1. Introduction
It is often said that some form of relativism must be accepted when it comes to value. Things are not good or bad absolutely but only relative to this or that person or, perhaps, group. Saying something along these lines, however, invites a challenge. Anyone who proposes to treat value as relative to particular agents (or groups), and to speak in terms of things being good or bad relative to this or that person (or group) will have to tell us how we are to understand such talk.
This challenge has been posed by, among others, Mark Schroeder. Schroeder notes that, unlike ‘good’ and ‘good for’, the expression ‘good relative to’ does not occur in everyday language. He goes on to ask whether we should therefore treat ‘good relative to’ as a purely theoretical notion, or if instead it is a notion of which we have some pre-theoretical grasp, one that we are able to have intuitions about.Footnote 1 If ‘good relative to’ is not a purely technical notion (on a par with ‘electron’ in physics), there should be some way for us to talk about what is good relative to whom in terms that are familiar to ordinary speakers of English. As Schroeder rightly points out, attempting to do this by understanding ‘good relative to’ in terms of ‘good for’ would be a mistake.Footnote 2 But Schroeder does not say precisely why this would be a mistake. I argue that this is so because whilst ‘good for’, like ‘good’, is a value notion – a notion that is used to make first order, evaluative claims – ‘good relative to’ is a value theoretical notion used to characterize a certain second order thesis about the relative nature of value. Unlike ‘good for’, ‘good relative to’ is not intended to capture anything in our first order, evaluative and normative thought and discourse, but instead a certain philosophical idea about value. But even though ‘good relative to’ is something of a term of art, this does not mean that it is a purely technical notion of which we have no pre-theoretical grasp. As I will try to show, it is possible to put good-relative-to talk in familiar terms using everyday language. The way to do this, I claim, is to understand relative goodness as goodness from a point of view or perspective. Schroeder acknowledges this possibility but claims not to understand what is meant when it is said that something is good from a point of view. I agree that talk of points of view is itself open to more than one interpretation, and I will distinguish the interpretation that I have in mind from some others. Being clearer about what we mean by ‘good relative to’, however, will not be enough since conceptual analysis will take us only so far. What is needed in addition to an analysis of that kind is a positive proposal about what determines or fixes the relevant perspective. Focusing on individual persons, I suggest one candidate for the role: what is good relative to a person or from that person's perspective or point of view is determined by that person's pro and con attitudes. Whilst this subjectivist view is not the only candidate open to value relativists I think it is what many such relativists have had in mind.
2. The Notion of ‘Good Relative’: A Couple of False Starts
As noted by Schroeder, one way in which we might make sense of good-relative-to talk in more familiar terms involves speaking about it in terms of different points of view or perspectives. This, I believe, is the correct way of understanding good-relative-to talk. But Schroeder claims not to understand what is meant when it is said that something is good from a point of view. He points out that talk of points of view is itself open to more than one interpretation.Footnote 3 Some of the interpretations that Schroeder considers and rejects involve the idea that evaluative claims can be true from some points of view whilst false from other points of view. In my attempt below to elucidate what it is for something to be good from a point of view, I do not rely on any relativism about truth. It is the nature of value and not the nature of evaluative truth that relativists of the kind I have in mind wish to understand in relative terms; it is value and not evaluative truth that is the object of the relativisation. It may, of course, often be the case that an evaluative claim appears true from one person's perspective whilst at the same time appearing false to another person, leading one of them to form the belief that the claim in question is true and the other to form the belief that the claim is false. But this is quite another matter from holding that the claim really is true from one perspective whilst simultaneously being false from another perspective. The kind of value relativist that I argue can make sense of goodness relative to perspectives is not committed to that conclusion.
Another interpretation that I want to guard against is one that equates ‘good relative to’ with ‘good according to’. What is good according to you is what you think or judge to be good, and even though there is disagreement over what precisely is involved in an evaluative judgment, we all have at least some idea of what it is to have such judgments. Whatever it is to be good from a perspective, it cannot be that. If it was, there would be no need to introduce the novel concept of ‘good relative to’ at all, since we already have ‘good according to’ which we all understand fairly well. What we are after is what it is for something to be good relative to a person and not what it is for a person to judge that something is good.
3. Is Relativism Inherently Subjectivist?
It may perhaps be tempting at this point to understand ‘good relative to’ in terms of attitudes such that what is good relative to you is analysed in terms of what you favour (and what is bad relative to you is what you disfavour). But here it is vital to distinguish between two different kinds of analysis corresponding to two different questions: ‘what does “good relative to” mean?’ And ‘what is it for something to be good relative to someone?’ Even though my own positive proposal about what it is for something to be good relative to a person – the nature of relative goodness – essentially involves the person's attitudes, this is not something that can be learnt from a conceptual analysis; it is not something that can be extracted from the meaning of ‘good relative’. The kind of value relativist that I have in mind is not committed to any specific semantic or linguistic analysis. In particular, she is not committed to holding that ‘X is good relative to P’ means ‘P favours X’. Realizing this will also help us to see that ‘good relative to’ is not an inherently subjectivistic notion. There is conceptual room for objectivist as well as subjectivist versions of value relativism. Eric Mack is one example of someone who explicitly defends an objectivist form of value relativism.Footnote 4 Even though I think that objectivist forms of value relativism are false, I don't think that such views can be ruled out on conceptual grounds. Analysing the meaning of ‘good relative to’ we will find, I believe, that it just means good from a perspective or point of view. This is as far as we get by conceptual analysis and about this objectivists and subjectivists (and relativists and non-relativists) can agree. To move beyond this we need a different kind of analysis; we need a positive proposal about what determines or fixes the relevant perspective. It is here that subjectivists and objectivists diverge. How precisely to demarcate subjectivist from objectivist theories about the nature of value is an issue that cannot be fully resolved here.Footnote 5
In what follows, whilst leaving conceptual room for objectivist forms of value relativism, I will take a subjectivist stance. Even though there is no necessary connection, there is a natural affinity between subjectivism and relativism. Even the objectivist relativist Mack concedes that ‘there are plausible links between subjectivism and agent-relativism. Subjectivism is a natural expression of the more general agent-relativist idea that what is valuable is valuable in and through its relation to agents'.Footnote 6 Toni Rønnow-Rasmussen writes that ‘“good relative to a” often stands for what is good from a's (subjective) perspective’.Footnote 7 And Schroeder writes that some philosophers ‘wrongly suspect that agent-relative value is a kind of subjective value’.Footnote 8 I believe that value subjectivists are in a better position than are value objectivists to make sense of relative goodness.
4. Relativism is About the Nature of Value, Not About What is Valuable
The perspectives that subjectivists have in mind when they speak of things as being good or bad relative to this or that perspective are the attitudinal points of view of individual subjects, as determined by their (actual, present) pro and con attitudes. On Stephen Darwall's interpretation of Thomas Hobbes, for instance, Hobbes holds that in desiring something ‘we ascribe to it a property, that of being good (something we ought to achieve), that it does not literally have’.Footnote 9 There is nothing relativistic about that. Darwall ascribes to Hobbes the view that all deliberation begins in the agent's desires, but not from premises about her desires. The agent reasons not from the premise that she has a certain desire, but from a premise that she accepts in having a desire. ‘[T]his premise is something normative – that something would be good, that she is to or ought to do something. The agent has these normative thoughts because she has desires. They are the “appearances” of her desires'.Footnote 10 Further, ‘Thought and discourse about good and evil encode an agent's view of things in deliberation, from the agent's perspective provided by her desires.'Footnote 11 Darwall's Hobbes is an example of a value relativist of precisely the kind that I believe can make sense of relative goodness. But the view requires further elaboration.
This is how I wish to understand this form of relativism: From the first-person point of view of an individual subject S who favours (has a pro attitude towards) an object O, O is good simpliciter, or perhaps good for someone, not good-relative-to-S. Unlike ‘good’ and ‘good for’, ‘good relative to’ has no application on this level. It is only when we move to the third-person point of view that we will come to recognize – so the value theoretical relativist holds – that all values are by their nature relative to particular points of view. It is thus the nature of value that is claimed to be essentially relative, and it is not being claimed that there necessarily must be anything relative in the contents of any of our evaluations. Relativism is a second order, value theoretical view about the nature of value, not a first order, evaluative or normative view. Value theoretical relativism, then, does not force evaluating subjects to value anything in either explicitly or implicitly relative terms.
Insofar as the agent ascribes a value-property to the object of her desire (which need not be the case, it is enough that she has the desire in order to give rise to a relative value) it will be the property of being good simpliciter, or the property of being good for someone, not the property of being good-relative-to-the-agent, or some such thing. This sort of value relativism does not imply that the meaning of ‘X is good’ is ‘X is good relative to me [the speaker]’, or anything along such lines. The conclusion often ascribed to relativist views, that disagreements in evaluative matters are impossible because evaluators that apparently disagree are in fact systematically talking past each other, is avoided by this sort of relativism.
Consider in this context an objection posed by Jan Österberg:
A person who subscribes to [subjectivism] will sometimes conclude that one and the same object both is, all things considered, [finally] good and also, all things considered, [finally] bad. This absurd conclusion will be reached whenever one person [finally] desires a state-of-affairs and another person [finally] desires its non-occurrence.Footnote 12
The subjectivist relativist will respond to the objection in the following way. Whenever one person finally desires a state of affairs and another person finally desires its non-occurrence, the state of affairs in question is finally good relative to the one person and finally bad relative to the other person. The alleged contradiction evaporates since from no person's perspective is one and the same object simultaneously all-things-considered finally good and all-things-considered finally bad. If, from any particular individual's first-person perspective, an object, O, is all-things-considered finally good, it is never the case that O is at the same time and from that same perspective all-things-considered finally bad. The absurd conclusion is thus avoided. Krister Bykvist also recognizes that relativism ‘does not necessarily imply the absurdity that p is better than q, and q is better than p. For arguably, agent-relative theories claim that the relevant “better than”-evaluation is not the two-place “p is better than q” but the three-place “p is better than q relative to agent A”’.Footnote 13 Note, though, that on the relativist view as I understand it ‘p is better than q relative to agent A’ is not itself an evaluation. Valuing is essentially tied to the first-person points of view of individual subjects whilst good-relative-to claims belong in the third-person perspective. The evaluations, as made from within the first-person perspectives of A and B are simply ‘p is better than q’ and ‘q is better than p’ respectively. There need be nothing relative whatsoever in the contents of A's and B's respective evaluations.
5. Relative Is Not Relational
As David Gauthier recognizes, ‘on both … absolute and … relative conceptions [of value] it is possible to distinguish what is good for some person from what is straightforwardly good’.Footnote 14 Rønnow-Rasmussen similarly observes ‘that something is good relative to a leaves open whether the goodness in question is good, period; or good-for; or some other kind of goodness'.Footnote 15 The idea that things can be good only relative to the first-person perspectives of individual subjects – which I outlined above – should thus not be confused with the fundamentally different idea that all goodness is goodness-for, or as in any way standing in conflict with the idea that things can be straightforwardly good or good simpliciter. Connie S. Rosati seems to conflate goodness-from-a-perspective with goodness-for when she asks,
After all, what could be more obviously true than that things can be good or bad, not merely in an absolute sense, but for us? What could be clearer than that our lives can go better or worse, not just, to borrow Henry Sidgwick's memorable phrase, from the point of view of the universe, but from our point of view?Footnote 16
Later on in the same paper, however, Rosati recognizes a distinction between relative and relational. She says that the notion of good-for ‘does not involve a relativisation of goodness or normativity. Good for P is not goodness relative to P but a distinct relational value’.Footnote 17 As I have insisted above, the idea that things are good or bad only from the subject's first-person point of view need not form any part of the contents of her value judgments. ‘Good relative to P’ does not capture anything in our first order, evaluative thought and discourse, but instead a certain second order idea about the relative nature of value. ‘Good for’, by contrast, is meant to capture something in our first order, evaluative thought and discourse. Hence, what is good from someone's point of view or relative to a person's perspective is fundamentally different from what is good for a person.
Goodness-for is an example of a relational value, which should be contrasted with straightforward goodness, or goodness simpliciter, which is a non-relational value. Recognizing both relational and non-relational values is something that both relativists and non-relativists can do. One might perhaps be forgiven for thinking that the very idea that things can be good simpliciter would somehow be inimical to relativism, but a defender of a relativist conception of the nature of value along the lines outlined above is not, I argue, disqualified from analysing what it is for something to be straightforwardly, or non-relationally good. Holding that all values are relative to the perspectives of individual subjects – that things are good only from particular points of view – does nothing to threaten the idea that things can be good simpliciter relative to a particular perspective. The thesis that the nature of value is relative should thus not be confused with the fundamentally different thesis that all values are relational, and neither does a commitment to the former thesis imply a commitment to the latter. Indeed, as we will see below, relativists of the kind I have in mind will most naturally reject the thesis that all values are relational.
6. Goodness-Type Monism and Goodness-Type Dualism
According to a classification proposed by Rønnow-Rasmussen, a distinction can be drawn between goodness-type monism and goodness-type dualism.Footnote 18 Goodness-type monism comes in two varieties. Moorean monists claim that all values are non-relational; all goodness is straightforward goodness, or can be fully understood in terms of, or reduced to, straightforward goodness. Hobbesian monists claim that all values are relational; all goodness is goodness-for, or can be reduced to goodness-for. Goodness-type dualists recognize both relational and non-relational values and claim that neither of these types of value can be fully understood in terms of the other. The divide between these forms of monism and dualism is independent of the debate over relativism. Dualism and both forms of monism are (at least in principle) open to relativists and non-relativists alike.
G. E. Moore held that all goodness is fundamentally of the straightforward type. He expressed his inability to even understand how it could be otherwise.
In what sense can a thing be good for me? It is obvious, if we reflect, that the only thing which can belong to me, which can be mine, is something which is good … When therefore, I talk of anything I get as ‘my own good’, I must mean either that the thing I get is good, or that my possessing it is good.Footnote 19
Moore argued that the only ways in which we can possibly understand what it is for something to be good for a person is in terms of (1) it being simply or straightforwardly good that this person has or gets this thing, or (2) that the thing that the person has or gets is straightforwardly good (whether or not it is good that she has or gets it), or (3) what this person believes to be good in one of the first two senses (what I have called ‘good according to’ above).
Whilst Moore didn't explicitly operate with a distinction between on the one hand relational/non-relational, and on the other hand relative/non-relative, it should be clear that what he was attacking in the quoted passage is the idea of relational value (specifically of goodness-for) and not the idea that things are good relative to perspectives. Moore was hostile also to the latter idea and he tied his scepticism of the notion of good-for to his objectivism and absolutism when he said ‘“My own good” only denotes some event affecting me, which is good absolutely and objectively’.Footnote 20 But irrespective of Moore's actual views, Moorean monists as I am thinking of them here advance a view according to which all values are non-relational – or at least that all values can be understood in terms of, or be reduced to, non-relational values, like straightforward goodness – but is silent with regard to the debate about relativism (and subjectivism). The Moorean monist position, as I understand it here, can thus be accepted by relativists as well as by non-relativists (and subjectivists as well as objectivists).
The name ‘Hobbesian monism’ was inspired by Thomas Hobbes' insistence that ‘one cannot speak of something as being simply good; since whatsoever is good, is good for someone or other’.Footnote 21 However, it is not obvious that Hobbes himself was a Hobbesian monist. He might well have been a goodness-type dualist or even a Moorean monist! On a different possible reading of Hobbes, what he is expressing here is not the idea that all values are relational – that all goodness is goodness-for – but rather the idea that all values are relative to individual subjects. Saying the latter is fully compatible with any of the three positions on the relationality or non-relationality of value that I have distinguished above (goodness type dualism, Moorean monism, and Hobbesian monism). Holding that all values are relative to subjects' perspectives is compatible with saying that from a particular subject's perspective there can be both relational and non-relational values, or that there are only non-relational values, or that there are only relational values. On this reading, Hobbes was giving voice to a form of relativism rather than to what Rønnow-Rasmussen calls Hobbesian monism. Perhaps what Hobbes was trying to say was not that all goodness is relational but rather that nothing can be good from a third-person point of view, but only from the various first-person points of view of individual subjects. In the same paragraph as the quoted passage, Hobbes also says that ‘since different men desire and shun different things, there must … be many things that are good to some and evil to others' and good is said to be ‘relative to person, place, and time’.Footnote 22 My aim here is not exegetic, and I do not wish to anachronistically ascribe to Hobbes the modern distinctions with which I operate. But I do wish to insist on keeping the two views separate. That is, Hobbesian monism (which may or may not have been accepted by Hobbes himself) should not be confused with relativism (which, I believe, can be more safely ascribed to Hobbes). Whilst relativism is compatible with Hobbesian monism, it is equally compatible with Moorean monism or with goodness-type dualism.
A potentially better example of a Hobbesian monist than Hobbes himself is Christine Korsgaard. She has recently defended the thesis that all values are relational, or, as she puts it, ‘the essentially relational nature of the good’:
I think that the notion of ‘good-for’ is the prior notion [prior to the notion of ‘good’]. Or, to put it a better way, I think there is something essentially relational about the notion of the good itself. … I think there is such a thing as the good, only because there are creatures for whom things can be good; that is, creatures who can welcome or reject the things that they experience. In fact, I think that the idea of something's being good without its being good-for someone should be rejected as unintelligible.Footnote 23
Perhaps we should distinguish two different but related ideas being expressed here. The first is that the notion of ‘good for’ is prior to the notion of ‘good’, that goodness itself is essentially relational. It is this that makes Korsgaard a Hobbesian monist. The following two claims testify to her allegiance to this view: ‘[G]ood in the final sense is a relational notion – a form of good-for’ and ‘All final goodness is essentially goodness-for the being whose final good it is'.Footnote 24
The second idea being expressed in the block quote from Korsgaard is that things can be good only by being good for someone; nothing can be good without being good-for. Subjectivist relativism, on the other hand, I believe, naturally leads to the arguably common sense view that lots of different things are good that are not good for anyone in particular. This is so because we tend to favour some things for their own sakes even when we don't favour them for someone's sake. The common sense view is compatible with relativism – the view that things are good (or good for) only relative to points of view.
On Korsgaard's view, there must always be someone involved in the object of value, some person or animal for whom things can be good or bad, someone who can be made better or worse off in some way. This seems to set up substantial, first order restrictions on what can be valued. The relativist, by contrast, holds the fundamentally different thesis that values are always relative to the first-person points of view of individual subjects. On this picture, there need not be anyone involved in the object of value for whom it is good, and as such this raises no substantial restrictions on what can be valued.
A final comment on Korsgaard: She says that ‘the final good is something essentially relational, because it exists in relation to the consciousness, or the point of view, of a conscious being’.Footnote 25 Here I believe that she conflates the distinction between, on the one hand, the claim that all values are relational, that all goodness is goodness-for (Hobbesian goodness-type monism), and, on the other hand, the claim that value is relative to first-person points of view, that things are good and bad only from the perspectives of subjects. Value relativists hold the latter, but they need not (and probably do not) accept the former; by no means does the former follow from the latter. That an object, O, is valuable relative to a subject, S, does not imply that O is good for S, or that it is good for any other person or conscious being. Acknowledging that O is valuable relative to S leaves it open whether the value is relational or non-relational. How relativists can analyse relational value is a discussion for a separate occasion.
In response to Korsgaard, the relativist may reasonably insist that it is not the idea of something being good without being good for anyone that is unintelligible, but rather the idea that things are good without being good relative to subjects' points of view. Relational values such as goodness-for are no less mysterious than non-relational values such as goodness simpliciter as long as they are held to be absolute. It is not that things can be good without being good for anyone that is problematic but instead the idea that there are values (whether relational or non-relational) that are not relative to the perspectives of subjects.
I suspect that the initial plausibility of Korsgaard's claim that the idea of things being good without being good for someone should be rejected as unintelligible is entirely parasitic on the overwhelming plausibility of the more general demand that value must on some level have something to do with conscious beings. And, of course, relativists satisfy this general demand as value is essentially tied to perspectives of individual subjects. Once it is understood that acknowledging that things can be good without being good for anyone in particular does not force us to deny that value is connected to conscious beings, the initial attraction of Korsgaard's claim disappears.
7. Conclusion
I have argued that good relative to a person should be understood as good from the perspective or the point of view of the person. But this analysis of the notion ‘good relative to’ leaves open questions about the full nature of relative goodness; about what it is for something to be good relative to a person's point of view. For this, a different sort of analysis in needed. I proposed such an analysis on which the relevant perspective is determined and fixed by the pro and con attitudes of the person in question. Goodness relative to should not be confused with goodness for as the latter is a relational value notion – a notion used to make first order, evaluative claims – and the former is a value theoretical notion, used to characterize a certain second order thesis about the relative nature of value. The proposed relativist view about the nature of value is compatible with all of three different views about relational goodness; that all goodness is relational, that all goodness is non-relational, and that there are both types of goodness.
