Why the Truth Arrives Thirty Years Late

Almost nothing in the next book was known at the time in the form it is known now.

The United States passed the Freedom of Information Act in 1966. The Church Committee reported on covert action in 1975. The CIA acknowledged organising the 1953 coup in Iran in August 2013, sixty years after the event, through an internal history released to the National Security Archive. Belgium's parliament concluded that members of its own government bore moral responsibility for the killing of Patrice Lumumba in 2001, forty years afterwards.

Call the gap the archival lag. It has three consequences, and the third one is the difficult one.

The first is straightforward. Public argument about a recent event is always conducted on the worst evidence that event will ever have. Whatever is being argued about this year will look different in thirty years, and the direction of the revision is not predictable, only its existence.

The second is that the lag changes who gets to write history. The people who were right early were right without proof, and were treated accordingly at the time. By the time the documents arrive, the argument has moved on and the vindication is a footnote.

The third is the one this chapter exists for.

The record tilts towards states that declassify. An open society generates a paper trail, an inspector general, a congressional committee, a freedom of information statute, journalists with standing to sue, and eventually a release. A closed one generates none of that. The result is that the best documented wrongdoing in the twentieth century belongs disproportionately to the countries with the strongest habits of self-disclosure, and absence of a record gets read by everybody as absence of an offence.

The United States appears four times in the next book. It appears there partly because it did these things, and partly because it writes them down, investigates itself in public, and eventually releases the files. The Soviet interventions in Hungary in 1956 and Czechoslovakia in 1968 were larger, more violent, and involved the open use of regular armies. The internal record of the decisions behind them remains far thinner, and no equivalent committee ever sat.

Neither of those facts cancels the other. That is the entire discipline, and it is harder than it sounds, because both halves have a constituency that wants only its own half quoted.

What follows practically is a reading habit. When you meet a well-documented case, treat it as the best studied rather than as the worst. When you compare a documented state with an undocumented one, notice that you are comparing two archives and not two records of conduct. And when somebody tells you that one country's misconduct is uniquely well established, ask whether that is a claim about the country or a claim about its filing system.

None of this is an argument for suspending judgement. The documents in the next book are real, they were written by the people who did the things, and they say what they say. It is an argument for knowing what kind of evidence you have, which is the only form of scepticism that survives being applied to your own side.