Rome's republic was designed by people who had thrown out a king and were determined not to get another one, and almost every feature of it is a precaution against a single individual acquiring power.
At the top were two consuls instead of one, holding office for a single year, each with the power to block the other. Below them a ladder of offices, each also annual, each also held by more than one man. Nobody could stand for the same post again for ten years. An army command expired when the year did.
Steering all of it was the Senate, a few hundred men who had held those offices and now sat for life. Formally the Senate only gave advice. In practice it controlled the treasury and the conduct of foreign policy, because a consul with twelve months in hand cannot fight a war that lasts three without the body that votes the money and remembers what happened last time.
And set against the Senate were the tribunes of the plebs, ten officials elected by the common citizens, whose bodies were declared sacred so that harming one was a capital offence, and whose central power was the ability to say no. A tribune could halt a law, an election, a levy, or the act of any magistrate, by standing up and forbidding it.
A Greek observer named Polybius, who lived in Rome as a political hostage in the second century, thought this was the secret of Roman success. He described it as a mixture of monarchy in the consuls, aristocracy in the Senate and democracy in the assemblies, each restraining the others. It is an elegant analysis, it has influenced constitution writers ever since, and it was produced by a foreigner explaining to defeated Greeks why the people who had beaten them deserved to win.
The republic did not begin as a partnership. For its first two centuries the offices, the priesthoods and the interpretation of the law were held by a closed group of families, the patricians, and everyone else, the plebeians, was locked out.
What the plebeians had was numbers and the fact that the army was raised from them. The tradition says they used it directly: on several occasions they simply walked out of the city in a body and camped on a hill outside, refusing to muster until concessions were made. The first of these withdrawals is dated to 494 before Christ, the last to 287.
The gains came slowly and they were real. Written law, so that judgements could not simply be announced by men who claimed to know the custom. Access to the consulship. Legal marriage between the two orders. And finally, in 287, a rule that decisions of the plebeian assembly bound the whole people, patricians included.
The written law came first and mattered most. Around 450 a commission produced the Twelve Tables, which were put up in the forum where anybody could go and read them. The text has not survived; we know it from quotations in later writers, and what they quote is a mixture of procedure, family law, debt, property and some startlingly harsh penalties. The point was not that the law was mild. The point was that it was published.
What emerged from all this was not democracy. It was a new governing class of the richest plebeian families merged with the old patrician ones, and the assemblies were weighted so that wealth voted first and in smaller units. But it was a settlement that most citizens had a stake in, and that stake is why Rome could ask so much of them.
Rome conquered Italy over roughly a century and a half, in a long series of wars against Latin neighbours, hill peoples, Greek colonies in the south and Celtic tribes in the north. That is not unusual. Plenty of ancient states conquered their neighbours.
What was unusual was what it did with them afterwards.
A Greek city treated its citizenship as a closed inheritance. Athens tightened its own rules in the fifth century so that both parents had to be Athenian. Rome went the other way. Defeated communities were bound to Rome by individual treaties, and the treaties varied: some places got full Roman citizenship, some got citizenship without the vote, some remained formally independent allies. Individuals could be granted citizenship. Freed slaves became citizens, which astonished Greek observers.
The allies did not pay tribute. They supplied soldiers. That is the mechanism that explains everything that follows, because it meant every conquest enlarged the pool of manpower available for the next one, and a defeat that would have finished a city state merely cost Rome a season.
It was not generosity and the allies did not always think it was a good bargain. They fought and died in Rome's wars, took a smaller share of the spoils, and had no vote on whether the war happened. When they finally revolted over exactly that in the first century, the war was savage, and Rome ended it by conceding the citizenship the allies had been demanding.
Carthage was a Phoenician foundation on the North African coast, older and richer than Rome, with a commercial empire and a serious navy. Between 264 and 146 the two fought three wars, and the second of them is the one that nearly ended Rome.
In 218 a Carthaginian general named Hannibal brought an army from Spain over the Alps into Italy, which nobody expected, and then destroyed three Roman armies in three years. The worst was at Cannae in 216, where he let his centre give ground until the Roman infantry had pushed into a pocket, then closed the wings and the cavalry behind them. The Roman army was surrounded in the open and killed where it stood.
Ancient writers put the dead at somewhere between forty eight and seventy thousand in a single afternoon, and those figures cannot be checked, but the scale is not seriously doubted. Rome lost a large part of its political class in that pocket, including a serving consul.
And it did not surrender. It refused to negotiate, refused to ransom the prisoners, raised new legions including some from slaves and men below the property qualification, and then avoided giving Hannibal another battle for years while wearing him down and attacking Carthaginian Spain instead. Most of the Italian allies stayed loyal, which is the return on the settlement described above. Hannibal won every battle he fought in Italy and could not convert it into anything, because he could not detach Rome's manpower from Rome.
The war was decided in Africa in 202 when a Roman commander who had learned from Hannibal used Hannibal's own encirclement against him at Zama. Fifty six years later Rome went back and destroyed Carthage entirely, killing or enslaving the population and demolishing the city. In the same year it destroyed Corinth, in Greece, as a warning.
By the middle of the second century Rome controlled the Mediterranean, and the republic began to come apart, and the two facts are connected.
The wars had been fought by citizen farmers serving long campaigns far from home. A man away for six years could not work his land, and many came back to find it sold. The wars had also produced captives in enormous numbers, and those captives were put to work on large estates owned by the men who had done well out of the conquest. So the small farms that had supplied the legions were being consolidated into big holdings worked by slaves, while the men who had won the wars drifted into the city.
That is the classical account, and archaeology has complicated it. Field surveys in parts of Italy find small farms persisting rather than vanishing, and the picture now looks more regional and less uniform than the ancient writers claimed. What is not in doubt is the arrival of slavery on a new scale, or the political crisis it fed. There were two enormous slave revolts in Sicily within forty years, and a third rising in Italy under a gladiator named Spartacus that took Rome two years to put down.
Provincial government made it worse. A governor held his province for a year or two, was effectively unsupervised, and had spent a fortune getting elected. Taxes were farmed out to private companies that bid for the contract and kept the surplus. The system produced spectacular extraction and occasional prosecutions, and the prosecutions were themselves weapons in Roman politics.
In 133 a tribune named Tiberius Gracchus proposed to enforce an old limit on how much public land any one man could hold, and redistribute the rest to citizens. It was a legal proposal, aimed at rebuilding the class the army recruited from. He pushed it through by unconstitutional means when it was blocked, and a group of senators led by his own cousin clubbed him to death in the street with the legs of broken benches.
That was the hinge. For nearly four hundred years Roman politics had been fought with vetoes, prosecutions and procedural obstruction. After 133 it was fought, increasingly, with violence, and everybody had seen that it worked.
Firm: the second century onwards. From the Punic Wars we have Polybius, who interviewed participants, and a growing body of inscriptions, treaties and coinage. The institutions, the treaty system and the wars are solid.
Very soft: everything before about 300 before Christ. Our continuous narrative of early Rome was written by Livy under the first emperor, four hundred years after the events, working from earlier writers who were themselves working from family traditions and priestly lists. The expulsion of the kings in 509, the details of the secessions, the heroic individuals: these are national legend with an uncertain historical core, and Roman writers of the late republic freely improved them to make points about their own politics.
Soft: the Twelve Tables. We have no copy. Everything quoted from them comes through authors writing centuries later, some of whom may be paraphrasing or misattributing.
Soft: the numbers. Cannae is the clearest case. Two ancient authorities differ by more than twenty thousand dead, and both had reasons to prefer a large figure.
A myth worth naming, because it is repeated constantly: the Romans did not sow the ruins of Carthage with salt. No ancient source says they did. The story appears in modern books from the nineteenth century onwards and has been traced to a scholarly guess that hardened into a fact. What they actually did was bad enough without it.
Contested: whether the republic was destroyed by its empire or by its own elite. The first reading says the institutions of a city state could not administer a Mediterranean and were bound to break. The second says the institutions were doing tolerably until a governing class decided that winning mattered more than the rules. Both readings are held by serious historians, and this book leans on neither, because the next chapter is where the question gets settled by events rather than by argument.