The functional, computational and neural basis of human survey knowledge - how does map-based and navigation-based knowledge interact?
Final Report Abstract
Survey knowledge allows people to point to distant locations and to find novel shortcuts. The project examined whether survey knowledge was memorized in a single cognitive map encompassing the whole environment or rather within multiple local units, for example, streets, rooms, or corridors, which are combined only when required. Several results speak against a cognitive map: Navigators did not integrate multiple views into one reference frame during navigation, but did so later when pointing to a target. Here the number of corridors walked was important for timing which should not be crucial when relying on a cognitive map. Furthermore, examinations into neural representations suggested that navigators did not spontaneously form a constant global map spreading across multiple rooms. Last, when walking through an impossible virtual environment encompassing a wormhole, participants adjusted their memory accordingly, however, not enough to represent it in a single cognitive map. Apart from these results, speaking against a cognitive map, other results contradict multiple local units as well. When pointing to a target not only the number of segments mattered, but also their length. In case of integrating segment by segment only the number should have mattered. An impossible virtual environment with a wormhole can easily be represented by local units which do not have to fit into a globally consistent coordinate system. However, participants adjusted their memory to the wormhole nevertheless, which was not expected. As a conclusion, none of the examined concepts of a cognitive map or multiple local units could fully explain the data. Consequently, new concepts have to be developed to explain human survey knowledge. Although people do not seem to form a cognitive map when navigating through an environment, our results indicate that they nevertheless sometimes do have a map in their head. However, such a mental map is learned from external paper or electronic maps, not from navigation. If available, external maps determine the way survey knowledge is represented in memory, both within novel learned environments as well as within our city of residency. When accessing such knowledge while located within their city of residency, navigators spontaneously adjust it to their current egocentric position. This might prepare them for later actions. In summary, this second line of research suggests that people flexibly use map and navigation based knowledge in order to solve their everyday navigation problems. Several public media reported about results from the current project. This included New York Times, Le Monde, Süddeutsche Zeitung, Deutschlandfunk, SWR, Geolino, Max Planck Forschung and Apotheken Umschau.
Publications
- (2011). Grid cell remapping in humans. 2011 Annual Meeting of the Society for Neuroscience
Pape, A-A., Wolbers, T., Schulze, J., Bülthoff, H.H. & Meilinger, T.
- (2011). When do we integrate spatial information acquired by walking through environmental spaces? 33rd Annual Conference of the Cognitive Science Society
Henson, A., Mallot, H.A., Bülthoff, H.H. & Meilinger, T.
- (2012). The Map in Our Head is Oriented North. Psychological Science, 23, 120-125
Frankenstein, J., Mohler, B.J., Bülthoff, H.H. & Meilinger, T.
