Poland and Lithuania formed a single republic, the Rzeczpospolita, through the Union of Lublin in 1569. At its peak in the early seventeenth century, it was one of the largest and most populous states in Europe, dominating much of Eastern Europe. Yet between 1772 and 1795, it was partitioned three times and then disappeared from the map for 123 years. It is common in history for a great power to lose territory after defeat in war, or to become a dependent state after its power fades. But it is rare for a country to be carved up by its neighbors as if it were spoils to be divided. Even the Ottoman Empire, called the “sick man of Europe” as its system neared paralysis, was almost torn apart by the Treaty of Sevres after World War I. Ataturk’s national movement reversed that crisis, and the republic survived under the Treaty of Lausanne in 1923. Even a dying empire reconfigured and mobilized itself at the final moment. Poland also resisted, but after prolonged structural weakening, it could not even mount a real war effort. Of course, the partitions of Poland had many causes, including geopolitics, economics, and religion. Here, I approach the partitions through the lens of protocol security.

Jan Matejko, Rejtan, or the Fall of Poland (1866). The painting depicts Tadeusz Rejtan blocking the door with his body during the Partition Sejm of 1773 in an attempt to stop ratification of the partition treaty. This essay looks back to where that helplessness came from. Image source is Wikimedia Commons.

The Formation of Privilege, and the Liberum Veto

At the time of the partitions, Poland was a monarchy. But it was very different from the hereditary monarchy we usually imagine. In 1572, the Jagiellonian dynasty ended without a direct heir, and the Polish throne effectively became elective. The nobles themselves voted to choose the king. It was not like the Holy Roman Empire, where only a small number of prince-electors voted. Instead, tens of thousands of nobles gathered in fields near Warsaw and cast one vote each. Repeated elections meant repeated bargaining. Any candidate who wanted to become king had to negotiate with the nobles each time, and the nobles extracted more privileges and further reduced royal power at every election. Power flowed in only one direction. The wide-ranging privileges accumulated by the nobility became known in Poland as Golden Liberty.

As is often the case, the weakening of royal power began with war and money. A key starting point was the Nieszawa Statutes of 1454. The Polish king summoned the nobles to arms before a major war against the Teutonic Order, but they refused to follow. Pressed by the situation, the king had to accept an arrangement that required the consent of local assemblies, or sejmiks, before making new laws or imposing taxes. In 1505, the principle of “Nihil novi” - “nothing new” - was formalized, preventing the king from making any new law without the consent of the Polish parliament, the Sejm. In this way, elected kings continued to accept unfavorable conditions. The liberum veto appeared at the peak of this process.

Polish parliamentary decisions had long been governed by a norm of common consent. It was a system that required agreement by all, but it was not originally fixed as a rule that a single objection could destroy the whole process. In 1652, however, one deputy’s veto actually dissolved the Sejm, leaving a precedent in which common consent could be broken by the refusal of one person. This became known as the liberum veto, meaning “free veto.” In practice, it meant that any deputy could block action. A deputy would shout “Sisto activitatem!” in Latin, meaning “I stop this activity,” or “Nie pozwalam!” in Polish, meaning “I do not allow it.” Over time, this hardened into a destructive mechanism that did not merely block one bill under discussion, but immediately ended the session and invalidated the legislation.

One fact matters here. The foundations of this system were created before Prussia and Russia, the powers that would later partition Poland, became strong. Prussia became a kingdom in 1701, and Russia emerged as a great power after Peter the Great in the early eighteenth century. By contrast, the Polish-Lithuanian Commonwealth reached its territorial peak around 1619. In other words, Poland created this custom when it was strong, when there was no overwhelming external enemy, and when slow decision-making still seemed tolerable. A strong Poland did not design institutional change with external enemies in mind. For the Polish nobility, the realistic enemies came from inside. They were a king who might become a tyrant and rival nobles. Polish institutional change hardened precisely to block those internal enemies. Unfortunately, this made it impossible to respond effectively when the national crisis of the mid-seventeenth-century Deluge, or Potop, began to weaken the state.

Looking Through the Lens of Computer Security

When we discuss security in computer systems, two properties often matter most. One is safety, the property that “bad things must never happen.” The other is liveness, the property that “good things eventually happen,” meaning that necessary actions eventually get handled. In a banking analogy, safety means that money cannot be taken from my account without authorization. Liveness means that a transfer I try to send will eventually go through, even if it is delayed.

The problem is that although it would be ideal to preserve both, distributed networks cannot perfectly achieve both properties at the same time. If safety is pushed to the extreme and every even slightly suspicious action is blocked, normal actions also stop. Conversely, if progress is prioritized, there is room for bad actions to slip in. This is a trade-off. And it is not merely a rule of thumb, but a mathematically proven limitation. This is known as FLP impossibility, named after the three researchers who proved it. In an asynchronous network, even one faulty node makes it impossible for deterministic consensus to guarantee both “everyone agrees on the same value” (safety) and “the system eventually decides” (liveness), although randomness or timing assumptions can work around the result. FLP is about the limits imposed by crashes and asynchrony. But here we can imagine a malicious failure mode. What if a node does not merely fail, but lies or deliberately betrays the system? In computer science, following Lamport’s work, this is called a Byzantine fault. How much of this kind of fault can a network tolerate? In many BFT (Byzantine Fault Tolerance) protocols used by blockchains, safety and liveness can be guaranteed as long as malicious nodes do not exceed one-third of the total, meaning at least two-thirds remain honest. Beyond that threshold, the system collapses.

Now let us view the state as a network, and its legal order as a protocol. Each noble participating in decision-making is one node. The unanimity system containing the liberum veto was a protocol that sacrificed almost the entire legislative function, or liveness, in order to prevent bad changes, or preserve safety. The problem was that this network did not exist alone in the world. Liveness could be destroyed not only when one noble deliberately stalled or opposed action on his own. In blockchain, there is a concept called a bribery attack. If an external third party offers a node a bribe larger than the reward it would receive for behaving honestly, the node defects and both sides profit. In this interpretation of the state as a network, Prussia, Austria, and Russia were external actors that could intervene from outside the network. For them, bribing just one among the many Polish nobles who possessed the liberum veto was a cheap and effective way to paralyze the entire state. The nodes were not uniform in size, either. Some nobles enjoyed greater influence and wealth. These were the magnates. In practice, foreign powers preferred to co-opt only part of the magnate class, which made influence easy to exercise. This is the structural reason why a hegemonic state in Eastern Europe, despite its territory and population, collapsed without being able to mobilize its strength or fight anything like a proper war. Just before the partitions, Poland had a population of roughly 14 million, far more than Prussia’s roughly 5 million, but its standing army had been capped at about 20,000 since 1717, and it could not properly issue a mobilization order, leaving it helpless before neighbors that could mobilize armies in the hundreds of thousands.

Russia’s Protocol Exploit

Once Polish power had weakened, Russia maximized the vulnerability of this protocol. After the Deluge in the mid-seventeenth century and the Great Northern War in the early eighteenth, Poland’s population and economy had been damaged, and Russia had already gained enough influence to shape Poland’s kings and parliaments for decades. Instead of immediately annexing Polish territory, Russia chose to make the vulnerability permanent. In 1768, Russia coerced the Polish Sejm into turning the customary liberum veto into an immutable legal provision, and appointed itself the guarantor of that system. It hard-coded an old remnant of the system that had become a bug. Direct territorial annexation creates resistance, and resistance creates expensive war. But changing the system’s rules so that the system paralyzes itself is a low-cost form of control. The First Partition took place in 1772, but its ratification still proceeded through the Sejm in 1773 as a formality. Because the partition itself could also have been vetoed if treated as an ordinary parliamentary matter, it was deliberately handled through a special form, a confederated Sejm, where decisions were customarily made by majority vote. Later, while Russia was tied down in other wars, the Polish Sejm carried out a major protocol upgrade by adopting Europe’s first modern written constitution in 1791, and tried to abolish even the liberum veto. But conservative factions invited Russia back in and overturned the constitution, rolling the system back to an old version whose liveness was broken. At that point, the cost of protocol attack had become extremely low. Honest nodes did not even need to be bribed, because nodes already connected to the enemy could betray for free, while the enemy only had to stand behind them. For Russia, there was no immediate need to annex a Poland that had already been paralyzed into a de facto protectorate. In the end, however, Russia, Prussia, and Austria divided Poland among themselves three times between 1772 and 1795, and Poland disappeared from the map for 123 years.

What Are We Really Asking?

This kind of decision-making structure is not confined to historical Poland. The most familiar veto power today is found in the United Nations Security Council. The five permanent members can each stop the Security Council on their own. It is, in effect, a liberum veto whose holders are limited to five. It does not paralyze randomly the way Poland’s did, but the design philosophy is the same. It is built around an internal enemy, namely conflict among great powers, and therefore it can still stop in the face of external threats that would require all five to move quickly together. Decision-making structures and policies have always been part of a continuing experiment. What was right then may be wrong now. Still, one thing remains clear. Whenever we create or repair an institution, we must ask again. What does this change aim to do? What is it trying to prevent? What situation does it assume? And what remains if that assumption breaks? Poland answered the first question perfectly (the enemy is the king), but when history changed the answer, it was not even given the chance to ask again.