Loving the mess: navigating diversity and conflict in social values for sustainability

This paper concludes a special feature of Sustainability Science that explores a broad range of social value theoretical traditions, such as religious studies, social psychology, indigenous knowledge, economics, sociology, and philosophy. We introduce a novel transdisciplinary conceptual framework that revolves around concepts of ‘lenses’ and ‘tensions’ to help navigate value diversity. First, we consider the notion of lenses: perspectives on value and valuation along diverse dimensions that describe what values focus on, how their sociality is envisioned, and what epistemic and procedural assumptions are made. We characterise fourteen of such dimensions. This provides a foundation for exploration of seven areas of tension, between: (1) the values of individuals vs collectives; (2) values as discrete and held vs embedded and constructed; (3) value as static or changeable; (4) valuation as descriptive vs normative and transformative; (5) social vs relational values; (6) different rationalities and their relation to value integration; (7) degrees of acknowledgment of the role of power in navigating value conflicts. In doing so, we embrace the ‘mess’ of diversity, yet also provide a framework to organise this mess and support and encourage active transdisciplinary collaboration. We identify key research areas where such collaborations can be harnessed for sustainability transformation. Here it is crucial to understand how certain social value lenses are privileged over others and build capacity in decision-making for understanding and drawing on multiple value, epistemic and procedural lenses.


Introduction
Social values inquiry draws upon a rich range and depth of theoretical traditions, each with its own assumptions related to how values are conceptualised, elicited and related to other constructs. This paper concludes a Special Feature of Sustainability Science that has brought together a broad range of these traditions. We seek to build bridges across these traditions, considering their diverse social value lenses and areas of tension between them. We embrace the 'mess' of diversity, yet also bring an innovative framing to this mess to support and encourage active transdisciplinary collaboration for social values as a key concern of the environmental social sciences (Chan et al. 2018, Ives andKendal 2014;Kenter et al. 2014aKenter et al. , 2015Kenter et al. , 2016aKronenberg 2014;Pascual et al. 2017;Parks and Gowdy, 2013;Rawluk et al. 2017;Raymond et al. 2014Scholte et al. 2015;van Riper et al. 2017).
Researchers and practitioners conceptualise social values in ways that connect to particular understandings of the world based on history, culture, geography, experience, and embodiment (Williams 2014). This means that no single disciplinary framework can fully integrate the many understandings of social values. We adopt a post-normal view grounded in epistemic pluralism that suggests there is no 'one correct way' of conceptualising social values; each provides a limited perspective to be scrutinised in democratic debate and decision-making (Funtowicz and Ravetz 1993;Ainscough et al. 2018). Post-normal science addresses complex, wicked problems, where facts are uncertain, stakes are high, and decisions are urgent. Today's deeply challenging environmental sustainability issues provide a prime example. Here, the choices about what and how we research are inherently normative, because all problem descriptions partially result from the value lenses through which issues are viewed. Diferent lenses give rise to competing knowledge claims, which can be addressed through deliberative processes of knowledge co-production that extend peer review from expert-only to a transdisciplinary community also involving practitioners, policy makers and citizens (Funtowicz and Ravetz 1993;Strand 2017).
In this paper, we develop a novel transdisciplinary conceptual framework to help navigate the messy reality of social values research and practice. Central to this framework is a consideration of social values as lenses of worthiness: lenses of what is considered to matter. Underpinning these lenses sit diverse meta-lenses, which explain how values are conceived and assessed. We highlight two key types of meta-lens; the epistemic lens and the procedural lens ( Fig. 1). We identify fourteen dimensions along which the diferent social value lenses and meta-lenses of diverse theoretical traditions can be discriminated (Table 1). Armed with this framework, we then investigate central areas of tension between diferent social value theoretical traditions, identifying key avenues for future research. These tensions emerged from a deliberative global expert workshop in York, UK, in June 2018 (Box 1; also see Eriksson et al. 2019), to which authors representing each of the papers within this Feature were invited. The papers were submitted in response to an open call for contributions (Raymond et al. 2018). They draw on a wide variety of theoretical bases, highlighting the importance of social values as a boundary concept (Kenter 2016a;Steger et al. 2018). This is the irst paper to present a comprehensive framework for drawing together knowledge across the wide  (Table 1) of value associated with the three types of lenses. The value lens is depicted in two parts, with broad, transcendental val-ues guiding speciic, contextual values and their indicators. While value lenses and objects of worthiness are depicted as separate entities, whether they are assumed separable will difer per epistemic lens (hence porous boundary of value lens) range of traditions that consider social values of sustainability, from more realist to more relativist, from positivist to diverse interpretivist and critical approaches. However, rather than rehash well-known spectra of ontology and epistemology, we identify speciic tensions to more precisely investigate key pinch points in relation to social values and valuation. While we focus on environmental sustainability, our contribution may also beneit other ields where social values are a key concern, such as energy and health. The exploration of tensions between theoretical traditions is an opportunity for personal and collective growth and a means for advancing scholarship, not least because it highlights diferent understandings of and approaches to social values that may not be self-evident when those from diferent backgrounds collaborate. Tensions and lenses need to be explicitly and rigorously considered if the goal is to incorporate a diversity of worldviews into environmental decision-making, as proposed by, for example, the Intergovernmental Science-Policy Platform on Biodiversity and Ecosystem Services (IPBES) (Pascual et al. 2017;Díaz et al. 2018;Christie et al. 2019b). The Intergovernmental Panel on Climate Change has long been heavily dominated by natural science worldviews, but also here there are demands for a more diverse knowledge base and challenging implicit social value lenses (Hulme 2011;Ford et al. 2016). Through a mutual recognition of diferences, viewing sustainability issues through diferent lenses of social values provides a richer and more comprehensive picture and can ofer a more inclusive and more relevant value-evidence basis for sustainability transformation. Thus, we clarify issues at the frontier of social values for sustainability in the light of these tensions, providing a forward-looking and constructive agenda for transdisciplinary engagement with sustainability science.

Box 1: Deliberation process within the author team
An open call for Special Feature abstracts was publicised in February 2018 in Sustainability Science (Raymond et al. 2018). Forty-seven submissions were received, of which 18 were selected by the co-editors (CR, AR, CvR, DK, JK) based on criteria including academic quality from peer review of abstracts, disciplinary and geographic diversity, and gender balance. An author from each paper was invited to attend a workshop at the University of York, UK, funded by the Valuing Nature Programme. The goals of the workshop were to identify linkages across papers, facilitate deliberation on broader social values knowledge across diverse disciplines, and synthesise this new knowledge in a collective article. A diversity of perspectives were represented, including environmental science and ecology, human geography, sociology, psychology, ecological and mainstream economics, anthropology, philosophy, and business and religious studies.
The irst day of the workshop involved ive sessions, each containing three paper presentations and 45 min discussion. At the conclusion of each session, workshop participants added their individual relections to an online interface. The co-editors distilled these insights into ive sets of topics: (1) value scales and hierarchies; (2) integration, plurality and conlict; (3) values and value-activation; (4) power and value change; and (5) overall conceptualisation of social values. The following day, participants deliberated on these topic sets in small subgroups over the course of two discussion sessions facilitated by the co-editors, allowing each participant to engage with two sets in detail and learn from the group's insights. Each subgroup focused on identifying areas of tension between diferent knowledge traditions and their lenses, and future directions for research. A inal plenary session further reined tensions, and identiied key conclusions and cross-linkages. Following the workshop, reports from each subgroup were prepared, which were further elaborated into mini-articles by subgroups of the authors through collaboration and discussion via email and videoconference. The lead author (JK) and co-editors then iteratively reorganised the partially overlapping content of these outputs into the structure of lenses, dimensions and tensions as presented here.

Social value lenses and dimensions of social values
There are many understandings of social values. Central understandings include values as overarching principles, values pertaining to a common good or society as a whole, and values that become shared through processes of socialisation, including deliberation and internalisation (Kenter et al. 2015;van Riper et al. 2018;Ishihara 2018). Diverse knowledge and appraisal traditions each harbour one or more social value lenses. These lenses articulate both what values are focused upon and how their sociality is envisioned. The lenses of diverse traditions can be characterised and diferentiated along multiple dimensions of social values, such as the scale of values or the process by which they are elicited (Table 1). For example, a research tradition may focus on values at the societal scale, expressed by a social unit larger than an individual (e.g. a local community) and/or through a social process (e.g. a group workshop). In addition, different traditions harbour meta-lenses ( Fig. 1), comprising speciic theories and bodies of scientiic or local and indigenous knowledge that articulate diferent perspectives on social valuation, with their own epistemologies and explicit or implicit meta-values: values about values, for example, about how values should be aggregated ). Meta-lenses thus frame the social perspective and position of the viewer with respect to how social values and their dynamics are perceived and expressed. We consider social value lenses and meta-lenses to be a dynamic medium of perception, articulation and understanding through which the world is interpreted and evaluated: they are therefore open, relexive and responsive, and not ixed, unidirectional or unchanging.
Meta-lenses also help us understand how social value lenses are associated with and applied to diferent purposes, exempliied by the diverse papers in this Special Feature. While a complete discussion of the knowledge paradigms embedded within difering meta-lenses of diverse traditions is beyond the scope of this paper, the teleological or purposeful nature of social value lenses can be explained by two key types of meta-lens: the epistemic and procedural meta-lens ( Fig. 1), or, for brevity, simply epistemic lens and procedural lens. The epistemic lens considers how we harbour, create and know 'value', as well as the philosophical orientation of knowledge traditions that guides researchers' social value lenses. The procedural lens describes the types of processes used to attain and explain social values.
To help understand similarities and tensions between diferent social value traditions, we consider a range of dimensions of their value, epistemic and procedural lenses (Table 1). Here we build on foundational work by Kenter et al. (2014aKenter et al. ( , 2015, who developed a framework for diferentiating between types of social values according to how they have been conceived in diferent traditions. At the basis of this lies diferentiation between broad, transcendental and speciic, contextual values and their indicators. Transcendental values are broad notions of what is important in life; they are not attached to a context-speciic object of value and serve to guide speciic, contextual values. For example, a broad desire for harmony with nature might lead us to express context-speciic opinions of value around conserving a particular habitat, which may be assessed in terms of quantitative or qualitative indicators through some form of valuation. It is important to note that transcendental values, while they transcend the speciic contexts where we ascribe importance to particular things, are by deinition generalisable or divorced from cultural contexts more broadly. For example, what constitutes harmony, nature, and harmony with nature might be conceived of quite diferently between diferent cultures. In practice, transcendental and contextual values can be closely entwined, particularly where transcendental values are strongly embedded in particular relationships (including with non-humans) and not meaningfully separable from the importance ascribed to those relationships (Gould et al. 2019).
This nomenclature extended research by Rokeach (1973) that diferentiated 'held' values (i.e. guiding principles and life goals) and 'assigned' values (i.e. opinions on the values of particulars), where the irst were thought to predict the latter, both through the process of deduction and a relational realm of felt experiences (Schroeder 2013). However, Kenter et al. (2015) noted that opinions on the values of particulars could semantically be both held and assigned, and that values might thirdly refer to measures and other indicators. Further, the terms held and assigned have come to be associated with a particular knowledge tradition, which makes implicit epistemic assumptions that values are pre-formed, discretely observable mental entities. These assumptions are shared by some but contested by other traditions (we will discuss this further in the next section). Thus, transcendental and contextual should not just be seen as new terms for Rokeach's old concepts, but along with indicators as a way to distinguish between the three main meanings of the word value that does not seek to prejudice towards the implicit epistemic assumptions associated with the idea that values can be held or assigned. Kenter et al. (2015) also discriminated between diferent types of shared and social values along dimensions of value provider, scale, intention, and elicitation process. To provide for a framework that can tackle knowledge of a broader multidisciplinary nature, such as relected in this feature and other important recent knowledge developments in the ield (e.g. a special issue of Current Opinion on Environmental Sustainability on relational values; Chan et al. 2018), we add further dimensions and organise them in relation to the value, epistemic and procedural lenses ( Fig. 1; Table 1). We add two dimensions associated with the value lens: value frame and value justification. These dimensions categorise values in relation to framings of how the world matters to people and diferentiate between intrinsic, relational and instrumental values. Within the epistemic lens, abstractness, constructedness, normativity and rationality denote whether values are considered abstract or place-based, pre-formed or constructed, descriptive or normative, and which conception of rationality justiies them. Associated with the procedural lens, the closely related dimensions of aggregation, integration and power denote diferences in the way that value plurality and conlict are perceived and managed.
As an example of the relations among value, epistemic and procedural lenses, and some of their associated dimensions, in this feature O'Connor and Kenter (2019) investigated a particular type of social values, 'articulated intrinsic values', focusing on marine ecosystems employing ethnographic stakeholder interviews. The social value lens was the worthiness of the 'more-than-human' world, relecting other-regarding values on the dimension of intention, individual and communal values in terms of scale, and intrinsic values in relation to justiication (Table 1). The underpinning epistemic lens in this research was interpretivist and perspectivist. This lens characterised values as place-based and situationally constructed on the dimensions of abstractness and constructedness, and the dimension of normativity highlights an epistemic lens that seeks to emancipate the more-than-human world. The authors deployed a procedural lens along the dimensions of elicitation and aggregation that highlighted the purpose of the exercise as feeding into a deliberative democratic process that should be used to weigh and aggregate the diferent values expressed. The procedural lens thus emphasised meta-values of participation, deliberation and, in relation to the power dimension, fairness in terms of procedural and recognition justice, through which the social value lens of articulated intrinsic values was considered.
We do not claim this new set of dimensions fully and inally articulates all possible lenses and meta-lenses across the vast diversity of social values literature. However, it relects a substantial extension and evolution of understanding from Kenter et al. (2015), which was largely grounded in ecological economics, to a much more comprehensive interdisciplinary underpinning.
Combining and comparing social values across theoretical traditions can lead to tensions, because these traditions utilise diferent social value lenses and meta-lenses relecting diferences in the way values are conceived, elicited and applied. Following Goldstein (2015), a commitment to conceptual and theoretical openness in interdisciplinary teams generates conceptual tension at various levels, to diverse degrees, and to variable efects. In turn, tension and conlict open up established theories and concepts for dialogue and revision. Lenses and tensions are closely related, because lenses can be seen as a key source of tension, or conversely, are themselves characterised by one's position in areas of tension. Thus, the notion of lenses and tensions across different dimensions provide a useful means of scafolding to 'frame the mess' of diversity in the broad ield of social values.
Tensions can arise at the level of the content of values, value lenses, and meta-lenses. For example, a typical conlict between pro-development and pro-conservation values is not just a matter of valuing diferent things, as what values are included will be diferent depending on the dimensions of the value lens used, such as its scale (e.g. individual, communal, societal), and on the epistemic and procedural meta-lens underpinning it, such as in terms of what value justiications are considered and how questions of value aggregation and power are addressed.
We discuss seven key areas of tension. We focus on the following, gradually shifting emphasis from ontology of social values to their application: (1) social values as aggregated from the individual scale vs being pre-aggregated at a social scale (related to the value provider, scale of values, and aggregation dimensions); (2) social values as discrete, pre-formed and held vs being embedded, implicit, and constructed through their manifestation in deliberation and action; and (3) values as static or changeable (both related to the constructedness dimension); (4) social values through a descriptive vs normative lens (intention and normativity); (5) the relations between social, shared, relational, intrinsic and instrumental values (frame and justification dimensions); (6) tensions relating to value integration; and (7) tensions in the degree to which power is acknowledged in navigating value conlict (power dimension).

Scales and aggregation: the relations between individual and social values
The irst area of tension arises from a basic question: what makes social values social? Although interpreted diferently, essentially the idea of values being social relates to society. This raises the question of how society and its values are represented, particularly whether societal values are considered an independent construct or an aggregation of individual values. Thus far, most lenses have either focused on individualistic or collective indicators, and there is only limited understanding of the relations between them (Kenter et al. 2014a).
Some social values cannot be reduced to the individual scale of expression. As an example, take the UK Marine Policy Statement, which formalises a "shared vision" of "clean, healthy, safe, productive and biologically diverse oceans and seas". 1 This signiies shared social values across value lens dimensions: the statement was made by governments to represent society as a whole and express transcendental values at the social scale, established through a social process. Individuals are socialised: therefore all individual values relect social values to a certain degree (Kenter et al. 2015). Individuals may also experience, represent and enact shared values such as expressed by the policy statement above. However, how can individual values be aggregated to form social values?
The relationship between individual values and social values at a societal scale (i.e. as value to society in terms of contextual values, and as values in relation to society in terms of transcendental values), can be thought of in at least ive ways, which inform diferent social value lenses (Fig. 2). In Fig. 2a, at the contextual value level, the aggregate of individual and social values are diferent but may overlap, while at the transcendental level people may express multiple sets of potentially overlapping and clashing values (e.g. consumer values vs citizen values; Kenyon et al. 2001). The second perspective is that of a nested diagram (Fig. 2b), which indicates that any method of aggregating values, whether through analytical approaches or deliberative processes, is bound to exclude some, typically because of power relations (Ernstson 2013;Hockley 2014;Orchard-Webb et al. 2016). Under this model, social values are always a subset of the pool of individual values and rarely approximate the totality of pooled values. The third and fourth igures (Fig. 2c, d)   Further research on the interrelations between individual and social values is needed within and across each of the ive models in Fig. 2. Also, comparative research between the overlapping, nested, causal, and dynamic perspectives will be of particular value in considering what factors inluence the diference between (aggregate) individual and (preaggregated) social values, and how values transfer between these levels. Moreover, the use of more than one model will likely provide added insight into complex and contested issues, as investigation of shared values is particularly salient in situations of social conlict and disagreement among interest groups (Kenter et al. 2014b). All of these models are sensitive to the diferences between aggregate individual and social values, and as such, it is important for policy and practice to recognise what might inluence degrees of diference, and how these diferences relate with associated lenses used to assess values. Greater understanding of the diferences between aggregated individual values and social values will also enable researchers to identify appropriate methods for establishing a more comprehensive perspective.
Extending the ontological tension between (aggregated) individual and social values, we further complicate this relationship and distinguish social values across multiple levels of value providers. Relationships between individual and social values function within complex systems can be organised hierarchically . Previous research has identiied and grouped values provided at the individual, group (extended) community, and whole culture and society levels (Manfredo et al. 2014;Kenter et al. 2015) that accommodate interactions within these hierarchies. In line with arguments that values 'scale up' to higher levels , the values expressed by groups are in part an aggregate of individuals' values but may also be entirely new 'emergent' phenomena (Fig. 3). In addition to this ontological tension between individual values and values of larger social units, there are practical tensions between values that exist at diferent provider levels. This tension is generated by value hierarchies in inding sustainability solutions, as well as processes for aligning values across multiple scales such as the need for decision-making processes to prioritise between the values of smaller versus larger collectives. This is further complicated by diferent procedural lenses on the commensurability and compatibility of values and lenses (see section "Constructedness and value change: social values as stable vs changeable"), and on how to navigate conlict and address power issues (see section "Value integration and rationalities").
We distinguish two mechanisms by which values are transferred between levels. The irst is socialisation, and it occurs over extended periods of time (Ishihara 2018) as well as when values are formed in shorter-term social processes, such as group deliberation (Kenter et al. 2016c;Kendal and Raymond 2019). The values that emerge from socialisation can be solidiied through social learning and social norms that regulate practices within a collective (Irvine et al. 2016). The second is internalisation. Over time, individuals observe interpersonal dynamics and adjust their orientations to align with a group (Calcagni et al. 2019. This is grounded in personal relections and intraindividual deliberation. Together, these mechanisms can yield changes in systemic understanding of others' values, improved capacity for individuals to recognise their own orientations, and knowledge of why changes in values occur at diferent levels of social organisation (Kenter et al. 2016c).

Abstractness and constructedness: social values as discrete and pre-formed vs embedded and constructed
An in important tension between diferent epistemic lenses revolves around whether values are believed to exist as discrete entities, pre-formed and held by people, or as only coming into existence when manifested, including in deliberation (Ravenscroft 2019) and as 'lived values' in individual and collective behaviour (Brear andMbonane 2019, Gould et al. 2019). In terms of contextual values, valuation researchers have pointed out these are frequently poorly formed in unfamiliar environmental contexts (Jobstvogt et al. 2014a, b;Urama and Hodge 2006). However, there is also a tension in the conception of transcendental values as: (1) held as, (a) relatively singular and stable across a human lifespan, or (b) multiple sets of contextually activated values; or (2) not held but constructed and manifested in response to individual, group and social-ecological context. This tension also relates to the dimension of abstractness associated with epistemic lenses and discussed in more detail by Rawluk et al. (2019). This dimension clariies whether values are seen to be: (1) distinctly isolated as an abstract, discrete entity The tension between epistemic lenses that see social values as abstract, discrete and held vs embedded, situationally constructed and manifested has important implications for social valuation: from the irst perspective, associated with for example social psychology, conventional economics and public participation GIS, values are considered as self-existing mental entities that can be isolated and interrogated. In other perspectives, associated with humanities and deliberative ecological economics, values are understood as embedded in cultural and institutional contexts. Here the language of value 'capture' becomes inappropriate (Ravenscroft 2019). These latter perspectives are also less likely to see values within a power vacuum, rather considering them as part of an institutional setup shaped by discursive structures of power and knowledge-we will return to questions of power in "Value integration and rationalities". Some synthesis between the two positions is possible through the concept of proto-values , where people neither hold fully formed values nor are they an evaluative tabula rasa. Proto-values mediate between the transcendental (broad) and contextual (speciic) concepts of values, and between the abstract and pre-formed and constructed and situated. They are not fully formed values but exist as a broad value-inclination or attitude that becomes more moulded by and embedded within context through a key set of institutional and contextual process factors, which can include the lens and meta-lenses of the particular social values tradition. Proto-values provide an avenue for allowing some generalisation, whilst acknowledging valuation as a process of value formation that is highly context-dependent. However, the concept is in need of further development and empirical exploration.
Constructedness and embeddedness also raise questions about the social and spatiotemporal scales within which this embedding is situated. Scales inluence how research is conducted and looking through difering spatiotemporal value lenses can yield conlicting perspectives on sustainability solutions (Gunton et al. 2014). Future research should be sensitive to the efects of spatial and temporal variation in values and focus on mechanisms that can bridge multiple spatiotemporal lenses. Deliberative and interpretive participatory mapping exercises could consider how spatially explicit social values are culturally and institutionally embedded. This approach could also provide insight on how values map onto the geographies of relevant environmental conditions, evaluate how group deliberation can synthesise values across a range of spatiotemporal scales, and reconcile mismatches between scales of people's values and ecosystem processes. Furthermore, the degree to which values are seen as isolatable from the contexts of place, time and culture will inluence the types of interventions that are considered: whether it makes sense to develop generalised interventions focused speciically on encouraging pro-environmental values, or whether they should be highly situated and placebased, or focus on a value formation process that is geared towards activation and translation of proto-values to particular contexts.

Constructedness and value change: social values as stable vs changeable
A further tension related to the epistemic dimension of constructedness is whether values are perceived as stable or changeable. This is of particular importance and increasing debate within the sustainability ield, because the degree to which values are pre-formed and stable will more generally determine the usefulness of interventions targeting values (e.g. mindfulness or targeted deliberations) as a strategy for sustainability transformation. This debate within the context of pro-environmental value and behaviour change is most relevant to consideration of transcendental values. These are generally seen as more stable than contextual values (Schwartz et al. 2012), yet they are expressed to diferent degrees depending on the salience of issues (Trope and Liberman 2010) and centrality to the evaluator's identity (Stets and Burke 2000). Both across and within traditions such as social psychology, deliberative ecological economics and sociology, diferent procedural and epistemic lenses conlict in terms of their perceptions on how easily transcendental values can be changed (e.g. Manfredo et al. 2017;Raymond and Kenter 2016;Everard et al. 2016;Ives and Fischer 2017). Others argue that a notable gap between transcendental values and actions (Kollmuss and Agyeman 2002) makes this mission irrelevant. That is, while values may activate certain behavioural intentions, constraints posed by people's environments limit their expression. In contrast, the positive psychology literature (Raymond and Raymond 2019) does not focus on value change but instead on individuals acting congruently with their values, with congruence associated with higher wellbeing and psychological health. This literature brings a strong focus to how values are operationalised and behaviourally manifested in diferent contexts and to awareness raising processes to deliberate on and express values within context, including speciic decision-making processes. Drawing on Bardi and Goodwin (2011), awareness raising represents a 'priming' process for value change and/or expression. Awareness raising processes can be considered across two pathways: a healthy values pathway whereby certain value types are associated with healthy outcomes for the individual, and a value activation pathway which considers whether selfidentiied values are congruently expressed . Mindfulness, operationalised as (1) awareness ('what is mindfulness'), (2) skill (mindful awareness of values in decision-making) and (3) mindset (mindful orientation), is a way to promote wellbeing and sustainable behaviour through the pathway of value activation. Mindfulness has thus emerged as an important process variable to understand the elicitation and expression of values (Wamsler et al. 2018) with clear relevance for sustainability science. However, thus far the value lens of positive psychology has almost solely focused on internalisation with individuals. To act as mediator for value change at the communal, cultural and societal level, mindfulness also needs to be linked to socialisation processes.
Sustainability science is increasingly focused on the causes and efects of change, and values can be conceived of as both a driver and an outcome of that change. Societal values form the foundation of institutional rules and knowledge systems that are part of managing and governing natural resources (Gavin et al. 2018). At the same time, complex environmental change such as climate change can become a catalyst for changes in values (O'Brien and Wolf 2010). Crisis triggered by natural hazards shortens even more the feedbacks between values as drivers and outcomes. As such, the opportunity space for responses to risks is delineated and shaped by deliberated, reconciled societal, communal and group values, but at the same time, crisis may be the most rapid trigger for radical changes in our principles and life goals, and this in turn is likely to afect contextual values. When the consequences of environmental changes become evident for people, they may become more aware of the diverse values of nature, compared to 'normal' times. Crises of natural resources or climate change thus can become opportunities to form and (re-)connect to shared values of nature if the focus is on how to think and act together towards these values.
Further research is needed that considers to what degree and how rapidly transcendental values can change, why a focus on contextual factors and values may not be suicient (IPBES 2019), how does value change 'ripple out' ) to the societal and cultural level, what interventions are most efective at achieving such change, and to what degree value change acts as a precursor to or an efect of changing behaviour. There is also a need for interdisciplinary scholars to reconcile the approaches of value change and value congruence, notably if wellbeing as a construct is considered as much a process as an outcome, and to relate individualistic processes such as mindfulness more strongly to social values, socialisation processes and social outcomes.

Normativity and intention: valuation as descriptive vs normative
The next tension relates to whether the formation and understanding of social values is perceived as teleologically normative: a critical, emancipatory, and potentially transformative afair ( (Ravenscroft 2019) or derived from people's lived experience (Brear and Mbonane 2019). However, this raises important questions of procedural and recognition justice, of what, and whose perspectives should be included within consideration of the common good and by what criteria this can be validated. Though the emancipatory tradition typically focuses on maximising inclusion (Lo and Spash 2012;Orchard-Webb et al. 2016), this does not mean that all individual values should be included or aggregated, for example where they do not serve society as a whole (Sagof 1986), or are incompatible with sustainability (Menzel and Green 2013).
Interestingly, a similar tension between descriptive and normative exists with regard to relational values, that can be discussed as a matter for observation (Calcagni et al. 2019;Klain et al. 2017) or an agenda for inclusion and emancipation of non-scientiic knowledge (Gould et al. 2019;Stålhammar and Thorén 2019). However, while relational values are not generally put forward as 'better' than instrumental ones, the normative tradition clearly advocates social values as more desirable than individual ones for the purpose of decision-making, as long as the conditions of procedural and recognition justice are reasonably satisied (Howarth and Wilson 2006;Irvine et al. 2016;Kenter et al. 2016b;Ravenscroft 2019;Zografos and Howarth 2010). This relates also to the dimension of intention on the value lens: when valuation is seen as transformative, it is unlikely that this purpose seeks values to be more self-regarding. Rather, social values are seen as oriented towards more other-regarding perspectives and the consideration of society as a whole (and which may also include non-humans; O'Connor and Gould et al. 2019). From this perspective, the individual self-regarding preferences associated with market decisions should not be transferred to the social value domain of the group, or public policy (Mauss 1954;Lo and Spash 2012;Kenter et al. 2014a;Irvine et al. 2016;Ravenscroft 2019). As Sagof (1986, p303) put it: "Why is it good in itself that a person who wants a Mercedes succeed in getting one? Having a preference is a reason for the person who has it to try to satisfy it. Why should it be a goal of public policy, however to satisfy that preference?".
However, the clear division of self-regarding consumer vs more other-regarding citizen values is seen by some lenses as an abstraction, such as in care ethics and relational axiology, where identities and values are seen as deined by embodied, reciprocal relationships and values, and where self and other are not discretely separated as diferent objects of regard (Held 2006;Muraca 2016;Gould et al. 2019).
The tension relating to the intention and normativity dimensions of value and epistemic lenses can in part be resolved by recognising that diferent positions along these dimensions typically correspond to diferences at value lens dimensions of scale, provider and concept and the procedural lens dimension of process. Normative meta-lenses are particularly focused on shared and social values in the sense of value to society, formed through a shared social process, and/or expressed by non-individual value-providers. Transcendental values, particularly those relating to environmental sustainability and social justice, are important in the sense that they are seen as ends that steer those processes and that play an important role in contextual value formation, but they are not generally the primary objective of study. There is a goal of providing evidence for interventions, but the social valuation itself can also be seen as an intervention to transform values and/or behaviour or challenge existing institutions. In contrast, descriptive meta-lenses more typically focus on either social values as aggregated individual values, or social values in the sense of transcendental values, mostly by individual providers, and the relations between transcendental and contextual values and behaviour to provide evidence for exogenous interventions. In a small number of studies, the two approaches have been successfully combined where descriptive approaches inform or are integrated with consequent deliberation on the common good (Kenter 2016b;Kenter et al. 2016b;Raymond et al. 2014;Raymond and Kenter 2016;Borsuk et al. 2019). Martino et al. (2019) illustrate that the need for descriptive vs normative valuation also depends on endogenous characteristics of diferent environmental goods, and how these characteristics interact with social institutions. Further research may consider more deeply when descriptive and normative approaches are most appropriate and when combinations of both add particular value. There is also a need for more explicit evaluations of recognition and procedural justice in critical social valuation, and to what degree, and under which conditions, the transformative objectives of social valuations are met.

Value justiications and frames: shared, social and relational values and our relationship with the natural world
In recent years, the increasing emphasis on social values within the sustainability ield has arisen in parallel with increased attention to relational values, particularly with regard to ecosystem assessment (e.g. Chan et al. 2018Chan et al. , 2016Muraca 2016;IPBES 2016;Christie et al. 2019b). Several contributions to this feature have considered relational values and their relation with shared, social, instrumental and intrinsic values (Calcagni et al. 2019;Stålhammar and Thorén 2019;Gould et al. 2019;O'Connor and Kenter 2019). Both relational and social strands of thinking evolved, at least in relation to ecosystem assessment and valuation, from increasing recognition of the limitations of mainstream economic valuation and its instrumental value assumptions. This opened up a fuzzy ield of nonmonetary, social, cultural or sociocultural values, largely associated with the study of cultural ecosystem services and to some degree indigenous and local knowledge systems, that used a wide array of methods but without much attention to underpinning value, epistemic and procedural lenses (Raymond et al. 2014;Scholte et al. 2015). Two interventions signalled diferent directions: Kenter et al. (2014aKenter et al. ( , 2015 focused on clarifying the concept, dimensions and types of shared and social values as critiques of the pre-formed, individualist and self-regarding assumptions of mainstream valuation. While shared and social values were considered largely synonymous, social values tended to emphasise social scales whereas shared values tended to refer to social value providers and the outcomes of collective value formation. This discourse (further developed in a special issue of Ecosystem Services, October 2016, mostly by authors associated with the UK National Ecosystem Assessment follow-on; Kenter, 2016a) articulates strongly the social nature of values and the long and short-term processes for socialisation and internalisation of values, with particular regard for integrating deliberative and interpretive approaches as a preferred methodology for assessing shared values (e.g. Orchard-Webb et al. 2016;Edwards et al. 2016;Ranger et al. 2016). Relational values, in the sense of values pertaining to meaningful, non-substitutable relationships between people and their environment, were considered, but primarily from a perspective of their shared-ness within groups, communities, cultures and societies.
In a diferent intervention, Chan et al. (2016) deined relational values as preferences, principles, and virtues pertaining to relationships. They argue that in practice the instrumental and intrinsic value concepts central to environmental ethics did not capture what matters most to people, and that a distinct bridging concept was needed. This concept has since been elevated to central importance in IPBES (Pascual et al. 2017;Díaz et al. 2018;Christie et al. 2019b). As with social values, the scope of relational values is broad and fuzzy (Stålhammar and Thorén 2019). Relational values can refer to the ethical nature of value as being anthropocentric, yet non-instrumental, in the sense of not open to trade-of (Díaz et al. 2015;Himes and Muraca 2018); or it can relate to the content of transcendental or contextual values as pertaining to relationships (Klain et al. 2017;Gould et al. 2019); or it can refer to a 'relational ield' as the source of value, rather than the value object or subject (Muraca 2011(Muraca , 2016. Importantly, while the dichotomy between intrinsic and instrumental is typically conveyed as a major tension in environmental debates, Stålhammar and Thorén (2019) point out that these value types are somewhat caricatured, and that environmental ethics has long had more nuanced interpretations of instrumental and intrinsic values that are inclusive of relational value justiications (e.g. Naess 1973). As such, the ambition of the relational intervention is perhaps more pragmatic than theoretical, in advancing recognition of the relational nature of how people talk and think about values (Chan et al. 2018).
Relational values as a boundary concept has thus strategically focused on the dimension of justiication. In contrast, the challenges posed by shared and social values to instrumental values are not resolved by developing a noninstrumental concept, but by pointing to the importance of the collective level, understanding the intersubjectivity of values, and development of pluralistic boundary concepts and processes for sharing, aggregating and integrating values that are inclusive of multiple value justiications. As such, shared and social values, and relational values, are complementary constructs both essential for inclusive valuation. Furthermore, the two are closely related; it is hard to imagine any relational values that are not in one dimension or another shared or social (Ishihara 2018), and relationships with nature and place are central in environmental social value formation processes (Kenter 2016a;Ranger et al. 2016;Edwards et al. 2016;Ainsworth et al. 2019;Calcagni et al. 2019;Gould et al. 2019).
An altogether diferent approach to inclusively communicating values is presented by O'Connor and Kenter (2019), who build on O'Neill et al. (2008) to develop the Life Framework of Value, which moves beyond value justiications to consider valuation in terms of diferent frames. Here, values are presented simply as what matters, and in relation to the environment this can be framed as living from, with, in and as the world (Fig. 4). Living from relects the value of the world as a resource, providing for our sustenance and livelihoods. Living in the world points to its role as place, as the stage for our lives. Living with the world points to non-human nature as the other, with whom we co-exist, and which has its own purposes, patterns and cycles. Living as the world points to more-than-human nature as self, with nature and non-humans constitutive of our bodies, psyche Importantly, while relational values may be particularly associated with living in and as frames, and intrinsic and instrumental values with the with and from frames, the diferent justiications straddle the frames, pointing to the entwinement of multiple ethical categories in our common experience. For example, a farmer clearing forest for shifting cultivation may be seen through a living from frame, but his livelihood is also likely to be the source of meaningful, non-substitutable relational values, and his clearing activities could support the intrinsic good of biodiversity (e.g. Bayliss-Smith et al. 2003). The authors note that "O'Neill's way of phrasing values in relation to 'living' intuitively imbues a sense of egalitarianism between diferent values" and "its elegance incites a natural inclination towards including each of the categories [in valuations and decision-making]" (p. × 2). Diferentiating between value frames may be an easier way of communicating nature's values to a broad audience than through more abstract value justiications.
However, further research and debate is needed to better align the need to communicate values efectively in a way that resonates with citizens and policy makers, such as through the Life Framework and relational values, with rigorous explanation of the relationships between diferent categories. Further, research is needed that builds on discourses and approaches associated with shared and social values to ind ways of resolving practical tensions in sustainability practice between diferent value justiications and frames, enabling more efective value integration.

Value integration and rationalities
Diferent forms of value integration were presented in this Special Feature. Papers examined the integration of diferent types of values, including for example across diferent scales of values ) and provider Fordham and Robinson 2019) and across diferent value justiications (Christie et al. 2019b;O'Connor and Kenter 2019;Kronenberg and Andersson 2019). Integration becomes more complex when aligning between epistemic lenses that difer in terms of abstractness and constructedness , for example between values that are lived or embodied, where value is seen as dynamically situated ) and more objective approaches where values are seen as stable across situations. Such questions point to an urgent need to consider new forms of value integration. Gunton et al. (2017) argued that we need value frameworks that can integrate the place of interest and the scale and subject of interest. They proposed a suite of considerations for valuing ecosystems (e.g. social, economic, aesthetic, jural, sensory, symbolic), to be compared with diferent stakeholder groups and across diferent types of places. These frameworks point to the diference between concept and method integration (Kronenberg and Andersson 2019;Guerrero et al. 2018), although arguably method integration needs to be underpinned by conceptual integration, at least if one wishes to avoid unconscious pragmatism where no attention is given to how tensions along diferent lenses are resolved (Raymond et al. 2014). Most social values for sustainability papers do not discuss the interface between conceptual and methodological integration, and this is an important avenue for future research.
Value integration can achieve diferent levels or purposes. Kronenberg and Andersson (2019) distinguished full integration through a single measure, comparison of results from selected combinations of methods, and parallel use where there are multiple diicult to compare sets of values. The level of value integration achievable will depend on the degree to which values and value dimensions are deemed logically and axiologically commensurable and value indicators and valuation methods technically compatible.
Diferent integration levels tend to be informed by different epistemic lens perspectives, including with regard to rationality. Full integration and its demands for commensurability is strongly tied to instrumental rationality (Lockwood 2005). Here the focus is on the choice of suitable indicators where diferent values are traded-of in the search for an optimal outcome. Values are treated as commensurable by looking to measure them according to a common scale and thus aggregated into a single value indicator, such as through monetisation and many forms of multi-criteria analysis. More limited combination approaches can be linked with bounded rationality, where doing well enough rather than optimising choices is inevitable in many contexts (Simon 1984). Here, incommensurable values may be compared ordinally or nominally (e.g. improvement vs degradation). Values from diferent methods can also be brought together on the basis of communicative rationality, where values are weighted on the basis of the force of argument rather than analytical criteria, and the realisation of ideals of non-coercion and inclusivity determines to what degree outcomes are rational (Habermas 1984). Here, plural values may be thought of as weakly comparable, which means they should only be compared in terms of practical judgement, rather than on a common scale (Martínez-Alier et al. 1998). For example, O'Connor and  propose the integration of intrinsic, instrumental and relational values through discussion structured according to the multiple Life Frames, where people are invited to relect on how diferent policy scenarios might afect both people and the more-than-human world. Irvine et al. (2016), Kenter (2016a) and Ravenscroft (2019) point out philosophical challenges around deliberative value integration, relating to how much diferent representations and value criteria count, and urging further empirical investigation of how deliberative valuations can act as new democratic spaces for integration based on social learning and communicative rationality.
Given these diferent value integration rationalities, how can policymakers recognise the diverse values of nature? Here, we need to accept the rationality of 'value pluralism' in that value diversity is an outcome itself. Diverse authors (e.g. Larmore 1987;Kekes 1993) have argued that conlicts between values are not always resolvable. The notion of moral conlict (Stocker 1990) suggests that ethics need not always be action guiding. Instead, respecting plurality involves recognition of diverse pathways of policy formation and implementation. This will require a shift in the culture of policy making and associated capacity building to promote awareness of diverse value traditions and practice in grappling with multiple value lenses and meta-lenses. In terms of deliberation, the principal outcome sought may not be the kind of consensus solutions associated with Habermasian ideals, but providing a forum for plurality, capacity and trust-building. Furthermore, even where values cannot be agreed, it may still be possible to agree on shared meta-values around how to deal with divergent or conlicting values and lenses.

Values, conlict and power
The consideration of plural values and the challenges of integrating them with each other and into decisions raises key questions of power in navigating such conlicts. Despite real consequences, the interplay between values and power continue to be neglected, especially in empirical valuation studies. This is in part due to the multifaceted nature of both values and power. Power can be both overt and almost imperceptible and exercised through hegemonically privileging certain lenses and meta-lenses (Foucault 1980;Lukes 2005). Power dynamics can inluence whose values are expressed or recognised, and which values emerge in contexts, though this is not necessarily transparent. Researchers and practitioners of sustainability science must become attuned to recognising and navigating power as expressed through values and the lenses by which we examine them.
The interplay between social values and power can occur in many ways. A dominant scientiic framing of sustainability privileges one way of knowing, which can depoliticise inherently political challenges (Sletto 2008). Examples include the concepts of the Anthropocene (Haraway et al. 2015) and sustainability itself (Ferreira 2017), which homogenise social drivers apolitically. Consciously or unconsciously privileging one set of social value lenses over others can manifest in social-ecological injustices (Collard et al. 2018). Certain values (e.g. economic, moral, religious, scientiic, etc.) of particular groups (e.g. diferent social classes) will be favoured in policy and decision-making through exercise of power, for example through privileging of economic value above all else (Demaria 2010). In contrast, lenses and languages of valuation, and consequently values, associated with indigenous and local knowledge systems are often ignored in decision-making (Martinez-Alier 2009; Christie et al. 2019b). Further, meta-values encapsulated in dominant lenses, including ideas of social memory and how the future should be (Rawluk and Curtis 2016), can cause people to silence values that don't match expectations, including in deliberative processes (Orchard-Webb et al. 2016;Lo and Spash 2012).
There are many ways in which power can be exercised in order to direct, control or regulate the conduct of people, in overt and subtle ways (Foucault and Rabinow 1997), for example, through discursive strategies of power-knowledge embedded in diferent 'governmentalities', such as Sovereignty, Discipline, Neoliberalism and Truth, as an art of government (Foucault et al. 2008). According to diferent technologies of power exerted in a historical context, these governmentalities afect the values that people are able to adopt in their lives. Given that values are crucial aspects of the choices, decisions, and behaviours of people related to sustainability, the interplay between how power is exercised, the values that people adopt, and the construction of individuals' identities is key to understand environmental governance and its outcomes (Agrawal et al. 2005; D'Alisa and Kallis 2016).
While ontological and epistemological diferences can be a source of contestation , tensions around power inevitably arise in relation to any form of social values assessment in practice, though are often not acknowledged. In particular, there is a need for more attention to power relations in diverse processes of value formation, socialisation and internalisation, such as by Calcagni et al. (2019) who consider the impact of communication and market strategies inluencing value creation on social media. Even in deliberative value formation characterised by ideals of non-coercive communicative rationality, such ideals can only be approximated, as in the real-world, unconscious power relations cannot be fully ironed out (Orchard-Webb et al. 2016;Kadlec and Friedman 2007). Intractable controversies inevitably remain, demanding speciic, contextual answers at the level of practice (Pellizzoni 2001). Further, in sustainability practice, an important barrier for realising pro-sustainability social values are people's limited power and control to change their unsustainable practices resulting from unmet wellbeing-related needs (Brear and Mbonane 2019;Huxley and Yiftachel 2000).
Thus, sustainability that manifests social-ecological justice requires centring on both social values and power.
If other-regarding transcendental values that underpin the ethos of sustainability, such as equity, generosity and care, are to be promoted, there is a precursory requirement to transparently observe diversities of values and needs alongside privileging mechanisms of power. In the ield of valuation, scholars may need to become more comfortable with relational and post-structuralist meta-lenses, since power is observed more easily through these (Foucault 1980). In line with Geels et al. (2017), Smith et al. (2005) and Everard et al. (2016), considering values through multiple value, epistemic and procedural lenses is critical because socio-technical transition pathways towards sustainable systems imply necessarily value-oriented governance systems, which are afected by the interplays between technologies of power, the institutional system, and the processes of pro-sustainability value socialisation.

Conclusions
In this paper, we have considered key theoretical and practical tensions in the burgeoning ield of social values of sustainability. These tensions relate to important dimensions of values that characterise the lenses and epistemic and procedural meta-lenses through which diferent traditions conceive and perceive these values. Key avenues for future research relating to these tensions include: 1. Exploration of relations between collective and individual values, and the dynamic internalisation and socialisation processes by which values transfer up and down between individuals and multiple social scales of value provider; 2. Investigation of crisis-triggers for pro-sustainability value-change and levers for 'rippling out' changes; 3. Conceptual development and empirical exploration of proto-values; 4. Application of interventions based on value awareness, activation and congruence within sustainability contexts and their upscaling from individualistic to social; 5. Evaluation of values-based interventions that take a generalised vs place-based perspective; 6. Further development of the Life Framework as a novel way of organising and communicating why the natural world matters; 7. The interface between value, conceptual and methodological pluralism, value integration and comparative and combined use of multiple rationalities for valuation; 8. Deliberative mechanisms to address conlicts between values at diferent spatiotemporal and social scales, between diferent value justiications and Life Frames and between diferent value and epistemic lenses; 9. More explicit evaluations of recognition and procedural justice in critical social valuation, and under which conditions transformative objectives of social valuations are met; 10. The interplay between how power is exercised and the values that people adopt across diferent institutions and contexts; 11. The development of new languages of valuation that are better relective of relational, constructivist and poststructuralist perspectives; 12. Understanding mechanisms whereby certain lenses are privileged over others in diferent decision-contexts, and capacity building for understanding and drawing on multiple value, epistemic and procedural lenses in decision-making.
The large number of dimensions of values to which these questions point relect that sustainability issues are by and large complex and wicked problems. Addressing such issues requires us to navigate transcendental and contextual values at multiple spatiotemporal scales, between individuals and collectives, across diferent potentially conlicting value justiications, frames and rationalities, and with close attention to power relations in such conlicts, both within and between diferent value articulating institutions. Efective navigation requires charts, beacons and experience. This paper has sought to scout the terrain providing a multidimensional interpretation of the messy social values landscape. Such a map is crucial in communicating with fellow travellers where one is, in the sense of what values one is articulating and from which vantage point. Understanding of tensions provides beacons to shed light on crucial areas of conlict, where we need to pay particular attention in our journeys of sustainability science and practice. At these points, experience of engaging not just with the landscape and its map but with fellow travellers becomes vital, as the terrain is too challenging for any tradition to tackle on its own. Crucially, all values around sustainability have a social dimension. A juxtaposition between individual instrumental values and social, shared, cultural, non-instrumental or relational values is thus not helpful-rather we must help each other understand what dimensions of the value landscape we are viewing and through what lens. At these junctions, by loving the mess and enjoying the thrill of exploration, conlict can become a space of creative dynamism where new concepts, methods and approaches can be born. The mess does not need resolving but engaging with. This requires building capacity with researchers and practitioners: learning to navigate and learning to love, by embracing the plurality of how we conceive and articulate values in research, decision mechanisms and boundary spaces-all are ultimately social processes of valuation.