This is the first in the reading-group series. Contributions welcome (from members of the network or beyond) in the comments. The post itself gives an overview of the article.
Hardimon, Michael O. (1994) Role Obligations. Journal of Philosophy. 91(7): 333-363.
This is a modern classic for anyone working on the ethics of roles, and is often cited in discussions of special obligations and the like. The writing is beautifully clear, and this a good place for us to start if only because one of Hardimon’s ambitions is to bring the topic of roles in from the wings of normative philosophy and onto centre-stage where many of us agree it belongs.
His other main ambition is to get clear on what is and isn’t right about what he calls the ‘standard view’ of role obligations, which he summarizes in three claims (p. 337):
- Role obligations are of two kinds, ‘contractual’ and ‘non-contractual’;
- Contractual role obligations are acquired by signing on for the roles from which they derive;
- Non-contractual role obligations are extremely problematic, if they exist at all.
He takes these in reverse order. His conclusions:
On (3), role obligations can be normatively binding despite being non-contractually entered into and unchosen – just so long as the roles themselves, and the institutions to which they belong, are ‘reflectively acceptable’. On (2), the reason-giving nature of many contractually entered-into roles are better thought of in terms of role-identification by the occupant than by comparison with ordinary acts of promising. And on (1), there is more in common between contractual and non-contractual role obligations than we might suspect once we spot that role occupants are motivated in both cases by role identification.
Hardimon covers a lot of ground in reaching this conclusion, and the journey will be of at least as much interest as the destination to members of this Network. Of note, for example, are:
- His objections to what he calls the Doctrine of Perfect Adequacy, a caricature commonly used to dismiss role-centric ethical thinking (section II)
- His claim that roles animate our moral lives because they bridge the gap between us as individuals and the social institutions we inhabit (e.g. pp. 341-2)
- His suggestion that there is less individual choice involved in voluntarily entered-into roles than one may suppose, because “roles, contractual and noncontractual alike, are institutionally defined clusters of rights and duties” (p. 354)
- His view that the obligations of a friend are importantly distinct from obligations of a family member or employee (section I)
- His description of role identification as potentially transformative, but not necessarily ‘engulfing’ of self (pp. 358-60).
A final point worth remarking on: Hardimon does not adopt the virtue-ethics approach to roles that has become more prevalent since 1994 (see e.g. the Special Issue mentioned in a recent post). Virtue thinking barely figures here, even as something he rejects. With the benefit of 23 years of hindsight, that might be seen as a serious lacuna. A more positive way of seeing it is that Hardimon offers us a lucid alternative candidate to a virtues approach to social roles.
ROLE ETHICS NETWORK 
To kick things off… If Hardimon’s account is a good one it would be a shame not to extend it to the role of friend. Happily, the two reasons Hardimon gives for setting it aside (p. 336) seem weak. First reason: friendship is not a persisting institution. ‘When two individuals become friends, a new friendship comes into being. Should they cease to be friends, their friendship will cease to exist; it cannot be resumed by anyone other than those particular individuals’ (p. 336). But that would be true of a family of two, and wouldn’t be true of a family-sized friendship group. So unless the numbers count, he should include friendships alongside families. Second reason: the rules of friendship are informal and elastic, unlike institution rules. But he includes gender roles (p. 349) in his discussion, the rules for which are likewise often negotiable up to a point.
Test comment per Alex’s request. (But I’ll post something substantive soon!)
Hi Alex –
I agree that the first reason is a bad one. As for the second reason, judging by what he says on the page you reference, it’s not clear to me that he thinks gender roles (e.g. woman, man) are institutional roles so much as he thinks that members of certain gender roles occupy various institutional roles (e.g. daughter, mother, wife). But if he does think gender roles are institutional roles and they are elastic in the way friendship is, he could decide to preclude gender roles from being institutional.
I think the larger questions, then, are, ‘what makes a role an institutional role?’, and ‘what constitutes an institution?’. If we knew that, we would know whether to put friendship/gender in the ‘institutional role’ category.
But rather than ask you what you think about this (though I am curious!), I want to ask you why “it would be a shame not to extend it [H’s account of institutional roles] to the role of friend.” Why does it matter whether it’s an institutional role instead of a non-institutional role? What is lost by its not being institutional?
Here is a thing that occurred to me on rereading the text, and on which I’d be interested to hear other people’s thought. Hardimon seems to describe reflective acceptability and role identification as mental processes that individuals go through on their own. But in the examples there are numerous pieces of conversations, questions asked by others, etc. That made me wonder whether we should think of reflective acceptability and role identification as SOCIAL processes in a much deeper sense (this is not incompatible with what Hardimon writes, was far as I can tell, he just doesn’t seem to discuss this, unless I overlooked it). This seems to cohere quite well with the phenomenology of roles: we discuss with others what the roles are, how binding they are, etc.; often, it is when we encounter others, who refuse to accept certain roles or fill them in very different ways, that we only ever start to think about our own roles. It seems hard to imagine how a process of reflective endorsement would happen without some social element (could be with an imaginary dialogue partner or a character in a novel or so – but that would seem extreme cases…).
Hello everyone!
On the question of whether or not friendship is an institution/friend is a social role in Hardimon’s view, and whether/why that matters. I wonder if Hardimon overdoes the supposed informality of the friendship role, the elasticity of friendship norms and the supposed susceptibility of these norms to individual foibles, and from here then (erroneously) derives the supposed non-institutional status of the friend role. For Hardimon, the terms and conditions – and perhaps the norms – of particular friendships are negotiable and elastic for the particular individuals in the relationship, in a way that they are not for the individual in the institutional role of President of USA (although actually even some POTUS norms now look to be under serious threat of reshaping from a particular incumbent, but that’s another issue). Although Hardimon outlines this largely empirical point as a ‘furthermore’ following the denial that friendship is a self-reproducing structure, it looks as though the latter is supposed to support the former.
But I confess I have never really grasped why Hardimon uses friendship as an example of a non-institutional role. Notwithstanding the (alleged) elasticity of the concept and the idiosyncrasy of particular individuals, friendship it looks like a structure that is ‘self-reproducing’ at least insofar as a) there are norms and obligations of friendship that transcend particular friends, and b) the social roles that interest Hardimon are, for him, sets of duties and obligations (or at least for him the duties and obligations of role r are inelimenable from the description of role r). In answer to Reid’s question to Alex, this is the extent to which it might matter. That is, there are norms and obligations of friendship- not just those of ‘friends’. Whether there is, a la Hardimon, a useful narrower sense of ‘institutional role’ and whether it matters a great deal in this context, I am less sure.
*As an aside: what would Hardimon say we are talking about when we use ‘friendship’ as a mass noun, if not some kind of institution? That is, how do we grasp that concept without thereby grasping a kind of institution, in the broadest sense of its encompassing a certain set of norms?
Reid, I was just voicing a simple thought that it would be a shame to limit the coverage of an otherwise good account for bad reasons, if they are bad. Are they bad? Well, on that second reason: there does some to be an ambiguity in ‘institution’. I’m sure there are many, but in particular, we say that a particular nuclear family – the Bakers of Lime Avenue – is an institution that can be added to with a new baby, or that can cease to exist when the children grow up and the parents divorce, but this is different from the sense in which the institution of the nuclear family survives the demise of the Bakers. Anyway, H seems to mean the first sense, but as Sean asks, why should we limit ourselves to that sense? (Gender roles are also, arguably, not institutions in the narrower sense, save in the respect Reid suggests. Again, an unnecessary limitation?)
On elasticity, of friendship or any other role: how flexible a role is can sometimes be in the eye of the beholder, or rather, can depend merely on how fine grained their focus. There are different ways of being the Dean of Faculty, e.g. one can take this to include being hands on and supportive, or leading by example, etc. (different ‘leadership styles’), but on the other hand there is only one relatively simple obligation: support the mission of the University by leading the Faculty.
That seems right to me Lisa: compatible with what he is up to but not explicit in the text. There is some acknowledgement of a role for social shaping when he talks about contractual roles being ‘interpretive’ (p. 355) rather than entirely pre-shaped. But here the interpretation seems to be fairly lonely process – not an ongoing exchange of views – and the result is just variability in ‘ways’ of being in the role rather than in deeply rooted features of the role’s shape and function arrived at through discussion, disagreement and example (‘role models’). But I’d have thought his account lends itself to extension (and improvement) in this way.