Prosthetic Memory: A Camera That Gives Back Lost Moments

Anyone who has spent time playing the photographer at a family event can affirm how a camera carves up scenes, altering the memories we hold of them from then on. A human photographer, of course, is not a passive observer: Before we click, we calculate, choosing important moments and composing them just so. What would happen if we relinquished those calculations to someone—or something—else?

Someone to Watch Over Me

In the late 1990s, the Microsoft Research Group in Cambridge, U.K., began developing a simple digital camera designed to take over just that job. The camera is diminutive enough to nestle comfortably in the palm of the hand and light enough to be worn on a thin cord around the neck. It can be programmed to take a picture at regular intervals, such as once every 30 seconds—and it can do something a little more subtle. Built-in sensors keep constant tabs on surrounding levels of light intensity and color, the wearer’s body temperature, and his or her movements and acceleration; small changes in any of these conditions can be set to trigger an automatic shutter release. The camera is alert for those moments that are likely to be significant, making it ready to capture the joyful approach of someone whose face makes the heart race, the blood boil—ready for the first glimpse of what lies beyond a just-opened door.

Since the camera has no viewfinder or display, its wearer cannot choose which scenes it records or dictate their composition. It is not a tool; it is more akin to a witness. By the end of the day, hundreds of low-resolution images have been taken, collecting the arc of the wearer’s experiences in a series of odd, unpredictable splinters that can be viewed on a computer as still images or patched together into a twitchy little film. Microsoft calls their project the SenseCam; I’d say it’s the ultimate antidote for an unobserved life.

The creators of the SenseCam meant to devise a sort of memory prosthesis, an external aid that does its best to mimic a person’s own ability to recall experiences. In order to match the viewpoint in the photos as closely as possible to a first-person perspective, Microsoft fitted the camera with a wide-angle lens and recommends that wearers position it at chest-height. You might think of it as a perfected version of human memory. The camera’s testimony, unlike that of a human mind, is impartial, infallible, and everlasting. Yet the images the camera forms are also random and incomplete. And although they seek to mimic a first-person perspective, they do not duplicate it.

Over the past ten years, the SenseCam has played a leading role in several studies conducted by clinicians, psychiatrists, and sociologists. Reading these reports, an odd truth emerges: The camera is at its most fascinating when its record diverges, either in tone or substance, from that of its user.

La Vie Quotidienne

In 2007, British researchers trained six citizens from South Wales in the use of the SenseCam and its software, and sent them off for one week to see what they could see. The researchers’ findings, published in the proceedings of the 2007 International Conference on Human Computer Interaction, are among the most charming and philosophical scientific results you may ever read.

The paper discusses a number of common threads in the study-subjects’ experiences as they viewed images captured over the course of the week. Not surprisingly, many reported that they were taken off guard by the sight of features in their lives that had seemed unimportant at the time: the way the road rolled steadily on as their bicycle wheels spun below during a commute, the specific gestures a lover used as he talked about his day. The participants found these images moving, perhaps because of the sense that if the SenseCam had not been there, they would have simply drifted softly into silence, like the thrum of a bell.

In the same way, subjects were forced to notice not just the major events of their lives, but those that—by sheer force of repetition—comprised the greatest number of photographs. One participant noted “how much I was in the car… how much you go shopping… how much of your day is taken up by washing up.” The reliable firing of the camera’s shutter, in other words, divided the day into units that could be counted. Its images presented people with an opportunity to measure their lives: this much time spent alone, this much with family, this much working, this much walking, this much sitting, staring, thinking. This is a very different camera from the ones that accompany people to dinners with friends and fly happily through the magnificent streets of foreign cities, pointed at the best moments of our lives.

The scientists noted two other SenseCam-related experiences that seem closely connected: First, subjects frequently described how the images rendered the details of their own lives “strange” to them, as if they had each been the lonely and fascinating protagonist of a silent film. Second, they discovered a heightened awareness of the lives of others. Having looked at their experiences from a disembodied perspective, people began longing to enter the experiences of other bodies. They became inclined to fasten their SenseCams to kites, so that they could enter the world of flight; dogs, so that they could discover the world as other creatures lived it; they offered their cameras to friends and neighbors, wanting to become—if only temporarily—someone other than themselves.

Memory-Maker

In 2004, a 63-year old woman named by her doctors “Mrs. B.” became the first person to use the SenseCam as a memory rehabilitation aid. Two years previously, Mrs. B. had suffered an acute episode of limbic encephalitis: an inflammation of the brain caused by the onset of a viral infection. Although she made a full physical recovery from the infection, the lesions Mrs. B. sustained in her hippocampal lobes caused her to experience marked anteretrograde memory impairment: an inability to construct lasting memories of events that happened after the illness. The neuropsychologists who were treating Mrs. B. hoped that, with the help of her husband, she would be able to use the SenseCam to form and retain memories of significant events in her life.

They were right.

Without the SenseCam, Mrs. B. could only remember events—no matter how interesting or unusual—for a few days at a time before their outlines would become diffuse. Writing down what had happened, and then rereading her own descriptions, enabled Mrs. B to recall some portion of the events for up to two weeks—but then they, too, would slip from her mind like sand. The SenseCam, on the other hand, dramatically intensified both the clarity and persistence of Mrs. B.’s memories.

Here’s what Mrs. B.’s physicians had her do. She would wear the camera around her neck on days when she and her husband agreed that an event was about to take place that was worth remembering—a trip to a new town or a visit with friends. She would then review the images it took on that day, discuss them with her husband, and return to them several times over the course of a two week period. This process resulted in an astonishing improvement in her long-term recall of autobiographical events.

By astonishing, I mean this: three months after her last viewing of any given set of SenseCam images, Mrs. B. could, on average, recall more than 75 percent of the items on a predetermined list of details about the events they depicted. She had gone from being a functional amnesiac to being someone whose memory, though faulty, was comparatively rich and reliable. And she had done it by combining her own limited recall with that of a tiny automatic camera.

The authors of the paper in which Mrs. B.’s case study is described refer to the Sensecam as “aiding,” or “improving,” her autobiographical memory, arguing that it “cues and consolidates” her own impressions. They consider it a helpful little sidekick for her brain, compensating for what she lacks.

Perhaps. But consider this final study. College-aged subjects were asked to work at remembering three childhood events: two that their parents reported had actually happened to them at school, and one that had not. Subjects who were shown a class photo (also helpfully provided by their parents) from the year in which the invented event supposedly took place were dramatically more likely to report that they remembered it happening quite vividly.

Merely showing people an old photograph, in other words—even though the image itself depicted something entirely different—made it far easier for researchers to implant false memories. If Mrs. B. pores over her SenseCam photos with her husband over and over, where, truly, are the memories she now makes coming from? It may not matter much, in practical terms; Mr. B. reports that his wife’s quality of life has improved significantly since she started wearing half her memory around her neck, and that they are able to share experiences in a way that they could not before. That is probably the most important thing.

Still, when you think about the power of the observer to change the observed, it’s hard not to wonder about the strange and unexpected ways in which this tiny little camera might be changing the lives of those who carry it.