Two Types of Social Norms

1 The Agency-Respect View

The agency-respect view offers a novel answer to the question of when and why, if at all, socially constructed norms are morally binding:

I argue that the moral normativity of socially constructed norms stems from our duty to give people agency respect: to respect their authentic commitments as agents, provided that those commitments are morally permissible and respecting them isn’t too costly for us. I show that this duty accounts for when and why the fact that a socially constructed norm requires something of us places us under an obligation to comply. (Valentini 2023, 82)

The agency-respect view is built on two premises, one ontological and one normative. The ontological premise is referred to as the agential-investment account and states the existence conditions for socially constructed norms, which are roughly that there is a general rule that is publicly and widely accepted, where acceptance is understood as belief in the content of the general rule and a commitment to act by it. The normative premise is referred to as the agency-respect principle and states the moral criterion, or principle, that distinguishes social norms that are morally binding from those that are not, as nicely summarized by Valentini in the quote above.

In addition to clear and compelling argumentation, Valentini offers a methodological discussion. Any view that seeks to answer the general question should meet two common desiderata: explanatory power and fit. She argues that the agency-respect view fares better than the competing accounts in meeting these two desiderata.

I limit this discussion to part one of the book, the development of the agency-respect view, and leave the application to the three problems in part two for another occasion. Let us turn to the conception of socially constructed norms.

2 Social Norms and Standards of Behavior

Queuing for the bus is a leading example of a socially constructed norm and is used throughout the book. Valentini recalls encountering the British queuing norm in a somewhat harsh fashion during a visit to the United Kingdom at a young age: Unknowingly not following the local norm but rather the Italian norm of waiting for the bus gave rise to sanctions ranging from strange looks to verbal scolding.

There are several other examples of socially constructed norms: “men ought to take their hats off upon entering a church,” “young people ought to offer their bus seats to the elderly,” “ladies ought to be served first,” “one ought not to use others’ property without their consent,” and “doctoral students ought to attend the doctoral workshop” (Valentini 2023, 20).

Valentini suggests that one thing these social norms have in common is that they are about specific standards of behavior. Another is that not acting by socially constructed norms of this type gives rise to moral reactive attitudes such as shame, guilt, and resentment (cf. Strawson 1962). Moral reactive attitudes are thus a way to single out the socially constructed norms that are the focus of attention in Valentini’s book. In an illuminating methodological discussion, she contends that there is no such thing as the right or best account of socially constructed norms. Instead, there are many different conceptions, and philosophers might just as likely be interested in norms like aesthetic norms or epistemic norms. For her purpose, the focus is on social norms that elicit moral reactive attitudes upon breaking them: “I simply restrict the scope of my analysis to a particular prominent class of such norms. These are norms the violation of which triggers moral reactive attitudes” (Valentini 2023, 48).

Valentini builds on Geoffrey Brennan’s effort to give an account of socially constructed norms: “Every A, in C, ought (not) to/may φ, where A is an agent-type, C is a context-type, and φ is an action-type” (Brennan et al. 2013, 3). So, to use one of the above examples, every doctoral student in a particular university’s philosophy department should attend the doctoral workshop. Starting from the general definition of a socially constructed norm, Valentini interprets it in a specific way and thus provides a particular conception of a socially constructed norm; the agential-investment account. It takes “public acceptance” and “commitments” as central notions: “socially constructed norms exist when a general action-guiding rule is widely and publicly accepted in a given context” (2023, 22).

For a general rule to be ‘accepted,’ one must believe in it and be committed to carrying it out. Valentini writes that the “acceptance of a general rule or requirement involves a belief in the content of the requirement, coupled with a commitment to the requirement functioning as a general standard of behavior” (2023, 28, my emphasis). For this general rule to be “widely accepted” means that “a sufficiently large number of individuals within the context under consideration must possess the right kinds of beliefs and commitments” (2023, 28). For the general rule to be public there must be mutual beliefs or what Valentini refers to as “common attitudes,” which “involve a group of individuals whose attitudes have the same content and who are also mutually aware of one another’s attitudes: each is aware that the others hold the attitudes, each is aware that the others are aware that the attitudes are held, and so on” (2023, 29). These are the existence conditions of socially constructed norms.

Let us return to another item on Valentini’s list of socially constructed norms to see how they fit this conception. For the bus norm (“young people ought to offer their bus seats to the elderly”) to exist, a sufficiently large number of people in, say, Stockholm must believe that young people ought to give up their seats to older people and be committed to that course of action. There are common attitudes about this general action-guiding rule in that people are mutually aware of one another’s attitudes regarding the bus norm.

These socially constructed norms come in different forms. According to Valentini, the most salient distinctions are between formal and informal socially constructive norms and between directive and power-conferring norms. Formal norms usually combine primary and secondary rules, while informal rules consist of primary rules (Hart 1961; Valentini 2023, 43–48). The directive, or prescriptive, norms are those that “directly govern our behaviour” while “power-conferring norms confer certain powers upon us” (Valentini 2023, 46). Another benefit of this work is that these different types of norms are included from the start as objects of analysis, in contrast to, say, first proposing a theory of formal social norms and then unsuccessfully trying to extend it to informal norms or proposing a narrower account of solely formal and directive social norms.

However, in addition to the above distinctions, there is another, equally important distinction that has been omitted from the discussion: the one between what I refer to as deontic social norms and teleological social norms (Burman 2023). Deontic normativity concerns what we can demand of each other, while teleological normativity concerns standards of excellence that we sometimes want to live up to and that others expect us to live up to. The most plausible interpretation of the list of socially constructed norms that Valentini starts from is that it concerns what we can demand of each other. The language of rights and obligations springs to mind: Given that the bus norm is in place, older people have a social right to a bus seat, and younger people have a social obligation to give them their bus seat. Alternatively, given that a traditional gender norm is in place, women have a social right to be served first, and men have a social obligation to wait to be served.

3 Social Norms and Standards of Excellence

Parallel to the development of the agency-respect view, there has been a simultaneous development in contemporary social ontology concerning social normativity. This section brings these two parallel discussions in conversation with each other. I use gender norms as an example since they are viewed as paradigmatic by the different philosophers involved in this discussion. In contemporary social ontology, Johan Brännmark argues that teleological normativity can be reduced to deontic normativity and that deontic powers explicated in terms of Hohfeldian incidents are the central building blocks of institutions (2019a, 2019b). By contrast, Charlotte Witt argues that deontic normativity can be reduced to teleological normativity and that the Aristotelian conception of a function is central to understanding social role normativity (2023). In Nonideal Social Ontology (2023), I take a middle position, arguing that teleological normativity and deontic normativity and their respective forms of social power – telic and deontic – are distinct because they have different existence conditions. This discussion presupposes the use of a deontic in a narrow rather than wide sense. We can take deontic in the wide sense to mean anything that is obligatory, permissible, or forbidden, while deontic in the narrow sense means actions that concern social or institutional rights and obligations.

Let us turn to teleological normativity. In contrast to the social norms discussed thus far, teleological norms are primarily about standards of excellence rather than standards of behavior (although these standards of excellence are connected to actions in a way to be explained). In an earlier article, Sally Haslanger introduces this Aristotelian notion of a norm with an everyday example of a paring knife and then extends the account to gender norms.

Something counts as a paring knife only if it has features that enable it to perform a certain function: it must be easily useable by humans to cut and peel fruits and vegetables. We can distinguish, however, between something’s marginally performing that function and something’s performing that function excellently. A good paring knife has a sharp blade with a comfortable handle; a poor paring knife might be one that is so blunt that it crushes rather than cuts a piece of fruit, it might be too large to handle easily, and so on. Those features that enable a paring knife to be excellent at its job, are the ‘virtues’ of a paring knife. (2012, 42–43)

Note that a standard of excellence rather than a standard of behavior is referred to and that the central notions are those of excellence, functions, and virtues. Generally speaking, once we have imposed a function on an object, it becomes possible to evaluate how well it performs that function, such as being an excellent, good, or bad paring knife. Likewise with people, once there is a standard of excellence in place, it becomes possible to evaluate how well a person fulfills her role, such as being an excellent, good, or bad president, bus rider, or woman.

In general, our evaluation of the goodness or badness of a tool will be relative to a function, end, or purpose, and the norm will serve as an ideal embodying excellence in the performance of that function. Likewise, masculinity and femininity are norms or standards by which individuals are judged to be exemplars of their gender and which enable us to function excellently in our allotted role in the system of social relations that constitute gender …. For each role there are performances that would count as successes and others that would count as failures; in general, one could do a better or worse job at them. The suggestion is that gender roles are of this kind; gender-norms capture how one should behave and what attributes are suitable if one is to excel in the socially sanctioned gender roles. (Haslanger 2012, 42–43)

In addition to what I have referred to as deontic gender norms that are about social rights and obligations, there are teleological gender norms about being a good or bad woman or man, as Haslanger points out in the above quote. Recall Valentini’s earlier example of ‘ladies first’: this norm can be characterized in terms of informal social rights and obligations; women have the social right to be served first while men have the social obligation to wait, but the notion of a good woman or a good man points to a different type of social norm, understood in terms of standards of excellence and functions.

Charlotte Witt has developed an original account of gender norms using a similar Aristotelian framework as Haslanger in The Metaphysics of Gender (Witt 2011). In Social Goodness, published just a few months prior to Valentini’s Morality and Socially Constructed Norms, Witt develops a general theory of social role normativity referred to as the artisanal model. According to Witt, arts and crafts are “intrinsically normative activities: each activity realizes a function that is associated with an excellence in relation to which the function is performed” (45). Other social roles, such as parent or woman, are also constituted by functions, and social normativity resides in these functions. In short, the notions of standards of excellence and functions play central roles in this conception of social norms.

The central concepts in the artisanal model are function, technique, and expertise. Witt uses the example of a carpenter, who ought to use a level in making tables and other material objects. Ideally, the carpenter is skilled; in other words, she has expertise, which like skill “refers to the technique possessed by an individual agent; it is a stable ability or power to perform an activity or function in a manner responsive to the relevant set of norms of techniques – to perform it badly, well, or in-between” (Witt 2023, 73). Her expertise has been developed over time by imitation and habituation, by learning from excellent carpenters and the techniques they employ: “Techniques are shared ways of working and acting within a community, and as such they are part of a community’s social knowledge and shared practices” (73). In this Aristotelian conception of functions, they are intrinsically normative. For example, the function of a house is to provide shelter; based on how well a house fulfills that function, it is a better or worse house. Witt argues that the artisanal model can be extended to social role normativity as such, to roles such as professors, women, men, and presidents:

Consider former U.S. president Donald Trump. His decisions, statements, and actions were criticized from many perspectives and for many reasons. For example, some of his actions on January 6, 2021, might have been illegal or might have been unethical in that they might have compromised his oath of office. However, in addition to the legal or the ethical criticisms, a central line of criticism focused on his social position and the idea that some of his actions or statements were unpresidential. They were not responsive to the norms associated with being president, with occupying that social position. Yes, it is wrong to use crude language about one’s opponent, but it is particularly wrong for a president. … Even though its exact content is debatable, there is a social role, a set of norms that attached to Trump because he was president, not because he endorsed them or promised to keep them, but because he occupied that very social position. (Witt 2023, 40–41)

The critique of being unpresidential is well understood in terms of a standard of excellence and the president being perceived as substandard, as failing to display the virtues appropriate to that role; this is teleological normativity. Critics will most likely respond with resentment: being unpresidential in the way Witt describes also elicits moral reactive attitudes. In addition to such attitudes discussed thus far and in Valentini’s book, there are also moral reactive attitudes like gratitude, love, and praise. Going back to Haslanger’s example of gender norms, it is often the case that being viewed as a good or even excellent woman elicits moral reactive attitudes like love and praise from others. Consequently, the type of social norm Haslanger and Witt have singled out is prominent according to Valentini’s own criteria for choosing what types of social norms to include and exclude in answering the central question of when and why, if at all, socially constructed norms are morally binding. In this way, Valentini’s conception of a social norm is too narrow even according to her own standard.

4 Teleological Social Norms and the Desiderata

A proponent of the agency-respect view might at this point object that teleological norms can be accommodated by its conception of social norms. Consequently, the account is not too narrow. However, I think this line of response is problematic for the agency-respect view due to the its desiderata – explanatory power and fit – and the distinction between a narrow and wide sense of the deontic. The agency-respect view equates the normative with the deontic in the wide sense, but this blurs the difference between deontic norms in the narrow sense and teleological norms. To show this, let us return to three recurrent examples Valentini uses in developing the agency-respect view: traffic light, barbecue, and non-proceduralist president (2023, 57–59). These examples are used to show that some socially constructed norms have moral normativity. The aim of traffic light is to show that there is a weak kind of moral normativity in the social norm of stopping at a red light even if no one would get hurt and one would actually better achieve one’s aim better of, say, getting home sooner by not stopping at the light. More generally, it is used to show that an agent can have a moral obligation (even if there is no rights violation) due to the existence of a social norm. Barbecue also suggests that there seems to be something morally wrong with using someone else’s pots and pans during a camping trip even if they never find out about it and the users are better off. More generally, barbecue is used to show that an agent can have a moral obligation not to violate another agent’s right due to the existence of a social norm.

Setting aside the interesting questions of moral normativity, consider what types of social norms are present in these examples: the driver has an obligation to stop at red lights, and the campers have an obligation not to use someone else’s equipment without their consent, while the owners of the equipment have the right not to have people use their property without their consent. In both cases, the social norms are deontic in the narrow sense of consisting in social rights and obligations. The same observation holds for Valentini’s own non-proceduralist president example, in which the president is acting ultra vires; that is, beyond his legal powers, breaching his institutional obligation not to use unwarranted domestic surveillance. Valentini’s interpretation is that the president is violating the ideal of the rule of law.

Non-proceduralist President: In the wake of the 9/11 terror attacks, President George W. Bush authorized the National Security Agency to track international calls and e-communications of people inside the US, without a court warrant. … Once this became known, the President was criticized for acting ultra vires, in violation of the 1978 Foreign Intelligence Surveillance Act (FISA), which prohibits warrantless domestic electronic surveillance. In an open letter to Congress, published in the New York Review of Books, a group of leading legal scholars and former government officials insisted that, to be lawful, such surveillance would need to be authorized by Congress. (Valentini 2023, 58–59)

This example fits nicely with the initial distinctions used by Valentini of formal and informal and directive and power-conferring: the Foreign Intelligence Surveillance Act is an example of a formal power-conferring norm. But in addition to criticizing the president for violating citizens’ rights, such as privacy, and not fulfilling his obligation to act in accordance with the rule of law, we might in Witt’s words criticize the president for being unpresidential, for not displaying the virtues appropriate to his role. In short, we might both offer a critique in terms of social rightness and social goodness and feel resentment with respect to both breaches. As Valentini and Witt have vividly illuminated, we are often engaged in both types of activities. In this way, having both types of social norms at our disposal fits the available evidence; it thus meets one of the desiderata.

I turn now to arguing that invoking social teleological norms in addition to deontic social norms meets the other desideratum – explanatory power – better than only having deontic norms in the narrow sense at one’s disposal. I have developed this argument with slightly different purposes elsewhere and it will be helpful to briefly revisit it here (Burman 2023, 180–81):

Sociologist Beverley Skeggs investigated gender and class norms in Formations of Class and Gender: Becoming Respectable (1997) by conducting interviews with British working-class women. Many of them worked in the home as housewives. Skeggs refers to the norm of good housewife in this context as someone who has an impeccably clean home, respectable clothes, refined language, and shows care and concern for others. One of the women interviewed shared her thoughts with Skeggs (1997, 3) after a Health Visitor inspected the interviewee’s home: ‘You know they’re weighing you up and they ask you all these indirect questions as if you’re too thick to know what they’re getting at and you know all the time they’re thinking ‘she’s poor, she’s no good, she can’t bring her kids up properly’ and no matter what you do they’ve got your number. To them you’re never fit, never up to their standards.' The woman notes a standard that she fails to live up to in the eyes of the Health Visitor – and perhaps even herself and that thus she is not a good housewife. (Burman 2023, 181)

Like the other examples of this second type of social norm – the teleological social norm – it is easy to imagine that the perceived failure to live up to the standard of a good woman triggers moral reactive attitudes like shame and guilt. And these two different the types of social norms can both reinforce and conflict with each other. For instance, an agent’s perceived excellence or failure with respect to the standard of womanhood might influence her social and institutional rights and obligations (cf. Burman 2023, 189), just as being perceived as an excellent researcher means that one might secure a new academic position with new formal rights and obligations, which is an example of teleological norms reinforcing deontic norms in the narrow sense. It thus seems that deontic and teleological social norms have different causal powers since they can either reinforce or conflict with each other. Causal power is a common criterion for existence, and thus it appears that these two types of social norms have different existence conditions. That different types of norms have different existence conditions is an argument Valentini uses to say that moral norms are different from social norms. Similarly, this argument can be used to say that deontic social norms and teleological social norms are really different types of norms.

Note also that these distinctions cut across each other, so that one can have formal and informal deontic norms and formal and informal teleological norms. For instance, in the Swedish research funding context, the criteria for someone being an excellent researcher and something being an excellent research environment have been formalized, whereas the standards of excellence concerning traditional and other gender roles are informal in Sweden.

The upshot is that having both deontic social norms and teleological social norms in our theoretical toolbox means increased explanatory power, which is the other desideratum of the agency-respect view. In sum, then, incorporating teleological social norms meets the two desiderata.

5 Standards of Excellence and Virtue Ethics

The agency-respect view consists of two parts, an account of socially constructed norms and a moral principle P that provides a criterion for when and why socially constructed norms are morally binding. This moral principle is called the agency-respect principle and states that “one has an obligation to respect people’s commitments (i.e., to give ‘agency respect to people’) provided those commitments are authentic, morally permissible, and respecting them is not too costly” (Valentini 2023, 88).

I agree with Valentini’s conclusion that this moral principle is plausible as a criterion for deciding when and why the socially constructed norms the agency-respect view assumes are morally binding. Recall the three previous examples of traffic light, barbecue, and the non-proceduralist president. All these cases, as well as the list of social norms, are couched in terms of rights and obligations (or the failure to respect one’s own obligations or other’s rights). Respect for a person’s or community’s commitments goes well together with a rights-based approach and a rule-based ethics. In fact, Valentini refers to the work of Kant and Rawls as inspiration. But does the agency-respect principle work as well for the social good as it does for the social right?

It is instructive to return to Witt’s recent work and her discussion of morally bad social roles like the excellent thief. She emphasizes the distinction between an internal and external critique, arguing that an Aristotelian framework has certain internal resources by which to criticize this social role by stating that it runs contrary to the social whole: “There is a straightforward sense in which a thief might be skilled at their ‘job,’ but the role itself is not a cohesive part of the social system” (Witt 2023, 122). To criticize other examples of morally bad social roles we need to invoke external, or moral, criteria, according to Witt. She does not elaborate on what kind of moral criterion she has in mind, but it seems like a natural fit with the Aristotelian conception of a social norm to suggest an ethical framework like virtue ethics to help decide which teleological social norms have moral force. For example, one might criticize traditional social norms of womanhood and manhood for being contrary to human flourishing. In short, this is an entirely different framework with respect to both the conception of social norms and the ethical framework appealed to in answering the general question of when and why social norms have moral force.

6 Conclusion

The agency-respect view consists of an ontological and a normative premise. I have questioned the plausibility of the former by arguing that the account of social norms assumed is too narrow. More specifically, I have argued that there is another type of social norm – the teleological norm – that also elicits moral reactive attitudes and that is relevant to answer the general question and thus for Valentini’s purposes. These teleological norms are pervasive and often reinforce or come into conflict with the type of social norm assumed by the agency-respect view. Questioning the ontological premise means that the normative premise is less plausible than it initially seemed: widening the conception of social norm to incorporate both types calls the moral criterion proposed by the agency-respect view into question. This is due to the fact that the agency-respect principle fits well with the deontic social norms assumed by this view, but it might not be easily extended to teleological social norms.

Until proponents of the agency-respect view have shown that their proposed moral criterion is applicable to teleological social norms, that view is incomplete. One option would thus be to show that teleological social norms can be reduced to deontic social norms and that the suggested moral criterion holds for both types of norms. This would keep the generality of the agency-respect view. But until this has been achieved, the general question is not fully answered. Another option would be to offer a pluralistic account of social norms (and possibly more than one moral criterion) to answer the general question. This would restrict the scope of the agency-respect view to one type of social norm. Regardless of which option to pursue, Valentini has already made a genuinely important contribution to the philosophical study of the morality of social norms.