Warp, Weft, and Way

Chinese and Comparative Philosophy 中國哲學與比較哲學

Moral Perfection in the Analects (and Beyond?)

As part of some commentary on Paul Goldin’s new book Confucianism, Bill Haines has noted the centrality of the idea of moral perfection in Goldin’s characterization of the basic convictions of the Confucian philosophical orientation. In a subsequent comment, he questions to what degree the ideal of moral perfection was actually held by Confucius. With Bill’s permission, I’m re-posting his comments on this latter subject here, because I think they deserve more attention than they may get buried deep in a comment string. Read on, and please direct your comments to Bill, the author of what follows.

Bill writes:

Paul’s list of five basic convictions in Confucianism has an elegant unity, in that its points center on two main ideas: “morality” and “perfection” – or on the one idea, “moral perfection.” But I wonder about the extent to which this is idea was shared by Confucius. There is at least some prima facie reason to suspect that the very idea may be profoundly at odds with Confucius’ views.

I don’t have an opinion yet about whether he actually did hold this idea, or about whether there is significant evidence either way; I haven’t thought about this question before today, nor read the Analects with the question in mind, or indeed read the Analects recently, so maybe this elaborate comment has an easy and obvious answer. Here I’ll just set out the prima facie reason to wonder.

The following oversimple view is familiar in the West: Morality has two parts. One part consists of fairly clear rules (e.g. don’t kill people, don’t steal) that can and should be completely obeyed, and that can be associated with the terms ‘right’ and ‘wrong’. Some of these rules may be associated with particular offices or positions, e.g. a borrower must return what she borrowed, and a captain must go down with the ship. The other part of morality consists of what might be called scalar values, and are more properly associated with such words as ‘bad’ and ‘better’ and ‘good’ and ‘excellent’. Examples might include virtues, like rén, wisdom, or courage; or other things like moral self-cultivation, working for peace, or succeeding in benefitting others. The more of the scalar things you do the better, other things being equal, and there may be no such thing as doing them completely or all the way.

At first glance the notion of moral perfection or flawlessness is more at home in the former part of morality; or in morality conceived on the model of that part. Here are three reasons. (1) We tend to find clear rules plausibly obligatory only when they are not beyond most relevant people’s powers to obey completely. (2) Also, human society seems to need a groundwork of clear rules to ground security, peace and cooperation, rules we can pretty much count on people to obey; and the analogous point is true on a smaller scale, e.g. in any particular organization. So it is common (if perhaps fallacious) to think of the rule part of morality as the elementary part, the part for total obedience; and the scalar part of morality as the proper focus for high aspiration. (3) Perhaps most importantly, complete obedience to any attractive set of rules is pretty much guaranteed to be conceptually possible, i.e. conceivable. In setting out the rules, one is automatically describing complete obedience, i.e. perfection by those standards. Nothing of the sort is true for scalar values. There is nothing unusual about holding a scalar value such that the idea of complete success or perfection is just as nonsensical as the idea of the greatest whole number. Hence one might well associate the idea of scalar morality, or high moral aspiration, with the rejection of the very idea of moral perfection.

It is popular recently to associate Confucianism, or at least Confucius, with “virtue ethics,” which is naturally associated with scalar values (better/worse rather than right/wrong), and with conceiving the whole of morality on the model of the scalar part. For that reason alone I think we should be careful about attributing the idea of moral perfection to Confucius and to Confucianism in general.

Not that a scalar moralist can’t believe in perfection. Indeed, virtue theory is sometimes associated with a natural teleology whereby morality is thought to be about trying to realize some ideal form for our species, or something like that, as Aristotle and Mengzi seem to have thought; and this view makes prima facie sense of the idea of moral perfection, at least in the abstract.

Still, even that kind of view can have difficulty with the idea of perfection. Think of the idea of actually drawing a perfectly straight or perfectly circular line, with pencil on paper. On the one hand there is in some sense a clear concept of perfection, and it is intelligible to aim at those perfections in our drawing. Those are the targets we should have our mental eyes on, not the drawing of an only slightly wiggly line. On the other hand, we just as readily grant that achieving such perfection is utterly impossible for anyone, even with compass and square. (I guess the image 不逾距 from LY 2.4 is not about ideals or models, but about boundaries that one can stay well within, as one may obey clear rules.)

The drawing problem arises from a tension between the ideals and the matter. Aristotle was troubled by a similar tension, I think, between his abstract ideals (the contemplative knower, e.g. God) and the limitations of the human species qua matter for that form. More familiarly, he may never have resolved the deep tension among his picture of individual perfection, his notion of humanity as an essentially social species (almost like bees), and his picture of a community’s perfection. Off the top of my head I don’t recall what signs there might be of either sort of tension in the Analects. I imagine Bryan Van Norden must have discussed such things in his book Virtue Ethics and Consequentialism in Early Chinese Philosophy, but I forget.

There is, however, a different sort of potential tension in the Analects that arises from a lack: Confucius seems to have lacked a term equivalent to our ‘morality’ (and most of what he says in the Analects may be specifically about the vocation of public service or even the role of ruler rather than about norms applicable across the board). I think he may pretty much have the idea of morality anyway, and I’ve argued in print that he does. But if we find him talking about perfection in one or another part or aspect of what we call morality, we can’t simply infer that he has thought of moral perfection. For one thing, moral perfection might involve trade-offs, a balance among morality’s parts; e.g. one might think morality counsels the occasional lie. For another thing, even if there is perfection in one part of morality, there may be another part of morality regarding which perfection makes no sense.

So: is the idea of moral perfection evidenced in or alien to the Analects? Anyone?

January 20th, 2012 Posted by | Confucianism, Confucius | 6 comments

6 Responses to Moral Perfection in the Analects (and Beyond?)

  1. Bill Haines says:

    Pursuant to a suggesting from Steve, I’m adding here a further comment I had posted in the thread about Paul Goldin’s book:

    ( Old Business: I was very obscure in my original comment above, points (1) and (2) of the paragraph “At first glance, ….” My little thought was only that when we think of the rule part or model of morality, we commonly think of complete obedience as being well within the reach of the ordinary relevant people, so that a focus on this part or model of morality would encourage and suit the very idea of a completeness or perfection in morality, or moral perfection if we conceive all of morality by that model. )


    I’ve read through Books 1-15 of the Analects looking for material on perfection; this is my report. It’s always great fun to read the Analects looking for something one hasn’t looked for before!

    If Confucius had the concept of moral perfection, it wasn’t the sort of thing he emphasized or focused on. His followers’ conspicuous deep puzzlement about how to conceive e.g. rén (9.11) strongly suggests that he was not interested in articulating standards of perfection. He was much more concerned with identifying what was important (for his interlocutor) to focus on for practical purposes; what is good enough. For example: 1.5 “The Master said, ‘To lead a country of ten thousand chariots: respect the tasks and be trustworthy; be frugal in expenditure and care for people; and call upon the people only in the proper seasons.’” Similarly e.g. 1.6, 2.17, 4.20, 6.8, 12.6, 12.7, 13.17, 13.19, 14.12; cf. 5.1, 5.8, 13.20-1, and 1.1.

    I haven’t found any persuasive sign that Confucius had the concept of moral perfection; I’ve found some signs that he had a different view. Here are the passages that seem closest to evidencing the idea of moral perfection, and some comments.

    9.22: “The Master said, ‘There are cases in which the blade springs, but the plant does not go on to flower! There are cases where it flowers but no fruit is subsequently produced!’”

    The image suggests a natural teleology and hence perhaps a concept of human perfection, which one might identify with moral perfection. But the absence of other passages making a similar suggestion is compelling reason to think that this was not Confucius’ framework.

    11.16 “… Shi oversteps, and Shang doesn’t go far enough … one is as bad as the other.”

    If one thinks virtue is a point simply between two other achievable points on an otherwise conceivable scale, that amounts to the idea that there is such a thing as moral perfection (at least as a theoretical ideal). But the absence of other passages suggesting a similar view is compelling reason to think that this was not Confucius’ framework.

    4.3 “The Master said, “If the will be set on virtue, there will be no practice of wickedness (無惡).” (Legge’s translation)

    2.4 “…After seventy I could follow my heart’s desire without stepping over (踰) the line.”

    How high a standard is the line he is referring to? We aren’t told. Confucius did often point out that he had flaws and limitations (5.9, 5.28, 7.17, 7.22, 9.8, 14.28); but we do not know how old he was when he said those things. (Confucian readers: if you think he was perfect during much of the time he taught, you must think that humility constrained him to deny it; so you must think the standard he articulates at the end of 2.4 is not perfection.)

    6.27 “The Master said, ‘The superior man, extensively studying all learning, and keeping himself under the restraint of the rules of propriety, may thus likewise not overstep what is right (可以弗畔).’” (Legge)
    12.15: “The Master said, ‘By extensively studying all learning, and keeping himself under the restraint of the rules of propriety, one may thus likewise not err from what is right (可以弗畔).’” (Legge)

    I submit that in these passages and 2.4, the image of the line or boundary-marker refers to a distinction between right and wrong actions, or recognizably right and wrong actions, rather than perfection vs. imperfection of character or desire. I suspect that the image of not stepping over the line is supposed to sound modest.

    8.1: “The Master said, ‘Tai Bo may be said to have reached the highest point of virtuous action (至德). Thrice he declined the kingdom, and the people in ignorance of his motives could not express their approbation of his conduct.’”

    The translation is redundant in a way; the word for “reached” does double duty as “highest point of.” The root image of arrival is operative here, and inherently suppresses the idea of levels of virtue, i.e. the idea of going farther than arriving at virtue at all. Perhaps we should render the phrase as “got all the way to virtue” or “really arrived at virtue”? The thought may simply be that Tai Bo’s virtue was a good solid example of virtue. “Tai Bo – now that’s virtue.”

    8.21: “The Master said, ‘I can find no flaw in the character of Yu (吾無間然矣). He used himself coarse food and drink, but displayed the utmost filial piety towards the spirits. His ordinary garments were poor, but he displayed the utmost elegance in his sacrificial cap and apron. He lived in a low, mean house, but expended all his strength on the ditches and water channels. I can find nothing like a flaw in Yü.” (Legge)

    It’s not clear to me that the point is that Yu has no imperfection of character. I submit that what is meant is Confucius sees nothing to point out as a problem.

    8.19: “The Master said, ‘Grand indeed was Yao as a sovereign! How majestic was he! It is only Heaven that is great (大), and only Yao corresponded to it (則之). How vast was his virtue! The people could find no name for it. How majestic was he in the works which he accomplished! How glorious in the elegant regulations which he instituted!’” (Legge modified)

    This last passage suggests that Yao was distinctly greater than both Tai Bo and Yu, so that Confucius seems not to have thought of these two as having reached the pinnacle.

    Further, in 8.19 Confucius seems also to be emphasizing that on the scale on which Yao is great, there is something greater than all human beings. The suggestion seems to be be that all human beings are born incapable of the highest greatness. So if the highest greatness is moral perfection, all humans are born incapable of moral perfection. Or at least, no sages to date have been morally perfect.

    Further, the image used to represent greatness is largeness (大), and this image does not even suggest the concept of perfection. To people accustomed to abstraction, this image vividly suggests ruling out the concept of perfection. Of course, we would need more evidence than that to prove that Confucius did not have the idea of moral perfection.

    Indeed what Confucius means by greatness here may differ from morality in being broader: possibly including things that arguably do not fall within the scope of what we mean by the word ‘morality’ (when we try to report Confucius’ views): certain talents and skills, for example, or knowledge, or success in generating good results. If that is so, i.e. if Confucius operates with a comprehensive concept of greatness that does not quite fit the topic we mean by ‘morality’, then it calls into further question the idea that phrases elsewhere that do not specifically mention morality or any other particular realm – e.g. “stepping over the line” or “going too far is as bad as falling short”– are about morality rather than about something a little different.

  2. Steve Angle says:

    There’s a lot that I find really stimulating in your thoughts here, Bill. As you might imagine, perfection (including moral perfection) is a subject that I take up more than once within Sagehood. As a starting point to engaging with you, here are two passages from the book. The first specifically concerns the Analects. I do agree with you that the idea of perfection, and/or specifically moral perfection (whatever exactly we take that to me; more later), is not explicit and may not appear at all, especially in early strata of the text.

    A number of critical changes take place within Confucian writings of the classical period.2 Since my main focus here is on the later Neo-Confucian time period, I will only glance briefly at the question of stages of development within the classical period itself. Suffice it to say that we have good evidence to believe that the Analects, thought to be the most reliable source for Confucius himself, was not composed all at once, and expresses differing attitudes toward the idea and content of the sage. In one relatively early passage, for instance, sagehood is associated with being variously skilled, and Confucius is at pains to distinguish it from the superior, because moral, status of “gentleman ( junzi 君子).”3 I will put off for now the question of the relation between sages and junzi; instead, note that elsewhere (and perhaps chronologically later) in the Analects, sagehood has become an elusive and mysterious state, surpassing even Confucius’s master virtue of humaneness (ren 仁).4 Its sense of mystery is only enhanced by various comments linking sagehood to Heaven (tian 天), the ultimate source of value for most classical Chinese thinkers [Chen 2000, 416–17]. It will retain an air of mystery ever after, though most Confucians will nonetheless insist that it can be attained by anyone. As Mencius famously asserts, “The sage and I are of the same kind.”5 (Sagehood, p. 14)

    One of the things I want to suggest here–especially in light of your comment #1 on LY 8.19 and 8.21–is that “perfection” can be taken in a while variety of ways. One might be not making any mistakes, and this fits well with your thought, in the original post, that perfection fits better with the rule-governed part of morality than with the “scalar” part. But another thing one may mean by perfection is acme, culmination, completion. I only halfway agree with your suggestion that the surpassing mystery of this means that we are incapable of the highest greatness: perhaps this is what (parts of) the Analects have in mind, but it seems not to be the view of many later texts.

    Speaking of later texts, since Paul was not speaking of the Analects alone, but of Confucianism more broadly, I think it is fair game to ask about subsequent developments. Another bit from Sagehood:

    Finally, sagehood from the classical period on becomes conceptually tied to supreme moral virtue. As one scholar puts it, “the Confucians attach the highest moral character to the sage, such that not only are sages supremely wise, but they are also the embodiment of perfect, transformed human relationships and morality” [Wang 1999, 29]. This same author points out that while there had been a connection between sages and “virtue (de 德)” in some pre-Confucian texts, this becomes complete or perfect virtue during the classical period. Thus Mencius contains a passage in which sages are said to be the “culmination (zhi 至)” of human relationships; another text speaks of sages as “the culmination of virtue (zhi de 至德).” Another way of expressing the same idea is Xunzi’s, who says that sages “exhaust ( jin 盡)” human relationships, meaning that sages fulfill such relationships completely.8,9 Sages thus come to represent the human achievement of moral perfection. (Sagehood, p. 15; references are to Mencius 4A:2; Zhou Li, Qi section; and Xunzi 21, respectively. “Culmination” is D.C. Lau’s apt translation [Lau 1970, 118].)

    I think I have more to say, but my daughters are calling….

  3. Bill Haines says:

    Hi Steve! Sorry for the delay, and thanks very much for the quotes.

    the Analects… expresses differing attitudes toward the idea and content of the sage. In one relatively early passage, for instance, sagehood is associated with being variously skilled, and Confucius is at pains to distinguish it from the superior, because moral, status of “gentleman ( junzi 君子).”

    I think you mean 9.6, in which an official associates sagehood with having many arts, and Zigong (perhaps tellingly) doesn’t think to challenge the idea. I read Confucius there as rejecting that view even of the sage, but there is certainly room for argument.

    I also think there are good reasons for Confucius to think the junzi should be a generalist that don’t depend on the idea that “junzi” is a purely moral concept or status; but here maybe the issue is just that you and I hear the word ‘moral’ in slightly different ways. I think of a junzi as someone qualified to lead, and one feature of the special role of a leader is to be a big-picture person.

    elsewhere (and perhaps chronologically later) in the Analects, sagehood has become an elusive and mysterious state, surpassing even Confucius’s master virtue of humaneness (ren 仁).

    Do you think 6.30 may be later than 9.6? The features of a sage mentioned in 6.30 look to me difficult rather than mysterious: “gave extensively to the common people and brought help to the multitude” (Lau).

    (Lau’s version has Confucius say, “Even Yao and Shun would have found it difficult to accomplish as much (堯舜其猶病諸).” I gather most other translators agree. Paul gives: “Could even [the sage kings] Yao and Shun find infirmity in someone [like that]?” (18), suggesting a claim about flawlessness or blamelessness. Conceivably it should be translated “Would even Yao and Shun have minded acting like that?” There are lots of new and intriguing readings of Analects passages in Paul’s book.)

    “perfection” can be taken in a whole variety of ways. One might be not making any mistakes, and this fits well with your thought, in the original post, that perfection fits better with the rule-governed part of morality than with the “scalar” part. But another thing one may mean by perfection is acme, culmination, completion.

    Point taken: one might think of morality mainly in terms that are scalar as “small” is scalar, while still implying a limiting case.

    And I’d put it this way: the concept of moral perfection is simple, though vague enough to admit radically different conceptions. To be morally perfect is to be as morally good as can be, in all that one is and does.

    A. Thus one kind of conception of moral perfection is that one is always doing one’s best. That is, what counts as “perfection” is handicapped (as a horse may be handicapped) by ignorance (or alternately, unavoidable ignorance) of circumstances, limited physical control, dilemmas encountered, bad upbringing, a genetic inability to empathise – there is much room for variation in conceptions. This sort of conception of moral perfection tends to suggest that the claim “All humans are capable of moral perfection” is a simple tautology, empty of content. In brief: ought implies can.

    B. Another conception of “as good as can be” takes the “can” not as applying separately to individuals, but as applying to the species (or some part of the species such as males, Greeks, etc.). The best that a human can be, then, is not automatically by definition something most people are capable of, or even something that anyone alive today is capable of. So the claim that we’re all capable of it certainly has substance. And this kind of perfectionist might think the relevant standards pertain most directly to character (presumably as relevant to action); or to actions; or to both, with neither sort of standard wholly reducible to the other.

    I only halfway agree with your suggestion that the surpassing mystery of this means that we are incapable of the highest greatness

    (I’m not sure what “this” you are referring to, and mysteriousness hadn’t crossed my mind.) Regarding Confucius’ statement in 8.19 “Only Heaven is great, and only Yao followed it,” my thought was that it says there is a level of greatness – that is, a level of the sort of quality we should aim at having – that no human has achieved, or (as is presumably suggested) ever could achieve, though it is not only conceivable but is in fact realized, albeit by a very different sort of being. And separately, 8.19’s image for greatness is of a vividly limitless quantity, which fact is in some slight but perhaps insignificant positive tension with the hypothesis that Confucius had articulated for himself the idea of perfect greatness.

    Even given all that, Confucius’ remark still leaves open the possibility that there is such a thing as the farthest a mere human can go in the direction of greatness. But once we start looking at things that way – understanding greatness in a non-species-relative way, and then handicapping for the limitations (imperfections?) of a species – is it then natural or unnatural to suppose that the greatest possible greatness for Smith is the same as the greatest possible greatness for Jones? What might make it natural is a general theory of greatness or goodness that makes it species-relative for other reasons, as Aristotle had. But then we’d have a little trouble with the idea that only Heaven is great. And even Aristotle might wonder who is the greater tennis player: Billie Jean King or Bobby Riggs.

    In fact human society may need diversity of characters, so that there can be no such thing as the set of virtues everyone ought to have. Many years ago Amélie Oksenberg Rorty suggested to me that human society may need people with a diversity of kinds of moral theory! I think she then published something to that effect, but I don’t know what or where.

    • Bill Haines says:

      Oops, I think I was wrong about the concept of moral perfection. I said “To be morally perfect is to be as morally good as can be…” – but a better account would be, “To be morally perfect is to be all the way good,” with the reference to possibility or capability being one way to cash that out.

  4. Bill Haines says:

    I guess my argument about 至 in 8.1 was pretty silly. I’d love to hear more about the image behind that term, and usage before Confucius, if anyone has something handy to say.

    I’m not sure exactly what Taibo is supposed to have done, but if it was refusing the crown three times as Caesar did in the play, from a position of not yet being monarch, it doesn’t look like an apt image of moral perfection to me. Still, yielding “all under heaven” is yielding not just a huge amount, it’s yielding what is in some sense everything.

    Other passages, such as 1.1, look to me like answers to one or another challenge about Confucius’ not being in office. Perhaps Confucius’ aim in the remark was not to hold up Taibo or this kind of action as an example illustrative of perfection, but rather to defend his own refusal to accept or pursue some job.

    . . .

    Steve here and Paul in the other thread have mentioned that the image of completion might underlie a conception of perfection, an acceptable use of the English ‘perfect’. That seems right. Point taken, Steve, about the two parts of morality. And as Paul said, 成 is about completion, so we should look for the idea of completion in its cognates as well.

    I’ll think out loud about the image of completing something, or something’s becoming or being complete.

    When I think of “completing,” I think first of making something: completing a pot or a house. Not much is implied about the quality of the product, except perhaps that it is not missing something that is essential to being a pot or house at all (or to being a halfway decent one), and also not such as to suggest that the process of production is still underway.

    Houses may differ from pots in this way: what counts as a “complete” one has to meet a higher standard than what counts as the outcome of “completing” one. A “complete house” is one that has a balanced array of the things that make a house a good one or a very good one. That’s still well short of what would inspire us to call something a “perfect house.” In any case, if we did call something a “perfect house” I suppose that if challenged we would be quick to grant that the phrase is hyperbole.

    Living organisms offer another image of completion and completeness. For most kinds of animal and perennial plant it might simply suggest adulthood (as in the modern use of “成人”), like a pot that has been completed. An adult human needn’t be good or healthy or strong or smart (and at least idiomatically, I think, “成人” doesn’t tell us whether she has any arms).

    If we emphasize a distinction between the contributions of nature and nurture, and think of nature as providing a blueprint while nurture provides only materials, we can have a concept of an organism’s completion that allows us to hypothesize that sometimes, or even usually, an organism that has a long and healthy adulthood never becomes complete. What other background views could support that hypothesis?

    One might think the relevant standard is the same within each species. But what would a perfect dog be like? And would a perfect human be male? I wonder when Chinese thinkers began to discuss what counts as a species. No doubt they were more concerned with vertical kinship than horizontal kinship. The example of dogs could function to drive a wedge between the idea of an individual’s blueprint and the species blueprint, and/or undercut the latter idea: surely the individual inborn mandate for the pup of dachshunds is not a mandate for dogs in general? The same goes for the observation of distinctive inherited physical features of humans.

    What conception of species or kinds is Youzi using in Mencius 2A2?:

    豈惟民哉 How could it be so only with people?
    麒麟之於走獸 The Unicorn male and female in relation to walking beasts,
    鳳凰之於飛鳥 The Phoenix male and female in relation to flying birds,
    泰山之於丘垤 Mount Tai in relation to hills and mounds,
    河海之於行潦 The River and the Sea in relation to gutter puddles:
    類也 They are of a kind.
    聖人之於民 The Sage in relation to people:
    亦類也 Also of a kind.
    出於其類 Sprung from their kind,
    拔乎其萃 Rising above the many.
    自生民以來 Since people have been living,
    未有盛於孔子也 There has never been a greater than Confucius.

    Turning to another point: the images here don’t suggest to me completeness, or an upper limit. Rather they positively suggest limitless room for improvement. (盛 at the end: does it suggest overflowing? Filling a vessel beyond full?)

    One might think of morality by analogy with more specialized excellences, departments in which one might be 善.

    I wonder whether on the whole these would have tended to suggest an inherent upper limit. What it is to be a great runner, a great warrior, a great tennis player, a great historian – these suggest to me the absence of an upper limit, though I’m not absolutely sure about the historian.

    A compleat angler has the full range of available arts, skills, knowledge, and (perhaps) tools. (Confucius woudn’t have thought of relativity to stages of technological progress.) I guess one can be a compleat angler without being a truly great angler, and perhaps vice versa. There are dimensions to angling excellence that don’t seem to have upper limits, such as speed in finding the best spots in unfamiliar lakes and streams. But for angling, unlike running, one thinks less about (a) “how much speed” than about (b) “how little time.” Unlike quantity (a) for a runner, quantity (b) for an angler has a clear limiting case whose difference from close approximations is not even potentially interesting.

    Archery has two salient dimensions: how far the archer can shoot, or the strength of the archer; and how far from the target the arrow lands. The former dimension is saliently limitless; the latter is not. There’s indirect evidence that Confucius’ preferred conception of good archery lacks the former dimension (LY 3.16). I don’t know if that’s potentially telling about limits in his idea of morality.

  5. Bill Haines says:

    A separate discussion of moral perfection in the Analects continued here, and Alexus has put up a very interesting and relevant post over at Unpolished Jade.


Leave a Reply

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