Volume 12, Number 2, 2003
Seeing is believing. Or maybe not, according to Helene Intraub, a professor of psychology who has conducted extensive experiments in perception, memory and visual illusions.
In the course of her research, Intraub has discovered that people tend to share a common error when remembering a photograph of a scene. They remember parts of the scene that were not in the photograph but were likely to have existed just beyond the boundaries of the view, a phenomenon she refers to as "boundary extension."
The National Institute of Mental Health has awarded Intraub a four-year, $739,000 grant to continue her work in understanding these constructive errors and the beneficial role they may play in everyday perception.
"When we think of memory errors, we often think of them as failures," Intraub says. Memory errors that are detrimental have been studied for many years, particularly in the context of eyewitness testimony, she says, adding there are cases in which witnesses incorrectly remember the details of a crime, not because they are lying but because of errors in memory.
"There is a lot of research on what is called the 'misinformation effect,'" she says. "People can honestly confuse newly provided information with what they originally saw."
In contrast, what is so intriguing to Intraub about boundary extension is that "although it is certainly error, it is an error that is apparently shared by everyone, and it is an error that usually provides an excellent prediction of what really did exist beyond the boundaries of the view," she says.
Boundary extension occurs under conditions in which one would expect memory to be quite accurate, she says. For example, when viewers were shown a series of close-ups and then were shown the identical pictures a short time later, they tended to deny they were the same, remembering instead that the original pictures had shown a more wide-angle view than they actually did. This error can occur as soon as one second after viewing a picture.
Artists who have participated in these experiments make the same error, Intraub says, and recent research from other laboratories has shown that both young children and elderly viewers also tend to make the error.
Strangely enough, warning people about the phenomenon does not prevent its occurrence. In fact, Intraub and her students continue to be prone to the error themselves. "Why would memory fail so consistently and in the same way for so many people? Why should such an error occur at all?" she asks.
Intraub suggests that one can begin to answer these questions by considering a classic puzzle in the field of visual perception.
"Although we all have the impression that we see everything at once, we actually don't," she says. "The human observer can only see a very small part of the world in sharp detail at any one time."
Intraub says the human field of view encompasses about 200 degrees of visual angle in width and 150 degrees in height, but out of that entire field, only about 1-2 degrees of visual angle--an area about the size of a thumbnail held at arm's length--can be seen with high acuity.
"Although the world around us is continuous, perceptual input is made up of a succession of discrete inputs," she says. "Our eyes scan a scene by making movements that rapidly shift the position of the high acuity region of vision from one location to the next. Eye position can shift as quickly as three to four times per second. Although we don't realize it, while the eyes are in motion vision is suppressed. The brain must rely on memory for what was present in the last glimpse."
Intraub says cognitive scientists have puzzled over the question of how people perceive a smooth and continuous world when visual input consists of a succession of discrete snapshots in which only the central region can be clearly seen.
To provide a coherent mental representation of our continuous surroundings, she says, the brain has adapted by ignoring the visual boundaries of a given view and extrapolating beyond its edges.
"To the mind, looking at a photograph of a scene is like looking at the world through a window," she says. "The world continues beyond the window frame."
Intraub and her students also have developed experiments in which research subjects exhibit the same type of boundary extension in viewing three-dimensional scenes using common household objects. In a variation on those experiments, blindfolded participants who feel the objects, their layout and their relation to the boundaries also remember having touched the area just beyond the borders.
As people move their eyes to shift focus, or move their hands to explore a new area, the mind gets there in advance of the senses, Intraub says.
Because of these findings, she says she believes boundary extension "is a very general aspect of perception. Whether we sample the world through eye movements or hand movements (when vision is not possible), the mind must be able to provide a coherent representation of a continuous world."
Intraub presented her work at the College of Arts and Science's "Alumni College" events during Homecoming weekend.
--Neil Thomas, AS '76