The advantages of amnesia
From the Internet to the iPod, technology is bringing rapid advances in memory. What society needs now are new ways to forget.
IMAGINE CARRYING AROUND an entire research library on an iPod. Such a feat suddenly seemed feasible as of earlier this month, with the news that
Parkin's development, which should reach consumers within a few years, is the latest installment in our eternal quest to preserve and supplement the human memory, which has taken us from the cuneiform slab to the cassette tape to the Ginkgo biloba tablet washed down with the day's first cup of coffee.
As digital-storage capacities reach seemingly boundless proportions, however, some thinkers are becoming nervous about the unintended consequences of memory technology. Certainly Google's enormous reserves of user information, stored in dozens of secretive data centers across the world, and the literally photographic memory of the Internet Archive, which preserves billions of defunct Web pages for posterity, are enough to leave anyone rattled. New forms of memory are permanent and accessible from anywhere. As their reach grows, scholars are asking if now - perhaps for the first time in human history - we need to find ways to forget.
"We used to have a system in which we forgot things easily and had to invest energy in remembering," says Viktor Mayer-Schönberger, an associate professor of public policy at Harvard's Kennedy School of Government. "Now we're switching to a system in which we remember everything and have to invest energy in order to forget. That's an enormous transformation."
Jorge Luis Borges envisioned the risks of perfect memory in his famous story "Funes the Memorious," about a man gifted with unlimited recall, and paralyzed by it. Perhaps not even Borges, however, could have imagined our present capacity to accumulate and preserve memory in digital form - or the powerful impact it is already having on individual lives, as temporary indiscretions become part of the permanent record. "What you do online is potentially there forever," says Coye Cheshire, an assistant professor at the School of Information at the University of California, Berkeley. "Delete if you want; ask Google to take down that one unflattering photo - but it's still saved, archived, somewhere."
The personal costs of this reality are clear, but there may be broader social costs as well. "What a lot of people forget - no pun intended - is that forgetting is hard-wired," says Mayer-Schönberger. "Cognitively and sociologically, we've never had to develop the capacity to forget or to put things in temporal perspective, because forgetting was built in biologically."
In a working paper posted online last spring, "Useful Void: The Art of Forgetting in the Age of Ubiquitous Computing," Mayer-Schönberger hypothesized a number of future big-picture costs of an unforgetting era. The assumption that every online comment and transaction is preserved somewhere, never to be forgotten, could suppress public speech and civic participation in ways that we could never calculate. There is also the contradiction whereby "personal" information - from private e-mails to sensitive identifying data - is indefinitely available on a remote server.
In "Useful Void," Mayer-Schönberger argues that we can - and should - retrain ourselves to forget. He argues in support of "data ecology," whereby a combination of law and software would apply an expiration date to certain types of sensitive data. Search engines, online vendors, cellphone companies, surveillance-camera services - all would operate under a legal obligation to wipe their data after a certain span of time.
Of course, such rules come with a significant loophole: if data can be copied, it can live forever. "My solution is not based on perfect technology," Mayer-Schönberger says. "What I want is for people to think about the life span of the information that surrounds us, to reintroduce into our consciousness the importance of forgetting."
Mayer-Schönberger is not the first scholar to argue that forgetting matters. A decade ago, sociologists Donald MacKenzie and Graham Spinardi suggested that human society might "forget" how to make nuclear weapons because, even if the technical information required to build them could not be eliminated, the tacit knowledge required ("embodied in people rather than words, equations, or diagrams") could wither and die.
The case for forgetting, however, presents significant practical and moral dilemmas. To "forget" how to make nuclear weapons "might sound promising, but the point is that we could also forget how to make medicines and other technologies," Cheshire says. "There are trade-offs associated with 'forgetting,' or with not even having to remember in the first place."
Codified forgetting also carries ugly associations of book-burning and other totalitarian cultural purges. But Mayer-Schönberger's proposal addresses a very specific, very recent phenomenon: the blurring of public and private space on the Web. Data ecology would, among other things, enable people to use the Internet without leaving a long-term trail of their every click and search query - acts that feel private, unrecorded, but are anything but.
"Consumers of Internet-based services do need to press for more humane uses of data retention," says Dan Visel, a fellow at the Institute for the Future of the Book in New York. This summer, Google,
According to Danah Boyd, a fellow at the Berkman Center for Internet and Society at Harvard Law School, the solution is not to fight the ubiquity of memory but to adapt. "No amount of structural intervention is going to combat this," Boyd says. "People, particularly younger people, are going to come up with coping mechanisms. That's going to be the shift, not any intervention by a governmental or technological body."
Alessandro Acquisti, an assistant professor of information technology and public policy at the Heinz School at Carnegie Mellon University, points out that as digital memory accumulates, the problem of over-remembering might take care of itself. "The cost of downloading and storing digital information keeps decreasing; the cost of analyzing it is also decreasing through data-mining technologies, but the big question is whether these costs are decreasing at the same rate," Acquisti says. "If there's no incentive to analyzing the information, it still exists, but it's basically protected by information overload.
"And then," he says, "we go back to forgetting, the way our ancestors did."
Jessica Winter is a writer in New York