The Test of Agreement

On manufactured doubt, scientific consensus, and the freedom to question

He was a lovely man. Bright, curious, newly arrived in Cambodia, and plainly excited by the prospect of discovering Angkor. Much of what he had learned about the temples, he told me, had come from YouTube.

His first question was almost painfully familiar.

“How old do you think the temples are?”

I cringed.

Not because it was a foolish question. It was perfectly reasonable. But I had heard versions of it before, and knew the kind of conversation that might follow: one in which centuries of archaeological scholarship become merely another opinion, and the extraordinary claims of a stranger on the internet are granted much the same standing as the work of people who have devoted their lives to understanding the past.

Perhaps I was being unfair. He had only asked a question.

But I wondered what had happened to the knowledge he had gone looking for.

We know a great deal about Angkor. Not everything, certainly. Its history remains full of unanswered questions, and new discoveries continue to alter our understanding of the civilisation that built it. But the broad chronology is well established. The great Angkorian capitals flourished between the ninth and fifteenth centuries. Angkor Wat itself was constructed in the early twelfth century, during the reign of Suryavarman II.

These are not dates someone simply decided upon. They emerge from inscriptions, architectural relationships, historical records, archaeological excavation and the patient comparison of evidence gathered by generations of scholars.

And sometimes that evidence surprises us.

In 2019, archaeologists published the results of excavations within the enclosure of Angkor Wat. Radiocarbon dating revealed evidence of activity in the landscape during the eleventh century, before the generally accepted construction date of the great temple.

It was an important discovery. It suggested a more complicated history of the ground upon which Angkor Wat would eventually rise. The researchers examined its implications, revised aspects of the site’s occupation history, and acknowledged that further evidence would be needed to overturn the accepted chronology of the monument itself.

This is what responsible enquiry looks like.

Yet imagine how easily the discovery could be transformed into something quite different.

Angkor Wat older than archaeologists admit.

A genuine finding becomes a sensational claim. A legitimate uncertainty concerning earlier activity becomes supposed evidence that the entire chronology of the temple is doubtful. The distinction between the age of a place and the age of a building disappears.

Into that enlarged space of uncertainty, almost any speculation can be introduced.

The process is familiar far beyond archaeology. Find a disagreement at the edge of established knowledge. Exaggerate its importance. Present the necessary caution of serious researchers as evidence that they know nothing. Then offer an explanation unburdened by the requirement to account for what has already been discovered.

The more astonishing the claim, the greater its chances of attracting attention.

And the person who has spent years carefully investigating a difficult question must now compete with someone who has spent an afternoon inventing an answer.

There is nothing wrong with speculation. Every fruitful hypothesis begins somewhere. Nor is there anything wrong with questioning established conclusions. Much of our knowledge exists because someone refused to accept an explanation that could no longer withstand scrutiny.

The difficulty begins when questioning becomes an end in itself, when the discovery of uncertainty is treated as a victory rather than an invitation to investigate further.

The political writer Jonathan Rauch describes a related form of manipulation that he calls consensus spoofing. Through intimidation, coordinated amplification or the suppression of dissenting voices, a relatively small faction can make its position appear more widely accepted than it is.

We are unusually susceptible to this. What others seem to believe influences what we feel confident enough to say, and sometimes even what we think we have seen.

If everyone in a room appears to agree, it becomes harder to disagree. If enough people fall silent, their silence begins to resemble agreement. Each person’s hesitation becomes part of the apparent evidence confronting the next.

Social media can magnify this distortion enormously. A thousand repetitions may originate from a single unsupported assertion. A handful of unusually loud voices may dominate what appears to be a public conversation. An algorithm need not distinguish truth from falsehood to learn that outrage, novelty and conflict hold our attention.

But there is an opposite manipulation, equally corrosive.

Instead of creating the impression that everyone agrees, one can create the impression that nobody really knows.

Find a few dissenting voices. Place their assertions beside the conclusions of an extensive body of research. Give each the same apparent weight. Emphasise every unresolved detail, every historical mistake, every disagreement between specialists, until the accumulated evidence of decades appears to be little more than one side of an argument.

One technique manufactures agreement. The other manufactures doubt.

Both exploit our dependence upon other people’s knowledge. Both confuse the number and prominence of voices with the reasons we have for believing them.

And both can leave us profoundly disorientated.

The temptation, then, is to distrust consensus itself. To imagine that independence of mind consists in refusing whatever most people believe.

But the habitual contrarian is no more independent than the habitual conformist. Both allow the positions of others to determine their own. One follows the crowd; the other defines himself by standing against it.

Neither has necessarily examined the evidence.

Here we encounter one of humanity’s great intellectual achievements, and one of the most widely misunderstood.

Scientific consensus is not, in its proper form, an agreement to stop asking questions. It is the provisional agreement that emerges because questions have been asked repeatedly, competing explanations have been tested, and conclusions have survived sustained attempts to discover where they fail.

Its authority lies not in how many people accept it, but in what the evidence has survived.

Peer review is part of that process, though hardly a guarantee of correctness. So are independent investigation, replication where possible, comparison between different kinds of evidence, and the willingness to revise conclusions when better explanations emerge.

Science is practised by human beings, with all their ambitions, prejudices, loyalties and mistakes. Institutions can become complacent. Entire disciplines can pursue mistaken ideas. An established consensus can be wrong.

But the achievement of science is not that scientists have learned to agree.

It is that they have developed ways of disagreeing that make knowledge possible.

An explanation must do more than appeal to our imagination. It must answer to something outside itself. What observations support it? What discoveries would count against it? Does it explain the evidence better than its rivals? Can the criticism itself withstand examination?

Not every historical claim can be tested in a laboratory. We cannot repeat the construction of Angkor Wat as an experiment. But we can investigate the traces its builders left, compare independent lines of enquiry, test proposed chronologies against new findings, and abandon explanations that fail to account for what we discover.

The essential discipline remains the same.

Our beliefs must remain vulnerable to reality.

This sounds straightforward until we consider how little of the world’s accumulated knowledge any one of us can verify personally.

I cannot examine every archaeological excavation, repeat every scientific experiment, or independently establish the safety of every medicine I take. Neither can you.

We depend upon other minds.

The question is not whether to trust, but how to trust well.

A reliable institution does not deserve confidence because it declares itself authoritative. It earns confidence through the openness of its methods, the quality of its evidence, the independence of those who examine its claims, and its willingness to acknowledge and correct mistakes.

That confidence need not be absolute to be rational.

There is a profound liberal principle here. Freedom of thought requires that no person or institution be beyond criticism. Yet the right to challenge a conclusion does not entitle every challenge to equal credibility.

A dissenter may be right against a thousand authorities. History gives us good reason to preserve that possibility. But the dissenter must still bring evidence. Being unpopular does not make an assertion true, any more than being widely accepted guarantees its truth.

And scepticism must be prepared to examine itself.

If every contrary finding is dismissed as fabricated, every qualified critic declared corrupt, and every failed prediction reinterpreted as further evidence of concealment, a belief has become impossible to correct.

It may offer the exhilarating sensation of thinking independently. But it has abandoned the very discipline that gives independent thought its value.

This is what troubles me most about the culture in which so much of our information now circulates. It is not simply that falsehood travels easily. It is that the patient means by which human beings distinguish better explanations from worse ones can themselves be made to appear suspect.

The freedom to question is precious. So is the possibility of arriving at an answer.

I have thought again about the man I met at Angkor.

I do not know what he believed about the temples. I had heard his question, not his argument. Perhaps he was genuinely asking for an informed answer, and nothing more.

My cringe, I realise, was itself a small failure of the discipline I have been defending. I had recognised a pattern, drawn an inference, and begun judging the man before I possessed much evidence about what he actually thought.

Experience may have made my suspicion understandable. It had not made it certain.

That is worth remembering, particularly when we believe ourselves to be on the side of reason.

There is so much to discover at Angkor. The builders and the kings, the inscriptions and the great reservoirs, the changing religious life of the city, the people whose houses have vanished while the sanctuaries of their gods remain standing. Each answer opens another question. Each discovery gives us a little more of a world that would otherwise be lost.

Knowing when a temple was built does not diminish its mystery. It allows us to begin imagining the real human lives that brought it into being.

The man had arrived with curiosity. Whatever he had encountered on his screen, he had come to the place itself, where there was still so much for him to learn.

His question deserved something better than my impatience, and something better than another opinion.

It deserved the pleasure of finding out.