From whence do moral claims derive their force? Stephen Finlay (2014) has an answer: Moral claims are elliptical and deflatable to end-relational claims, meaning that they provide information concerning the probability and necessity between some world-states (concerning means) and some other pre-reflexively desired world-states (our ends). In this thesis, I argue that Finlay’s end-relational theory provides a plausible, if radical solution, to our metaethical ‘wish list’. I will present what I see as the Finlayian view’s three central premises: analytic reductionism, Humean psychology, and relational morality, and consider objections to each of them, concluding that while the view may adapt in response to the challenges, it may end up a different view, and perhaps strain our reflective understanding of our moral practices too far to accept by the time we are done.
Firstly, I sketch two affective pictures by which we should consider our investigation. In the second and third sections, I establish David O. Brink’s problem of our metaethical ‘wish-list’, and argue that Finlay’s end-relational view is an attractive solution. In the fourth, fifth and sixth sections I focus on challenges and problems to the three central premises of Finlay’s view.
Introduction
From whence do moral claims derive their force? Stephen Finlay (2014) has an answer: Moral claims are elliptical and deflatable to end-relational claims, meaning that they provide information concerning the probability and necessity between some world-states (concerning means) and some other pre-reflexively desired world-states (our ends). In this thesis, I argue that Finlay’s end-relational theory provides a plausible, if radical solution, to our metaethical ‘wish list’. I will present what I see as the Finlayian view’s three central premises: analytic reductionism, Humean psychology, and relational morality, and consider objections to each of them, concluding that while the view may adapt in response to the challenges, it may end up a different view, and perhaps strain our reflective understanding of our moral practices too far to accept by the time we are done.
This thesis is structured in six sections. In the first section, I sketch two affective pictures by which we should consider our investigation. In the second and third sections, I establish David O. Brink’s problem of our metaethical ‘wish-list’, and argue that Finlay’s end-relational view is an attractive solution. In the fourth, fifth and sixth sections I focus on challenges and problems the three central premises of Finlay’s view.
1. Two Affective Scenarios
How does it feel, to have moral claims leveraged against us, or to leverage moral claims? Two affective pictures are put forward to the reader, which will be referenced throughout the thesis.
Scenario Number 1: The Dropper
It is late at night. The wind howls. You come across a masked figure, dangling something over a bridge. With horror, you realise that what the masked figure is dangling, teetering, about to drop, is That Which Absolutely Should Not Be Dropped. With even more horror, noticing the large and sequinned ’D’ emblazoned across their back, you realise that you have encountered The Dropper, a local supervillain that has been terrorising your town by dropping That Which Absolutely Should Not Be Dropped off of the local bridges. “Stop!” you yell. “You ought to not do that! It’s bad!”
“Oh. Hello there,” cordially remarks The Dropper, still dangling the object. “Can tell me why I shouldn’t? I’m perfectly open to considerations.”
You contemplate what to say or do.
Scenario Number 2: Neighbourly Ned and Nelly
The work day is finally over. It is a swelteringly hot day. You arrive home, however, with budding joy. The time has come for That Which Makes it All Worthwhile. Just as you begin, you are interrupted by a smart rap on the door. As you open it, you are greeted by the sight of your neighbours, Ned and Nelly, both smiling fixedly in the heat. “Ah, good,” Ned says.
“Oh. Hello, th—” you begin, but Nelly jumps in.
“You must stop immediately! You ought to not do it. It’s bad. Very bad. You should stop. In fact, you must.” She finishes. Ned continues smiling determinedly behind her, with an air of projected consternation. You are quite certain, that while your neighbours mean well, it is them who is wrong, but perhaps they know something you don’t.“Can you tell me why I shouldn’t?” you ask. “I’m perfectly open to considerations.”
You see them contemplate what to say or do.
2. David O. Brink’s Metaethical Wish List
We have a wish list of properties or features that we think distinguishes morality. However, it appears that not all of them can simultaneously be true at once. In this section, I set up Brink’s Puzzle of The Rational Authority of Morals (1992), to which I will go on to argue Finlay’s end-relational theory is an attractive answer.
There are four premises, Brink (1992, 1) puts to us, that we might intuitively think are true about morality. Call this proposal, that we classically think of morals as composing of these four premises Classic Morals.
Classic Morals entails:
- Moral Universalism (MU) - That moral requirements apply to all agents, independently of their ends.
- Motivational Internalism (MI) - That moral requirements provide agents with motivating reasons for action
- Agent Particular Reasons (AGP) - That motivating reasons for action are dependent on an agent’s particular ends.
- Possibly Divergent Ends (PDE) - That it is not the case that all agent’s ends necessarily converge.
Moral Universalism refers to the intuitive premise that moral claims are inescapable by simply declaring oneself to not care about them, or that moral claims are limited to a certain type of person or agent. Motivation Internalism refers to the idea that a key element in moral claims is that it provides an actual motivationally sound reason to act in a particular way, allowing us to recommend, prescribe, or prohibit certain types of behaviour. Agent-particular reasons refers to the assumption that for something to be able to motivate an agent in a particular way, it must ‘plug in’ and must affect an agent’s interests in some way. Lastly, Possibly Divergent Ends refers to the fact that it is intuitively the case that different agents care about different things, and to different degrees. Whilst in many cases agents may be aligned in our ends, or preferred states of the world, it appears not to be a necessary fact that between any two agents, their ends are necessarily aligned.
The issue is, however, that all four of these premises cannot be true at the same time. Indeed, if Moral Universalism is true, then moral claims apply to all agents. If Motivational Internalism is also true, then these moral claims provide all agents with motivating reasons for action. If, simultaneously Agent Particular Reasons is also true, that motivating reasons must stem from ends, this therefore means their ends are aligned, entailing Possibly Divergent Ends is false. A similar story may be told with any collection of the three premises - the fourth must always be jettisoned.
The choice of which premise to deny, then, Brink argues, yields four different ‘baskets’ of metaethical solutions.
A. (~MU, MI, AGP, PDE) Relativistic or Relational Solutions - Deny Moral Universalism
Solutions in this basket all involve a version of denying moral universalism, resulting in a form of relativism, in which it is not the case that all moral claims apply equally to all agents, in all conditions. It is difficult, perhaps, to even consider morals as applying differently to different people, or to certain acts at some times and some at others, so engrained in us is the idea of moral universalism. Solutions in this basket, therefore, will have to come up with some story, that allows for negotiation explains our moral intuitions.
B. (MU, ~MI, AGP, PDE) Externalist Solutions - Deny Motivational Internalism
Externalist solutions involve denying that moral claims in themselves are explicitly capable of motivating action. In solutions in this category, motivation and morality come apart. As such, it is feasible to acknowledge something as morally right, without this generating any action-motivating reasons. Equally, it is possible for someone, although they have been acting entirely rationally according to their motivations and interests, and have no reason to act otherwise, to be labelled immoral. This is perhaps a difficult pill to swallow, and the theorist advocating for this solution will have to establish a sound definition of what it means for something to be moral or immoral, to account for its classical relation with motivation.
C. (MU, MI, ~AGP, PDE) Agent-Neutral Reasons Solutions - Deny Agent Particular Reasons
Solutions that deny agent-particular reasons must instead posit agent-neutral reasons for action; reasons that motivate action, despite not connecting in any way to the furthering of an agent’s particular interests or ends. Russ Shafer-Landau (2003, 207) puts forth, for example, that regardless of an agent’s interest or desire, their ability to prevent someone from experiencing excruciating pain provides them with a reason to do so. A peremptory challenge to this, however, is that one might think that an agent has a reason only in virtue of their perceiving the world in which they do help as being constitutively better than a world in which they did not. As such, it seems like their desire for a particular world-state may have played a role in their motivation. In other words, it seems like the proponent of agent-neutral reasons must posit reasons which motivate an agent to perform an action even when that agent evaluates the world-state in which they do perform the action and the world-state in which they do not as equally desirable, in other words, that their action elicits no discernible effect on a criterion they value, and this, perhaps, appears difficult
D. (MU, MI, AGP, ~PDE) Metaphysical Egoist Solutions - Deny Possibly Divergent Ends
To deny Possibly Divergent Ends, solutions in this basket must instead insist upon the necessary intertwining of agents’ interests. One version of such a solution is to deny the separation of individual agent’s interests, by arguing from something like a Platonic Good, that all agents are ultimately striving for or would benefit from. As such, any moral claims are relative to ‘our (singular) interest’, The Good. The second strategy, which Brink’s own preferred solution falls under (1992, 18), results from attempting to dissolve the psychological separation between agents. All interests, therefore become ‘our (shared) interests’. This is done by arguing for an arguing for an Aristotelian conception of other-selves, in which one treats oneself well by treating others well. One might worry, however, that on the first method, finding such a universally uniting end is difficult if not impossible, and secondly, on the second method, that such extension of psychological continuity to all people is a) supremely unintuitive and b) runs into issues when agents have contradictory, or oppositional ends.
As such, given the four premises were each initially agreed to have intuitive force, each ‘solutions-basket’ must endeavour to account for or ‘explain away’ the premise that it denies. In light of the difficulties regarding externalist notions of motivation, agent-neutral reasons, and metaphysical egoist solutions, in what follows, I will argue that Finlay’s end-relational view, falling under the relativistic solutions basket, provides an attractive solution to the problem.
3. Finlay’s End-Relational Solution
Finlay, in his analysis, aims to provide a balanced approach to the objectivity and practicality of moral claims (2014, 2). By objectivity Finlay refers to the fact that moral claims appear to obtain independently to our interests; morality appears in some way to be involved in the realm of facts. By practicality, Finlay refers to the ability moral claims appear to have to express our motivations and desires, for us to do things with moral claims (2014, 4). Objectivity, he argues, is represented most strongly by metaethical primitivists (his preferred label for metaethical non-naturalists), who hold that basic moral properties such as goodness are mind-independent, ‘primitive’ and ‘irreducible’ to other concepts. A primitivist argues that claims such as “X is morally good” are claiming that it is a primitive fact that “X is good” that obtains independently of any perspective, and that it is irreducible to any other concept. Practicality, on the other hand, Finlay argues, is most strongly exemplified by metaethical expressivists, who hold that moral claims are not even facts at all, but rather expressions of an agent’s desires, or motivations. An expressivist argues that claims such as “X is morally good” express a favourable disposition by the speaker towards X, such that they are motivated towards it or desire it.
The problem with the primitivist picture, is it is unclear how normative facts become known to us, and are capable of motivating us. Given their disconnection with the natural world, it appears primitivists are beholden to a “spooky pricking of their thumbs” to account for this (Finlay, 2020, 137). On the other hand, the problem with the expressivist picture is that disagreement and negotiation break down into brute assertions of preferences. The Finlayian picture, then, in attempting to meet both of these problems, seeks to offer a conception of morality centred on facts that are expressed by agents by virtue of their desires and are capable of motivating other agents by providing information in relation to their desires.
With this framework, Finlay’s conception follows with very little surprise. On his picture, moral claims are relational between certain facts and ends (desired world-states). But what facts, and what ends? On the Finlayian picture, moral claims are elliptical for facts concerning the probability- relation of one world-state, to another desired world-state, or end. Ends, under the Humean psychology that Finlay employs, are pre-reflexively desired world-states; states of affairs that we desire to make actual (2014, 2). Ends are hierarchical and come in two forms: final ends, which are irreducible and basic, and instrumental ends, which are in the service of other, more final ends. On the end-relational view, when one proclaims “X is morally good” they are claiming “X counterfactually raises the probability of salient end E, compared to other salient options_”,_ where both the salient end and salient option set of actions are pragmatically provided via the context of the utterance (2014, 39).
Why ‘relational’ and not ‘relativistic’? Finlay wishes to highlight that normative terms are incomplete relational predicates, akin to “old, tall, fast” and “cold” (2014, 21). A temperature may be cold for a particular climate, but the same temperature warm for a different one. Rather, describing something as ‘cold’ or employing any relational predicates relies on context for the salient relation. Equally, Finlay seeks to extend such an analysis to moral terms, such that certain salient context clues, namely the desired ends of either speaker or hearer, are referred to elliptically.
Armed with the concept of elliptical reference to end-relational probability facts, Finlay extends his descriptivist semantic analysis to further normative terms such as “ought”, “must”, and “reasons”. Building off of the notion that one ‘ought to do what is best’, and that ‘best’ is the superlative form of good; that one ‘ought to X’ is elliptical on the Finlayian picture to: ‘X, of the set of salient possibilities, raises the probability of the salient end E the most’ (2014, 73). ‘Must’ the stronger term, by comparison, is analysed in terms of modal necessity, the claim ‘One must X’ is elliptical for the claim that ‘In all worlds in which salient end E obtains, X must also obtain’. Finally, ‘reasons’, Finlay argues, are to be understood as “explanations why” (2014, 90). Normative reasons, therefore, are to be understood as why some X is good, or why an agent ought to X, which on both of Finlay’s previous readings, results in an explanation why “X raises the probability of the salient end E” occurring or why “X raises the probability of the salient end E occurring the most”.
So far, Finlay has proposed analytic reductions of normative terms down to descriptive probability relations regarding ends. It seems like the objectivity challenge has been met. But how does Finlay account for the practicality, or motivational and expressivist side of moral claims? He argues, via appealing to pragmatics, that agents, when they speak, are endeavouring in virtue of their Humean psychology to accomplish their conversational ends as best as possible (2014, 122). Consequently, on the Finlayian picture, moral claims, as factual probability claims, are made from a motivated perspective of contingent Humean ends, in order to motivate a hearer from their contingent Humean ends (2014, 129). As such, Finlay’s conception is both quasi-expressivist, in that moral claims are used to express preferences or contingently desired ends, and quasi-realist, in that they concern actual, mind-independent probability facts.
With the specificities of his theory so defined, Finlay then argues that the end-relational view is capable of accounting for the problems that usually stem from solutions in the relativist basket, such as ‘all things considered’ judgments, categorical claims, and disagreement. All-things-considered judgements about what one ‘ought’ to do, on the Finlayian picture (2014, 154), are the result of decision-theory style weighing of the payoffs in accordance with multiple ends. Furthermore, while categorical claims, which we might think of intuitively being explicitly _non-_relational to any ends, are instead on the Finlayian picture a particular speech act, that of moralism (2014, 187). The speaker of a categorical claim elliptically leaves out the relation to the salient end, thereby expressing their own bare preference, exerting psychological force on the hearer by pragmatically signalling that the salient end is so obvious and ubiquitous that it need not even be referenced, thereby prescribing it be made so (Finlay, 2014, 187). A final worry for solutions in the relational basket is that they make disagreement impossible, due to seemingly opposing moral claims merely being relational to different ends. Here, Finlay appeals to the quasi-realist and quasi-expressivist elements of his theory, in order to articulate four avenues of available disagreement, in either preference or belief (2014, 270). On the quasi-expressivist side, agents may disagree outrightly, about their preference as to what ought to be done in a particular situation or may disagree fundamentally, holding ultimately inconsistent final ends. On the quasi-realist side, agents may disagree instrumentally, regarding the actual probability effect of a particular action upon a salient end, or may disagree rationally, about what probability-related facts it is rational to believe given a certain set of available information.
Let us return to our affective scenarios, and illustrate this via reference to The Dropper (and thereby implicitly Neighbourly Ned and Nelly). On the end-relational theory, when you declare to The Dropper that they “morally ought to stop” you are declaring that ‘[Their stopping] counterfactually most raises the probability of salient end E obtaining’. Already, via your expressed preference of action in this situation, you disagree outrightly. Which end is salient in this scenario? Presumably, at the moment of your speaking it is obscure, thereby causing The Dropper to ask if you can provide a reason. As a reason, on the end-relational view, is an explanation why the probability of the end is affected by the action, presumably when supplied this will make the end explicit. If you deliberately left the end implicit for pragmatic psychological force by indicating your assumption of its ubiquity, you may have been engaging in moralism, the end-relational explanation of a categorical claim. If, through your providing of reasons, and subsequent discussion with The Dropper, you come to find out that your irreducible final ends are different, then you disagree fundamentally. Alternatively, if you find an end that you are both aligned to, but disagree with The Dropper’s methods (i.e. dropping) for accomplishing those ends, perhaps because you know something they do not, then you disagree instrumentally. Finally, if you disagree with the appropriateness of the Dropper’s action, even if you were in their situation, with their ends, and their information, then you disagree rationally, and may thereby think they are getting the ‘all-things-considered’ balance of their own ends, incorrect. With various avenues of disagreement available, and with an account of categoricity, and all-things-considered judgement, Finlay’s end-relational theory, therefore, accommodates many of the challenges typically aimed towards solutions in the relativist basket. This serves to defuse some of our intuitive worries in light of denying Moral Universalism and thereby increases the theory’s attractiveness in light of the prima facie difficulties of externalist, agent-neutral, and metaphysical egoist solutions.
In each of the following three sections, I will take aim at one of the three central premises of Finlay’s end-relational account: I) Analytic Reductionism - That moral claims may be reduced to non-normative propositional content. II) Humean Psychology - That agents necessarily attempt to optimally fulfil their end. III) Relational morality - That morality may be understood as relational, instead of simpliciter. I will argue that: I) Finlay’s account, in instances of moralism, relies on a mysterious ‘psychological pressure’ that somehow comes from the expression of ends that we do not share, relying on a ‘spooky pricking of thumbs’ that he claims his theory avoids. II) That his reliance on Humean psychology deflates moral claims down to an automata-esque information exchange III) That his account is far more revisionary in regards to concepts such as blame, punishment and evil than we might initially think, and while technically neutral on the possibility of making first-order moral claims from any system, is most certainly not neutral on their force.
My contention is that the end-relational view may survive as a plausible explanation and useful metaethical theory, but in the lengths it must go to adapt to the challenges, becomes quite the revisionary tale, that may perhaps strain our reflective understanding.
4 . Analytic Reductionism
In this section, I survey two challenges to Finlay’s premise of analytic reductionism - that our moral concepts may be reduced to other concepts, and that our moral claims may elliptically be referring to these other concepts. Firstly, I bring up the challenge brought up by Richard Joyce (2011, 527), what I call the Problem of Explicitation; If moral claims are elliptical for implicit end-relational claims, then they should identically function if those end-relational claims are made explicitly. However, given that they do not function identically if these end-relational claims are made explicit, this therefore suggests that moral claims may not be elliptical for these claims. Finlay’s rejoinder is that if ellipticity is pragmatically permitted, then indeed in some cases it is to be expected, and this accounts for the changing of motivational force as in moralism. However, I argue that it remains unclear how, in moralism, a speaker’s moral claim retains the ‘psychological force’ to affect motivations that Finlay claims that it does. This, I call the Problem of Preference Osmosis. I will argue, instead, that given the endorsement of Humean psychology, it is a more elegant solution, to propose that even though in moralism the salient end may not be instrumentally shared by the hearer, to the extent that it affects their behaviour, it must ‘plug in’ to a higher-order end of theirs.
If it is true that moral claims function elliptically, Joyce argues, referencing implicit end-relational claims, then they should continue to function exactly as so though end-relational claims are instead expressed explicitly. Joyce (2011, 527) and Finlay (2011, 353) agree that there exist some examples of normatively motivating claims that function elliptically, such as an All Blacks rugby captain saying “You ought to pass, after their commit”; the implicit claim being “Pass, after their commit, so we win the game”. However, Joyce argues, such an analysis does not extend to moral claims. If a Nuremberg trial judge were to keep relativising their condemnations of war criminals with “…by our moral standards” or “for our ends” (Joyce, 2011, 527) the moral force of condemnation would be entirely and scandalously diminished. Given the premise that, if truly reducible as Finlay claims, explicit end-relational claims should function identically to elliptical and implicit end-relational claims, then the Nuremberg trial example serves as a counter-example to Finlay’s theory and suggests that there may be something different going on in such moral claims.
There are two options to push back on Joyce’s argument, as I see it. The first is to deny that it really would be so scandalous if the judge explicitly relativised their claim, and the second is to come up with some sort of explanatory story, explaining the difference in apparent illocutionary force between the relativised and unrelativised utterances.
On the first option, I argue that one might agree that the Nuremberg trial judge’s utterance, in suffixing them with “by our moral standards” diminishes the force of their utterance, but claim that this is due to the fact that “by our moral standards” or “for our ends” is a vacuous and content-less addition, that is already assumed. In other words, one might agree that tacking on “for the sake of our interests in general” weakens a claim, but maintain that making explicit a specific interest maintains the moral force. To this end, it is worth noting that many of the Nuremberg judgements were explicitly relativised, in that many charges levelled against individuals were explicitly “crimes against peace” (Sayapin, 2014, 148). On the end-relational picture, this puts forth a particular end that aligns with people’s contingently desired world state: peace. In contrast, writing the blank cheque of “for our ends” (whatever they may be) is less motivational. The charge, therefore, is that the wrongness in force, is merely the wrongness of the misidentified end.
Finlay puts forth a version of this first defence, arguing that the judge’s condemnation might instead be read as relational to “promoting general human well-being” (2008, 543). Whilst this end is at least slightly more specific than the blank cheque of ‘for our ends’ or ‘by our standards’ it appears to me to be still too vague to be strongly motivating, though the reader may disagree.
Finlay’s main line of defence, however, is to take the second option and provide an explanatory reason as to why there is a difference in force between the explicitly versus implicitly relativised claims. His central move relies on the pragmatic assumption of hearers that speakers are speaking optimally in order to accomplish their conversational ends. From here, he argues that once ellipticality is accepted, in certain situations, and for certain speech acts it will be expected. Thus, to speak explicitly, will in fact change the very nature of the speech act, due to the pragmatic assumptions of the hearer. In other words, the hearer will search for a reason that the speaker conveyed the message in long-form, instead of short-form. ‘Take your shoes off’, Finlay argues, is straightforwardly understood (2014, 149), whereas “Take your shoes off your feet” has a different pragmatic effect, despite one presumably being elliptical for the other. The long-form articulation makes us pause to wonder why it was necessary, whether there is something about the speaker’s ends involving taking our shoes off of our feet specifically, such that they had cause to mention it. Finlay therefore accounts for the difference in force between elliptical and non-elliptical speech acts via the pragmatic expectation that speakers are speaking optimally to accomplish their conversational ends.
The previous example, however, uses an imperative. How does all of this relate to unrelativised moral claims? Here, Finlay employs his account of moralism. A speaker, by speaking as though their ends are so ubiquitous and are shared, in their lack of need to reference them, displays a ‘show of force’ that is not present in the explicit version. By speaking in such a fashion, Finlay argues, even in the presence of hearers that do not share the ends of the speaker, the speaker puts “psychological pressure on [them] to share that preference” (2014, 189)(2008, 357)(2011, 543), “demanding they make it true” (2011, 187) as well as pragmatically expressing the force of their own preference and an “intolerance of refusal” (2014, 187). The Nuremberg judge who explicitly relativises their condemnation then, Finlay argues, fails to express this strong preference or elicit the same psychological force. The end-relational theory, then should be considered to survive The Problem of Explicitation.
Finlay’s defence, however, has brought up a mystery. What exactly accounts for the power of moralism? Indeed, we were initially marketed the end-relational theory in virtue of its avoiding of the ‘spooky pricking of thumbs’, or the motivational power of non-naturalistic facts, and yet in the case of moralism, we seem to be ‘spookily pricked’, psychologically pressured, by preferences that are not our own. On the end-relational view, as it currently stands, I argue, it appears that hearers fall under the sway of ‘preference osmosis’, merely by being exposed to another’s preference.
I argue, that instead of relying on a mysterious ‘preference osmosis’, given the previous endorsement of Humean psychology, it is a much more elegant solution to simply come up with an explanatory story of expressed preferences that can be accounted for in terms of ‘plugging in’ to a hearer’s ends. Whilst this means that we give up categoricity, deflating it down to contingently defeasible pseudo-categoricity, it has the advantage of a consistent account of motivation. On the Humean picture, it appears as though we can have our motivation forces be non-mysterious, or categorically/always motivational, but not both. Contrast these two cases:
Ice Cream (Interview): It is your turn at the ice cream store, but you are still pondering what flavour to buy. On an old TV in the corner, a large crowd of bikers is being interviewed by a local news channel, “Look, I’m a nice guy,” you hear one of them say, “I like strawberry ice cream, you ought to get it”.
Ice Cream (Behind You): It is your turn at the ice cream store, but you are still pondering what flavour to buy. The store is absolutely packed full of bikers. “What’ll it be?” Asks the attendant. “I like the strawberry” rumbles a deep voice, which as you turn around, realise belongs to the largest, toughest, most grizzled biker, “you ought to get it”.
In both of these cases, the prescription, and preference for strawberry ice cream has been heard, from (let us say) the very same biker. But is it true that these prescriptions exhibit equal motivational force? I think not. And I think the reason why not, is elucidated by the question: In Ice Cream (Behind You), do your chances of experiencing ‘trouble’ post-ordering in any way change based on whether you do or do not choose the strawberry ice cream? I think they do. This is because we are forced to pragmatically presume the illocutionary force of the biker’s expressed preference, forced to ponder why they said that, and what they might be even slightly more likely to do about it, if we disregard their expressed preference. Equally, even the bare expressed preferences of Neighbourly Ned and Nelly may be enough to elicit motivational force from us, given that is usually in one’s interests or ends to be on pleasant terms with one’s neighbours. Notice that this motivational force evaporates, once we have decided to move; their preferences, and what they might be able to do about them, are no longer our problem.
Of course, one might point out that the expression of bare preferences, as in the case of Ice Cream (Interview) are still capable of motivating us, this is why we have advertisements full of celebrity endorsements. But even this, I argue, can be subsumed under ends-based information, in which they present that if we desire to be That Kind Of Person, then we ought to buy This Kind of Product. Equally, the motivational force evaporates, or even goes in the other direction if the preferences of a person at odds with our self-idealisation are expressed. Moralism, or categorical claims, on this view, therefore functions as a signalling method, in order to convey an agent’s unwillingness to persuasion or high commitment to a course of action, and its motivational force is derived from the way this ‘locked in’ course contingently affects ends of the hearer. Therefore, the end-relational view has survived both the Problem of Explicitation and the Problem of Preference Osmosis, though we have exchanged mysterious categorical motivation for the non-mysterious, contingent Humean variety.
5. Humean Psychology
In this next section, I focus on analysing the theoretical repercussions of Finlay’s reliance on Humean psychology, the principle that people will and do endeavour to optimally accomplish their pre-reflexively unchosen ends. I will argue that reliance on this premise serves to functionalise human decision-making, making it no surprise when moral claims turn out to be a particular form of information exchange. I will argue that this results in a potentially strange, though logically coherent affective picture, of Automaton Valley. Finally, I will examine the necessity of an explanatory story for cases of akrasia, or weakness of will.
Finlay relies on an “Instrumental Law of Pragmatics”; that people will always act optimally to accomplish their conversational ends (2014, 123), for much of his argumentation about how a hearer is able to deduce the salient end and elliptical content of claims. But where could this principle but from the broader Humean principle; that people will always act optimally for their ends in general.
Certainly, one could attempt to come up with an explanatory story for why only in the bounds of conversation people optimally attempt to accomplish their ends, but such a restriction seems a) ad-hoc and b) not endorsed by Finlay (2014, 17).
The upshot of this, however, I argue, is that while the semantic content of moral claims may be as Finlay says, (“X raises the probability the most for salient end E”), the pragmatic content, once the quiet part is said out loud about our Humean assumption, is actually much stronger. Given that we retain the assumption that people will act in an optimal fashion to accomplish their ends, we are really pragmatically expressing “You will do X, if if you a) realise that X raises the probability the most of E and b) E is one of your ends”.
I intend to convey the form: ((Realise Optimality of(X for E) & ‘Has end’ (E) → Will(X) ). This is similar, but Humean, to Philippa Foot’s (1972) system of hypothetical imperatives.
This is to say that if we can demonstrate to someone that X is the action that will counterfactually raise the probability of salient end E the most, and salient end E is one of their ends, then it necessarily follows that they will pursue X. Equally if someone says that will not do X, we can deduce that they either do not think that X raises the probability the most of E obtaining of their ends, or they are not aiming at E, or both. As such, once we ‘bake in’ the normative, necessarily motivating Humean assumptions, we are left with a pragmatic content that, I argue, has served to functionalise, as in, reduce down to a function, human decision making.
Finlay insists that the semantic content of moral claims should be understood as regarding probability relations regarding certain world states, and this allows him to argue for the truth of seemingly categorical claims even if they do not align with the ends of the hearer. Even if they were not motivating, he argues, to the Nazi on trial at Nuremberg, because they did not share our ends, it is still true that their actions lowered the probability of our end of “general human well-being” (2011, 543). But, notice that the same may be said about what I have argued the pragmatic content of such utterance to be under the end-relational picture, due to the conditional form. There is nothing stopping the Nazi from agreeing that if they held the same end, and they also agreed that their actions were not optimal for this end, then they would not have engaged in such actions. While attributing this conditional relationship to the semantic content of claims may further strain our reflective understanding, my point is merely that this functionalist pragmatic content is exactly what one must think they are doing when they make moral claims under the end-relation view. Under the end-relational theory, claims, including moral claims, are made by an agent, for the purpose of their own ends, by providing information to another agent, which if it ‘plugs in’ to their ends will affect their behaviour, under the Humean assumptions.
There is something of a shell-game going on with the functional concept, of which an agent’s behaviour is merely the result (this concept is borrowed from Finlay (forthcoming, 3) though it is employed here to different effect).
It seems that there is something closely related to morality that is fundamental, irreducible, motivational, and mysteriously known (cf. Shafer-Landau, 2003, 210). On the end-relational picture, this functional concept is ‘ends’ however, the same result may be accomplished by any other concept which does the same motivational work (cf. Finlay, 2014, 251). As such, the same functionalist picture is achieved if agents necessarily act in accordance with their ‘all-things-considered’ reasons, in response to their perceived balance of ‘non-natural’ normative facts, or on the balance of their conative affections.
This functionalist picture, however, leads us to a logically consistent, but perhaps uncanny result. It is possible for one to ‘wake up’ on the Finlayian functionalist picture, and realise that there is nothing ‘behind the eyes’ of the agents around them but the whirring of ends-calculations, and the exchange of linguistically-transmitted probability-information. This information transmitted, of course, only when green-lit by the agent’s ends-maximising function. On such a view, then, one has awoken in Automaton Valley. Such uncanniness is the result of the total objectification of agents, such that they behave in predictable and rational ways in accordance with their information inputs, pre-reflexively given ends, and outputted behaviour.
On the other hand, such a theory is eminently useful for accomplishing one’s ends, if it is true that agents behave in such a manner. If they do, it follows that in order to affect an agent’s outputted behaviour one must provide them with inputs in the form of information that they are not currently basing their actions on, and this accounts for the normative effect of the conveyed probability relations. We might put this as being an infnormative process.
There are two options I will consider, that might give us reason to reject the functionalist story, both refuting that agents necessarily act as optimal automata in pursuing their ends, in other words; rejecting the premise of Humean psychology. The first will focus on first-personal phenomenology, and the second on akrasia.
The first option is to deny the accuracy of the description from a first-person perspective and say that phenomenologically, it does not feel like one is merely calculating and pursuing their ends, but is able to weigh, choose and evaluate between them (cf. Jones, 2003, 2018). Indeed, contrary to the functionalist automata picture, this appears to be what Finlay endorses, as he frequently mentions agents having to ‘choose’ between various ends (2014, [39, 142, 154]) suggesting an evaluation. Therefore, the first-personal phenomenology of our ability to evaluate our decisions, and seemingly ‘weigh up’ our ends, may give us reason to reject this functionalist picture as inaccurate.
A rejoinder, however, that the Humean functionalist may offer, is it is unclear by what metric we might evaluate and choose which of our ‘ends’ to follow. The issue goes like this: If we, in evaluating what ends we ‘choose to follow’, merely choose that which is in accordance with our higher-order ends, then this is no choice at all, but rather is the computation of our ends upon themselves, sorting themselves into higher and lower order ends (cf. Street, 2012, 51). Thus, this feeling of ‘choosing’ is merely a derivate of this process of computation of our pre-reflexive or ‘bequeathed’, ends. If, on the other hand, there is some other sort of metric we have for choosing our ends, such as those with a good ‘reason’ or those supported by the ‘facts’, then it seems like that concept, instead of ends, is our functional concept, as it ultimately determines how we choose, and therefore the functionalist can still tell their computational story. Finally, we might think that perhaps we ‘just choose’ our ends, without any metric whatsoever, but this seems to imply that we just choose randomly, i.e. are not concerned with any criteria for the outcome, and this does not seem characteristic of choosing at all. Therefore, the functionalist can tell an explanatory story about the phenomenology of choosing, that it merely arises due to the computational process of one’s ends (or other functional concept) applying recursively to themselves, sorting themselves into hierarchical order. Such a revisionary tale of our concept of choice, however, may strain our intuitions.
The second option, as a natural counter-example for any theory that posits necessary behavioural outputs from inputs, is akrasia, or ‘weakness of will’. In akratic cases, we might imagine an agent that a) recognises themselves as having a particular end E, b) recognises that X will most raise the probability of E obtaining, and yet still c) fails to pursue X. Perhaps an agent, Amy, recognises herself as having the end of running a marathon in November, acknowledges that they ‘ought’ to go for a run, relational to their end, and yet when the alarm rings at 5:30 am, they do not. Akratic cases therefore pose a counterexample to the end-relational view, as they threaten the premise that one’s Humean ends necessarily provide motivation, which Finlay relies on for the motivational power of moral claims.
There are two answers, I argue, the Humean psychology endorser may provide. Indeed, Finlay appears to endorse versions of both, at different times. The first is to insist that proper identification of ends necessitates motivation (Finlay, 2008), and therefore the akratic agent has misidentified their ends (cf. Finlay, 2014, 193). In this instance, though Amy, identified her ends at 9:30 pm as running a marathon in November, and therefore deduced she ‘ought’ to go for a run the following morning, ‘morning Amy’ evaluated her ends of sleeping in as higher than contributing to her higher-order end, and therefore acted appropriately. This solution, however, does not account for the feelings of misgivings we may experience on akratic occasions. This, however, might be explained by a split between identified ends in judgement, and operative preferences (Finlay, 2014, 193). In this sense, though Amy thinks at 5:30 am that her end is marathon running, unbeknownst to her, she is now acting in accordance with her sleep-in end. Whilst this accounts for misgivings we may feel from a split in recognised ends versus operative ends, it has the unintuitive upshot that we may be unaware of the ends we are acting, which seems contra, I posit, to our usual phenomenological experience.
The second answer, is to weaken the Humean principle, thus that one is only necessitated to endeavour to bring about the end, not necessarily succeed (cf. Finlay, 2008, 184). This is to deny total control of the outcome of one’s behaviour. Just as we would not judge the Humean hypothesis to be incorrect if someone attempting to backflip failed while intending to succeed, or a narcoleptic person fell asleep despite intending to stay awake, perhaps we are to equally consider Amy to be earnestly endeavouring to go for a run, despite ultimately failing to do so. The advantage of this answer, I argue, is that it distinguishes between akratic cases of sleeping-in from genuinely enjoyable sleep-ins, in that in the akratic case, there is a part of ourselves that is genuinely trying not to, or recognises us as performing non-optimally against our ends. Perhaps this is the part that is responsive to Humean preferred-world-state/end considerations, and that may be addressed with probability-facts in end-relational moral claims.
Whilst neither of these arguments I take to be conclusive on the topic, I take them to indicate the possibility and necessity for a functionalist view such as Humean psychology to extend its revisionary story to cases of akrasia, if it is to rely on the internal motivation of its functional concept.
I have therefore argued that Finlay’s reliance on his instrumental pragmatic principle, that people will optimally pursue their conversational ends, in fact belies his commitment to the stronger Humean principle that people will optimally pursue their ends. This was argued to functionalise agential decision making, leading us to the uncanny situation of Automaton Valley, in which agents are objectified, necessarily pursuing their ends, occasionally through conveying probability-related data in the form of ‘moral’ claims. This was argued to be logically coherent, and even useful for accomplishing one’s ends, but uncanny. Furthermore, I have suggested that in response to the challenge of akrasia, the Finlayian functionalist must tell a revisionary story, giving up either our ability to always know our ends or always control to always our actions. As such, the end-relational theory may survive consistently by adapting, growing stronger, perhaps, but stranger.
6. Relationality
In this section, I discuss two versions of the prime counterexample to relativistic metaethical theories: The Problem of Obvious Evil. Again, I argue that the end-relational view survives, but must extend its revisionary story. Furthermore, I argue against Finlay’s claim that the end-relational theory is neutral regarding first-order moral claims, suggesting that it is an undermining ‘grey’ sort of neutrality, based on the contingency of their motivational force. This contingency, that were we constituted differently, so too would our values be, results in a sense of ‘vertigo’, ungroundedness or irony, to which I offer two avenues out of, attempting to reframe the uncanny Automaton Valley to the cannier Glade of the Destined. I explore whether evolution, in light of the benefits of cooperation, might guarantee us on some level having shared ends, and conclude that it may make overlap more likely, but not guaranteed.
The Problem of Obvious Evil is a Moorean argument from certainty, seeking to reductio a relativistic theory. Any theory that sanctions such an obviously evil act, it argues, must be wrong. The first version focuses on the motivational internalist, or what I have described as the functionalist nature of the end-relational theory (Shafer-Landau, 2003b). It argues that:
- If it is the case the agents are merely pursuing their ends, then they might not have any reason not to act ‘evilly’.
- And yet, we might know that such agents can be blamed, punished, and called evil for their actions.
- If an agent had no other reason not to perform an action, we cannot blame them, punish them, or call them evil for doing so.
Therefore the argument goes, we should reject a), and claim on the grounds of our certainty for our intuitions of b) and c) that there must have been multiple reasons available to the agent. Given that they had no other options, it would be like blaming, punishing or calling evil a poker-playing-robot, programmed to make the best moves with the cards it is dealt.
The end-relational theory, I argue, can survive if it merely rejects c), and offers an explanatory story for how agents can be blamed, punished, and called evil for their actions, even if they were’ just following their ends’. Blame is the easiest to accommodate on the Finlayian functionalist picture, given the objectification of agents, one may casually blame an agent, as one might blame a poker-playing robot having a better hand than them as the reason they lost. Equally, punishment, on the ends-relational picture, is not retributive, but merely a deterrence penalty that affects the decision-theoretic expected-payoffs of actions such that agents change their behaviours. Evil is perhaps trickier, and following Finlay’s method, we might expect that someone who is ‘evil’ is superlatively ‘very bad’ such that they maximally lower the probability of our ends obtaining. But given that someone can lower the probability of our ends consistently merely through their own error, and not be ‘evil’ I propose that an end-relational concept of evil is more likely applied to agents who unabashedly pursue final ends that fundamentally differ from our own, and therefore maximally lower the probability of our own most fundamental ends.
One might respond, however, that this reply merely dodges the question, this account does not account for the distinct emotional response, associated with blame, punishment and evil. In a certain important sense, there is no use getting angry with the agent, because they were merely acting in accordance with their ends, and the means available.
But this is not necessarily the case. It seems like a contingent manner whether anger is appropriate for the situation, contingent, one might think, on whether it serves one’s ends by acting as a deterrence (Mcdermott, Lopez, and Hatemi 2017). This is to say that we may continue to view anger, or any emotional response as instrumentally serving our higher-order ends, and therefore still appropriate, and may cultivate it to occur in other similar situations (cf. Jones 2018). The end-relational view, therefore, by providing a revisionary story such that it is appropriate to blame, punish, call evil, and in certain end-serving situations ratify the appropriate emotions, is able to deny premise c), and therefore survive the first version of The Problem of Obvious Evil.
The second version of The Problem of Obvious Evil focuses on the truth of different moral statements. If the end-relational theory licenses the acts of an atrocious person, merely by relativising it to their ends, then given we know the act is morally wrong, we ought to reject the end-relational theory. Here, examples are numerous, Nuremberg Nazis (Finlay, 2014, 184)(2011, 527), Saddam Hussein (Finlay, 2014, 208) Gyges; an invisible Platonic criminal (Joyce, 2011, 524), practitioners of female genital mutilation (Wong, 2023, 49), a coherent but torture-inclined Caligula (Gibbard, 1999,145) and, of course, we have used The Dropper. The Problem of Obvious Evil hears that moral claims are made in relation to ends, and, pointing at the most dastardly ends in sight, asks “Even those ends!?”. The end-relational theory must therefore accommodate our intuitions of condemnation if it is to survive.
Finlay’s answer is relatively succinct and will be familiar to us from the Nuremberg Nazi example. Yes, he argues, claims could be made relative to those agents’ ends. It is possible, he writes, that in speaking with his generals, Saddam Hussein, relational to their salient end of quelling the Kurdish resistance uttered a true proposition in saying “We ought to use chemical weapons on the Kurds”. To disagree, would be to argue that chemical weapons were not the best salient option to raise the probability of the Kurdish resistance being quelled. But, he argues, there is nothing stopping us from making claims relative to our ends, such as “general human well-being” (2011, 543), utility (2014, 254), or even Kantian universifiable maxims being upheld (2014, 255). Indeed, if someone were to ask us “Ought Hussein have used chemical weapons?”, we may violently exclaim “No!”, relational to our (presumably shared) salient ends. As such, Finlay argues that the end-relational theory survives the second version of The Problem of Obvious Evil, arguing that it remains “neutral in regards to which ends are ‘moral’” (2014, 254) and that we can persist in making all of our classic, condemnatory first-order moral claims.
Whilst the end-relational theory allows us to condemn acts in relation to our ends, we might still think that, frankly, the Problem of Obvious Evil demands more, demands that we be able to make claims that do apply to Hussein, or Caligula, or The Dropper. I argue, therefore, that it is exactly this neutrality in regards to ends, that means that we cannot continue going on making our first-order moral claims on the end-relational view, at least with the assumption that first-order frameworks provide the right answer for what an agent ought to do. On most ‘normal’ uses of these first-order frameworks, I argue, these realist and universalist assumptions appear built-in, and it is these assumptions that provide a reason for agents to make such claims; they are capable of motivating others in terms of generating the ‘right’ action for them. On the end-relational view, however, making such a claim lacks this guarantee of motivation, or even of ‘right’ action, one must merely appeal to the brute luck of whether the receiver so happens to have been ‘dealt’ the ends of ‘global utility’ or ‘universifiability’. The claim that under the end-relational theory, one can continue making classic first-order claims, I argue, is in the vein of claiming that “Even though we have broken up, we can still keep hanging out”, or “Even though the dog died, we can still keep him”. Even though one can keep making first-order claims, their motivational power, the reason one would do so, is gone. Given how unlikely it perhaps is that someone truly pre-reflexively desires the ends of ‘global utility’ or ‘universifiability’ it is unclear if the rational end-relational theory endorser would ever make such claims.
Unless! One way, I argue, of vindicating first-order moral claims under the end-relational theory is to provide a genealogical or contractrarian story of how they in fact do ‘plug-in’ to our ends. On this picture, first-order frameworks such as utilitarianism or Kantianism may function as crystallised ‘negotiation frameworks for preferences/ends’ (cf. Harman, 1975, 16 and Railton, 1986, 192). In this sense, such systems, even though on occasion they may constrain an individual’s ability to realise their ends, through the reduction of cooperation-costs in virtue of many agents signing on to an ‘agreed’ ruleset, in the long-run allow agents to fulfil more of their ends than they would otherwise. In this sense, even an agent who endorses the end-relational theory might make such first-order claims, given that they might fairly assume them to be held as instrumental ends by another agent. Therefore, even under the end-relational view, we might continue to make classic first-order claims, provided we consider them as contingently held instrumental ends. If, however, we still hold with certainty that a proper metaethical theory must vindicate our intuition that we should be able to leverage motivational moral claims at all times to all agents, though this seems a tall order, we must reject the end-relational view as ultimately the motivational force of its moral claims still rests on the basis of contingently held ends.
The fact that the end-relational theory, holds that there is no justifying force more fundamental than our contingently held ends, may result in a sense of vertigo (Street, 2017, 148) or ungroundedness - a sense that there is nothing ‘holding up’ or justifying that our functional concept is worthy of orienting our behaviour. Sure, we might think, we have our final ends, but to what end ought we pursue our final ends? (Finlay, forthcoming, 14) Certainly, we might have our reasons, but do we have good reasons for having them? (Korsgaard, 1996) (cf. Scanlon, 2014, 123). It seems like, on the end-relational view, with its most fundamental authority being contingent ends, there is no way of guaranteeing that our ends are the ‘right’ ones. Indeed, I now confess that the purpose of placing Neighbourly Ned and Nelly right next to The Dropper at the start of the thesis, was an attempt to prompt the reader to ponder, “What if I were the coherent Caligula?” If certain world-states (ends) are preferable solely because we just so happen to prefer them, or have them as ends, well, then we may worry they are not really preferable at all, we just so happen to be constituted to think that they are. If the contingency of our ends implies they are in fact value-less, and all our ends are contingently held, then we have found ourselves at nihilism. This, however, is certainly grounds for another argument from certainty. Holding as certain that we do have ends that we value, we may therefore either reject that our ends are held only contingently, or reject the implication that contingency entails loss of value.
I will examine versions of both arguments, hoping to demonstrate that the end-relational theory, despite its fundamental reliance on contingently held ends, need not collapse into vertiginous nihilism.
The first method is to argue that even were our contingently held ends to be considered unworthy of orienting our behaviour around, our ends are not contingently held, they are necessarily held, as the necessary outputs of a selection of inputs. This is to say, that on the basis of ‘one’s upbringing’/ ‘the history of humanity’ / ‘the universe since the big bang’, it necessarily follows that the specific person that one is, necessarily has the values that they do. This is capable of being universalised (Hare, 2013) thereby claiming that any individual, such that the specific selection of inputs occurred, would have the same outputs, the same ends. These ends are therefore considered justified as being acted upon.
There are two primary issues with this argument. The first is that it seems like no matter what selection of inputs one chooses, whether it is one’s upbringing, or the history of the universe, the open question remains “Why couldn’t that have gone differently?”. Then, no matter what is asserted antecedently and argued to necessarily result in the first input occurring, the same question may be asked, ad infinitum (this retains the same form as Korsgaard’s (1996) argument for the justification of reasons). It simply seems like it all could have gone differently. It seems we either commit to an infinite chain of justification for our ends (in this particular argument, causal justification), or we commit to a self-justifying bedrock. The second problem is that such an argument results in the axiomatic justification of whatever values one happens to have. The Dropper, just as much as we can, can point to the long chain of causal necessity resulting in them having their ends. This argument, even though it claims to universalise, universalises to so specific a scenario, such that it fits exactly one agent, and therefore ‘justifies’ whatever ends they happen to have, therefore failing to differentiate between justified ends and unjustified ends, therefore failing to abate the vertigo.
The second method is to argue that even if our ends are contingent, this does not imply any lack of value or unworthiness in orienting our actions around them. What is interesting about the previous rebuttal of the regress of infinite causation argument, is that it applies equally, to all causally impacted things. Therefore, if our ends are taken to be causally generated (and if we think they have been evolutionarily shaped, it seems like we do), then our end’s causal status is no more troubling than anything else that we encounter. Certainly, if the world were so constituted that the sunset was not beautiful, it would not be, and people would not stop to watch it. But it is, and so they do. Equally, were the people that are our parents, partners, or children, not our parents, partners, or children, we would love others instead of them, but they are, and so we do (Street, 2012, 57). All of this is to deny that the contingency of our ends matters for how justified we are in orienting our behaviour around it. If our ends, values or perceived reasons were different, they would be different, but they are not. This is the nicer version of the affective picture of Automaton Valley, perhaps ‘The Glade of the Destined’, in which, the world being as it is, is such that people have the ends that they do. Of course, the glade is host to all of our baddies, such as The Dropper, too. What’s more, we have no more way than before of judging between justified ends and not. This merely seems to speak to the impossibility of us attempting to simultaneously declare that ends (or our functional concept) are the most fundamental valuing-concept, and then attempting to reach beyond them, to further licence the values they accord. As we cannot reach beyond whatever our functional concept is, it seems we must merely accept, whatever it, on balance suggests (cf. Scanlon, 2014, 123) (cf. Finlay, 2014, 256) This solution, then, in denying the effects of contingency upon the justifiability of our ends is only a halfway solution. We may accept that our ends are justified in orienting our behaviour, but in doing so, we license all ends. This does not mean, however, that we must sit idly by, and let The Dropper drop, indeed, according to this argument, we are equally justified in making The Dropper stop.
What is the role of evolution in all this? If we are asserting causal stories for the development of ends, it’s certainly intuitive that there might be a role for it to play. Theorists such as Michael Tomasello, (2014) Kim Sterelny and Ben Fraser (2017) have proposed that our ability to cooperate, and ability to form ‘moral frameworks’ of behaviour modulation may have been a key evolutionary advantage for the human species. What does this mean? Can we go up to The Dropper and tell him that he is doing the wrong thing, in virtue of how cooperation and collaboration conferred evolutionary advantage? I do not think so. Our shared evolutionary past does not guarantee that one’s ends align with others, after all, perhaps their neurons ‘just fused like that’. However, if it is true that cooperative ends did confer an evolutionary advantage, then, as with other evolutionary advantage-conferring characteristics such as snake detection (Öhman, Flykt, and Esteves 2001), we might expect to find them widespread.
Indeed, perhaps this explains our intuition, that there simply must be some understanding, some reason or shared end we can offer another person (Joyce, 2011, 525, and implicitly in Shafer-Landau, 2003b, 207), in that this is an adaptive behaviour, encouraging us not to give up on someone too early. Therefore, whilst not in itself exerting any normative force, the fact that evolution may have been selected for cooperative dispositions should give us hope that, in more cases than not, we will be able to find shared ends.
The end-relational theory, therefore survives both versions of The Problem Of Evil, allowing us to continue to morally condemn, salient to our ends. This gives up necessary moral motivation, instead relying on contingently motivational ends. I have argued that our lack of justification from beyond our contingently held final ends may result in a sense of vertigo, but need not lead to nihilism, as we may claim that contingency need not bear on our ability to orient our behaviour around our ends. In this sense, we may reframe Automaton Valley to the Glade of the Destined. Furthermore, I have argued that while the evolution’s potential selection for cooperative ends in itself confers no normative force, it should give us reason to hope that our ends overlap.
Conclusion
Ought we accept the end-relational theory? Well, on the end-relational theory, we should consider whether such acceptance will counterfactually raise the probability of our ends obtaining. Presumably, an optimal metaethical solution to David O. Brink’s Puzzle Of The Rational Authority of Morals will do just that, and in this thesis I have argued that Stephen Finlay’s end-relational theory provides a more attractive option than might initially be thought, in face of the prima facie difficulties of ‘motivational externalist’, ‘agent-neutral reasons’ or ‘metaphysical egoists’ solutions. Though solutions in the relativist basket face their own series of challenges, such as accounting for categoricity, and disagreement, these were accounted for via the quasi-realist, quasi-expressivist nature of the end-relational theory, in which moral claims are elliptical for probability facts which ‘plug-in’ to an agents pre-reflexively given Humean ends. From there, challenges were surveyed to the three primary premises of the view: Analytic Reductionism, Humean Psychology, and Relational Morality. It was seen that the end-relational theory survives The Problem of Explicitation through the pragmatic force of elliptical speech acts, but I argued that in order to not rely on mysterious ‘Preference Osmosis’ we should adopt a more narrowly Humean conception of the force of categorical prescriptions, that they function as signalled ‘course of action’ ‘lock-ins’. The end-relational account’s reliance on Humean motivational internalism was argued to result in much stronger pragmatic force than was initially proposed; that given their Humean psychology once agents realise certain facts, they will act in certain ways. This was argued to lead us to the uncanny, but not inconsistent picture of Automaton Valley in which all agents merely pursue ends, and exchange information. Furthermore, it was argued that the end-relational view could account for akratic cases, but such conceptions came at the sacrifice of either always knowing our ends, or always controlling our actions. I argued that the end-relational theory can survive two versions of the Problem of Obvious Evil, by offering us explanatory stories about blame, punishment and ‘evil’ though these conceptions might strain our reflective understanding. Furthermore, we saw that the end-relational theory is neutral, but I argued in an undermining fashion, to classical first-order claims, due to the final authority, on the view, of contingent ends. Furthermore, I argued this need not lead us to nihilism, and that these contingent ends retain their value. Finally, it was argued that evolution, whilst not conferring normative force as a concept itself, may have selected for cooperative dispositions, therefore giving us reason to hope that The Dropper, Ned, Nelly, and ourselves, align in some ends.
I sincerely express my gratitude to Holly Lawford-Smith, Peter Chilcott, Joanne Becker, James Chilcott, Julian Grimm, the University of Melbourne Philosophy Faculty and the 2023 Honours Cohort, for their roles in the actualisation of this thesis. It has been and remains a pleasure.
References
Brink, David O. 1992. “A Puzzle About the Rational Authority of Morality.” Philosophical Perspectives 6: 1–26. https://doi.org/10.2307/2214236.
Finlay, Stephen. 2008a. “Motivation to the Means.” In Moral Psychology Today: Essays on Values, Rational Choice, and the Will, edited by David K. Chan, 173–91. Philosophical Studies Series. Dordrecht: Springer Netherlands. https://doi.org/10.1007/978-1-4020-6872-0_10.
———. 2008b. “The Error in the Error Theory.” Australasian Journal of Philosophy 86 (3): 347–69. https://doi.org/10.1080/00048400802001921.
———. 2011. “Errors Upon Errors: A Reply to Joyce.” Australasian Journal of Philosophy 89 (3): 535–47. https://doi.org/10.1080/00048402.2010.510531.
———. 2014. “Confusion of Tongues : A Theory of Normative Language / Stephen Finlay.,” January.
———. 2020. “Reply to Worsnip, Dowell, and Koehn.” Analysis 80 (1): 131–47. https://doi.org/10.1093/analys/anz087.
———. forthcoming. “Beyond Normativity: Can Metaethics Escape Samsara’s Wheel?”
Foot, Philippa. 1972. “Morality as a System of Hypothetical Imperatives.” The Philosophical Review 81 (3): 305–16. https://doi.org/10.2307/2184328.
Gibbard, Allan. 1999. “Morality as Consistency in Living: Korsgaard’s Kantian Lectures.” Ethics 110 (1): 140–64. https://doi.org/10.1086/233207.
Harman, Gilbert. 1975. “Moral Relativism Defended.” The Philosophical Review 84 (1): 3–22. https://doi.org/10.2307/2184078.
Jones, Karen. 2003. “XI. Emotion, Weakness of Will, and the Normative Conception of Agency.” Royal Institute of Philosophy Supplements 52 (March): 181–200. https://doi.org/10.1017/S1358246100007967.
———. 2018. “Towards a Trajectory-Dependent Model of (Human) Rational Agency.” In The Many Moral Rationalisms, edited by Karen Jones and François Schroeter, 0. Oxford University Press. https://doi.org/10.1093/oso/9780198797074.003.0013.
Joyce, Richard. 2005. The Evolution of Morality. Cambridge, UNITED STATES: MIT Press. http://ebookcentral.proquest.com/lib/unimelb/detail.action?docID=3338603.
———. 2011. “The Error In ‘The Error In The Error Theory.’” Australasian Journal of Philosophy 89 (3): 519–34. https://doi.org/10.1080/00048402.2010.484465.
———. 2012. “Enough with the Errors! A Final Reply to Finlay.” Unpublished, 1–15.
Korsgaard, Christine M. 1996. The Sources of Normativity. Cambridge: Cambridge University Press. https://doi.org/10.1017/CBO9780511554476.
Mcdermott, Rose, Anthony Lopez, and Peter Hatemi. 2017. “‘Blunt Not the Heart, Enrage It’: The Psychology of Revenge and Deterrence,” December.
Öhman, Arne, Anders Flykt, and Francisco Esteves. 2001. “Emotion Drives Attention: Detecting the Snake in the Grass.” Journal of Experimental Psychology: General 130 (3): 466–78. https://doi.org/10.1037/0096-3445.130.3.466.
Railton, Peter. 1986. “Moral Realism.” The Philosophical Review 95 (2): 163–207. https://doi.org/10.2307/2185589.
Sayapin, Sergey. 2014. “The Crime of Aggression in International Criminal Law : Historical Development, Comparative Analysis and Present State / Sergey Sayapin.,” January.
Scanlon, T. M. 2014. “Reasons and Their Strength.” In Being Realistic about Reasons, edited by T. M. Scanlon, 0. Oxford University Press. https://doi.org/10.1093/acprof:oso/9780199678488.003.0005.
Shafer-Landau, Russ. 2003a. “Moral Rationalism.” In Moral Realism: A Defence, edited by Russ Shafer-Landau, 0. Oxford University Press. https://doi.org/10.1093/0199259755.003.0009.
———. 2003b. “Reasons Internalism 7.” In Moral Realism: A Defence, edited by Russ Shafer-Landau, 0. Oxford University Press. https://doi.org/10.1093/0199259455.003.0008.
Sterelny, Kim, and Ben Fraser. 2017. “Evolution and Moral Realism.” The British Journal for the Philosophy of Science 68 (4): 981–1006. https://doi.org/10.1093/bjps/axv060.
Street, Sharon. 2012. “Coming to Terms with Contingency: Humean Constructivism about Practical Reason.” In Constructivism in Practical Philosophy, edited by James Lenman and Yonatan Shemmer, 0. Oxford University Press. https://doi.org/10.1093/acprof:oso/9780199609833.003.0003.
———. 2017. “Nothing ‘Really’ Matters, but That’s Not What Matters.” In Does Anything Really Matter?: Essays on Parfit on Objectivity, edited by Peter Singer, 0. Oxford University Press. https://doi.org/10.1093/acprof:oso/9780199653836.003.0006.
Tomasello, Michael. 2014. “The Ultra-Social Animal.” European Journal of Social Psychology 44 (3): 187–94. https://doi.org/10.1002/ejsp.2015.
Wong, David B. 2023. “Moral Relativism and Pluralism.” Elements in Ethics, January. https://doi.org/10.1017/9781009043496.