How we remember objects in the real world

The Parthenon in Nashville, Tennessee, is a full-scale reconstruction of the rather more famous monument atop the Acropolis in Athens, Greece. We visited it with our daughter Nora a few years back:

As you can see, it's a dramatic building, dominating the landscape of the otherwise ordinary city park in which it sits. So, when we're confronted with such a massive landmark, do we use it to organize the surrounding area as well? Several studies have shown that we do pay attention to the surroundings of objects in order to remember their location. If we memorize the locations of a number of objects in a room, and later are asked to imagine ourselves back in a particular spot in the room, we're more accurate pointing in the direction of one of the objects if it's in front of us, rather than behind. If the room is rectangular, we're more accurate pointing to objects that are oriented in a direction parallel to the walls of the room—even if we memorized the objects when we were facing in an oblique direction.

So both the orientation of the viewer and the surrounding area can impact our memory for object locations. But what about when we're outside? Can features in the landscape similarly influence our memory? A team led by Timothy McNamara conducted a study in Centennial Park where the Nashville Parthenon is located to investigate these issues. Since the Parthenon is such a large, regular object, the team suspected it might influence how people's memory for other objects in the park. Volunteer participants in the experiment were blindfolded and driven to the park. The blindfolds were removed and the participants were led on one of two paths through the park. They were instructed to memorize the location of eight objects along the path. The key to the experiment was the orientation of the paths: one path was aligned with the shape of the Parthenon, but the second one was misaligned—rotated by 45 degrees, so participants were always walking at diagonals compared to the monument. I've created a diagram of the area using a satellite image from Terraserver (that's the Parthenon in the middle of the picture):

Notice that the items the participants memorized were located at the intersections of both of the paths. Participants were told that they were allowed to stop and turn their heads as they walked, but they should keep their bodies facing forward along the path at all times. They were led through the path twice, and by the end of the learning phase, they could recall the order in which they had seen each of the objects.

Next, they returned to the laboratory and were tested on the relative locations of the objects using a series of questions, all in the same format. For example, they might be instructed to "imagine you are standing by the bench and facing the tree. Point to the frame."

The results were as follows:

This chart compares the direction people imagined themselves heading to the errors they made pointing to the locations of other objects. Note that higher values on the chart correspond to less accurate location memory. Participants who followed the misaligned path were nearly always less accurate than those who walked the path aligned with the Parthenon—the only exceptions were facing 45 and 225 degrees, where the results were statistically indistinguishable. Why not in those cases? It may be that participants were using other landmarks to frame their memories—for example, the lake, which cuts at about a 45-degree angle on the lower right of the map.

The participants following the aligned path did the best when their imagined views aligned with both their path of travel and the walls of the Parthenon. There was not a comparable effect for those who followed the misaligned path.

McNamara et al. claim that theirs is the first experiment examining real-world outdoor frames of reference. They argue when a viewer's path is aligned with a landmark, the landmark becomes an important aid to navigation and memory. Even when recalling a view that was directly along their path of travel, the misaligned group was nearly always less accurate than the aligned group. But the wide disparity in accuracy throughout the various viewing angles suggests that many factors are involved. Other landmarks such as the lake probably also influenced the results.

McNamara, T.P., Rump, B., & Werner, S. (2003). Egocentric and geocentric frames of reference in memory of large-scale space. Psychonomic Bulletin and Review, 10(3), 589-595.