Pick up two books on the French Revolution. One opens in the royal court, with ministers and their letters. The other opens with a hungry crowd and the price of bread. Both are accurate about what happened. They differ in what they decided mattered.
That difference has a name, historiography: the history of how history has been written, and the study of the different approaches historians take. It is the subject that turns a reader of history into a critic of it. A graduate does not memorise a list of schools. They get a reflex: when someone tells you what caused something, ask what sort of historian they are, and what a different sort would have asked instead.
The lineup below is a tour of ten answers, across more than two thousand years and four continents. Each has its best case and its critics, and I have tried to give both.
Western historiography begins with two Greeks who disagreed about nearly everything. Herodotus (born in Halicarnassus, about 484 BCE; his Histories date from around the 430s–420s BCE) called his work an "inquiry" (historiē, the Greek word that gives us "history"). He travelled, asked, and wrote down what he heard: Egyptian priests, Persian customs, Scythian burials, tales of the Greco-Persian wars. He was candid about method: at one point he says his job is to report what is said, not necessarily to believe it all (7.152). Cicero later called him the father of history; Plutarch called him a liar. The truth is in between, and modern debate continues: some stories have been confirmed by archaeology, and the classicist Detlev Fehling (1989) argued that some of his named sources were literary inventions.
Thucydides (about 460 – about 400 BCE) was an Athenian general who failed to save a city and was exiled. He wrote of the Peloponnesian War between Athens and Sparta (431–404 BCE), left it unfinished, and set out a different ambition. He wanted a work that would be "a possession for all time" (1.22, in Crawley's 1874 translation), strictly contemporary, tested against witnesses, and focused on power and politics. His "truest cause" of the war was the growth of Athenian power and the fear it created in Sparta (1.23). His famous speeches are reconstructions that keep to the general sense of what was said, as he admits. His account of the plague of Athens (2.47–54) describes symptoms so that future readers could recognise it.
| Herodotus | Thucydides | |
|---|---|---|
| Question | What did people do, and how do they differ? | Why did the war happen, and what does it show about power? |
| Evidence | Many peoples’ reports, travel, wonder | Eyewitnesses, his own presence, constructed speeches |
| Weak point | Credulity, or invented sources (Fehling) | Narrow: leaves out women, religion, much of the economy |
| Legacy | Ethnography, comparison, the world as the subject | Political history; later read as a founder of "realism" in international relations (see 35/14) |
Notice the pattern: the first historians already disagree about what history is for. That argument never ended.
Sima Qian (born about 145 or 135 BCE, died about 86 BCE) wrote the Shiji, the Records of the Grand Historian, in 130 chapters, finishing it around 91–90 BCE. He took over the project from his father, Sima Tan, who served as court astrologer-historian. When Sima Qian defended a general in disgrace (Li Ling, in 99 BCE), the emperor had him castrated; he chose to live in order to finish the book. The structure he invented was to last: annals of rulers, chronological tables, treatises on subjects such as rites and rivers, accounts of noble houses, and a long section of biographies. These biographies include assassins, merchants and entertainers as well as ministers. Dynastic histories in China kept modelling on his form for two thousand years. He travelled, read the archives and compared accounts, and sometimes left two versions side by side when he could not decide.
His critics began early. The Han historian Ban Gu complained that he placed Daoist thought over Confucian. Modern scholars debate his early chapters, and here his method has been put to the test: oracle-bone inscriptions found in 1899 and studied by Wang Guowei (1917) confirmed most of the Shang kings in his list, while the earlier Xia dynasty has no comparable confirmation and is debated.
Ibn Khaldun (Tunis 1332 – Cairo 1406) wrote the Muqaddimah, an introduction to a world history, in 1377. He argued that historians keep repeating absurd reports because they never test them against how societies work: how many soldiers can a place feed, how do people behave in groups. His own theory centres on asabiyyah, "group feeling", the solidarity of tribes on the margins that lets them conquer settled states. Once in power, he said, the new dynasty softens in luxury, its solidarity fades, taxes rise, and a new hard group takes its place. He reckoned this cycle at about three generations of forty years, around 120 years.
A toy model of one idea, not Ibn Khaldun's full theory and not a prediction. A generation is taken as 40 years, following the Muqaddimah. Critics point out that his cycle was drawn from North African and Middle Eastern cases and does not fit every state.
Without JavaScript: with founders at 100, a loss of 40% a generation and a challenger at 30, group feeling runs 100, 60, 36, 22, so the dynasty lasts three generations, about 120 years, before the challenger is stronger.
Europeans in the nineteenth century celebrated him as a forerunner of sociology. Many Arabic and Islamic scholars note that he was also a working judge and a diplomat inside the system he described, and that he read the Greek and Islamic traditions together. Classical Social Theory (35/75) picks up his argument.
Leopold von Ranke (1795–1886) is the figure textbooks use for "scientific history". In the preface to his first book (1824) he said that history was often supposed to judge the past and instruct the present, but his work would only show "how it essentially was", wie es eigentlich gewesen. (Translators have chosen "as it actually was", but eigentlich carries more of the sense "essentially", and he may have meant something more modest than the slogan suggests.) He went to the archives of Venice and Vienna, for diplomatic reports written by people who were there. He preferred these to later narrators, and at his seminar in Berlin he taught students to compare sources and judge them. In Ranke's world the main actors were states and the Great Powers, and the main evidence came from state archives.
His critics have been grouped under a single question: can the historian ever just show it? E. H. Carr, in What Is History? (1961), said history is "a continuous process of interaction between the historian and his facts", and that facts speak only when the historian chooses to call on them. Hayden White, in Metahistory (1973), argued that historians give their material a story shape (romance, tragedy, comedy, satire) and that the evidence alone does not decide between shapes. Peter Novick's That Noble Dream (1988) traced how American historians repeatedly believed in objectivity and repeatedly lost it. The defenders reply that the facts still constrain: Richard Evans, in In Defence of History (1997), argued that some accounts are simply better supported, and that an interpretation which ignores evidence is not just another viewpoint.
Most working historians now sit somewhere between: the evidence limits what you can say, and nobody sees without a position.
Karl Marx wrote that "men make their own history, but they do not make it just as they please" (The Eighteenth Brumaire, 1852). His 1859 sketch of history said the way a society produces its living (the economic base) shapes its laws, politics and ideas (the superstructure). Marxist historians took this to mean that class conflict drives change. In France, Georges Lefebvre (The Coming of the French Revolution, 1939) and Albert Soboul read 1789 as a bourgeois revolution. In Britain, a group of Communist-party historians formed in 1946, including Christopher Hill, Eric Hobsbawm and E. P. Thompson. In Trinidad, C. L. R. James's The Black Jacobins (1938) told the Haitian Revolution as a revolution in its own right. In the United States, Eugene Genovese wrote about the enslaved and their masters in Roll, Jordan, Roll (1974).
Thompson's The Making of the English Working Class (1963) moved the focus. He wanted to rescue weavers, croppers and artisans from what he called "the enormous condescension of posterity", and argued that class is not a thing but a relationship that people make through experience, and that culture matters. That is one of the roots of social history and "history from below", which asks about ordinary people's lives, and which also grew from the Annales (next) and from women's, labour and later gender history.
The critics: that a base-and-superstructure model is too deterministic; that Soviet-era "Marxist" history was often orthodoxy with a state behind it; and that the French "revisionists" (Alfred Cobban's 1954 lecture "The Myth of the French Revolution", François Furet's 1978 Interpreting the French Revolution) said the supposed bourgeois class did not fit the evidence and the revolution was better understood through politics and language. A different criticism, from outside, is that social history can end up as a thousand local studies with no overall story.
Use the tool to see how one event is questioned by each approach.
A summary for orientation. Each school has internal disputes and none is a single voice. The books named are representative, not exhaustive.
Without JavaScript: Ranke-style political history asks what ministers decided; Marxist history asks which classes gained; revisionists ask whether those classes existed; Annales-style history asks about bread prices and structures; social history asks about the crowd; Atlantic history asks about Saint-Domingue.
In 1929 two French historians, Marc Bloch and Lucien Febvre, founded a journal in Strasbourg, Annales d'histoire économique et sociale. They wanted history that took in geography, economics, belief and daily life, not just kings and treaties. The most famous product is Fernand Braudel's The Mediterranean and the Mediterranean World in the Age of Philip II (1949), much of it drafted from memory while he was a prisoner of war in Germany. Braudel divided time into three speeds. The longue durée is the slow history of geography, climate and mentalities, which moves over centuries. The conjuncture is the middle tempo of economic and social cycles. And events are the quick surface: battles, treaties, assassinations, "foam on the waves". The deep currents, he argued, constrain what the foam can do. Later Annales historians turned to mentalities and to individuals: Emmanuel Le Roy Ladurie's Montaillou (1975) rebuilt a mountain village from the Inquisition register of the bishop Jacques Fournier in the early 1300s.
Below is an invented series of grain prices over 240 years. It has a slow rise, a roughly 35-year swing, and four bad-harvest spikes. Change the window to see which story you tell.
Invented, illustrative data (price index with a trend of 25 points per century, a 35-year cycle and four spikes). Real historians such as Ernest Labrousse used actual price series from archives.
Without JavaScript: looking year by year you see four sharp spikes (events); averaging over about 9 to 15 years removes the spikes but keeps the 35-year swing (conjuncture); averaging over 35 or more years averages the swing away too and leaves only the slow rise (longue durée).
Critics said the Annales had little room for politics, individuals and chance, and that a vision of "total history" can swallow everything. Braudel's own geographic determinism was challenged. Many of the third generation moved toward culture and away from numbers.
"Subaltern" is a word Antonio Gramsci used for groups outside the ruling class. In 1982 the historian Ranajit Guha and others in India launched the Subaltern Studies series, aiming to write the history of peasants, tribal peoples and workers as actors in their own right, and not only as followers of the nationalist elite. Guha's Elementary Aspects of Peasant Insurgency in Colonial India (1983) read colonial police and court records against their grain to find rebels' own ideas. In 1988 Gayatri Chakravorty Spivak's essay "Can the Subaltern Speak?" asked the sharper question of who is heard when the archive was written by the powerful. Edward Said's Orientalism (1978) had argued that Western scholarship made "the East" an object of its own imagining. Dipesh Chakrabarty's Provincializing Europe (2000) asked historians to stop treating Europe as the silent standard against which other histories are measured.
The critics are varied. Rosalind O'Hanlon and David Washbrook (1992) argued that the project drifted from material history toward cultural criticism and risked essentialising the people it hoped to free. Sumit Sarkar, a founding member, said in the late 1990s that it had gone away from its early aims. Of Said, Bernard Lewis and Robert Irwin argued that he misrepresented the scholars he criticised.
Global history goes in a different direction: widen the frame. William McNeill's The Rise of the West (1963) tried to tell human history as one story of contact. Kenneth Pomeranz's The Great Divergence (2000) compared China's Yangzi delta and England around 1750 and argued that they looked similar in many measures of living standards and markets. What set England apart, he said, was coal close to its factories and access to the land and labour of the Americas. Critics disagree: Robert Allen stresses Britain's high wages and cheap energy; others stress institutions, science or culture, and some argue Pomeranz's comparison flatters the Chinese side. Daniel Woolf's A Global History of History (2011) and Sebastian Conrad's What Is Global History? (2016) show how many traditions there are, and note the risks: broad history depends on others' specialist work, and can slide into elegant generalisation. The Columbian Exchange (06/09) is a good example.
Which school asked it? Test yourself on six lines I wrote in the manner of each approach.
The six lines are invented examples of the way each approach talks, not quotations.
Without JavaScript: line 1 is Annales; 2 is Ranke-style political history; 3 is social history from below; 4 is Marxist; 5 is subaltern; 6 is global history.
Histories differ because historians ask different questions: about power, class, deep structures, ordinary lives, the silenced, or the whole world. Each approach sees something the others miss and has critics who have a point. The skill is to name the approach, not to crown one.
Try this: pick a news story about a past event this week. Ask which kind of historian would tell it that way, and what one other approach would ask.
Sources. Herodotus, Histories (7.152 on reporting what is said); Thucydides 1.22–23, 2.47–54 (Crawley translation, 1874); Fehling, Herodotus and His "Sources" (1989); Sima Qian, Shiji, and Burton Watson's English translations; Wang Guowei (1917) on oracle bones and the Shang kings; Ibn Khaldun, The Muqaddimah (trans. Franz Rosenthal, 1958); Ranke, Geschichten der romanischen und germanischen Völker (1824); Carr (1961); White (1973); Novick (1988); Evans (1997); Marx (1852, 1859); Thompson (1963); Cobban (1954); Furet (1978); Braudel (1949); Le Roy Ladurie (1975); Guha (1983); Spivak (1988); Said (1978); Chakrabarty (2000); O'Hanlon and Washbrook, Comparative Studies in Society and History (1992); McNeill (1963); Pomeranz (2000); Allen on British wages and energy. The price series, dynasty model and quotes-in-the-manner-of are invented.