New Issue of Dao, and Article Discussion

(Moved to top for article discussion as a featured post – March 12, 2014)

The latest issue of Dao has been released. And in keeping with our new collaboration with the journal, one article has been set to free access:

Filial Obligations: A Comparative Study, by Cecilia Wee

Well, first, I have to thank Cecilia Wee for the very stimulating piece. I hope I have not misunderstood or misrepresented too much of it in the following remarks. I look forward to her comments and discussion by all.



In this comparative study, Cecilia Wee considers some contemporary analytic philosophical discussion of the nature of filial obligations – i.e. obligations that children have to their parents – and argues that Confucian ideas about such obligations are not captured very well by those analyses.

Wee on Western Analytic Philosophical Theorizing

Simon Keller’s (2006) piece, “Four Theories of Filial Duty,” frames Wee’s exposition, for the most part, of contemporary analytic thinking about the topic, though she also draws on a number of others. As Wee explains, there are, in one sort of approach, two types of account (my lettering):

Approach in Relation to other Obligations

A) filial obligations are analogous to other sorts of obligations
B) filial obligations are identical (in kind) to other sorts of obligations

I assume, not having looked carefully at those accounts myself, that part of the strategy of such analyses is to provide understanding of filial obligations through either some better understood category of obligation or (inclusive “or”) some broader, more general category of obligation under which they belong.  Candidates for those other categories – which Wee calls RSOs (“relevant special obligations”) include (my numbering):

1)   obligations of debt
2)   obligations of gratitude
3)   obligations of friendship

So, analytic philosophical accounts attempt analyses of filial obligation as either being analogous to or identical in kind with those three categories. As Wee points out, claims of identity (in kind) are stronger claims than those of an analogy – an analogous relationship would involve some kind of modeling, though only through appropriately relevant similarities. Hence, if there are significant reasons to doubt even the weaker claim of analogy to a type of obligation, then that would undermine any stronger claim of identity.

A different approach altogether is one that is not based on relation to other obligations:

C) filial obligations are in their own (independent) category of obligation

That is to say, filial obligations should be categorized in their own type of “special obligation” different from 1-3.

As Wee details, Keller and/or others argue that filial obligations are significantly disanalogous to obligations of debt, gratitude, and friendship. I find those arguments fairly convincing, except with the case of obligations of friendship. We can discuss the exact points of disanalogy that are raised with respect to any of these obligation types if anyone wishes, but I’ll skip here to what I don’t find convincing about Keller’s argument. I’m not sure whether Wee agrees with Keller on this point, so I hope she will say something about it. As Wee reports, Keller thinks the serious point of disanalogy with obligations of friendship is this (Wee 89):

…when friendships end, the nature of the obligations to one’s former friend would change quite drastically. However, this is not true of one’s obligations to one’s parents Keller notes [and here, Wee quotes Keller]:

There are some perfectly legitimate explanations of why you do not have certain duties to your friends which are not respectable when applied to filial duties. You cannot explain your failure to look after your parents by saying “Look, they’re great people…but over the years we’ve taken different paths. I went my way, they went theirs; it seemed the relationship wasn’t taking us where we wanted to go…” You are stuck with your filial duties, in a way that you are not stuck with your duties of friendship. (Keller 2006: 264)

Keller’s argument seems to be appealing to intuitions about filial duties. If so, I can imagine that there is some significant divergence from the one on which he relies, especially in the Western world. The explanation that he dismisses seems like one that lies behind much of cultural practice outside of, say, Confucian societies. In any case, I’m not sure common-intuition based arguments are convincing without some empirical evidence.

Keller’s alternative theory is a form of C above, with the relevant category formed by the obligations arising from “special goods” that are only obtainable through the parent-child relationship. Because some goods are only providable by a parent to a child or a child to a parent, special duties are engendered by that relationship. I have to agree with Wee that Keller’s examples – which seem to do all the argumentative work – are not very convincing (e.g. keeping in touch regularly, ensuring needed care, providing advice based on close knowledge of the other, etc). As Wee points out: “…the examples provided by Keller above, all of which emerge from a long and caring intimacy, may not show sufficient disanalogy with lifelong friendships….” (Wee 89)

Wee on Confucian Filial Obligations

As Wee construes Confucian filial obligations (CFOs), they also differ in significant ways from obligations of debt, gratitude, or friendship but for somewhat different reasons than the filial obligations of a more Western sort that Keller has in mind. Wee takes CFOs to be more formally specified to a significant degree, for example, than obligations of debt (Wee 90-91), taking the cue from texts like the Analects and Mencius and regarding CFOs to be highly ritually circumscribed. More strongly, Wee points out, CFOs require the honoring of obligations to be expressive of certain attitudes and feelings in ways that obligations of debt do not (92). Again, in comparison to obligations of gratitude as well, CFOs differ in requiring expression through specified rites according to Wee. Finally, CFOs differ from obligations of friendship primarily because, unlike those, the appropriate attitudes required in childrens’ honoring of CFOs to parents are reverence and deep filial respect.

Wee seems to favor an account of CFOs that involves special goods as Keller does with analytically understood filial obligations. However, her argument for that (on pp. 94-95), seems to be less about special goods that are only obtainable in a child-parent relationship and more about the “overdetermination” of CFOs by being rooted both in mutuality (as in friendship) and reciprocity (as in gratitude). She cites P. J. Ivanhoe as providing the argument that CFOs may be rooted in both of those rather than one or the other.

But suppose we grant Ivanhoe and Wee that CFOs are overdetermined in that way, through the following “very distinctive elements that make up” the child-parent relation:

  • “mutuality” – not between friends as equals, but shaded by guidance and admonition from parents, and respect and reverence in children
  • “boundless gratitude” – borne of both the gift of life and parental care

Wee thinks this allows us to conclude that “Given the very distinctive elements that make up this relation, there is nothing else that it is like” (Wee 95). It isn’t clear to me, however, that this is the right conclusion. It seems more sensible to conclude, given the ways in which CFOs are, by hypothesis, overdeterminedly rooted both in mutuality – as in friendship – and in reciprocity – as in gratitude – that CFOs can be analyzed either as significantly analogous to friendship or to gratitude, while allowing the caveat that the mutuality involved is slightly different from that of friendship (in the case of gratitude, there doesn’t seem to be any difference from other cases of gratitude). That would leave the specificity of ritual form and the appropriateness of accompanying attitudes as the remaining marks of difference between CFOs and obligations of gratitude or friendship. They may make for a difference – but do they make for significant differences?  Here I’m not sure any more what we’re after in the exercise and I’ll finish my comments with following questions:

Q1) If we analyze CFOs by modeling them on either obligations of friendship or of gratitude, while providing adequate caveats about the special expression-forms and feelings of friendship or gratitude that seem to characterize the child-parent relation, haven’t we done enough to understand CFOs adequately?

Q2) What is gained by claiming instead that CFOs should be regarded as belonging properly in their own category of obligation if in citing their “overdetermination” as obligations, we make references to obligations of friendship and gratitude that help to explicate their status as obligations?

(I have not included discussion of Wee’s points about CFOs being “role-obligations” (in pp. 95-96). Of course, if they are relevant to trying to answer my questions at the end, or if anyone is interested in discussing those, that would be great.)

Hannah Pang detail

16 replies on “New Issue of Dao, and Article Discussion”

  1. Thanks, Cecilia, for the essay and your willingness to share it with us, and to Manyul for these initial comments. I think that Manyul raises some stimulating challenges to the way that Cecilia formulates her conclusions. I’d like to raise a somewhat different kind of questions. Cecilia’s excellent objective is to uncover shared assumptions among Western, analytic writers on special obligations (including filial obligations), and to show how Confucian approaches differ. It seems to me, though, that there is an assumption underlying the whole essay that needs to be examined: namely, that we should understand Confucian talk about filiality in terms of “obligations.”

    In “Translating (and Interpreting) the Mengzi: Virtue, Obligation, and Discretion” (JCP 37:4), I spent a bit of time taking Bryan Van Norden to task for the degree to which “obligation” talk infected his “virtue” talk. One of the definitions of virtue that Bryan uses is: “Mengzian benevolence is a disposition toward agent-relative obligations involving the well-being of others.” I comment:

    In thus building the notion of obligation into his understanding of virtue, Van Norden is at the very least bucking the trend in recent discussions of virtue. Michael Slote has argued that a virtue ethics “makes primary use of aretaic terms” (like admirable, excellent, virtuous) rather than “deontic terms” (like right or wrong, permissible, and obligatory), which depend on how acts accord with rules. Virtue ethics “either treats deontic epithets as derivative from the aretaic or dispenses with them altogether.” If we look at two recent definitions of virtue, we will see what Slote means. According to Christine Swanton, a virtue is a dis- position to respond to items within its field (she also calls this “the demands of the world”) in an excellent way. Robert Adams says that virtues are persistently excellent ways of “being for the good.” It is arguable that if we look at earlier sources of Western virtue ethics such as Aristotle or Hume, we will also not find an idea of “obligation” as key to their understandings of virtue.

    If we look at P.J. Ivanhoe’s widely-cited essay on filially, “Filial Piety as a Virtue” (in Walker and Ivanhoe, eds., Working Virtue), we will notice that P.J. mostly avoids talk of “obligation” when he is giving his favored account of filial piety. Now I admit that this is complicated, and also admit two other things: that some talk of obligation, in a secondary way, can fit with a virtue-centered ethic; and that some talk of obligation may be apt for Confucian ethics. Explicit, ritual-based obligations might be (defeasible, based on “discretion”) important practical sources of character development. (For example, P.J. writes, “The role-specific obligations of parents, even more than the procreative functions of biological parents, establish the special bond that is the true basis of filial piety” (p. 303).)

    Of course, not everyone is convinced that one or more types of virtue ethics provides the best model for understanding early Confucian ethics, and I haven’t provided an argument for it here. (If I were to look for one in Cecilia’s paper, it would probably be based in the way that the core “inner feeling” aspect of filiality seems to be something of an afterthought on her obligation-based account.) My goal instead is to open up a conversation about whether, indeed, there is as much distance as I have claimed between an obligation-based account and a virtue-based account.

  2. Thanks to Cecilia and Dao and to our leaders for this wonderful new program!


    Steve writes,

    there is an assumption underlying the whole essay that needs to be examined: namely, that we should understand Confucian talk about filiality in terms of “obligations.”

    I have a similar worry, though I haven’t gone back to see whether the paper explicitly makes that assumption or instead can be read as speaking only about the obligations in filiality rather than about filiality tout court.

    The worry that came to my mind about the assumption may be slightly different from Steve’s. I take it Steve’s worry is mainly about whether (filiality or) Confucian filiality can be conceived exhaustively in terms of something like rules. My worry is mainly that the term ‘obligation’ gets its meaning from the metaphor of owing or debt. Indeed, Wee sometimes uses the phrase “what we owe our parents” as though it is synonymous with, indeed explanatory of, the phrase “filial obligations” (e.g. p. 84). The model of owing strongly suggests (but does not require) that (a) the main value of living up to the requirements is value for the obligees, and (b) the obligees can cancel the obligations by releasing the obligors. Like Wee, I’m not sure this is the right way to think about Confucian filiality. For example, LY 1.2 focuses on filiality’s training effect on the agent, and LY 2.21 talks about other broad social effects. More broadly, I wonder if one could argue that Confucianism conceives filiality mainly as a species of ritual propriety. Do departed parents benefit from one’s filiality?

    (‘Obligation’ may be similarly awkward in connection with friendship.)

    Maybe this worry of mine is misplaced. Maybe if we look at the broad social tradition rather than only at the earliest texts, the image of owing predominates. I wouldn’t know.


    It seems to me that in the Confucian tradition, filiality (xiào 孝) is thought to be continuous with, or indeed to include (as a part or a kind), a certain approach to one’s ancestors, elders, official superiors, and predecessors-in-office. It isn’t just about one’s parents. If Confucian filiality-toward-parents is to be likened to a wider set of relations than child-to-parent relations, one should look first to these others.


    There is maybe a specific reason to distinguish the question of the relation between Confucian filiality and such relations as friendship and debt, from the question of the relation between Confucian filiality and such relations as Confucian friendship and Confucian debt (not to mention the question of the Confucian view of the relation between filiality and such relations as friendship and debt): What such relations as friendship and debt amount to, or can amount to, may be different in a society that stresses Confucian filiality, from what such relations amount to in the kind of society contemplated by the anglo or European discussions of such relations.

    From Wee’s discussion on the bottom of p.93 I can’t tell whether she’s talking there only about very early China, or about friendship in the bulk of the Confucian social tradition.


    When Wee speaks of the idea that filial obligations are “identical” with obligations of friendship, she does not of course mean that we have filial obligations toward all our friends. I suppose she means filial obligations are a species of the obligations of friendship – i.e. that child-parent relations are a species of friendship. But I’m still not sure what that means. For surely there are different kinds of friendship. One species of friendship may differ from others and involve different obligations, yes? So the claim that child-parent relations are a species of friendship would not imply that child-parent relations are exactly similar to other friendships. Pointing to differences would not refute the species claim.

    But Wee rebuts the species claim by pointing to differences. So maybe I’ve misunderstood what she means by “identical.”

    • Hi Bill; your #4 is basically the same point I try to make with my Q1. I think Wee does mean species identity (or, as I called it, identity in kind) so that if filial obligations are or are very much like obligations of friendship (or gratitude), then they share significant characteristics with other subspecies of friendship (or gratitude) obligations, though not necessarily all. Maybe I didn’t make that point clearly enough above.

    • I see! I didn’t understand the phrasing.

      The “identity” terminology seems to claim to point to some definite idea, something that gets us beyond mere questions of degree, and I don’t see that it succeeds in pointing to or indeed reflecting any such idea.

      As this terminological point is very unimportant, I’ll dilate upon it.

      Saying that two sorts of things are “identical in kind” suggests to me either that (a) the two sorts/kinds are in fact the same kind, or that (b) things of the two sorts are all members of some third kind – but (a) is too strong, and (b) too weak, to be what anyone wants to point to here. If one means that (c) one sort is a species of the other (the other is a genus of the one), then using the term ‘identity’ seems unmotivated and thus misleading.

      Perhaps what Cecilia meant to point to by the “identity” wording is this idea: that the obligations of friendships (other than filial relations) are all and only certain forms of obligation, and filial obligations are all and only those same forms – as we might say the obligations of friendship in Perth are just the same as the obligations of friendship in Melbourne: they’re exactly similar, differing only in involving distinct individuals and meals.

      That is, in respect of obligations, (i) all friendships are exactly similar to each other and to filial relations; or (ii) if friendships (other than filial relations in case those are friendships) admit a range of variation in respect of forms of obligation, then fililal relations admit just the same range. Note that if indeed friendships admit a range of variation in respect of forms of obligation, then the obligations of some common species of friendship are likely not to be the identical with the obligations of friendship in sense (ii): the obligations of the species won’t vary over the same range as the obligations of the genus range over.

  3. If one wishes to avoid using terms such as obligation, duty, role ethics, or the three debts Manyul cites from Cecilia’s paper, would examining filial piety from the perspective of flesh be a viable alternative? This would allow us to give an account of the biological inter-dependency of parent and child while pointing out that as both parties age, the filial relationship fluidly and naturally inverts itself–the child assumes “responsibility” for the parents while she herself becomes a parent to her own children. What is more, do current accounts of filial piety explain what makes this a decidedly human relational experience and if not, what solutions, if any, are available?

    • How about examining filial piety from the perspective of non-human animals in Nature? While the instinct to protect and nurture the young is impressively evident “in the wild”, filial piety is non-existent. This should call into question not only the western ethical analysis of “child to parent obligations” but also – and especially – the Confucian teaching of filiality (xiào 孝).

      It would be interesting if Cecilia Wee’s comparative study could go beyond the confines of conventional philosophical boundaries.

  4. Just a few things in response to some overlapping concerns of Steve and Bill above:

    If you think of Confucian filial piety as mediated to a significant degree through ritual forms and practices, then talk of duties makes sense even if ritual propriety is conceived aretaically. Excellence in ritual propriety involves performing certain rituals dutifully, even if that requires subtle understanding of how to adjust those rituals to circumstances. More generally, perhaps, virtue ethical approaches to certain kinds of virtue — justice, friendship, and here, filiality — can’t ignore the roles that performance of duties and keeping one’s obligations, or understanding what one owes to particular others play in the exercise of those virtues.

    I think Bill is right that too much emphasis on concepts of owing or obligation can shift the terms of analysis in such a way as to distort our account of filiality (or of friendship, for that matter…). So, more properly, Cecilia’s analysis should be construed as interesting mainly for understand obligations as a subset of relevant items involved in understanding filial piety better, not as constituting the whole of filial piety. I think she’s pretty clearly offering the narrower analysis and not offering something more holistic.

    A few other things, more tangential:

    I’m not ever all that comfortable analyzing and categorizing types of obligation because I’m never sure what kind of analysis we’re left with at the end. Are we merely providing some kind of coherent account of the concepts involved or are we trying to provide a justifying account of the actual hold that those obligations have on us? I assume it is usually the former. If so, that’s interesting up to a point. Often, as someone interested in ethical theory, I care more about the latter. Given the conception of duties or of obligation within the virtuous exercise of filial piety, do those duties or obligations really have a hold on me? Should I care about my filial piety, for example, more than my happiness? Does my happiness really depend essentially on the exercise of the virtues, as eudaimonists claim? These are large questions — I’m only mentioning them rather than actually asking them in this context.

  5. Hi all – I’m so sorry for not coming in earlier on the interesting debate that has been going on. I’ve been out of town for a conference at Seoul National and just got back late last night. (Manyul – do you know Sukjae Lee?)

    Hope to get my thoughts together in a couple of days when things are a bit more settled, and maybe offer some responses and discussion to the various interesting points, e.g., on friendship and obligations.

  6. Dear all – Thanks so much for all the stimulating and very thoughtful comments on my paper. They provide much food for thought. I don’t think, though, that I’ll be able to address all the issues that have been raised. So I shall offer answers to a conspectus of some of the issues.

    Perhaps it might be helpful to outline first why the paper was framed in the way that it was. The question of whether there are special obligations (as opposed to duties owed to all other persons) has had considerable discussion in the analytic literature, and in particular in political philosophy. In the course of looking at the literature while working on another project, I came upon some discussions in the analytic literature (e.g., Jeske, Jecker, Dixon, Blustein) that discussed the issue of special obligations specifically in respect of friends, families etc. Their papers examined special obligations in terms of gratitude, friendship etc.

    I then also came upon Philip Ivanhoe’s widely cited paper in Working Virtue (which I shall hereafter call WVP). (By the way, pj sent me a very kind note when the paper first came out, and said he liked its ‘lucidity and insight’.) Bill mentioned that he thought that Ivanhoe avoided the use of ‘obligation’ in that paper. But I think pj does use that word (see e.g.,p.298 third para.) It was moreover also clear that pj was in WPV framing the discussion in terms of special obligations of gratitude, debt etc., not unlike the Western discussion. Indeed, he cites the Western literature in the discussion (see, e.g., his footnotes on p.4 and 5), and he uses the same categories of special obligation as the analytic literature – viz., debt, gratitude etc.

    While pj sought to elucidate filial piety in terms of this framework, my paper had a slightly different project – to contrast the Western and Eastern views of filial piety by examining how their respective accounts of filial obligations would stack up against the usual special obligations. Obviously, this was a pretty large project (and there were times I thought it could only work as a book-length project!).

    With this background in place, here are some responses to the issues mentioned in the various posts.

    1. Steve and Bill raised some questions on whether a virtue-theoretic account would work better than an account focusing on obligations. I agree that this is an interesting question – but it is part of a much wider debate. I think my paper should be taken in situ – that is, as part of a present ongoing argument and discussion in the existing literature on the nature of filial obligation, and how it might be seen in relation to other species of special obligations. The paper thus already assumes that we are looking at the issue of filiality in terms of obligation, rather than some other framework.
    The further issue of whether filiality should in fact be framed in terms of virtue-ethics or obligations is a much broader issue and requires much more discussion. I don’t think it can be adequately addressed within this particular paper. But I hope the possible lack of resolution within the larger debate does not negate my small contribution to an ongoing smaller debate involving special obligation.

    My point above also provides a sort of response to some of Manyul’s comments. The discussion is framed in these terms because the original discussion in the analytic literature, and also pj’s paper, were framed in these terms.

    2. There was also a question as to whether filial obligations are analogous to or identical a given special obligation. Manyul suggested that they might be thought of in terms of species identity. This is a really interesting question which would bear further examination. All I can say is that I found in the existing literature itself a certain lack of clarity – viz., filial obligations were sometimes treated as identical, and sometimes as analogous, to other obligations. You’ll appreciate that the paper is already a pretty long one as it stands. I didn’t feel that I could address this issue in detail, so I simply took the weaker claim (viz., that it was analogous) to hold unless specified. (See p. 85 para 2) But I think it would be great if someone could look further into the issue.

    3. It was also asked if my paper was limited to early Confucianism or more broadly focused. It was indeed mainly focused on early Confucianism (as is some of pj’s paper). As I’ve mentioned, it was evident that it was becoming a sprawling project so I thought it best to limit the discussion to the early Confucians.

    4. Perhaps the best way to end is with Manyul’s two thoughtful questions. Answer to Q1 – Yes, we can analyze CFOs in terms of friendship and gratitude, but there is a question of whether this would exhaustively cover the nature of the CFO. I would argue that it does not. This is because the sort of gratitude and friendship by the child forms, so to speak, a sum that is greater than its parts. It’s not just gratitude per se or friendship per se – but friendship tinged with respect and reverence, and boundless gratitude, both coming together. ‘Gratitude-plus’ and ‘friendship-plus’ then form an amalgam so distinctive that there is nothing else that it is like.
    Answer to Q2 – Yes, there is indeed overdetermination in the sense that the child would have an obligation to the parents either in terms of gratitude OR in terms of friendship. Either of these two, independent of the other, would render the child as having an obligation to her parent. Nevertheless, I think there is room for a new distinctive category for the reasons cited above. Namely, that the amalgam of the two (and perhaps the depth of each) for one’s parents makes it an altogether a distinctive relationship.

  7. Thank you, Cecilia! I understand now. CFOs are indeed a huge topic, and I think your project in the paper makes a great deal of sense. Just a few footnotes:


    “Bill mentioned that he thought that Ivanhoe avoided the use of ‘obligation’ in that paper. But I think pj does use that word (see e.g.,p.298 third para.)”

    I haven’t seen the paper. What Steve said above was, “P.J. mostly avoids talk of ‘obligation’ when he is giving his favored account of filial piety.”


    “1. Steve and Bill raised some questions on whether a virtue-theoretic account would work better…”

    That was just Steve, but we each expressed worries about the view that (Confucian) filiality can be understood entirely or mainly as fulfilling certain obligations. (For example, I think duties other than obligations may play a big role.) If I understand you now, you’re saying the paper doesn’t involve that view; it simply talks about the obligations that are part of filiality without saying how big a part of it they are.

    I don’t think of this as a question about what philosophical or grounding “framework” for us to use; I think of it as a question about the broad-strokes shape of the thing itself as it has been understood.


    “3. It was also asked if my paper was limited to early Confucianism or more broadly focused. It was indeed mainly focused on early Confucianism (as is some of pj’s paper).”

    I would distinguish between the scope of the claims and the scope of the evidence offered. My question was about the time-scope of the evidence about friendship cited at the bottom of p. 93.

    My sense is that the comments on filial obligations and duties in the Analects and the Mencius are scanty and express a range of deeply differing outlooks, and sometimes maybe are just designed to get Mencius out of a jam. I think filial piety became a core issue for Confucianism only after Confucius’ death. Also I’m not sure how far the Shijing – as we read it – counts as evidence about early Confucian philosophy, though it may be good evidence about early Chinese social practice.

  8. Sorry, I seem to be unable to agree with very much of the above discussion on the grounds that it all seems to be premised on the assumption that human beings are all ‘rugged individuals,’ as is often claimed in the West, rather than Confucian ‘inter-related types.’ One of my dictionaries quotes the following:
    “Shun by his filial piety [xiao] was able to [ke = let go, enable, allow] harmonize with [xie = bring into accord with] his father, stepmother and stepbrother” [New Complete Dictionary; 93].”

    So, when a child achieves [xie] with his elders (allowing xiao as wide a scope as you like beyond fathers alone), this is to realize dynamic harmony [he] in the given external circumstances which in turn results in the internal experience of moral delight [le]. Xiao is the way of relating to one’s elders in a delightful way and the more often this experience occurs, the deeper the relationship becomes. The same for friendship, older and younger sibling relationships and husband and wife relationships. Accordingly, a child will continually try to adjust its xiao relationship with its parent (“I just want to make them proud of me,” etc.) by behaving with li in better and more effective ways—and also by remonstrating (as much as this may be possible) with the parent to improve their relationship. Often this is easy, sometimes hard. So there is, as has been said, some obligation involved here as, until such time as the child is able to achieve this external harmony/internal delight with its parent, the child usually continues to try, even after the two come to be estranged. And similarly with other human relationships: Of course, non-family friends (including co-workers and colleagues) and spouses are typically those others who are, at least at first, open to establishing this experience of harmony/delight (this thickening of the inter-relationship), but there is not the same ‘obligation’ to continue when and if ‘harmony/delight’ fails to be mutual. Guan Zhong did it by switching sides and joining Duke Huan, thereby re-establishing ‘harmony’ and becoming a sort of hero to Confucius.

    • That made sense to me until you got to Guan Zhong. First, which version of the tale are you referring to–and then how does Guan Zhong’s behavior exemplify xiao?

    • Hi David,

      Which part of the discussion above seems to you to be premised on the “rugged individual” idea? I’m not seeing it. Thanks!

  9. If I understand David right, he’s saying we should avoid talking about relationships in a one-sided way, by concentrating (in this case) on the child and the filial obligations on her. Rather the relationship should be understood at all times as including both parties.

    I don’t really agree with that point, because the Confucian tradition is so full of talk of self-cultivation and moral praise and blame directed at the individual. If a child fails to be filial, Confucius wouldn’t say, “My goodness, that parent child relationship is not very harmonious today!” Confucius would say: “You’re a bad son.” So while I agree that in Confucianism the relationships are primary and can never be dissolved, there’s still a lot of room within that tradition to talk about individuals’ contributions to the relationship.

    More generally, as I read the paper and Manyul’s comments, I was wondering about psychological reality. Whether or not it is true that “filial obligations are debt obligations,” it certainly seems that we don’t experience them primarily in that way. Of course, how we experience relationships is highly culturally conditioned, but surely we experience relationships primarily as personal connections? I don’t want to go so far as to say each relationship is unique, but I do think that your relationship with your parents is probably unique.

    The consequence of this would be particularly salient for education. If a developing mind experiences the relationship with their parents as a relationship, it would be relatively ineffective to try to educate the young person through generalised obligation-related or friendship-related rituals, even if filial obligations are (or can be analysed as being) nothing but a particular type of obligation/friendship obligation. So to the extent that Confucianism is concerned with shaping families and societies, it must pick out filial relationships as special.

  10. Paul, I was trying to come up with a case in which it turns out to be all right to switch sides, something which can’t (or can’t easily) be done in family setting, but can be done in a socio-political setting. The story I have in mind is that Guan Zhong did not die with his lord, Duke Huan’s older brother, Prince Jiu, but instead jumped ship to Duke Huan’s side. Though the justification for his switch is dicey (though arguably a clear case of quan 權 = moral counter-balancing), what he eventually did was to organize and peacefully settle disputes with the city state of Qi by creating the ba system, and so saving the Chinese people from wearing their hair hanging down loosely like wild men, and their clothes from being folded to the left side. This success was a kind of justification for his refusal to die with his lord which Confucius appreciates [Lun-yu, 14:16, 14:17]. Yes, I am aware of Confucius’ criticism in LY 3:22.
    Philip, what I wanted to concentrate on is of course the two-sidedness of relationships, and of course Confucius would throw a filial failure in jail (along with his father), but what I wanted to emphasize was that the way in which to tell whether or not a real relationship is going on is not by judging whether or not certain obligations or debts have been properly fulfilled, but on whether, for the people involved, a xie (a bringing into accord with) has occurred, a dynamic harmony [he] has been realized in the situation between them and a contented, overwhelming feeling of moral delight [le] has washed over the people involved. It is not that the child ought to feel some sort of xiao debt which he can then use to motivate himself to do better (or that the father must feel an ai commitment and use that to express himself better)—these are what I am calling ‘individualistic’ as compared with ‘related-istic’– but that it is by experiencing this wave of delight that both sides are motivated to do more of that kind of relating to each other. Not obligation, but moral delight. And, by the way, the better both get at performing li (not rote ritual, but with the heart-mind engaged [LY3:12]), the better their inter-relationship tends to become. Maybe think Muppets. Or think ‘crises bring you closer together,’ and that sort of thing, etc., etc.

  11. Thanks Phil! That’s really helpful to me.


    One thing I like about David’s first remarks is that he points out that whether the notion of obligation applies may depend on whether the parties are not fully realizing the ideal filial relation. (This point is similar to Kant’s point that the concept “ought” or moral requirement has its main application in cases where there is some motivation not to do what one ought — which, he says, is always or almost always for all human beings, though not for e.g. God.)

    David doesn’t say explicitly whether he thinks people who are in the full swing of ideal filial relationships with their parents still have filial “obligations.” But I suppose that’s a fairly trivial terminological point one way or the other. Maybe that’s because I think that even the Confucian tradition sees perfect filial relationships as quite rare.


    Here are three propositions:

    1. There is more to having filial piety than fulfilling obligations.
    2. There is much more to good filial relationships than the younger parties’ filial piety.
    3. There is much more to good filial relationships than the younger party’s fulfilling obligations.

    As for 1, most of us above are on record as either pushing for 1 or at least explicitly not denying it.

    As for 2, I think it’s obvious enough to go without saying.

    As for 3, it’s transparently implied by 1 and 2, and is also pretty obvious on its face. I imagine that everyone in the discussion agrees with it.

    All that is true never mind whether we’re talking about Confucian or American filial relationships.

    (However, if by “filial relationship” one means the kind of relationship that would make filial piety appropriate, then arguably the “filial relationship” (in Confucianism) is approximately the bare fact that the one person is a parent of the other. And in that case the “filial relationship” is simply different from – not more or less than – anybody’s fulfilling obligations.)


    Of course it’s a little odd, because one-sided, to speak of a relationship as being a “filial” relationship.


    David, here are three different views:

    4. For Smith to have an obligation to X, she must lack any other motivation to do X except a certain sense of obligation.
    5. For Smith to have an obligation to X, she must have some slight motivation that pushes away from doing X; she must at least be potentially vulnerable to some temptation not to X.
    6. For Smith to be behaving perfectly, she must be doing everything she does from the feeling that it is morally required.

    I think 4 is actually a strange and implausible view; I think maybe nobody accepts it. But it sounds as though maybe, David, your objection is based on your accepting 4, or your thinking that some others here accept it.

    I think 5 is controversial, though so far as I know it hasn’t received a great deal of discussion. I think few people think it matters much, because none of us is motivationally perfect.

    I think 6 is accepted by a very small minority of Western philosophers, and that pretty much all of that minority accept it only on condition that it’s understood to allow that we may also have other compelling motivations to do the same thing.


    Phil, when you write

    Whether or not it is true that “filial obligations are debt obligations,” it certainly seems that we don’t experience them primarily in that way. Of course, how we experience relationships is highly culturally conditioned, but surely we experience relationships primarily as personal connections?

    I wonder whether you are eliding the question “how do the obligations feel” with the question “how does the relationship feel”?

Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.