Investigating sense of place as a cultural ecosystem service in different landscapes through the lens of language

The concept of ecosystem services is increasingly important for measuring both tangible and intangible benefits that humans obtain from ecosystems. Much research on ecosystem services focused on more tangible services. Intangible cultural ecosystem services, such as sense of place, are often neglected, but in the context of highly populated, increasingly urbanized countries, maintenance of cultural ecosystem services is an important policy objective. One of the challenges of integrating sense of place into the framework of ecosystem services is that it is not linked to abstract notions of ecosystems, but tied to perceived landscape features such as mountains, or rivers. In this study, we used free listings and interviews with visitors to investigate perceived landscape features and sense of place through the lens of language in five different landscape types in Switzerland. Within each landscape type, we selected two study sites to quantitatively and qualitatively compare landscape descriptions. Using text processing, we show that terms for landscape features were more similar within the same landscape type, suggesting that people perceive differences between these landscapes that they express in language. However, in general, elicited concepts related to sense of place were similar across landscape types. Thus, our results show that we can use free-listing descriptions to distinguish landscape types, but we found limited differences in the language used to describe sense of place. Our findings offer insights into exploring sense of place in different landscapes and contribute to ongoing efforts for refining the definitions and standardizing assessments of cultural ecosystem services.


Introduction
Ecosystems are essential for human well-being through the benefits that people derive, directly or indirectly, from ecosystem functions (Costanza et al., 1997). One, increasingly popular, way of quantifying these benefits, is through ecosystem services, defined as benefits provided by ecosystems that contribute to making human life both possible and worth living (MA., 2005). Research on ecosystem services has steadily increased (Costanza et al., 2014;Fisher, Turner, & Morling, 2009), partly driven by the recognition that such quantification may provide a useful tool in both monitoring and scenario development, and as an input to policy decisions, for example, in the European Union 2020 Biodiversity Strategy (EU Commission., 2014). Ecosystem services are commonly classified into four groups: provisioning services such as drinking water and food; regulating services such as erosion and flood control; supporting services such as nutrient cycling, and cultural services, defined as nonmaterial benefits obtained from ecosystems, including aesthetic beauty, recreation, cultural heritage value and, the focus of this paper, sense of place (Chan, Satterfield, & Goldstein, 2012;MA., 2005). Although the importance of cultural ecosystem services (CES) is generally recognized, their assessment lags behind that of more tangible services (Benayas, Newton, Diaz, & Bullock, 2009;de Groot, Wilson, & Boumans, 2002;Feld et al., 2009;MA., 2005). In the Millennium Ecosystem Assessment, for instance, ten cultural services are defined, of which only three were assessed (spiritual values, aesthetic values, and recreation), while the other seven (including sense of place) were not (MA, 2005). Literature reviews (Hernández-Morcillo, Plieninger, & Bieling, 2013;Milcu, Hanspach, Abson, & Fischer, 2013) showed that where cultural services were assessed, studies often focused on cultural services that could be investigated through relatively easily quantifiable proxies, such as tourist visits for assessing recreation, or the number of images on social media as a measure of aesthetic beauty (Feld et al., 2009;Figueroa-Alfaro & Tang, 2017;Fuller, Irvine, Devine-Wright, Warren, & Gaston, 2007;Tenerelli, Demšar, & Luque, 2016;Yoshimura & Hiura, 2017;Zoderer, Tasser, Erb, Lupo Stanghellini, & Tappeiner, 2016). One of the key challenges involved in CES research is linking CES to space. This has been attempted, for example, by counting georeferenced user-generated images on photo sharing sites as proxies for scenicness (Gliozzo, Pettorelli, & Haklay, 2016;Yoshimura & Hiura, 2017), by participants entering points or polygons associated with certain CES on maps and in Geographic Information Systems (GIS) (Brown & Weber, 2012;Fagerholm, Käyhkö, Ndumbaro, & Khamis, 2012) or by assessing landscape types and associated values through visitor surveys that are mapped in a GIS (van Berkel & Verburg, 2014;Zoderer et al., 2016). However, moving beyond relatively simple proxies and developing a better understanding of CES such as sense of place requires complementary competencies and methods drawn from the social sciences and humanities (Chan, Guerry, et al., 2012;Daniel et al., 2012;Schaich, Bieling, & Plieninger, 2010).
One of the reasons why sense of place has remained an elusive concept that is often omitted from CES assessments may be that multiple theories, concepts and empirical approaches exist in different disciplines in the social sciences and humanities, even within a single field of research. However, this matter is only seldom touched upon when introducing sense of place (Hausmann, Slotow, Burns, & Di Minin, 2016;Williams, 2014).
Furthermore, studies investigating the links between landscape features and CES often assume the landscape features to which these services are attached to be culturally universal (Oteros-Rozas, Martín-López, Fagerholm, Bieling, & Plieninger, 2017), and less attention has been paid to describing and comparing terms for culturally recognized landscape features of different groups of people in different landscapes (Lofland, 1975;Stedman, 2003). Focusing on culturally recognized landscape features is important, as one of the major challenges in integrating cultural aspects into the framework of ecosystem services is that such values and meanings are not linked to abstract notions of ecosystems, but tied to perceived and culturally recognized landscape features, such as lakes, mountains, or rivers (Kirchhoff, 2012). Recent work in linguistics and landscape ethnoecology highlights however, how the terms people use to refer to features of the landscape are non-universal, often culture-specific, and difficult to translate, with implications for methods attempting to link CES to landscapes (Burenhult, 2008;Johnson & Hunn, 2010;Mark, Turk, Burenhult, & Stea, 2011). The focus of many linguistic and ethnoecological studies has mostly been on indigenous cultures and endangered languages, and relatively few studies have addressed variation of landscape terminology in European languages (e.g. Mark, Smith, & Tversky, 1999;Wartmann, Egorova, Derungs, Mark, & Purves, 2015;Williams, Kuhn, & Painho, 2012). One recent example, which has also resonated with the public, is work by Macfarlane (2015), who collected over a thousand 'terms for the land' used in different British dialects and languages. The terms range from those describing small-scale landscape features such as smeuse ('the gap in the base of a hedge made by the regular passage of a small animal'), to entire landscapes, for example, the expression gruffy ground, referring to the surface landscape left behind by lead-mining in Somerset (Macfarlane, 2015). Such landscape terms thus carve out identifiable units from a landscape infused with cultural meaning and significance. Investigating landscape terms thus fulfills an important requirement of cultural ecosystem service assessments, namely, the identification of units based on local understandings of groups of people to which cultural, as well as other values can be assigned (Kirchhoff, 2012).
So far, little research has focused on identifying landscape terms and investigating differences in sense of place as a cultural ecosystem service between different landscapes. In this paper, we address this gap by investigating the following research questions: • How are different landscapes described by visitors in situ and to what extent is sense of place expressed in these descriptions?
• What methods are suitable for eliciting and analyzing such information?
• To what extent do these descriptions capture formal landscape classifications?
In addressing these questions, we aim to develop and empirically test a set of methods for eliciting descriptions of sense of place and comparing those between different landscapes. The originality of our approach lies in the combination of qualitative and quantitative approaches revolving around language to compare different landscapes based on the identified landscape terms, and in building a semantically rich understanding of sense of place based on language. We explore these questions through empirical work studying visitors' descriptions of five contrasting landscape types in Switzerland. Given the multitude of theoretical conceptualizations and empirical approaches for studying people-place relations, in the following, we provide an overview of related literature and outline the conceptual framework we adopted in this paper.

Background
The relation between people and places has received considerable attention in different research areas. A myriad of different concepts, definitions and approaches exist between and even within research fields. To situate our own research within the broader theoretical debate, we introduce concepts and approaches that have been applied for studying people-place relations. To aid readability, we broadly divide our literature review into three main research areas: environmental psychology, human geography and social anthropology.
Place identity was defined as that part of our identity that relates to place (Proshansky, 1978). Other conceptualizations of place identity include the notion of 'place-referent continuity', the process by which people maintain continuity of their identity via specific places that are of emotional significance to them, and 'place-congruent continuity', as the generic and transferable characteristics of places that help maintain continuity (Twigger-Ross & Uzzell, 1996). The concept of place dependence captures how a place can be important because of its functional value (Stokols & Shumaker, 1981). For example, people may depend on a place because it allows them to realize particular recreational activities (Jorgensen & Stedman, 2001;Moore & Graefe, 1994). Altman and Low (1992) conceptualized place attachment as containing emotional components (affect, feeling, emotions), cognitive elements (thoughts, knowledges, beliefs) and practices (actions and behavior). The forth concept, sense of place, was considered by Jorgensen and Stedman (2001) as an umbrella concept subsuming place identity, dependence and attachment.
Despite considerable theoretical and empirical work, the hierarchical organization of these concepts and their linkage remains controversial (Giuliani & Feldman, 1993;Hernández, Carmen Hidalgo, Salazar-Laplace, & Hess, 2007;Hidalgo & Hernández, 2001). One critique of this body of literature is that perhaps less attention has been paid to the concept of place with respect to the specific objects or features in a place or landscape to which people are attached (Hidalgo & Hernández, 2001;Jorgensen & Stedman, 2011;Lewicka, 2011).