Spatial Memory - Long-term Spatial Memory

Long-term Spatial Memory

Spatial memory recall is built upon a hierarchical structure. That is to say that people remember the general layout of a particular space and then "cue target locations" located within that spatial set. This paradigm includes an ordinal scale of features that an individual must attend to in order to inform his or her cognitive map. Recollection of spatial details is a top-down procedure that requires an individual to recall the superordinate features of a cognitive map, followed by the ordinate and subordinate features. Thus, two spatial features are prominent in navigating a path: general layout and landmark orienting (Kahana et al., 2006).

People are not only capable of learning about the spatial layout of their surroundings, but they can also piece together novel routes and new spatial relations through inference. Yet, this field has traditionally been hampered by confounding variables, such as cost and the potential for previous exposure to an experimental environment. Thankfully, technological leaps have opened a new, albeit virtual, world to psychologists.

A cognitive map is "a mental model of objects’ spatial configuration that permits navigation along optimal path between arbitrary pairs of points." This mental map is built upon two fundamental bedrocks: layout, also known as route knowledge, and landmark orientation. Layout is potentially the first method of navigation that people learn to utilize; it’s workings reflect our most basic understandings of the world.

Hermer and Spelke (1994) determined that when toddlers begin to crawl, around eighteen months, they navigate by their sense of the world’s layout. Indeed, it would seem that a sojourning toddler’s world is a place of axial lines and contrasting boundaries. McNamara, Hardy and Hirtle identified region membership as a major building block of anyone’s cognitive map (1989). Specifically, region membership is defined by any kind of boundary, whether physical, perceptual or subjective (McNamara et al, 1989). Boundaries are among the most basic and endemic qualities in the world around us. These boundaries are nothing more than axial lines which are a feature that people are biased towards when relating to space; for example one axial line determinant is gravity (McNamara & Shelton, 2001; Kim & Penn, 2004). Axial lines aid everyone in apportioning our perceptions into regions. This parceled world idea is further supported items by the finding that items that get recalled together are more likely than not to also be clustered within the same region of one’s larger cognitive map. Clustering shows that people tend to chunk information together according to smaller layouts within a larger cognitive map.

Boundaries, though, are not the only determinants of layout. Clustering also demonstrates another important property of our relation to spatial conceptions. This is that spatial recall is a hierarchical process. When someone recalls an environment or navigates terrain, that person implicitly recalls the overall layout at first. Then, due to the concept’s "rich correlational structure," a series of associations become activated . Eventually the resulting cascade of activations will awaken the particular details that correspond with the region being recalled. This is how people encode many entities from varying ontological levels, such as the location of a stapler; in a desk; which is in the office .. Alas, layout has its flaws too. One can recall from only one at region at a time (a bottleneck).

A bottleneck in a person's cognitive navigational system could be disastrous, for instance if there were need for a sudden detour on a long road trip. And yet, people are still capable of getting place to place functionally. Lack of experience in a locale, or simply sheer size, can disorient one’s mental layout, especially in a large and unfamiliar place with lots of overwhelming stimuli. In these environments people are still able to orient themselves, and even find their way around using landmarks. This ability to "prioritize objects and regions in complex scenes for selection (and) recognition" was labeled by Chun and Jiang in 1998. Landmarks give people guidance by activating "learned associations between the global context and target locations." Mallot and Gillner (2000) showed that subjects learned an association between a specific landmark and the direction of a turn, thereby furthering the relationship between associations and landmarks. Shelton and McNamara (2001) succinctly summed up why landmarks, as markers, are so helpful: "location...cannot be described without making reference to the orientation of the observer."

It is fairly clear that people use both the layout of a particular space, as well as the presence of orienting landmarks in order to navigate. Yet, psychologists have yet to explain whether layout affects landmarks or if landmarks determine the boundaries of a layout. Thus, this concept suffers from a chicken and the egg paradox. In fact, McNamara has found that subjects use "clusters of landmarks as intrinsic frames of reference," which only confuses the issue further.

People perceive objects in their environment relative to other objects in that same environment. In other words, landmarks and layout are complimentary systems for spatial recall. However, it is unknown how these two systems interact when both types of information are available. Thus, we have to make certain assumptions about the interaction between these two systems. For example, cognitive maps are not "absolute" but rather, as anyone can attest, are "used to provide a default...(which) modulated according to...task demands." Psychologists also think that cognitive maps are instance based, which accounts for "discriminative matching to past experience."

These assumptions could soon be validated. Advances in virtual reality technology have pried open the door to this enigmatic field. Now experimenters find themselves creating scenarios that were impossible to imagine fifteen years ago. Virtual reality affords experimenters the luxury of extreme control over their test environment. Any variable can be manipulated, including things that would not be possible in reality.