For years the Turing Test has been used to compare humans with computers. Now sociologists are using it to compare humans with each other.
This weekend marks the centenary of Alan Turing's birth. Turing was one of the greatest computer scientist of all time. In a 1950 paper that outlined what has come to be known as the Turing Test he offered a way out of endless philosophical speculation about whether computers could ever be classed as 'intelligent.' He said that if human judges ask interview questions of a hidden computer and a hidden person and cannot tell the difference after five minutes, the computer should be considered intelligent. Nowadays, programmers compete yearly for the Loebner Prize, which is won by the computer that is most often mistaken for a human.
But the Turing Test's application is no longer limited to questions of artificial intelligence: Social scientists too are getting in on the action and using the test in a completely new way -- to compare different human subjects and their ability to pass as members of groups to which they do not belong, such as religious and ethnic minorities or particular professional classes. With the Turing Test, sociologists can compare the extent to which subjects can understand people who are different from them in some way.
In the words of sociologists, what they're now studying is called "interactional expertise." The easiest way to understand what interactional expertise entails is to contrast it with a more common idea, contributory expertise. Contributory experts are the typical array of professionals (physicists, chemists, lawyers, economists, musicians etc.) who develop specialized knowledge and skill through formal education and long experience.
Interactional experts, by contrast, are not primary practitioners. They learn about a field primarily by talking with the people who have acquired contributory expertise. The new claim is that linguistic socialization enables interactional experts to acquire enough tacit knowledge to see the world from a contributory expert's perspective. Their existence defies the cliché that understanding a person necessitates walking a mile in his or her shoes. Interactional experts can do more than talk the talk -- they can 'walk the talk' or, really, 'talk the walk' by offering authoritative technical judgments, making inside jokes, and raising devil's advocate questions that revolve around ideas normally known only to specialists.
Wherever deep interdisciplinary collaboration takes place across technical fields, interactional expertise is used.
Wherever deep interdisciplinary collaboration takes place across technical fields, interactional expertise is used. Science and technology journalists (if they go very deeply into specialist subjects) may become interactional experts, while others can be found among sociologists and historians of science and technology. Project managers need interactional expertise to excel at their jobs because it puts them in a position to understand and talk with different technical groups in a way that will generate respect. Activists may develop interactional expertise, but readily find powerful insiders are primed to use their lack of credentials against them.
While interactional expertise is not a new phenomenon, it is a new concept. Now, using a variation of Turing's test, researchers are beginning to show what interactional expertise can do.
* * *
Turing based his test on the Imitation Game, a parlor game in which men pretended to be women (or women pretended to be men) and the judge had to determine who said what. In the mid-1990's, sociologist Harry Collins started to use the idea as a research tool. In the very first experiments, Collins hypothesised that female judges would be better at spotting men pretending to be women than men judges. Not recognizing Turing's inspiration embodied the outdated gender divided society of his time, the experiment failed to reveal any meaningful differences.
In the 2000s, however, Collins, together with colleagues at Cardiff University, tried new Imitation Games to study three things: 1) whether color-blind people could pass an imitation game as color-perceivers, 2) whether those without perfect pitch -- the ability to recognise and name a musical note just as most of us can recognise and name a color -- could pretend they had perfect pitch, and, 3) whether the blind, proper -- at least those who had lost their sight in early childhood -- could pretend to be sighted. The new subject matter gave rise to interesting results that followed a particular pattern.
Consider the blind. They spend their whole lives immersed in sight-dominated societies that speak sight-dominated languages. Based on interactional-expertise theory, their exposure to this language should enable them to make the same judgments as sighted people, even where they are discussing things they have never seen, such as the bounce of a tennis ball, its relationship to the line and how hard it is to call it 'in' or 'out'. By contrast, because sighted people lack immersion in blind society, their attempts to pass as blind should come across as more caricatured than authentic. Rather than extrapolating from blind people's actual discussions of their experiences, sighted people are inclined to imagine by subtraction, guessing, unconvincingly, what it would be like to go through life without seeing.
The other cases yielded similar results: the color-blind were better able to pass as color-perceivers than vice versa, while those with perfect pitch were better able to pass as those without perfect pitch. The reversal of polarity when the color-blind were compared to the pitch-blind was exactly what would be expected. These experiments were a proof of concept, establishing the Imitation Game as a research tool that can reveal interactional expertise in action.