Our commentary focuses on navigating multilayer environments and extends Jeffery et al.'s view to an architectural and environmental psychology perspective.

The functions of buildings are generally organized horizontally, probably reflecting the constraints that humans encounter: A horizontal plane is neutral to the axis of gravity and allows for stable walking, sitting, and storing of objects. Humans and buildings “inhabit” the same “two-dimensional” ecological niche, and buildings stack floors on top of one another. As a consequence, the structure of typical buildings is highly compatible with the “bicoded” representation: Whereas the horizontal plane is continuous (albeit subdivided by corridors and partitions) and in line with the floors, the vertical axis is discontinuous and discretized; that is, floors are on top of one another, with only local connections via stairs or elevators. Unless one has a view along a multi-storey atrium, the vertical dimension is visually limited to the current or directly adjacent floor. Verticality is presented as “contextual,” at ordinal rather than metric scale, and perceived indirectly or derived by inference processes.

Tlauka et al. (Reference Tlauka, Wilson, Adams, Souter and Young2007) describe a systematic bias in vertical pointing between floors. Across several studies in our group we have observed a pattern in the process of pointing that links to the bicoded representation: Pointing appears to be based on categorical and discretized rather than continuous information, visible in smooth horizontal pointing and stepwise vertical pointing. Participants report counting floors, rather than making spontaneous judgments (Hölscher et al. Reference Hölscher, Büchner, Meilinger and Strube2009). This is further illustrated by longer response times for the vertical component of the pointing gesture.

Environmental researchers such as Weisman (Reference Weisman1981) and Carlson et al. (Reference Carlson, Hölscher, Shipley and Conroy Dalton2010) differentiate between four types of environmental characteristics that impact navigation in buildings: (1) visual access to decision points, (2) architectural differentiation (the likelihood of confusing visual scenes), (3) layout geometry, and (4) signage. These features can be seen in the light of the bicoded representation, with vertical integration as the central challenge. Several studies (e.g., Soeda et al. Reference Soeda, Kushiyama, Ohno, Takahashi and Nagasawa1997) illustrate that multilevel public buildings are inherently difficult to navigate, and wayfinding complexity can be traced to mentally linking individual floors of the building. Participants expect a consistent corridor layout across floors and are disoriented by layout differences between floors. Buildings in which the floors are partially horizontally offset from one another, such as the Seattle Public Library (by architect Rem Koolhaas; Carlson et al. Reference Carlson, Hölscher, Shipley and Conroy Dalton2010), provide an additional challenge for the mental alignment of corresponding locations across floors.

Vertical connections form salient elements in the mental representation of the building and are usually limited in number. Level changes are a key source of disorientation about one's heading and position in a building (Hölscher et al. Reference Hölscher, Meilinger, Vrachliotis, Brösamle and Knauff2006), especially when a staircase is offset from the main corridor axis, enclosed by walls, or requires many turns. An atrium in the center of a building can provide visual access between floors that helps people correctly align floors in their memory representations (Peponis et al. Reference Peponis, Zimring and Choi1990), especially when the staircase/escalator/elevator is visually linked to this vertical core. Views to the outside with a distinctive global landmark can further support integration of cognitive maps between floors.

The architectural community has developed spatial analysis methods to predict human movement patterns and cognitive representations, namely “space syntax” (Conroy Dalton Reference Conroy Dalton2005; Hillier & Hanson Reference Hillier and Hanson1984) and “isovist” approaches (Benedikt Reference Benedikt1979; Wiener et al. Reference Wiener, Hölscher, Büchner and Konieczny2012), capturing visual access and layout complexity alike. It is noteworthy that these approaches concentrate on single-floor environments and that current space syntax analysis treats multilevel connections only in a simplified fashion (Hölscher et al. Reference Hölscher, Brösamle and Vrachliotis2012). While this reflects a correspondence with the bicoded view, more fine-grained analysis of vertical connectors in building analysis is called for.

Human indoor navigation is strongly driven by the semantics of floor layout and expectations about the positioning of meaningful/typical destinations and landmark objects (Frankenstein et al. Reference Frankenstein, Büchner, Tenbrink, Hölscher, Hölscher, Shipley, Belardinelli, Bateman and Newcombe2010), as well as by expectations of regularity, good form, and closure (Montello Reference Montello, Shah and Miyake2005). Human navigation is further guided by symbolic input like signage and maps. Hölscher et al. (Reference Hölscher, Büchner, Brösamle, Meilinger, Strube, McNamara and Trafton2007) have shown that signage and cross-sectional maps explicitly designed to highlight multilevel relations and connections between floors help to overcome vertical disorientation, making first-time visitors almost as efficient in wayfinding as a benchmark group of frequent visitors.

Based on their experience with typical buildings, humans develop specific strategies for handling vertical layouts (e.g., central-point or floor strategies; Hölscher et al. Reference Hölscher, Meilinger, Vrachliotis, Brösamle and Knauff2006). Jeffery et al. refer to Büchner et al.'s (Reference Büchner, Hölscher, Strube, Vosniadou, Kayser and Protopapas2007) finding that more people preferred a horizontal-first route. This could be partly due to the fact that the horizontal distances were longer than the vertical here (as in many buildings) and that people tend to follow the longest line of sight (Conroy Dalton Reference Conroy Dalton2003; Wiener et al. Reference Wiener, Hölscher, Büchner and Konieczny2012). Furthermore, people adapt their strategies to properties of the building; the aforementioned floor strategy can then be replaced by a “line-of-sight” strategy, depending on the connections in a complex building (Hölscher et al. Reference Hölscher, Büchner, Meilinger and Strube2009).

Finally, we wish to point to the need for understanding the role of inter-individual differences in handling verticality, along the distinction of route versus survey knowledge, and the degree to which humans are able to coherently integrate cognitive maps of stacked floors. Studies cited by Jeffery et al., as well as our own, are compatible with separate, planar representations of spatial information. However, a subset of participants in our experiments consistently reports imagining the environment from an external perspective, with walls and floors like glass (“a glass doll house”). This appears to require a consistent representation of the vertical dimension, and these people tend to report allocentric rather than egocentric orientation strategies, as well as high levels of spatial abilities. Jeffery et al. state that investigating multi-floor environments alone is insufficient to identify to what degree surface-dwelling animals can build true volumetric mental maps. Virtual-reality simulations of such environments – for example, tilting a building layout by 90 degrees or comparing multilevel navigation to “vertical floating” along tunnels – provide useful extensions in this direction. In studies currently under way, we focus on the role of individual differences in how the vertical component is integrated.