The Normativity of Moral Contractualism: A Hegelian Solution
A thesis submitted to the University of Manchester for the degree of Doctor of Philosophy (PhD) in the faculty of Humanities
2019 Fred H Horton
Social of Social Sciences
Contents
Section Page
Abstract 5
Declaration 6
Copyright Statement 7
Dedication 9
Acknowledgements 10
Introduction 11
Chapter 1: An Instrumentalist’s Answer to the Question of Normativity 19
1.1 Introduction 19
1.2 Rational Choice Theory 21
1.3 A Disposition to Comply 31
1.4 What Sort of Motivation? 37
1.5 Conclusion 43
Chapter 2: Scanlonian Contractualism 45
2.1 Introduction 45
2.2 Scanlon’s Contractualism: An Overview 46
2.3 Scanlon’s Constructivism 53
2.4 Relativism and Convergence 60
2.5 Conclusion 70
Chapter 3: The Normativity of Deliberative Contractualism 71
3 3.1 Introduction 71
3.2 Deliberative Contractualism 72
3.3 The Question of Normativity 76
3.4 Some Concerns for Southwood’s Account 79
3.5 Conclusion 89
Chapter 4: Relational Formalism and German Idealism 91
4.1 Recap: The Shortcomings of Actually 91
Existing Contractualism
4.2 A Second Person Solution? 95
4.3 Thompson’s Alternative 101
4.4 Fichtean Themes in Contractualist Ethics 107
4.5 Conclusion 116
Chapter 5: A Hegelian Solution? 118
5.1 Introduction 118
5.2 Hegel’s Dialectic of Recognition 122
5.3 The Hegelian Thought Applied to Contractualism 134
5.4 Objections 140
5.5 Conclusion 154
Conclusion 156
Bibliography 158
Word Count: 60,569
4 Abstract
The University of Manchester
Fred H Horton
The Normativity of Moral Contractualism: A Hegelian Solution
June 2019
Moral contractualism is a branch of normative ethics that holds that our moral obligations to each other are determined by the principles that we would agree to in certain hypothetical, idealised circumstances. This thesis takes an in-depth look at this approach, in particular an aspect of it denoted the ‘question of normativity’. The question of normativity is concerned with how the conditions of the circumstances of the agreement come to constitute moral reasons for real world agents. It argues that existing contractualist approaches to the question of normativity fail to provide a satisfactory answer to the question of normativity, because the accounts of practical reason relied on to do so either fail to explain the binding of our moral requirements, or do so only at the expense of undermining contractualism’s ambitions to provide an account of morality’s foundations. It then attempts to provide a positive solution to the question of the form of normativity in the form of an account of agency inspired by the thought of G. W. F. Hegel. The claim is that an approach to agency inspired by what I call the ‘Hegelian Thought’ provides us with the tools for a plausible account of practical reason; one that makes sense of the interpersonal nature of this practical reason and is not hostage to a defeasible commitment to a substantive value.
5
Declaration
No portion of the work referred to in the thesis has been submitted in support of an application for another degree or qualification of this or any other university or other institute of learning;
6
Copyright Statement i. The author of this thesis (including any appendices and/or schedules to this thesis) owns certain copyright or related rights in it (the “Copyright”) and s/he has given The University of
Manchester certain rights to use such Copyright, including for administrative purposes.
ii. Copies of this thesis, either in full or in extracts and whether in hard or electronic copy, may be made only in accordance with the
Copyright, Designs and Patents Act 1988 (as amended) and regulations issued under it or, where appropriate, in accordance
Presentation of Theses Policy with licensing agreements which the University has from time to time. This page must form part of any such copies made.
iii. The ownership of certain Copyright, patents, designs, trademarks and other intellectual property (the “Intellectual Property”) and any reproductions of copyright works in the thesis, for example graphs and tables (“Reproductions”), which may be described in this thesis, may not be owned by the author and may be owned by third parties. Such Intellectual Property and Reproductions cannot and must not be made available for use without the prior
7 written permission of the owner(s) of the relevant Intellectual
Property and/or Reproductions.
iv. Further information on the conditions under which disclosure, publication and commercialisation of this thesis, the Copyright and any Intellectual Property and/or Reproductions described in it may take place is available in the University IP Policy (see http://documents.manchester.ac.uk/DocuInfo.aspx?DocID=2442
0), in any relevant Thesis restriction declarations deposited in the
University Library, The University Library’s regulations (see http://www.library.manchester.ac.uk/about/regulations/) and in
The University’s policy on Presentation of Theses
8
Dedication
To my Dad
9
Acknowledgements
This work was funded through the University of Manchester’s Philosophy Studentship. For valuable support and discussion, I would like to thank my main supervisor, John O’Neill. And various stages throughout this project John has been joined by Steve de Wijze, Richard Child, Thomas Smith and Stephen Ingram. All have provided very helpful comments and encouragement about my work. Special thanks should also go to Joel Smith and Alan Hamlin for reading on commenting on drafts of the early parts of this thesis. I have also benefitted immensely from the comments and questions of those who attended the MANCEPT seminar and the Manchester Philosophy Research seminar (twice). The Philosophy and Politics communities at Manchester have been a further source of support and ideas. Lastly I would like to thanks my family, especially my mother, father, grandmother, Anne, Klaudia, and Tim for their support and encouragement.
10 Introduction
Moral contractualism is, roughly, the idea that what makes an action permissible or impermissible is whether or not a principle that licensed it would be accepted or rejected by hypothetical agents in idealized circumstances. Various versions of this moral theory have been advanced, and the differences between them are largely explained by the assumptions made about the details of these circumstances and the nature of the agents in question. While these assumptions will have an impact on the principles that are licensed, this is not the focus of this thesis. Rather, the focus here is on what I am calling ‘question of normativity’: how do the specific conditions of the hypothetical agreement constitute moral reasons for real-world agents?
Each account of contractualism relies on a particular conception of practical reason to answer this question. That is, they each have an account of the kinds of considerations that provide an agent with a reason to perform action ϕ. The conception of practical reason relied on must explain why the hypothetical agents would reach an agreement, the nature of that agreement, and how that agreement comes to provide (strong, binding) reasons for real-world agents.
The contention of the first part of this thesis is that the conceptions of practical reason adopted by the different contractualist approaches are, in distinct ways, inadequate to the task of explaining the relationship between what constitutes a reason in the hypothetical circumstances and what constitutes a reason in the real-world. That is, they fail to answer the question of normativity.
The second part of the thesis attempts to provide a positive solution to the question of the form of normativity in the form of an account of agency inspired by the thought of G. W. F. Hegel. The claim is that an approach to agency inspired by what I call the ‘Hegelian Thought’ provides us with the tools for a plausible account of practical reason; one that makes sense of the interpersonal nature of this practical reason and is not hostage to a defeasible commitment to a substantive value.
Before outlining the plan of this thesis, I will say a bit about moral contractualism more generally and explain some terminology in a bit more detail.
11 *
Contractualism has a long historical pedigree, and while the basic idea behind the thought experiment that characterises it is relatively stable, the details and purpose of it vary. It has been used as a means of grounding political authority,1 to determine the nature of the political institutions that we should strive for,2and as the basis for determining the obligations we have, as individuals, to each other.3 The basic idea that these various approaches share is that we ought to be governed by what we would agree to in certain hypothetical idealised circumstances. Rousseau wondered what we would will if we had the general interests of the wider community at heart. Rawls considered what kind of political arrangements we would want if we had no idea what our position or status would be in the community.
These latter approaches have much to offer, but they are not the focus here. This thesis is concerned exclusively with moral contractualism; that is, the code that would be agreed to in order to regulate individual conduct. A successful contractualist account will explain what we ought to do in terms of what would be agreed in these hypothetical circumstances presumed. So the reason I should return the money I borrowed when I said I would is because if a group of agents were determining the code we ought to live by it is highly likely that a principle of repaying debts agreed to on time would be endorsed.4
But any moral contractualism that provides answers to such question must also explain why the requirements that are derived through this process are binding on the agents they apply to. It is here that we should introduce the idea of practical reason. Practical reason concerns our reasons for action; that is, what we have reason to do and in virtue of what is this the case. There are a variety of accounts of practical reason and they have
1 Rousseau (2002) 2 Rawls (1971) 3 Gauthier (1986), Scanlon (1998), Southwood (2010) 4 A small terminological point: what I am calling moral contractualism is sometimes divided into ‘contractarianism’ and ‘contractualism’, with the former being used to describe the kind of instrumental accounts described in Chapter 1 and the latter the more substantive accounts described in Chapter 2. In order to avoid confusion, I denote all theories that attempt to ground our individual moral requirements in facts about what we would agree to in hypothetical idealised circumstances as ‘contractualism’.
12 a range of strengths and weaknesses.5 For the purposes of this thesis the most important distinction between the different accounts of practical reason is between versions that are ‘formal’ and those that are ‘substantive’.6 The former type of account would hold that we have reason to ϕ in virtue of some aspect of our agency as such. An example of this would be the kind of instrumental practical reason discussed in Chapter 1. It is a feature of our capacity as agents to use the best means available to achieve a given end. A substantive account of practical reason explains the agent’s reason to ϕ through an appeal to the substantive value that is realised by ϕ-ing. Such an account would hold that we ought to perform an action to maintain a friendship because friendship is part of what makes a good life.
In moral contractualism, an account of practical reason plays two crucial roles. First, it must provide a plausible account of how such an agreement could come to be. For example, an account of moral contractualism based on an instrumental account of practical reason might explain the moral requirement to repay money owed in virtue of the fact that self-interested agent would presumably favour a principle which required money to be paid in full and on time. It would presumably be beneficial for agents in general if such a norm is in place so there is therefore a reason to agree to such a norm.
But an account of practical reason must also play another role. Not only must it explain why an agreement is reached, it must also explain why real-life agents have a reason to stick to that agreement. For example, it might be the case that almost every rational agent would agree to a norm that prohibits the murder of one’s wealthy relatives in order to gain their wealth. But suppose an individual found themselves in a position where they could murder their real-life relatives and get away with it (they are reasonably certain of this). An account of practical reason must be able to explain why such an individual should continue to comply with the principles agreed by the hypothetical contractors. This is what I call the ‘Question of Normativity’. It is crucial that the account of practical reason that provides an answer to this is the same as the account of practical reason that explains the nature of agreement itself. If this were not the case, our reason for having a
5 See Millgram (2001) for a survey of these 6 This terminology is taken from Southwood (2010)
13 moral principle and for complying with a moral principle would rest on different groundings.
It might be wondered how great a problem this is. If our reasons to be moral are distinct from the reasons grounding the fact that a particular act is prohibited, what is the harm? Such a stance might be reasonable for some moral theories but not, I claim for contractualism. A divine command account of morality might explain the content of our moral obligations in a very different way from our reasons to fulfil them. But contractualism is a constructivist ethical theory. It gains its validity from facts about the kinds of agents that we are, in particular the way we respond in certain hypothetical, idealised situations.7 If the force of our moral requirement derives from a different account of our status as reasoning creatures, then the contractualist project begins to look incoherent.
This might seem too strict. Why demand that the force of the agreement derive from the same place as the content of that agreement? Here I introduce another assumption that will guide this thesis: that the moral contractualism(s) under investigation aim to provide an account of morality’s foundations. Facts about what we would agree to are taken to be an ‘explanatory rock bottom’. The aim of the contractualists considered is to provide an account that obviates the need for any other ethical theory. This is not a conceptual truth about any moral contractualism. It might be that our moral obligations get their force from being ultimately grounded in God’s will. But it is also possible that we cannot know God’s will directly, so it might be thought that what we would agree to in hypothetical circumstances would be a best approximation of this. Such an account of contractualism would be epistemic, rather than foundational, in nature. The hypothetical contract would be a device for discovering the best way to live, but the normative force that requires us to live in this way would ultimately derive from elsewhere.
It is assumed through this thesis that we are concerned with a moral contractualism that is looking to provide a foundational account of morality. This is supported by the textual evidence provided by the three main thinkers it assesses in the first part. It is also, I would argue, a more interesting and ambitious project to investigate. This does, however require
7 I will not defend this approach to ethical thinking in general. See Enoch (2005) for some criticisms of this approach
14 consistency between the determination of the content of the hypothetical agreement and the reasons agents have to comply with it.
Before providing an outline of the thesis I should note a couple of limitations to this approach. First, contractualism is best thought of as an account of interpersonal morality. As the title of Scanlon’s book indicating, it is primarily concerned with what we owe to each other. Contractualists have had a hard time explaining our moral requirements concerning nonhuman animals and the natural environment.8 Because it is grounded in an account of what certain agents would agree to, it seems to only apply to those who are capable of agreement. Various contractualists have attempted to offer solutions to this problem.9 I will not address this issue in this thesis.
Another problem for contractualism that won’t be addressed here is the question of whether contractualism is too demanding an ethical theory.10 As with some variants of utilitarianism, it is sometimes argued that contractualism demands an unreasonable impartiality in its ethical requirements, one that renders it less plausible as an account of what we actually do owe to each other. Again, I will not address this worry here.
I mention these concerns to emphasise the point that the argument here is not to be interpreted as a full defence of contractualism as a normative ethical theory. I am interesting in the specific problem of the normativity of the hypothetical contract. The position argued for here is less a defence of the straightforward statement ‘Contractualism is the most plausible normative ethical theory’ and more the conditional statement ‘If contractualism is going to be a plausible ethical theory, then it needs to rely on the following account of practical reason.”
Plan
Chapters 1, 2 and 3 deal with currently existing accounts of moral contractualism and consider their responses to the question of normativity and find them to be insufficient
8 See, for example, Hooker (2000) pp. 66-70 9 See Scanlon (1998) p. 183 and Southwood (2010) pp. 107-116 10 Ashford (2003 provides an overview of this kind of worry, although ultimately defends contractualism against this charge.
15 in this regard. Chapters 4 and 5 consider what would be required to provide a contractualism that can provide a sufficient answer.
Chapter 1 considers a contractualist ethics grounded on an instrumental account of practical reason. This is done through a reading of David Gauthier’s Morals by Agreement. Gauthier’s instrumental account of practical reason is informed by Rational Choice Theory. The idea here is that the rational agent (as defined by rational choice) will have a reason to agree to the a hypothetical contract that divides the surplus generated by cooperation in a fair way, where the surplus generated by cooperation is understood to be the benefit we all gain from abiding by moral principles.
I argue here that Gauthier’s account fails to provide a satisfactory answer to the question of normativity. That is, the account of practical reason assumed to underpin the agreement itself is inconsistent with the account of why agents have reason to comply with the agreement. I also argue that Gauthier’s attempt at a solution to this problem - his argument that a rationally motivated agent will develop a disposition to comply with the agreement – is unstable and cannot be squared with any plausible metaethical account of moral motivation.
Chapter 2 considers a different attempt by a contractualist to answer the question of normativity, where the agents imagined are not the self-interested utility maximizers of rational choice, but rather agents who are motivated to find justifiable ways to live with one another. This is done through a reading of Thomas Scanlon’s work in this area.
Scanlon has received a lot of criticism for seeming to bake morality into the assumptions he depends on to try and derive our moral obligations to each other. I attempt to give as charitable a reading of Scanlon as possible, but conclude that he can only answer the question of normativity through positing a substantive independent account of value as providing our reasons to comply - which undermines the ambition to provide an account of morality’s foundations and invites the redundancy objection – or by failing to provide a plausible answer of to the question of content; of what it is that we owe each other.
Chapter 3 focuses on a more recent attempt to provide an account of moral contractualism – that of Nicholas Southwood’s ‘Deliberative Contractualism’. Southwood attempts to answer the question of normativity through an appeal to an
16 interpersonal account of practical reason. An interpersonal account of practical reasontakes agency as relational in nature. He argues that this relation entails obligations of deliberative citizenship. The thought is that if we stand in a certain relation to others, then we can explain why we must comply with the hypothetical contract that would be agreed to by agents motivated by the norms associated with deliberative citizenship, without having to posit the independent source of value that undermined Scanlon’s efforts.
However, I argue that Southwood is ultimately unsuccessful here. First, I claim that Southwood has given us insufficient reason to accept the existence of such a relational account of agency. Second, I argue that even if such a relation was assumed, it is not clear that this entails the norms that Southwood thinks it does. Southwood offers an interesting account of what a successful contractualist answer to the question of normativity would look like, but the specifics of his solution look objectionably ad hoc.
Chapter 4 begins to lay the groundwork for my positive solution. It argues that Southwood was correct to posit the need for an interpersonal account of practical reason to answer the question of normativity, even if his specific solution is unsuccessful. I will look at two ways to think about what this relation would entail, arguing in favour of that of Michael Thompson.
The second part of this chapter will lay the groundwork for Chapter 5 by considering how the insights of German Idealism might be appropriated by a contractualist ethics. It will do so by assessing Stephen Darwall’s appropriation of Fichte. I argue that while this provides some insight as to how such an appropriation might work, there are still issues with this attempt to think through the kind of interpersonal practical reason contractualism requires.
Chapter 5 takes the ideas outlined in the previous chapter and adopts them to fit the work of a different German Idealist. I consider Hegel’s account of agency and mutual recognition, and argue that such an account can provide the foundations for the sort of account of practical reason that contractualism requires to answer the question of
17 normativity.11 I then apply this to an example from Chapter 2 in order to show how such an account is more successful.
Adopting Hegel in this way is going to be controversial, so I spend some time at the end of this chapter dealing with anticipated objections to this approach. I consider methodological objections to the effect that I am guilty of the same kinds of moves that I criticise earlier contractualists for. I look at objections to appropriating Hegel to a context that runs counter to his own project. And I look at the more general objection to relying on an account of agency to try and ground our moral requirements.
If the argument is successful then we will have a new way to think about moral contractualism’s attempts to deal with the question of normativity. It might also suggest new ways to adopt the insights of the German Idealists, and Hegel in particular, to modern ethical debates.
11 Another terminological note: Hegel doesn’t tend to speak in terms of ‘agency’ as such, but I will treat him as doing so for the purposes of this this and keep potentially interchangeable (but also potentially ambiguous) conflicting terms like ‘agent’, ‘subject’, ‘person’ etc. to a minimum.
18 Chapter 1: An Instrumentalist’s Answer to the Question of Motivation
In Morals by Agreement David Gauthier provides an account of why the rational agent, as understood by rational choice theory (RCT), would comply with the moral principles derived from the hypothetical contract. This means demonstrating that there are reasons that such a rational agent would accept not to reason rationally (in the sense assumed here) in every choice situation. He claims that there are RCT reasons to develop a disposition to reason in an alternative manner when deciding whether to comply with the hypothetical agreement. But such a disposition cannot simply be posited, it must be explained. Given that this new way of reasoning involves compliance with the agreement, and moral principles are determined by the contents of the agreement, an account is required regarding the connection between these moral facts and the motivation of the rational agent. Such a connection might be either necessary (the internalist account moral motivation) or contingent (the externalist account of moral motivation). However, it shall be argued, neither option works for Gauthier. The establishment of the agreement cannot provide a credible explanation of the transformation required in the agent’s normative relation to moral facts for an internalist account to hold, and the only contingent desire Gauthier can plausibly appeal to collapses back into the kind of prudential commitment to moral norms that Gauthier wants to avoid. In conclusion, answering the question of normativity requires the establishment of a connection between the moral facts (as determined by the contract) and the moral motivation of the agent, and the metaethical accounts on offer cannot be made to be consistent with the account of practical reason assumed by Gauthier.
1.1 Introduction
The reasons for drawing on a contractualism framework have been outlined in the introduction. In summary, it was argued that the contractualist approach provides the best means of determining the content of what, to borrow a phrase, we owe to each other.12 I will give a brief account of what how these are determined below. But for the most part
12 Gauthier describes his approach as ‘Contractarian’. As discussed in the introduction, I will call all the ethical theories that use the device of the hypothetical contract ‘contractualist’ to avoid confusion.
19 I will assume that the hypothetical bargaining process does yield a set of principles that correspond to the kinds of moral facts we intuitively take to be uncontroversial.13 Furthermore, it is taken that the agents involved will not, ceteris paribus, be mistaken about the nature of the principles that are derived from the agreement. This is to say that, for example, each party to the agreement has an understanding of what the concept of a ‘promise’ involves in the sense that they recognise that entering into a promissory relation with another generates certain expectations regarding their future action, and that if they make a promise they are thus expected to keep it. However, the central question for present purposes is what reason the agent has, given the assumed account of practical reason that applies to them, to continue to comply with the requirements of the hypothetical contract. The agent’s understanding of the content of what is expected of her need not, prima facie, mean that the agent is committed to act on this expectation, that is, to comply with the requirements of the hypothetical contract.14
This chapter considers the question of whether a Hobbesian version of contractualism can give a plausible answer to the question ‘why be moral’; why should the agent be committed, or bound to, the results of the hypothetical agreement? That is, given the instrumental account of practical reason assumed by such an approach, what reason does such an agent have to continue to comply with the derived requirements It will proceed through a reading of David Gauthier’s Morals by Agreement,15 since it is arguably the most sophisticated and best-known attempt to derive moral requirements from (neo) Hobbesian assumptions. Gauthier’s ambitious aim is to provide the rational – in the narrowly defined sense of rationality advocated by rational choice theorists – case for the
13 It is not the purpose of this thesis to use contractualism as a means of determining more contentious ethical issues. The focus here is on the question of normativity. It might be argued, of course, that any normative ethical theory with aspirations of settling contentious ethical questions needs to have a satisfactory answer to the question of normativity. 14 I have been careful in this example not to refer to the moral properties of promising in order to emphasise the separation of the expectations of others and moral obligations to others. Typically, we think of the act of promising as entering into a moral relation with another, but at this stage I will ignore this. The metaethical position of motivational internalism, which I will come back to in this chapter, holds that someone who appears to understand a moral requirement but is not motivated to act accordingly is conceptually confused: they don’t really understand the requirement. So, in the example, a person who understood the concept of a promise as a moral obligation but who felt no motivation to keep her promise could not be accounted for on the internalist account. Since I am looking at how a moral obligation could be binding it seems safer to assume that the agent does not understand the moral properties of the norms as yet. 15 Gauthier (1986)
20 agent to be moral. Although large parts of the book are devoted to analysing the nature of the agreement – that is, what is expected from the parties to the agreement – the focus in this chapter will be on Gauthier’s case for why a rational agent who, motivated as they are by their individual subjective utility function, would forgo the action that is in their best interests in order to comply with the norms established by the hypothetical agreement. Gauthier’s answer to this is that such a rational agent will develop a disposition towards compliance where their decision about how to act in each choice situation is not determined by the narrow motivational assumptions of rational choice theory (RCT).
Section 1.2 will outline the Rational Choice framework which Gauthier is working within, provide a necessarily brief overview of Gauthier’s project, and introduce the specific problem posed by Hobbes’s Foole – why should the results of the agreement be binding on the agent? Gauthier’s response to the problem will be briefly outlined too. Section 1.3 will unpack Gauthier’s answer to the Foole – that it is rational to form a disposition towards compliance with the agreement – by drawing on Martin Hollis’s framework of different levels of preference. The debate between motivational internalists and externalists will then be introduced and their relevance for Gauthier explained. Section 1.4 will draw on this debate to argue that Gauthier cannot plausibly ground the development he needs on either account of moral motivation
1.2 Rational Choice Theory
Gauthier’s aim is to show that it is rational to be moral; that it is rational to fulfil our obligations. This is the case since acting morally is taken to be equivalent to acting according to the principles agreed to in the hypothetical bargaining process contractualists assume determines what we owe to each other. By ‘rational’ in this instance we are referring to a specific type of rationality derived from rational choice theory (RCT), a form of modelling social and economic behaviour favoured by the neoclassical tradition of economics and increasingly prevalent in non-economic domains such as ethics. The use of rational choice theory is Gauthier’s way of cashing out what
21 an instrumental account of practical reason looks like. There follows a brief account of the central tenets of RCT.16
The aim of RCT is, “is to understand sociopolitical relations and institutions as the instruments created and used by mutually disinterested and rationally self-interested agents in the attempt to maximize the degree to which they can successfully pursue their particular ends and satisfy their particular preferences, whatever those might be.”17 It is characterised by its
1. Instrumentalism: This has two facets. It refers to the type of means-ends reasoning involved. The concern is not with how preferences are formed but how the agent acts so as to satisfy those preferences. In addition, methodologically, relations and institutions such as social norms and things like promises are viewed as the dependent variables when explaining social phenomena. These institutions and relations are to be explained by the independent variables: antecedently defined individual selves.
This has the implication that these selves cannot, prima facie, be regarded as social beings. To do so would be to presuppose that which RCT purports to explain. If it is true that we have a reason to cooperate this must be explained through the reasons of the independent variables: the antecedently defined selves. This brings us to the second tenet of RCT
2. Individualism: This refers to both the specification of individual action as the fundamental term in social explanation and a motivational theory about the kinds of reasons that rational individuals respond to that is sufficient to generate such explanations. The motivational theory posits that individuals are to be understood as rationally self-interested maximizers of utility. This ‘economic rationality’ is manifested by the structure of the agent’s preferences, which are defined by various axioms, namely:
16 The account provided follows Neal (1988), 636-639 17 Ibid., 637
22 • The axiom of completeness: For any two bundles of goods, X and Y, either X ≥ Y or Y ≥ X. This axiom assumes that the relevant bundles are comparable for the agent.
• The axiom of transitivity: For bundles X, Y, and Z, if X ≥ Y and Y ≥ Z then X ≥ Z
• The axiom of non-satiation: If bundle X* contains more of good x (and nothing else) than bundle X (which also contains only x) then X* will always be preferred to X
Furthermore, when explaining social phenomena, and ethical theories, all agents are aware of the fact that all the other agents they will interact with are also economically rational.18
3. Subjectivism: The ‘good’ for the agent is constituted by the content of his or her preferences. There is no external moral worth and only a ‘common good’ in the sense of a collection of individual goods.19
Gauthier in fact employs stricter starting assumptions than this account of RCT suggests. Whereas the above concerns the structure of the agent’s preferences, Gauthier additionally posits that the content of these preferences is non-tuistic. As Susan Dimock explains, the assumption of non-tuism “implies that the utility function of each individual, as a measure of her preferences, is strictly independent of the utility functions of those with whom she interacts.”20 It terms of Gauthier’s project this means that the agent is only concerned with the portion they will receive of the surplus generated by cooperating. So whatever the other agents get out of the hypothetical agreement has no impact, at least initially, on the utility of the agent in question. This is not, clearly, meant to be an accurate
18 This knowledge of the other agents’ rationality does not, of course, extend to the content of their preferences. 19 This last assumption might appear to stack the deck against any attempt to show that it is rational – in the RCT understanding of the word – to be moral from the outset. But a rational choice theorist like Gauthier would not object to the idea that the common good is merely the collection of individual goods. For the purposes of this chapter the pertinent question is whether Gauthier is successful in showing that this common good can motivate the individual to act cooperatively – i.e. adhere to the norms derived from the agreement – even when this seems to go against the individual good of the agent. 20 Dimock (1999), 252
23 reflection of how real people actually think. But it is necessary for Gauthier to begin from such assumptions since he wants to show that it is rational for an individual who has no reason ex ante the bargaining process to cooperate has such a reason ex post the agreement and will comply with the norms derived from this agreement.
To see why this is so it is helpful to consider Amartya Sen’ distinction between Sympathy and Commitment where the former concerns instances where our subjective well-being is directly affected by the knowledge of another’s well-being and the latter describes cases where our subjective well-being is not impacted, but we are nonetheless moved to help or cooperate with another.21
So, for example, Peter might not much care for animals. He has no pets, has never wanted one, doesn’t like being around them and in general has no positive or sentimental feelings towards them. In addition, Peter did eat several bacon sandwiches in his youth, an activity which gave him immense pleasure. Yet Peter has reasoned that, whatever his private feelings about animals and his appetite for bacon sandwiches, animals are sentient creatures with interests of their own, and therefore he is obligated not to harm them. Thus Peter is a strict vegan.
Compare the case of Peter with the case of Jessica. Jessica loves animals and can’t bear the thought of their suffering. The thought of eating meat makes her feel physically sick. Jessica is also a vegan.
The relative moral praiseworthiness of Peter and Jessica is not at issue here. Yet in Sen’s terms, Peter’s reasoning bears a closer resemblance to commitment, whereas Jessica’s reasoning can more plausibly be described as sympathetic. Given that an account of obligations must explain why we ought to do that which we might not feel sympathy for, it thus seems that a sense of obligation is more active in Peter’s case than in Jessica’s.22
The above example is used simply to draw attention to the fact that any study of moral obligation that is to be of interest has to feature an obligation that is somewhat difficult
21 Sen (1977) 22 A cynic, or an economist who subscribes to Revealed Preference Theory, might claim that Peter’s choice can be described in terms of his utility gained as a result of the satisfaction he gets from following his moral judgement. The problem with such an approach is that it seems to imply that any choice the individual makes can be re-described in terms of utility maximization. If this is the case then the RCT approach is in danger of becoming a vacuous explanation.
24 to keep. If fulfilling the obligation results in a net utility gain (or avoids a net utility loss) then this is not hugely informative about how the obligation binds us. For given the assumption of what agents have reason to do, if the agent gets a utility gain from avoiding meat regardless of the existence of an obligation to do so, then we cannot explain the bindingness of this obligation in terms of this account of practical reason. Gauthier recognises this, and thus adds non-tuistic starting assumptions to the traditional RCT picture of the agent’s motivation. As Dimock notes, “To argue that people ought to act morally because doing so will maximize their moral preferences would be a trivial result. And it would fail to show that we have sufficient non-moral reasons to act morally, which is what the contractarian wants to show.”23
*
Gauthier continues in the (neo-) Hobbesian tradition of seeking to explain how agents with such a preference structure can exit the state of nature and come to an agreement in order to improve the situation for all.24 ‘Exiting the state of nature’ here means avoiding a situation where no moral obligations hold. There is a moral free-for-all. He sets up the problem in the form of the classic Prisoner’s Dilemma game, as illustrated below.25
23 Dimock (1999), 257 24 I have included the parenthesised prefix ‘neo’, which might have been ‘quasi’, due to the fact that Hobbes was interested in the exit from conditions of social anarchy. Gauthier’s interest, and that of this thesis, is how we are to avoid moral anarchy. An avoidance of the former need not entail a positive account of moral obligation, which is the concern here. Such an alternative approach is advocated by Buchanan (1975) and, perhaps more controversially, by Hobbes himself (1991). See also Narveson (2010) on the distinction between Hobbesian and quasi-Hobbesian approaches, 509 25 Gauthier (1986), 79-8-2
25
Each agent is faced with a strategic choice.26 They can either cooperate or defect. The numbers represent the expected payoffs for each agent based on the collective decision of the two. They can be thought of here as representing the gains each agent achieves from the cooperation that is taken to be equivalent to acting according to the principles of the contract. Player A thinks about what player B will do and makes her choice accordingly. If player B cooperates then A’s best response is to defect since 5 > 3. If player B defects then player A’s best response is also to defect since 1 > 0. The payoffs are symmetric so player B’s best responses mirror player A’s. So for each player defect is the dominant strategy. This results in a stable ‘Nash equilibrium’ of ‘defect-defect’ where each player receives a payoff of 1. This equilibrium is stable but it is not Pareto optimal, since both players would be better off if they had cooperated.27 However, due to the rational choice assumptions outlined above, neither player has a reason to change their response from defect to cooperate. Each player can, at this point, only reason about their individual strategy, and thus can only move up or down or to the side. In order to move diagonally from ‘defect-defect’ to ‘cooperate-cooperate’ a joint-strategy would be required – that is, some sort of agreement to the effect that both will make this move.
26 A strategic choice is contrasted with a parametric choice. In the former case each agent incorporates their expectations of what the other will do. The latter case concerns instances where each agent takes his choice to be merely one variable among others. See Gauthier (1986), 23 27 ‘Pareto optimality’ refers to a situation where it is impossible to make any player better off without making another worse off. ‘Defect-Defect’ is Pareto inefficient since both players could be made better off at no extra cost
26 So far we have been treating this scenario as a one-shot prisoner’s dilemma game where the temptation to act selfishly, while hoping the other cooperates, would be greater than in a repeated game, where the players will have to continually interact as members of the same community. This means that those who become known as uncooperative are likely to be excluded from the benefits of cooperation. It is here that the idea of the agreement becomes salient and we can see how the prisoner’s dilemma goes to the heart of the problem that the contractualist enterprise endeavours to solve. If the players can reach an understanding whereby both (all) agree to engage in cooperation then each individual will be better off over the long run. Such an arrangement can be seen as an illustration of Rawls’s description of society as a ‘cooperative venture for mutual advantage’.28 Both (all) players would be better off if they (a) reached an agreement to cooperate and (b) trusted each other to adhere to such an agreement.29 Although the latter is the focus of this chapter a there follows a brief description of Gauthier’s solution to the former.
*
Given that each agent is aware that a joint strategy will yield a greater payoff for her than the individualistic approach outlined above, we assume that each has a reason to engage in the bargaining process. But given the structure of the motivation of each agent, and thus what they have reason to seek, Gauthier notes that,
“[O]f course each endeavours to have the joint strategy chosen that is most favourable to herself, minimizing the costs of her restraint and maximizing the benefits she receives from the restraint of others. In reaching agreement on a joint strategy, then, each individual sees herself engaged in a process of bargaining with her fellows. Through
28 Rawls (1971), 84. Rawls, like Gauthier, conceived of an account of practical reason as being consistent with the tenets of RCT. However, he was concerned with determining the optimal political institutions based on this account, not with individual moral conduct 29 In Gauthier’s terms (a) concerns ‘the internal rationality of cooperation’ – under what conditions is agreement rational? – whereas (b) concerns ‘the external rationality of cooperation: given that (a) is satisfied (agreement has been reached), it remains to be shown that it is rational to act cooperatively.
27 bargaining, individuals arrive at a basis for cooperative interaction that enables them to relate costs and benefits.”30
Here we come to the Gauthier version of the hypothetical bargaining process that characterises contractualist ethics. It is assumed that each ‘player’ has certain desires and needs such that cooperation with the other would be beneficial for them. It is also assumed that a cooperative process will yield a better situation for all involved. Furthermore, it is assumed that when entering the bargaining process each agent makes an initial claim which must, naturally, be greater than the non-cooperative outcome (which is taken to be the initial bargaining position). There will follow another round in the bargaining process where each will have to make a concession. It is argued that the agent with the lower relative concession must concede. Or, put another way, the person who has the smaller potential benefit is under less pressure to yield. If there is an optimal outcome requiring equal relative concessions from each person then the rational course of action is to select that outcome. In essence there is always a reason to move towards the point of equal relative concession, or as close as possible, rather than abandoning the bargaining process. The hypothetical agents have no reason to abandon the cooperative venture.
“Extending this rule to bargaining among several persons, we claim that the principle should state that given a range of outcomes, each of which requires concessions by some or all persons if it is to be selected, then an outcome be selected only if the greatest or maximum relative concession it requires, is as small as possible, or a minimum, that is, is no greater than the maximum relative concession required by every other outcome. We call this the principle of minimum-maximum, or minimax relative concession.”31
The principle of minimax relative concession constitutes Gauthier’s answer to the question of how rational agents are to reach an agreement, and the nature of that agreement. They reach such an agreement because given the RCT assumptions they always have a reason to seek cooperation, and the nature of the agreement will be the
30 Gauthier (1986), 128 31 Ibid., 137 The fact that it is a relative concession is designed to allow for agents with unequal endowments to enter the bargaining process.
28 one that is most beneficial to the widest number.32 However, as Gauthier acknowledges, the rationality of reaching such an agreement does not entail the rationality of keeping this agreement. Gauthier frames this as the need to provide an adequate response to the challenge posed by Hobbes’s Foole.33 Given that it has been shown that it is rational to agree to minimax relative concession, is compliance with this agreement rational? In terms of the prisoner’s dilemma above, if an agreement has been reached on the joint strategy cooperate-cooperate, why would the rational agent not then defect in order to gain the larger individual payoff that this would generate? It seems that the more profitable approach for the agent is to appear to be willing to agree to minimax relative concession and then to defect. In terms of moral contractualism, the account of practical reason under investigation here might explain the existence of moral requirements, but that same account of practical reason seems unlikely to be able to explain individual agent’s reasons for respecting these requirements.
The immediately obvious answer would be that, given that the moral and social world cannot plausibly be construed as a one-shot prisoner’s dilemma game, an agent with a propensity to defect would soon be identified as such and subsequently excluded. Hobbes’s Foole contends that in order for it to be rational to comply with the agreement the utility payoff at the time of the choice must be greater than that of defecting. In Gauthier’s terms, the Foole advocates straightforward maximization as the rational strategy. That is, comply with moral requirements when it is obviously beneficial to do so, but not otherwise. But such an approach would soon lead to exclusion for the individual or, if a majority reason in this way, to the breakdown of the cooperative venture. So, given this, Gauthier argues that the rational decision for the agent is to develop a conditional disposition towards constrained maximization.34 A constrained maximizer makes each decision based on the optimal outcome of the joint strategy, rather than their individual expected utility payoff. This disposition is conditional on the expectation that the agent is confronted with (mostly) fellow constrained maximizers, so that she is not played for a ‘sucker’.35
32 I leave aside complications such as individuals starting from different positions of power and influence 33 Hobbes (1991) 34 Gauthier (1986), 167 35 The need to develop a conditional disposition is based in part on the uncertainty surrounding the intentions of others. If intentions were transparent then straightforward maximizers would
29 Gauthier is keen to stress that it is a disposition, rather than a policy, that the rational agent ought to develop. This is to avoid the ‘sensible knave’ objection of Hume, which holds that while honesty might be the best general policy, there will be exceptions to this rule. While we might all benefit from the norm of truth-telling occasional situations might arise where an individual either finds the benefit of violating this norm outweighs the cost of exclusion from a particular venture, or is confident enough about not being caught that they view it as rational to defect, if only on this occasion. But if the individual reasons this way then he has not developed the rational disposition that Gauthier seeks:
“[C]onstrained maximization is not straightforward maximization in its most effective disguise. The constrained maximizer is not merely the person who, taking a larger view than her fellows, serves her overall interest by sacrificing the immediate benefits of ignoring cooperative arrangements in order to obtain the long-run benefits of being trusted by others. Such a person exhibits no real constraint. The constrained maximizer does not reason more effectively about how to maximize her utility, but reasons in a different way.”36
The difference between a adopting a policy and a disposition, in Gauthier’s terms, is that the latter involves an actual change in the kind of reasoning involved. It seems reasonable to say that such a change is necessary in order for the moral requirements derived from the contractual process to be binding. A person who generally keeps her agreements because of the reputational payoff from doing so, who is not averse to risking this payoff by the occasional transgression, might be thought of as generally reliable but it would seem more accurate to describe her fulfilling of her obligations as prudential rather than moral. If Gauthier can justify his claim that it is rational to develop this disposition then it would seem that he has provided a compelling explanation of why self-interested utility maximizers have a reason to comply with the moral requirements derived from the contract. But before we congratulate him on this, we need to take a deeper look at what
immediately be identified and excluded, so then it really would be irrational to be a straightforward maximizer. But Gauthier recognises that to assume transparency would rob the argument of much of its interest. So instead he assumes translucency of intentions, where over time those with a disposition to cooperate and those without will be identified and treated accordingly. Ibid., 174 36 Ibid., 169-70, c.f. 182
30 developing such a disposition involves, and whether it is in fact rational, on the RCT understanding of the what this entails, to do so. This is the topic of the next section
1.3 A Disposition to Comply
Given what we have said about the instrumentally rational agent’s reasoning, simply positing that such an agent will develop a disposition to cooperate is obviously not enough. We need to be clear on precisely what having such a disposition involves, and to do this we need to understand the nature of the motivation the constrained maximizer must have when faced with the choice of whether to cooperate or defect.
A useful framework for doing so is provided by Martin Hollis.37 A straightforward maximizer who has agreed to the principle of minimax relative concession can be understood to have a ‘first-order’ preference towards defecting once the agreement is operational. But an agent who acts on such a preference would soon find himself excluded from the agreement, and in a worse situation than had he cooperated. He thus constructs a second-order preference, where he cooperates for reasons that might best be understood in terms of deferred gratification: he realises that acting on the basis of utility maximization at each and every choice decision is counter-productive in the long run, so instead he chooses to cooperate.
It should be clear from the previous section that an agent who reasons in this manner cannot be said to have developed the disposition that Gauthier requires. Such an individual is still motivated by the expected utility of his individual strategy and not the optimality of the cooperative joint strategy. Having such a second-order level of preference cannot answer the sensible knave objection. Someone with such a policy who reasons that exclusion from certain activities is a price worth paying for not keeping his promises might well decide to abandon the norm of promise-keeping. Put another way, if Gauthier’s notion of the disposition amounted to nothing more than this second-order
37 Hollis (1993)
31 level of preference then he would not have shown that it is rational to fulfil one’s moral obligations.38
Hollis recognises this and thus introduces a third-order level of preference, one where the agent is no longer conceived as ‘Ulysses binding himself to the mast’.39 Instead she considers the moral norms derived from the agreement to be valuable for their own sake: she is concerned with the optimality of the group rather than her own utility. This is the type of preference that Gauthier is referring when he claims that,
“Only the person truly disposed to honesty and justice may expect fully to realize their benefits, for only such a person may rationally be admitted to those mutually beneficial arrangements—whether actual agreements or implicitly agreed practices—that rest on honesty and justice, on voluntary compliance.”40
An example might demonstrate the importance of this move from the second-order level of preference to the third order. A wealthy businesswoman, in a small town where most people know each other, frequently walks past a homeless man on the street who, it just so happens, cannot speak. She is not moved by his plight, that is, she has no first-order preference to give him some money. However, this is town with an implicit moral norm that one should help those less fortunate than oneself, so the woman is aware that such attitudes as those exhibited by her first-order level of preference would not only be frowned upon, but might lead to her exclusion from certain other activities, activities which she considers to be valuable – that is, she gains a utility payoff from being involved in these activities. Perhaps she will not be able to take part in joint business enterprises which will enhance her company’s reputation and profitability.
Given this, the woman has constructed a second-order preference to give some change regularly to the homeless man. It seems clear that her actions in this situation cannot be described as being morally motivated. In Gauthier’s terms she is still concerned only with the payoff of her individual strategy, rather than the utility of the homeless man or the optimality of the group. She is a sensible knave. It is likely that on a quiet day when no
38 Such an agent might in fact end up fulfilling all of his obligations. However, since Gauthier is attempting to show that it is rational to be moral, the agent would be doing so for the wrong reasons. 39 Ibid., 45 40 Gauthier (1986), 182
32 one is around to see her she will not donate, and the homeless person, being unable to speak, cannot inform the rest of the society about her duplicity. The important point is that this type of behaviour is entirely consistent with Gauthier’s development of a disposition towards morality if by this we mean only a movement from first to second- order preferences. It is clear that this will not suffice. The woman might be complying with the hypothetical agreement but her doing so is not morally motivated. That is, assuming that the hypothetical contract has produced a requirement to give money where one can to those in dire need, the woman complies with this conditionally on the longer- term expectation that there will be a net utility pay off for doing so. So it seems that in order for a rational agent to truly become disposed to comply with the moral requirements derived from the contract, Gauthier needs a third-order level of preference, one where the woman in the example will be moved to give the homeless man some money even if nobody is around to see her act of generosity.
Yet it must be remembered that the conditional nature of the disposition makes it important for Gauthier that this third-order preference is still defeasible. He does not want the constrained maximizer to be committed to cooperation regardless of the fact that she is surrounded by straightforward maximizers who will cheat her at every available opportunity. Likewise, the honest man who is surrounded by liars and promise-breakers must have the ability to, if not become a liar, at least not engage with these people in any enterprise where honest dealing is a prerequisite. If this was the result of Gauthier’s operation then the substantive results of this procedure – to be moral to the point of irrationality – would come into conflict with the formal aims of the starting premises – to reason instrumentally according to rational choice principles. He would have to argue that it is rational to be irrational. Gauthier’s aim is to show that it is rational to be moral so we cannot admit a result which dictates that one should be moral even to the point of it going against one’s rational interest.
*
It might be thought that I am stacking the deck against Gauthier here. If being a constrained maximizer means being a straightforward maximizer in disguise – that is, adopting a second order preference to comply – then we cannot say that the moral
33 requirements derived from the contract truly ‘apply’ to the agent concerned. But if they completely cease to reason rationally (in the manner defined above) then I accuse Gauthier of undermining his own project. Yet the textual evidence, and the nature of reasoning involved, do suggest something has to give. It might be that achieving the substantive aim Gauthier sets himself is simply inconsistent with the formal principles he begins with.41 But before reaching such a conclusion, we should investigate if there is another way to cash out this idea of developing a disposition to comply with the requirements of the contract
Let’s return to the possibility of the agent developing third-order level of preference to comply with the requirements derived from the contract. We can think of this not as an unthinking commitment to a particular doctrine, but rather an internalisation of the requirements themselves. So far we have only given a negative description of what such a preference would involve. In order to understand what such a preference is, rather than what it isn’t, we need to examine the relationship between making a moral judgement and the motivation to act on such a judgement.
It is at this point that it is necessary to introduce the metaethical debate between motivational internalists and externalists concerning reasons for action. This is a complex subject, with many different interpretations of the two positions, some of which will be mentioned below. Broadly however, it centres on what Michael Smith calls the ‘striking fact’ of the reliable connection between an agent recognising a moral fact, or making a moral judgement, and being motivated to act accordingly.42 Briefly, internalism claims that if an agent judges an action to be morally right, that is, they form a moral belief, then ceteris paribus this agent will be motivated to act on this belief: there is a necessary connection between morality and action. If a person agrees with the claim that it is morally wrong to eat meat but still regularly enjoys a fillet steak the internalist would say that such a person was either insincere in their agreement with the claim or that they are conceptually
41 Indeed, I think this is the case. But this is skipping ahead. 42 See Smith (1994), 71 and passim. More will be said on the connection between moral facts and moral judgements below.
34 confused. Smith, who argues for an internalist approach, would accuse them of being practically irrational ‘by her own lights’.43
The externalist about moral motivation holds that the forming of a moral judgement is insufficient to motivate action: a moral belief must be supplemented by a desire or some other affective state in order to motivate action: the connection between a moral judgment and action based on that judgement is only contingent. So our steak eater would not be practically irrational or conceptually confused. They could agree that eating meat is wrong but still do so provided that they do not have the sort of desire – perhaps to be a virtuous person – that compels them to act upon their moral judgement. The externalist appeals to the Humean idea that beliefs alone cannot motivate action; they must be complimented by some contingent desire. The internalist argues that moral beliefs have particular qualities that compel the agent in a way that, for example, an aesthetic judgement cannot.
The debate between internalists and externalists centres on what constitutes a reason for action. But there are many different types of reasons for action and not all of the literature uses a consistent taxonomy for these reasons.44 Bernard Williams distinguishes between internal and external reasons. Roughly speaking, the former take the form ‘A has a reason to ϕ’ while the latter take the form ‘There is a reason for A to ϕ’.45 Clearly the presence of an external reason does not entail the existence of an internal reason. Prima facie, an internal reason statement is falsified by the absence of some appropriate element of the agent’s motivational set S.46 An externalist about reasons for action might hold that there is a reason for Jane to read ‘War and Peace’ independent of what she thinks about the situation but unless there is some aspect of Jane’s set S such as the desire to be well read or to pass her exam she will not, according to the internalist
We can further distinguish internal reasons into normative reasons and motivating reasons.47 A motivating reason can be thought of simply as a link between desire and action. If I have
43 Ibid., 62. Alexander Miller criticises Smith for adapting a too strong version of internalism, one that seems to rule out the possibility of moral deliberation. See Miller (2003), 227 44 See, for example, Hieronymi (2011), 409-414 which provides a summary of the variety of terminology. 45 Williams (1981b), 101 46 Ibid., 102 47 This distinction follows Smith (1994), 131-132
35 a desire to eat the cake in front of me then I have a motivating reason for doing so.48 A normative reason concerns that which the agent values: for example the projects that she considers to be worth her time. While a motivating reason is necessary for action a normative reason can impact on a motivating reason: I might have a desire (a motivating reason) to leave the rubbish from my picnic on the grass when I have finished eating rather than search for a bin to put it in but I value (have a normative reason for wanting) the maintenance of the environment so I don’t do this. Likewise Jane might have a normative reason to read ‘War and Peace’. She values being a well-read person. If she does not have a subsequent motivating reason to do so then, according to Smith, she is practically irrational. Smith subscribes to a version of internalism which posits that if the rational agent has a normative reason for action then, ceteris paribus, she will be motivated to act on that which she values.49
External reasons are taken to be unrelated to motivation. For the purposes of the present argument we can think of external reasons as justificatory reasons. At this stage this is not to be understood as providing a (convincing) justification for the particular agent. It is rather that there exists a reason, independent of the agent’s inclinations, for the agent to engage in a certain activity. In a society where a knowledge of Russian literature is highly prized and even helpful in one’s career we might say that there is a justificatory reason for Jane to read ‘War and Peace’. But this does not imply that Jane has a normative reason to do so; the projects that are important to her might not include following society’s norms or having a well thought of career.
48 This is not to be confused with explanatory reasons, which might refer to psychological or neurological states which are viewed from a third-person perspective. A motivating reason refers to a first-person perspective. It can, in principle, be articulated: for example ‘I want to ϕ’. 49 The ceteris paribus clause is significant. For Smith this stands for ‘absent weaknesses of the will’ such as depression – where the agent is not incapable of being motivated by that which she values. Miller has taken Smith to task on this point, noting the danger of the connection between normative and motivating reasons collapsing into triviality – it seems to suggest that the agent will be motivated to act on her values except in cases where she isn’t. See Miller (2003), 221. However, for present purposes we can safely ignore this criticism. Since, as will be argued, the crucial connection for Gauthier to establish is not between normative and motivating reasons we can assume that the rational agents he is concerned with do not suffer from such weaknesses of the will. Thus, if the agent accepts that she has a normative reason for an action she will have a corresponding motivating reason.
36 The concept of the justificatory reason lends itself well to the situation Gauthier finds himself in when addressing the problem of compliance: moral requirements are defined by the results of the hypothetical agreement and thus constitute the justificatory reasons for the agent.50 From this perspective, there is an external/justificatory reason why the agent ought to comply with the agreement: the optimality of cooperation. But in order for the agent to be motivated by such a reason she must value the optimality of cooperation; that is, she must have a normative reason to comply with the agreement.
It should be becoming clear why the connections between internal and external reasons are important in understanding what the development of Gauthier’s disposition to cooperative involves. In order to move from a second to a third-order level of preference the rational agent’s normative reasoning must alter so that instead of her values corresponding to the RCT assumptions outlined in section 1.1, she values the optimality of the group. That is, in order for the agent to develop the disposition Gauthier requires for his argument to work, her normative reasons must correspond to the collective’s justificatory reasons. If Gauthier can appeal to existence internalism then this will be unproblematic since a moral fact, when recognised as such, would necessarily provide a normative reason to act accordingly.51 If, however, Gauthier cannot plausibly rely on such a connection, then he must turn to an externalist explanation of moral motivation, that is, he must appeal to a contingent desire in the agent’s motivational set to explain the commitment to the moral norms he hopes to derive. The next section will investigate whether Gauthier can rely on either account.
1.4 What Sort of Motivation?
50 That is, society provides justificatory reasons for agent. Whether or not she accepts these reasons has not yet been established. 51 Existence internalism is traditionally distinguished from judgement internalism. See Schueler (2010). The former relates to moral ‘facts’ while the latter concerns moral judgements. Which position one might advocate depends in part on one’s attitude to moral realism. Emotivists deny that there are such things as moral facts or even beliefs; a moral statement is always an expression of a desire, thus judgement internalism is trivially true. For present purposes it is assumed that moral facts do exist and are represented by the terms of the agreement. It is further assumed that the rational agent is capable of recognising aspects of the agreement that apply to the particular choice situation she is facing – that is, she is capable of recognising moral facts. The impact of such recognition (or judgment) on her motivation is still to be investigated.
37 The previous section introduced existence internalism as the claim that there exists a necessary connection between moral facts and normative reasons. Moral ‘facts’ are to be equated with the contents of the hypothetical agreement, in Gauthier’s case these facts will have been arrived through the principle of minimax relative concession. It has also been assumed that that the rational agents Gauthier is concerned with will not be mistaken about these facts: they will recognise which aspects of their lives the agreement is relevant to. So an agent’s ‘moral’ beliefs correspond to the moral facts of the agreement. In addition, they will not be prone to weaknesses of the will: if they accept a claim as a normative reason then they will be motivated to act upon this reason. Existence internalism requires going against the basic Humean intuition that a belief cannot motivate without a contingent desire to act on that belief. So it must claim that moral facts have unique properties which enable them to motivate action without the need for such a desire. Gauthier equates moral facts with the content of the agreement so if he is to rely on existence internalism he has to show that the agent’s beliefs about the nature of the agreement have such unique properties. How can he do this?
He might claim that these unique properties concern the fact that the rational agent now forms his beliefs as a ‘we’ rather than an ‘I’ because she now reasons as part of a joint strategy rather than an individual strategy. But this needs to be justified. Recall that Gauthier needs to show that it would be rational for the agent to develop a disposition to reason as a ‘we’ rather than an ‘I’ – that is, he has to show why the agent will value the joint strategy, and thus optimality, ex post the agreement. However, given Gauthier’s acknowledgement that Hobbes’s Foole must be answered, it is difficult to see how Gauthier can do this with an appeal to existence internalism
We can summarise the problem Gauthier faces thus:
1. The optimality of the contract does not, prima facie, provide a sufficient reason for the agent to comply with the agreement.
And
2. The RCT preference structure, and non-tuistic content of these preferences can, at best, explain how the agent might construct a what section 2.1 defined as a second-order level of preference to comply.
38 But
3. This type of reasoning is not the sort Gauthier is trying to derive. It is prudential rather than moral reasoning. In additional it does not seem sufficient to guarantee the stability of the cooperative venture.
So
4. Gauthier needs to show that the rational agent will move to a third-order level of preference – that is, they will develop a disposition such that they value the optimality of the joint strategy rather than the utility payoff of the individual strategy
Regarding internalism
5. To appeal to existence internalism would be to claim that there is a necessary connection between society’s justifying reasons (as represented by the contents of the agreement) and the agent’s normative reasons
6. If this were true then the disposition Gauthier seeks would be generated by the agreement itself
However
7. We have seen from 1) that the optimality of the contract is insufficient to ensure compliance.
Existence internalism posits that an agent who recognises a moral fact will be disposed to act accordingly. But since moral facts in this instance correspond to the contents of the agreement – contents which take their form as a result of a bargaining process determined by agents with the preference structure outlined above – recognising a moral fact is not in itself sufficient to explain why the agent would now have a different kind of normative relation to the contents of the agreement which their RCT motivation. Simply positing the disposition Gauthier needs tells us nothing about how this disposition is generated. And 1) – 7) above shows that it cannot be generated by the agreement itself.
39 In addition, it seems trivial to point out that Gauthier cannot rely on an existence internalism that holds ex ante the bargaining process. It is taken that there are no moral facts to speak of at this stage; agents are still in the state of nature, where this is defined by the absence of morality. So we cannot assume a connection between the motivation of the agent and a moral fact which hasn’t yet come into existence. Furthermore, to posit that this link between judgement and motivation emerges ex post the agreement would contradict the central point of internalism: that a moral fact necessarily motivated an agent who judges it as such.
*
So it seems clear that Gauthier cannot appeal to an internalist explanation of the connection the external/justificatory reasons generated by the agreement and the normative reasons of the individual agent. We now consider the possibility of an appeal to an externalist account of moral motivation. Recall that the externalist denies the necessary connection between a moral fact being recognised by the agent and the agent’s being motivated to act accordingly and instead argues that this judgement must be accompanied by a contingent desire. If Gauthier is going to appeal to this type of explanation we must discover what kind of contingent desire could fulfil the role he requires, and whether his account gives us a reason to suppose that such a desire could be active.
One possibility is that which Smith considered in his argument against externalism. Smith criticises the externalist for the implication that the agent is motivated to ‘do the right thing’ where this is read de dicto rather than de re.52 A de dicto reading of the sentence ‘I am going to give a prize to the best student in the class’ would imply that the person uttering the sentence is not yet aware of which student this might be. A de re reading would suggest that the person speaking knows the identity of the student (Jennifer) and is referring to her (knowingly) as ‘the best student in the class’. In the moral case ‘do the right thing’ when read de dicto would mean that I have a desire to do the ‘right thing’, whatever that
52 Smith (1994), 73
40 might turn out to be. So the recognition that eating meat is wrong must be supplemented by the desire to not do wrong. A de re reading of the same situation would indicate that I do not eat meat because I consider it wrong,53 I do not need an additional desire to ‘do the right thing’ because I am directly concerned with the issue at hand (in this case the well- being of animals).54
The plausibility of Smith’s attack need not detain us. At issue here is whether Gauthier can appeal to an externalist account of moral motivation, so we need to see whether a de dicto desire to ‘do the right thing’ or to‘act justly’ can play the role of the contingent desire that Gauthier needs to make an externalist explanation work. A plausible de dicto reading of ‘doing right’ in Gauthier’s case might be:
A. Defecting from the agreement when dealing with a constrained maximizer is wrong
B. I do not want to be the type of agent who does wrong
C. Therefore, I will endeavour to develop a disposition not to do wrong to those who are willing to cooperate with me
But it is difficult to see where such a desire as B could arise. Ex ante the agreement the agent is taken to be motivated by the RCT assumptions outlined in section 1.1. As was seen there, the agent has no standing commitment to the common good. So a desire to ‘do the right thing’, whether this is read de dicto or de re, would not make sense. Nor can we appeal to an antecedent desire to be a cooperative agent since the only thing of value at this stage is based on the agent’s own subjective preferences.
So it seems that the motivation for B must come from some part of the bargaining process or the results of this process. But, as we know, the agent’s reason for entering the bargaining process is in order to gain from the surplus generated by cooperation. This is
53 A de re reading naturally lends itself to an internalist explanation of moral motivation. 54 Smith is adapting the ‘one thought too many’ argument outlined by Williams (1981c). The claim here is that the man who deliberates on what is ‘the right thing to do’ when deciding whether to either save his drowning wife or a similarly drowning stranger has ‘one thought too many’; he does not have the appropriate direct concern for her. Smith extends this thought to attack externalism more generally. He thinks that the agent who consults the de dicto reading of ‘what is the right thing to do’ is guilty of fetishizing morality rather than exhibiting the appropriate direct concern for the person or situation under consideration.
41 the reasoning that underpins the principle of minimax relative concession: there is always a (utility-based) reason to make a (relative) concession rather than walk away. So the agent’s desire to reach an agreement is still motivated by individual utility considerations.
Another possibility is that it is the nature of the agreement itself that establishes B. But the facts of the agreement can only inform the agent’s moral beliefs. The agent now knows what is expected of her thanks to the agreement. She understands moral ‘facts’ such as the fact that breaking promises is wrong, in the sense that she realises that it transgresses joint-strategy optimality to do so. But this is only a claim about beliefs, and thus on the externalist account of moral motivation is insufficient for compliance. To claim that the rational agent’s reasoning goes “X is wrong, therefore I do not wwant to X’ is simply existence internalism restated. To say ‘X is not optimal, I do not want to do act in a non-optimal manner, therefore I do not want to X’ is just another version of A to C above. We still cannot explain where the desire not to do wrong comes from.
So it seems that the agent can understand optimality, she sees that she benefits from optimality, and is aware of an external justificatory reason not to defect – that is, she understands A. But we have no reason to presume that there exists a commitment to morality based on the agent valuing optimality.55
Since it does not appear to be possible for Gauthier to appeal to a contingent desire which takes the form of the agent valuing norms for their own sake it looks like the only alternative is to explain this desire with reference to the RCT assumptions about the agent’s motivation noted in section above.
In doing this, and supposing that the agent is able to take the long view and will himself to develop a disposition towards compliance, it does not seem that Gauthier can counter Derek Parfit’s objection that while it might indeed be rational to develop such a disposition, given that the agent’s normative values are determined by the RCT account of motivation, the development of this disposition does not entail that it is rational to act on this disposition at the time of choice.56 To this we might add the question of why any
55 An agent who values fairness might seem appealing in this instance since it would make sense of the conditional nature of the disposition that Gauthier wants to derive – as a desire to cooperate with like-minded agents and punish free-riders. But the emergence of this desire can be explained no more easily than the valuing of optimality. 56 Parfit (2011), 433-444
42 rational agent who is aware of the way other rational agents reason, and assumes that these make up the preponderance of the agents she will interact with, would reject cooperation with straightforward maximizers for utility-based reasons yet cooperate with other CMs on the same grounds. There is no way, on this reading, to distinguish between a second and third-order level of preference. Hume’s sensible knave objection has not been answered.
It seems that there is no contingent desire that Gauthier can appeal to in order to explain the movement from second to third-order levels of preference. The only answer to the question why be moral he can supply, if he draws on an external account of moral motivation, ends up looking like a type of prudential reasoning.
1.5 Conclusion
It appears that neither motivational internalism nor externalism can be rendered consistent with Gauthier’s claim that the rational agent will develop a disposition to adhere to moral norms. To assume an internalist approach would be to contradict Gauthier’s own acknowledgement that Hobbes’s Foole requires an answer: that establishing the optimality of the agreement is insufficient to motivated compliance. And the only plausible desire Gauthier can appeal to when drawing on an externalist account of moral motivation ends up collapsing back into the type of instrumental individualistic reasoning that Gauthier is seeking to escape. Of course, such an account raises the question of why we should assume that the agent will cheat the agreement. But the point is that the rational choice assumptions that Gauthier deploys can provide us with no confidence that the rational agent will not cheat.
Gauthier’s argument is very useful because it elaborates the conditions that need to be met in order for the agent’s narrow rational reasoning to be transformed into the kind of reasoning required for a moral obligation to determine the agent’s choice. It also provides a very helpful illustration of the division between the reasons agents might have to enter into a hypothetical agreement and the reasons they have to live according to the principles derived from such an agreement. It has been argued here that this division cannot be bridged when relying on the rational choice assumptions that Gauthier relies on. If
43 neither of the main accounts of moral motivation can be rendered consistent with a rational choice framework then it seems that we must accept either that the only reason an agent has to make good on his commitments is prudential, or we need to abandon the rational choice framework when dealing with the question of normativity in contractualist ethics.
44 Chapter 2: Scanlonian Contractualism
This chapter looks at Scanlon’s account of contractualism and in particular his answer to the question of normativity (or the ‘question of motivation’ as he describes it). It argues that this must be understand through his account of the way that the objective reasons that agents have prior to the hypothetical bargaining process are mediated through this process thus issuing moral requirements. An understanding of the answer to the question of normativity necessitates an understanding of what Scanlon calls the ‘question of subject matter’. Having established this connection, I turn to an objection against this question of subject matter. I argue that Scanlon cannot provide a compelling answer to this objection without sacrificing his ethical theory’s ambitions to provide an account of morality’s foundations. Scanlon’s answer to the question of normativity is thus found to be insufficient.
2.1 Introduction
The previous chapter considered the viability of a contractualist ethics that relies on an instrumental account of practical reason. It did so through a close reading and criticism of David Gauthier’s argument in Morals by Agreement. It was argued that the solution offered by Gauthier to the question of normativity was unconvincing. This chapter will consider a quite different form of contractualist ethics, which relies on a distinct account of practical reason. This is Thomas Scanlon’s contractualism, as developed initially in his 1982 article ‘Contractualism and Utilitarianism’ and developed in more detail in his 1998 book What We Owe to Each Other.57
The literature often refers to the Gauthier’s approach to contractualism and Scanlon’s as Hobbesian contractualism and Kantian contractualism respectively.58 This is encouraged by the respective authors’ acknowledgement of their intellectual lineage.59 It is also true that this characterisation makes a certain amount of intuitive sense given the two authors’ accounts of reason as being determined primarily by individual self-interest (Gauthier)
57 Scanlon (1982) and Scanlon (1998) 58 See, for example, Darwall (2003) and Southwood (2010) 59 Scanlon (1998) p. 5 and Gauthier 1986 passim
45 and the need to live in ways that are justifiable to others (Scanlon). Nonetheless, the account of practical reason that can be most plausibly ascribed to Scanlon seems rather un-Kantian in many respects. As I will argue below, the most plausible way to read Scanlon’s account of practical reason is one that can be best described as rooting our reason for complying with the demands of the contract in the substantive value of living in conditions of mutual recognition with others. This is supported not only by the textual evidence that this is indeed what Scanlon had in mind, but also due to the fact that the alternative appears to suggest that we end up with implausible answers to the question of subject matter in his moral theory. As we shall see, however, such an account leads Scanlon into methodological difficulties when attempting to answer the question of normativity.
I will proceed by first providing an overview of Scanlon’s account of reasons. This is important to understand given his presentation of the hypothetical contract as a means of mediating between the substantive reasons that agents have (conceptually) prior to this process. I will then turn to the details of his contractualist ethics, paying special attention to the connection between the ‘question of subject matter’ and the ‘question of motivation’. I will claim that, somewhat in contrast to Gauthier, a firm grasp on the former is essential to a firm grasp on the latter.60 This will have implications for my critical move against Scanlon. I will consider whether the contractualism outlined in the previous sections can avoid an objectionable relativism in our moral requirements. I argue that it can’t, and that while a solution can be offered through appeal to the substantive value of living in conditions of mutual recognition – and there is textual evidence that Scanlon endorsed such a move – this comes at the cost of providing a non-redundant answer to the question of normativity.
2.2 Scanlon’s Contractualism – An Overview
60 I don’t mean to suggest that there is no connection between the two in Gauthier. Clearly, an understand of the account of practical reason Gauthier invokes is crucial in accounting for why an act is wrong. But it is easier to separate disputes about whether a particular act is wrong in Gauthier’s case than it is in Scanlon’s. This point should be clearer by the end of the chapter.
46 The primary distinction between the kind of contractualism advocated by Scanlon and that advocated by Gauthier lies in the fact that it is not assumed that the agents who reach a hypothetical agreement are motivated to do so by the type of self-interest outlined in the previous chapter. Rather, Scanlon emphasises the motivation as being one defined by the search to find ways to live that are justifiable to others. Quite what this motivation amounts to, and the acceptability of this kind of stipulation, will be discussed below. First, I will present a more general overview of Scanlon’s contractualist ethics.
According to Scanlon’s contractualism, “an act is wrong if its performance under the circumstances would be disallowed by a set of principles for the general regulation of behavior that no one could reasonably reject as a basis for informed, unforced general agreement.”61 In part, the question of whether a principle can be reasonably rejected depends on the motivation of finding such principles, and he explicitly disavows any such motivation being primarily about self-interest, but rather stressed the importance of justifiability to others.62 This formulation actually contains the answer to the two questions that Scanlon takes to be crucial for any moral theory to answer.
The first of these is the ‘question of subject matter’. As he puts it,
“In moral judgements, as in mathematical ones, we have a set of putatively objective beliefs in which we are inclined to invest a certain degree of confidence and importance. Yet on reflection it is not at all obvious what, if anything, these judgements are about, in virtue of which some can be said to be correct or defensible and others not. The question of subject matter, or the grounds of truth, is the first philosophical question in both morality and mathematics.”63
That is, the question of subject matter can be understood as addressing what is it that makes a moral judgement true or false. Suppose that a principle is proposed which dictates access to certain civic rights on the basis of sexual orientation. The question of subject matter focuses on what it is for such a proposal, or more accurately a principle licensing such a proposal, to be morally objectionable. Scanlon’s claim is that the contractualist formula provides the best account of why this is so. Such a principle is
61 Scanlon (1998) 153 62 Ibid., 5 63 Scanlon (1982) 125
47 objectionable not because of its impact on the total welfare of society, nor because it would be rationally rejected by any self-interested person choosing from behind a veil of ignorance. Nor is it solely because it imposes a burden on those who lose out on civic rights as a result of their sexual orientation, but rather because the principle(s) that ground such a proposal are not suitable “to serve as a basis of mutual recognition and accommodation.”64 The proposal to limit the civic rights of people with a certain sexual orientation would be wrong, and it would be wrong in virtue of the fact that a person affected by this limitation in their rights could reasonably object to the principle which licenses the limitation of these rights. Wrongness, on this account, is inconsistency with principles that could not reasonably be rejected.65
The second question concerns the normativity of moral judgements. Given a certain answer to the question of subject matter, it remains to be seen what reason the agents the moral judgement affects has to care about and act on the judgements. Although Scanlon refers to this matter as the ‘question of motivation’, this has the potential to be misleading, since it might suggest that the crucial question is how moral facts can ‘get a grip’66 on the agents’ psychology. Scanlon explains his approach to the question:
“[W]hat must an adequate philosophical theory of morality say about moral motivation? It need not, I think, show that the moral truth gives anyone who knows it a reason to act which appeals to that person’s present desires or to the advancement of his or her interests. I find it entirely intelligible that moral requirements might correctly apply to a person even though that person has no reason of either of these kinds for complying with it… what an adequate moral philosophy must do, I think, is to make clearer to us the nature of the reasons that morality does provide, at least to those who are concerned with it… It must make it understandable why moral reasons are ones that people can take seriously, and why they strike those who are moved by them as reasons of a special stringency and inescapability.”67
64 Scanlon (1998) 194 65 My thanks to Tom Smith and Richard Child for pressing me to clarify this point 66 In Korsgaard’s phrase (1996) 67 Scanlon (1982), 126-7
48 When Scanlon claims that moral requirements ‘might correctly apply’ to a person independent of their desires and interests, this entails that the person in question has a normative reason to comply with moral facts – specifically to avoid actions that are determined as wrong by the contractualist formula. So when Scanlon speaks of the question of motivation, he is not referring to how a particular agent comes to be motivated to comply with moral requirements, but why the reasons morality provides are so important. As Pamela Hieronymi puts it, “why avoid morally wrongful action as such?”68 Why does the fact that a given action is morally prohibited provide an agent with a reason to avoid this action in virtue of its being morally prohibited?
In order to make sense of the claim that moral requirements ‘apply’ to agents irrespective of their desires it follows that it must be possible for the agent to ‘have a reason’ which does not depend on their subjective psychological states. While this point is implicit in Scanlon’s 1982 article, he acknowledges in his later What We Owe to Each Other that he hadn’t, at that point, fully appreciated the importance of this to his account of the normativity of moral requirements. On the motivating force of seeking to act in ways that are justifiable to others, he notes:
“Many people pressed me to say whether, on my view, a person who lacked this desire would have any reason to avoid acting wrongly, and to explain how I would account for the fact that lacking this desire is a particularly serious fault. In addition, it became clear that the accounts I wanted to offer of the structure of reasoning about right and wrong, and of the relation between this part of morality and other values, were much more naturally put in terms of reasons. It was very difficult, perhaps even impossible, to present these accounts adequately within a conception of practical reasoning that took it to be a matter of figuring out how to fulfil various desires and how to balance these desires against one another.”69
By making reasons, rather than desires, the central part of his account of the normativity of moral requirements, Scanlon’s aim is to provide a foundation for the idea that moral requirements are ‘inescapable’ – that they do not cease to apply to us if we have no desire to comply with them. In making this move he deploys the further assumption that
68 Hieronymi (2012) 104 69 Scanlon (1998) 7
49 reasons are practically basic, or fundamental – that is, they are not explainable in terms of other concepts. “Any attempt to explain what it is to be a reason for something seems to me to lead back to the same idea: a consideration that counts it favour of it. ‘Counts in favour how?’ one might ask. “By providing a reason for it’ seems to be the only answer.”70
It should be noted that in denoting reasons as practically basic Scanlon is not suggesting that there is nothing more that can be said about the reasons we have. The point here is that we, as individuals, have a set of considerations that might or might not constitute reasons for us. It might be that we are mistaken about the considerations that provide us with a reason. Or it might be that one reason is outweighed by another such that a pro tanto reason turns out to not be a sufficient reason for a course of action. Scanlon’s contractualist ethics is premised on the idea that individuals already have reasons but what makes the considerations that count in favour of some course of action a moral requirement is that these reasons that individuals have could be passed through the contractual process.
Treating reasons, rather than desires, as practically basic is not only necessary in order to argue for moral requirements being inescapable. It also seems to better explain the phenomenological experience of, for instance, why we feel we ought to tell a friend some unwanted news.71 To say that we ‘desire’ to do so seems odd, unless we treat desire as merely a component part of action. If that was the case then ‘desire’ would be part of an explanatory account of the agent’s motivating reasons – some psychological fact about the agent. And yet, a more substantive account of desire does not seem to accord with why we ought to tell our friend the unwanted news. It seems plausible to suggest that there is nothing desirable about doing so. However, it does seem that we have a reason to tell the friend the bad news independent of our desires concerning the matter. Scanlon does not deny that desire can play a role in an agent’s reasoning, but thinks it is more accurate to say that we desire something because we have a reason to want it than to say we have a reason to want something because we desire it.
70 Ibid., 17 71 Ibid., 39
50 But this emphasis on the reasons that agents have, and the phenomenological plausibility of this account, should not mislead us into thinking that the subjective states of the agents concerned are the determining factor of what we owe to each other. It is crucial to Scanlon’s contractualism that someone can have a reason without acknowledging that they have a reason. Thus he distinguishes between ‘operative reasons’ – the reasons that the agent takes themselves to have in favour of adopting an attitude or performing an action – from ‘reasons in the standard normative sense’ – the reasons that the agent actually has.72 To borrow an example from Hieronymi, Erin might believe that the meeting is over, and therefore she has an operative reason to leave the room. But the meeting is not over, she was mistaken in her belief. That is, she did not in fact have a ‘reason in the standard, normative sense’ to leave the room because the proposition that her belief was based on – ‘the meeting is over’ – turned out to be false.
What this example shows is that an agent can have a reason to adopt an attitude or course of action even if there is nothing in their subjective state at the time which motivates them to do so. This does not demonstrate that there is no connection between normative reasons and the subjective states of the agent. It might be the case that normative reasons are given by the desires that the agent would have if they were fully informed. Returning to the previous example, presumably there is nothing in Erin’s psychology, that corresponds to a preference for leaving meetings early, so were she to have access to the correct information, then she would not form the intention to leave the room at that time.73
Such an account is still ‘subjective’ in the sense that it is still ultimately determined by a connection to the agent’s internal psychological states. By contrast, an objectivist about reasons at least allows for the possibility that we can have reason that does not have this
72 Ibid., 19. Henceforth I will refer to ‘reasons in the standard normative sense’ as ‘normative reasons’ 73 This is essentially the Williams (1981) account of what it is to have a normative reason. According to this account, an agent has a reason to ϕ only if ϕ-ing would advance some element in the agent’s ‘subjective motivational set’. The agent might not be aware of this reason, but provided there is a ‘sound deliberative route’ from their subjective motivational set to becoming motivated to adopt the corresponding intention they can still be said to have a reason. Parfit thinks that Williams’s conception of normativity is ‘too weak’ (2006, 334). I will not dwell on this here, but the important point is that this conception of normative reasons is dependent on subjective conditions – that is, the agent’s motivational states. As will be explained below, Scanlon’s is not.
51 direct connection to our subjective states. Sticking with the example above, Erin might be rather blasé about meeting etiquette, but the cultural setting in which she is situated might have strong prohibitions on leaving meetings before the chair signals the end of the meeting. Here we might say that Erin has an objective normative reason to remain in her seat, regardless of her desires or the information she has about what stage the present meeting is at. Or, to take a moral example, the question of whether it is true that I should pay Sylvia back the money I owe her might be answered independently of how I feel about the proposition.74 As should be clear from the discussion above, Scanlon is an objectivist about reasons; he holds that the reasons that we have are independent of our subjective states. This is true of our non-moral reasons, and it is true to an extent about our moral reasons too. But there is a complication here. While Scanlon is a realist about reasons he is a constructivist about ethics. Our moral reasons do apply us independent of our subjective feelings about the situation. But these are determined by the contractual process – a constructivist method. What determines the moral reasons we have, i.e. what determines whether it is true that I ought to pay Sylvia back the money I borrowed from her, is determined by the hypothetical contract. A more precise account of how this works will be the topic of the next section
2.3 Scanlon’s Constructivism
Scanlon takes a realist stance on reasons for action but a constructivist stance on morality. He holds that agents have reasons and these are not entities that are capable of being broken down into more fundamental parts. Furthermore, he is an objectivist in the sense that the question of whether a purported reason statement such as ‘I ought to take an umbrella out’ does not necessarily depend on my subjective states. Yet one can agree with Scanlon in all of this without subscribing to his constructivist ethical theory. A moral account is constructivist when the normative truths that are derived from it are determined by some idealised process of deliberation or agreement. Ronald Milo defines contractarian constructivism thus as:
74 I borrow this example from Thompson (2004) and will return to it, and Thompson’s framework for thinking about the question of normativity, in chapter 4.
52 “hold[ing] that moral truths are most plausibly construed as truths about an ideal social order, rather than the natural (or some curious nonnatural) order of things. It is true (or is a fact) that a certain kind of act is wrong, for example, just in case a social order prohibiting such acts would be chosen by rational contractors under suitably idealized conditions.”75
Such an account will be familiar from the discussion of Gauthier in chapter 1. The contractors are ‘ideal’ in the sense that they are (narrowly construed) rational agents operating in specified conditions. Scanlon’s contractualist theory is also constructivist in the sense that our moral reasons are determined by facts about what would be agreed to in certain hypothetical circumstances. But rather than the idealisation being focused on the kind of rationality Gauthier appeals to, Scanlon’s agents are committed to finding justifiable principles for the general regulation of behaviour. And this is determined by the relative strength and nature of the reasons that the hypothetical contractors have.
So, to return to the earlier example. Say I have borrowed some money from Sylvia and there is some debate about whether I should pay this back or not. We might say I have a prima facie reason to avoid paying the money back. Sylvia has a reason to demand the money back from me. We might say that the statement ‘Fred ought (all things considered) to pay Sylvia back the money he borrowed’ is true or false depending on the relative strength of the reasons that we each have. According to Scanlon’s contractualism, the determination of whether this statement is true or false lies in the question of whether a principle licensing Fred not paying the money back could be reasonably rejected by agents seeking principles for the general regulation of behaviour.76 Assume that it is the case that I ought (all things considered) to pay Sylvia back the money I borrowed. What makes this the case is that a principle licensing my not paying back the money could be
75 Milo (1995) 184 76 It might seem unduly messy to present the principle in this way. I present it as such because Scanlon presents his contractualism first and foremost as a study of the property of moral wrongness. Whether it is right to perform an action can be ambiguous between it being merely permissible or in fact obligatory, whereas if an act is wrong it is simply impermissible. Of course, contractualist ethics has plenty to say about our positive responsibilities too, but if we are strictly applying the formula then must interpret this as it being ‘wrong not to ϕ’, rather than it being ‘not wrong to ϕ’. In Scanlon’s case, the first would be cashed out as ‘any principle permitting the avoidance of ϕ-ing could be reasonably rejected…’, whereas the second would be cashed out as “any principle prohibiting ϕ-ing could be reasonably rejected…’
53 reasonably rejected by a set of hypothetical contractors who were working towards a set of principles for the general regulation of behaviour.
Assuming that it is true that I ought to pay Sylvia back the money I have borrowed, and that this true in virtue of the factor that a principle licensing my not paying the money back could be reasonably rejected in the contractual process, then this applies to me regardless of my feelings about the matter. This might seem puzzling, since the determination of the principles depends on the reasonableness of the various objections that might be levelled at them, and this reasonableness in turn depends on the reasons that agents actually have. Given that I (Fred) have a reason not to pay Sylvia back the money I owe her, we might ask what it is about the contractual process that determines that I am wrong about this and Sylvia is right to demand her money back
It is here that we must introduce Scanlon’s concept of ‘generic reasons’.77 As discussed above, the contractual process in Scanlon is the method through which moral principles are determined, and this determination is based on whether or not said principles could be reasonably rejected by agents committed to finding justifiable ways to live with one another. Whether or not a principle is reasonably rejectable or not depends, in turn, on the reasons that real agents have. But we have yet to address the standpoint from which the relative strength of the reasons put forward as a basis for rejection of a principle would be assessed. This is the standpoint from which the abstractor contractors deliberate on the principles to govern moral life.
This is where ‘generic reasons’ come in. My reason not to pay Sylvia back the money I owe her might be down to my selfishness and/or poverty. But I am not present in the contractual process, so it is highly unlikely that these reasons are going to constitute a reasonable rejection for the principle of paying the money one owes back on time. Rather, a representative of people in my situation might put forward a reason for leniency when returning debts owed, and a representative of people in Sylvia’s situation might but forward a reason to adopt a stricter principle for repayment. These reasons are ‘generic’ in that they reflect the reasons people have in the typical situations they find themselves in. As Scanlon puts it:
77 Scanlon (1998) 202-206
54 “Since we cannot know… which particular indiv iduals will be affected by it in which ways… our assessment cannot be based on the particular aims, preferences, and other characteristics of specific individuals. We rely instead on the commonly available information about what people have reason to want. I will refer to this as information about generic reasons… Generic reasons are reasons that we can see that people have in virtue of their situation, characterized in general terms, and such things as their aims and capabilities and the conditions in which they are placed.”78
These reasons are ‘generic’ in the sense that they arise out of the general circumstances of the situation. They need not be the reasons that the majority of people within a given society have. They might be reasons that a small subset of the society have in virtue of some features that the small subset share that the rest of society do not.
The upshot of this is that in the case of the borrowed money it is the reasons that generic lenders and borrowers have to endorse or reject a certain principle regarding repayment. The relative importance of each reason or claim, and whether an objection to the adoption of a particular principle is reasonable or not, is determined by the impact it has on the generic versions of the people concerned. Not receiving repayment in a timely fashion might cause stress for the individual, it might render them unable to fulfil other obligations and plans. These might be thought to outweigh the financial insecurity typical borrowers have in the time and place concerned.79 It is important to note that the factors that go into the adoption of the principle and whether or not an objection to that principle is reasonable are not exhausted by considerations of well-being. While a good proportion of the relevant factors concern the sharing of the burden of adopting these principles, Scanlon also allows considerations of fairness and other moral considerations to enter into the process.
78 Scanlon (1998) 204. The fact that the reasons are determined by the general culture agents find themselves in will be an important factor in my argument against Scanlon later in this chapter. 79 I am assuming, for the sake of argument, that the contractual process will determine that the money ought to be repaid in a timely fashion – i.e. that it would be wrong for me not to do so. But I am not committed to this. It might be the case that a more lenient principle is adopted. This will depend on the prevailing social circumstances of the time and place and thus the generic reasons that are brought to the process. Scanlon is aware and to a degree endorses the relativist implications of this in (1998 328-361). I agree that this malleability is an attractive feature of his contractualism. However, as I argue below, it has the potential to create difficulties for him when combined with the kind of account of practical reason he relies on.
55 It might be thought that such an approach opens Scanlon up to a very obvious objection, and it is indeed an objection that has frequently been levelled at Scanlon. This objection tends to take one of two forms, both of which focus on the fact that the idea of ‘reasonable rejectability’ has moral content.80 This is most obviously the case in situations where the determination of what is reasonably rejectable hinges on a conception of the unfair burdens being placed on one of the parties.
Sometimes this objection is referred to as the ‘circularity objection’, sometimes the ‘redundancy objection’. The difference between the two is more one of emphasis rather than anything substantive. Both focus on the issues that seem to arise from the fact that ‘reasonable rejection’ contains moral content, but the circularity objection concentrates on the lack of explanatory adequacy purportedly involved in such an approach, since we appear to end up explaining our fundamental moral reasons through an appeal to concepts like fairness which contain moral reasons within them. Scanlon’s contractualism is, on this reading, methodologically incoherent.
The redundancy objection makes a similar point, but rather than the focus being on what we might think of as the formal methodological issues with doing this, the emphasis is on the fact that it seems that there is another sort of ethics involved. The idea is that whatever uses the contractualist method might have, it is ultimately parasitic on some other source of value.81 If either objection held they would be pretty devasting for contractualism as an account of morality’s foundations. Given the argument I want to make against Scanlon, namely that he faces a dilemma between an unacceptable form of relativism and positing a source of value independent of the contractualist method, the redundancy version of the objection will be the more pertinent one to keep in mind.
80 See, for example, McGinn (1999), Stratton-Lake (2003) 81 Some, such as Pettit (2006) have encouraged Scanlon to embrace the redundancy objection and make his contractualism an epistemic device for discovering the best way to promote overall welfare. This takes the idea that the best utilitarian outcome might be one where the majority of people do not think in utilitarian terms. On this account the promotion of well-being would be the ultimate source of value, but the contractualist thought-experiment is the best means of determining how to realise this value. Needless to say, any contractualist who adopted this approach would sacrifice any claim to be offering an account of morality’s foundations.
56 Yet Scanlon is quite prepared to accept that there is a measure of circularity in his account. He does not claim to be offering a reductive theory of morality, where moral wrongness is to be explained in terms of something else, such as well-being. Instead Scanlon is offering what he calls a ‘holistic’ account of morality.82 When an individual is deciding on whether a consideration constitutes a reason for action they have cause to rely on various background judgements about the various other reasons they have.83 Likewise, when our hypothetical agents are attempting to find principles for the regulation of conduct, their rejection of a potential principle will depend on certain other moral judgements which are held fixed. But while it is the case that we have to rely on some ‘standing judgements’ in order to determine a particular principle, it is important for Scanlon that these judgements can be assessed using the contractualist method. What he does deny is that we can stand outside of morality and assess all of these standing judgements at once from a nonmoral point of view.
It might be thought that this response concedes too much. It might also be thought that, given that Scanlon makes this concession, I am being unfair in arguing against him on the assumption that he is providing an account of morality’s foundations. But I think these responses miss the mark. They assume that an account being ‘holistic’ and not reductive mean that it cannot be foundational. Scanlon might be thought to concede too much were it not for his stipulation that the various principles we live by can be subject to the contractualist test, just not all at once. This is his explanation for the moral content going into the judgment of whether a principle can be reasonably rejected or not. What this means is that not every principle can ultimately be explained in terms of something else. This can be contrasted with Gauthier’s approach, which can be understood as being reductive since our moral principles, it is argued, can ultimately be reduced to and explained by considerations of a (narrowly construed) rationality. Yet Scanlon’s account is still (potentially) foundational because of the possibility that every principle can be subject to the test and it there is still room to argue that these moral principles get their force in virtue of their passing that test.
82 Scanlon (1998) 216 83 Ibid., 67
57 It is also crucial, as Michael Ridge has pointed out, that the reasons that are offered by the hypothetical agents are agent-relative rather than agent-neutral – that is, they must refer to the reasons that apply to the particular agent in question, rather than, say, a reason to maximise happiness in general which applies to all agents.84 If Scanlon were to rely on the latter then there might be some substance to the accusations of redundancy. The standing judgements that are relied on to give content to the definition of ‘reasonable’ must be capable of being evaluated in the same way (although they need not be at the same time) as the principle under consideration. Being evaluated this way means that if there is a reasonable case to reject a principle this is because of the burden that is imposed on the agent in question. If agent-neutral reasons based on agent-neutral values like the general duty to respect the environment were permitted, then it is not clear how these judgements could be assessed by Scanlon’s contractualist method. Scanlon’s contractualism would not be holistic; it would depend on facts about our moral reasons that are prior to and independent of reasonable rejectability itself.85
Ridge thinks that Scanlon does not face this problem since he does rely on only agent- relative reasons. This means that Scanlon is unable to give a full account of our reasons to act morally in the wider sense – encompassing the natural environment and non- human animals for example. Scanlon’s contractualism focuses on the narrower sense of morality – hence ‘What We Owe to Each Other’.86 This dependence on agent-relativity is what Scanlon has in mind when he refers to his account as being ‘avowedly heteronomous’ when compared with Kant’s.87 While in the latter’s account moral motivation is derived from rational agency, Scanlon’ contractualism is an attempt to show
84 Ridge (2001) 85 It should be noted that Scanlon speaks in terms of personal and impersonal reasons, rather than agent-neutral and agent-relative reasons. These might not be exactly the same entities, but for our purposes the distinction is not important. What matters is that the reasons that are admissible to the contractual process must be the reasons of (generic) individuals, not reasons that apply to all agents in virtue of being agents as such. 86 It has been suggested that contractualism can accommodate the natural world and non-human animals through something like a ‘trustee’ relationship. This is controversial, but not something I will address here. I take it that if my argument against Scanlon’s account of practical reason is successful then this will apply to any attempt to extend the scope of contractualism to the non- human world too. 87 Scanlon (1998) 6 He also has in mind the variety of reason that people bring to the contractual process. As he discusses in (2011), his can be thought of as a ‘bottom up’ account of what we owe to each other, in contrast to a ‘bottom down’ more traditional Kantian approach. The sort of reasons involved here have been discussed above and will be returned to below
58 that individuals have reason to take the already-existing reasons of others seriously and treat them with respect.
The fact that the reasons that go into the contractual process are personal or agent- relative means that Scanlon can, on the face of it, avoid the redundancy objection because he does not posit the independent source of value that would be required in order to claim that there is a moral requirement for all agents, in virtue of being agents, to respect the environment.88 Here we can see the connection between what he calls the question of subject matter and the question of motivation. The question of subject matter concerns what makes it the case that I should repay Sylvia the money I owe her – that is, that any principle that licensing her demanding the repayment in a timely fashion could not be reasonably rejected by agents seeking principles for the general regulation of behaviour. What gives this moral requirement its force is the fact that it is based on the reasons that the hypothetical contractors already have along with the commitment to determine which of these reasons have overriding force – a determination that occurs through the contractual process.
*
The purpose of this section has been to establish the connection between Scanlon’s answer to what he calls the question of motivation (and I call the question of normativity) and what he calls the question of subject matter; that is, between what makes it the case that an action is wrong and what gives this fact its forces in determining what we ought to do. As has no doubt become apparent, the separation between these two questions is not as simple as in the Gauthier chapter. In the latter’s case there is taken to be a way of dividing up duties such that rationally motivated agents would agree to this, but it is a whole other question whether such rationally motivated agents will stick to this agreement. Because Scanlon posits that the hypothetical contractors are motivated to
88 Again, it is possible that there are contractualist moral requirement to respect the environment but these would be derived from the agent-relative reasons that individual agents have to value the environment, not a value that applies to all agents as such.
59 find justifiable ways to live with one another it might seem that he assumes the question of normativity away.
But this would be a mistake. Scanlon’s answer to the question of normativity is determined in part by the assumption of a willingness to find justifiable ways to live with one another, but it also relies on the existence of the substantive reasons of real agents (conceptually) prior to the contractual process. The contractual process constitutes our moral reasons by adjudicating between the competing claims that these substantive reasons give rise to. They gain their force by being the reasons people already have as mediated through this process. This is how the reasons of one become the reasons of the other.
What this suggests is that the determination of the what constitutes a moral requirement, and how it is constituted, is inextricably connected to the answer to the question of normativity. It is the nature of the agreement that gives the moral requirements their force. But this means that we must turn our attention to the nature of this agreement. As has been discussed, the reliance on pre-contract substantive reasons does give Scanlon’s contractualism an attractive degree of cultural relativity. But, as I shall argue below, this runs the risk of too strong a relativity or indeterminacy of principles, that is unless the process is supplemented with an additional source of value that invites the redundancy objection and thus undermines Scanlon’s contractualism’s claim to offer an account of morality’s foundations.
2.4 Relativism and Convergence
So far I have argued that Scanlon’s answer to the question of normativity depends on the objective conditions that need to be met in order for a (moral) reason statement to be true. This in turn is determined by the idealised response of the agents in Scanlon’s contractualist situation. This idealisation amounts to the agents seeking to live in ways that are justifiable to one another, and whether a principle can be reasonably rejected or not depends on whether it is compatible with this aim. This section will argue that this account runs into problems when we apply it to certain cases. It seems that we either end
60 up in a moral stand-off, or we find that Scanlon’s contractualist approach ends up endorsing some rather dubious moral principles.
Imagine a ‘traditional’ society where women play a subordinate role to men. Assume further that this has been the case for some time, such that this subordination is ingrained and accepted without resentment as part of everyday life. That is, the men have no wish to alter the situation and neither do the women in this society. Then imagine within this society a relationship between a husband and a wife. The wife is unusual in thinking that her husband ought to treat her with more respect. The husband is not abusive but, in keeping with the wider norms of the society, tends to treat his wife as subordinate. Perhaps he is dismissive of his wife’s opinions on certain matters of politics or the domestic division of labour. The wife feels that the husband should afford her more respect in these matters, the husband does not. I make the intuitive appeal that there is a ‘right’ way of resolving this matter, namely it is the case that the husband ought to treat his wife with more respect. That is, the husband has a (moral) reason to alter his behaviour towards his wife. The question that needs to be addressed is whether or not Scanlon’s contractualism can provide us with a convincing account of why this is the case.
We can say that the wife has a reason for wanting her husband to treat her better, and the husband has a reason for wanting the situation to remain unchanged. This is in keeping with how Scanlon would approach the question of normativity. Recalling his account of reasons, we realise that the relevant parties have their (objective) reasons that conceptually pre-exist the contractual process. That process is meant to adjudicate between these reasons and determine their relative strength. That is, the contractual process should determine whether and how the reasons of one become the reasons for the other.
It might seem at first that this won’t be a problem for Scanlon. Both have their reasons for wanting a particular outcome, but the woman can point to the manifest unfairness of the current arrangement. The man’s pleas that the arrangement actually works quite well for him because he likes to be the head of the household would not seem to be adequate to outweigh his wife’s reasons. That is, a principle that licensed the man ignoring his
61 wife’s demands could be reasonably rejected, but a principle which stipulated an equal share of the burden could not.
But this isn’t how Scanlon’s contractualism works. As discussed above, it is not the individual reasons of the two people directly affected that determine which objections count as reasonable and which don’t. Rather, it is the generic reasons that are offered by the hypothetical contractors who are taken to be representatives of the individuals concerned. In this case this would mean a representative of the class of husbands and representatives of the class of wives. And the reasons and claims that are brought by these representative are ‘generic’ in the sense that they are determined by the ‘commonly available information about what people have reason to want’. In this case, this will be information about what people within this society who find themselves in the class of husbands and people who find themselves in the class of wives. This means that the sort of claims that would typically be offered in the hypothetical bargaining process that determines how the reasons of one become the reasons of another are determined by what people who find themselves in these respective classes have reason to want. And, as we have seen, it is assumed that the relative positions of men and women dictate that neither have reason to object to the situation.
It might be objected that this criticism of Scanlon’s method conflates the question of whether the representative of the wife would object with whether she could object. Perhaps it is true that a typical representative wouldn’t attempt to veto a principle that allowed the husband to continue to treat his wife as subordinate, but it surely it is the case that she could. And if she were to, wouldn’t that constitute a reasonable objection to the aforementioned principle?
It seems not, at least not on the basis of what Scanlon says about how we should think about the claims and reasons that go into the process of determining the principles that govern our behaviour, and whether or not an objection to these principles is reasonable or not:
“I suggest that judgments about what it is or is not reasonable to do or think are relative to a specified body of information and a specified range of reasons, both of which may be less than complete. For example, the reasonableness of a belief or action may be assessed relative to the agent’s beliefs at the time and the reasons he sees as relevant. But
62 the grounds of assessment may be broader than this, provided that the reasons and information in question are available to the agent.”89
Scanlon takes it to be a virtue of his approach that he does not need to posit ideally rational agents, or agents with access to perfect information, in order to determine moral requirements. As repeatedly shown, his account takes the reasons that real people have and treats the contractual process as a means of determining our moral requirements based on these reasons.
At this point it might be asked whether the reasons involved need to be as ‘generic’ as Scanlon stipulates. Perhaps they should be less reflective of the prevailing culture and norms of the time and place and more relevant to the individuals concerned. Scanlon stipulates that the reasons of the hypothetical contractors should be personal, so perhaps the wife has compelling personal reasons to demand better treatment that transcend the prevailing norms and attitude of the time and place.
The problem with this response is that it underestimates the importance of generic reasons in Scanlon’s account. If the reasons are not generic, then it seems that we are headed for a moral stand-off. Presumably the husband has reasons to want the situation to continue unchanged and presumably he thinks that these are compelling. The idea that the contractors can come to an agreement on the principles that should govern behaviour depends on the contractors abstracting away from their day-to-day concerns and considering the reasonableness of the claims put forward given a whole other range of standing judgments which are determined by the prevailing culture of the time. Abandoning this method and concentrating only on the personal reason of the individuals concerned robs the contractualist method of its ability to avoid moral stand offs.
Suppose, just for a moment, that an answer was provided to this worry; that is, the representative of the wife somehow comes to argue for the better treatment of women in this position. Even if this is the case, we are still owed an explanation of why a claim based on this reason is going to be determined as weightier than the husband’s claim to be able to carry on with the situation as it is. Even if an appeal is made to the fairness of
89 Scanlon (1998) 32
63 the distribution of power and duties, that very concept of fairness has to have gone through the contractual process at some point. And the factors that go into determining this notion of fairness have, by stipulation of Scanlon’s method, to have been determined by the prevailing reasons and norms of the society in question.
It might also be thought that this criticism of Scanlon’s method overlooks the subtlety of his move in stipulating that it is reasonable rejection that matters in his contractualist process rather than reasonable acceptance.90 The thought here is that we would avoid the kind of situation I am describing because while certain particularly self-sacrificing groups might be said to reasonably accept what could otherwise be thought of as a principle that imposed an unfair burden on them, they could also reasonably reject such a principle, and by making the latter central to the contractual process Scanlon avoids these kinds of worries because this will exclude these kinds of situations.
But the determination of what is reasonable, and whether or not a representative of a group would reject it or not, still depends on the generic reasons that group has, which in turn is still determined by prevailing social conditions. In a situation where there is wider information about the possibility of different norms it might be the case that a contractual process foregrounding rejection over acceptance could weed out such dubious principles as the one that favours the husband, but this is not the situation we are considering.
The worry here is that Scanlon’s contractualism’s reliance of the substantive reasons that real people have leaves it unable explain why the wife in the example is correct in demanding better treatment from her husband. One option for Scanlon would be to simply bite the bullet and accept that his contractualism might be able to resolve some disputes within a society based on the prevailing norms of that society, but it is unable to mediate between disputes which take aim at these very norms themselves (or it will always resolve such disputes in favour of the status quo). Indeed, at certain points he does seem to embrace the fact that the different standards of right and wrong in different cultures might lead to different principles being adopted through his contractual process.91
90 See Scanlon (1982) p.111 91 See Scanlon (1998) p.342 and Chapter 8 passim on this
64 But if such a stance extended to the kind of example being discussed it would seem to render us with a rather unambitious model of contractualism. Nor is it one that Scanlon would endorse:
“It often seems evident that, whatever ‘they’ or some of them, may actually think (they may have become accustomed to harsh treatment, for example, and think it inevitable), they in fact have the same reasons that we do for wanting not to be treated in these ways. A relativism that fails to take these reasons seriously may be put forward as broad-minded and tolerant, but it in fact shows a lack of respect for the people in question. The view I am defending allows for this counterargument, because what it takes as fundamental is not what people actually think or want, but what they have reason to want. But it is also true that what people have reason to want depends on the conditions in which they are placed, and among these conditions are facts about what most people around them want, believe, and expect.”92
Scanlon is clearly confident that his method will enable us to explain why the wife in the above example is right and the husband should adjust his behaviour. But the key phrases in the above quote that might support this contention are ‘they in fact have the same reasons we do for not wanting to be treated in these ways’ and the claim that a relativism that failed to acknowledge this ‘in fact shows a lack of respect for the people in question’. It is worth pressing the issue of whether Scanlon is actually entitled to appeal to such notions.
The first key phrase highlighted in the passage above regarding Scanlon’s potential rebuttals to this worry concerns the claim that the kind of contractualist ethics that ended up endorsing the kind of situation ‘in fact shows a lack of respect for the people in question’. We can interpret this as the claim that even if the representative contractors only possessed the generic reasons that derived from the prevailing conditions of the society in question, and these in themselves were not enough to explain why a principle licensing the continuation of the status quo could be reasonably rejected, the fact that this kind of contractualism stipulates that the contractors are driven to find justifiable ways to live with one another is a manifestation of the equal respect alluded to by Scanlon.
92 Ibid., 340-341
65 Suppose we grant that this is the case, and the representatives of the husband and wife in the contractual process can be taken to be so motivated. We are still entitled to ask what this equal respect amounts to. The contractual process does allow for objections to be raised based on the interests of the affected parties, so equal respect doesn’t mean always conceding ground when a request is made by the other party. Perhaps it means treating the claims of the other with a certain level of seriousness, and not merely looking at them as obstacles for securing the best possible arrangement for oneself (or one’s own group).93 Presuming this is the case, it is not clear that this will lead to a principle being adopted that prohibits the subordination of the wife, or women in general, if the reasons and claims that are brought to the contractual process are determined by the prevailing conditions of the time. The representative of the husband might treat the demand of the representative of the wife with a certain level of seriousness, but this does not necessarily lead to convergence on a principle which would prohibit the continued subordination of the wife’s position. To assume that it would commits us to assuming that the generic reasons brought forward by the representative of the wife will be agreed to have outweighed the generic reasons of the husband which, as we have seen, is implausible in the setting we are considering.
Perhaps we should think of equal respect as a kind of ‘meta-principle’ such that all principles that are derived from the hypothetical contract must be consistent with the treatment of individuals (and their representatives in the bargaining process) as worthy of equal respect, where this means not just having a right to participate (or be represented) in the process but as being worthy of equal respect in terms of the behaviour that is licensed by the process.
The problem with this solution is that it seems that Scanlon must be committed to this meta principle itself passing through the contractual process, and given the circumstances that we are considering we have no good reason to assume that this will happen. For, recall, the necessity of invoking the meta principle arose from the possibility of a subordinate position for particular groups not being rejected in the bargaining process. This subordinate position was licensed in virtue of the fact that a certain level of
93 This latter approach is certainly a plausible rendering of Gauthier’s version of the hypothetical bargaining process discussed in chapter 1
66 subordination was viewed as acceptable. If the same procedure is applied to the search for the meta principle that all other principles must be consistent with a substantive kind of equal respect in behaviour, then it is difficult to see how this would be adopted by the hypothetical contractors while the lower order principle (of a wife’s subordinate position) would not.
Let’s turn to the ‘they in fact have the same reasons we do for not wanting to be treated in these ways’ claim. In this example, such a claim seems to suggest that women in this society have reason to demand different treatment even if they are not aware of these reasons. This might seem quite plausible, given Scanlon’s commitment to an objectivist account of reasons – that is, an account of reasons that is not necessarily connected to the subjective states of the individuals concerned. Recall that we are not discussing the individual wife here – she has a subjective reason to want her husband’s behaviour to change. Rather, it might be the case women in general in this society have a reason to demand better behaviour from their husbands even if they are, for the most part, unaware of this. In this case, might not the representative of these women in the contractual process argue for the respect of these reasons?
Even supposing that it is true that the women in this society have these (somewhat inert) reasons prior to the hypothetical bargaining process, it should be remembered that it is central to Scanlon’s argument that it is the result of the contractual process that provides the moral reasons with their force. Given the stipulation that the reasons that enter the process are personal, it is difficult to see how a representative advocate for claims that are not subjectively felt would pass the reasonable rejection test but a representative advocating for the opposite principle based on claims which are subjectively felt wouldn’t. Unless, that is, the force of the respective claims gathers its strength not from the hypothetical bargaining process but somewhere else. This, however, invites the redundancy objection back into the picture. For if the fact that the claims of one group are taken to trump the claims of the other is rooted not on the basis of what would be decided in the process but due to some other consideration that we have gone outside contractualism in order to explain the principle that apply to the agents. Contractualism would no longer be providing the fundamental explanation of why this requirement apply to the agents concerned.
67 It seems, then, that Scanlon is left with two options, neither of which are palatable according to the terms he has set himself. The first is simply to posit that we have reason to live in conditions of equal respect – where this means that the behaviour licensed would not be consistent with the prevailing norms of the imagined society – and this reason is derived from a source of value that is independent of the hypothetical bargaining process. Surprising as it may be, there is textual evidence to support such a reading of Scanlon. In a discussion comparing the reasons provided from his contractualism to the reasons provided by friendship, Scanlon states:
“The contractualist ideal of acting in accord with principles that others (similarly motivated) could not reasonably reject is meant to characterize the relation with others the value and appeal of which underlies our reasons to do what morality requires. This relation, much less personal than friendship, might be called a relation of mutual recognition. Standing in this relation to others is appealing in itself—worth seeking for its own sake. A moral person will refrain from lying to others, cheating, harming, or exploiting them, ‘because these things are wrong.’ But for such a person these requirements are not just formal imperatives; they are aspects of the positive value of a way of living with others.”94
As I will discuss in chapter 5, I think Scanlon is on the right lines in highlighting the importance of the relation of mutual recognition. But the way that it is presented here suggests that standing in this relation has a value that is independent of the contractual process and more akin to the value of friendship; that is, as one of a set of things that make up a good life. Appealing to this kind of value might not be objectionable in different circumstances, but appealing to this kind of value as a means to explain why particular principles must be adopted invites the redundancy objection. For if that is the reason that certain principles are picked ahead of others, then this suggests that the contractualist method is an epistemic device for finding the best way of to achieve this independently valuable state of affairs. Such a reading supports Nicholas Southwood’s
94 Scanlon (1998) 162
68 contention that Scanlon’s contractualism is based on a ‘substantive theory of practical reason’ and cannot provide an account of morality’s foundations.95
The other option is to stipulate something about the agency of the contractors involved in the process. Given that advocating an independent source of value would invite the redundancy objection and fixing the bargaining process some other way to ensure the only principles consistent with a strong understanding of equal respect pass would be objectionably ad hoc, perhaps we can stipulate more about the agents themselves. We have seen an example of this in the previous chapter: Gauthier posited that the agents in the process were rational (on his understanding of this term). Perhaps there is a conception of rationality that could be appealed to so that any representative of women in this society might not be unreasonable in accepting a principle that entrenched her subordinate position but would be irrational.96
The problem for Scanlon is that invoking a particular kind of rationality as well as the notion of reasonableness in determining moral principles begins to look a lot like he is stacking the deck in favour of the outcome he is looking for. A further issue with this kind of solution is Scanlon’s own hostility to it. In his ‘How I am not a Kantian’ paper, Scanlon contrasts Kant’s ‘top-down’ account of reasons with his ‘bottom-up’ version, stating that “I do not believe that the idea of rational agency is rich enough to yield all the claims about reasons that seem evidently correct”.97
The contrast here is between an account such as his that treats our moral reasons as having their character in virtue of the reasons that we have passing through the contractual process, and the Kantian account where the moral reasons we have are
95 Southwood (2010) p.75 I agree with Southwood that this is the most plausible way of interpreting Scanlon based on the textual evidence, although this suggests to me that the label ‘Kantian’ – applied by Southwood to Scanlon – is not entirely appropriate. See Scanlon (2011) in response to Parfit (2011) on the relation between his work and a more properly ‘Kantian’ contractualism. I discuss Southwood’s positive contribution to contractualist ethics in the next chapter. 96 This solution would seem to require a different kind of rationality to Gauthier’s since the latter’s is instrumental in that the ends are taken as given and rationality concerns the best means to achieve those ends. Given that the representative of the woman in the case we are considering is likely to have somewhat unambitious ends, it seems that a ‘thicker’ conception of rationality would be required. See Parfit (1984) on the different conceptions of rationality at play here. 97 Scanlon (2011) p. 124
69 directly derived from our status as rational creatures.98 It might be true that starting with the reasons we have prior an abstract process of bargaining might yield a richer and more varied account of what we owe to each other, but in the absence of a more detailed account of the agents involved or an appeal to a substantive account of value independent of the contractual process, Scanlon will struggle to answer the question of normativity.
2.5 Conclusion
I have argued that the worry of cultural relativism is more pressing for Scanlon than he appears to realise. This is because of its impact on the question of subject matter, which is inextricably tied up with his answer to the question of normativity. Scanlon’s account of contractualist ethics has some attractive features, but this chapter has attempted to show that these attractive features rest on an account of practical reason that can only answer the question of normativity through some methodologically dubious measures. Scanlon is right that contractualism rests on the relation between agents looking for ways to live together, but his reliance on a substantive conception of the value of doing so undermines his project. The next chapter will turn to an approach that tries to avoid the methodological pitfalls that Scanlon encounters, while maintaining the importance of the relation at the centre of a contractualist ethics.
98 To anticipate the later argument, I do not believe that the two accounts are as exclusive as Scanlon seems to here. Scanlon considers his approach to be an alternative to reasons being derived from a Kantian conception of agency. As I will argue in chapter 5, a more Hegelian account of agency can be understood as already placing us in a social space of reasons, and this conception of agency has implications for how we navigate this space of reasons and how we can think of the contractual process, but this is skipping ahead
70 Chapter 3: The Normativity of Deliberative Contractualism
This chapter provides an overview of Nicholas Southwood’s account of Deliberative Contractualism and raises some concerns about how it deals with the question of normativity – that is, what reason do agents have to comply with this version of the hypothetical agreement. The first two sections are descriptive. Section 3.1 provides an account of the structure of deliberative contractualism, and section 3.2 shows how Southwood attempts to answer the normative question. The third section raises three worries for Southwood: It is argued that the notion of agents having reasons of compliance in virtue of their role as deliberative citizens invites a redundancy objection. It is then claimed that the idea of the inescapable relation does not appear to generate the kind of normativity that Southwood needs. I hold that even if we can establish such a relation, Southwood does not give us adequate reason to think that this relation entails the kind of deliberative rationality he identifies. This is followed by a brief summary.
3.1 Introduction
The previous two chapters have investigated the two most prominent accounts of modern moral contractualism, in particular their answer to the question of normativity. Despite relying on very different accounts of practical reason, it has been argued that neither was successful in showing how the requirements derived from the hypothetical bargaining process apply to the agents concerned in the way we take moral requirement to apply to real agents. Here, we turn to a more recent attempt to provide a new version of moral contractualism, namely Nicholas Southwood’s Deliberative Contractualism and the Foundations of Morality.99 Southwood attempts to overcome the difficulties inherent in the previously discussed account of contractualism by positing a distinctive kind practical reason: a relational formalism rooted in what he calls deliberative rationality. While I argue that this attempt to solve the question of normativity is ultimately unsuccessful, I believe that Southwood has made a strong contribution to contractualist ethics by highlighting the way the question if normativity might be solved. In later chapters I will attempt to use this insight and build on it to provide a solution of my own to the question of normativity. Here, however, I will take a closer look at Southwood’s account. Despite
99 Southwood (2010)
71 its strength I argue that Southwood hasn’t provided us with enough of an argument for supposing that we do in fact stand in the sort of relation he takes to be crucial to provide a solution. Moreover, given the lack of a solid foundation for positing this relation it is not clear that the set of norms he holds to be constitutive of ‘deliberative rationality’ would actually hold in the hypothetical bargaining process based on the assumptions he makes. It is thus argued that Southwood makes a valuable, but ultimately unsuccessful, contribution to contractualism’s question of normativity.
3.2 Deliberative Contractualism
An Account of Morality’s Foundations
Nicholas Southwood’s Deliberative Contractualism and the Foundations of Morality is an attempt to offer a novel version of contractualist moral theory as an account of morality’s foundations.100 As will be familiar by now, ‘Moral Contractualism’, very broadly, refers to moral theories that account for what we ought to do with the kind of code that would be agreed to by agents in certain hypothetical, idealised circumstances.101 It shares some features with, but is distinct from, political contractualism as described by the likes of Rawls and Barry, where the device of the hypothetical contract is used to determine the most just political institutions.
Not every thinker who self-describes as a moral contractualist need be offering the same kind of account as Southwood. Contractualism might be thought of as an epistemic guide for how we should go about discovering what we ought to do. In such a theory, our moral reasons gain their stringency from some other source, perhaps from facts about well-being, but using the contractualist thought experiment is the best means of working out how best to promote well-being. Contractualism can be considered as an analysis of moral concepts, a causal-historical explanation of moral norms, or a heuristic device to render vivid what morality requires of us.102
100 (Oxford: OUP, 2010) 101 Here, and in Southwood’s book, ‘contractualism’ is inclusive of what is sometimes referred to as ‘contractarianism’: approaches which use the device of the hypothetical contract but take their lead from Hobbes, and attempt to ground morality through rational self-interest. David Gauthier is the prime example of this approach. 102 Ibid., 4
72 In contrast, the account Southwood is offering is one of morality’s foundations.103 It is offering a constitutive explanatory relation. This kind of contractualism holds that if we have, say, a moral rule that prohibits the murder of wealthy relatives, then we have this rule because such acts, or principles licensing such acts, would not pass the hypothetical contract test. This is an ultimate explanation, and, as Southwood notes, “To give an ultimate explanation of the moral facts requires us to reach explanatory rock-bottom.”104
The various moral contractualists’ accounts of morality are distinguished in large part by their distinct accounts of practical reason. As will be recalled from Chapter One, an instrumentalist like David Gauthier holds we ought to abide by the code that would be agreed to by agents seeking to maximize their own utility.105 A more substantive account is offering by Thomas Scanlon, as discussed in the previous chapter, who provides an account of the code that agents would agree to when seeking to live in ways that are justifiable to one another.106
Southwood offers version of contractualism that he describes as ‘less partial’ than the Hobbesian approach exemplified by Gauthier, and with a “more procedural but still normatively rich account of practical reason” than the approach represented by Scanlon.107 According to deliberative contractualism, “morality’s foundations are to be located in facts about what common code we would agree to live by if we were, as I shall put it, perfectly deliberatively rational.” To be deliberatively rational is, according to Southwood, to engage in discussion and bargaining with our fellow contractors while following the appropriate deliberative norms, where those norms are taken to be presuppositions of the activity of deliberation itself.108 So, just as in Gauthier’s account moral requirements correspond to the code we would agree to in a hypothetical bargaining process if we were instrumentally rational utility-maximizers, in Southwood’s account moral requirements correspond to the code we would agree to in a hypothetical bargaining process if we were deliberatively rational.
103 Southwood is very explicit that this is what he is doing. As discussed in the introduction, I take the various contractualist I assess in this thesis as attempting the same thing 104 Ibid., 11 105 Gauthier (1986). See Chapter One for my criticisms of this approach 106 Scanlon (1998). I discuss this in Chapter Two 107 Southwood (2010) 87 108 Ibid.
73 Deliberative Rationality
As with other variations of contractualism, what is agreed to, and our reasons for complying with this agreement, is determined by the assumptions made about how our idealised agents would behave in hypothetical circumstances, and the sorts of reasons they will respond to. In Southwood’s account, three conditions must be satisfied in order for a decision to be considered deliberatively rational109
1. Agents must engage in deliberation with the relevant others prior to making the decision
2. In doing so they must fully comply with relevant deliberative norms
3. The decision must be based on the outcome of such deliberation
Condition 1 suggests a distinctive account of practical reason which will be discussed below. For present purposes, it is worth noting that it is not possible to be deliberatively rational through reflection alone – one must actively engage with others in a deliberative back-and-forth. “In short, according to the deliberative conception, rationality is necessarily an interpersonal achievement.”110 Condition 3 implies a certain ‘basing relation’, the investigation of which would take us too far afield; it will simply be assumed for the purposes of this chapter. Simply put, if a decision is reached through following the procedure described, then the outcome of this will determine our moral requirements. The question of whether agents must comply with the norms (i.e. the second condition) Southwood identifies will be discussed below.
According to Southwood, the interpersonal engagement which is constitutive of the right sort of deliberation with others has three main aspects to it, and each of these aspects have certain norms that are associated with them. First, interpersonal engagement has a communicative aspect. In order for there to be mutual understanding and for the respective parties to be kept abreast of the relevant evolution that deliberation occasions there must
109 Southwood acknowledges that part of the inspiration for his account comes from Habermas’s idea of Discourse Ethics. The main difference between the kind of approach Habermas is concerned and the approach under consideration here is that the former is concerned with the conditions of actual, real-life deliberation, rather than the hypothetical procedure that characterises moral contractualism. See, for example, Habermas (1990) 110 Southwood (2010) 90
74 be a free and frank exchange of information. For this to be the case certain communicative norms must hold. There must be a norm of sincerity: the information that is communicated must be believed to be accurate (it might be based on mistaken beliefs). A norm of effective transmission must hold: a good faith attempt by the communicator must be undertaken to make oneself understood. The communicatee must follow a norm of openness – a good-faith attempt to understand what is being communicated to them. Norms of relevance and rectification (attempting to overcome obstacles to effective transmission) must be adhered to by communicator and communicatee.
Second, interpersonal engagement has a discursive aspect. The aim is towards argumentation, persuasion, and ultimately consensus. The good performance of this aspect of deliberation requires that certain discursive norms are adhered to. The norm of persuasion requires that deliberators attempt to influence each other rationally, and not through bribery, coercion or other non-rational means. Counterposed to this is the norm of adaptiveness: deliberators must be capable of being persuaded. Closely linked to both of these is the norm of reciprocity, which holds that deliberators “endeavour to present considerations for and against in a manner that their fellow deliberators are capable of regarding as normatively salient.”111
Third, interpersonal engagement has a reflective aspect – agents’ hopes, beliefs and desires must be coherent. Although the kind of reflection that produces such coherent sets of hopes, beliefs and desires is not sufficient for deliberative rationality, it is necessary. One has to know the content of one’s own mind, and be prepared and able to rethink this content when appropriate. Reflection is oriented towards self-exploration and self-interrogation, suggesting reflective norms of internal exploration and internal coherence: respectively working out as precisely as possible what the individuals think, and endeavouring to eliminate contradictions.112
Southwood allows desires and moral judgements to act as inputs into the deliberative process. He also allows aggregative reasons; that is, reasons that gain their purported weight through the impact they have on multiple agents. This stands in contrast to
111 Ibid., 95 112 Ibid.
75 Scanlon, who enforced a requirement of strictly personal reasons.113 Any judgement in principle can make it into the deliberative situation, but if the deliberative norms outlined are followed it might not remain there for long, since it will be subject to “relentless examination, scrutiny, and critique. It is a place where one will be called upon to justify one’s views and then to have one’s reasons picked over with a fine-tooth comb.”114
The purpose of this section has been to explain the type of account that Southwood is providing and to provide an overview of his account of deliberative contractualism, distinguishing it from previous versions of the idea in the moral context, and to elaborate his account of the deliberative rationality that defines deliberative contractualism. The next section looks at his attempt to explain the normativity of deliberative contractualism.
3.3. The Question of Normativity: What Reasons are there to Comply with Southwood’s Contract?
Having described the basic structure of Southwood’s deliberative contractualism, we can now turn to Southwood’s attempt to address the normativity of the contract, which is based on his interpersonal account of practical reason. The precise details, and some issues that arise with them, will be looked at in more depth in section 3, but for now it will be useful to outline how and why he thinks his account is successful in this respect. Southwood begins his discussion of how deliberative contractualist agreements provide normative reasons thus:
“I shall suggest that an account of the normativity of deliberative contractualist principles is already implicit in the account of deliberative agency [outlined above]. Roughly, this account holds that to have the capacities constitutive of deliberative agency presupposes that we stand in a certain normatively significant relation to others, and that we have
113 I think Scanlon is right here as long as we are committed to providing an account of morality’s foundations. If aggregation of the claims made by the hypothetical contractors was allowed, then it seems more natural to think that the requirements get their force from the total well-being promoted by adopting a particular principle, rather than the agreement of the agent involved. I leave this to one side, however, since I want to make a different kind of argument against Southwood. Furthermore, if my argument against Southwood fails, and the aggregation point is a serious worry, then it seems that Southwood could adjust his account without it losing its distinctive character. 114 101
76 reasons that express the demands of this relation. The reasons that deliberative contractualist agreements give us are reasons of this kind.”115
Put another way, part of what is involved in being a deliberative agent is that one has the capacity to adopt a certain standpoint or point of view, which Southwood calls the interpersonal of view. This is distinct from the personal point of view, since it necessitates the “openness and receptivity to the points of view of others and a willingness to engage discursively and reflectively with them.”116 Yet it is distinct from the impersonal point of view, since one does not ascend to the role of detached observer but is instead an active participant “within a process in which one exercises and is subject to the exercise of a special kind of shared normative power. In adopting the interpersonal point of view one is simultaneously a co-legislator and co-subject among individuals with whom one shares an authority as a deliberative agent.”117
Southwood acknowledges the debt of this interpersonal account to Darwall’s concept of the “Second Person Standpoint”.118 Claims on others are rendered valid through our shared normative power, and it follows that we must stand in a normatively privileged relation to others. There is a presupposition of sharing a common second-personal authority, competence and responsibility in virtue of being free and rational agents. “What this means is that simply in virtue of being a creature that possesses the capacities required in order to be a deliberative agent, our relation to others is relevantly transformed.”119
It is worth stressing here that it is a relation rather than a relationship that is so transformed. This relation is not akin to something like friendship with its shared histories and sympathies. Nor is it dependent on a shared worldview. A useful, but imperfect, analogy is with the connection we have with fellow democratic citizens: we cannot be said to have a relationship with all of them but we do stand in a certain relation to them, a relation that carries with it some normative significance. The idea is that in a similar way we share certain commitments as fellow deliberative citizens. Yet it is an
115 Ibid., p.119 116 Ibid., p.125 117 The personal and impersonal distinction is derived from Nagel (1986) 118 Darwall (2006) The relevance of Darwall for Southwood will be returned to below. 119 Southwood (2010), p.127.
77 imperfect analogy since the deliberative relation that Southwood envisages is wider in scope than democratic citizenship, indeed the more limited scope of democratic citizenship might be part of its value.120
Southwood characterises this account of practical reason as one of relational formalism. We do not have reason to comply with the hypothetical agreement because doing so is the best way to realise some substantive value.121 Rather, it is in virtue of being democratic citizens that we have such reasons, reasons “that are normative expressions of the relation of deliberative citizenship – reasons that express what it is to be a proper deliberative citizen. Call these reasons of deliberative citizenship. To be a proper deliberative citizen requires honouring our shared authority and mutual accountability. The only way to do this is to live according to the common code that we would agree to live by if we were perfectly deliberatively rational”.122
It is important to note that these are reasons of deliberative citizenship, rather than reasons to be a deliberative citizen. When Southwood says that the reasons we have are ‘normative expressions’ of the relation of deliberative citizenship, he is appealing to the idea that there are reasons of X that express what it is to be a proper X. The X in question might be friendship, parenthood, or, as in this case, deliberative citizenship.
There is a certain similarity to what Southwood calls rational formalism, which is mainly associated with Kant. Here the reasons that we have are taken to be formal requirements of rationality – not dependent on contingent desires nor any particular value. The difference between this kind of Kantian rational formalism and the account of practical reason associated with deliberative citizenship lies in the fact that the reasons we have are expressions of a certain relation. They are formal requirements of deliberative citizenship, but they are not formal requirement of individual rationality. Hence the term ‘relational formalism’.
120 Ibid., pp.127-8 121 This is discussed in relation to Scanlon below. 122 Ibid., pp.128-9 Southwood also claims that we not only have such reasons to abide by the common code, but also to ensure others do too, by, for example, punishing the non-compliant. This might well be implied if his argument is successful, but goes beyond the scope of this chapter. I will stick to looking at whether his account can provide reasons for the individual agent to comply.
78 We are now in a position to present Southwood’s relational formalism in more schematic terms, which should be helpful when thinking about the concerns I raise in the next section:
1. Being an agent with certain capacities means we are in an inescapable relation called deliberative citizenship.
2. In virtue of being deliberative citizens, we have reasons that are ‘normative expressions’ of this relation.
3. These reasons correspond to the code that would be agreed to if deliberatively rational agents constructed such a code.
4. By hypothesis, this code constitutes our moral obligations.
5. Therefore, all agents with these capacities are bound by moral obligations.123
This section and the previous have been dedicated to presenting Southwood’s account of deliberative contractualism and his attempt to answer the question of normativity. The rest of this chapter will be concerned with the difficulties such an account faces.
3.4 Some Concerns for Southwood’s Account
We now turn to the more critical part of the chapter. I will raise three potential worries for Southwood. Briefly: I want to question how we are supposed to cash out the idea of our reasons being ‘normative expressions’ of the relation agents allegedly stand in. There are moments in the book where Southwood appears to equivocate on how this is supposed to work: are we to understand these normative expressions of the relation as something akin to role-based reasons, or should we think of them as reasons that apply to all agents, in a manner more akin to Kantian formalism. I argue that the latter is the only plausible way of interpreting this idea of normative expression, given Southwood’s commitment to providing an account of morality’s foundations, although I briefly consider whether the former would work for Southwood if he dropped this commitment. I then turn to the idea that we are in an ‘inescapable relation of deliberative agency’. Southwood does not develop this idea in the text, but he does acknowledge his debt to
123 This is my interpretation of Southwood’s argument, although I was helped by R. Eric Barnes’s review of his book. See Barnes (2012)
79 Darwall’s ‘second-person standpoint’. I investigate whether appealing to this is a legitimate move for Southwood, and whether it will get him the normativity he requires. Finally, I question whether, even if these objections are met, and we are in an inescapable relation of something like deliberative citizenship, this means that we are wedded to the norms that Southwood identifies. I claim that Southwood has not provided us with reason to believe this is the case.
Are ‘Normative Expressions’ Role-Based Reasons?
The first worry I want to raise is perhaps the one that Southwood can most easily meet. Yet it seems instructive to raise it as it will highlight the kind of account that Southwood is committed to, which will be useful in the later stages of this chapter.
The worry relates to precisely how to cash out reasons we have that are ‘normative expressions’ of the relation of deliberative citizenship as claimed in the second premise of the argument in the previous section. It seems as though there are two ways of interpreting this claim:
(A) The reasons we have are normative expressions of our particular role as deliberative agents
(B) The reasons we have are normative expressions of our agency as such. These are reasons that apply to all (competent) agents.124
It might seem from the previous sections that the answer is clearly (B). After all, we have stipulated that is an inescapable relation that applies to agents who have certain capacities, and Southwood refers to his account as ‘relational formalism’. This suggests that the reasons we have are reasons that derived from agency as such, at least as it is understood by Southwood. But there is some textual evidence which suggests that Southwood has in mind something more like (A), so it seems worth investigating whether Southwood can appeal to something like (A) to cash out premise 2. He seems drawn to this kind of
124 I have ignored the question of inclusion for the purposes of this paper, largely because I am concerned with reasons for compliance, rather than to whom the obligations are owed. Southwood discusses these matters in pp.107-116 of Deliberative Contractualism. Still, it is clear from the argument that agents to whom the argument applies are taken to have certain capacities, hence the need for the ‘competent’ in parentheses.
80 interpretation particularly when explaining what it means to have reasons of deliberative citizenship, for example:
“Just as the friend towards whom one has knowingly failed to be a good friend has a reason of friendship to resent one for having thwarted a legitimate expectation of friendship, so too does the deliberative citizen towards whom one has knowingly failed to be a good deliberative citizen have a reason of deliberative citizenship to resent one for having thwarted the legitimate expectations of deliberative citizenship.”125
Quite how the roles we get their normative forces is a disputed area.126 On some accounts there must be a voluntary taking up of the role and the responsibilities associated with it. Others hold that certain roles feature natural duties. More hold that the process of identifying with a certain role generates normative reasons for the agent. Consequentialists argue that certain people occupying certain roles leads to greater welfare overall.
Yet it is not clear that any of these accounts of role-based reasons can help Southwood cash out the idea that the reasons we have are normative expressions of our role as deliberative citizens. The idea of roles getting their normativity as a result of being voluntarily entered into seems to be a non-starter for Southwood, given that he holds that the reasons that are normative expressions of the relation apply to all competent agents. The reasons that are associated with voluntary roles, such as friendship, can be avoided if one decides to renounce the friendship. It is, of course, possible that someone might still be blameworthy for renouncing the friendship. But such instances would suggest that we have reason to be a good friend, whereas Southwood is looking for a way to cash out the idea that we have reason of deliberative citizenship; reasons that we have in virtue of occupying the role. Similar worries would arise if the normativity of the deliberative citizenship were analogous to role-based reasons in virtue of identification with the role, since it is difficult to account for how these reasons still hold when one fails to, or no longer does, identify with the role.
Neither can role-based reasons based on consequentialist considerations or natural duties provide a suitable model for the normativity of deliberative contractualism. The
125 Ibid., 130 126 See, for instance, Stefan Sciaraffa (2009) or Hardimon (1994) for more on this type of normativity. My thanks to Thomas Smith for a helpful discussion around these issues.
81 consequentialist might claim that the private should obey the sergeant because having such a command structure bring about the best protected society. But making such an appeal on behalf of deliberative contractualism would invite accusations of redundancy. Southwood is very clear that he is offering an account of the morality’s foundations, so if his explanation of our reasons to comply with the demands of deliberative contractualism was rooted in the beneficial consequences of doing so, then he would undermine this claim. The same point applies to appealing to natural duties. Perhaps parents do have natural duties to their children, but appealing to the natural duty to comply with deliberative citizenship invites the question of where this natural duty comes from. However we cash out the purported natural duties involved in parenthood, the kind of reasons and norms that are associated with deliberative rationality do not seem amenable to the same treatment. For deliberative norms seem to be only applicable to competent adults and are, in a sense, an achieved form of rationality. This might be a problem for deliberative contractualism in general, but it is difficult to square such an account with a natural duty interpretation of the normative expressions implied by the role of deliberative citizenship.
So it seems that there is no way for us to cash out the second premise in the argument in terms of (A), yet this can be dealt with by simply assuming that Southwood always meant to be understood more along the lines of (B). The next section will turn to this interpretation, but it might be worth noting that an (A) interpretation would be open to Southwood if he dropped the stipulation to be providing an ‘explanatory rock bottom’ account of morality’s foundations.
Phillip Pettit, in his criticisms of Scanlon’s contractualism, holds that the latter can provide a plausible account of wrongness relative to a system of co-deliberation, but fails to provide an account of the grounds of absolute wrongness.127 Pettit’s suggestion is that the system of co-deliberation be justified on consequentialist grounds, and the principles of contractualism could be justified as necessary to the continuing functioning of this system. If Southwood were to drop his ambition to offer an account of morality’s foundations, then such an argument might be available to him. It is possible that the deliberative norms he identifies are necessary to such a system, and our reasons to comply
127 Pettit (2006), p.84 and passim
82 might be grounded in something like an (A) interpretation of our roles as deliberative agents. That is, there is (consequentialist) value in abiding by the code deliberatively rational agents would agree to, and we have role-based reasons as deliberative citizenships in virtue of the value of our performing this role. This is an interesting line of thought, but one that Southwood would be unlikely to want to investigate much further, given the ambition of his project.
How Inescapable is the ‘Inescapable Relation’
As outlined in section 2, Southwood holds that we are in an inescapable relation with other agents, one he calls deliberative citizenship. Whether being in such a relation entails being committed to the deliberative norms Southwood identifies will be the subject of the next section. For now, the focus is on the idea of the inescapable relation itself. Southwood does not develop this claim in his book, but he does acknowledge his intellectual debt here to Stephen Darwall’s concept of the second-person standpoint. Indeed, Southwood describes his account as a ‘special instance’ of Darwall’s concept.128 Providing a detailed critical account of Darwall’s ideas would be beyond the scope of this chapter129, but I will attempt to sketch a brief account, and consider how useful it can be to Southwood’s claims.
Darwall holds that by engaging in certain forms of address, which he calls second person reasons, with one another we presuppose certain relations of authority and competence between each other. Second person reasons are forms of agent-relative reasons that are addressed between one party (or more) to another. These are contrasted with agent- neutral reasons which hold independent of any relation between individuals.130 All second personal reasons are agent relative, but not all agent-relative reasons are second personal, since one can have a reason that applies only to oneself without it involving a relation with another person or persons. The forms of address involved in second person reasons include not only explicit demands but also ‘reactive attitudes’ such as blame and resentment. Such attitudes to one another presuppose the authority to make demands,
128 Southwood, (2010), 126 129 Although I will return to Darwall in the next chapter when I consider how to think about how a contractualist ethics based on a relational formalism might work, along with Darwall’s appropriation of Fichte to this end. 130 I discuss this distinction in more detail in chapter Two
83 and provide second-personal reasons to comply with these demands – that is, the reasons that are provided are reasons that presuppose these relations of authority and competence.131 Darwall’s account rests on a ‘circle of irreducibly normative concepts’132. The claims and demands, second-personal reasons, authority and responsibility are explicable in terms of, and only in terms of, each other and not in a reductive sense.
Darwall thinks that having some capacities, and engaging with others in a certain way, presupposes certain aspects of our relations with each other, in particular our authority to make demands and hold each other to account. Southwood, as noted, views his account as a ‘special instance’ of Darwall’s second-personal standpoint.
“What is special is its legislative dimension: it involves being in a position to render such claims valid by exercising a shared normative power, clearly we must already stand in a certain normatively privileged relation to others; we must already have the power in the first place… What this means is that simply in virtue of being a creature that possesses the capacities required to in order to be a deliberative agent, our relation to others is relevantly transformed.”133
Darwall spends much of his book establishing the relation between the various part of the circle of irreducibly normative concepts. In effect, he is establishing a reciprocity thesis. Taking the second-person standpoint implies relations of accountability, competence, and second-person reasons. But this kind of reciprocity thesis, even if correct, does not show that we are actually in this circle. This standpoint might still be ‘rationally optional or even illusory’. Darwall’s solution is to point to our reactive attitudes, such as blame: when we experience these kinds of reactive attitudes, we implicitly make demands on the other. These demands presuppose that we take the second-person standpoint.134 So just by making the sorts of claims that we make all the time, we enter into a particular relation with others, a relation that is, in fact, inescapable.
131 Darwall, (2006), 8 132 Ibid., 12 133 Southwood, (2010) p.126 134 Darwall’s cashing out of the second person standpoint that we must take involves an invocation of Kant’s realm of ends. Southwood is not committed to such a realm as the basis of contractualism. Rather he wants to use Darwall as an illustration of the kind of relation involved in is interpersonal account of practical reason. Darwall thinks of the standpoint we must stand in as populated by agents would obey the Categorical Imperative. Southwood thinks we must adopt
84 Yet it is questionable whether the sort of claims that are involved here can plausibly be construed as being enough to get the kind of normativity that Darwall (and Southwood) want. According to Darwall, when we have a reactive attitude like blame, we are making an implicit demand, and in doing so we are providing a reason for the agent to perform some action (or desist from performing an action). Yet it is not clear why we should think that this provides anything other than the weakest pro tanto reason for an agent.
A pair of examples might help to illustrate this.135 A student requests a meeting with a professor, and according to Darwall, and Korsgaard, the professor at least has an obligation to respond to the student’s request, even if it is to politely decline. Most would probably agree that this is the case. But surely this obligation to respond is not grounded in some general relation that all agents, or all members of the ‘moral community’ co-exist in relation to each other. The reason the professor has to respond is due to the institutional context within which the request is made. It is due to the role of the professor that they have a reason to respond. How this role provides a reason for the professor to respond can be cashed out in a number of ways, depending on one’s view of the nature of these kinds of obligations. However, as the previous section attempted to show, making an appeal based on these kinds of roles does not appear to be an option for Southwood for explaining the nature of the reasons that derive from deliberative citizenship.
Consider a case in a less obviously institutional context – a salesperson attempting to stop you in the street. Clearly a demand is made which invites a response, and it is possible that failing to respond might provoke some reactive attitude such as resentment. But quite what this tells us about the nature of the relation that we stand in to the salesperson is something of a mystery. Suppose, for the sake of the example, that the person is representing a particularly obnoxious organisation, a far-right news service, say. This might seem to make the example seem unfairly loaded, but the point I am trying to bring out is that issuing a demand is not the same thing as issuing a justified demand, and it seems like the latter is needed for the kind of inescapable relation that Southwood needs
the position of deliberatively rational agents. My thanks to John O’Neill for pressing me to clarify this. 135 Darwall mentions this case in (2006) p.42, but the example comes from Korsgaard (1996) p.141
85 for his argument. For if the demand, or the reactive attitude, has no justification, then it is difficult to see in what case it is inescapable.
This is not meant to be a knockdown objection to Darwall’s idea of the second-person standpoint. But it does raise questions about the kind of inescapable relation to which Southwood is appealing. Returning to the example of the salesperson, perhaps we are under some kind of ‘obligation’ to understand what they are staying to us – perhaps then we really are unable to ignore them, in the most literal sense. But if this is all that is meant by our standing in an inescapable relation to all competent agents, then it is not clear how Southwood is going to be able to use this obligation to ground any more demanding moral requirements. After all, the claim is that the reasons that we have are ‘normative expressions’ of this relation. But if these reasons are simply the normative expressions of being in a relation where we are ‘obliged’ to understand the request, then it is difficult to see how we get from here to the need to comply with the demands of deliberative citizenship. Such a move might be more plausible when we think back to the other example. Perhaps the reason the professor has to respond to the student is a normative expression of their relation. But these seems to be more akin to role-based reasons (however we seek to cash this out), and as we have seen, deriving the normativity of deliberative contractualism from these sort of reasons does not seem to be a legitimate move for Southwood.
The Deliberative Norms
Suppose, however, that we were able to establish some kind of inescapable relation, and this relation was such that it was plausible to say that it implied we should act according to some kind of hypothetical contract. There might still be a problem for Southwood in justifying his claim that being in this relation, and having reasons that are normative expressions of this relation, means that we have reasons that correspond to the code that would be agreed to by deliberatively rational agents, as Southwood defines this. We might still have cause to question premise 3 from the argument outlined above.
It might have been noted that Southwood’s deliberative norms do appear to contain a substantial amount of moral content. As we have seen, these deliberative norms include communicative norms such as sincerity (honesty, authenticity), openness, effective transmission, effective reception. It seems that these norms express moral values –
86 sensitivity to others, reciprocity, understanding, empathy. Discursive norms include persuasion (ruling out coercion etc.), reciprocity (implying respect), adaptability (a morally desirable willingness to be accommodating/flexible). But if the deliberative norms that Southwood identifies turn out to be there because they are morally desirable qualities, then Southwood would again be open to a circularity worry objection. We would be advocating following the code that morally praiseworthy people would agree to. Compare the instrumental contractualist justification:
(1) We ought not kill our wealthy relatives because doing so would violate an agreement we all have (instrumental) reason to adhere to
With
(2) We ought not kill our wealthy relatives because doing so would violate an agreement that moral agents would agree to.
Whether or not one thinks that principles such as (1) can provide a suitable account of morality, it seems pretty clear that an account based on them will have a much more plausible claim to an ‘explanatory rock bottom’ than (2).
Southwood is aware that this might be a worry, and in response notes that we are evaluating conduct “against norms that are presupposed by, and applicable to one in virtue of, one’s occupying a certain role, the role of deliberator.”136 It is true that some of these norms are also morally forbidden. But this does not show that deliberative norms are parasitic on moral norms any more than prudential norms are parasitic on moral norms when they overlap. He offers the example of the suicide bomber, who seems to violate quite distinct prudential and moral norms, and yet we still seem confident in having an independent grip on both sets of norms. One can adhere to, just as one can violate, different kinds of norms at the same time without losing one’s ability to assess them and their implications individually. He further notes that it is hardly surprising that those who were deliberatively rational would choose moral norms that bear some resemblance to deliberative norms. It would be strange if deliberatively rational contractors chose principles that allowed for widespread insincerity. The crucial point, he argues, is not about the resemblance between the moral norms and deliberative norms,
136 Southwood (2010), p. 182
87 but rather whether we can have an independent grip on deliberative normativity. In support of this, he notes that it seems possible that a moral nihilist would not be rationally barred from making assessments about deliberative rationality.137 That is, it doesn’t seem inconceivable that the moral nihilist can hold forth on deliberative rationality, and which are the best deliberative norms, while still being unmoved by moral considerations.
This seems, on the face of it, a reasonable point. It is possible that a moral nihilist might have some opinions on the best way to deliberate. It is somewhat less clear that the moral nihilist would think that Southwood’s particular deliberative norms are the right way to go about deliberating. The worry can be clarified when we consider Barnes’s distinction between being a good deliberator and being a good negotiator.138 A good deliberator would be someone who follows Southwood’s deliberative norms – the kind of norms that imply certain levels of respect and reciprocity towards others – but a good negotiator might be willing to be somewhat more manipulative, perhaps even coercive. They might be more willing to engage in some strategic posturing to get the best deal that they can. The norms involved with being a good negotiator might be something like ‘always begin with an unrealistic target’, or ‘be prepared to deceive if you are sure of not being detected’.
I am not trying to suggest that being in such a relation does in fact entail that agents ought to abide by the code that good negotiators, rather than good deliberator, would agree to. But I am suggesting that given the different possible interpretations of what being in this inescapable relation involves, we need an independent reason to follow the route of deliberation. Failure to provide such an independent reason, opens Southwood up to ad hoc objections: the reason for choosing deliberative norms over negotiator norms seems to be largely because it is easier to come up with an account of why deliberatively rational agents will comply with the contract rather than instrumentally rational agents (for there seems to be little difference between agents who follow negotiator norms and the kind of instrumentally rational agents who appear in Gauthier’s account, apart from the fact that they enter into the contract for reasons of individual gain, rather than being in an inescapable relation).
137 Ibid., p. 183 138 Barnes (2012), p. 817
88 Can Southwood provide an independent reason for choosing his deliberative norms? He doesn’t do so in the text, and it is difficult to see how he can with appealing to something like the kind of substantive value discussed above. To take a different kind of example, consider democracy. Maybe our justification for democracy is based on the fact that democracy best realises the substantive, independent value of individual self-realisation. And perhaps it can be argued that a democracy where people are encouraged to follow something like the kind of deliberative norms that Southwood identifies would best promote that value. The analogous case in contractualism would be the claim that a contractualism based on Southwood’s deliberative norms would best promote, say, individual well-being. However, as we have seen, this would mean Southwood abandoning his ambition for deliberative contractualism to provide an account of morality’s foundations, and so this would not be acceptable.
Conclusion
This chapter has aimed to provide an overview of Southwood’ deliberative contractualism and his answer to the question of deliberative contractualism’s normativity. It has attempted to show that all the major premises in the argument for deliberative contractualism’s normativity are questionable. In brief, it is not clear precisely what the inescapable relation Southwood identified consists in, and whether it can generate the kind of normativity he needs. It also seems questionable whether the idea that the reasons we have in virtue of this relation can be best thought of as role-based reasons, as Southwood seems to suggest at various stages. And even if we can come up with a plausible argument for our being in this relation, it is not clear that being in such a relation entails being bound by the deliberative norms that Southwood identifies.
Some of the worries presented in this chapter could be dealt with if Southwood gave up his ambition to provide an account of morality’s foundations, and instead offered something closer to an epistemic contractualism; a device for discovering independent moral truths. Or perhaps his deliberative norms could serve as a guide for a system of co-deliberation which is justified by another kind of substantive value. But I suspect that Southwood would be reluctant to take this approach, and perhaps he is right be so. It is not inconceivable that an independent reason could be found for preferring his deliberative norms over something closer to negotiation. And while I do not believe his
89 account of the inescapable relation satisfactory, I suspect he is right that something like an interpersonal, or second-personal, account of practical reason is needed to provide a satisfactory answer to question of the normativity of the hypothetical contract, given the difficulties that previous proponents of this approach have had in answering this question.
Southwood’s account ultimately fails to answer the question of normativity, but it provides an opening for thinking about what a successful answer to this question might look like. I will return to this idea in the next chapter.
90 Chapter 4: Relational Formalism and German Idealism
The first three chapters of this thesis have argued that the conceptions of practical reason supporting the prominent accounts of moral contractualism are unable to answer the question of normativity. But they have yielded important insights into how we might approach this question. This chapter will lead off from Southwood’s idea that moral contractualism must rely on an account of practical reason that is interpersonal in nature, and that such an account must be formal in the sense that it is inescapable – it needs to be understood as part of our agency as such. It will begin with a recap of the arguments of the previous three chapters and explain how these negative arguments suggest the form that any positive attempt to explain the kind of practical reason a contractualist ethics requires must take. That is, a contractualist ethics, if it is to answer the question of normativity requires a formal interpersonal account of practical reason. I will then turn to two ways of thinking about the kind of inescapable relation that is implied by such a formal account; namely that of Stephen Darwall and that of Michael Thompson. I will argue in favour of the latter’s as the best basis for thinking about the kind of relation an account of practical reason that will provide an answer to the question of normativity should be based on. Nonetheless, I will then turn to Darwall to look at his appropriation of Fichte for establishing the existence of this inescapable relation. While useful in illuminating the kind of move that needs to be made, I argue that this has its shortcomings. This will set up the next chapter, where I discuss the possibility of appropriating Hegel’s account of agency and mutual recognition as the basis of a formal interpersonal account of practical reason.
4.1 Recap: The Shortcomings of Actually Existing Contractualism and the Need for a New Kind of Practical Reason
This section will start with a recap of what can be thought of as the negative part of my thesis’ argument – the insufficiency of the various other accounts of contractualist ethics to provide an answer to the question of normativity. The three previous chapters have argued that any contractualist ethics that attempts to provide an answer to the question of normativity must rely on a particular conception of practical reason. This amounts to an account that answers the question of in virtue of what do agents have a reason to:
91 (A) Agree to the conditions of the hypothetical contract (however these conditions are cashed out)
(B) Adhere to these conditions when faced with moral decisions
The focus in this thesis is on the various accounts of contractualist ethics’ answer to (B): the question of normativity. However, as has been shown, the various accounts’ explanation of (B) must also be compatible with their explanation of (A).139
It has been further assumed that all the accounts considered aim to provide a version of a contractualist ethics that accounts for morality’s foundations. That is, an agent ought to avoid ϕ-ing because a principle licensing ϕ-ing was rejected in certain hypothetical circumstances or, equivalently, ϕ-ing is wrong because it would be rejected in such circumstances.140 Contractualism does not necessarily have to be concerned with morality’s foundations. It is conceivable that a contractualist could concede that the force of moral commands comes from a different source of value, such as the value of promoting the well-being of all persons. A contractualist might put forward the thought experiment of agents trying to come to an agreement concerning the general regulation of behaviour as an epistemic device to discovering the best way to promoting this wellbeing. But the contractualists considered do not endorse such a concession. They are concerned with providing an ‘explanatory rock bottom’ account of morality’s foundations,141 and I will continue to assume that an account of morality’s foundation is the guiding aim of contractualists in general.
Chapter 1 considered David Gauthier’s attempt to answer the question of normativity. Gauthier subscribes to an instrumental account of practical reason. This is cashed out in the form of a Rational Choice Theory account of what the rational agent has reason to
139 This is most clear in Chapter One, where the posited motivation that compels the rational agents to reach an agreement is found wanting in explaining their reason to continue to comply with this agreement. 140 ‘Wrong’ here is understood to refer to acts that can be characterised as wrong that occur between agents or persons. It concerns what Scanlon calls a narrow version of morality. See Scanlon (1998) for more on this. To repeat the point made in the introduction, I will set to one side the difficult questions that all contractualists must face concerning how they can account for our responsibility to non-human animals. I think that these questions do pose genuine problems for all varieties of moral contractualism, but that is not the issue I am focusing on in this thesis. 141 In Southwood’s (2010) phrase p. 11
92 do. That is, the reason we have to avoid ϕ-ing is that doing so will lead to a loss of our subjectively defined utility. By assuming such an account of practical reason, we can explain why agents would enter into an agreement in certain hypothetical idealised circumstances. Perhaps it is better for the average agent’s prospects (defined by their subjective utility function) if we have (say) a rule prohibiting the murder of wealthy relatives in order to get an early inheritance. That is, Gauthier’s account of practical reason seems to provide a satisfactory answer to (A). But if such an agent found themselves in a situation where they were able to kill these relatives and be virtually certain to get away with it, it is not clear than an instrumental account of practical reason can explain why the agent should not take up this opportunity, so the instrumental account of practical reason cannot provide a satisfactory answer to (B). Gauthier is aware of this problem, and attempts to offer a solution in the form of the claim that it is rational for such agents to develop a disposition to comply with the contract’s requirements independent of whatever short-term utility gain they might get out of not doing so. I argued that this solution is unstable. Moreover, any attempt to overcome this instability is going to be inconsistent with any plausible metaethical account of motivation; a crucial blow for an account that attempts to explain our moral reasons in terms of subjectively defined utility functions.
Chapter 2 looked at Thomas Scanlon’s account of contractualism, and his proposed answer to (B). It was argued that Scanlon’s answer to the question of normativity is best understood as relying on a substantive account of practical reason. This was supported by textual evidence and an investigation into the connection between Scanlon’s realist account of reasons and his constructivist ethics. Yet this account of practical reason was found to be problematic when applied to a contractualist ethics that attempts to provide an account of morality’s foundations. Suppose that ϕ-ing is wrong according to his contractualist formula. Then, according to this interpretation of Scanlon’s account of practical reason, our reason to avoid ϕ-ing is given by the substantive value of living in conditions of mutual recognition with others, which just is acting according to Scanlon’s contractualist formula. Yet Scanlon’s account faces a dilemma. If the substantive value of living in conditions of mutual recognition does not apply to all agents – that is, if it is possible for agents to reject the value for themselves of living in such conditions – then Scanlon’s contractualism fails to explain the special bindingness of our moral obligations:
93 he fails to answer (B). If, however, the claim is that this substantive value is something that applies to all agents, then we need a further explanation as to why this is so. I argued that Scanlon cannot do so without invoking moral reasons, moreover he must invoke moral reasons that are independent of the contractual process that he describes, thus inviting redundancy objections and failing to provide an account of morality’s foundations.
Chapter 3 assesses Nicholas Southwood’s deliberative contractualism. Here, the agent’s reason to ϕ is given by an interpersonal account of practical reason. More precisely, this account of practical reason holds that we have a reason to ϕ because ϕ-ing is what a deliberatively rational agent would do, where being deliberatively rational is understood to mean complying with various deliberative norms. Crucially, being deliberatively rational is not something agents are supposed to have some (otherwise explained) reason to be (as was the case in the argument against Scanlon), but rather is the result of their standing in an inescapable relation to other agents. This chapter argued that this inescapable relation is poorly defined, and even if granted it is not clear that it entails a commitment to the deliberative norms that Southwood identifies. Southwood faces the dilemma of, like Scanlon, appealing to an independent source of value to motivate the need for these norms, or his contractualism risks collapsing into something like the kind of instrumental approach criticised in Chapter 1.
A feature that Gauthier and Southwood’s account of contractualism share is a commitment to what Southwood calls a ‘procedural’ account of practical reason. The reasons that agents have arise as a result of facts about their agency – they have a reason to ϕ because ϕ-ing is implied by the types of agents they are. In Gauthier’s case, the ‘type of agents they are’ is cashed out as the kind of instrumentalist utility maximizers posited by rational choice theory. For Southwood, ‘the type of agents they are’ is cashed out as deliberative citizens whose status as deliberative citizens requires them to act according to the deliberative norms he discusses: an account he describes as relational formalism. These approaches stand in contrast to accounts which hold that an agent has a reason to ϕ because ϕ-ing realises some substantive value. In Scanlon’s case, this substantive value is living in conditions of mutual recognition with other agents. If the argument of chapter 2 is accepted, any contractualist approach which finds itself appealing to a substantive
94 account of practical reason in this way will face the same problem that Scanlon does; namely, redundancy objections and the impossibility of providing an account of morality’s foundations. It follows that if a contractualist account of morality’s foundations can answer point (B) from above, it must be based on a procedural account. But if the argument of chapter 1 is accepted, then any individualist procedural account of practical reason will face the same problem as Gauthier.
If the above is accepted, then by a process of elimination we can say that Southwood is correct about the kind of account of practical reason that any version of moral contractualism must be based on in order to answer (B). That is, it must be interpersonal and it must be inescapable. However, as noted above, Southwood is vague on the nature of this relation, and doesn’t provide a satisfactory account of what it entails. My intention, and the positive intent of this thesis, is to provide a more plausible grounding for thinking that such a relation exists and that it can entail facts about what we have reason to do, and this can be useful for thinking through how a contractualist ethics might answer the question of normativity. If accepted, this will point us towards a solution that is not objectionably ad hoc (Southwood), open to redundancy objections (Scanlon), or an implausible account of our reason to continue to comply with the requirements of the hypothetical contract. To anticipate, I will argue that the insights of the German Idealist Johann Gottlieb Fichte and, especially Georg Wilhelm Friedrich Hegel can help us reach such a solution. But before looking at these figures’ potential contribution, I will look in greater depth at what such an interpersonal account might look like.
4.2 A Second-Personal Solution?
According to Darwall, when we address a demand to another person we are attempting to provide them with a second-personal reason. The notion of a ‘second-personal reason’ is best illustrated with the example Darwall uses repeatedly.142 If my foot is on top of your toe and is causing you pain, we can think of two ways of explaining why I have a reason to alter this situation by lifting my foot off your toe. It might be drawn to my attention that my foot being on top of your toe is causing pain to you, and your being in
142 Darwall (2006), p. 5 and passim
95 pain provides a me with a state-of-the-world-regarding (or agent-neutral) reason to take the appropriate action to stop causing you pain. There might be such reasons for me to move my foot, but they would not be second-personal in nature. Such reasons are agent- neutral in that they apply to any agent who is able to take action to relieve your pain, and do not only apply to the agent who is directly causing the pain.
This is contrasted with a reason that derives from a claim or a purportedly valid demand. This might be a demand from the person whose foot is being stood on, or a member of the moral community. The crucial point is that the reason that I have to remove my foot from your toe (and we assume this is a moral reason) results not from some feature of the world – the amount of pain that exists, for instance – but rather from the fact that you have issued a demand to me. Such reasons are agent-relative, in the sense that an explanation of the source of the reason – “I have a reason to move my foot because…” – must refer to the agent(s); they apply to the agent in the specific situation in particular, not to any agent able to (say) alleviate the pain felt. Such reasons, issuing from the claims or demands of the other person(s) are described as second-personal. Darwall does not claim that all the reasons agents have are second-personal, but he does hold that the concept of moral obligation is irreducibly second personal.143 He holds that it is a conceptual truth that we are morally obligated to do is what we are warrantedly held responsible for. Holding another responsible involves the second-personal address of a demand and bids for the recognition of the authority to address it.
This is the case not only where explicit demands are addressed, but also where implicit demands are addressed through our various ‘reactive attitudes’ such as blame and resentment. Whether a second-personal reason is provided depends on certain normative felicity conditions being met.144 These include both parties having a common
143 Ibid., 11. C.f. ch.5 91-118. This might seem like a controversial claim, but I won’t defend it here. Even if the idea of moral obligation being second personal might seem questionable to some, a contractualist ethics account of moral obligation seems less so. If moral obligation is bound by what we would agree to in certain circumstances, then it seems that must correspond to the duties that we owe each other. Matters might be complicated if we took a non-foundational approach to contractualism; if, for example, contractualism was thought of as an epistemic device to find the best means of promoting an independent source of value. Then the concept of moral obligation might not be irreducibly second personal. But I am assuming this possibility away. 144 This notion of felicity conditions being met is taken from Austin (1975). Roughly the idea is that in order for a certain speech act to take place certain social conditions must be in place.
96 competence, authority, and therefore responsibility towards one another.145 Because second-personal reasons have these felicity conditions the act of addressing them entails certain moral commitments, namely that in issuing such a demand we commit ourselves to a shared point of view from which both parties can see that the claim issued by A constitutes a reason for B to ϕ. This kind of shared point of view should be fairly familiar from the previous chapters’ discussion of the various existing forms of moral contractualism. Such a ‘shared point of view’ can be understood as the genus of the species of the interpersonal point of view that Southwood uses to ground his account of practical reason.146
I will come back to some potential problems with this account below, but it seems worth pausing to consider how the second-person standpoint fits in to a contractualist ethics.147 If I am interpreting Darwall correctly, the making of demands in the second-personal sense is taken to imply accountability. This accountability is not just to one another, but to the moral community in general. Such accountability to the moral community implies a set of rules governing conduct towards one another. Darwall holds that such rules are best thought of through a Kantian realm of ends: “‘a regulative ideal that we employ to make sense of our ethical thought and practice”.148 A contractualist picture, on this kind of reading, is a way of cashing out how such a regulative ideal would work; so we might understand this by having a regulative principle along the lines that an act is not permissible if a principle prohibiting it could not be reasonably rejected by these members of the realm of ends, to provide a Scanlonian version of the idea.
*
Throwing a glove on a table counts as a challenge to a duel in certain cultures, but this is only the case in virtue of the prevailing norms and understanding of the time. 145 Ibid., pp. 4-12 146 Southwood (2010), p. 126. What makes Southwood’s account a ‘special instance’ of Darwall’s idea is the purported entailment of deliberative rationality from standing in this relation. See the previous chapter for my criticisms of this move. 147 Darwall, (2006), ch.12 pp. 300-320 148 Darwall, (2007) pp. 52-69., especially p. 64. The realm of ends referred to here is not straightforwardly Kantian. Chapter 9 of The Second Person Standpoint is devoted to arguing that Kant’s notion of autonomy must be understood second-personally. This does not seem crucial to my concerns here though. See Korsgaard (2007) for some criticisms of this move
97
While the above necessarily glosses over many of the subtleties of Darwall’s account, it hopefully provides an overview of the general idea and its relevance for contractualism. But the idea that the inescapable relation that we stand in to one another is grounded in the claims of agents is not without its critics. Going back to the example of your hurting me by standing on my foot, we might think that the reasons you have not to stand on my foot exist prior to any claim being made by me. Of course, Darwall does not argue that these reasons derive from explicit claims made by agents. Rather, it is the disposition of the agent whose foot is being stood on to have the reactive attitude of blame or resentment which implicitly makes demands on the other person.
But as R. Jay. Wallace notes, this doesn’t entirely deal with the problem. The person whose foot is being stood on, causing them pain, might be so worn down by life that they cannot even summon the energy to have a reactive attitude such as blame or resentment. If this is the case, and the implicit or explicit claim is taken to be constitutive of the relation concerned, then this still presents a problem for Darwall. Furthermore, they might have the misfortune of living in a community where nobody else is exercised about such indignities either. If this were the case, then the Darwallian idea that the relation we stand in to others derives from our reactive attitudes seems to imply that there is no reason for you not to stand on my foot.149
Wallace holds that it makes more intuitive sense to suppose that the reasons we have not to stand on each other’s feet exist prior to and independent of any claims made by the affected agent or the moral community. That is, we have these reasons independent of any claims made, with explicit demands or implicit reactive attitudes. This can be further illustrated by considering the role of consent. I might give you permission to inflict some pain on my foot – perhaps if you are a surgeon performing an operation. But if, as Darwall holds, the reasons I have to avoid inflicting pain on you derive from reactive attitudes, then this seems harder to understand. The problem is that on Darwall’s model the reason not to inflict pain on your foot is created by/derived from the reactive attitudes that I have. If, on the other hand, the reasons I have not to inflict pain on your
149 Wallace (2007), 27
98 foot exist prior to any claims or attitudes of yours (leaving aside, for the moment, how and why this is the case), then the notion of consent makes more sense. The idea is that there are standing reasons not to inflict pain, and that in certain circumstances these can be vetoed or rescinded by the rights holder, seems more intuitively plausible than the idea that we are free to inflict pain except in those cases where the reactive attitude directs us otherwise.150
Darwall does have a response to such objections. First, he re-emphasises the point that the ‘claims’ in question need not be actual or explicit claims, but rather simply agents’ proneness to certain reactive attitudes. So it is not the making of a claim but the existence of a disposition to make such a claim. But such a response would not answer the question of the agent who is worn down by life and has no proneness to such attitudes. It seems that it is correct to point out that the reasons we have cannot be solely based on the reactive attitudes of actual individuals, and Darwall appears to acknowledge this.151
This is where the idea of the moral community comes in. As stated above, the moral community is not, in Darwall’s account, an actual community of human beings. It is understood to be a regulative ideal along the lines of Kant’s formula of the realm of ends. As he puts it,
“We might… understand the moral community as being prone to the reactive attitudes in a contractualist way, for example, taking it that moral demands are ‘in force’ if no one could reasonably reject principles that would warrant them, or if these principles would be chosen by representatives from a point of view that expressed the idea of respect for all persons as having equal second-personal authority.”152
In this case, the reason you have to avoid standing on the person worn down by life’s foot does not depend on her reactive attitude at all, but rather the reactive attitudes of those agents in the realm of ends who are motivated by the Categorical Imperative.
150 Ibid., 29-30 It should be noted that Wallace’s objection seems to appeal to the favourability of a more Scanlonian approach to the reasons we have. Scanlon would presumably agree that we have these standing reasons prior to any claim being made. The contractualist method is deployed to explain when these standing reasons have particular force, when they can be overturned, and so on. Chapter 2 deals with the limitations of such an approach. 151 Darwall, (2007), 64 152 Ibid., 65
99 Yet this response does not seem completely satisfactory, at least if we remember that the aim is to provide an interpersonal account of practical reason to ground the normativity of the contract. First, it seems in danger of doing away with the interpersonal aspect altogether. If the reasons we have to avoid standing on people’s feet derive from the moral community, and this moral community is understood along the lines of a Kantian realm of ends, then it is not clear what role the reactive attitudes of real people play at all. But it is the existence of the reactive attitudes that we have and what is implied by the existence of the reactive attitudes that puts us in the ‘circle of irreducibly normative concepts’. Darwall spends much of the book establishing.153 Put another way, if the obligations we have to each other are established by legislators in the realm of ends, then what is the need for the second person? Darwall’s account seems less amenable to providing the basis of an interpersonal account of practical reason rooted in the agent- relative reasons of individuals and more like an agent-neutral account of the reasons we have that derives from the Kantian realm of ends.
A related worry, specific to using this as a basis for a contractualist ethics, is that the normativity of the demands that result from this standpoint seem to derive from Kant’s categorical imperative. If our reason to avoid causing pain to others is grounded in our being rational agents, and being rational agents means adhering to the categorical imperative, then we might ask what the role of the contract is in all this. Part of the whole point of any contractualist ethics is the claim that our moral requirements are given by what we would agree to in certain hypothetical idealised circumstances. On such an account, we ought to avoid ϕ-ing because ϕ-ing would violate the principles that would derive from such an agreement. But if the wrongness of an action is rooted in the categorical imperative, which is itself rooted in facts about what it is to be a rational agent, then it seems that the impetus to act morally comes from this fact and not from facts about the hypothetical agreement. It might be that thinking through a contractualist thought experiment is the most promising way to discover how best to express the idea of having equal respect for all, but if the reason to have equal respect for all is established
153 I introduced this idea of the circle of irreducibly normative concepts in Chapter 3. I will return to what is implied by the reactive attitude that Darwall makes central to his account in my discussion of his appropriation of Fichte below.
100 antecedently, then the contractualism we end up with is not an account of morality’s foundations.
4.3 Thompson’s Alternative
Thus far, I have argued that Southwood is correct to claim that in order to provide a compelling answer to the question of normativity a contractualist ethics requires a formal interpersonal account of practical reason. The limitations of Southwood’s specific version of this were discussed in the previous chapter. In this chapter, I have looked at the inspiration for Southwood’s model in more general terms, namely Stephen Darwall’s account of second-personal normativity. I have noted some limitations that would be involved in a move to use this kind of account as the basis for an interpersonal account of practical reason for contractualism. I will have reason to return to Darwall shortly, since he is not only the inspiration for Southwood’s account but also provides a clue as to how some of the insights of German Idealism can be incorporated into the question of normativity. For now, however, I want to turn to another way of framing the kind of formal interpersonal relations that are needed to form the basis of the kind of practical reason that contractualism requires. I argue that Michael Thompson’s account of ‘bipolar’ normativity provides a more helpful way of thinking about how to approach this kind of practical reason. I claim that it is superior to Darwall’s as the basis for contractualism, but still leaves some unanswered questions; questions that I will attempt to answer later on by drawing on the insights of Fichte and, especially, Hegel. But it is to Thompson that I now turn.
Thompson’s approach is, like Darwall’s focused in a particular kind of wronging – where A is wronging B and B has a claim against A. He calls these relations of bipolar normativity. Bipolar normativity is distinguished from ‘merely monadic normativity’ – this can be understood as the kind of normativity that applies ‘in general’ – i.e. not to a specific person.154 This distinction can be understood as equivalent to the agent-relative and agent-neutral kinds of wrongness identified in the earlier discussion of Darwall.155 As he
154 Thompson (2004) p.338 155 These distinct kinds of normativity apply to non-moral domains as well as the moral domain.
101 puts it, the puzzle he is concerned with is what makes the bipolar thoughts, such as “I should pay Sylvia back the money I owe her” true.156
Thompson’s suggestion concerning how we explain how bipolar normative thoughts come to be true is through his definition of persons. When two agents are to be represented in true bipolar judgements of a certain type, they are defined as standing together under a particular dikaiological order (meaning the responsibilities they have are to one another), that agent is rendered a person. ‘Person’ here is distinguished from ‘agent’, with the latter being understood as a being that operates on the strength of practical reasons. Any normativity for an agent in this sense would derive from the agent’s own ends: “If all else is left out of the account, pairs of agents will at best provide materials adequate for an application of game theory.”157
Thompson’s concept of a person, in contrast, is always relational. X is a person in relation to Y, in the same way that X is a sister in relation to Y. How this relation works depends on the wider circumstances in which the ‘persons’ operate. What Thompson calls the ‘manifold of persons induced by a given dikaiological order’ determines how persons within such an order are related to one another.158
This idea can be used in (and is perhaps best illustrated by) nonmoral cases. An example would be persons who are legal citizens of a certain territory. In virtue of being a legal citizen of a certain territory, other citizens (read: persons), incur certain responsibilities towards me, perhaps giving way in certain traffic situations. That is, I incur a bipolar obligation to you to give way in a particular situation in virtue of the fact that we are both citizens of this particular territory. We can see that this is the case since in the event that I don’t do so, and in failing to do so cause an accident, I will end up with a legal requirement to pay compensation to you as a fellow citizen. I might end up having to pay
156 Ibid., 337 157 Ibid., 353 And as I argued in chapter 1, a contractualist ethics based on such an approach cannot adequately answer the question of normativity. I should note that although Thompson draws a distinction between ‘persons’ and ‘agents’, with the latter understood in a narrower sense more akin to the kind of rational utility maximizer discussed in chapter 1, I will continue to allow that there might be other conceptions of ‘agency’. 158 Ibid., 354
102 some sort of compensation to a non-citizen, but the circumstances will have changed and as a result the legal requirements will have changed.
Another example would be the ‘manifold of persons induced a by given dikaiological order’ being the players in a game of chess. Here we can substitute the relational concept ‘player’ for the concept of ‘person’. Being a player of chess means incurring certain obligations to the other player who I am constitutively related to, such as not moving a rook diagonally. If I fail my obligations here I have wronged the other player as another player of the game of chess.
The difference between these kinds of bipolar obligations and the kinds of obligations described by Darwall concerns the fact that they do not derive from the claims, implicit, dispositional, or otherwise, of agents or from the ‘moral community’ in general.159 Instead, the claims can be understood as being constitutive of the order which the relevant people find themselves in and the relation they stand in to one another in virtue of being within that order. Yet in the examples given, it is clear that the relevant authority derives from the rules of chess, or the authority of the state (however this is cashed out). The question that Thompson asks, and the question relevant to the present discussion, is what manifold of persons is induced when assessing the statement that ‘I ought to pay back Sylvia the money I borrowed from her’ in the moral sense.
That is, when considering the truth or otherwise of the sentence:
‘I ought not move a rook diagonally when playing a game of chess’ we can determine that it is true in virtue of the fact that if I were to do so I would not, in fact, be playing chess at all. The rule that prohibits the moving of rooks diagonally is constitutive of the game itself.
In a somewhat similar way, the legal reading of the sentence
‘I ought to give way when entering a roundabout’
159 At least where this is understood as a Kantian realm of ends
103 Is true in virtue of the fact that I am a citizen of a country that has these laws, and if I disobey these laws with the consequence of bringing harm to you, then I will owe you compensation in virtue of you being a fellow citizen.
However, as the current investigation is concerned with contractualist ethics, we are interested in the what makes the following sentence true:
‘I ought (morally) to pay back Sylvia the money I borrowed from her”
What makes the previous two statements true is the fact, in Thompson’s terms, that the individual in each case is a person in the manifold of persons induced by taking part in a game of chess, or being a citizen of a particular territory. As Thompson puts it, “In what ‘many’ does our just agent implicitly locate herself as ‘one’ in thinking the thoughts of justice? What puts her into connection with Sylvia, thereby inducing a class of others to whom she might intelligibly be similarly linked?”160 We can think of this as an attempt to work through the relation implied when we make such moral statements, and thus to help us understand what we are doing when we assess these statements as being true or false.
Thompson considers Humean, Aristotelian, and Kantian attempts to answer these questions – each interpreted as an attempt to provide an account of the manifold of persons induced when make a moral judgement. He finds them all wanting. The details of his arguments against these approaches need not detain us, since the aim here is to determine whether this way of thinking of our interpersonal relations can help us answer the question of normativity that faces any contractualist ethics. The question I am concerned with is whether Thompson’s framing of the relation involved in determining the truth of purported bipolar moral obligations is a useful way to conceive of the kind of relation involved in contractualist ethics. More specifically, does Thompson’s idea of the order of normativity brought about by the manifold of persons provide a more useful framing than Darwall’s moral point of view based on the second-personal reasons implicit the claims implied by various agents?
160 Ibid., 359. When Thompson talks of ‘justice’ this is to be understood in its wider sense of morality in general, rather than the political arrangements of a particular state
104 I believe it is. If it is true that we do in fact stand in a certain relation to one another – that there is a universal manifold of persons to whom we can appeal to when trying to determine whether it is true that I ought (morally) to give Sylvia the money I owe her, then such an account would not seem to be prone to the problems facing Darwall that were discussed above. Issues of consent would not arise since the relations are taken to exist prior to any implicit or explicit claims being made. That is, if it is the case that it is impermissible to inflict pain expect in special cases where permission is granted, then this would be better explained by our standing in a relation to others that entails the commitment not to do so, rather than the prohibition on doing so being rooted in claims that real agents address to each other, whether implicit or explicit.
Furthermore, we would not lose the interpersonal aspect – which seemed to be at risk in Darwall’s response to Wallace – since the purportedly moral reason that we have still derives from the real relation between myself and Sylvia. And while we might stand in such a relation prior to any contractualist thought experiment, we might still need contractualism to help us determine that the statement ‘I ought to pay Sylvia back the money I owe her’ is in fact true – whether this is cashed out in terms of reasonable rejection or acceptance of principles, so the moral normativity of the statement would still derive from the contract. On such an account, it follows that not only would we have reason to be guided by such principles, but we have an answer to Hobbes’s Foole.161 The account of practical reason that necessitates the contractualist mode of thinking is consistent with the reasons we have to continue to comply with the principles derived from this mode of thinking. In the same way that I have a reason to stick to the rules of chess in virtue of being a part of the manifold of chess players, I have a reason to stick to the rules of the contract in virtue of being part of the manifold of persons who, in virtue of this inescapable aspect of our agency, are subject to it.162
If the above is accepted, then we have good reason to seek an account based on Thompson’s relational account of bipolar obligations rather than Darwall’s account that
161 To recall the discussion of Chapter 1 162 This claim – that we can think of the rules of morality as constitutive of the order we find ourselves in in a similar way that the rules of chess apply to players – is not without its critics. I will return to deal with objections to this kind of move after I fill out my positive account of what makes these the statements that represent the rules of contractualist morality true; that is, after the discussion of what Hegel can contribute to our understanding of the relation involved.
105 relies on a Kantian realm of ends to explain our standing towards one another, at least for the purposes of a contractualist ethics. The purpose of the above discussion is to establish Thompson’s account as the superior framework for thinking about what we owe to each other. But even if this is accepted, a Thompsonian approach to assessing statements like ‘I ought (morally) to pay Sylvia back the money I owe her’ still depends in it being the case that Sylvia and I do stand in such a relation – that we are part of the ‘manifold of persons’ – such that such statements are in fact true.
*
I have argued that we need an interpersonal account of practical reason to solve the problem that any contractualist ethics faces when attempting to answer the question of normativity. This must be a ‘formal’ or ‘procedural’ account (to follow Southwood’s terminology) rather than substantive – that is, the moral reasons we have, which are determined by the contractualist process, must apply to us as a result of an inescapable aspect of our agency, where this is understood to be interpersonal in nature. The previous section argued that Michael Thompson’s account of bipolar obligations, rather than Darwall’s account of the second-person standpoint, provides us with the best model for thinking about these relations. Yet it remains to be seen how we can argue for the manifold of persons that is necessary for Thompson’s account to go through. We need an argument to show that we do in fact stand in such relations to one another. It is my contention that the best place to start looking for such an argument is in the work of Fichte and Hegel. I will begin by returning to Darwall’s account, since while Darwall does appeal to a Kantian realm of ends to establish the connection between his account and contractualism, he also attempts to use Fichte to establish the sort of relation that pushes us into the interpersonal point of view that is required to answer the question of normativity. While Darwall’s appropriation of Fichte is very useful in suggesting how a German idealist account of agency might show how taking the reasons of the other seriously is constitutive of agency, which is, as I have argued, necessary for a contractualist ethics to answer the question of normativity, I claim that there are still issues with this approach. The rest of this chapter will be devoted to presenting this appropriation of Fichte and the issues with this approach. The next chapter will present
106 my alternative approach of invoking Hegel to try and establish the kind of relational formalism involved, and my argument for the superiority of this approach
4.4 Fichtean Themes in Contractualist Ethics
If a contractualist ethics is going to provide an answer to the question of what makes it the case that the claim ‘I ought to pay Sylvia back the money I owe her’ is true, such that this provides me with a compelling reason to pay this money back, then the contractualist must lean on what I, following Southwood, am calling a formal interpersonal account of practical reason. A formal interpersonal account of practical must be such that it is an inescapable aspect of our agency as such that we stand in a certain relation to each other, in this case Fred stands in a certain relation to Sylvia. The Thompson model outlined above gives some indication of how this might work. What makes it the case that Fred ought to pay Sylvia the money back is the fact that they are agents whose agency is defined by the relation they stand in to one another. Just as in the legal case the relation of the two parties to the accident (citizenship) entails that one owes the other compensation, so in the moral case the relation Fred and Sylvia stand in to one another entails that they are subject to the rules derived from the hypothetical contract, namely (let’s assume) that it is true that Fred ought to pay Sylvia back the money he owes her.
Of course, in the legal case the existence of the relation is confirmed by the reality of the state that confirms the citizenship of the two parties and enforces the rules that govern this citizenship. In the moral case we cannot, prima facie, appeal to the existence of any such entity. What is needed is an account of why agents like Fred and Sylvia do in fact stand in such a relation; is there a moral community that can play a similar role to the state in the legal case? The jurisdiction of such a relation will be far wider than the community presupposed in the legal example. What is needed is an account of why agency as such is relational in this way, and how this fact has implications for what we owe to each other.
I want to argue that such an account can be provided by drawing on what I am calling the ‘Hegelian Thought’ about agency and mutual recognition. This will be spelled out in
107 much more detail in the next chapter, but briefly the idea is standing in a relation of mutual recognition is a necessary and constitutive feature of agency as such, and this has implications for how we treat the reasons of the other. This in turn provides a motivation for adhering to the principles that would be agreed in a hypothetical bargaining process in a way that is similar to Scanlon but, I argue, on a firmer footing.
But first I want to motivate this move by looking at a similar attempt to invoke the insights of German Idealism in order to ground an account of the obligations we have to one another. This means returning to Darwall’s The Second Person Standpoint. Darwall’s discussion of what he calls ‘Fichte’s Point’ provides an illuminating account of how the German Idealist account of freedom and recognition might be appropriated for a contractualist ethics. My earlier criticism of Darwall’s approach remains and I think there are reasons to prefer Hegel’s account of freedom and recognition to Fichte’s. Nonetheless, I will now present Darwall’s account as an example of how this appropriation might work, before moving on to some criticisms.
Darwall spends much of the book establishing the conceptual connections between his ‘circle of irreducibly normative concepts’. Addressing a second-personal reason to another entails that the other has second-personal competence, which entails certain relations of responsibility and authority. It is only in the discussion of ‘Fichte’s point’ that Darwall moves to argue that we do in fact stand in such relations to one another.163 As discussed above, Darwall’s suggestion was that once this is established, along with the conceptual connections between his circle of irreducibly normative concepts, then determining the content of the demands of this relation will involve a test of reasonable acceptance or rejection – i.e. the contractualist solution.
Before turning to how Darwall makes use of Fichte, it will be useful to give a brief overview of Fichte’s account, in particular the role of the other in establishing agency and the role of recognition in this account. As was the case with German Idealism in general, the overarching concern was the status of knowledge.164 The relation between Fichte’s practical philosophy and his theoretical account of knowledge are somewhat complicated by his prioritization of the former over the latter. Suffice to say that his general project
163 Darwall (2006) pp. 252-276 164 See Dudley (2007) as an example of an overview of this general project
108 was to provide a foundational account of the necessary conditions of self-knowledge, and this account involves the transcendental deduction of standing in a certain relation to the other, which in turn forms the basis of Fichte’s account of ‘natural right’.165
Fichte conceives of the I/self as a self-positing activity. The idea of an I/self being ‘self- positing’ rests on a distinction Fichte makes between the self as activity and the self as entity. The self is defined by activity, but in order to become self-conscious is must posit itself as a self.166 The self is conceived of as activity because Fichte views the self as by definition thinking and thinking as always active. Fichte comes to this view as a result of his attempt to derive the possibility of knowledge from a minimal starting point. While he, following Descartes, views something akin to representation, understood as something like an awareness of the awareness we have as the most suitable minimal starting point, it is only this direct awareness of awareness that we have access to. If the minimal starting point of the self is this awareness of awareness, then the idea that ‘thinking is acting’ seems more plausible. Furthermore, it is not too much of a stretch to view the positing of a being that is engaged in this act of thinking/acting – a ‘me’ or self that is doing it.167
It is necessary to understand this conceptual separation between self-as-activity and self- as-posited because it is Fichte’s account of the practical unification of the two that brings the other and recognition into the picture. The idea is that certain essential activities of the I itself—in particular, those though which it determines itself as an individual agent— presuppose interaction with other I’s and therefore the employment of concepts of an I that is not our own as a condition of possibility of our even own self-awareness.
165 Since Darwall only attempts to use Fichte’s insight into agency and recognition as a means to justifying a quite distinct normative ethics (and I am attempting a similar move with Hegel), I will not get into the details of Fichte’s account of natural right. The following discussion draws heavily on Wood (2014b) and Wood (2014c) 166 “That whose being or essence consists simply in the fact that it posits itself as existing, is the self as absolute subject. As it posits itself, so it is, and as it is, so it posits itself; and hence the self is absolute and necessary for the self” (Fichte 1982 p.98) Fichte goes on, “To explain, one certainly hears the question proposed: What was I, then before I came to self-consciousness? The natural reply is: I did not exist at all; for I was not a self. The self only exists insofar as it is conscious of itself.” (Ibid) 167 C.f. Wood (2014b) p. 202
109 Fichte’s strategy is to show that the self-positing I is an ‘incomplete thought’ – that is, there are a series of other thoughts that are necessary for its completion. And it is Fichte’s contention, which he elaborates through a series of transcendental steps, that among these necessary thoughts is awareness of other I’s – “No thou, no I; No I, no thou”168
We start from the self-positing I as activity, and then move to the claim that forming a concept of this activity means distinguishing it from an opposing activity—that of the object of the ‘not-I’. Thus far, the not-I is indeterminate. An active I ‘finds itself’ only as willing and this willing takes the form of striving against the material world. In striving against the material world, we ‘find’ ourselves as individuals. But this finding itself is still an incomplete thought – it is not possible without an experience, as an object, of the free activity constituting my individual I. ‘As an object’ can be understood in terms of the above discussion of our conception of ourselves as selves. And this experience, according to Fichte, is not possible without the instantiation in my consciousness of a distinctive concept, which he calls the concept of a ‘summons’ (Aufforderung). And this is a concept, he argues, that requires for its completion a reference to its origin in a rational being other than myself.
For Fichte, agency requires the unity of the self-as-activity and the self-as-posited. Only then can the self conceive of itself as a self, which is the minimal starting point for the self as agent.169 The connection between this conception of the self and the self as a free agent rests in Fichte’s claim that one’s individuality depends not on facts that distinguish me from others, but in possibilities of acting through which I actively determine who I am – that is, my individuality is fundamentally normative. Part of what it is to be a free individual
168 Fichte (1982) pp. 172-173. There is some discussion on the practical implications of such points. One’s stance on these is determined by what one takes Fichte (and Hegel when he makes a somewhat similar move) to be doing. See Stern (2012) on the interpretation of this kind of move as a kind of argument against solipsism. If true, then this might present a difficulty for adopting these thinkers for practical purposes. Stern argues that this is a more plausible reading of what Fichte is doing than Hegel. Since Darwall treats the Fichtean move as having practical implications, and since I will argue that Hegel provides a better foundation for such purposes, I won’t pursue this here. 169 I do not intend to make any strong claims about the existence or otherwise of metaphysical freedom. Whether or not such a freedom exists, the conception of oneself as free – as being an agent capable of acting on reasons – seems both necessary (if not sufficient) to such freedom and an essential part to how we think about normative ethics. Thus, I believe I can leave this question to one side.
110 is to have the capacity to choose between possibilities for reasons. This entails the capacity to subject oneself to a norm; that is, being the kind of agent who is able to assess the range of choices concerning what they are to do and make a decision is to say one acts for reasons. To say that one acts for reasons is to say that one has the capacity to subject oneself to a norm, for acting for a reason is implicitly to commit one to something like “in situations where A holds, I will ϕ in virtue of X.”
Fichte thus connects the concept of the individual I with the concept of an end or goal (Zweck) of activity.170 It is through this that we make a transition from ‘determinability to determination’ – from plural possibilities to the free determination of who I am.171 This transition involves setting an end for myself, constituting my reason for making this transition. But this concept of an ‘end’, Fichte claims, can only present itself to me in the form of a ‘summons’, and the concept of the summons requires the actual existence of another rational being as an explanation.
Before proceeding it is important to note that ‘summons’ doesn’t, for Fichte, have the coercive connotations associated with the English translation. Wood notes that ‘invitation’ might be a better fit, since it is more suggestive of the fact that we are not, on Fichte’s account, compelled to comply with it.172 But as ‘summons’ is the usual translation, and since this is the term Darwall uses in his argument, I will stick with that.
How is a summons a transcendental condition of free activity? To understand this, we need to consider the role of the other in our capacity to act for reasons. There are two fundamentally different ways in which the world might be given to us as agents who act in it: through causal necessity, which restricts our freedom, and through our (normative) reasons for acting. Fichte’s transcendental method moves us from the positing of the I to the necessary positing of the not-I as co-constitutive of it. The I-as-activity strives against the material world of the not-I. But if the not-I of the material world had no other agents, then the only thing that would hinder the striving of the I would be causal
170 I interpret ‘end’ or ‘goal’ here as being compatible with teleological accounts of such activity. The end in question might be to act so as to do one’s duty. The important point is that the agent acts for a reason, thus subjecting herself to a norm. 171 Wood (2014b) p. 206 172 I mean that we are not compelled in the sense that English people are likely to feel compelled when they receive something they identify as a summons.
111 necessity. The not-I as the other provides a different sort of hindrance, since they will inevitably issue competing claims on the material world. These claims correspond to an invitation/summons to the I. At the present stage we can leave aside the precise response the agent may have to this summons. The important point is that it is through this summons that the agent becomes aware of her capacity to act for reasons at all.
The summons is required in order for the agent to become aware of her capacity to act for reasons at all. But this can only be the case if the agent recognizes this other as having the capacity to issue a summons to respond to these reasons. The condition for these invitations is the mutual recognition of each other’s rationality:
“Thus the relation of free beings to one another is necessarily determined in the following way, and is posited as thus determined: one individual’s knowledge of the other is conditioned by the fact that the other treats the first as a free being (i.e. limits its freedom through the concept of the freedom of the first). But this manner of treatment is conditioned by the first’s treatment of the other; and the first’s treatment of the other is conditioned by the other’s treatment and knowledge of the first, and so on ad infinitum. Thus the relation of free beings to one another is a relation of reciprocal interaction through intelligence and freedom. One cannot recognize the other if both do not recognize each other; and one cannot treat the other as a free being if both do not mutually treat each other as being free.”173
Fichte’s cashing out of this idea is that we can only be fully rational creatures in a social context. He emphasises the role of education as a kind of concrete summons and the role of community. Yet the ethical content that Fichte derives from this is a rather thin account of natural right, similar to the realm of property rights that Kant establishes in the Metaphysics of Morals. This is due, in part, to his narrow understanding of freedom. I will return to this when I discuss Hegel’s criticisms of Fichte.
*
173 Fichte (2000) p. 42
112 In order to see how this idea of the summons is relevant to thinking about the question of normativity we should return to Darwall. Darwall appropriates what he calls ‘Fichte’s Point’ by interpreting the summons along the lines of the reactive attitudes that implicitly provide second-personal reasons to the other. The idea is that the kind of address involved in the reactive attitudes that we engage in presupposes the mutual second- personal competence (read: agency/freedom) that is required in order to address such demands. Such attitudes/demands can be understood as summons – I ‘summon’ (or ‘invite) you to remove your foot from mine – and this means that these presuppositions hold. This, in turn, seems to suggest that we must not make demands that violate each others’ second-personal authority – to make a demand which, if met, would encroach on the other’s second-personal competence.
In Darwall’s hands, Fichte’s point is used to argue that practical freedom is not simply the ability to set ends but also to engage with the other second personally which means committing the agents concerned to limiting their external freedom through the ‘principle of right’:
“Although agency is no doubt assumed in the background somehow in any deliberation, it is only from the second-person standpoint that the addresser’s free agency (and that of addressees) must be posited, that is, brought into their reasoning as a premise.”174
Having laid the groundwork for the relationship between the various normative concepts of authority and accountability, Darwall invokes ‘Fichte’s point’ as a means of ‘breaking into the Kantian circle.’175 The upshot of this, according to Darwall, is that the authority to make these kinds of claims is always subject to a test of ‘reasonable’ acceptance or rejection. Darwall’s cashing out of this involves invoking the idea of a Kantian realm of rational beings, and suggests that such a line of argument might be useful as providing a foundation for a Scanlonian conception of contractualism. I have discussed what I take to be the limitations of Darwall’s suggestion above. However, the metaphor of breaking into the circle of normative concepts is illuminating in that a similar move is required in order to explain the existence of the relation that is required in order to explain the truth of the statement “I ought to repay the money I borrowed”. The Fichtean thought might
174 Darwall (2006) 256 175 Darwall (2013) 245
113 be used to show how this relation is presupposed by our interactions with one another; that is, our interactions with each other demonstrate that this relation exists. Our reactive attitude are, in Darwall’s hands, a kind of Fichtean summons.
Here, I want to turn to the limitations of appropriating Fichte as the means of establishing the kind of relation that can form the kind of formal, interpersonal account of practical reason that a contractualist ethics requires, and in doing so motivate the move from Fichte to Hegel as the German Idealist who might provide a better basis for such an account.
For both Fichte and Hegel, recognition is a key concept in understanding the agent’s relation to the other and thus agency as such. Fichte makes use of the idea of the summons to explain the importance of the concept of recognition. Hegel does not rely on this notion of the summons. I will give an account of the kind of freedom implied by Fichte’s summons and an account of the theory of recognition here. Hegel’s account will be presented in the next chapter.
As described above, for Fichte the summons can be understood as an invitation to respond to the claim of another agent on the material world. This summons results in an awareness that the claims of the agent are constrained not only by the causal necessity imposed by the physical world, but also the competing claims of other agents. It is this awareness that is constitutive of our status as agents who respond to reasons. But this can only happen if we recognize the other (agent making this competing claim) as having a status that is, in some sense, similar to us – namely, as the sort of agent who can and does make such claims. At the very least, the agent is forced to reformulate their claim on the material world in terms of reasons that the other could (in principle) accept.
We can think of this in terms of Darwall’s repeated example of you standing on my foot. My reaction to this – whether in the form of a rude exclamation, pointed stare, or polite request to desist – can be understood as a summons (or invitation) to reconsider the position of your foot. But this only works to the extent that I recognise you as the sort of agent who has the competence to respond to such claims, and your recognising me as the sort of agent who has the authority to provide you with second-personal reasons which merit a response. Darwall does not make recognition central to his account, but the
114 normative concepts he uses – accountability, authority &c – do appear to be consistent with the kind of recognition outlined in Fichte.
But notice that this way of thinking about what Fichte is doing is consistent with the reasons of the agent being of the egoist variety. ‘My claim on the material world is more important than yours’ is a viable reason, on the basis of what has been said so far. It might be that my request provides you with a reason to lift your foot off of mine, and this might be rooted in your recognition of my status as a reason-giver whose request merits a response. But your reason to comply with my request might be given in virtue of the strategic calculation that it is better not to escalate the situation arising from the competing claims on the material world. By refusing to move the position of your foot, you might render me less cooperative and this might in turn hinder your future claims on the material world. The other is recognized in Fichte’s account as a rational agent. But this recognition as a rational agent only extends to the ability to make such claims on the material world. Standing in such a relation of mutual recognition might mean recognizing the other is the kind of being that can make such claims, and such claims are limitations of the agent’s own freedom.176 And it might be further recognized that asserting one’s own freedom might involve concessions to others to assert their freedom, where this is understood as making claims on the material world. Such recognition might lead to a certain level of cooperation.
This kind of reasoning will be familiar from the discussion of Hobbesian contractualism in Chapter One. It was argued there that such an account of practical reason was unsuitable as a basis for answering the question of normativity. Alternatively, we could seek a way of adjudicating between such competing claims that avoids these problems. This is Darwall’s strategy in his attempt to connect the claims that he takes to be implicit in our reactive attitudes to the Kantian realm of ends, the limitations of which as a basis for contractualism were discussed above. By contrast, my approach is to see what happens when we take a closer look at the kind of recognition that is assumed to be working here in the necessary relation that is posited between the agent and the other. It is my contention that a Hegelian understanding of the concept of recognition will better
176 In this, the agent making such claims in not a huge advance on the constrained maximizer discussed in Chapter 1
115 enable us to answer this question that a Fichtean understanding of the concept does in virtue of the reasons it gives us for the necessity of the other agent ‘counting’ in a specific way. The other counting in this specific way entails that the kind of Hobbesian instrumental attitude towards the reasons and interests of the other is not compatible with the account of agency assumed. It is to Hegel’s approach that we turn to in the next chapter.
4.4 Conclusion
The purpose of this chapter has been largely to set the scene for the positive contribution that will be offered in my investigation of moral contractualism’s question of normativity. The first part of the chapter sought to recap the negative arguments made against various current contractualists attempts to answer this question. The various failures to do so, it was argued, are instructive in suggesting the kind of practical reason that must be relied on in order for a positive answer to the question to be supplied.
The rest of the chapter considered what such a positive approach might look like. It considered two approaches thinking through the kind of relation that such an account of practical reason would involve. It did so by looking at the accounts offered by Darwall and Thompson, arguing in favour of the latter’s as a better basis for moral contractualism. But such a framework is not much use unless we have a basis for presuming that we do, in fact, stand in the kind of relation that it assumes. I thus turned to Darwall’s attempt to invoke Fichte to argue that we do, in fact, stand in the relation that such an account of practical reason requires. I highlighted some shortcoming with invoking Fichte in this way.
Despite my criticisms, I think that Darwall and Southwood are correct to posit the kind of relation they do, and the former is right to look to German Idealism as a way of substantiating this relation. But I want to suggest that there is a better alternative. It is to this alternative that the next chapter will be addressed.
116 Chapter 5: A Hegelian Solution?
The previous chapters have sought to establish the need for a formal interpersonal account of practical reason if moral contractualism is going to answer the question of normativity. It has also been argued that invoking the ideas of German Idealism is a plausible way of positing the existence of this relation and working through the implications of it. This chapter picks up this idea and investigates whether Hegel can be adopted in this way. A brief overview of this kind of argument that is going to be made will be provided. This will be followed by an exposition of the early Hegel’s account of recognition, including a discussion of the famous Battle of Life-and-Death and subsequent Lordship and Bondage. It will then be argued that the insights derived from Hegel’s account can be appropriated to suggest the form of practical reason that a contractualist ethics must appeal to. Such an appropriation of Hegel for such purposes will no doubt be controversial, so I will spend some time at the end of the chapter dealing with objections to such a move.
5.1 Introduction
If the arguments presented in the previous chapters are accepted, then any contractualist ethics which hopes to answer the question of normativity must be grounded in a formal interpersonal account of practical reason. Formal, in that it does not depend on any substantive account of value to provide agents with reasons for action – the reasons that agents have are constitutive features of agency as such. Interpersonal, in that the reasons of the other form a constitutive element of the factors that go into determining the course of action that the agent has (most) reason to do.
Perhaps the best way of illustrating the kind of argumentative move that is being made here is through an analogy with the kind of instrumental account of practical reason discussed in Chapter 1. The motivation for attempting to ground a contractualist ethics on such an account of practical reason runs something like:
1. Instrumental practical reason is a necessary feature of agency
117 2. Ethical theories should be as parsimonious as possible
3. Therefore, if a contractualist ethics can be grounded in instrumental practical reason, we should adopt this rather than a less parsimonious account.
The first premise is uncontroversial. On any conception of what it is to be a being with agency, such a being must be capable of determining the best – as in most efficient, least resource-depleting – method for achieving a given aim. It is difficult to imagine an account of agency which does not contain something like this criterion, at least any account of agency which includes at least some reference to rationality. The second premise seems at least prima facie plausible. If we are faced with an ethical theory A and ethical theory B, which are similar in most respects but A includes less (metaphysical) commitments than B, then it seems we have at least a prima facie reason to see if ethical theory A can provide a plausible account of what we ought to do.
Chapter One demonstrated the shortcomings of a contractualist ethics that is grounded in such an instrumental conception of practical reason. However, it does provide us with some insight as to what a formal account of practical reason must look like. Chapter 2 investigated how a less parsimonious but more in-tune-with-our-intuitions account of practical reason might fit into a contractualist ethics. Here, it was argued that Scanlon’s contractualism faced an unpalatable dilemma of being unable to explain how the (generic) reasons of one can become the (generic) reasons of the other, or committing to a substantive value conception of practical reason that left him unable to adequately answer the circularity and redundancy objections that have often been levelled at him. Nonetheless, it was argued that his idea of the hypothetical contract as a means of mediating between the objective reasons that agents already have in order to determine how the reasons of one can become the reasons of the other is a useful heuristic for thinking about the contractual process. And chapter 3 argued that while Southwood’s account of deliberative contractualism provided us with an idea of what a successful answer to the normative question based on a formal interpersonal account of practical reason might look like, he does not provide sufficient grounds for explaining why we should adopt this, and thus his solution looks somewhat ad hoc. Nor did he convincingly show why his account of contractualism does not collapse back into the kind of instrumental account advocated by Gauthier, with all the problems this entails.
118 This chapter will seek to overcome some of the issues faced by previous iterations of contractualism. In order to do this, I will introduce what I will call the ‘Hegelian Thought’. My thesis is that what Hegel has to say about agency and mutual recognition can give us independent grounds for supposing that a formal interpersonal account of practical reason is not objectionably ad hoc.
In order to understand the move being made here it is worth recalling the discussion of Thompson in chapter 4. Thompson’s way of framing the matter is through an account of bipolar normativity.177 Thompson’s concern is with what makes statements like “I ought to pay Sylvia back the money I owe her” true, where the ‘ought’ in this case is to be understood in moral terms. His suggestion was that we should approach this question in a way that is analogous to the legal ‘should’ interpretation of “I should pay Sylvia back the money I owe her.” The statement is true in virtue of Sylvia’s status as a person, where ‘person’ is understood as a relational concept that entails being part of a particular jurisdiction. But whereas determining the truth or falsity of the legal interpretation of the sentence is relatively unproblematic – the statement is true because the action takes place within a set of circumstances where (say) a particular state has legal authority – the moral case raises more difficult questions.
This idea might be appropriated for a contractualist ethics if it is the case that being an agent as such is interpreted as a relational concept that entails that those who fall under this concept are part of the jurisdiction of the requirements derived from the hypothetical contract. This would mean that what we have reason to do is determined by the relation we stand in to others, which is turn can be thought of as being mediated by the hypothetical contract. In order to make such a claim, we must make a plausible case that agency is relational in this sense and this relational agency is such that it entails facts about what we owe to each other. The contention of this chapter is that an understanding of Hegel’s account of the centrality of mutual recognition to freedom and reason allows us to make such claims for agency. I call this the Hegelian Thought.178
177 A reminder to the reader that I am conveniently ignoring the ethical issues that an account that makes ‘bipolar normativity’ so central faces – especially responsibilities to non-human animals and the environmental. 178 The phrase ‘Hegelian thought’ is designed to indicate that while this idea is inspired by Hegel, it might be a bit of a stretch to define the move being made here as advocating a kind of ‘Hegelian
119 The claim is that an understanding of agency informed by the Hegelian thought allows us to posit the formal interpersonal account of practical reason that is required for a contractualist ethics to answer the question of normativity. A rough sketch of an argument demonstrating how this idea fits into a contractualist ethics would be something like the following:
(1) Agency entails freedom
(2) Freedom entails mutual recognition
(3) Mutual recognition entails answerability
(4) Answerability entails complying with the requirements of the hypothetical contract.
Premise (1) is relatively uncontroversial, but the particular understanding of agency and the Hegelian conception of freedom involved will be outlined below. If agency entails something like a Hegelian understanding of freedom, then the next step is to show that this kind of freedom is only possible when the agent is standing in a relation of mutual recognition with the other (2). Standing in this relation with the other necessitates responding to the reasons of the other in a certain way (3). And, it will be argued, responding to the reasons of the other in a certain way is equivalent to abiding by the standards of the hypothetical contract (4).179
I will proceed by outlining Hegel’s account of agency and recognition in his Encyclopedia and Phenomenology of Spirit. The aim here is to explain Hegel’s account of the nature of the two concepts and the connection between the two. I will attempt to show why Hegel’s account of this is superior to that of Fichte’s (as appropriated by Darwall) discussed in the previous chapter. I will then turn to how such an account might be thought of as
contractualism’. The insight taken from Hegel here is based on one small part of his much larger system. Furthermore, Hegel himself was quite hostile to the idea of the social contract. I will return, briefly, to the question of how Hegelian the ‘Hegelian’ solution I am offering is below. Thanks to John O’Neill for a useful discussion around this naming issue. 179 It will be noted that the move from 3 to 4 is equivalent as Scanlon’s understanding of adhering to the contract as being equivalent to living in conditions of mutual recognition with others. But the steps leading up to 3 are distinct, and designed to avoid some of the objections that Scanlon faces as discussed in chapter 2.
120 providing the kind of practical reason that can fit a contractualist ethics. I will then turn to some objections to this kind of approach and attempt to offer responses to them.
5.2 Hegel’s Dialectic of Recognition
For Hegel, as for Fichte, standing in a certain relation to the other is necessary in order to stand in a certain relation to the self. Both take this relation to be constitutive of the agents’ status as reasoning creatures. That is, the agents’ status as reasoning creatures entails a certain kind of freedom. An important component in this status as reasoning creatures is autonomy. Hegel does not treat autonomy as exhaustive of his conception of freedom, but it is a necessary part of it. Hegel treats autonomy in much the same way as Kant: it is conceived as a break from nature and natural causation. It is negative in the sense of being free from external influence, and positive in terms of being self- determined, spontaneous, and independent.180
However, for Hegel autonomy is a mediated concept, and it is mediated by the relation that the agent stands in to the other. More precisely, true autonomy requires that the agent stands in a certain relation to herself – she must treat the self-as-other. But adopting such a relation to the self requires standing in a certain relation to the other. Before turning to the specifics of this relation and its consequences, we should first look at why Hegel holds that autonomy must be mediated in the first place. That is, why not treat autonomy as something inherent to the agent as an individual?
To address this question we must turn to Hegel’s famous discussion of Lordship and Bondage. But first, a brief word on a difference of method between Fichte and Hegel. Fichte proceeds transcendentally from that which cannot be doubted (like representation or awareness of awareness) through a series of derived necessary steps which leads us to claims about our agency and our relation to others. Hegel proceeds by way of contradiction. That is, he begins with how he takes ‘ordinary consciousness’ to face the world and themselves and then proceeds through a series of steps to show how each
180 Williams (1997) p. 81
121 position that ordinary consciousness takes is self-defeating, but in being self-defeating it points to a new form or ‘shape’ of consciousness.181
The simplest way to grasp this method is through his discussion of sense certainty. In the beginning of the Phenomenology Hegel imagines what might be thought as the most basic explanation that the relation between consciousness and the external world might be. Sense-certainty can be understood as the claim that consciousness has immediate access to this external world. It can simply point to ‘this’ experience of colour or ‘that, over there’. But this attempt at describing the relation between consciousness and the external world is undermined by the way it attempts to do so. Consciousness has to make use of a conceptual apparatus that involves a kind of mediation between consciousness and the world of experience in order to articulate what it claims to be its unmediated experience of the world.
Sense-certainty is cited here merely as an illustration of the method deployed. The precise theoretical aims of the Phenomenology need not detain us here. It is a complex account of the conditions necessary for consciousness to make claims about itself and its relation to the world, both at an individual developmental level and a socio-historical level. And, in fact, the account of freedom and recognition that is most relevant for the purposes of this thesis is spelled out in more detail in Hegel’s Encyclopedia. But it is in the Phenomenology where this method of a shape of consciousness undermining itself and moving to the next stage is articulated most clearly. The self-defeating nature of standing in a certain way towards the world (and more relevantly for our purposes, the other agent) will become apparent below.
As is the case in Fichte, for Hegel, a conception of oneself that goes beyond the purely subjective is taken to be necessary for freedom and agency. This is to say that the self must stand in a certain relation to herself. Hegel’s way of working through this idea and the various possible ways of conceiving it is through his introduction of the “I = I”, or
181 This is part of Hegel’s general method of highlighting and then working through the intractable difficulties that present themselves to consciousness as it tries to establish the relation between itself and the world. See Stern (2002) p. 23 for more on this.
122 the self-identity of ego, described by Williams as ‘the basic principle of German idealism’.182
This notion of the “I = I” might seem puzzling to contemporary analytic philosophers. Yet for Hegel, this is not merely a tautology. Rather, it describes the relation of a self to itself (this will be elaborated below). Nor is it a static relation set for all time. Rather, there are various possible forms that this relation can take, and ultimately in order to achieve the self-relation of a full, positive freedom, the agent must (conceptually) pass through three stages. Before explaining these stages I will address what is meant by a relation of a self to itself (or self-relation) with reference to the kind of practical reason being aimed at here and Terry Pinkard’s very helpful elaboration of this idea in his account of Hegel’s conception of History.
Part of the idea that the kind of practical reason we are relying on is formal is that the reasons that we have are capable of presenting themselves or ‘showing up’ (to borrow Pinkard’s terminology) in virtue of our agency itself.183 ‘Showing up’ is Pinkard’s gloss of the kind of idealism that can be ascribed to Hegel. This is not a claim about the fundamental ontology of the world being exclusively mental, but rather the fact that the way certain things present themselves or ‘show up’ to the agent is co-constituted by the type of creature they are. What ‘shows up’ for the particular being concerned depends on the purposes of the being in question. Lettuce shows up for rabbits as food because of the kinds of creatures that rabbits are, while rabbits show up as food for foxes because of the kinds of creatures that foxes are.
What distinguishes humans from non-human animals is rooted in the kinds of things that ‘show up’ for the former. Things show up for humans that do not for non-human animals – states, art, ethical requirements and so on. And the fact that such things as states, art and ethical requirements show up for humans is rooted in the particular relation that
182 Williams (1997) p. 70 183 Pinkard (2017) p. 7. I want to stress the ‘capable of’ part of this sentence. It is naturally part of any contractualist ethics, and indeed any account of practical reason, that the reasons agents can fail to present themselves to the agent. There is a difference between having a reason and perceiving oneself as having a reason. However, I take it to be a necessary condition of being subject to a moral requirement (and thus having a moral reason) that one is capable of conceiving oneself as having such a reason. In this sense, it is necessary that we are the kinds of agent for whom reasons are capable of ‘showing up’ for us.
123 humans, as self-conscious creatures have to themselves. And the particular relation that self-conscious beings have to themselves involves what Kant would call an apperceptive life – or, in a more Hegelian idiom, we are self-conscious animals in virtue of falling under the concept of self-conscious life by bringing ourselves under that concept (i.e. ‘self- conscious life’) – that is, ‘the concept giving itself its own reality’. As Pinkard notes, “This conception of subjectivity’s apperceptive self-relation comprises more or less the ground floor of Hegel’s metaphysics of subjectivity”.184 Furthermore, it is in the very nature of this self-consciousness that we become different kinds of creatures – i.e. rational agents.
So, the “I = I” is a self-relation, namely a relation of consciousness to itself (i.e self- consciousness). And the nature of this self-relation is what makes us rational animals – i.e. animals that respond to reasons. But as is always the case with Hegel this relation is not some ahistorical transcendental truth about human nature, but an achieved socio- historical state. It is here that we can return to the various forms that the relation can take and the conditions that lead to these various forms.
The fact that the self stands in a certain relation to itself is a necessary condition for agents to realise their capacity as rational creatures but, according to the Hegelian, this fact is not sufficient. This relation can take various forms. Or, as is presented in the Encyclopedia, it must pass through three stages, and the agent only achieves the fullest, most positive state of freedom and therefore rationality when they reach the stage where this self-relation is mediated through standing in a relation of mutual recognition with the other. We must conceptually pass through the ‘lower’ stages before attaining ‘universal self-consciousness’ or ‘objective spirit’ and thus being an agent subject to universal norms. The three stages as described in the Encyclopedia185 are (to be elaborated more fully below):
184 Ibid., p. 11 185 While the dialectic of the Life and Death Struggle, followed by Lordship and Bondage in the Phenomenology is the most well-known account of the failure of mutual recognition in Hegel it is in the Encyclopedia that we get a fuller picture of what overcoming the situation of dominance and servitude will involve. See Williams (1997) p. 74-75 for more on this. Williams argues that the Phenomenology version of the dialectic of Lordship and Bondage has been overrepresented in accounts of Hegelian recognition. While important, it is ultimately about the failure of recognition, or an insufficient recognition. More relevant for my purposes is the fact that the Encyclopedia
124 1. Individual (or desiring) self-consciousness. Here the interaction with the world and the other is mediated by desire. This stage of consciousness’s relation to itself is characterised by “the certainty of itself as the being in the face of which the object has the determination of something only seemingly independent but is in fact a nullity.” 186
2. The stage of the relationship of one self-consciousness to another self-consciousness. Self- consciousness is no longer merely individual self-consciousness, “but in it there already begins a unification of individuality and universality.”
Here, there is an encounter with the other, and this encounter is a necessary step in order to objectify oneself and become conscious of freedom. While the method for getting here is different, the kind of freedom and reasoning is akin to that in Fichte. “The other makes possible the separation of autonomous freedom from mere desire.”187
3. Universal self-consciousness. Here, self-consciousness ‘relinquishes its particularly’ and ‘becomes reason’.188 In the terms outlined above, this is the point where the other comes to ‘count’ in the way that is, I claim, required by the contractualist.
Before going into more detail on each stage a caveat is in order. Such a reading would suggest that the first stage, where the agent’s relation to the object is mediated by desire, is a stage that each individual must go through in order to attain true freedom. This might in fact be the case, but it is irrelevant for the purposes of the argument being made here. Moreover, it is not what Hegel was attempting. As Hegel makes clear later on when discussing the battle of life and death, he is working through the implications of what would happen were the institutions, such as the state, that we take for granted, taken away. In this case, given that the purpose is to argue for an account of practical reason provides a clearer account of what overcoming this failure would involve, whereas in the Phenomenology the dialectic of Lordship and Bondage is followed by a historical account of Stoicism and Scepticism as illustrations of the differing ways that consciousness can relate to itself and the world. 186 Hegel and Inwood (2007) §425 p. 152 Zusatz 187 Williams (1997) p. 72 188 C.f. Williams (1997) p. 71
125 which takes standing in a relation of mutual recognition to be a formal, constitutive part of this account of practical reason, we can view Hegel’s procedure as an investigation of what would happen to our status as reasoning creature were such a relation absent. Thus, we are not committed to the historical or developmental veracity of this account’s depiction of the emerge of freedom and agency (if it were read that way)
In the first phase of the self’s relation to itself and its object the meditation of that relation is desire. And in its primitive sense, desire is nothing more than an urge.189 At this stage the possibility of the encounter with the other is abstracted away. Everything that appears before consciousness is a ‘nullity’ – the object does not have independence in the way that the self in question perceives itself to have. What appears before consciousness is mediated by desire, in the form of an urge, and this amounts to a decision whether or not to ‘consume’ it, which we can read as bringing under the will of the subject.190
If taken too literally it might seem that we are describing the life of some pre-linguistic human-type of species. However, to reiterate, this is an exercise in abstraction and what is being abstracted is any kind of relation to the other. It is conceivable, on this account, that the agent being described still has access to language.191 Yet any kind of reasoning capacity that she might have is mediated and determined by desire. It might be conceivable for such an agent to articulate her reason for consuming this item, rather than that. But to the extent that she engages in normative reasoning at all, it is limited to the best means of fulfilling these desires and urges.
What is more important for our purposes is that in such a state the agent does not appear to be capable of possessing what Kant would call the ‘apperceptive spontaneity’ that seems necessary to engage in any higher order reasoning. Or, as Hegel puts it, “in its individuality it does not correspond to the universality of its concept.”192 That is, there is nothing in its conceptual apparatus which enables it to appreciate itself as a particular kind of being (a self-conscious reasoning person). So far, this is in keeping with the
189 Hegel and Inwood (2007) § 426 p. 154 190 Ibid., Zusatz 191 It might be conceivable, yet in fact impossible. It might be the case that no agent existing in isolation could have any kind of linguistic capability. This seems very plausible to me but, to reiterate, this is only an abstraction from relations with the other; a thought experiment designed to work through the conditions required for freedom and agency. 192 Ibid., § 427 p. 156
126 Fichtean discussion above of the need to perceive of ourselves ‘as an object’ as well as a subject. This is necessary in order for the “I = I” not to be an empty tautology, but a particular kind of self-relation. Moving from this empty conception of the “I = I” (or ‘abstract identity’ in Hegel’s terms) to a kind of self-relation requires difference, and this difference requires the other.193
I will describe the encounter with the other as it appears in Hegel in both the Encyclopedia and the earlier Phenomenology.194 While the former goes into more detail regarding the aftermath of Lordship and Bondage as is ultimately more important for my argument, the latter goes into more detail about the nature of the struggle and its aftermath. I will present this as it appears in Hegel first, before outlining what I claim can be taken from this for our conception of practical reason and contractualist ethics.
The beginning of the Self-Consciousness section of the Phenomenology corresponds to the first stage outlined in the Encyclopedia. At this stage self-consciousness is a ‘simple self-identity’. Similar to Fichte’s discussion prior to the introduction of the Summons, at this stage the relation between the agent and the external object is mediated by desire and limited by causal necessity. At this point consciousness is purely ‘for itself’ and lacks ‘objective realization’. The conditions are not there for it to ‘bring itself under the concept of itself’ – which is necessary to freedom and agency. But whereas Fichte solves the issue through the transcendental deduction of the summons, Hegel shows how the shape of consciousness as simple self-identity, or desiring self-consciousness, is an insufficient account of agency through a consideration of the consequences of applying this shape of consciousness to an encounter with another consciousness like it – that is, one that appears to aspire to ‘independence’.
This encounter itself involves a radical decentering of the self. The I is objectified under the gaze of the other. But whereas Fichte treats this encounter as inevitably involving a summons and thus establishing the respective agents as reason-givers and receivers,
193 C.f Williams (1997) p. 72 194 I should point out at this stage that I am interpreting the encounter with the other as an encounter with an actual, existing other agent. John McDowell (2003) has disputed this, arguing that the Battle of Life and Death and Lordship and Bondage can be interpreted as internal to one agent. See Houlgate (2009) for a response. If McDowell were right then this might be problematic for my argument, since I want to derive moral principles from the relation we stand in to (real) others. His position seems to be in the minority in Hegel scholarship, however
127 Hegel emphasises the role of struggle. Having been shocked into the possibility of self- being other, Hegel’s agent proceeds to try and reassert itself as ‘universal’. This means establishing its freedom in the eyes of the other. That is, it seeks the recognition of the other.
It might be wondered why the agent seeks to reassert itself as ‘universal’, and why this means demonstrating its freedom in the eyes of the other. What Hegel is doing is working through the implications of continuing to mediate the relation between oneself and the external world (now including this other apparently independent object that is the other) by desire. He does not treat the encounter with the other as establishing each of us as reason-givers and receivers. Instead he works through what happens when the agent/self-consciousness continues to work in the realm of ‘pure ego’.195 But prior to the encounter with the other, the agent’s desire has been driven to consume the natural world, and has been limited only by physical necessity. In the encounter with the other, an object that appears to be capable of being consumed or submitted to the agent’s will, but is not necessarily going to be so submitted, is encountered. The desire for the recognition of the other is driven by the desire to maintain the understanding of itself and its relation to the world where the external world exists ‘for consciousness’. This means attaining recognition of the agent’s ‘universality’.
But given that (it is assumed) the other who is encountered is like the agent, they will reason in the same way. Both seek one-sided recognition of each other’s universality. The attempt to secure this one-sided recognition here results in the life-and-death struggle.
“The presentation of itself, however, as the pure abstraction of self-consciousness consists in showing itself as the pure negation of its objective mode, or in showing that it is not attached to any specific existence, not to the individuality common to its existence as such, that it is not attached to life. This presentation is a twofold action: action on the part of the other, and action on its own part. In so far as it is action of the other, each seeks the death of the other. But in doing so, the second kind of action, action on its own part, is also involved; for the former involves the staking of its own life. Thus the relation
195 Williams (1997) p. 72
128 of the two self-consciousnesses is such that they prove themselves and each other through a life-and-death struggle.”196
Both seek to assert their universality and freedom and both seek to deny this of the other. But if one actually does manage to kill the other, then the source of recognition that is sought would disappear. So the victor in the struggle would permit the loser to survive, but live in a condition of servitude.
Consuming or killing the other would deny the agent the means to secure the recognition that, it is claimed, is needed. In the dialectic the loser withdraws, fearing his own life and submits to servitude, which the victor accepts in order to secure the recognition that is sought. But submitting the other to servitude – “stripping him of his possibilities and reducing him to my own possibilities”197 – proves to be self-defeating as well. The recognition that is craved can only come from an independent consciousness, since in order to recognise the independence of the agent one must have some conception of what this independence involves, and an independent consciousness cannot be in servitude.
What has happened in the encounter is the realization that the agent’s status as an independent being is no longer limited only by the causal necessity of the empirical world but also by the response of this other, purportedly independent agent. Hegel is working out the implication of continuing to take a stance towards the external world (now including this other self-consciousness) that is mediated by desire. The battle of life-and- death and the subsequent Lordship and Bondage are the results continuing this kind of mediation. Had the battle resulted in the death of one, then it seems that the victor would
196 Hegel 1977) §187 pp. 113-114 The motivation for the move from desire to the life-and-death struggle is a matter of debate in the secondary literature on Hegel. See Stern (2002) 76-83 for an overview of the different positions. I read this passage in conjunction with the discussion in the Encyclopedia, such that we begin with the freedom that we take ourselves to have and consider what would happen if this was removed. The life-and-death struggle can be understood as the implication of not standing in the relation of mutual recognition that we take ourselves to stand in when we ordinarily think about what we owe to each other. But this should not be understood as an appeal to the substantive value of not living in a Hobbesian state of nature. As will be shown, even the resolution of a life-and-death struggle in favour of one of the participants is self-defeating in establishing the freedom required for full agency.
197 Williams (1992) p. 155
129 return to the state of desiring self-consciousness. But given that the other lives, the situation has changed. Recognition is now in play, although at this stage it is an impoverished and one-sided recognition. The agent’s freedom and agency is now mediated through the will of the other, but in distinct ways for each agent: one dominating the other’s will and one dominated by the other’s will. Neither are sufficient for full freedom. Hegel’s elaboration of this situation in the Encyclopedia is worth quoting in full:
“Here, then, only one moment of freedom emerges, the negativity of egotistic individuality; whereas the positive side of freedom attains actuality only when, on the one hand, the servile self-consciousness, liberating itself both from the individuality of the master and from its own individuality, grasps what is in and for itself rational in its universality, independent of the particularity of the subjects; and when, on the other hand, the master's self-consciousness is brought, by the community of need and the concern for its satisfaction obtaining between him and the bondsman, and also by beholding the sublation of the immediate individual will objectified for him in the bondsman, to recognize this sublation as the truth in regard to himself too, and therefore to submit his own selfish will to the law of the will that is in and for itself.” 198
We can interpret the situation of the Lord and the Bondsman as attaining a kind of negative, impoverished freedom. The master is able to impose his will, but he is unable to attain the authority that he seeks because he does not recognise the slave who is imposed upon. The slave attains a higher freedom than that of the first stage of individual self-consciousness since blindly following his urges or desires is no longer possible. But, of course, such a situation could not plausibly be described as freedom.
Again, it is worth stressing that the life and death struggle and lordship and bondage are not to be understood as phases that individual consciousness’s must go through in order to attain freedom, or at least that is not the claim here. Hegel makes this clear:
“To prevent possible misunderstanding with regard to the standpoint just outlined, we must here add the remark that the fight for recognition in the extreme form indicated here can only occur in the state of nature, where men live only as individuals; by contrast
198 Hegel and Inwood (2007) §435 p. 162
130 it is absent from civil society and the political state because what constitutes the result of this combat, namely recognition, is already present there.”199
For Hegel, the rational state is one with norms and institutions that enable individuals to live in conditions where mutual recognition is maintained. We can understand the necessity of the battle of life and death as an illustration of what would be the consequence of a situation when these norms and institutions no longer hold. The argument here is that an account of practical reason rooted in mutual recognition can be understood in the similar way. The absence of such recognition would mean an impoverished account of the kinds of reasoning creatures that we are. Just as for Hegel living in conditions of mutual recognition means living in a rational state, the contractualist can argue that following an interpersonal account of practical reason rooted in mutual recognition means complying with the principles of the contract. But this is to skip ahead.
Returning to the situation of Lordship and Bondage, we can understand this as Hegel working through the next phase of what happens when we abstract away the norms and institutions that we live in. On the encounter with the other, individual or desiring self- consciousness is forced to move to a stage where each agent knows that the authority of her reasons depends on the recognition of the other. Living in conditions of lordship and bondage means that there is an acknowledgement of the necessity of the authority of the other, but this is still lacking.
Lordship and bondage each contain elements of what is necessary for true freedom, but neither is able to attain real/true freedom.200 The Lord achieves a kind of recognition of the authority of his reasons and is able to satisfy his desires. But this recognition is only from one who is not recognised and therefore the authority granted is impoverished. Moreover Hegel, like Kant, viewed the ability to pursue desires as a less-than-full sense of freedom.201 The Bondsman is provided with the resources to act on reasons that go beyond his egoistic desires but the reasons provided here are not of the right sort. As Williams explains:
199 Ibid., §432 p. 159 200 I will return to what this kind of freedom entails below 201 C.f. Patten (1999) pp. 48-51
131 “The ‘problem of slavery’ is not its relativity to the other per se but the other to which it is relative. The ‘problem of slavery’ is really a problem of the inherent arbitrary incoherence of mastery; its choice of domination and coercion is self-subverting. The slave, in being subject to the will of the master, is not obedient to an inclusive, universal will, but only to another particular will that remains immediate, particular, egocentric and coercive.”202
Both only achieve a kind of ‘negative’ freedom. But as the above suggests what is needed for full agency is for a more ‘positive’ freedom to be in play. Inwood’s commentary on the crucial § 435 (the transition from the second to the third stage, that of ‘universal self- consciousness’) passage is instructive here:
“The other, ‘positive’ ingredient of true freedom is submission not to the will of another individual but to the ‘universal rational will that is in and for itself’… This is the laws and constitution of one’s state. It is called ‘will’ because Hegel believes, under the influence of Rousseau and Kant, that it is what everyone at bottom wills. It embodies a rationality that expresses the rational core of the human mind, so that submission to it is genuine freedom.”203
Giving up the ‘particularity’ of desiring self-consciousness and submitting to the will of the other is a condition of full freedom for Hegel. But this cannot be a one-sided imposition of will as is the case in Lordship and Bondage. Real freedom, or universal self-consciousness, occurs only when the relation is mediated by mutual recognition between the agents concerned. Again, the aim here is to provide plausible grounds for accepting a formal interpersonal account of practical reason, so the claim that it is the following of the ‘laws of one’s state’ that constituting this can be ignored. Instead, like Southwood, we can think of the rules of the contract of normative expressions of this account of practical reason.
To defend the claim that the account of recognition outlined in the Phenomenology and Encyclopedia can be appropriated without being committed to the concrete ethics of the state alluded to in citations above would be beyond the scope of this thesis. However,
202 Williams (1997) p. 78 203 Hegel and Inwood (2007) p.447
132 the claim that the kind of mutual recognition discussed here translates into the moral bargaining envisaged in the hypothetical contractual space might still strike some as controversial. It is to this that I now turn.
5.3 The Hegelian Thought Applied to Contractualism
It might be accepted that the above shows that Hegel’s account of the necessity of standing in a relation of mutual recognition to agency, and thus our status as reasoning creatures, is superior to Fichte’s. It provides a more methodologically satisfying account of the necessity of standing in this relation since it works through the alternatives and shows them to be self-defeating. It is also suggestive of the necessity of a more ‘positive’, or thicker kind of freedom than that considered by Fichte, which implies a stronger conception of what it means to stand in such a relation. But the question remains how this account can be used in a contractualist ethics. What does the superiority of one German Idealist’s account of the necessity of mutual recognition tell us about the kind of practical reason needed for a contractualist ethics to answer the question of normativity?
Here, we should recall the framework provided by Thompson. The aim is to establish the conditions under which a statement like ‘I ought (morally) to pay Sylvia back the money I owe her’ is true. If we adopt Thompson’s framework to understand this then what makes such a statement true is the fact that I stand in a certain relation to Sylvia in virtue of the jurisdiction we both fall under. What we need is a reason to think that we stand in a relation to Sylvia that entails that we stand under the jurisdiction of morality, or (by hypothesis) the principles derived from the contract.
Thompson’s framing of the issue of bipolar normativity bears a strong similarity to Dean Moyar’s discussion of the ‘Reasons Identity Condition’ in his account of the metaethical significance of ‘Hegel’s conscience’.204 Moyar frames the issue as an investigation into the conditions where an agent’s motivating reasons and justifying reasons are identical. Motivating reasons are the reasons that I act upon, justifying reasons are those that I
204 Moyar (2011) p.44
133 actually have, independent of my beliefs and desires.205 So an investigation into the conditions under which the RIC holds can be understood as determining when the content of the belief that I should return the money I borrowed to Sylvia is identical to the content of the reasons that I actually have.
Moyar invokes Bernard Williams’s discussion of internal and external reasons as an example of what a solution to the reasons identity condition might look like.206 Williams (via Moyar) would argue that the Reasons Identity Condition is explained by the fact that our justifying reasons are always (and only) true insofar as they can be traced back to our subjective motivational set, which is made up of our immediate desires and standing purposes. I am not committed to this internalism about reasons. Yet it is reasonable to suppose, following Scanlon’s approach to contractualism (see Chapter 2) that there is a set of competing claims, and the purpose of the contract is to determine the truth or otherwise of the claims involved. If it is the case that the claim that ‘I ought to pay Sylvia back the money I owe her’ is true in virtue of facts about the hypothetical bargaining process, then, in Moyar’s terms, the Reasons Identity Condition is satisfied.
Moyar’s Hegelian explanation of the conditions under which the Reasons Identity Condition is satisfied is when the standing purposes of the agent are ‘nested’ within a broader community of institutional purposes. A classical Hegelian example of this would be my reason to help my daughter with her homework being because I love her, but this reason gets its justification by being embedded within the broader context of the value of the family, which in turn is valuable in virtue of its role within the rational state.
One reason that the kind of classical Hegelian (via Moyar) example wouldn’t work, at least in the simplified version outlined above, for the purposes of this thesis is that it would raise the question of why the family (and the state) occupies this role as an institutional provider of reasons/purposes for the individual. The immediate worry would be that such an account is committed to a conception of practical reason that relies on the substantive independent value of the state. The unsuitability of such an approach was discussed in chapter 2. Yet the idea of the reasons of the individual gaining
205 Justifying reasons can be treated as equivalent to the normative reasons discussed elsewhere in this thesis. 206 Williams (1981)
134 ‘objectivity’ in Hegelian or satisfying the reasons identity condition in Moyar’s terms in virtue of being nested within a community of purposes (read: reasons) has some resonances with the contractualist approach. What is required is for the nesting of the agent’s reasons within this wider communal space of reasons to be a constitutive feature of agency itself. This would form the basis of a formal interpersonal account of practical reason and thus avoid the substantive value worry Scanlon faced (along with the truncated classical Hegelian example above).
The above discussion of Hegel should help us put some meat on the bones of the basic argument sketched above:
(1) Agency entails freedom
(2) Freedom entails mutual recognition
(3) Mutual recognition entails answerability
(4) Answerability entails complying with the requirements of the hypothetical contract.
The Hegelian claim is that standing in a certain relation to oneself is a necessary and constitutive feature of agency as such. We have seen that this involves standing in a relation of mutual recognition with the other. This is necessary for each to establish their agency and themselves as a person as such. But we still need to say more about (3) and (4); that is, what the implications of standing in this relation are for the reasons we have and our relation to the reasons of the other.
According to Hegel, for the agent to stand in such a relation to herself as mediated through mutual recognition with the other amounts to standing within a ‘shape of life’ – a concrete order of thoughts, or an ‘Idea’.207 It is within this shape of life that the agent seeks authority for her beliefs and actions. That is, she is compelled to locate herself within the ‘space of reasons’.
‘The space of reasons’ is an abstraction. It is part of the ‘realm of shadows’ that we must access when we require clarification but mostly go about our day unaware of.208 But since
207 Pinkard (2017) p.20 208 Ibid., c.f. Hegel (2010) p.37
135 the space of reasons is necessarily social, this means engaging with the other, and this will inevitably lead to conflict over what the space of reasons requires. Because the space is also a space of authority, this means taking on the burden of justifying ourselves to each other (and to ourselves), “all carried out within the idea of a rational human life”.209 So the space in which we reason is a social space. Gaining authority for an action involves taking on a responsibility for it, and this means justifying it to the other in the space of reasons.210 But supposing that this is true, we have said nothing about under what conditions this justification will be adequate to the other agent. It might be that the set of considerations that agent A thinks is a suitable justification for performing a particular action is not suitable by the lights of agent B or, indeed, the wider community.
It will be recalled that this was a concern for Scanlon’s account as discussed in chapter 2. Scanlon held that the hypothetical agents would bring ‘generic reasons’ to the contractual space, and these reasons would be based, to an extent, on the norms and beliefs that characterized the environment that they find themselves in. But when these norms and beliefs are such that certain groups end up in a subordinate position, then we are going to have difficulty providing a contractualist justification for some highly plausible moral principles.
It is worth returning to look at the relativism objection that I levelled at Scanlon in chapter 2. We considered the example of the man and his wife who go about their business in a culture where it is considered normal for women to have a subordinate position to men. This was a problem for Scanlon because the reasonable rejection test of his contractualism stipulated that what determined whether a principle could be reasonably rejected or not was were the generic reasons that were brought to the hypothetical bargaining process by the representatives of the people who would be affected by the principles derived from this process. So in the case of the man and his wife the claim that ‘the man should be kinder to his wife’ would not pass the reasonable rejection test on the basis of the generic reasons that the representatives of the relevant parties affected would have.
209 Pinkard (2017) p.23 210 To be clear, this doesn’t actually have to happen. The key question is whether the action could be justified if the agent was challenged about it.
136 As was argued in chapter 2, there seem to be two possible solutions to this unfortunate situation for Scanlon. The first is to posit that the representative of the woman has reasons to demand equal respect that she is not aware of and which carry enough weight to overturn any claims the representative of the husband might have to the maintain the terms of the relationship under prevailing social conditions. But it is difficult to see what justification there could be for assuming this given that this cannot be determined by the bargaining process itself (since the results of the bargaining process are determined by the generic reasons of the representative which are in turn determined by the prevailing social conditions of the time). Given this, it seems that if we want to say that the reasons the woman has to demand more equal treatment outweigh the reason the husband has to demand the continuation of the current arrangement, then we have to posit an independent source of value to explain this. This means that the redundancy objection re-emerges, and Scanlon is no longer offering an account morality’s foundations.
The other option is to build certain requirements into the bargaining process itself. Examples of the kind of requirements might be that the agents in the bargaining process are perfectly instrumentally rational (as was the case in Gauthier’s contractualism) or deliberatively rational (as was the case in Southwood’s contractualism). However, as has been argued, Southwood’s deliberative rationality appears objectionably ad hoc. It is not obviously a feature of agency that all rational agents require. Gauthier’s instrumental rationality is more plausible in these terms but, as was argued in Chapter 1, it is unsuitable for answering the question of normativity. Scanlon’s requirement is that the principles determined by the bargaining process be ‘reasonably’ decided, where the determination of what is reasonable depends on the objective reasons that the people affected by the principle have. This was due to his realism about reasons combined with his constructivism about ethics. The point of the contract is, according to Scanlon, to determine the relative strength of the reasons agents actually have in order to work how the reasons of one become the reasons of the other.
But in taking the Hegelian Thought seriously, we can see how this problem might be resolved. For Hegel, it is a feature of agency that we must operate within a space of reasons. For Scanlon, the contract is a means of mediation between the generic reasons that real agents have. These reasons are antecedent to the mediation process of the
137 contract. This is true of Hegel too (the space of reasons is not invoked unless a concrete difficulty gives us cause to do so). The difference lies in the nature of the mediation between the reasons that the real agents have.
Scanlon assumes that the mediation of the reasons is determined by the prevailing norms of the time and the hypothetical contractors’ commitment to finding a set of justifiable principles to govern general behaviour. But as we saw in chapter 2, this does not, on its own, explain the man’s reason to treat his wife better. Given the assumptions made, the contractual process may very well allow a principle that maintains the status quo without a ‘reasonable objection’ being sustained. Scanlon appears to be left to appeal to the substantive, independent value of living in conditions of mutual recognition. But this means inviting the redundancy objection.
But the Hegelian Thought suggests that living in conditions of mutual recognition is not some positive value that is independent of the contract but a constitutive element of our status as independent reasoning creatures in the first place. Agency involves freedom and authority. But freedom and authority are social achievements: they are only achievable through recognition. That is, they must receive their authority through the social space of reasons. Naturally, this sometimes leads to conflicts, even conflicts of good faith. Pinkard invokes Thompson to cash out what is going on here:
“One can think of them as analogous to agents belonging to two different legal systems confronting each other and each demanding of the other that the other recognize— bestow or acknowledge the binding status of—the former’s own legal system as binding on himself.”211
In the case of the husband and wife (or their representatives in the bargaining process), we can think of the agents as subscribing to different sets of norms. That is, the husband takes himself to have a reason to argue for the continuation of the present arrangement and the wife has a reason to demand a change in the arrangement. But the husband’s dismissal of the wife’s norms would be contrary to living in conditions of mutual recognition. It would be an assertion of power akin to the ‘resolution’ of Lordship and Bondage. The representative of the husband would not be subscribing to living in
211 Pinkard (2017) p.27
138 conditions of mutual recognition because he would not treat the other as ‘counting’ in the right way. In Scanlon’s terms, the principle licensing the continuation of the status quo could be reasonably rejected not on the basis of the generic reasons the representative parties have, or because it is not the best way to achieve the independent value of living in conditions of mutual recognition, but because it is contrary to the agents’ relation to themselves and each other as free, independent beings. The husband has a reason to change his behaviour because to fail to do so would be to fail to live in conditions of mutual recognition and thus diminish himself as an independent being. He would not be able to claim authority for his principles on the basis of a recognition from one who is not so recognised herself.
Thus we can return to the argument above. Point (3) of that argument held that mutual recognition entails answerability. What the Hegelian Thought shows us is that this answerability is of a particular kind; not one that views the other as a competitor for claims on the material world who it might be better to reach an accommodation with, but rather answerability as a free and independent agent like ourselves. This entails an engagement within the social space of reasons, and the imperative to find ways to live that are justifiable to one another and in keeping with this understanding of mutual recognition. However, assuming that we are looking for general principles rather than operating on a case by case basis, this is just a description of the contractual process (the Scanlonian version of this anyway). And so we have a justification for our move from (3) to (4) in the argument above. That is, we have a reason to comply with the requirements of the contract that is rooted in the relational aspect of agency as depicted here.
5.4 Objections
Thus concludes my positive attempt to use what I call the ‘Hegelian thought’ to explain why we have an independent reason to posit the kind of interpersonal account of practical reason that is necessary to answer the question of normativity, and how this would work in a contractualist ethics. Appropriating Hegel in this way is going to be controversial, so I now turn to some objections to doing so. I divide these into three broad categories. First, whether the account I offer runs afoul of the kinds of objections I levelled at previous contractualist attempts to answer the question of normativity. Second, how far
139 the account I am offering can reasonably be described as ‘Hegelian’. Third, is the form of constitutivism about the reasons we have acceptable.
Is this solution any better than the alternatives?
The first objection that I want to deal with concerns the idea that the account I am offering is not, in fact, formal in the way that I have argued is necessary in order for any account of practical reason to answer the question of normativity. This objection might proceed in two ways. The first is to dispute that the relation is ‘inescapable’ in the way I present it as being. The second is to concede this but deny that this relation entails the reasons that I take it to.
Taking the first worry first, sometimes Hegel can be read as offering an implausibly strong account of the role of mutual recognition in agency. That is, there is a temptation sometimes to view the argument as implying that there could be no self-consciousness at all without standing in this relation of mutual recognition with the other.212 If the claim that we have the moral reasons we do rested on such an account, then it would indeed be implausible. But it does not. We are not concerned with making a claim about the necessary conditions for the formation of the ego. The concern instead is with the conditions that are necessary to achieve the status of a particular kind of reasoning creature. The claim is that this is a social achievement, and it is a constitutive part of this social achievement to stand in a certain relation to one another.
This might give rise to a different kind of objection to do with strength. This is the idea since the relation is taken to be inescapable we can’t, by definition, leave this relation. If we cannot leave this relation, and what we ought to do is determined by this relation, then can we ever do what we ought not to? If this was an implication of the claim that we stand in this relation, then this would be a serious problem for the account on offer.
But this is a misreading of what it means for a relation to be ‘inescapable’. If it is true that the relation we stand in to each other is normatively inescapable, and this means that we have reason to adhere to a certain principle, then we can still choose to disregard this
212 See Pippin (2000) pp.155-6
140 principle. What we bring about by doing so is the invitation to a particular form of condemnation. This is of a kind with going against one’s own rational (in the narrow, Gauthier type of understanding of what this means) interests. I might have a certain end in mind but act in such a way as to guarantee that I will fail to bring this end about. T,here is nothing about being subject to the norms of rationality that prevents me from doing this, but I will stand accused on being irrational if I act in this way. A similar type of accountability is at work in the case of the Hegel-inspired interpersonal account of practical reason that I am advocating. It is still perfectly possible to act against what such an account of practical reason requires, but doing so will invite a certain sort of condemnation.
Is it the case that the case of relation that I am discussing is as uncontroversially a part of human experience as the kind of means-ends rationality that Gauthier uses to build his version of contractualism open? Clearly not. The idea that our agency is fundamentally interpersonal does not have the kind of philosophical pedigree that individualist accounts of action and agency do.213 Nor can it honestly be claimed to be as parsimonious as an account that only relied on the narrower form of rationality. I have to bite this particular bullet, but still argue that it is possible to use Hegel to show that given that we inevitably must reason in a social environment, to act in a way that is contrary to standing in a relation of mutual recognition is to undermine aspects of our status as reasoning creatures as such.
How does such an inescapable account of agency provide us with a reason to ϕ?
I appeal here to the idea presented by Southwood and discussed in Chapter 3 as the reasons of the contract being thought of as ‘normative expressions’ of the relation involved. In Southwood’s case, the idea was that we stand in an inescapable relation of deliberative rationality and as such are subject to the principle that would be agreed to by agents who followed the norms associated with deliberative rationality. The results of this
213 But more relational accounts of agency, autonomy, and action have attracted interest in other spheres as well. The MacKenzie and Stolar (2000) collection looks at feminist perspectives on relational autonomy. Julius (2016) considers taking the other as an a priori account of our acting together. And Laden (2005) has offered a (differently) Hegel-inspired challenge to the notion that our social reasons are reducible to individual reasons
141 hypothetical agreement could be thought of as ‘normative expressions’ of this inescapable relation.
My argument against Southwood in chapter 3 was that the inescapability of the relation was poorly defined and even if granted it it wasn’t clear that it entailed a commitment to the norms Southwood associated with deliberative rationality. I have attempted to invoke the Hegelian Thought as a better way of thinking about the inescapability of this interpersonal relation. But even if this is accepted, it might be wondered whether we can appeal to the notion of ‘normative expressions’ of this relation as equivalent to principles derived from a hypothetical bargaining process where the contractors respond to reasons based on mutual recognition.
I believe we can. To return to the example of the man and his wife. One way to think of the wrongness of the man continuing to treat his wife as subordinate is that a principle prohibiting this is a normative expression of a relation mediated by mutual recognition. A principle which did licence the continued subordination of the wife could not be described as such a normative expression, since any bargaining process looking for principles which were consistent with standing in a relation of mutual recognition would not endorse such a principle.
Even if all the above is granted, another objection might emerge. If the relation is inescapable in the way that is claimed, and the reasons we have can be understood as normative expression this this relation, is there any real need for the contract at all? In the case of the husband and the wife, the wrongness of the husband’s behaviour seems to be rooted in the fact that he would not be standing in the relation of mutual recognition if he did so, not because doing so would be rejected in a hypothetical bargaining process.
In some respects, I might be able dodge this objection by pointing to the overall aim of this thesis. I have taken it as axiomatic that contractualist ethics has something to offer and focused on the kinds of conceptions of practical reason such an ethics must rely on to answer the question of normativity. That is, it is built into the assumption of the entire ethical project that we are searching for general principles to regulate behaviour, not looking for a tailored way to deal with each situation.
142 But this is an unsatisfactory response. Appealing to mutual recognition could be itself interpreted as such a universal principle. Instead I appeal to what Scanlon calls the ‘universality of reason judgments’. As Scanlon puts it, “Whenever we make judgments about our own reasons, we are committed to claims about the reasons that other people have, or would have under certain circumstances.” 214 Versions of this can be found in Kant, Fichte and Hegel. The idea is that being a rational agent means having the ability to subject oneself to a norm – that is, a commitment that in situations like this I will consistently act like that. The account of practical reason I am advocating here holds that our agency, and thus the norms that we subject ourselves to, is irreducibly social in nature. This means that we are compelled to subject ourselves to norms that are justifiable to each other, and for the norms to apply to others in similar situations. Thus, it is a consequence of standing in such a relation that we are required to adhere to the principles that would be agreed to by agents motivated to find such principles that are consistent with this relation.
This should make clear that the kind of contractualism being advocated here should not be susceptible to the objection that it is essentially an epistemic device for discovering the best way to promote source of value that is independent of it. As the previous paragraph argues, to stand in a Hegel-inspired relation of mutual recognition just is to act according to principles that agents who want to maintain this relation would agree to. Acting according to these principles is constitutive of standing in such a relation, not the best way to realise the independent value of such a relation.
Why Hegel though?
The next broad set of objections I want to consider concern how ‘Hegelian’ the Hegelian move I make to attempt to answer the question of normativity is. These might range from the accusation that the move I make is not Hegelian at all to the claim that the move is too Hegelian in that it invites the kind of cultural relativity that I criticise Scanlon for
214 Scanlon (1998) p.74
143 being too open to. There is also the small matter of the specific objections Hegel had to contractualism and the search for abstract principles in general.
First, it might be argued that I have cherry-picked the parts of Hegel that work with the move I am trying to make and ignored the elements of his work that contradict this move. This might be thought of as a particularly culpable move in the case of Hegel given his own view and those of many of his interpreters as being a systematic thinker such that the various parts of his system are intricately connected and influence each other.
The accusation of cherry-picking is undoubtedly justified. My aim has been to provide a solution to the question of normativity by appealing to what I call the ‘Hegelian Thought’ about the nature of our agency and the relation we stand in to each other. This was derived from a reading of two relatively small sections of Hegel’s Phenomenology and his Encyclopedia. The aim was to show that this way of thinking about agency is plausible in itself and a useful approach for the contractualist ethics that is the primary concern of this thesis. Answering the question of whether this account of agency can be taken out of the context of Hegel’s system as a whole would require a knowledge of how the various parts of Hegel’s system interact with each other. Providing such an answer is beyond the scope of this thesis, and the abilities of its author.215
While paying heed to Hegel’s whole system might not be required, the fact that Hegel himself was thoroughly opposed to the kind of social contract theory advocated by Rousseau might not be so easily side-stepped.216 Hegel’s later ethical thought conceived of our obligations as always mediated by the institutions we found ourselves in, in particular those of the rational state. His objection to conceiving of the state, and our obligations to one another within the state, on the basis of the social contract is that it imported the norms of civil society: namely exchange and ‘external relations’ of right into the state where they had no place. Given Hegel’s antipathy to the idea of a contract as mediating the ethical relations between members of a state, it might be thought to be
215 It is worth noting that there is a long scholarly tradition of studying the component parts of Hegel’s system without paying heed to their relation to each other. Wood (1990), for example, gives an account of Hegel’s ethical theory while explicitly rejecting his earlier metaphysical claims. 216 See Hegel (1991) §75 pp. 105-06. C.f. Williams (1997) pp 276-80 and Patten (1999) p.108-112
144 illegitimate to appropriate his earlier account of recognition to ground the principle derived from modern moral contractualism.
But I think this objection can be avoided too. Avoiding this also appeals to the point that appropriating a part of Hegel’s system need not commit us to his system as a whole. The fact that he thought that our ethical obligations are only realised in the concrete setting of an ethical State ought to be a further indication that this thesis is not making such a move. In addition, it appears that Hegel’s objection was levelled more at the kind of developmental account of the nature of the state than the hypothetical account of contractualism being discussed here.217 Furthermore, Hegel’s objection to the importing of the morals of civil society to the state was rooted in the instrumental values of exchange and self-interested bargaining. Such an objection might be forceful against Gauthier’s version of contractualism, but appears to be less so against the kind of account that is advocated here.
A further worry that the account on offer here cannot be called a Hegelian, or even a Hegel-inspired, account of what we owe to each other is that it is open to Hegel’s ‘empty formalism objection’ that he levelled at Kant. This is connected to Hegel’s suspicion of abstract ideas grounding ethics in general. More specially in Kant’s case the concern was that the formal and immaterial nature of the Categorical Imperative meant that it could never actually provide concrete imperatives concerning what we ought to do, at least not without bringing in other non-formal principles to bear on the procedure of deriving these imperatives.218 The worry here would be that in appealing to a ‘relational formalism’ I would, in Hegel’s eyes, be making a similar mistake.
But the account of practical reason that I advocate to answer contractualism’s question of normativity does not attempt to provide a purely formal account in this sense. As is the case in Scanlon, agents can still be understood has having reasons (conceptually) prior to and independent of the contract. But the reason we have to enter into the bargaining process and find ways to live justifiably with others arises from the fact that our agency is inextricably bound up in the social space of reasons and so acting rationally requires taking this space of reasons seriously. Doing so involves standing in a relation of mutual
217 Patten (1999) p.119 218 See Stern (2012b) pp 75-76 on this.
145 recognition with the other, and this means according the other the kind of freedom we take to be necessary to and constitutive of our own status as independent agents. The concrete reasons of the agents play a role in the contractualism on offer, but the mediation of these reasons is abstracted from these concrete concerns.
These objections are worries that the account provided is not ‘Hegelian’ enough. I now turn to concerns that the account is too Hegelian. As has been repeatedly stated, the account of agency and our status as reasoning creatures put forward is a social achievement. It is also, in virtue of being a social achievement,219 a historical achievement. Given that our agency is socially, and therefore historically mediated, might it be the case that principles that are derived from this account of agency are going to be culturally relative in a similarly objectionable way to Scanlon’s?
It would be pointless to deny that there is going to be an element of relativity in any Hegel-inspired ethics. So when agents find themselves navigating the social space of reasons these reasons will be determined by the social norms and structures of the time. But my above discussion of why we can explain why the husband’s continuing his treatment of his wife as subordinate is wrong hopefully demonstrates that it is not objectionably relative. Although the reasons that are brought to the contractual process are ‘generic’ in being socially and historically mediated, the means of mediating them must be consistent with the agents standing in a relation of mutual recognition. So, while the kind of contractualism being advocated here might involve a degree of relativity, it is hopefully not objectionably so.
But it might be countered that the notion of mutual recognition is itself a product of a given social and historical epoch. Thus, it seems that the kinds of reasoning creatures that we are is itself historically and socially contingent. Hegel notes that the way we go about binding ourselves to various principles, and the way we offer and consider reasons is itself a product of the Reformation.220 If it is the case that the types of reasoning creatures we are is historically and culturally specific, then it follows that any principles which are derived from facts about the kinds of reasoning creatures we are will also be contingent.
219 And in keeping with the Hegelian approach in general. 220 See Moyar (2010) p.24
146 This is an objection where I am happy to bite the bullet. Contractualism is a species of constructivist ethics. All constructivist ethics depend to a greater or lesser degree on what human agents are like, or how they would respond in certain situations. If the facts about the kinds of creatures we are change, then the facts about what we have reason to do will change too. I am not claiming that the sorts of contractualism considered here can provide an answer to the question of normativity that applies universally across all possible worlds, but rather one that is applicable to the world as it is today.
The Shmagency Objection
I now want to turn to a different kind of objection, one that has been levelled at accounts that try to root our moral reasons in facts about our agency. This is an objection that has been addressed to a variety of accounts, not just accounts that account for agency as a social achievement. What accounts of this kind, whether based on an individualistic conception of agency or an interpersonal one, share is the idea that certain norms or rules follow as a result of being the kind of agent described. This is known as Constitutivism. The objection is addressed to the claims about what follows from the particular account agency. It is often known as the ‘Shmagency Objection’, after the paper that David Enoch wrote.221
I will present the argument as if it was a specific objection to my account, for while Enoch does not address his objection to a relational account of agency, I assume that if the objective is forceful against individualistic accounts of agency it will also be forceful against interpersonal accounts. Suppose that I have established that we do stand in a certain relation to one another and this relation is an inescapable feature of agency. I might then present the argument:
1. All agents stand in a particular, inescapable relation to one another.
2. This relation is constitutive of agency
221 Enoch (2006)
147 3. Maintaining this relation involves the performing, and the avoiding, of certain acts
4. Failure to perform, or avoid, certain acts means we have violated this relation and so violated our own agency
5. Therefore, we ought to act in the way that the maintenance of this relation requires
Enoch is concerned with the inference from premises 1-4 to the conclusion. Even if all were true, 5 would not be justified. Enoch’s claim is that whatever is taken to be constitutive of agency, whether it is standing in relations of mutual recognition or some other account,222 it is always open to the agent to simply state that they don’t care about the fact that their agency is constituted in this way. They are quite happy being a ‘shmagent’ - that is, one who is similar but subtlely different to being an agent
At times it is a little unclear precisely what a shmagent is. Enoch claims that a shmagent is simply an agent who lacks the aim of being an agent (Enoch 2006, 178). But this seems ambiguous between two understandings of what aiming to be an agent involves. First, it might be the case that being a shmagent is just like being an agent except for the fact that a shmagent lacks the explicit goal of being an agent, where this is understood in a de dicto sense (c.f. Smith 1994). On this understanding, then, someone who fails to do what is constitutive of agency, however this is cashed out, might well shrug off any criticism to the effect of ‘that’s not what an agent would do’. On the other hand, someone who is a shmagent might be one who lacks the aim of agency, where this is understood in a de re sense. To take a Korsgaardian example of what this might mean, suppose that agency is to be understood as having a consistent practical identity.223 On this kind of constitutivism, my reason not to throw a kitten off a motorway bridge is given by the fact that doing so would be a violation of my practical identity. I would, in some sense, be acting against the person that I am, and the values that I endorse. If being a shmagent means not having this aim, whether explicitly articulated or not, then the ‘why should I care’ response seems highly questionable. This, of course, depends on how convincing
222 Another example would be Velleman’s (2006) that we have reasons of agency in virtue of agency being tied up with the search for self-understanding. 223 See Korsgaard (1996) on this.
148 we find the particular conception of agency in question, whether this is derived from Korsgaard or a different constitutivist. But if being a shmagent means a more substantive rejection of certain ways of living, then Enoch would need to do more than simply posit the possibility of being a shmagent rather than being an agent. The problem with the first possibility – of shmagency being simply a matter of not being concerned with labels – is that if there were norms that could be derived from agency, then it is not clear why the shmagent would be exempt from them. If there are certain norms that constitute building a good house, then someone who wants to build a shelter for themselves would do well to follow them. Someone who is not concerned about whether the shelter they build is called a house or a shmouse would presumably still care that it meets the standards that are required for the new construction to be stable. Nor does Enoch seem to think that being a shmagent means lacking the capacities that we take to be constitutive of being an agent. In short, it is slightly unclear what Enoch means by this.
Applied to the account on offer in this thesis, the above suggests that an agent has a reason to perform some action, and failure to do so would be a violation of the relation we stand in to others, then Enoch’s objection would be pressing if read de dicto – that is, if the claim was that we would no longer be able to think of ourselves as standing in this relation. However, this is not what is being argued for. I assume that if the relation holds, then this is better read de re. This means that if an agent has a reason to do something for their friend then this is because failure to do so would undermine the specific friendship itself , not because it goes against what might be thought of as ‘friendship’ etiquette.
One of the common objections to Enoch’s shmagency objection is that it assumes that agency is an enterprise like any other.224 If I am playing chess, I am bound by the rules of chess. I may not move my rook diagonally. But I might decide that I am not, in fact, playing chess. I am playing another game, which uses the same pieces, and in this game you are allowed to move your rook diagonally. Call this game ‘shmess’. The point is that the player can step back from the game of chess, and deliberate about whether to continue to play chess (according to the rules), or play shmess, just leave the board altogether and do something else. In fact, agency is not like any other enterprise – when
224 See Ferraro (2009) pp. 308-312
149 deciding whether or not to participate in any other enterprise (say chess) we are always already taking part in the enterprise of agency.
If so, then the shmagency objection is rendered incoherent because we cannot stand outside agency and consider whether or not to be an agent or a shmagent. Just by doing so we are involved in the enterprise of agency. As Velleman argues:
“To ask ‘why should I have the aim of making sense?’ is to reveal that you already have it. If you don’t seek to do what makes sense, then you are not in the business of practical reasoning, and so you cannot demand reasons for acting or aiming.”225 (Velleman, 2009, p.137)
By raising this kind of challenge to agency, you show yourself to be a devoted follower. The shmagency challenge assumes that there is an external standpoint to agency where agency can be assessed. But this is not possible. This can be thought of as analogous to epistemological skepticism, where skeptics are shown to defeat themselves by using the rules of inference, and thus failing by their own lights. Similarly, the shmagency challenge relies on agency, and thus ignores the truly inescapable nature of agency.
Enoch does have responses to this. One response involves the claim that defences of constitutivism which invoke the inescapable nature of agency are committed to the fallacy of deriving an ought from an is. This will be investigated in the next section. But he also thinks that this kind of challenge is erroneous due to the error of anthropomorphising the ‘skeptic’ who raises the shmagency objection. The issue the constitutivist faces is not being confronted by an actual skeptic. The skeptic is rather the embodiment of the constitutivist’s commitments.
Enoch illustrates this the problem through the example of the ‘paper skeptic’. Suppose a philosopher argues that all paper submissions are intellectually corrupting and should be jettisoned in favour of books. Yet she makes this argument in a paper. The question is whether this latter fact means that she is wrong. It might well seem that she is in trouble, due to the inconsistency between her claims and her actions, but this does not mean that those who want to defend paper submissions are out of the woods yet. They still need to come up with a substantive response to the challenge. Simply pointing to the
225 Velleman (2009) p. 137
150 inconsistencies of the critic is nothing more than an ad hominem attack. Enoch calls this making the mistake of the ‘adversarial stance’.
The analogous problem for the constitutivist is clear. The fact that the skeptic (in this case the one who advances the shmagency objection) relies on agency to make this challenge does not mean that the constitutivist has no challenge to answer. The mistake, according to Enoch, is to anthropomorphise the shmagency challenge and then ad hominem attack the imagined opponent who makes it.
Yet there is something not quite right about this argument through analogy. The response to the shmagency objection is to deny the possibility of the shmagent. Even asking the question about whether to be an agent or a shmagent involves agency. Enoch’s thought seems to be that this is a dodge of a similar kind to that of the philosophers who point out that the paper skeptic submitted her argument in a paper, rather than engaging with the substantive criticisms. But the paper skeptic presumably could have made her argument in a non-paper format. The fact that she didn’t might show that she was careless or inconsistent, but she might still be correct about the paper format. And the reason why she might still be correct about the paper format is that it is possible for academic philosophy to move away from that format. She is arguing for a change that can, in principle, take place.226 But the point of the inescapable objection is that we cannot decide whether to be agents or shmagents because we are always already agents. Just by entertaining the question we are presupposing the agency involved which makes it a false choice. If the paper skeptic’s use of the paper format showed that it was impossible to conduct philosophy outside the paper format, then the defenders of the status quo might be correct to use this as evidence against her position. This is not the case, and so it is fair to label their response as ad hominem. But this does not apply to the agency case.
Enoch has suggested while we might not be able to opt out of agency altogether, we can participate in an alienated, or half-hearted manner.227 If we think back to the chess analogy; suppose that, for whatever reason, we cannot exit the game of chess. We are stuck, and forced to make some moves. But we still have room to decide how we go
226 This is essentially an ‘ought-implies-can’ point. I take this to be an uncontroversial principle for present purposes. My thanks to Stephen Ingram for pressing me to clarify thsis 227 Enoch (2006), p. 188
151 about playing chess. We might play without having the aim of winning, or we might seek to circumvent the rules – say, by moving the rook diagonally when our opponent is not looking. Perhaps a shmagent can be understood as an agent who, while unable to exit agency, takes part in it in a similarly alienated way. Thus, we might be inescapably in a relation that is mediated by mutual recognition, but we still have no reason not to treat this with a certain amount of contempt, perhaps by complying with the moral requirements in a haphazard and half-hearted way.
Ferraro’s response to this is to first to deny that one who is pretending to play chess is actually playing chess.228 This seems questionable, since it seems to mean that the parent who goes easy on their child while playing is not in fact playing chess at all.229 This might be true, but it seems a controversial claim to hang much on. Ferraro also notes that the analogy is, once again, flawed, since whether or not ‘pretending’ to play chess amounts to playing chess, ‘pretending’ to be an agent must involve agency. Pretending, after all, involves intentional action.
But for present purposes we should perhaps grant Enoch that agency is something like being stuck in a chess game that you cannot leave.230 If this is the case, then it might be thought that we have some reason, if only a prima facie one, to attempt to engage in the rules of the game.231 This might be thought of as making the best of the situation. But Enoch is not convinced by such claims. His response is to point out that the fact that we are engaged in such an activity is not sufficient for there to be a reason to continue in it.232 This is not only an attempt to derive an ought from an is, such as some natural fact, but a particularly egregious one. While he is not convinced by the idea that some natural
228 Ferraro (2009) p. 313 229 C.f. Enoch (2011) p. 214 230 ‘Cannot leave’ is potentially misleading. One can permanently exit agency through suicide, or take a drug to temporally induce a loss of agency (c.f. Parfit 1984). Even if one takes either of these steps, there will have been an exercise of agency involved. One who decides to exit agency, whether permanently or temporarily, is not indifferent to agency. (See Ferraro 310) 231 The ‘only prima facie’ clause might be significant. One of the issues with trying to ground normativity with agency is that even if such a move is successful, the norms that are derived from the conception of agency might be too weak to say anything interesting about what we ought to do (morally). This will, of course, depend on how much you pack into your conception of agency – the more you pack in, the more substantive the norms you will get out. But, naturally, the more you pack in the more objections you will face. 232 Enoch (2011) p. 215
152 fact, like ϕ-ing causing me pleasure, is enough to justify continuing to ϕ, he finds it much more convincing than the claim that ‘I am already ϕ-ing’ can be used to justify continuing to ϕ. The fact that I cannot, in fact, desist from ϕ-ing doesn’t change the fact ϕ-ing is bad. Enoch gives the example of the person who recognises that they really shouldn’t care about how their local sports team performs, but because of childhood indoctrination she cannot help herself. Her being unable to help herself in caring doesn’t mean that she should care.
But it seems that this response again overlooks the significance of the distinction between agency and other ordinary enterprises, like supporting a sports team. Even if a person cannot in fact stop caring about their local team, this does not mean that there is a good reason to do so. Yet someone who reasons to this point, who thinks that it would be better if she didn’t care about her team’s success, presumably could imagine what it would be like to not care about this team, even if she can’t actually attain this state. Likewise, it seems difficult to imagine oneself not being an agent.
The purpose of bringing up the shmagency objection is not to settle the debate about constitutivism. But by showing that some of the examples Enoch invokes as a comparison don’t quite work as analogies, hopefully the force of this objection is somewhat muted. Part of the insight of German Idealism is that we are not only always already acting in a space of reasons, this space is irreducibly social. This brings forth the necessity of recognition and justification. Perhaps the agent in this relation might have reason to ignore this, but this in turn suggest the necessity of further justification.
5.5 Conclusion
The previous chapter attempted to establish the framework through which we can think about the kind of account of practical reason that can answer the question of normativity. This chapter constitutes the positive contribution of this thesis to this question. It has argued that it is a feature of agents as such that they stand in a certain relation to others, and this relation entails standing in conditions of mutual recognition with the other. This implies the need for engaging with the other in a way that respects their agency and
153 freedom, which in turn suggests acting according to the principles that would be agreed in the hypothetical circumstances that characterises contractualist ethics.
I attempted to support these claims by invoking the early Hegel’s account of recognition. I provided a exegesis of this account and demonstrated how it can be appropriated by a contractualist ethics. I then raised a number of objections to this account and offered some responses to them, paying particular attention to the Shmagency objection that all attempts to derive our moral requirements from facts about our agency must face. Given how unusual adopting Hegel for the purposes of this thesis is, there will doubtless be more. I hope, however, that I have made a prima facie case for an adoption of this kind, and shown another way modern analytic ethics can gain something from German Idealist thought.
154 Conclusion
Having dealt with some anticipated objections to my proposed solution to the question of normativity, I will here simply sum up the argument of this thesis and say a brief word about its possible implications.
The purpose of this thesis has been to show that current account of moral contractualism struggle to answer the question of normativity, and lay the foundation for a potentially more successful response. I will summarize the steps here:
First, I considered David Gauthier’s account of moral contractualism as an instance of a contractualist theory based on an instrumental conception of practical reason. While such an account can explain why agents would enter into an agreement, I argued that it fails to provide a plausible explanation of why real-world agents conforming to such a conception of practical reason would continue to adhere to this agreement.
Second, I looked at Thomas Scanlon’s account of moral contractualism. Scanlon attempted to answer the question of normativity from the assumption that agents have reason to find justifiable ways to live with one another. While a very attractive account in many respects, I argued that Scanlon struggles to avoid an objectionable relativism without positing an independent source of value that undermines contractualism’s ambitions to provide an account of morality’s foundations.
Third, I investigated Nicholas Southwood’s Deliberative Contractualism. Southwood posited an interpersonal account of practical reason based on what he calls ‘relational formalism’. I argued that this provides a useful insight into what a successful solution to the question of normativity might look like, but Southwood’s account of the nature of this relational formalism is underdeveloped and the norms he derives from it seem somewhat ad hoc.
Fourth, I attempt to lay the groundwork for what a more successful interpersonal account of practical reason might look like. I consider the sort of frameworks offered by Darwall and Thompson, arguing in favour of the latter’s. I considered Darwall’s appropriation of Fichte for the contractualist cause, arguing that while this provides a useful heuristic for such an approach, it suffers from limitations inherent to Fichte’s ideas.
155 Finally, I offered my attempt at a solution to the problem, by invoking Hegel’s account of agency and mutual recognition. I provided a description of Hegel’s account and explained how it can be applied to contractualism. I then considered several objections to such a move and offered some responses.
*
I turn now to some limitations and implications of this work. It should be noted that even if the account offered here is successful, it still wouldn’t solve some of the other issues that face moral contractualism that were discussed in the introduction. There is, for example, nothing that has been said here that seems likely to do a better job of explaining our obligations to nonhuman animals than other accounts of contractualism. It might be necessary for moral contractualism to limit its ambition to just the interpersonal aspect of morality,233 rather than everything we think of as ethical.
Time has prevented me from providing a fuller account of what a contractualism that takes the Hegelian Thought seriously might look like. I argued for such a contractualism doing a better job of dealing with the man and his wife example, but it would have been interesting to consider the applications of this kind of contractualism to other ethical issues. Such an activity would have been interesting in itself, but it would also have provided a useful test for the kind of contractualism that is suggested by the Hegelian thought. If an answer to the question of content based on the sort of practical reason suggested here was found to be strongly contrary to our widely shared intuitions, then this would be a problem for this account.
The hope motivating this thesis has been to highlight the nature of the question of normativity and offer a solution to it. If the argument is accepted it might lead to further research in contractualist ethics and perhaps interpersonal practical reason. It might also pave the way for further engagement in the wider resurgence of Hegelian thought in analytic philosophy, in particular in ethics and the possible reconciliation of Hegel and the kind of liberal theories of which contractualism is an archetypal example.
233 And even that might be questionable, given that there are persons who would seem to struggle to stand in a relation of mutual recognition to others in any meaningful sense
156
Bibliography
Ashford, Elizabeth., (2003). “The Demandingness of Scanlon’s Contractualism”, Ethics, 113(2): 273–302
Available at https://www.jstor.org/stable/10.1086/342853 [Accessed 27/06/2019]
157 Austin, J. L. (1962) How to do Things with Words (London: Oxford University Press)
Barnes, R. Eric. (2012) ‘Contractualism and the Foundations of Morality’, Australasian Journal of Philosophy, 90 (4), pp. 815-818
Available at https://doi.org/10.1080/00048402.2012.728613 [Accessed 06/06/2017]
Brady, Michael. (ed.) (2011) New Waves in Metaethics (New York: MacMillan)
Brooks, Thom. (2012) Hegel’s Philosophy of Right (Oxford: Blackwell Publishing)
Buchanan, James. (1975) The Limits of Liberty, (London: University of Chicago Press)
Darwall, Stephen (2003) Contractarianism/Contractualism (Oxford: Blackwell Publishing)
Darwall, Stephen. (2006) The Second Person Standpoint: Morality, Respect and Accountability (Cambridge MA: Harvard University Press)
Darwall, Stephen, (2007) “Reply to Korsgaard, Wallace, and Watson”, Ethics (2007) 118 (1) pp. 52-69
Available at http://www.jstor.org/stable/10.1086/522018 [Accessed 28/02/2017]
Dimock, Susan. (1999) ‘Defending Non-Tuism’, Canadian Journal of Philosophy 29 (2), pp. 251-274
Available at https://doi.org/10.1080/00455091.1999.10717513 [Accessed 15/02/2015]
158
Drier, James (2006) (ed.) Contemporary Debates in Moral Theory (Oxford: Blackwell Publishing)
Dudley, Will. (2007) Understanding German Idealism (Stocksfield: Acumen)
Enoch, David (2005) ‘Why Idealize’ Ethics, 115 (4) pp. 759-787
Available at https://www.jstor.org/stable/10.1086/430490 [Accessed 21/06/2019]
Enoch, David. (2006) ‘Agency Shmagency: Why Normativity Won’t Come from What Is Constitutive of Action’ The Philosophical Review, 115 (2) pp. 169-198
Available at https://www.jstor.org/stable/20446897 [Accessed 18/06/2017]
Enoch, David. (2011) ‘Shmagency Revisited’ in Brady (2011) 208-233
Ferraro, Lucas. (2009) ‘Constitutivism and the Inescapability of Agency’ in Shafer- Landau (2009) 303-333
Fichte, J. G. (1982) Science of Knowledge with the First and Second Introduction edited and translated by Peter Heath and John Lachs (Cambridge: Cambridge University Press)
Fichte, J. G. (2000) Foundations of Natural Right According to the Principles of Wissenschaftslehre edited by Frederick Neuhoser trans. Michael Baur (Cambridge: Cambridge University Press)
159
Gauthier, David. (1986) Morals by Agreement (Oxford: Clarendon Press, 1986)
Gauthier, David and Sugden, Robert. (eds.) (1993) Rationality, Justice and the Social Contract: Themes from Morals by Agreement (Hertfordshire: Harvester Wheatsheaf, 1993)
Habermas, Jurgen (1990) Moral Consciousness and Communicative Action, trans. Christian Lenhardt and Shierry Nicholsen. (Cambridge, UK: Polity Press)
Hardiman, Michael O. (1994) ‘Role Obligations’, The Journal of Philosophy 91 (7) pp. 333- 363
Available at http://www.jstor.org/stable/2940934 [Accessed 14/08/2013]
Hegel, G. W. F. (1991) Element of the Philosophy of Right edited by Allen Wood trans. H. B. Nisbet (Cambridge: Cambridge University Press)
Hegel, G. W. F. (1997) Phenomenology of Spirit trans. A.V. Miller (Oxford: Oxford University Press)
Hegel G. W. F. (2010) The Science of Logic trans. George Di Giovanni (Cambridge: Cambridge University Press)
Hegel, G. W. F. and Inwood, Michael (2007). Hegel’s Philosophy of Mind trans. W. Wallace and A.V. Miller (Oxford: Oxford University Press)
160 Hieronymi, Pamela. (2011) ‘Reasons for Action’, Proceedings of the Aristotelean Society (2011) 111 (3), pp. 407-427
Available at https://www.jstor.org/stable/41331558 [Access 27/06/2019]
Hieronymi, Pamela. (2012) ‘Of Metaethics and Motivation: The Appeal of Contractualism’ in Wallace et al. (2012)
Hollis Martin, (1993) ‘The Agriculture of the Mind’, in Gauthier and Sugden (1993)
Hobbes, Thomas. (1991) Leviathan (Cambridge: Cambridge University Press)
Hooker, Brad. (2000) Ideal Code, Real World (Oxford: Oxford University Press)
Houlgate, Stephen. (2009) ‘McDowell, Hegel and the Phenomenology of Spirit’, The Owl of Minerva 41 (1) pp. 13-26
Korsgaard, Christine M. “Autonomy and the Second Person Within: A Commentary on Stephen Darwall’s The Second‐Person Standpoint”, Ethics (2007) 118 (1) pp. 8-23
McDowell, John. (2003) ‘The Apperceptive I and the Empirical Self: Towards a Heterodox Reading of "Lordship and Bondage" in Hegel's Phenomenology’, Hegel Bulletin 24 (1) pp. 1-16
Available at https://doi.org/10.1017/S0263523200001774 [Accessed 23/06/2019]
McGinn, Colin. (1999) “Reasons and Unreasons”, The New Republic, May 24, pp. 34–38
161
Millgram (2001) Varieties of Practical Reason (Cambridge MA: MIT Press)
Miller, Alexander. (2003) An Introduction to Contemporary Metaethics (Cambridge, Polity Press)
Milo, Ronald. (1995) ‘Contractarian Constructivism’ The Journal of Philosophy 92 (4) pp 181- 204
Available at https://www.jstor.org/stable/2940922 [Accessed 20/06/2019]
Moyar, Dean. (2011) Hegel’s Conscience (Oxford: Oxford University Press)
Nagel, Thomas (1986) The View from Nowhere (Oxford: Oxford University Press)
Narveson, Jan. (2010) ‘The Relevance of Decision Theory for Ethical Theory’ in Ethical Theory and Moral Practice 13 (5) 13-25
Available at http://dx.doi.org/10.1007/s10677-009-9193-3 [Accessed 21/12/2014]
Neal, Patrick (1988): Hobbes and Rational Choice Theory The Western Political Quarterly, 41 (4) pp. 635-652
Available at http://www.jstor.org/stable/448487 [Accessed 21/09/2014]
O’Connor, Timothy & Sandis, Constantine (eds.) (2010) A Companion to the Philosophy of Action (Chicester: Blackwell Publishing)
162 Parfit, Derek. (2011a) On What Matters: Volume 1 (Oxford: Oxford University Press)
Parfit, Derek. (2011b) On What Matters: Volume 2 (Oxford: Oxford University Press)
Patten, Alan. (1999) Hegel’s Idea of Freedom (Oxford: Oxford University Press)
Pettit, Phillip. (2006) ‘Can Contract Theory Ground Morality?’ In Drier (2006) Contemporary Debates in Moral Theory (ed.) James Dreier pp.77-96
Pippin, Robert P. (2000) ‘What is the Question for which Hegel’s Theory of Recognition is the Answer?’ European Journal of Philosophy 8 (2) pp. 155-72
Available at https://doi.org/10.1111/1468-0378.00106 [Accessed 27/06/2019]
Pinkard, Terry. (2017) Does History Make Sense? Hegel on the Historical Shapes of Justice (London: Oxford University Press)
Rawls, John. (1971) A Theory of Justice (Cambridge MA: Harvard University Press)
Ridge, Michael. (2001) ‘Saving Scanlon: Contractualism and Agent-Relativity’, The Journal of Political Philosophy 9 (4) pp. 472-481
Available at https://doi.org/10.1111/1467-9760.00137 [Accessed 27/06/2019]
Rousseau, Jean-Jacques. (2002) The Social Contract and The First and Second Discourses edited by Susan Dunn (London: Yale University Press)
163 Scanlon, Thomas M. (1982) ‘Contractualism and Utilitarianism’ in Sen and Williams (1982) pp. 103-128
Scanlon, Thomas M. (1998) What We Owe to Each Other (USA: Harvard University Press)
Scanlon, Thomas M. ‘How I am not a Kantian’ in Parfit (2011b) pp. 116-139
Scanlon, Thomas M. (2014) Being Realistic About Reasons (Oxford: Oxford University Press)
Schueler, George Frederick. (2010) ‘Motivational Internalism and Externalism’ in O’Connor & Sandis (2010)
Sciaraffa, Stefan. (2009) ‘Identification, Meaning, and the Normativity of Social Roles’, European Journal of Philosophy 19 (1) pp 107-128
DOI: 10.1111/j.1468-0378.2009.00375.x [Accessed 27/06/2017]
Sen, Amartya. (1977) “Rational Fools: ‘A Critique of the Behavioral Foundations of Economic Theory’: Philosophy and Public Affairs, 6 (4). 317-344
Sen, Amartya and Williams, Bernard. (eds.) (1982) Utilitarianism and Beyond (Cambridge: Cambridge University Press)
Shafer Landau, Russ (ed.) (2009) Oxford Studies in Metaethics Volume 4 (Oxford: Oxford University Press)
164
Smith, Michael. (1994) The Moral Problem, (Oxford: Blackwell Publishing)
Southwood, Nicholas (2010) Contractualism and the Foundations of Morality (Oxford: Oxford University Press)
Stern, Robert (2002) Hegel and the Phenomenology of Spirit (London: Routledge)
Stern, Robert. (2012a) ‘Is Hegel's Master–Slave Dialectic a Refutation of Solipsism?’, British Journal for the History of Philosophy, 20 (2), pp. 333-361
Available at http://dx.doi.org/10.1080/09608788.2012.664026 [Accessed 18/02/2013]
Stern, Robert. (2012b) ‘On Hegel’s Critique of Kant’s Ethics: Beyond the Empty Formalism Objection in Brooks (2012)
Stratton-Lake, P., 2003. “Scanlon’s Contractualism and the Redundancy Objection”, Analysis, 63 (1) pp. 70–76.
Available at https://www.jstor.org/stable/3329143 [Accessed 20/06/2019]
Thompson, Michael. (2004) ‘What is it to Wrong Someone?’ in Wallace et al. (2004)
Velleman, J. David. (2006) Self to Self: Selected Essays (Cambridge: Cambridge University Press)
165 Wallace, R. Jay. et al (eds.) (2004) Reason and Value: Themes from the Moral Philosophy of Joseph Raz (eds.) (Oxford: Oxford University Press)
Wallace, R. Jay. (2007) ‘Reasons, Relations, and Commands: Reflections on Darwall’ Ethics 118 (1) pp. 24-36.
Available at https://www.jstor.org/stable/10.1086/522016 [Accessed 21/03/2018]
Wallace, R. Jay, Kumar, Rahul and Freeman, Samuel (eds.) (2012) Reasons and Recognition: Essays on the Philosophy of T.M. Scanlon (Oxford: Oxford University Press
Williams, Bernard. (1981a) Moral Luck, (Cambridge, Cambridge University Press)
Williams, Bernard. (1981b) ‘Internal and External Reasons’ in Williams (1981a)
Williams, Bernard. (1981c) ‘Persons, Character, and Morality’ in Williams (1981a)
Williams, Robert R. (1992) Recognition: Fichte and Hegel on the Other (Albany: State of New York University Press)
Williams, Robert R. (1997) Hegel’s Ethics of Recognition (London: University of California Press)
Wood, Allen W. (1990) Hegel’s Ethical Thought (Cambridge: Cambridge University Press)
166 Wood, Allen W. (2014a) The Free Development of Each: Studies on Freeedom, Right and Ethics in Classical German Philosophy (Oxford: Oxford University Press)
Wood, Allen W. (2014b) ‘Fichte’s Intersubjective I’ in Wood (2014a)
Wood, Allen W. (2014c) ‘Fichtean Themes in Hegel’s Dialectic of Recognition’ in Wood (2014a)
167