Using Partial Relationships as a Basis for Ethics: An Ancient Strategy

Abstract

Over the past few decades, the topic of partiality has received increasing attention in ethical debates. Philosophers have enquired into why we treat people differently from others when we share a special relationship with them and why many of us believe that we should treat those people differently. Is that belief justified, i.e., can ‘duties of partiality’ be given a philosophical foundation? And if so, what role does the ‘special relationship’ play in such a foundation?

The reference to ancient theories has always played an important part in discussions of partiality. Since ancient ethics gives much more attention to relationships than does modern moral philosophy, it is an obvious source of inspiration for any engagement with partial reasons or duties. Modern accounts of friendship, for instance, are heavily influenced by Aristotelian ideas. Yet I believe that recent developments in the debates on partiality call for another look at ancient ethics. It makes sense to ask what similarities exist between the modern ethics of partiality and ancient ethical projects, what modern theories can learn from ancient authors in this field, and how changing views on the role of partiality in ethics might change our views of ancient ethics as well. It is questions like these that this special issue aims to discuss.

The following paper serves as an introduction to that discussion. It explains why one might say that ancient philosophers already presuppose what modern ethicists want to establish: a relational approach to duties or reasons. It considers in which respects the ancient treatment of partiality is comparable to modern accounts and in which way it is not. And it sketches some features of the ancient ethics of partiality.

The paper falls into three parts. In the first part, I will distinguish between two projects within modern debates on partiality: the project of making room for partial duties and the project of letting partial duties do ethical work. I will argue that this distinction offers a useful basis for approaching ancient ethical theories. It gives us a first idea of their attitude towards that topic. In the second part, I will take a closer look at three ancient arguments from different contexts, concerning, respectively, the abolition of the family, our duties towards strangers, and vegetarianism. I will show that these arguments share a strategy of using partial relationships as a basis for ethics. In the third part, I will summarize some features of that strategy.

1. The Ethics of Partiality: Two Projects

There is a lot of discussion on what partiality is and how partial reasons or duties should be defined.¹ Here is what I will assume for my purposes. Though I will speak of duties rather than reasons in the following, I will not work with an ambitious notion of a moral duty. A duty, as I understand it, is a normative requirement: something someone has to do or is expected to do, for instance in connection with a certain role he or she is playing. A partial duty, as I understand it, is a duty grounded in a special relationship, a duty that I do not have towards anyone but only towards those with whom I stand in such a special relationship. Duties towards family members and friends are paradigm cases, and I will focus on these in what follows. But of course there are other relationships that might be taken to ground partial duties, which includes, for instance, the relationship created by a promise (Jeske 2001). Put in a nutshell, if someone asks, “Why should I give A some money?” and the answer is: “Because he is your brother,” then this answer works on the assumption of a partial duty. But now imagine someone asks, “Why should I give B some money?” and the answer is: “Because she suffers from starvation, and you should relieve suffering wherever you find it. This makes the world a better place.” This answer obviously works on the assumption of an impartial duty, a duty we are taken to have towards anyone who is suffering, independently of the relationship in which we might stand to that person. Impartial duties are thus grounded, not in relationships, but in the intrinsic features of those towards whom we have those duties.

I have mentioned above that partial duties have found increasing interest in the ethical debates of the past few decades. The discussion is already highly complex, but for the purposes of this introduction, I suggest to distinguish between two basic projects. This will help us to explicate the notion of an ethics of partiality as well as to highlight similarities and differences between ancient and modern approaches.

The first project is that of making room for partial duties. Simply put, this project is a reaction to the predominance in modern ethics of a view of morality as strictly impartial, as it is most consistently advocated by utilitarianism. On such a picture, adopting the moral point of view precisely means adopting an impartial point of view and thus ignoring the relationships in which we stand to other people (think, for instance, of Henry Sidgwick’s “point of view of the universe” [Sidgwick 1907, 382]). Partial duties, if there are any, would fall outside the realm of morality; partial behavior, such as caring for your children because they are your children, would appear problematic from the moral point of view. Such behavior could only be justified if it could be shown, from an impartial perspective, that the world will be a better place if everyone cared for their nearest and dearest, for instance because these are the persons whose needs one knows best.² But as many have pointed out, this is counterintuitive. In fact, it seems morally blameworthy to not give one’s children special treatment and to care for them just as much as for any other child in the world.³ The aim of the first project is to do justice to this intuition and to defend partial duties as part of an adequate moral theory. This defense usually comes with a rejection of the reductionist strategy just sketched. It insists that partial duties cannot be given an impartial foundation.⁴

The aim of the second project, by contrast, is to let partial duties do ethical work and treat them, not merely as a part, but as a basis of an adequate moral theory, thus creating an ethics of partiality in a more ambitious sense. The philosophers of this project would assume that our partial or relational duties, the duties towards family and friends, tell us something important about duty or duties more generally. When Diane Jeske, for instance, discusses the question how political obligations could be grounded (Jeske 2001), she starts by considering two uncontroversial sources of special obligations: promises or contracts, on the one hand, and intimate relationships among family members and friends, on the other. Her strategy is, basically, to ask which of these uncontroversial cases offers a ‘model’ for the controversial case of politics.⁵ And when Jay Wallace (2019) develops a novel theory of moral obligation, this theory is partial—more precisely, relational—through and through, for it takes such obligation to be a relational requirement, i.e., something we owe to someone in a nontrivial sense.⁶

How does ancient ethics relate to that? Here are some observations to begin with. None of these observations should be very controversial, though the details definitely allow for disagreements. First, ancient ethical theories, at least before Cicero’s De officiis, are not built around the concept of duty but around the concept of the good, which is conceived as an end.⁷ More precisely, ancient philosophers did not think about ethics along the Kantian lines of acting from duty against inclination. This is a commonplace, and I think it is basically correct. So nothing of what I am going to say in what follows is meant to turn ancient ethics into a deontological project. Second, this does not mean, however, that ancient philosophers did not know the concept of a duty in the nonambitious sense I have suggested (something that someone is required to do) (see Visnjic 2021, ch. 1). The life of, say, a 4th century BCE Athenian was not a life without duties or obligations. A male citizen of Athens was obliged to do military service, for instance. If he failed to do so, he would have been blamed or even punished. And more generally, one might speak of a duty to obey the written or unwritten laws of one’s polis.⁸ The third observation is that ancient ethical texts make reference to partial or relational duties on a regular basis. They talk about keeping one’s promises, returning what one has borrowed, obeying one’s father and helping one’s brother.⁹ In Plato’s Crito, to give an example, the duty to obey the laws that I have just mentioned is discussed as a relational duty. It is compared to the duty to obey one’s father (50e–51a) and taken to be grounded in an implicit promise or agreement between the citizen (Socrates) and the laws of his polis (Athens) (49e ff.). This is not so different from the kind of reflection that Diane Jeske engages in.

Finally, what about impartial duties? On the one hand, I doubt that we can find in ancient ethics anything like the above principle “I should relieve suffering wherever I find it” or “In whatever I do, I should make the world a better place.” This is just not how deliberation was conceived in ancient ethical texts. On the other hand, it would again be wrong to assume that ancient philosophers did not know the concept of impartiality or that our ancient Athenian was never expected to act impartially. Most notoriously, impartiality was taken to be a requirement of distributive justice. As Aristotle explains, a just distribution is one according to merit (NE 5.3, 1131a20–29; Politics 3.9). It treats equals equally and unequals unequally and thus ignores the relationship between distributor and distributee. We also know from several contexts that a judge qua judge was expected not to favor his friends—though it is telling that the situation is usually described as a dilemma. It is not understood that the role of an impartial judge should have more weight in our decisions than the role of a friend. As Aulus Gellius relates, there must have been a tradition of treatises on this topic.¹⁰

Though all these observations would require further discussion, they seem to indicate already that a comparison between ancient and modern approaches to partiality is no simple matter. I believe, however, that it is possible to identify at least one clear contrast between those approaches. To do that, we need to draw on our first project. This project, remember, is one of making room for partial duties. It is directed against the predominance of a picture of morality as impartial. Now one way to characterize that project is by a certain distribution of the burden of proof. Due to the predominance of the impartial viewpoint, the burden of proof is taken to be on those who defend partial duties as part of an adequate morality (otherwise it would be no ‘defense’). In ancient contexts, by contrast, there is no such predominance. In the example of the judge I have alluded to, it is precisely not taken for granted that the impartial viewpoint of a judge should have more weight than the partial perspective of a friend. This is something that needs to be argued for. Accordingly, we find nothing like our first project in ancient ethics, which is my first claim regarding the comparison between ancient and modern approaches to partiality. Ancient philosophers do not doubt the existence of partial duties. For someone like Aristotle, it is obvious that not helping one’s brother is a terrible thing to do, more terrible in fact than not helping a stranger (NE 8.9, 1160a5–6). He also believes that gifts are most laudable if given to a friend, not to someone in need, as we might suppose (NE 8.1, 1155a7–9). Furthermore, it is not uncommon for ancient ethical treatises to reflect on the hierarchy of obligations that corresponds to the hierarchy of our relationships (Plato, Laws 4.717a–18b; Cicero, Deofficiis 1.160). There is no need to ‘make room’ for partiality in ancient ethics, for it is already there.

What is more, considered from a modern perspective, the distribution of the burden of proof seems somehow reversed. Take Plato’s Euthyphro.¹¹ In the opening scene of this dialogue, Euthyphro is introduced as planning to charge his father for the murder of a servant. This leads to a discussion about the nature of the pious (to hosion). The beginning of the dialogue leaves no doubt that Euthyphro has to justify himself for this plan: not just because the case is complex and difficult to judge but also, and mainly, because the culprit is his father. The servant, by contrast, was a stranger (allotrios: 4b). Thus Socrates says: “Is then the man your father killed one of your relatives (tôn oikeiôn tis)? Or is that obvious, for you would not prosecute your father for the murder of a stranger” (4b). To be sure, this question need not express Socrates’s or Plato’s considered judgement. Plato’s Socrates is notorious for his revisionist inclinations. But the question indicates, I think, the dialectical situation an ancient ethicist would start from. That we should treat oikeioi differently than allotrioi is here taken for granted. The challenge is to argue against this partial attitude. So, we observe a distribution of the burden of proof which is opposite to the one presupposed by the first project of the modern ethics of partiality. Euthyphro has to justify himself for not treating his father partially. And this is not the only example. Three hundred years later, for instance, Cicero is attacked for putting the needs of the republic above his friendship to Caesar when approving of the latter’s assassination.¹² Here as there, partiality is taken to be the normal case (cf. Dover 1974, ch. 6). This, I take it, is most characteristic for the ancient attitude towards partiality. When reflecting on duties, ancient philosophers did not think in terms of “I have to relieve suffering wherever I find it” but rather in terms of “I have to help my brother.”

Now, if that is correct: if ancient ethical theories regard partial or relational duties as default duties, then it seems natural to think that they somehow presuppose what the second project outlined above wants to establish, namely a relational perspective on questions of morality. My second claim is that this is in fact what they do. The aim of the next section is to explicate this claim and thus the idea of ancient ethics as an ethics of partiality. How do partial duties do ethical work in ancient philosophy? By considering three examples, I will try to outline a first answer to this question.

2. Using Partial Relationships as a Basis for Ethics: Three Examples

In his historical “Introduction” to the volume Partiality and Impartiality, Brian Feltham asserts that Plato’s ethical perspective is, unlike Aristotle’s, impartial in both structure and content. One of the many reasons he offers for this assertion is that the Guardians in Plato’s ideal society will be “limited in respect of personal attachments” (Feltham 2010, 7). This obviously refers to one of the most notorious aspects of Plato’s Republic, namely the idea that within the class of the Guardians there should be no families: wives and children are held in common, as Plato puts it (Republic 5.457cd). This abolition of the family is our first example. Let’s take a look at it.

One of the core problems that ancient political philosophy has to deal with is stasis or social conflict. In Plato’s view, such conflict is caused to a large extent by the usual distribution of loyalties (another cause being the existence of private property). People tend to favor their relatives over other citizens and to put the interests of their family above those of the political community. As Feltham correctly notes, this has something to do with personal attachments, but it surely also has something to do with partial duties. Plato himself mentions as obvious that we have to show our parents respect, solicitude, and obedience (463d). One might say, hence, that Plato takes partiality to be an important part of the problem which plagues so many cities of his time and that he thinks one can increase the unity of a polis by operating against partiality through abolishing family structures. Feltham’s example seems well chosen.

In an interesting way, however, and viewed from a different perspective, Feltham is wrong and his example does not support the claim that Plato’s ethics is impartial (2010, 8).¹³ For if we take a closer look at Plato’s argument, we notice the following. The ideal projected in Republic 5 is that every member of the group of the Guardians should regard every other member of this group as somehow related or belonging to them (Plato uses in this context the word oikeios that we have already met in the Euthyphro; see 463bc). For instance, all future Guardians of a certain age should regard all male persons of a certain age as their fathers and treat them, accordingly, with the same respect: the respect, namely, that we owe to our fathers. The important point, then, is that the basis of this same respect is not that children learn to take a neutral perspective and ignore the relationship in which they stand to other people. The children of Plato’s ideal state do not adopt Sidgwick’s point of view of the universe. The basis of the same respect is rather the idea of equal partiality. The future Guardians are raised in such a way that they develop to all male persons of a certain age the same relationship that they would develop in a regular family to their own father. As we know, this requires some radical measures, such as separating the newborn from their mothers right after birth. And with Aristotle, we may doubt that these measures work.

Aristotle’s attack against Plato’s Republic, as launched in Book 2 of his Politics, is in fact interesting, for it helps us to further highlight the difference I have just sketched. Roughly put, Aristotle’s worry is that Plato’s project will precisely lead to the neutral perspective I have characterized as impartial. People will no longer feel related to anyone and will thus no longer care about each other (Politics 2.3). They will not act in the partial way that Plato seeks to establish. By focusing on the psychological basis of our partial behavior, and by charging Plato of ignoring that basis, Aristotle precisely demarcates the strategy of Plato’s Republic from an impartial viewpoint. Impartiality would be a mistake.

We can now see two things. First, we need to be careful in determining the way in which Plato’s suggestion is impartial. It is impartial in the (weaker) sense that the adults will make no difference among the children with regard to care, nor the children among the adults with regard to obedience. But it is not impartial in the (stronger) sense that this care or obedience would not be that of, say, a father towards his son, and vice versa. When considering how to behave towards a group of people, it makes a difference whether we ignore the relationship in which we stand to the members of that group or take ourselves to stand to each of those members in exactly the same relationship. The actions resulting from those considerations might be the same (though this is not necessarily the case), but they will have a different rational and motivational basis. Second, we get a first impression of what it means to let partial duties do ethical work. Plato does not want to call into question the special duties of respect, solicitude, and obedience we owe to our parents. Rather, he wants to direct those duties to a larger group of people. He follows what is sometimes called in the literature a strategy of ‘extended partiality’.¹⁴

Let us turn to the second example: helping strangers. It should not come as much of a surprise that interaction with strangers plays no crucial role in ancient ethical thought. This is particularly true for pre-Hellenistic Greek thought, i.e., for Plato and Aristotle, with its focus on the polis and its communities. What we do find in different, especially later contexts, however, is the following. We find, first, a reference to a number of common rules or duties that are taken to guide that interaction and request us to help strangers (though it is clear that we have stronger obligations towards those who are closer to us). Cicero’s De officiis, for instance, relates that one should not keep others from fresh water, one should allow them to take fire from one’s fire, and one should show them the way (De officiis 1.52).¹⁵ Second, we find the claim that human beings are in fact inclined to help each other, including strangers, which can be observed, for instance, when we travel.¹⁶ In that spirit, a summary of Peripatetic ethics, i.e., ethics in the wake of Aristotle, asks the following questions:

(T1) Who would not, if he were able, rescue a human being he saw being overpowered by a wild animal? Who would not give support to a person perishing through poverty? Who, if he came across a spring in a waterless desert, would not by markers indicate it to those who come by the same route? (Stobaeus, Doxography C = Sharples 2010, 113)

It is true that such remarks are quite rare. As I just said, ancient philosophy shows greater interest in what we owe to the members of our political community than how we should or do behave towards strangers. The examples we possess, however, have some interesting features in common, although they stem from different contexts. For one thing, they usually take the duties towards strangers or the inclination to help them to be based on a natural community, kinship or friendship among all human beings, which means: they refer to a relationship. For another, they often connect this community, kinship or friendship with the community or friendship between parents and their children. Aristotle, for instance, names those cases as two examples of a friendship existing by nature (notice his use of the word oikeios in this context):

(T2) And there seems to be a natural friendship of a parent for a child, and of a child for a parent, and this occurs not only among human beings, but also among birds and most animals. It also seems to exist naturally among members of the same species (tois homoethnesi) in relation to one another, particularly among human beings, which is why we praise people who are lovers of humanity (philanthrôpous). And one can see in one’s travels how akin (oikeion) and friendly (philon) every human being is to every other. (Aristotle 2014: NE 8.1, 1155a16–22)

The Stoics are even more explicit. They call the parent-child relationship the principle or ‘starting point’ (archê) of the community of all human beings (Porphyrius, De abstinentia 3.19).¹⁷ In a notoriously elusive argument, they describe a series of ever increasing communities that links the nuclear family, on the one hand, with the community of all human beings, on the other (Cicero, De finibus 3.62–67).¹⁸ A similar, but still interestingly different, kind of series appears in a fragment by Hierocles the Stoic.¹⁹ Hierocles claims that each one of us is as it were encompassed by many circles. In the center of those concentric circles is the mind, then follows the circle of the body, then those of the closer and the wider family, of further relatives, local residents, fellow-tribesman, etc. The widest circle finally encloses the whole human race. The message of this much-discussed fragment is that “it is the task of a well-tempered man, in his proper treatment of each group, to draw the circles together somehow towards the centre, and to keep zealously transferring those from the enclosing circles into the enclosed ones” (Long and Sedley (1987) = LS 57G5). Hierocles’s advice on how to achieve that task obviously ties on Platonic ideas. He tells us that we should address people from the wider circles as we usually address those from the closer ones, i.e., we should call cousins brothers, uncles fathers, etc.

This sketch of some passages concerning our duties towards strangers again tells us something important. To begin with, it affirms what has been said in Section 1 about the burden of proof. However the arguments work in detail (which is in each case a matter of debate): that parents should and do care for their children is taken for granted. It is the uncontroversial starting point of an argument on how strangers should be treated. Since this argument refers to the community of all human beings, one might presume, prima facie, that it relies on an impartial strategy, based on the intrinsic feature of being human (or the feature of being someone in need of help, as in T1). But in fact, this reference is again made within a partial framework. For the texts speak of a relationship among human beings, namely that of being oikeios, and they take the relationship between parents and their children, perhaps the most partial relationship there is, as a model. The alleged duty to help strangers is grounded in the idea that our relationship towards strangers can be modelled on our relationship to family members.²⁰

So, obviously, we are again dealing with a strategy of extended partiality. The ‘extension’, however, seems to work differently than in the first example. A comparison between Plato’s abolishment of the family and Hierocles’s concentric circles is telling in that respect. Plato creates a situation in which from the outset the group of the fathers, for instance, is much bigger than in regular families. The children of his ideal state are not supposed to know, let alone care about, who among that group are their biological fathers, nor the adults about who are their biological children. In this way, and by way of further regulations, Plato aims to control both the scope and the natural development of partial relationships. Hierocles, by contrast, starts from a situation in which different relationships are already established. The people he addresses know perfectly well that their cousins are not really their brothers. Although he unfortunately does not provide us with any details, it is clear that his argument describes an extension in a more literal sense: an extension which is the result of a deliberate reflection on how we should behave towards other people. For all structural similarities, the two arguments proceed from quite different psychological preassumptions.

Let us now turn to our third example: vegetarianism. This example is taken from the work De abstinentia, in which the third-century CE philosopher Porphyrius offers arguments against the consumption of flesh. Porphyrius belonged to the circle of Neoplatonists around Plotinus who advocated not just a philosophy but a way of life, which included vegetarianism. De abstinentia is addressed to a Platonist who had abandoned vegetarianism and is now confronted with a battery of arguments that should change his mind. This battery, however, is not Porphyrius’s own but a collection from different sources.

Very early in his treatise, Porphyrius sketches the argument of those who defend the consumption of flesh and the killing of animals (1.4). These ‘carnivores’ draw on a premise that has already been mentioned, namely, that justice requires equality. They claim that, since animals are irrational and thus unequal to human beings—unequal in a relevant respect—it is not unjust to treat them differently and make use of them in all different ways for our purposes. Interestingly, the contrast between equal and unequal is here spelled out in terms already familiar to us, namely in terms of being oikeios or allotrios. Someone who says it is unjust to kill animals treats as family, the carnivore says, what is really a stranger. This is perfectly in line with the assumption of Plato’s Euthyphro (almost 700 years earlier) that there is a difference between killing a stranger and killing a family member.

One way to argue that animals must not be killed is hence to show that they are in a relevant way equal or similar to human beings, which means they are oikeios. There is, the vegetarian says, a community of animals and human beings. In Book 3 of De abstinentia, we find such an argument, drawing on Aristotle’s student Theophrastus.²¹ This argument has a strikingly similar structure to our helping-strangers example. It describes a series of natural communities which begins in the family and ends with the community at stake (here, between animals and human beings). In contrast to the helping-strangers example, however, it starts with the relationship, not between parents and their children, but between siblings,²² and it is not interested in how we behave towards family members but rather in the grounds for someone’s being oikeios. Here is a sketch of how the series proceeds (3.25): (i) Siblings are related (oikeios) by nature because they stem from the same mother and father. (ii) The same is true for blood relatives, who are descendents of the same ancestors. (iii) Sometimes, the citizens of a polis are called related. This, however, is no natural relationship but due to their sharing the same land and living in one community. (iv) If we say that Greeks are related to Greeks, Barbarians to Barbarians and, in general (v) humans to humans, we either mean that they have common ancestors or that they share nutrition, habits, and race. (vi) Humans and animals are related insofar as the principles of their bodies are by nature the same. ‘Principles’ does not refer to the elements of which plants consist as well, but to stuff like skin, flesh, and blood, which are only to be found in animals. Though some animals have higher developed souls, the principles are the same.

This argument is difficult to reconstruct, especially because it emphasizes that the reasons for calling someone oikeios are different in different cases. I could think of at least two options how this emphasis is to be understood. Either the idea is that, since there are different reasons for calling someone oikeios, which need not include common ancestors, there is nothing wrong with calling animals oikeion on the grounds mentioned in (vi). Or the idea is that, though there are different reasons for calling someone oikeios, all those reasons refer to some kind of principle or starting point. No matter how different human beings are, we call them oikeios because there is a common basis; any differences between them are due to later developments. Mutatis mutandis, the same idea can be used to reveal a relationship between human beings and animals.

No matter which, if any, of these options is correct, the basic thrust of the argument seems clear enough: animals should not be treated differently from human beings because they share the same relevant features. This idea is once again spelled out in terms of the partial question we have met in Plato’s Euthyphro and Republic, in Aristotle and the summary of Peripatetic ethics, in Stoic sources and Cicero: the question, “Who is my relative?” It is this question which seems to mark the ancient way of letting partial duties do ethical work. The argument itself then follows the familiar strategy of extended partiality. It starts, as in the helping-strangers example, from the common assumption that there is a relevant difference between ‘us’ and ‘others’ (strangers or animals, respectively) and tries to show that this assumption is unfounded. Yet unlike the helping-strangers example, the example of vegetarianism does not appeal to a feeling of familiarity but rather to our practical reason. If we understand that being related can have different meanings, we understand that there is no reason to treat animals differently than human beings with respect to killing. Extension here means to identify a relevant criterion.

3. Features of the Ancient Ethics of Partiality

In the previous section, I have presented a rough sketch of three examples that can be read as three ancient attempts to let partial duties do ethical work. Though from very different contexts, these attempts show remarkable similarities in their general outlook and strategy. So it seems legitimate to draw from them some general lessons about the ancient ‘ethics of partiality’ in the sense of the second project described in Section 1. Which features of that ethics have emerged? First, in all three examples partial duties are treated as default duties. That parents should take care of their children, while children should obey their parents, is not called into question. It is rather deployed in the sense that a reference to those default duties is the starting point of a discussion of more controversial cases. Working with models or paradigms seems to be one of the most characteristic features of the ancient ethics of partiality. Second, the discussion of controversial cases is structured by the contrast between the concepts oikeios and allotrios (it is remarkable how consistently ancient debates on what we owe to others proceed along those lines). Its guiding question is, simply put, who does, and who does not, belong to the sphere of the oikeios, where partial duties are taken for granted? So an important part of the argument is the attempt to analyze and identify kinships. Third, the attempt to identify (relevant) kinships is usually connected with the claim that the group of oikeioi is wider than first assumed. This is basically the strategy of ‘extended partiality’, as it has been described in the literature. Notice, however, that the three examples discussed in Section 2 can help us characterize that strategy more precisely.

To begin with, although the ‘extension’ is supposed to result in a state of equal partiality (or ‘weak’ impartiality, as one might call it), it is important to demarcate that result from the introduction of an impartial point of view (or ‘strong’ impartiality). This demarcation does not concern our behavior (the content of our duties) towards others, for both equal partiality and impartiality imply that we do not prefer one individual over the other. It rather concerns the motivational and rational basis of our behavior. In the case of extended partiality, this basis is taken to be similar to that obtaining, for instance, in family relationships; it is the purpose of the ‘extension’ to reveal that similarity.

Although this structure is the same in all three, and further, examples, we have observed a remarkable degree of flexibility within that structure. Depending on which relationship serves as a starting point, and which feature of that relationship is taken to be most relevant, the structure can be filled in in different ways and the extension can take different routes. Understanding those differences is, I think, one of the most important tasks of a more comprehensive account of the ancient ethics of partiality. It makes a difference, for instance, whether the ‘extension’ is the result of an arrangement that the people involved do not realise, as in Plato’s Republic, or a conscious process that requires reflection on the part of the agent, as in Hierocles or Porphyrius. While the first route trades on the instinctive and emotional basis of partiality that it tries to replicate more or less directly, the second route leads over the agent’s deliberation on what they have reason to do. It invites the agent (i) to regard her relationships, if not from an impartial point of view, then at least from a third-person perspective, (ii) to identify the feature which grounds the partial duties of those relationships, and (iii) to understand that, since the same features exist between her and, say, all human beings,²³ it would be inconsistent to treat other human beings differently. That we should help strangers is, then, a rational requirement rather than the expression of a feeling of affinity.

On this basis, we can develop a more precise understanding of the objection that appears on a regular basis at least since Aristotle’s criticism of Plato’s Republic, namely the objection that partiality rests on a special relationship which cannot be ‘extended’ so easily to other people.²⁴ As levelled against the first route, the objection has a psychological thrust. It doubts that the instinctive and emotional aspects of, say, a parent-child relationship can be replicated without also replicating the family life that grounds and nourishes those aspects. As levelled against the second route, the objection seems to point at the following, different concern. It is part of the strategy of extended partiality that, for instance, the duty to help a brother is something like a brute fact: it is taken for granted that we should help our brothers. The duty to help a stranger, by contrast, is a rational requirement, based on our deliberation about normative reasons: we would be inconsistent if we helped our brother, who is oikeios, but not a human being, who is oikeios as well.²⁵ The thrust of the objection would then be that the rational and motivational basis is precisely not the same in both cases.

To conclude, let me mention one problem that, for lack of space, could not be treated in this paper. This is the problem of the hierarchy of duties. The ancient ethics of partiality seems to work with the unquestioned assumption that the closer the relationship, the stronger the duty (cf. once again Aristotle on helping brothers or strangers). Sometimes, however, ancient philosophers wish to argue that a wider relationship comes with a stronger duty. This is the case, for instance, in Cicero, who tries to show that we have particularly strong duties towards the res publica (De officiis 1.50–58). This argument appears like a ‘test case’ for the ancient ethics of partiality that a more complete account would have to discuss. It would be particularly interesting to consider how this concern to do justice to our special duties towards the state, as a wider community, relates to the modern concern to do justice to the impartial point of view.²⁶

REFERENCES

Footnotes