DEEP READING · READ 941
On the Origins of Genocide and Crimes Against Humanity · Philippe Sands · 2016
The two heaviest words the world has for prosecuting atrocity — crimes against humanity and genocide — were not deduced from legal principle. They were coined by two men who pushed them into the same indictment at Nuremberg in 1945. Those two men came from the same city, studied law at the same university under the same professor, each lost almost his entire family in the Holocaust, and argued for opposite things: Lauterpacht said the law can only protect the individual, because protecting groups means adopting the killers' own categories; Lemkin said the law must protect the group, because people are killed precisely for belonging to one. Sands tells this as a family detective story, but the argument underneath is blunt: who gets protected today still depends on which of those two men won.
Philippe Sands is a British barrister who has argued before several international courts and a professor at University College London. East West Street (2016) won the Baillie Gifford Prize, Britain's leading award for non-fiction. The form is unusual: half intellectual history of international law, half the author's investigation into his own family. Sands's grandfather, Leon Buchholz, was born in Lviv, lost almost everyone, and would never speak about it. The street in the title runs through Zhovkva, a small town near Lviv where Lauterpacht's family lived — and where Sands's own great-grandmother came from.
To see what these two charges weigh, you have to see what international law looked like before 1945. Traditional international law governed relations between states — war, treaties, borders, diplomatic immunity. What a government did to its own citizens inside its own borders was sovereignty, nobody else's business, and international law had no vocabulary for it. Lemkin liked to put it this way: sovereignty cannot mean the right to murder millions of innocent people. At the time, in practice, it more or less did.
So the real breakthrough of Nuremberg in 1945 was not that victors tried the vanquished — that is ancient. It was that for the first time, in legal form, a regime's crimes against its own citizens became something the outside world could prosecute. Once that shell was cracked, everything after it — the International Criminal Court, universal jurisdiction, the human rights treaties — had a foundation. It is also why the trial remains contested: it was a great advance and unmistakably victors' justice, with the Soviet Union on the bench and the Katyn massacre quietly set aside. Sands does not dodge this: Nuremberg is simultaneously international law's beginning and its original sin.
Lemkin, born in 1900 to a Jewish farming family in the Polish-Belarusian borderlands, studied law in Lviv. By his own account, what set his life's course was a 1921 trial in Berlin: Soghomon Tehlirian, an Armenian, shot dead Talaat Pasha, a former Ottoman interior minister and a principal architect of the Armenian genocide, and was put on trial for it. The student Lemkin asked his professor: this man is tried for killing one person, while the man he killed had a hand in killing a million and no court on earth could touch him — how does that work? The answer, in substance, was: sovereignty. (An honest note: this comes from Lemkin's own retrospective account, and some historians think he made his life look more coherent than it was. It does, though, state exactly the problem he spent his life on.)
At a 1933 conference in Madrid he proposed two prototype offences — "barbarity" (exterminatory attacks on groups) and "vandalism" (destruction of a group's culture). Nobody listened. Then in 1944, in exile in the United States, he published Axis Rule in Occupied Europe and minted a new word: genocide, welding the Greek genos (race, tribe) to the Latin -cide (killing).
The coinage was itself the strategy. His judgement was that a crime without a name does not get prosecuted and does not get remembered — that moral attention is allocated by vocabulary. That is worth a pause: the sheer political weight the word "genocide" carries today is the measure of how right he was. And his own understanding of the crime was far broader than mass killing: a group can be destroyed by murder, but also by taking its children, banning its language, burning its churches and libraries. What he wanted to protect was a group's existence as a way of living.
Lauterpacht was born in Zhovkva in 1897 and also studied law in Lviv, where both men sat in the criminal-law classes of Juliusz Makarewicz. He went on to Vienna, then to England, became professor of international law at Cambridge, served as one of the most important legal advisers to the British prosecution, and ended his career as a judge of the International Court of Justice. His life's argument was a single sentence: the ultimate unit of international law is the individual — not the state, not the nation.
He opposed Lemkin's route fiercely, and not out of coldness but out of fear. If the law takes the group as the thing it protects, it confirms in law that a person belongs first to a group — which is exactly the Nazi premise. The Nazis killed people because they had first sorted them by race; answer that in the language of races, and you have written their taxonomy into the founding code of civilisation, and possibly fed the next round of group hatred. He had a practical worry too: make the group an element of the offence and proof becomes a nightmare, because you must show that what the killer meant to destroy was the group, not the people in front of him. History has borne him out almost exactly.
So he pushed in the other direction: get "crimes against humanity" into the indictment, so that mass murder, enslavement, extermination and persecution of civilians become prosecutable in themselves — regardless of which group the victims belonged to, and regardless of whether the killers were their own government. A murdered person should not first have to prove membership of a collective before the law will take an interest. That sounds like common sense now. In 1945 it was a proposal.
| Lauterpacht | Lemkin | |
| Protected unit | The individual | The group |
| His charge | Crimes against humanity | Genocide |
| His worst fear | The law sorting people into races for the killers | The law missing why they were killed |
| At Nuremberg | In the indictment and in the judgment | In the indictment, not in the judgment |
| 1948 | 10 December: Universal Declaration of Human Rights | 9 December: Genocide Convention |
| The cost | Cannot say what is distinctive about identity-based killing | A threshold almost nobody can clear |
Two documents adopted a day apart: two winners of the same argument. Eighty years on, neither has swallowed the other.
The fourth figure is Hans Frank — Hitler's personal lawyer and later Governor-General of occupied Poland, whose territory included Lviv. The families of both Lauterpacht and Lemkin died on land he governed. Sands uses him to make a cold point: the most thorough atrocities are often administered by the people most at home in legal form. Frank was a professional jurist who kept dozens of volumes of official diaries, and those diaries became a principal source of evidence against him; in one speech he said openly that the Jews had to be annihilated (paraphrased). He was tried at Nuremberg, sentenced to death, and hanged in October 1946.
Sharper still, Sands goes to meet Frank's son Niklas — a man who has spent his life publicly repudiating his father — and Horst von Wächter, the son of another senior Nazi, who has spent his life looking for exculpatory readings of the same kind of record. Same archives, same bloodline, two sons arriving at opposite conclusions. This has nothing to do with jurisprudence and is among the strongest passages in the book, because of what it implies: faced with evidence, people first decide who they intend to be, and then decide what the evidence means.
The 1948 Genocide Convention adopted Lemkin's word and then fitted it with two locks whose consequences are still with us.
The first lock is intent. The Convention requires the intent to destroy, in whole or in part, a national, ethnical, racial or religious group as such — what lawyers call dolus specialis, special intent. It is not enough to prove someone killed; you must prove that while killing, he meant to eliminate the group itself. That is close to the hardest mental state in all of criminal law to establish: it asks you inside the defendant's head. The result is plain in the case law — convictions for crimes against humanity are numerous, convictions for genocide are rare. The first came only in 1998, in the Akayesu case at the Rwanda tribunal, and the finding of genocide at Srebrenica was litigated for years.
The second lock is which groups qualify. The Convention lists four — national, ethnic, racial, religious — and political groups were dropped during drafting, with Soviet objections part of the reason. That produces an absurdity: on the face of the text, killing you for your ancestry is genocide; killing the same number of people for their politics is not. The legal characterisation of the Khmer Rouge killings in Cambodia ran into precisely this wall.
We are left with a strange position: the gravest charge is the one hardest to establish. And that has a side effect — public debate keeps collapsing into a semantic fight about whether something "counts" as genocide, and arguing about the word is frequently the best available excuse for not yet acting. What Lauterpacht feared arrived faster than he expected.
This is the observation most worth carrying out of the book. Once genocide is treated as the crime of crimes, it establishes a ranking among atrocities, and a ranking has three effects.
First, victims are forced to campaign for a label. To have their suffering taken seriously, survivors and diplomats must prove that what happened to them was that word; anything else is second tier. Remembrance turns into a semantic lawsuit. Second, mass killing that does not fit the definition looks less serious. People destroyed for their politics or their class, or civilians killed by bombing or by famine policy, have no place at the top of this grammar. Third, it reinforces group identity itself — Lauterpacht's original worry: the more the law measures suffering by ethnicity, the more people are pushed back into ethnicity to understand themselves.
Sands has a side. He leans toward Lauterpacht, worried that a law built on groups feeds the thing it was meant to eradicate. But he concedes honestly that Lemkin caught something Lauterpacht could not: those people were not killed incidentally, they were killed for who they were, and a law that says nothing about that misses the entire meaning of the crime. The book issues no verdict, because the argument has none.
The fourth thread is Leon, Sands's grandfather: born in Lviv, arriving in Paris around 1939, silent for the rest of his life. Sands chases two family mysteries — how his one-year-old mother was carried from Vienna to Paris in 1939 by a woman nobody in the family could name (he eventually identifies her as Elsie Tilney, an English missionary), and why his grandmother did not travel with them and stayed on alone in Vienna. To the second he never gets a certain answer, only a possible explanation.
The thread is not decoration; it is part of the argument. Sands is claiming that two apparently pure legal categories were in fact shaped by particular losses — both Lauterpacht and Lemkin learned during the Nuremberg proceedings that their families had been wiped out (Lemkin lost relatives by the dozen) — and that the route each man chose is hard to separate from the way each of them lost. It is the most moving thing in the book and the thing to be most careful about: the weight of a narrative can make you feel a problem has been settled when it has only been felt.
Straightened out, the book does one thing: it uses four men and one city to turn "where did international human rights law come from" into a causal chain you can touch.
The chain runs like this. A city that changed hands seven or eight times in thirty years (Lemberg, Lwów, Lviv); one university; two Jewish students; one criminal-law professor. Twenty years later that same territory falls under a third man, Hans Frank, and the students' families are systematically murdered there. A few years after that, both students appear — each with a newly minted charge — in the same indictment at Nuremberg, where one man's word enters the judgment and the other's stays on paper. Then, on consecutive days in December 1948, each of them collects an international instrument of his own. And for the eighty years since, every argument about who deserves protection and under what name has run back and forth between the two lines they drew.
How is any of it established? By archives and legwork — indictments, trial transcripts, Frank's diaries, private letters, travel documents, photographs, plus the cities, archives and surviving descendants Sands visits himself. This is a book driven by evidence rather than theory. Its final claim is correspondingly modest: not "who was right," but "this choice has never actually been made, and every time you use that word or withhold it, you are voting on it."
Misreading 1: that genocide means "killing a great many people." It does not. Numbers are not an element of the offence; intent is. In principle killing few can qualify if the intent to destroy the group is proved, and killing very many does not qualify without it. Likewise, "ethnic cleansing" is not a crime in international law at all — it is a political and journalistic term, and the underlying acts have to be charged as crimes against humanity, war crimes or genocide.
Misreading 2: that Nuremberg convicted anyone of genocide. It did not. The word appears in the 1945 indictment, but the 1946 judgment convicted nobody on that basis — it was not yet a crime in positive law. Lemkin left Nuremberg the loser; his victory came with the 1948 Convention.
Misreading 3: that the two men were duelling rivals. Much of the book's narrative pull comes from that structure, and it should be discounted: there is no evidence the two had any substantial dealings, and their positions were less binary than the pairing suggests. Compressing a complicated scholarly disagreement into a two-man duel is a literary decision. The same discount applies to Lemkin's account of his own intellectual origins — a gifted storyteller's retrospective on his own life is not a reliable source line by line.
Criticism of the book itself. The most common concerns genre: it is memoir and detective story more than legal analysis. You finish it feeling the weight of the argument, but you learn relatively little about how the two categories are actually applied and reshaped in modern practice — for instance that crimes against humanity originally had to be linked to war at Nuremberg, a restriction later dropped. The second concerns stance: Sands is Lauterpacht's advocate more than a neutral narrator, braiding his grandfather's silence, his own convictions as a human rights lawyer and his jurisprudential preference into a single line, so that part of the persuasion runs on empathy rather than argument.
Against the concept of genocide itself, the scholarly criticism is sharper. Historians including A. Dirk Moses argue that elevating genocide to "the crime of crimes" has made other forms of mass killing look tolerable by comparison — repression in the name of security, the bombing of civilians, famine policy — because they fall outside the definition and therefore sink in the moral ranking. In other words, a word invented to make atrocity visible may make other atrocities harder to see. It is the strongest rebuttal to Lemkin's route, and it is an extension of exactly what Lauterpacht worried about eighty years ago.
The other side deserves the same honesty: crimes against humanity is no cure-all. It long lacked a dedicated convention of its own (efforts toward a treaty are recent), it is unevenly incorporated into domestic law, and by design it cannot say that these people were singled out for who they were — which, for survivors, is often the whole of it. The honest conclusion is that what each road drops is precisely what the other catches, and the world has been making do with both ever since.
1. The two heaviest words for prosecuting atrocity were not deduced from principle but coined by two men: Lauterpacht's crimes against humanity protects the individual, Lemkin's genocide protects the group.
2. Both came from the same city (Lviv), the same law faculty, the same professor's classroom, and each lost nearly his whole family — and they reached opposite conclusions. The book's methodological claim is that legal concepts grow out of biographies.
3. Before 1945 what a government did to its own citizens was sovereign business, and international law had no word for it. Nuremberg's real breakthrough was cracking that shell, not victors trying the vanquished — while carrying its own original sin, with the USSR on the bench and Katyn set aside.
4. Lemkin's founding question was simple: a man is tried for one murder while another helps kill a million and no court can touch him — on what basis? His answer was to coin a word, because a crime without a name is not prosecuted, and moral attention is allocated by vocabulary.
5. Lauterpacht's objection was just as strong: a law that protects groups writes the killers' taxonomy into the founding code of civilisation — the Nazis sorted people by race before they started killing them.
6. The Genocide Convention was adopted on 9 December 1948, the Universal Declaration of Human Rights on the 10th. Two documents a day apart: two winners of one argument, and eighty years later neither has swallowed the other.
7. The Convention fitted genocide with two locks: you must prove special intent to destroy the group as such — about the hardest mental state in criminal law to establish — and political groups were excluded, so killing for ancestry qualifies while killing the same number for their politics does not.
8. Hence the paradox: the gravest charge is the hardest to establish, and public debate stalls for years on whether something "counts" as genocide — arguing about the word being the best available excuse for not acting yet.
9. Once there is a crime of crimes there is a hierarchy: victims must campaign for a label, killings outside the definition look lesser, and the more the law measures suffering by ethnicity the more people are pushed back into it. Moses puts it harder: a word coined to make atrocity visible can make other atrocities invisible.
10. The book delivers no verdict because the argument has none: Lauterpacht cannot say what is distinctive about being killed for who you are, and Lemkin set a threshold almost nobody can clear. What each road drops is what the other catches — and every time you use that word, or withhold it, you are voting in a choice that was never finished.