Does language determine how we think about spatial relationships?

Monday, June 24, 2013 - 9:45am

Crowdsourcing helps answer questions

Imagine that you are asked to describe the layout of a room. Do you think that the language you speak influences your perception of the room? For example, in English the same preposition is used to refer to both horizontal and vertical support of objects.


We would say, “The book is on the table” as well as “The picture hangs on the wall.” Other languages, however, make different distinctions; in German and Dutch, auf/op is used for horizontal support (as in the book is on the table) while an/aan is used for vertical connections (the picture hangs on the wall). Does this mean that speakers of German and Dutch also think about the room differently by making finer distinctions between the spatial relations between objects?


This is just one example of many that researchers collected who are interested in the relation between cognition and language, a research field known as linguistic relativity (and associated with the famous Sapir-Whorf Hypothesis). The quintessential question that drives these researchers is: Does the language we speak shape the way we think?


Researchers in the Human Factors in GIScience Lab are exploring linguistic relativity for spatial cognition, that is, how thinking about space may be influenced by the language someone speaks. Importantly, they add a geographic perspective to this fascinating question by exploring how formalisms from the area of spatial information theory can help to explain whether or not language influences humans understanding of space. They recently published a paper showing that humans understand (some) fundamental concepts of space and time in the same way regardless of their native language (comparing English, Chinese, and Korean).


“Many characteristics of humans’ spatial environments (e.g., gravity) are universal across the globe and are considered central to human cognition, for example, to understand concepts like “up” and “down”. Other concepts, however, are linguistically and/or culturally influenced like the on example above. Research on linguistic relativity tries to disentangle universal from language specific concepts,” notes Alexander Klippel, who led the research.


In addition to Klippel, the team includes Jan Oliver Wallgrün, Jinlong Yang, Jennifer Mason, and Eun-Kyeong Kim from Penn State as well as David Mark from SUNY Buffalo, an internationally recognized expert on cross-cultural studies.


In their current study on how humans understand spatial relations, participants were shown 23 unique images of two shapes (a circle and a “candy cane” shape) in different positions relative to each other, for example, not touching,  slightly overlapping, and completely surrounding (Figure 2). The shapes represented an oil spill (candy cane shape) and a protected habitat zone (circle shape) and each participant performed two tasks:

Task A: Create any number of categories and place the images into these categories

Task B: Label the categories linguistically by using both short and long descriptions.


Figure 2

Figure 2.


The analysis showed that there were no statistically significant differences among English, Chinese, and Korean language speakers in either the number of categories they created or the amount of time they took for Task A.  (The results from Task B are being analyzed for another paper.) The overall category structure participants created in all three languages was also comparable and can be grouped into three distinct spatial relations for all languages: 

1: overlap, oil spill is partially in the protected habitat zone

2: surrounded: oil spill is completely inside protected habitat zone

3: separate: oil spill is completely outside protected habitat zone


“These results show a strong agreement across all three languages on what can be considered the most fundamental distinctions with respect to overlapping spatial regions,” Klippel explains.


In addition to addressing a prominent research question in the cognitive sciences Klippel's team also used a novel research method referred to as crowdsourcing. Usually laboratory research relies on fairly homogenous pool of participants: local college students. Now employing crowdsourcing through online tools can result in a more diverse pool of participants and highly efficient data collection. This experiment used Amazon Mechanical Turk (AMT) to collect data. “There are some challenges to using crowdsourcing, but AMT allowed us to set high standards for participation, pre-test, and adapt the instrument and reward structure for the most reliable results.”

"We are very excited about this study; beyond the results on how people fundamentally understand spatial relations, the crowdsourcing method we used opens new avenues for behavioral research in the area of spatial cognition,” Klippel notes, adding, “formal approaches developed in the spatial sciences allow us to precisely characterize various aspects of our spatial environments. As such, they are ideal for formulating scientific hypotheses which spatial aspects are important for humans and whether language influences the aspects considered most salient.”

This research is part of an NSF funded project (#0924534) that addresses the way people categorize spatial and temporal information and how human cognition can be formalized using approaches from the area of spatial information theory.


Further resources:


Klippel, A., Wallgrün, J. O., Yang, J., Mason, J. S., Kim, E.-K., & Mark, D. M. (2013). Fundamental cognitive concepts of space (and time): Using cross-linguistic, crowdsourced data to cognitively calibrate modes of overlap. In Conference on Spatial Information Theory, COSIT 2013.


Recently published overview articles on related research:


Klippel, A., Li, R., Yang, J., Hardisty, F., & Xu, S. (2013). The Egenhofer-Cohn hypothesis: Or, topological relativity? In M. Raubal, A. U. Frank, & D. M. Mark (Eds.), Cognitive and Linguistic Aspects of Geographic Space - New Perspectives on Geographic Information Research (pp. 195–215). Berlin: Springer.


Klippel, A. (2012). Spatial information theory meets spatial thinking - Is topology the Rosetta Stone of spatio-temporal cognition? Annals of the Association of American Geographers, 102(6), 1310–1328. Retrieved from