How to Read Ancient Primary Sources

Reading an ancient source is a strange kind of conversation. The person speaking to you may have lived more than two thousand years ago. The text may survive only because generations of people copied it by hand. The edition in front of you may have been reconstructed from manuscripts written centuries after the author died, and unless you read the original language, somebody has already made thousands of decisions about how those words should sound in English.

That sounds like a lot of distance between us and the ancient world, and it is. But these sources are still one of the main ways we get there.

Herodotus and Thucydides. Livy and Tacitus. Caesar writing about his own campaigns. Plutarch writing about people who had been dead for centuries. Speeches, letters, poems, inscriptions, laws, coins, papyri, graffiti and the things archaeologists recover from the ground.

The important thing is not working out which sources are “good” and which ones are “bad”. It is working out what each source can actually tell you.

Primary does not mean true

One of the easiest mistakes to make is to hear the words primary source and assume that means especially reliable.

It doesn’t.

A source can be contemporary with an event and still be mistaken, incomplete, self-serving or deliberately misleading. An eyewitness only sees what is in front of them. A politician has interests. A general may have a reputation to protect. A public inscription may have been commissioned precisely because somebody wanted a particular version of events remembered.

The opposite is also true. A writer living centuries later may preserve information taken from earlier works that have since disappeared.

So it is better to think of primary source as a description of the source’s relationship to the question you are asking, rather than a stamp saying “trust this”.

Homer is a good example. The Iliad is enormously important evidence for Greek epic tradition, ideas about heroism and the social assumptions built into that tradition. It is not a contemporary battlefield report from the Trojan War. If we are trying to reconstruct the Late Bronze Age world behind those stories, archaeology and contemporary documentary evidence have to do most of the historical work.

The same text can therefore be brilliant evidence for one question and poor evidence for another. Once that clicks, ancient sources start making much more sense.

First work out what you are actually reading

Before you ask whether a source is believable, ask something more basic:

What kind of thing is this?

A history is not the same as a speech. A biography is not the same as a legal decree. Epic poetry is not an inscription. A private letter was written for a very different reason from a monument set up in the middle of a city.

Genre matters because it changes what the source is trying to do.

Herodotus investigates wars, peoples, customs, geography, stories and competing explanations. Thucydides builds a tightly argued history around a particular war. Plutarch is interested in biography and character. Aeschylus is writing tragedy for performance. An emperor listing his achievements on stone is making a public statement about his rule.

You cannot read all of those in the same way.

Ancient writing also does not sit neatly inside the categories we use today. Historians could be literary writers as well as investigators. Biography could carry moral lessons. Poetry could preserve memories of the past. Political speeches could become literature.

The label on the book helps, but it is only the beginning. You still need to ask what the author was trying to make that kind of writing do.

Ask how the author could know

Once you know what sort of source you have, ask one of the most useful questions in the whole process:

How did this person get their information?

Sometimes the answer is fairly straightforward.

Julius Caesar took part in the campaigns he describes in the Gallic War. Thucydides lived through the Peloponnesian War and was personally involved in it. Josephus fought in the Jewish revolt against Rome before surrendering and later writing about the conflict.

That proximity matters. They knew things that somebody writing generations later could not know firsthand.

But proximity does not solve everything.

Take Caesar. He knows his army, his campaigns and many of the decisions being made because he was the commander. That gives his account enormous value. It also means the man explaining Caesar’s conduct is Caesar.

So when you read him, you do not need to decide whether he is either an eyewitness or a propagandist. He can be both a very well-informed witness and someone with strong reasons to shape the story.

Ask what he genuinely knew firsthand. Ask what he chooses to explain. Notice how responsibility is distributed. Look at how his opponents are described. Pay attention to what gets passed over quickly. Then compare his version with whatever else survives.

That is usually more useful than giving the source a simple reliability score.

Other authors are much further removed from what they describe. Arrian wrote about Alexander the Great roughly four centuries after Alexander’s death. That sounds like a serious problem, and it is something we have to take into account. But Arrian also drew heavily on earlier accounts, particularly those associated with Ptolemy and Aristobulus, whose histories are now largely lost.

He was not an eyewitness. But simply counting the centuries between author and event does not tell us enough. What matters is the route by which the information reached him.

Herodotus shows why this gets complicated

Herodotus is one of the best writers to learn this with because the Histories constantly reminds us that information has to come from somewhere.

He tells us about places he visited, people he questioned and stories he collected. He preserves local traditions, gives different versions of events and sometimes tells us which one he finds more convincing. At other times he passes on a story while making it fairly clear that he is not entirely sold on it himself.

That does not mean everything in Herodotus happened exactly as he says. It does mean that reading him becomes much more interesting once you stop trying to sort every passage into “true” or “false”.

Suppose he gives us a story about something that happened generations before his own lifetime. You can ask who might have told him that story, how far he was from the original event, whether different communities remembered it differently, and whether another source gives us another version.

Even when we cannot prove that the event happened exactly as told, the story may still tell us something important about the people who kept telling it and what they thought the past meant.

Oral tradition does not become worthless because it travelled by word of mouth. It just means we have to remember that stories can change as they move from person to person and generation to generation.

Herodotus is evidence for events, but also for memory, storytelling, cultural encounters and the way people in the fifth century BCE were already arguing about the past.

Bias is not the end of the conversation

“Biased” gets used so often in discussions of ancient sources that it can become almost meaningless.

Of course ancient writers had perspectives. So do modern writers. Everybody selects information, decides what matters and tells the story from somewhere.

The useful question is what that perspective is doing.

Tacitus wrote about the Roman emperors as a senator deeply interested in power, political freedom, corruption and the relationship between emperors and Rome’s governing elite. That matters when you read him because it shapes the things he notices and the questions he keeps coming back to.

Recognising that perspective does not mean we put Tacitus down and declare him unusable. It helps us understand what we are reading.

The same goes for Caesar. His account of Gaul matters partly because he was there, and partly because he had reasons to shape how those events were understood.

Even propaganda can tell us a great deal. An emperor’s monument may not give us an impartial assessment of his reign, but it can show us how he wanted that reign to be seen. A hostile portrait of a ruler may preserve exaggeration and gossip, but it can also show us which accusations had political force.

So instead of stopping at “this source is biased”, try asking:

“What does this writer want me to see?”

That usually tells you more.

Thucydides reminds us that speeches are not transcripts

Ancient histories often contain long speeches.

They can feel wonderfully immediate. Suddenly Pericles, a Spartan envoy or an Athenian general seems to be standing in front of us, speaking straight out of the fifth century BCE.

That feeling is useful, but it can also mislead us.

Thucydides tells us that reproducing the exact words of speeches was difficult. His explanation suggests that he aimed to give what was appropriate to the circumstances while remaining as close as possible to the general sense of what had been said. Scholars still argue over exactly how that should be understood in practice.

What we should not do is treat the speeches as if somebody had been taking shorthand notes in the assembly.

When you get to one of them, it helps to read on two levels at once. What might this tell us about the political arguments being made at the time? And why has Thucydides placed this speech here, in this form, at this point in his history?

The speech may tell us something about the speaker. It also tells us something about Thucydides.

Livy was writing about a past that was already ancient

Livy gives us a different problem.

He was writing during the age of Augustus, but the beginnings of Rome lay many centuries behind him. Even for Livy, early Roman history belonged to a very distant past.

For later periods he could draw on a substantial Roman historical tradition. Further back, the evidence becomes much harder to handle. Earlier historians, family traditions, public records, legends and stories all fed into the way Romans understood where they had come from.

That means a story in early Livy may be useful in more than one way.

It might preserve something about an early event. It might preserve traditions that developed around that event. Or it might tell us most clearly about what later Romans thought courage, citizenship, family, religion or political conflict ought to look like.

Those questions are related, but they are not the same.

Once you get comfortable with that, early Roman stories become much more useful. You no longer have to choose between taking them literally and throwing them away. Sometimes the way a society remembers its past is part of the history too.

Plutarch is interested in character

Plutarch makes more sense once we stop expecting him to write modern biography.

His Lives contain wonderful material about Alexander, Caesar, Antony and many others, but most of his subjects lived long before he did. He also tells us quite clearly that character matters to him. A small remark, habit or personal anecdote may interest Plutarch because he thinks it reveals something important about the person.

That is why his biographies are full of memorable stories.

Those stories can be historically valuable, but vividness is not proof.

If Plutarch gives you an unforgettable anecdote about Alexander, ask where it might have come from, whether another source knows it and what part it plays in Plutarch’s picture of Alexander.

Sometimes he preserves material from earlier works we no longer have. Sometimes a story is impossible to verify. Either way, it can still tell us something about how that person came to be remembered.

There are times when “Why was this story worth preserving?” is just as useful a question as “Did this happen exactly like this?”

Silence can matter — but it is easy to misuse

Ancient history survives unevenly.

Some emperors have enormous literary traditions. Other rulers barely appear. We know a great deal about certain wars because an important historian happened to write about them, while huge parts of ordinary life receive very little attention from elite authors.

That creates a trap.

If a surviving writer does not mention something, it is tempting to assume it did not happen or did not matter.

Sometimes that is reasonable. Sometimes it is not.

Ancient writers had their own interests. Texts were lost. Archives disappeared. Literacy was uneven. The people whose books had the best chance of being copied for centuries were not representative of everyone who lived in the ancient world.

So a gap in the literary record is not automatically a gap in ancient life.

This is one reason inscriptions, papyri and archaeology matter so much.

Ancient people left us much more than books

When people hear ancient primary sources, they usually think first of famous authors. That makes sense, but literature is only part of the evidence.

Historians sometimes use primary source broadly, though it can be useful to separate written sources from the wider body of primary evidence that survives from antiquity.

An inscription might preserve a law, a military career, a dedication, a treaty, a gravestone or a public honour.

Papyri can give us tax returns, petitions, contracts, accounts and private letters: ordinary paperwork that was never meant to become history.

Coins can help with chronology, rulers, political imagery, circulation and monetary activity.

Graffiti gives us jokes, insults, names, advertisements, political slogans and declarations of love.

Archaeology gives us houses, rubbish, human remains, roads, workshops, temples, fortifications, pottery, food and whole landscapes.

This kind of evidence is especially important because it can bring us closer to people who barely appear in elite literature at all.

It can also complicate what the literary sources tell us, which is often a good thing.

If an ancient writer gives us one picture of a city and archaeology gives us another, we do not necessarily have to choose which one is “right”. They may be looking at different things, describing different periods or seeing the same society from completely different positions.

Some of the strongest ancient history comes from putting different kinds of evidence next to each other and seeing what happens.

The text in your hand has a history too

There is another layer between us and an ancient writer that is easy to forget.

For most classical works, we do not have the author’s original manuscript.

The text survived because people copied it, and then somebody copied that copy, and so on.

Different manuscripts can preserve different readings. Words can disappear, move, be repeated or become corrupted. Modern editors compare surviving manuscripts, quotations in other authors and other evidence in order to establish the best text they can.

So when you pick up a modern edition of an ancient work, you are not holding an untouched object passed straight down from antiquity. You are looking at the result of a very long history of copying, survival and scholarship.

Then translation adds another layer.

Greek and Latin words do not always have one neat English equivalent. Translators have to make choices about tone, rhythm, technical vocabulary, ambiguity, names and sentence structure. Two excellent translations can therefore sound quite different without either translator necessarily being wrong.

You do not need to become suspicious of every sentence you read. Just remember that the English wording is not invisible.

If a particular passage really matters, comparing two translations can tell you a surprising amount.

“Can I trust this?” is usually too simple a question

It is natural to ask whether we can trust an ancient source, but the question pushes us towards two answers: yes or no.

Ancient evidence is rarely that tidy.

A historian may be poor evidence for the exact wording of a speech but very good evidence for the political arguments surrounding it.

A biography may not prove that an anecdote happened exactly as told, but it may show us the reputation someone acquired.

Imperial propaganda may be useless as an impartial assessment of an emperor’s achievements while being excellent evidence for how imperial power wanted to present itself.

A poem may not reconstruct an ancient battle, but it can tell us a great deal about the society that kept telling stories about war.

So rather than asking whether a source is reliable in the abstract, ask something narrower:

Reliable for what?

That small change is one of the most useful habits you can develop.

Put the sources next to each other

Very few difficult questions in ancient history should depend on a single piece of evidence.

If Herodotus gives one version of events and another tradition gives us something different, that disagreement is worth looking at.

If the archaeology does not fit the literary account neatly, pay attention.

If two ancient writers tell essentially the same story, ask whether they are actually independent witnesses. One may be repeating the other, or both may ultimately depend on the same earlier source.

And if a later author preserves something an earlier surviving writer never mentions, the gap is something to investigate rather than a reason to accept or reject it automatically.

Once you start comparing sources, the evidence becomes much more interesting. You stop seeing ancient texts as separate containers of facts and begin to see how knowledge about the past was gathered, passed on, argued over and reshaped.

Seven questions to ask when you read an ancient source

You do not need to carry out a full scholarly investigation every time you open an ancient text. These seven questions will take you a long way:

1. What kind of source is this?
   History, biography, poetry, speech, inscription, letter, law, religious text, archaeological object?
2. When was it created?
   At the time, a generation later, or centuries afterwards?
3. How could the author know?
   Personal experience, eyewitnesses, archives, earlier writers, oral traditions, or sources that are now lost?
4. Who was it made for?
   A public audience, political supporters, fellow elites, worshippers, one private recipient?
5. What is it trying to do?
   Record, persuade, praise, condemn, entertain, commemorate, explain, justify?
6. What is missing?
   Whose voice do we not hear? What might the author not know, not care about or prefer not to tell us?
7. What happens when we compare it with other evidence?
   Other writers, inscriptions, coins, papyri, archaeology or material remains?

You will not always be able to answer all seven. That is fine. Sometimes knowing where the uncertainty is tells you almost as much as having a firm answer.

Read critically without becoming cynical

There are two easy ways to get ancient sources wrong.

One is to believe everything because the writer lived closer to the ancient world than we do. The other is to believe nothing because every writer had biases, agendas and limitations.

Neither approach leaves us with much history.

Ancient authors could be careful investigators, wonderful observers, political partisans, storytellers, sceptics, gossips and brilliant analysts. Sometimes the same author can be several of those things on the same page.

You are not reading just to catch them out. You are trying to understand what kind of evidence they have left us.

Herodotus becomes more interesting when you notice how he handles different traditions. Thucydides becomes easier to use once you understand what his speeches are doing. Livy becomes richer when you separate early Roman events from later Roman memories of those events. Plutarch makes more sense when you remember how much he cares about character. Caesar becomes more revealing when you keep his firsthand knowledge and his self-presentation in view at the same time.

That is the habit worth developing.

Read the source. Work out what it is. Ask how the author knew what they knew and why they were telling the story this way. Put it beside the other evidence. Notice what agrees, what doesn’t and what you still cannot know.

The ancient world did not leave us a complete record of itself. What survives is patchy, uneven and sometimes frustrating.

That is not a weakness we can remove from ancient history.

It is part of what makes learning how to read the evidence so important.