The Amazon service is named after a late 18th-century automaton that played chess as well as a human being, or at least as well as the human being that was cleverly concealed inside the device. [Image; Refs: 1 2] Eventually it was denounced as a hoax. Its successors have been more successful. IBM's Deep Blue, for example, famously beat reigning world chess champion Garry Kasparov in 1996. Arguably, however, these later machines also contained people in the sense that human beings were required to build hardware, program software, prepare input, interpret output, and so on. Simply by ascribing meaning to computation, people become an essential part of the process. What the mechanical turks do
—either the original or its namesake—is make it obvious that there are, as one formulation puts it, "humans in the loop."So what does this have to do with history? For one thing, automata have a wonderful history of their own. The human sciences, too, have consistently drawn on contemporary understandings of mechanical or physical processes to create metaphors for cognition. In The Closed World (Cambridge, MA: MIT 1996), Paul Edwards argues that
Cognitive theories, like computer technology, were first created to assist in mechanizing military tasks previously performed by human beings. Complete automation of most of these activities—such as aiming antiaircraft guns or planning air defense tactics—was not a realistic possibility in the 1940s and 1950s. Instead, computers would perform part of the task while humans, often in intimate linkage with the machines, did the rest. Effective human-machine integration required that people and machines be comprehended in similar terms, so that human-machine systems could be engineered to maximize the performance of both kinds of components. [147]
Setting aside the history of computers and the history of people-as-computers, digital history allows people to be integrated into computation in surprising and powerful new ways. Many of these will have implications for public history. Take the folksonomies that arise from tagging. As Mills Kelly noted in a recent post in edwired, opening up archives and museums to public tagging is a new way of allowing people to find their own significance in historical sources. "Then we'll see what the archive really looks like to the user," he says, "as opposed to the archivist." Such a move could not only give curators and archivists a better idea of what their patrons think, but give professors a better idea of what their students think, and give students of social cognition a better idea of how people think. (And, of course, large repositories of tags lend themselves to data mining.) Best of all, Kelly and his colleagues will be developing their new website on the collapse of communism in eastern Europe so that users can do their own tagging.
Tags: application program interface | data mining | digital history | folksonomy | public history | tagging
