Navigation auf


UFSP Sprache und Raum

Landscape and language – Exploring approaches to extracting landscape perception

How individuals and groups of individuals perceive and interact with landscapes has enjoyed a long history of research in cognitive, geographic and linguistic sciences. Understanding how persons experience their environment has been an especially prominent field of research due to the importance of the topic for policy and decision-making purposes. The literature has long identified the landscape and environment a person traverses through on a daily basis to have a direct impact on overall wellbeing. A rather large portion of research has been conducted on specific aspects of mostly urban landscapes (e.g. how do persons perceive individual elements of a city?) or specific natural features (e.g. what effect do waterbodies have on a person’s mood?), but there is a lack of detailed and diverse data for landscape perception analyses not linked to specific locations. Generally, research has revolved around collecting data through targeted surveys and/or interviews. Internet technologies, especially Web 2.0, have opened the door to new forms of data collection through user generated content (UGC) - more specifically volunteered geographic information (VGI) - such as using georeferenced short messages or image collections (e.g. Twitter, Instagram, Flickr ...). However, such passively sourced data is produced for multiple reasons, and content directly relevant to landscape perception is relatively rare. By actively encouraging participation with a specific aim, we will collect more diverse and semantically rich texts describing landscapes which can in turn be used in both research and policy.

The main aim of this dissertation is to develop applications to actively collect and analyse landscape perception data using a traditional citizen science approach on the one hand and a gamified approach (e.g. using entertaining elements as a primary form of motivation) on the other. We will collect in-situ natural language descriptions along with photographs as the literature has identified the use of texts as viable underlying datasets to gage individual perceptions and sentiments. People use written text as a symbolic representation of meaning to communicate how they perceive and interact with elements of the environment. We thus set out to collect and interpret such data and are especially interested in the geographical and linguistic characteristics and variances of landscape descriptions over space (e.g. How do sentiments towards, or perceived usability of similar land covers or land uses depend on location?).

PhD candidate

Manuel Bär


Ross Purves


URPP Language and Space