Chapter 4 of 16 · The great periods of history
Chapter 3. The five great periods
The division into five periods, the dates that separate them, the reason for each boundary, and what this convention files well or badly.
A filing system, and it has a name
Five great periods share out the past in books published in France: Prehistory, Antiquity, the Middle Ages, Modern Times and the Contemporary Era. In French they are called la Préhistoire, l'Antiquité, le Moyen Âge, les Temps modernes and l'époque contemporaine, and the English names used throughout this book translate those French names rather than replacing them with the divisions used in English-speaking countries. This is the division that French state education uses: the history syllabus published in 2026 takes it up in its landmarks. It is nonetheless neither ancient nor natural: nobody in the Middle Ages knew they were living in the Middle Ages, and the very name of that period designates a MIDDLE age, between two others. Such a name cannot have been given by those who lived under it.
This point deserves to be made before anything else, because it changes the way everything that follows is read. A period is not an object that is discovered, it is a filing system that is decided. The official history syllabus of the French state school system, published in 2026, counts it itself among the things to know: the division of historical time is the product of convention and is subject to debate.
A convention is not for all that a whim. This one is useful: it gives a shared vocabulary, it makes it possible to arrange a library, it saves starting the argument again at every sentence. This chapter gives it in full, with its boundary dates, the reason for each boundary, and the list of what it files badly. A tool whose faults are known is handled better than a tool believed to be perfect.
| Period | It begins at | It ends at | Approximate duration |
|---|---|---|---|
| Prehistory | The appearance of the first humans | The appearance of writing, in the fourth millennium BCE | Several hundred thousand years |
| Antiquity | The appearance of writing, in the fourth millennium BCE | 395 or 476 CE, depending on the boundary chosen | About 3,800 years |
| The Middle Ages | 395 or 476 CE | 1453 or 1492, depending on the boundary chosen | About 1,000 years |
| Modern Times | 1453 or 1492 | 1789 or 1815, depending on the boundary chosen | About 320 years |
| The Contemporary Era | 1789 or 1815 | It is not closed: it runs up to today | More than 200 years, and it keeps growing |
Three of the four boundaries in this table are double, and this is not hesitation on ODERSA's part. Only the one that opens Antiquity carries a single date. The French official syllabus itself writes its periodisation landmarks in this form: "1453/1492: end of the Middle Ages / beginning of Modern Times" and "1789/1815: end of Modern Times / beginning of the Contemporary Era (up to today)". Two dates separated by a slash mean: the cut is not an instant, it is an interval, and authors do not agree on which of the two dates best represents it.
What each boundary designates
A boundary is not chosen at random. It designates a dated fact judged to represent a wider change. The fact is precise, the change is not: it is from this gap that every disagreement about periodisation is born.
The boundary between Prehistory and Antiquity is the appearance of writing. The official syllabus places it in the fourth millennium BCE and writes it thus: the beginning of Antiquity with the birth of writing. The Bibliothèque nationale de France (BnF, the French national library) dates the invention of cuneiform writing to the course of the second half of the fourth millennium, between 3500 and 3300 BCE, in Lower Mesopotamia, and links this invention to a precise cause: the complexity of the new social order, tied to the development of the urban phenomenon, brought about the invention of a new tool of communication, namely writing.
The boundary between Antiquity and the Middle Ages keeps two dates: 395 and 476. The first is that of the division of the Roman Empire between an eastern part and a western part; the second that of the deposition of the last emperor of the West. The official syllabus keeps both and writes them together.
The boundary between the Middle Ages and Modern Times keeps 1453 and 1492. The first is the taking of Constantinople; the Bibliothèque nationale de France holds a manuscript from the second half of the fifteenth century that carries an account of it, drawn up from the report of one Jacques Tedaldi, and it has digitised it. The second is the first voyage of Christopher Columbus towards America; extracts from his logbook can still be read, published online by the Bibliothèque nationale de France, and the same year sees the making of the oldest known terrestrial globe, that of Martin Behaim.
The boundary between Modern Times and the Contemporary Era keeps 1789 and 1815. The first is that of the French Revolution, and in particular of the Declaration of the Rights of Man and of the Citizen, adopted on 26 August 1789, whose first article states: men are born and remain free and equal in rights. The second closes the period of wars that follows, and marks the moment when the states of Europe redraw their borders.
| Dates chosen | The fact designated | What is held against it |
|---|---|---|
| Fourth millennium BCE | The appearance of writing in Mesopotamia | Writing does not appear everywhere at the same moment: whole societies remain without writing for millennia, and they are then filed under Prehistory even though they are contemporary with peoples who write |
| 395 or 476 | The division of the Roman Empire, then the deposition of the last emperor of the West | Nothing changes for a peasant in Gaul the day after 476, and the Empire of the East carries on for another thousand years |
| 1453 or 1492 | The taking of Constantinople, then the first voyage of Columbus towards America | These two facts are European: they mark no break for China, India or West Africa on the same date |
| 1789 or 1815 | The French Revolution, then the end of the wars that follow it | Of the four boundaries, this is the one that rests on the history of a single country: it starts a period used far beyond France at a French event |
What this division files well, what it files badly
What it files well
- The history of western Europe, for which it was made.
- The distinction between societies whose texts we have and those whose texts we do not: it is the only boundary that rests on the nature of the sources.
- Libraries, museums and teaching syllabuses, which need stable shelves.
- Ordinary conversation: saying "in the Middle Ages" saves saying "between the fifth and the fifteenth century" at every sentence.
- Teaching, on condition that it is said to be a convention, which the official syllabus does.
What it files badly
- Everything that is not European: the boundaries of 1453, 1492, 1789 and 1815 mark no break anywhere else.
- Continuities: the life of a village changes far less from one century to the next than the name of the period does.
- Proportions: four periods share about five thousand five hundred years, and a single one covers several hundred thousand.
- Societies without writing that still exist today, which the word Prehistory files in a past where they are not.
- Slow changes: agriculture takes millennia to spread, and no boundary date can represent that.
Prehistory is the real problem of scale
An imbalance leaps out of the table of the five periods, and it has to be looked at squarely. The four periods that follow Prehistory share about five thousand five hundred years. Prehistory covers several hundred thousand, depending on where the beginning of humanity is placed.
Counting Prehistory from the oldest traces of our species, that is about three hundred thousand years, and five thousand five hundred for all the rest, Prehistory on its own represents more than ninety-eight per cent of the total duration, and the four other periods fit into the remaining two per cent. This count is a lower limit: the period begins with the first humans, well before our species, and the landmark of the domestication of fire, around four hundred thousand years ago, already falls beyond this count. Counting from the first humans would enlarge Prehistory's share still further. A division into five shares of which one is worth more than fifty times all the others put together is not a balanced division: it is a division that is mostly interested in the end.
This is not a manufacturing fault, it is a consequence of the sources. Texts are available for the last five millennia and almost nothing else for everything before them. A division made out of what is available cannot be proportional to real time. Knowing this changes the reading of every timeline: the first segment is always crushed, and it has to be restored mentally.
Specialists of Prehistory therefore have their own division, much finer, based on tools and ways of life rather than on dates. It does not replace the convention in five periods: it lives inside its first box. The next chapter gives it.
Four other ways of dividing time
By the nature of the sources
You separate what you have texts for from what you have only objects for. It is the only criterion in the convention that does not depend on a political event, and it is the one behind the Prehistory-Antiquity boundary. It has the advantage of saying what can be known, and not only what happened.
By techniques
This is the division used by prehistorians: age of chipped stone, age of polished stone, age of metals. It follows what people made, and therefore what is left of them. It neither begins nor ends on the same date in different regions, and that is exactly what is asked of it.
By long durations
Some historians prefer to follow what moves slowly: climate, roads, ways of farming, ways of counting. Seen like that, political boundaries become details, and whole periods join up across the dates of the convention.
By local histories
Every country has its own boundary dates, and they almost never coincide with ours. A division made in Europe for Europe cannot serve as a universal measure. Acknowledging this costs nothing and avoids telling the past of the whole world from a single window.
How to use the convention without going wrong
Three habits are enough, and they fit into three sentences. The first: say the name of the period and the date, never one without the other. "In the Middle Ages" means nothing to someone who does not know the convention; "in the Middle Ages, that is between the fifth and the fifteenth century" is understood everywhere.
The second: state which space you are talking about. The last four periods describe the history of western Europe. Used for China, for India, for Africa or for America, they file facts in boxes that were not cut for them, and the result is misleading even when every date is correct.
The third: treat a boundary as a signal, not as a wall. Nothing changes in people's lives on the day a period begins, since nobody knows that it is beginning. A boundary says that a change took place somewhere around that date, and the history of that change almost always occupies more than a century.
These three habits are exactly what the official syllabus asks for when it places the statement that the division of historical time is the product of convention and is subject to debate among the things to be learnt. Knowing a convention and knowing that it is one are two distinct pieces of learning, and the second is the more useful of the two.