‘Relational Values’ is Neither a Necessary nor Justified Ethical Concept

ABSTRACT

‘Relational value’ (RV) has intuitive credibility due to the shortcomings of existing axiological categories regarding recognizing the ethical relevance of people’s relations to nature. But RV is justified by arguments and analogies that do not hold up to closer scrutiny, which strengthens the assumption that RV is redundant. While RV may provide reasons for ethically considering some relations, much work remains to show that RV is a concept that does something existing axiological concepts cannot, beyond empirically describing relations people have to environmental areas and places.

KEYWORDS:

• Relational values
• instrumental values
• extrinsic values
• intrinsic values

1. Introduction

Many share a strong intuition that people’s relations to, for instance, environmental areas or places matter ethically and that neglecting such relations is a shortcoming and unjustified abstraction of ethical life and practical thought. Relational value (RV), which has recently ‘swept the scholarly discussions on ecosystem services’ (Stålhammar & Thorén, 2019, p. 1201) and generated much discussion in environmental ethics, has the potential to justify that intuition. RV also influences practice through the concept of ‘nature’s contribution to people’ used by The Intergovernmental Science-Policy Platform on Biodiversity and Ecosystem Services (Pascual et al., 2017).

But despite increasing popularity and discussion it is not clear what RV refers to and how it is justified. I seek to clarify what RV is and scrutinize how it is justified as an ethical concept. I will assume that RV is an axiological concept and that values are connected to reasons. Thus, X being valuable counts in favor of preserving X or holding an appropriate pro-attitude toward X. Value is thus used in a reason-implying sense (Rønnow-Rasmussen, 2022; Samuelsson, 2010). Hence, X being valuable gives us reasons in favor of X, which are not necessarily deontological reasons but rest on an axiological basis. This ethical conceptualization of RV differs from viewing RV as a descriptive and empirical concept observing that an agent values and relates to an area, place, or species in a specific way (for discussion, see James, 2022a; Stålhammar & Thorén, 2019). RV as a justified ethical concept requires us to ethically recognize relations to nature.

I will analyze RV as an ethical concept carrying ‘normative force’.¹ To some proponents of RV the approach may come off as overly analytical, and as continuing a problematic separation of relationally connected entities. Coeckelbergh (2012) objects to detached and deductive approaches in environmental ethics, and Muraca (2011) argues that conventional approaches emphasize atomization and separation at the cost of neglecting relational complexity. The analytical approach here is motivated by the great need for clarity regarding RV, being an increasingly used concept.

I will stipulate positions illustrating how RV relates to existing axiological concepts, which are taken to show how RV is a justified or needed ethical concept. RV is either an overarching axiological category that includes a variety of other concepts, or supplements existing axiological categories. To the former position, RV includes possibly conflicting concepts and functions as a short-hand for ‘value’. To the latter, justifications of RV are often based on shortcomings of existing categories, but those proposed shortcomings do not do analytical and conceptual justice to existing concepts. Others justify RV through analogies with existing ethical concepts such as eudaimonia, constitutive value, or extrinsic final value. I argue that these analogies fail, but even if they were successful, they defeat the justification of RV.

I will object to RV as an ethical concept and question whether RV does something that existing axiological concepts cannot. Much of what is described as RV should more appropriately be included in existing axiological concepts. To that extent RV need not be introduced. There are some merits to RV as some relations to environmental areas or places provide normative reasons. But that requires conventional concepts to specify which relations that matter and why.

Following this introduction, I clarify some of the entities on the axiological landscape relevant to RV and suggest two general justificatory positions, followed by a scrutiny of arguments for RV that are variations of those positions, before concluding with a summary.

2. RV, Other Axiological Concepts, and Two Justificatory Theses

Assessing RV as an ethical concept requires surveying where to place RV on the axiological landscape. Given the extensive discussions on concepts relevant to RV the overview here will primarily serve as a clarificatory overview for the subsequent analysis. I will propose two justificatory theses for how to situate RV along them.

2.1. Parts of the Axiological Landscape Relevant to RV

If X has value or is valued in an ethically justified way, this counts in favor of X. In the case of RV, it is not clear whether X refers to the valuing agent, the valued entity, or the relationship per se, or some form of combination (James, 2020, see however; Deplazes-Zemp & Chapman, 2021). We can make a distinction between those entities that are valuable due to some inherent characteristic, and those the value of which is dependent on factors external to the valued entity, resulting in the categories of non-derivative and derivative value (Baard, 2022; Crisp, 1998). In the non-derivative category, we find intrinsic and final values, and in the derivative category extrinsic and instrumental value. The entities in the non-derivative category are not dependent on external sources for their value but are valuable due to inherent characteristics, whereas the derivative entities stand in a specific relation to something of non-derivative value. Thus, a bearer of instrumental value stands in a causal means-end relation to something of final value (Baard, 2019; Rønnow-Rasmussen, 2022), whereas a bearer of extrinsic value need not stand in such a causal means-end relation despite its value coming from an external source. In contrast, the bearers of final and intrinsic values do not rely on external features.

Added to the above is the extent to which reasons and values are agent-neutral or agent-relative (Ridge, 2023), concerning the source and characteristics of value. An agent-neutral view of calling something ‘good’ is when we mean ‘roughly that there are certain kinds of fact about this thing’s nature, or properties, that would in certain situations give us or others strong reasons to respond to this thing in some positive way’ (Parfit, 2011, p. 38). In contrast, Crisp (1998) equals mind-dependence, where value depends decisively on the agent, with the thesis that there is no value without valuers. This conflates ‘x having value’ with the descriptive claim ‘x is valued’ (Crisp, 1998, p. 478). Many proponents of RV lay very close to forms of subjectivism about values or agent-relative reasons and neglect how both derivative and non-derivative values come in both forms (Baard, 2022; Sandler, 2012).

There are other axiological categories relevant to RV echoing the above, such as between personal and impersonal value (Rønnow-Rasmussen, 2011, 2022). A piece of paper with a poem written by one’s son carries personal value that differs from the more generally recognized value of a poem by Baudelaire. This grounds a distinction between ‘good for’ and good proper, where the latter category is good without reference to external features while ‘good for’ is not intrinsically good but neither is it necessarily instrumental (Rønnow-Rasmussen, 2011, 2022). The distinction between ‘good for’ and ‘good’ parallels a distinction between relational and non-relational value (Rønnow-Rasmussen, 2022; Rosati, 2009). Stating that ‘X is good for Y’ sets X and Y in a relation and tie the value of X to that relation. But ‘good for’ can be interpreted differently as X can be a good means to achieve goal Y, as in ‘arsenic is good for poisoning’ or boots are good for walking (Rosati, 2009, p. 221), X can relate to an end of a particular kind relating to Y’s functioning, or X can preserve or enhance Y having intrinsic value (Rosati, 2009). ‘good for’ can also relate more directly to living things and welfare subjects, and in such cases ‘X is good for Y’ mean that X promotes life, growth, health or reproductive success of Y, or benefit Y as a welfare subject (Rosati, 2009, p. 224ff). While all these establish a ‘good for’ relation between X and Y, they differ regarding the normative content of the relation and other ethical concepts that do the normative work.

By stating that ‘X is good for person A’, in the sense that ‘X is good according to A or is what A esteems’ one is making a descriptive claim and ‘a philosophically less interesting sense of “good for”’ (Rønnow-Rasmussen, 2022, p. 30), like the mind-dependence described above. A more challenging issue concerns the extent to which all values are potentially constituted by and relate to subjects or interests thereof (Rønnow-Rasmussen, 2022). How this relational understanding of values relates to RV will be critically unpacked below.

Central to my discussion will be needed or justified ethical concepts. I take a needed or indispensable ethical concept E to refer to instances when there is a reasonable judgment that something ought to be done (is obligatory), or refrained from (is prohibited), and where those statuses can only be justified with reference to E, or E plays a central role in the justification of them. For instance, a concept such as final value is needed to end an infinite regress of means and is thus necessary for practical reasoning. If there is a reasoned intuition that something ought to be done or refrained from and this cannot be justified by existing axiological categories, then RV is (possibly) needed if it is the relation between an agent and an environmental entity that provides that reason. Justification of E I take to mean that E is supported by or with reference to reasonable and justified ethical theories and have ethical action-guiding relevance.

2.2. Two Justificatory Theses of RV

At stake is whether RV does something that existing categories do not, meriting the introduction of RV. As some proponents of RV make substantial meta-ethical claims whereas others align RV with existing concepts, and some seem to do both, I will stipulate two theses justifying RV. They will be described at a somewhat general level in this section, and details of different versions of them will be analyzed in Section 3.

There are intuitively merits to RV that justifies adopting it as a normative concept. It is more inclusive than conventional categories by ascribing moral relevance to the many different relations that individuals and groups have to environmental areas and places. Rather than fostering a disengaged view on the environment and assessing its moral status by the application of principles detached from experience, recognizing the relevance of relations in one’s identity and community offers a normative foundation for RV (Muraca, 2011).

Chan et al. state that ‘although intrinsic and instrumental values are critical to conservation, thinking only in these terms may miss a fundamental basis of concern for nature’ (2016, p. 1463). Environmental ethicists are criticized for having ‘tried to determine the value of ecosystems for people by developing theories’ which ‘are too often developed without attention to the actual experiences of the people affected by decisions’ (Norton & Sanbeg, 2021, p. 710). Consequently, it seems as if RV is needed and justified due to the shortcomings of existing axiological categories.

But how does RV relate to the landscape described above? Some definitions of RV are excessively wide, such as RV being the ‘preferences, principles, and virtues associated with relationships’ (Chan et al., 2016, p. 1462). We can call this the ‘overarching thesis’:

Overarching thesis: RV is the name of a set containing many, possibly all, axiological concepts.

An additional example of this view is Muraca who include the ‘merely instrumental’ – or causal in the sense of ‘serv[ing] the particular aim-setting of individuals according to their preferences’ (Muraca, 2011, p. 384) – as part of RV, alongside intrinsic-eudaimonistic values. Both are part of functional-relational values which, together with fundamental-relational values, exhaust RV (Muraca, 2011, p. 384). Muraca thus provides a very broad set of RV that includes likely conflicting concepts, such as between preference satisfaction and fundamental values, defeating the analytical and systematic purpose of the concept. Similarly, the RV set proposed by Chan et al. (2016) includes potentially conflicting elements, such as those between preferences and duties.

If the Overarching thesis holds RV would exhaust much of applied ethics involving values. Norton and Sanbeg (2021) confirm that interpretation when they state that all values are relational, as does Himes and Muraca (2018) and Deplazes-Zemp who maintains that both instrumental and intrinsic values are relational (2023, p. 2). Intrinsic value/inherent moral worth and relational values are respectively the sole coordinates on Muraca’s (2011) ‘map of moral significance’. One can concede that most likely all, or at the very least the great majority of, axiological concepts provides reason to act or refrain from acting in one way or another to a valuable and hence morally relevant entity (including relations of the moral reasoners to themselves). In that regard axiological concepts concern relations. But does this mean that all ethics is relational and that axiological concepts serve to establish a relation between valuer and valued? If that is the case, why the need to specify ‘relational’ if all values are relational (see also Luque-Lora, 2023)? RV serves as a redundant middle-step.

Thus, to some arguments RV functions as an overarching concept that includes preferences, principles, and virtues and extrinsic final, eudaimonic, constitutive, and fundamental values as well as instrumental value. The Overarching thesis includes many of the categories of the axiological landscape surveyed above, seemingly based on the assumption that all, or most of, values are ‘good for’ or agent-relative. Yet, these forms of values are so disparate and some of them conflicting that it is questionable what analytical work RV does that the more generic concept ‘value’ does not. Moreover, more meta-ethical arguments must be provided before committing to the conclusion that all, or most of, values are relational or ‘good for’ and specifying in what sense they are so. Thus, there are strong objections to the Overarching thesis, but an alternative will be discussed in Section 3.

As a different approach RV is justified by being distinguished from intrinsic and instrumental values (Chan et al., 2016; Deplazes-Zemp & Chapman, 2021; Knippenberg et al., 2018; Norton & Sanbeg, 2021). Even Chan et al., proposing the Overarching thesis, argue that there is a need for a third category of values alongside intrinsic and instrumental values. Similarly, Deplazes-Zemp and Chapman suggest that RV is a complementary environmental value category (2021, p. 679).

RV is justified by the shortcomings of existing concepts which it supplements rather than includes. We can call this the Novel and indispensable thesis:

Novel and indispensable thesis: RV does something necessary to ethical reasoning that other concepts cannot.

It is claimed that to gain insight in RV ‘we should first of all shed the dichotomy of intrinsic/instrumental value’, and ‘if something does not have intrinsic value, its value does not need to be merely instrumental’ (Knippenberg et al., 2018, p. 41), implying that if something valuable does not have intrinsic value, instrumental value is the sole remaining candidate. According to this thesis, RV provides an indispensable supplementary axiological concept which provides ethical reasons that current concepts do not. The thesis implies that established values are non-relational neglecting that both intrinsic and instrumental values can be agent-relative (Baard, 2019) or subjective (Sandler, 2012), and that established concepts encompass relations (Luque-Lora, 2023).

In this section I have surveyed parts of the axiological landscape to find elements relevant for RV. I have also stipulated two justificatory theses based on the available arguments for RV serving an analytical purpose as we proceed. There is a tension between these two positions, as it seems that while the Overarching thesis suggests that all, or most of, values, are relational, according to the Novel and indispensable thesis no established concept can encompass relations.

3. RV as an Indispensable and Justified Ethical Concept?

In Section 3.1 I will scrutinize justifications of RV as a reason-implying concept that focus on shortcomings of existing axiological concepts. In Section 3.2 I will focus on arguments that provide a positive account of RV that do not rely as much on the shortcomings of existing concepts.

3.1. Is RV a Needed Ethical Concept?

In this section I will analyze arguments for RV justified through shortcomings of existing axiological concepts, ultimately finding them unsuccessful.

3.1.1. Wrongfully Conflating Extrinsic and Instrumental Values

The Novel and indispensable thesis justify RV by the shortcomings of intrinsic and instrumental values (Chan et al., 2016; Deplazes-Zemp & Chapman, 2021; Knippenberg et al., 2018; Norton & Sanbeg, 2021). While there is vagueness regarding what RV refers to it is clear what it does not refer to (James, 2022a): intrinsic and instrumental values, relative which RV is a third category of values.

This justification can be specified by something like the following argument where an entity, such as an area, place, or territory X is assumed to be valuable:

• P1. If something is valuable, then it has intrinsic or instrumental value

• P2. X does not have intrinsic value

• P3. X does not have instrumental value

• C. Therefore, X is not valuable

The conclusion contradicts the intuition that X is valuable. Proponents of RV argue that P1 is false as it omits other potential axiological candidates. RV is therefore a third potential category to explain and justify that value and solidify the conclusion that X is valuable, as was presumed.

But the falsity of P1 is not because it excludes RV specifically, but that it is too narrow. A fuller, though still incomplete, P1 would read: ‘If something is valuable, then it has either intrinsic or extrinsic value, or instrumental or final value’, as well as other axiological categories such as inter alia aesthetic value or the many senses of ‘good for’.

Moreover, P1 contains a false opposing pair which is costly in this context. Recall that Knippenberg et al suggest that ‘if something does not have intrinsic value, its value does not need to be merely instrumental’ (2018, p. 41), implying that these are the only alternatives. A central category that is often omitted or portrayed problematically in this context is extrinsic value. To distinguish instrumental from intrinsic value is ‘misleading, a false contrast’ (Korsgaard, 1983, p. 170). While the distinction between intrinsic and instrumental values is conventional in environmental ethics and rarely lead to confusion, the analytical opposite of intrinsic value is extrinsic value, whereas instrumental value is opposed to final values (Korsgaard, 1983; O’Neill & Sugden, 1992; Peterson & Sandin, 2013). Extrinsic value means that an entity is valuable by its relation to external properties, being valuable ‘in virtue of its relation to something that is not one of its parts’ (Rabinowicz & Rønnow-Rasmussen, 2000, p. 35), in contrast to intrinsic value. A central difference between external values and instrumental values is that a causal means-end relation between the bearer of instrumental value and the bearer of final value is a necessary condition, whereas such a causal means-end relation is not necessary for extrinsic values (Baard, 2019). For example, ‘rarity’ is a relational property that enhances the values of stamps, but that property is relational to the extent that it is an externally relational feature to other, more common, stamps , even if the stamp has instrumental value.

The consequences of equating extrinsic with instrumental are serious in this case. If one does not separate extrinsic and instrumental values, then one is forced to concede that all valuable things that are not intrinsic are means or instruments (Korsgaard, 1983, p. 171; see Deplazes-Zemp, 2023; Knippenberg et al., 2018). Importantly, this neglect a wide range of alternatives, such as non-instrumental, and non-intrinsic, accounts of extrinsic value. Extrinsic goodness merely recognizes ‘the value a thing gets from some other source’ (Korsgaard, 1983, p. 170). This external source can be a non-instrumental or non-causal relation. One could even claim that instrumental value is a specific form of extrinsic value where an entity gets its value from an extrinsic source, and that extrinsic source is a causal means-end link to something of final value. Furthermore there are several nuances to instrumental values often wrongfully labeled as ‘mere means’, but which can include other extrinsic factors in addition to the necessary means-end relation (Baard, 2019)., But causal relations do not exhaust the set of entities having extrinsic value.

Thus, two objections to the Novel and indispensable thesis are that it relies on incomplete premises, and conflates extrinsic and instrumental values. These errors do not justify introducing RV.

3.1.2. Justifying RV Through Stipulating a New Axiological Pair

A similar strategy as the above, but at a more detailed level and which initially keeps extrinsic and instrumental values separate, is provided by Deplazes-Zemp (2023) who establishes the following axiological pairs to identify a gap for justifying RV:

• (1a) intrinsic value in the sense of final value

• (1b) instrumental value in the sense of means, and

• (2a) intrinsic value in the sense of value that supervenes on internal properties of the object

• (2b) instrumental value in the sense of extrinsic value

Problematically, however, she includes ’extrinsic value’ under the heading of instrumental value (Deplazes-Zemp, 2023, p. 6), which risk conflating them. Deplazes-Zemp writes that the ‘conclusive pairs’ (1) and (2) give rise to conceptual problems ‘since the respective broad category’ (1a) and (2b) are only ‘vaguely defined, roughly as a category that includes those values that are not part of the respective narrow category’ (1b) and (2a) (Deplazes-Zemp, 2023, p. 8). Consequently, to Deplazes-Zemp intrinsic value (1a) is vague and defined negatively as that which is of non-instrumental value, whereas that which has extrinsic value (2b) is vague and defined as that which is non-intrinsic. The narrow definitions (1b and 2a) allow ‘us to account for the particular features of final extrinsic value as a separate category’ (Deplazes-Zemp, 2023, p. 8), a claim that will be challenged in Section 3.2.2.

By fusing the clear and non-negative definitions (1b) and (2a), Deplazes-Zemp identifies a gap. RV is needed to fill the identified gap between intrinsic value as supervening on internal properties (2a), and instrumental values understood as means (1b) (Deplazes-Zemp, 2023, p. 8). Deplazes-Zemp identifies the following forms of third-category environmental values that fills the gap (2023: 8ff) between (1b) and (2a) (Deplazes-Zemp, 2023, p. 13):

1. Environmental values in the environmental policy and conservation social sciences literature

2. Eudaimonic intrinsic value

3. Constitutive value

4. Inherent value

5. Cultural value

6. Philosophical conceptions of RV

Justifying the introduction of RV, Deplazes-Zemp argues that ‘it may be objected that we do not need a third value category, if we could just work with the complementary value pairs that includes a broad and a narrow category’ (2023, p. 8), referring to pairs (1) and (2) above. But to emphasize, she motivates the need for RV by suggesting that ‘the respective broad category’ (1a) and (2b) are only ‘vaguely defined, roughly as a category that includes those values that are not part of the respective narrow category’ (1b) and (2a) (2023, p. 8), which she takes as justification for establishing a new pair, identifying a gap, and filling that gap with RV.

Thus, vagueness and negative definitions are the alleged motivations for excluding the components and setting up a new pair of narrow and positive definitions, because for a ‘substantive understanding of environmental values and a nuanced differentiation between value categories, it makes sense to work with the two narrow categories’ (Deplazes-Zemp, 2023, p. 8). Constructing the pair and identifying a gap between them then motivates RV to fill this gap. This is consistent with the Novel and indispensable thesis as Deplazes-Zemp (2023) argues from the insufficiency of existing axiological categories, correctly identifies two different conceptualizations of them, but then places a third value category between one element from each pair motivated by their clarity and not relying on negations.

There are several challenges with Deplazes-Zemp’s strategy based on excluding concepts due to their vagueness and negative definitions. Consider vagueness first. While conceptual clarity is a philosophical virtue it has its limits. When there are vague concepts one has the choice of stipulating vagueness-preserving and vagueness-resolving definitions (Hansson, 2006). In the latter case, ‘the definiens will be more precise than the definiendum’ (Hansson, 2006, pp. 23–24). Yet, the desired level of vagueness is connected to use. Vagueness or negative definitions are not necessarily reasons for rejecting a needed ethical concept. For example, a concept such as ‘intrinsic value’ is central to ethical reasoning (Baard, 2022; Batavia & Nelson, 2017). Admittedly, it is a vague concept that is difficult to find a single crisp positive definition of, but it is central to ethical reasoning despite its vagueness, in the same way as final value is needed to end an infinite regress of means. However, vagueness-resolving definitions are usually ‘needed for scientific or technical standard-setting or for legal purposes’ (Hansson, 2006, p. 25). In ethics this is evident in the many stipulated definitions and discussions on intrinsic and final values.

There are at least two more possible objections to Deplazes-Zemp’s strategy. First, one can object that the excluded components are not vague. There are no shortcomings of definitions of neither final (Sandler, 2012) nor extrinsic (Korsgaard, 1983) value. Second, one can state that even if vague, they are just as vague as is needed for them to be used, and the needed usage is evident from the discussion of them, especially in environmental ethics. In contrast, instrumental value (1b), which Deplazes-Zemp keeps, have long been neglected and lacked definitions and in-depth analysis (see however Baard, 2019). Rønnow-Rasmussen expresses surprise at this neglect in general moral philosophy given how instrumental value plays an important analytical role and is ‘commonly introduced to explain the very notion of intrinsic value’ (2022, p. 17). It becomes questionable whether Deplazes-Zemp is justified in including instrumental value for reasons that it is clearly and narrowly defined.

Moreover, if vagueness and negative definitions give reason to discredit a concept, which is what Deplazes-Zemp suggests, then it is questionable whether there are good reasons to preserve RV given its persistent vagueness and the way it is defined as not being intrinsic nor instrumental (James, 2022a). Even if granting Deplazes-Zemp (2023) that there is vagueness to existing axiological categories, adding another vague concept that is defined negatively hardly seems to be the right remedy. But it should be emphasized that vagueness need not give one reason to reject a needed concept. The desired level of vagueness is connected to use. But for conceptual parsimony it must then be showed that RV does something that other concepts cannot, which is questionable.

Deplazes-Zemp readiness to justify RV by utilizing two components from two different opposing pairs, create a gap between two components that are not mutually exclusive, and then fill that gap with RV, is a questionable strategy.

3.1.3. RV and the Relational Core of Morality

In environmental ethics Coeckelbergh has criticized the highly deductive approach of moral status ascription to environmental entities in environmental ethics as a reductive approach to living nature, ‘removed from life with its relations and change’ (2012, p. 205). To Muraca, RV evades both objectivism and subjectivism about value, as RV is not located in neither subject nor object but in their relation, justifying RV by drawing on Heidegger’s being-in-the-world and Whiteheadian process philosophy (Muraca, 2011, p. 382). Conventional categories cannot accommodate relational complexity, having led to a neglect of relations (Muraca, 2011).²

Similar criticism is found in moral philosophy more generally. Darwall suggests that any ethical account that fails to capture relational and intersubjective components is deficient and morality is fundamentally intersubjective and relational (Darwall, 2006; see also Wallace, 2019). Relations form a ‘moral nexus’ of obligations and claims (Wallace, 2019). On Darwall’s account the second-person is reason-giving. Darwall objects to theories of normative reasons according to which there ‘exist independent normative reason facts’ that agents should correctly recognize (2006, p. 292), making ethical reasoning like theoretical reasoning based on correctly tracking features of the world. Such theories seem to be analogous to those who argue for objective values or agent-neutral reasons of different kinds requiring correctly identifying and deducing features of value-holders. Instead, he argues for intersubjective foundations of normative reason for acting in accordance with publicly available principles that free and rational agents would accept (2006, pp. 292–293). Darwall briefly states that moral obligations can be owed to non-human entities through an authority of trustees such as the moral community or by imputing animals (and possibly the environment) with proto- or quasi-personality (2006, p. 29) giving rise to claims that are like those that moral agents have to other persons as a form of respect for their moral standing and care for their well-being.

The position provides a foundational defense of RV which we may call the Alternative overarching thesis:

Alternative overarching thesis: Ethics inherently concerns relations which is recognized by RV.

But the thesis can only take us so far. In addition to sharing the problems of the original Overarching thesis, RV fails to specify which relations that give rise to ethical reasons. This differs from basing morality on relations in an intersubjective sense. Darwall’s (2006) second-person foundation for morality is one wherein the relations that count ethically are those of free and rational individual’s mutual recognition of each other’s dignity. But it is difficult to make analogous environmental cases as the arguments rely extensively on intersubjective second-person recognition, and it is the relation to other persons that counts morally. One should perhaps respect some entities out of respect for the persons to whom they have importance. One can also make a promise to care for an environmental area, but the claimant of that promise will be another person (or, possibly but ontologically more difficult, an environmental quasi-personality [Darwall, 2006, p. 29]). That makes the moral standing of that area derivative of the respect for persons and makes the introduction of RV questionable.

Interestingly, Darwall suggests that while ‘a person’s good is intrinsically normative’, it is so ‘for anyone who cares for her, herself included’ (2002, p. 20). This provides a shift to relational care-based reasons. Darwall suggests that this has potential for environmental ethics, ‘if, for example, we can sensibly care about nonsentient biological species, or natural places, for their own sakes, then these will have a welfare’ (2002, p. 21). Even if lacking desires or preferences, things will be good or bad for them, even if not to them (Darwall, 2002, p. 21). Rotari objects that there are several reasons to be cautious about this conclusion, one being that Darwall does not account for purely instrumental senses of ‘good for’ (Rosati, 2009, p. 209ff).

A threshold must be set for what relations that matter when introducing relations to non-persons unless all relations matter equally. Setting such a threshold requires introducing conventional ethical concepts. To illustrate, Deplazes-Zemp and Chapman (2021) asks us to imagine a forest to which we have had a relation since we were children:

We admire its beauty and the diversity of living beings it engenders, and we cherish it because it makes us feel healthy and connected to our home. We clearly ascribe value to this forest, but what type of value is this. (Deplazes-Zemp & Chapman, 2021, p. 670)

Deplazes-Zemp and Chapman suggest that RV can ‘capture’ what type of value this is, which neither intrinsic nor instrumental value can (2021, p. 670), again referring to the false opposing pair discussed above. Moreover, it seems highly uncharitable to state that there are no other categories available to capture beauty and health.

Deplazes-Zemp and Chapman’s claim suggests an empirical observation that an agent values a forest. That is, an observation that valuing occurs and the forest is ‘good for’ the agent in a descriptive sense. It is not yet an ethical argument that provides us with normative reasons. There are several axiological options wherein the relation could justify reasons showing that conventional concepts can accommodate relations, but also that the observation need to be supplemented with established concepts. Agent-relative reasons lay close at hand, as does subjective final value ‘created by valuers through their evaluative attitudes, judgments, and preferences’ which does not ‘exist prior to or independent from’ the valuers (Sandler, 2012, p. 18). Subjective value also comes in the forms of either preference value or the more temporally robust and fundamental integral value (Sandler, 2012, p. 23ff). While preferences can be transitory and superficial, integral values are well-established components of persons worldviews and value systems. Such integral values are non-substitutable, as is RV. More concepts are possible. That the forest engenders beauty and diversity suggests that it is extrinsic or instrumental, either for the moral agent or the goods provided by diversity (see Baard, 2022 for a discussion). Moreover, the forest can have natural or intrinsic value which do central justificatory work in ethical reasoning. These categories result in reasons to respect the entities of that relation as a part of a respect for persons.

That there are several alternatives shows that RV does not necessarily follow from Deplazes-Zemp and Chapman’s example, reducing our confidence in the need for RV, unless RV is an empirical concept having descriptive or epistemological relevance (see however James, 2022a; Luque-Lora, 2023; Stålhammar & Thorén, 2019). But if it is such an observation, conventional normative concepts are still needed to make the relation reason-providing.

In summary, it is not clear how advocates of RV want to situate RV relative existing categories and greater analytical and conceptual caution is required before justifying RV based on rejecting existing concepts or conceptualizations of ethics given the central function that these strategies have for the introduction of RV.

3.2. Is RV a Justified Ethical Concept?

Here I will scrutinize whether RV is a justified ethical concept by not focusing solely on the shortcomings of intrinsic and instrumental values, instead assessing positive arguments for RV.

3.2.1. RV, Eudaimonia, and Constitutive Value

RV as an ethical concept is sometimes justified with reference to other ethical concepts. The concept of eudaimonia is an example. Knippenberg et al. (2018) argue that RV encompasses how relationships have constitutive value, meaning as ‘an integral part of a greater valuable whole that does have intrinsic value, for example a flourishing human life or partnership’ (Knippenberg et al., 2018, p. 41). They distinguish nature-based eudaimonia from eudaimonia in a more general sense. A ‘nature-inclusive eudaimonia’ is ‘a flourishing life in which nature is an integral part (constitutive value)’ (Knippenberg et al., 2018, p. 41). This has the potential of setting the threshold of which relations that matter. That is, not just any relation merit RV in a reason-implying sense, but only those related to constitutive value.

Knippenberg et al. suggest that while nature-inclusive virtues are not essential (2018, p. 43) to a good life they are good for all lives for three reasons: A relationship with nature is a relation with our most fundamental Other; relationships with nature entail growth, health, and healing ‘that no other relationship can provide’; encounters with nature can co-constitute lives (Knippenberg et al., 2018, p. 43). Thus, Knippenberg et al constrain RV to eudaimonia, but take efforts to single out the nature-related eudaimonia, making RV a subcategory of non-essential virtues.

Others also relate to eudaimonia. Consider Deplazes-Zemp and Chapman’s example in the preceding section (2021, p. 670). The reason that the forest example is RV is allegedly explained by the following implication:

If B (e.g. the forest) matters to [agent] A in this way, then A values B in virtue of a particular relationship that A has with B. (Deplazes-Zemp & Chapman, 2021, p. 677)

‘In this way’ refers to many different suggestions: local knowledge, a feeling of home and place for a change, admiration of beauty, or making us feel healthy (Deplazes-Zemp & Chapman, 2021, p. 670). But it is not clear whether it is the conjunction of all of them, or a disjunction of either of them that ‘In this way’ refers to, or whether it is up to the agent to establish how ‘In this way’ ought to be stipulated. If it is up to the agent, it is difficult to separate RV from individual preferences, integral values, or forms of subjectivism about values.

Deplazes-Zemp and Chapman suggest that RV is ‘often associated with eudaimonia’ (2021, p. 677). This would change the implication to something like:

if B matters to A by enabling A to lead a good life (in an eudaimonistic sense), then A values B in virtue of what a particular relationship gives A.

But this move reduces the need for RV since it is a meaningful statement in environmental virtue ethics, and the ethical relevance does not depend on the relation but is derivative on its impact on virtues, which does the ethical heavy lifting.

Generally, the introduction of RV is questionable if eudaimonia and constitutive value can fill the same function. Moreover, one could claim that eudaimonia and constitutive value can fill that function better than RV (James, 2022a, 2022b). First, it is still unclear what RV actually is. In contrast, eudaimonia is an established concept justified by a virtue ethical framework, also discussed in environmental ethics. Second, at least one difference can be noted between RV and constitutive value that speaks to the benefit of constitutive value. Constitutive value refers to claims to cultural identity and meaningfulness and are as such non-replaceable (James, 2022a, 2022b), similar to subjective integral value (Sandler, 2012). This leads to ethical and political reasons to recognize and consider some relations to environmental areas, places, or territories. But constitutive value is a much narrower concept than RV. For instance, continued access to territories for reindeer herding is essential, and non-substitutable, to Sámi culture and identity, but while my memories of forests nearby to where I grew up may be a part of my identity and upbringing, the cases are not the same as inter alia my relation does not refer to a group identity and collective claims. Even if there are individual constitutive values, such as when a place contributes to a person’s well-being, several conditions must be fulfilled for it to count as constitutive value such as the place having specific forms of meaning (James, 2022b, p. 62). Moreover, such constitutive values primarily lead to obligations of taking the value into consideration (James, 2022b, p. 65) while the ethical status of it relative other values must still be determined. Constitutive value sets the threshold of moral relevance for relations higher than RV, which make it a more demanding ethical concept, but also not as relevant in as many cases as proponents of RV take it to be.

3.2.2. RV as Extrinsic Final Value, and Subjectivism About Values

A recent attempt at justifying RV as an ethical concept consistent with the Novel and indispensable thesis is provided by Deplazes-Zemp (2023). After introducing the questionable pairs discussed above, she suggests (Deplazes-Zemp, 2023, p. 13) that RV is analogous to what Rønnow-Rasmussen has called extrinsic final value (Rabinowicz & Rønnow-Rasmussen, 2000; Rønnow-Rasmussen, 2011, 2022), being an axiological category of its own. Deplazes-Zemp’s analogy fails but reveals that RV is limited to the relation between valuer and valued entity, contrary to how ‘relational’ is discussed by Rønnow-Rasmussen. This makes it difficult to distinguish RV from a descriptive concept that valuing occurs and subjectivism about values.

Deplazes-Zemp’s analogy to extrinsic final value is surprising for two reasons. First, if there already is an axiological category with which RV is analogous, this reduces the need for introducing an additional concept.

Second, it is not a given that the analogy holds. Rabinowicz and Rønnow-Rasmussen (2000) argue that there are final values that need not be intrinsic. Rønnow-Rasmussen challenges the coextension of something being valuable for its own sake (having final value) and something being valuable for the properties it has (intrinsic value) (2011, p. 3; see also Rønnow-Rasmussen 2022, 2022). Rather, something can have final value that is not intrinsic. Deplazes-Zemp mentions Rønnow-Rasmussen’s discussion on a mink coat. A luxurious mink coat is a thing that is valued for its own sake, but also due to the role it plays in the owner’s life. Similarly, a rare stamp can have extrinsic value due to it being rare, being an externally relational feature of the stamp, but still have final value (Rønnow-Rasmussen, 2022, p. 20).

To Rønnow-Rasmussen the mink coat and rare stamp are examples of objects that have final value due to external relational features understood as referring to relations to other objects or events. Deplazes-Zemp, when describing Rønnow-Rasmussen’s position, writes that extrinsic final value refers to ‘values that are at the same time intrinsic and instrumental’ (2023, p. 8). This is not obvious from Rønnow-Rasmussen’s work who bases it on distinguishing intrinsic and extrinsic but again shows the conflation between instrumental and extrinsic values.

Entities having extrinsic final value are ‘valuable for their own sake in virtue of at least some relational property that is not internal to the value bearer’ (Rønnow-Rasmussen, 2011, p. 5). Rabinowicz and Rønnow-Rasmussen define relational property as “externally relational property” (i.e. a property that an object has in virtue of its relation to something that is not one of its parts)’ (2000, p. 35). While the valuing agent is not one of its parts, so are numerous other things, and from Rønnow-Rasmussen’s examples the agent seems not to be the primary component as he is not arguing for subjectivism about values but rather for different senses of ‘good for’ and moreover suggests that ‘X is good according to A or is what A esteems’ is a purely descriptive sense of ‘good for’ (Rønnow-Rasmussen 2022, p. 30).

To Deplazes-Zemp (2023) and Deplazes-Zemp and Chapman (2021) RV solely encompass the relation between valuer and valued entity, which is a very different way of understanding relations than Rønnow-Rasmussen’s account of final extrinsic value. Rabinowicz and Rønnow-Rasmussen highlight an argument from O’Neill that an entity may have value ‘for its own sake in virtue of its relational features’ such as uniqueness or rarity (2000, p. 40). Qualities such as, but not limited to, being rare or unique does not refer to the valuing agent, but rather the relation of the rare or unique entity to other entities that are not amongst its parts. That is, it is the relation of the valued entity to other entities that are potential value-bearers that are ‘not one of its parts’ (Rabinowicz & Rønnow-Rasmussen, 2000, p. 35), that give rise to the final extrinsic value, not primarily the relation to the valuer which is what Deplazes-Zemp discusses (Deplazes-Zemp & Chapman, 2021, p. 674). This limited way of interpreting ‘relational’ is common amongst supporters of RV (James, 2022b, p. 125), but is problematic. The strategy results in some form of subjectivism about values or descriptive account of valuing which is not what is allegedly argues in favor of nor what RV is intended to refer to if it is to have ‘normative force’ (Deplazes-Zemp, 2023).

Relevant to an environmental context, Rabinowicz and Rønnow-Rasmussen write:

a wilderness may be valuable […] because of its being untouched by humans, which may be seen as a relational property 2000, p. 41)

Another class of such entities include when an entity ‘is valued for its own sake in virtue of its special relationship to a particular object, event, or person’ (Rabinowicz & Rønnow-Rasmussen, 2000, p. 41). The value of a work of art can have the relational property of being an original, rather than a copy, and its relation to the artist (Rabinowicz & Rønnow-Rasmussen, 2000, p. 41). Other examples include ‘Napoleon’s hat, a gun that was used at Verdun, etc’, as in such cases ‘a thing acquires a non-instrumental value’, due to its ‘causal relation to some person, object or event that stands out in some way’ (Rabinowicz & Rønnow-Rasmussen, 2000, p. 41). These are externally relational properties in the sense of relating ‘to something that is not one of [the entities] parts’ (Rabinowicz & Rønnow-Rasmussen, 2000, p. 35), such as wilderness being valuable for the relational feature that it is untouched by humans, which are not amongst its parts.

The examples are curious as part of a justification for RV. The value of a piece of art due to its relation to the artist, or Napoleon’s hat, is hardly analogous to the examples discussed in RV research. The condition of a ‘relation to some person, object or event that stands out in some way’ (Rabinowicz & Rønnow-Rasmussen, 2000, p. 41, my emphasis) is telling, person here relating to historical or cultural person rather than a valuing moral agent. It is not just any relation that matter, but it refers to a person, object, or event that stands out in some way. Those reasons may be historical, or cultural, and again sets the threshold higher than a mere biographical account of one’s subjective relation to a forest. As suggested above, there are possibilities for biographical versions of ‘good for’, such as the case of the paper on which one’s son has written a poem, but it could be argued that these are valuable primarily due to inter alia the paper standing in that specific relation to one’s son, in that case.

Much can be said about this and in specifying the conditions of final extrinsic value, but that is beyond the scope of this paper. Suffice here to conclude that the analogy between RV and final extrinsic value does not hold. First, it is not primarily the relation between a valuing subject and valued object that is referred to by final extrinsic value. Rather, if conditions such as rarity generates value, it is because of the relation between a rare entity and other entities not amongst its parts. Second, final extrinsic value refers to the value of an entity due to its relation to special conditions, such as historically or culturally important persons or events, even if biographical accounts are also possible. This is reminiscent of constitutive value, which sets a higher threshold for which relations that counts ethically. Both these conditions defeat the analogy and support to RV. Lastly, even if the analogy with final extrinsic value would be successful, it would be questionable whether there is a need for RV.

4. Summary Note

I have formulated three different theses for how RV could be situated relative existing axiological concepts, justifying its introduction. The Overarching thesis is problematic. If limited to axiological categories, it renders RV a shorthand for ‘values’ but adds little, instead it risks combining conflicting values under the same heading. To the Novel and indispensable thesis RV is an axiological concept that does something that existing concepts cannot. Thus, making that thesis successful requires showing how RV is an alternative to existing concepts due to their shortcomings. One form of this thesis relies heavily on conflating extrinsic and instrumental values which too quickly leads to rejecting them as insufficient to account for relations. In another sense, RV is directed to the general failure of value concepts to account for relations, which are at the core of ethics, giving rise to an Alternative overarching thesis. While recognizing that the (intersubjective) relational core of morality has some merit, it still fails to account for which values that matter ethically in the environmental context without introducing additional and more conventional values, and is difficult to apply to environmental entities. But there are also defenses of RV that are not based on the shortcomings of existing axiological categories, such as by referring to eudaimonic values or extrinsic final value, but the analogies fall short. Even if the analogies were successful they would defeat the purpose of introducing RV.

RV seems to refer to agent-relative reasons based on final non-instrumental value of an environmental entity for an agent or group and the entity contributes to the well-being of the valuers and, if including conditions of final extrinsic values, that entity is valuable due to its relation to other entities that are not amongst its parts. But this can be described and ethically reasoned about using existing concepts.

Greater attention must be paid to an honest and charitable understanding of existing concepts, as well as clarity regarding how these are defined and what axiological functions they fill in practical reasoning if RV is to perform a normative function beyond descriptively stating that valuing occurs for an agent or having a rhetorical function.

Acknowledgments

I would like to thank Alejandra Mancilla, and the participants of the Practical Philosophy Group at the University of Oslo. The article is part of the ERC funded project ‘Dynamic Territories’ (grant agreement No 948964).

Disclosure statement

No potential conflict of interest was reported by the author(s).

Additional information

Funding

The work was supported by the HORIZON EUROPE European Research Council [Grant agreement no. 948964].

Notes

1. Deplazes-Zemp suggests that ’normative force’ is the ’persuasive power that these values have in guiding moral agents in their interactions with non-human nature’ (2023, p. 19, footnote 19). While ’persuasive power’ could be interpreted as RV primarily having rhetorical functions, I will understand it as being about normative justification.

2. It may be suggested that philosophers discussing RV belong to such different philosophical traditions that inter-translatability is questionable. The views here all belong to the admittedly wide umbrella of environmental ethics and make use of similar axiological concepts, but admittedly differ in approaches and fundamental premises.