Troubles with Interpreters
In the twenty years since 9/11, we have seen the US military conquer Iraq and Afghanistan in a flash and with minimal casualties. On the territory of the former, it has (unwittingly, but nevertheless) given rise to the terrorist Islamic State. With great difficulty and losses, this entity was then suppressed. The infamous retreat, or rather flight, from Afghanistan is still vivid in our memories.
America, or rather the West, has unrivalled firepower, logistical support and information technology. The tactical superiority of the West is unquestionable. However, in terms of strategy, it is failing. American (or rather allied) forces are failing to win the domestic population to their side. Why? What went wrong?
If we compare Afghanistan to the roughly 50-year-old history of the Vietnam War, we find that the problem is chronic. In Vietnam, too, it was a failure to win the hearts and minds of the domestic population. The war for hearts and minds was lost long before the famous photograph of the helicopter on the roof of the embassy in Saigon was taken, and before the tragic events at Hameed Karzai Airport in Kabul, respectively.
A book by American historian Jill Lepore has recently been published with a remarkable theme. The book, called 'IF/THEN', is devoted to the story of Simulmatics. You've probably never heard of it. The firm, founded in the early 1960s, initially advised presidential candidate J.F. Kennedy. Based on a mathematical-statistical analysis of polling data, it advised Kennedy to be vocal about civil rights for the black population and to speak out about his affiliation with the Roman Catholic Church.
It was the right tactic, and it won Kennedy the election. Lepore notes that there is no evidence that Kennedy was actually fundamentally influenced by computer data analysis, but one way or another, Simulmatics entered the big world and began bidding for major government contracts.
By the mid-1960s, it was obvious that the Vietnam War was not going well. Simulmatics was commissioned to research and analyze the opinion of the Vietnamese public. Including suggestions on how to influence that opinion. A branch office was set up in Saigon, which at its height had dozens of employees. Of course, the company had to do its own public opinion research in war-torn Vietnam. It went about it in much the same way as it had done at home: it identified social groups and subgroups (there were 480 of them in the US) and sent its people into the field with questionnaires.
A team of researchers (Americans and local interpreters) went to work by military helicopter or automobile. Its members usually stayed overnight in American military camps. They even wore uniforms that, while different from US Army gear, looked obviously military. Then they went among the people and asked questions that would have sounded reasonable back home in America, but were inappropriate, even absurd, under the local circumstances. For example, "When you think about what is important in your life, what are your hopes and aspirations for the future?"
Most people answered, "I wish for peace."
When asked a follow-up question about what peace should look like -- whether peace under the Americans or under the Communists -- people were reluctant to answer. In general, respondents answered the way they thought the researchers would want them to answer.
It was clear to some of the Vietnamese employees at Simulmatics that such research was worthless. Who would answer delicate questions quite honestly to someone who had just stepped out of a military helicopter? "Certain standard U.S. public opinion research practices create the perception among Vietnamese peasants that this is a police investigation," argued one local Simulmatics employee.
The research suffered from other methodological flaws as well. Especially in the later stages of the work, when interpreters with a poor command of English were hired, who, moreover, abbreviated or distorted the questions in all sorts of ways. "The questions in the questionnaires are difficult to translate into vernacular Vietnamese," was another objection. Its author was Joseph Hoc, a Vietnamese Catholic and staunch ally of the Americans.
Hoc managed to infiltrate the social networks of several villages, including several under Communist control. He came up with a proposal to influence local shamans, holy men who had considerable influence in the countryside. (The proposal was unsuccessful.) But even he was not a completely credible source of information -- precisely because he was so desperate for the Americans to win. In general, it appeared that the interpreters were, quite logically, on the side of the Americans and therefore painted reality pink.
Simulmatics therefore continued to send optimistic messages to Washington, and the Department of Defense was satisfied. However, in reality, on the battlefield, things were going less and less well. Simulmatics finally went bankrupt in 1970.
The statistical methods that Simulmatics developed live on in the algorithms used by Facebook, Google, Amazon, and many others. However, when it comes to Afghanistan, it is clear that Washington has a lot of catching up to do in researching the opinions of nations in conflict zones.