stymieing the utopian solutions which history tells them are within human potential.

Indispensable skills of the microhistorians

By contrast, historians are uniquely capable of assessing diverse kinds of information streaming out of the exotic foreign country that is the past. “The arbitration of data…requires talents and training which no other discipline possesses..”, because such “skills…are often overlooked in the training of other kinds of analysts”. “[D]iscerning multiple sources of causality and ranking them” and “examining them from different perspectives and experiences to offer the fullest possible account” of complex phenomena such as climate change or income inequality are the province of historians, not scientists or economists.

Thus the abilities of a Jared Diamond, the interdisciplinarian par excellence with a command of a dozen different disciplines, pale in comparison with the “deep engagement” of a Nathalie Zemon Davis or a Robert Darnton. In “their intense reckoning with archives”, these social-cultural microhistorians of obscure, forgotten lives “had to grapple” with fairy tales, architecture, “old books and their illustrations”, and more conventional kinds of evidence, in order to write about the “shame-inducing charivaris of early modern France” and the “mystifying cat massacres of eighteenth-century Paris”. The “heights of sophistication” thus reached in the Short Past can be put to good use in our public debates about the past and the future.

But the sciences are self-correcting

I’m a big fan of The Return of Martin Guerre and The Great Cat Massacre, but would Robert Darnton or Nathalie Zemon Davis, as cultural historians, really add any more value to public debate than educated lay people in general ?

There have indeed been many thin models of universalising human behaviour based on evidence that’s historically shallow or excessively Eurocentric. But over time these tend to be self-correcting as debates rage within and between disciplines. Natural and social scientists argue about data all the time — about their provenance, their representativeness, their meaning, their limits. And they have always looked to historians with constructive and reliable methodologies for historical information.

Evolutionary psychology, for example, has matured from a field of poorly evidenced rampant speculations to one much more constrained by evidence and internal debate — enough to turn the irascible evolutionary biologist Jerry Coyne away from scepticism to cautious embrace. Steven Pinker took a pretty strong position on the historical trajectory of violence in The Better Angels of Our Nature, but there had long been a “war over war” within and between disciplines.

For decades the simple model of a single migratory radiation out of Africa had been accepted by archaeologists. Initially, genetic data reinforced this model, but the recovery of ancient DNA has shown that all non-Africans have some Neanderthal admixture ; some non-Africans have admixed with “Denisovans” ; and there must have been many hominid species across the Old World that had been undreamt of by archaeology. Thus the provisional paradigms of science keep finding complications and deviations.

Guldi cites the case of Thomas Piketty as powerful confirmation that social science is short-sighted and would be improved by the historian’s perspective. But her own intepretation of Piketty’s place in the intellectual history of economics demonstrates a fairly normal operation of science. An old hypothesis by Kuznets had held that in industrial economies inequality should initially rise with growth but diminish over time. This view was challenged by Piketty using both longer-run evidence and quite conventional economic methodology. (He has been criticised by “heterdox” economists precisely for being so conventionally neoclassical.) Model, empirical challenge, debate, more data brought to bear on the question : isn’t that how things are supposed to work ? But of course Guldi is unaware that the Kuznets hypothesis has been challenged before Piketty. For example, Gallup 2012 surveys the literature on the empirical tests of the Kuznets hypothesis and finds that the initial confirmations based on cross-sectional evidence were challenged as soon as cross-country panel data became available in the 1990s.

Historians can do “multicausal analysis” ? Who knew ?!

Here I gauge the analytical sophistication of Jo Guldi and David Armitage themselves, using a few examples from the book, as a way of suggesting that cultural-social historians of their type are ill-equipped to deal with serious policy issues. Guldi wisely stays clear of trying to argue against evolutionary psychology, except to mutter, perfunctorily, against biological determinism by noting “gender roles and systems of hierarchy show enormous variations in human history”.

(a) Example of shallow causal analysis

Guldi objects far less to climate scientists than to economists. Her problem with the former is mostly that she doesn’t much care for their “species thinking”, because it’s the Western elites and corporations who are to blame for climate change, not “humanity”. Guldi seconds Malm & Horborg‘s view that the Anthropocene only meaningfully began with the industrial era of steam inaugurated by elite capitalists, for which “at no moment did the species vote … either with feet or ballots, or march in mechanical unison”. Furthermore,

This may be a plausible argument for exonerating countries so poor as to produce virtually no carbon emissions, but a bad one for concentrating the blame on a “particular, small subset of western elite families and corporations”. Millions of consumers around the world in the 19th century did, in fact, “vote” for steam by buying cheap cotton clothing or buying other goods transported by rail or ship.

The bottom half of the US income distribution receives <20% of the total income. That’s still more than $3 trillion — a lot of carbon ! Even the bottom 5% of Americans have higher incomes than at least the bottom 60% of the world. This means, the average bottom-ventile American has got a carbon footprint many orders of magnitude greater than any Afghan or Malian villager. The bottom 5% of Germans would consume even more. And, of course, this global inequality of income was in part enabled by the Industrial Revolution.

(b) Example of naive historicism & extrapolationism

Guldi proclaims…

Markandya does have a plot of SO2 emissions per capita versus GDP per capita :

The above was not intended by Markandya as a statistical analysis of the issue, which is why he notes the “steep decline in the levels of SO2 since 1956…, without any serious impact on GDP per capita”.

Markandya did not extrapolate from the much smaller fluctuations in the longer history (unlike Guldi), because he must be aware of an important confound : the falling energy intensity of the economy resulting from technological change or an expanding service sector. During the 19th century, the amount of energy consumed per unit of GDP fell in Western Europe. From Kander, Malanima, & Warde (a book favourably mentioned by Guldi in The History Manifesto) :

Markandya is aware of this :

Nakicenovic (2002) documents the long-term “decarbonisation” (falling carbon per unit of GDP) of the US economy :

Since energy intensity clouds the causal picture, one cannot naively compare the trajectories of emissions and GDP in order to assess the impact of regulations on the latter. One might be on sounder footing in case of dramatic divergences such as after 1956 in the UK, which is what Markandya focused on.

(c) Plain old flakiness

Many passages in The History Manifesto degenerate into outright flakiness. For example, a section of chapter 3 discusses the possibilities of increased international governance. The “power of historical thinking…destabilise[s] conclusions about the best shape of institutions”. And in rapid order we are told, Samuel Huntington and Francis Fukuyama were wrong ; David Graeber has shown capitalism, historically, has used debt to put people into bondage ; the United Nations used to be more important in the 1950s and 1960s ; there has been a recent upsurge in mass protest movements and indigenous peoples’ movements ; etc. And because of these things, we know… well I’m not sure exactly, but it seems alternative, international modes of governance are more feasible than is reckoned by most people today.

Guldi also routinely commits the “genetic fallacy“, the logical error of confusing the origins of a thing with its current use, meaning, or implications. Thus, the staid metric known as “unemployment” was initially

Finally, may I suggest Guldi confuses the neo-Malthusianism of the ecologists, which really applies to all time, and that of the economic historians, which is only about the preindustrial past ?

Persson himself tells you that Clark “admits” human beings restrict their fertility decisions. Since A Farewell to Alms is cited in Guldi’s notes, she might have known Clark has got a whole chapter on fertility arguing that all human societies, contra Malthus himself, exercised “preventive checks” on fertility via delayed marriage, infanticide, birth spacing, contraception, etc.

Clark’s neo-Malthusianism relies heavily on income inequality. When per capita income is higher, this need not translate into increased population because income inequality is also higher. That means, most of the extra income is extracted by the elites and does not necessarily translate into better living standards for the ordinary person.

What’s not to like, for Guldi ?!

( The argument between Persson and Clark was not about the causes of “poverty and want”. The key bone of contention is what constitutes “subsistence income”. Persson interprets it to mean the absolute minimum for survival and Clark, any level of income, high or low, at which birth and death rates are equal. )