Empirical Perspective on Mencius-Xunzi Debate

Eric Schwitzgebel, one of our contributors, is having problems registering for a WordPress ID, so he has asked me to post this podcast of a lecture that he presented on the Mencius-Xunzi human nature debate. It is also cross-posted over at his blog The Splintered Mind:


Podcast of “An Empirical Perspective on the Mencius-Xunzi Debate about Human Nature”

… given to the Confucius Institute of Scotland on Jan. 19, available for listening or download here.

The podcast is audio-only, so you won’t see the overheads. I don’t think you need to see the overheads to understand the talk. But for completeness here they are (as MS PowerPoint 2003).

You may also be interested to see this article, which was part of the basis for the talk.



Comments welcome!

9 replies on “Empirical Perspective on Mencius-Xunzi Debate”

  1. This is some really interesting work, and it is definitely a good idea to think of the debate between these thinkers as focused on what is necessary for moral development rather that what we’d be like in a state of nature.

    I do think we have to be careful about aligning Mencius’ view with that of someone like Kohlberg too quickly. Kohlberg’s work on both moral development and moral education is very focused on cognitive and intellectual development, and he has little to say about either the development of moral sentiments or the role of learned behavior in spurring moral growth. In regard to behavior in particular, I think there is a pretty striking difference between Mencius and Kohlberg. Li and the cultivation of particular relationships in accord with pretty well defined social roles seems to point to a fundamental role for steering children toward conformity with external norms in Mencius. What’s distinctive about Mencius is his comparatively optimistic view that a) this behavior itself leads to the development of sentiments that are already present in children and b) this proper behavior can be elicited without too much coercion. (Xunzi is less optimistic on both counts.) What this suggests, I think, is that the dichotomy that Western child develomentalists take for granted–between external and internal sources of behavior–doesn’t fit too well with the Confucian picture.

    • Hi Dennis. But I think if you’re right, then the dichotomy between external and internal sources does fit well with at least Xunzi’s Confucian picture, though he is, as you say, less optimistic about the internal sources. My point here is just that there’s more than one “Confucian picture.” But I think I agree with your main point. There are also probably other reasons not to align Mencius with Kohlberg, for example, we don’t really know whether Mencius was basing his views on empirical observations or instead on the stipulative dictum that anyone who doesn’t feel sympathy for a child about to fall into a well isn’t really human. (I know that some people have criticized Kohlberg for scandalously sloppy empirical methods, but that’s a different point.)

  2. Hi Eric,

    I haven’t listened to the whole podcast, but I have read the paper, which is a nice exercise in reformulation, and very illuminating about Mencius, Xunzi, Hobbes, and Rousseau. What a shame that we had such a deluge of other topics right after your posting here.

    I have tried hard to come up with objections, and I think I’ve found two. First:

    As you say, it seems that Mencius and Xunzi disagreed on whether “human nature is good.” If their disagreement was real, what does it come to? (This was discussed in the April 21, 2008 string of Manyul’s blog.) You propose that the disagreement can be characterized in terms of a question that can be fruitfully explored empirically, viz.:

    “Must standards of good behavior be imposed on people from outside, by artificial means, as we might, say, impose the shaving of chins and legs, the stretching of necks, the dyeing of hair? Or does morality emerge without external imposition from normal processes of human maturation, drawing on the environment principally for nutrition and support?” (150)

    “The core question on which [Mencius and Xunzi] disagree is this: Is morality something imposed on people from the outside … or something that arises in the normal process of human development if people are encouraged to reflect …? In other words, is moral development a process more of indoctrination or self-discovery?” (152f)

    To the disjunctive question, my own answer is No. Nor do I think the truth is simply a mix of these two. Compare expertise in chemistry, or repairing bicycles well rather than poorly, or the virtue of navigating one’s neighborhood well. These are normally not imposed on us, nor do they emerge from reflection supported by adequate food, water, and shelter.

    I think there are two importantly different dimensions to the dichotomy as you present it. One is coercion versus freedom–as in the idea of morality’s being “imposed.” The other is about the source of structure–as in the conception of the environment as offering only material support, not direction or form.

    When a seed grows into a tree, the seed brings the form and the environment brings the matter, roughly. When a sculptor makes a “tree” from clay, the proximate source of the form is external, in the sculptor. (In both cases the farther source of the form is other trees.) Maybe sculpting is coercive. But when a vine grows around a tree, taking the shape of the tree, there doesn’t seem to be any imposition, and yet the vine’s environment is the source of the vine’s tree-shape.

    (a) Insofar as morality is prima facie pointless, like wearing neckties, the disjunctive question looks pressing. Why do people do that? Does it reflect a brute genetic impulse, or do we instead do it from brute social pressure? But (b) insofar as morality strikes us as a reasonable response to what is objectively important in the shape of things, then I think both disjuncts seem prima facie false.

    Between (a) and (b) is (c): one thinks one can isolate certain distinct parts of morality as basic to the rest, such as desires for some end (or some type of end) which the rest of morality somehow serves instrumentally; and one finds these parts prima facie pointless, like desiring to wear neckties. If one has view (c), then one is likely to find the “Natural or imposed?” question apt with specific regard to that basic motivation, though not otherwise with regard to morality.

    Consider a different kind of view, (d): a concern to do the things that are moral isn’t grounded in any basic sui generis motivations, but in our cognitive powers and processes — such as empathy and the experience that we tend to make cognitive mistakes if we don’t respect enough, don’t care enough about people, don’t try to see things from others’ points of view, etc.

    I think a common experience is that one acts in some way that is nice or not nice for others, and they reflect that back with a smile or a frown, a return invitation or a kick. Their action might be read as a bit of mild coercion or application of an incentive. But it might also be read as a kind of communication, a display of some of the significance of what one did. “See how it feels? That’s a taste of your own medicine!” When I better appreciate what it is like to be on the receiving end of my actions, I will act differently. That doesn’t mean I’m acting egoistically.

    Setting coercion aside, it is hard to know what it means to ask, for morality, whether the source of the structure is external or internal, since morality is shaped very much by the fact of our similarity.


    In your paper, you rightly say we shouldn’t count life without society as “natural,” because, for example, biologists “do not separate the ant from the colony or the wolf from the pack to see how they behave ‘naturally’” (148). You therefore propose this admittedly rough account of what it is for a trait to be natural to an individual:

    “A trait is natural to an individual just in case it arises in that individual through a normal process of development in a normal, nutritive environment, rather than as a result of injury, acquired disease, malnutrition or (especially) external imposition.” (148f)

    A potential problem here is that external imposition — for example, patterns of behavior enforced by threat and violence — would seem to be a normal and natural part of human social life, as it is among many other social mammals (more than injury and disease). It would seem to be a salient part of normal moral education. (Short of threat of violence there are other little digs and disappointments without which human social life is almost inconceivable, and which seem to “impose” things.) Of course we can only speculate about the usual conditions of prehistoric tribal life, but it seems likely that our brains evolved in contexts of mutual coercion. Our being hardwired to be receptive to such imposition might be thought to be a way in which we’re “naturally” moral. And it might be that any individual could learn to be moral by reading and reflecting because there has been enough coercion to set up examples of moral behavior and moral society.

  3. Thanks for the comments, folks!

    Dennis and Manyul: I agree that there are important differences, including the ones you mention, between Mencius and Kohlberg. It would be boring if everyone who thought that human nature was good had exactly the same view of moral development! I do think Mencius and Kohlberg share the inward-out model of moral development that makes it appropriate to say that they think human nature is good, despite other differences between them.

    Bill: Your objections are very thought-provoking! I agree that not all views of development may fit neatly onto the human nature is good/bad/somewhere-between framework that I’ve articulated. I’m not sure a (d) view isn’t an inward-out/good view, though. There wasn’t anything in my statement of the view that said that moral motivations had to be sui generis. Does it follow implicitly from something I’ve said? (Maybe, but I don’t see it yet.)

    On your second point, I’ve got some of the same trouble Xunzi does here, I think. I used to worry that since it was natural for human beings to make artificial things, his natural-artificial distinction collapses. Your worry about my view seems to be a version of that worry about Xunzi: Since imposition is a natural part of our environment, my distinction collapses. Here’s why I think that objection doesn’t work (if I’ve got it right). Even if imposition (the artificial) emerges naturally in human beings, we can still distinguish products of that imposition (artifice) from products of other processes. I think it’s going to be hard to get any distinction between natural and artificial going, with respect to human behavior or creations without some maneuver of this sort. Do you disagree?

  4. On my second objection:

    Even if imposition (the artificial) emerges naturally in human beings, we can still distinguish products of that imposition (artifice) from products of other processes. I think it’s going to be hard to get any distinction between natural and artificial going, with respect to human behavior or creations without some maneuver of this sort. Do you disagree?

    I agree with the latter point. But my worry wasn’t that coercion emerges naturally and so its consequences are natural. I was thinking rather that coercion is (because natural) normal, so that when you define “natural” traits as traits that “come from what’s normal rather than from e.g. coercion” (here I paraphrase), you’re building inconsistent standards into the definition. Maybe you meant traits that “come from what’s normal except coercion.” In that case I have less of an objection.

    But there’s still this: it may be that it is possible or even easy to learn to be moral without threat, i.e. without being threatened or coercively imposed upon, but only because any one individual can absorb without coercion the forms of behavior that have been coercively imposed on most people.

    And there’s still this: Unless I’m just reading carelessly and forgetting things, which I do all the time (especially to Steve), I think you don’t say enough about what you mean by “imposition” to suggest whether one should classify normal face-to-face interaction (the kind that’s necessary to support basic mental health) as imposing or not imposing attitudes and forms of behavior. It’s full of little carrots and sticks and threats and promises.

    Also I think there’s a difference, not just one of degree, between the coerced and the artificial. For coercion, threat is key. For the artificial, design is key. On the one hand, there can be threat that is directed without design. Think of Sidgwick’s idea that since (at least) we tend to dish out what we get, we all naturally enforce on each other modes of behavior that benefit ourselves; so that society tends to enforce general patterns of behavior that make people benefit those around them. And on the other hand, behavior can be influenced by design without threat, as when people are drawn by beautiful music. (Might morality be like a beautiful composition?)

    As for my first objection, which is largely unrelated to the second:

    I quite like my image of the vine. When the vine takes the shape of the strange tree it grows on, one might ask, how much is that because the vine’s genes include a blueprint for that shape, and how much is it because somebody imposes that shape on the vine? Or how much of each of those two is involved? The answer is that none of either of those is involved.

    I didn’t mean to suggest that anything in your menu of options implies that moral motivations are sui generis. One can hold what I called view (a) without having any views about moral motivation at all.

    My suspicion is not just that “not all views of development may fit neatly onto” the framework you articulated, but that most major sorts of human activity don’t fit approximately into that framework, and morality is one of the things that fits least well. On the other hand, I’m just not sure enough of what the framework comes to.

    At one of your poles there’s reflection, or self-discovery. But from what you say it’s unclear to me whether the mark of this pole is (I) reflection as opposed to lack of reflection, or (II) reflection pure of certain other influences. I don’t tend to associate reflection with self-discovery, myself; I tend to associate it with coming to understand all manner of things: the penal law, chemistry, math, my vices. Being imposed upon can involve lots of thinking, depending on what counts as imposition. Ask a rat in a maze, or a philosopher trying to get tenure. And discovering one’s own values might often require more time than thinking.

    I have to stop now, right in the middle of making my case. I’m sorry! And I fear I’ve said things I wouldn’t say if I had time to look back over your paper. I’ll be back in a couple of days.

  5. Thanks, Bill. Your comments are very helpful!

    * You’ve convinced me that I should at least tweak my statement of what makes a trait natural to be that it emerges from a normal process of development in a normal nutritive enviroment AND NOT as a result of injury, acquired disease, malnutrition, or external imposition. The “and not” leaves it open the extent to which the latter can belong to a normal, nutritive environment.

    * On your second point above: Maybe people differ, and maybe there are alternate ways. I acknowledge in passing at least the first of these options. But people who think human nature is good or human nature is bad do not endorse such views, so it’s fitting that a characterization of what it is for human nature to be good or bad does not sort those possibilities in either group.

    * I’ll let “imposition” be in the eye of the beholder (or the philosopher), as I’m letting “normal” and “nutritive”. Thus, “natural” will be in the eye of the beholder. That’s a feature, not a bug.

    * On your last point in the first part: Let me clarify that I use “imposition” not “coercion”. Then let me make the eye of the beholder move. Finally, let me add that if morality arises not from imposition, injury, disease, or malnutrition, and also not from not a result of a normal process of development, then the resulting view is neither a BAD nor a GOOD view.

    * I like your vine analogy too, so let me run with it a bit. The key thought I want to add is to remind you of my claim that a trait is natural to a species if it arises in normal members in a BROAD RANGE of normal enviroments. In a broad range of normal environments, vines will be green and will cling to surfaces. This is natural to the species. In a narrow range of enviroments, they will twist in [this particular shape]. So being green and clinging to surfaces is natural to vines but not twisting [in this particular shape]. However, twisting [in this particular shape] is a natural trait for that particular individual vine, on my account, because it meets the conditions for a trait’s being natural. So the species-level and individual-level yield different answers. A banzai situation is (or will probably be viewed as) imposition, so the shape will not be natural to that individual in that case.

    So here’s the question for your view: Is morality something that arises in human beings in a broad range of normal environments without imposition? Or does the environment have to be structuring in some relatively unusual way? If the former, you might be in the GOOD camp, after all — though maybe “inward-out” would still not be a fitting label.

    Good luck with tenure!

  6. Good discussion Eric and Bill. I wonder if “feel” plays a role. Maybe for Xunzi, morality felt imposed, whereas for Mengzi, it felt natural.

    I find myself thinking of Zhuangzi, when he says “Always rely on that which occurs of itself and not on an embellishment to life/nature” (常因自然而不益生也). He might not be critical of a morality that develops unconsciously and implicitly learned, (as opposed to explicitly taught/imposed).

  7. Hi Eric,

    1. Regarding my objection to your account of a trait’s being natural-to-an-individual: yes, switching to ‘AND NOT’ helps a lot, and is better than my proposed ‘except’.

    The vine analogy can be adjusted to make it relevant to the account of a trait’s being natural to a species: suppose Smith vines happen to be common only in continent Jones, which happens to be dominated by Jones trees, so that Smith vines normally take the shape of Jones trees …

    2. Regarding your proposed formulation of the human nature issue as an empirically explorable question, I want to propose that what you’ve formulated is still, in one or another sense, not an empirically explorable question. The senses I have in mind are illustrated by the following questions, each of which is in some sense not empirically explorable.

    i) Which do geese come from: barnacles or old rags? Or at least, which is closer to the truth?
    ii) Which of our social traits arise from our normal growth in our normal soil, and not from our ability to use language?

    3. The difficulty in exploring (ii) is just that one can’t tell what’s meant, because it’s so deeply counterfactual. (I’m not sure how redundant I’m being when I add:) Also one needs to know how direct or immediate is the “arising” that the questioner is talking about.

    Your question is deeply counterfactual insofar as “imposition” is hard to unravel conceptually from the normal support our environment gives us. For example, (A) a certain amount of imposition may be part of what it takes to give us various other traits that are not specific to morality: a natural language, personal hygiene, empathy with strangers, respect for others’ judgment. (If I smell bad, others may behave toward me in a way that discourages me from developing respect and care.) For another example, (B) it may be that basic mental health requires fellowship with people who are self-respecting enough to demand big and little things now and then: applying what is collectively called “social pressure.” (You’ve declined an invitation to say that such things aren’t cases of “imposition.”) Especially when we’re young and learning to socialize, the pushes and pulls of social life can be overt and visible. And (C) speaking collectively: the peace and security that allows people to learn the flute and look kindly on others may depend on people’s not behaving too violently or fraudulently, which may depend in turn on the core of the penal code.

    4. We can avoid these problems about counterfactuality, and other fine points about the notions of “imposition” and the ways of defining the other pole, by adjusting the formulation of the Empirical Question: Make it clear that when we ask “Is the truth X or Y?” we’re asking about certain extreme pictures involving nothing very subtle. (And if there is to be a follow-up, replace the quantitative follow-up “or is the truth closer to X or to Z?” with “or neither?” or “or what?”. That way we don’t have to worry about how to quantify degrees of closeness according to a cluster of diverse standards; we don’t have to work out any definite concepts to define an inquiry other than “X or Y?”)

    But then we’re at barnacles and rags.

    5. “I use “imposition” not “coercion”. … I’ll let “imposition” be in the eye of the beholder.

    I said “coercion” because I thought “imposition” implied it and I wasn’t sure what else “imposition” might imply. To be more accurate to the word I should probably have said “coercion or force or pressure,” thus muddying the waters at that pole. On the de-muddying side, maybe “imposition” suggests conflict and frustration. I’m not sure “imposition” implies artificiality of what’s imposed, though I think you take artificiality to be central to what you mean. Well, something can be artificial in that its design was thought up, or in that what puts it into effect is human effort. I suppose arithmetic is artificial in both senses.

    6. I had sketched the idea that perhaps any individual could have become moral without imposition, but only because most people have been made moral by imposition. You replied, “People who think human nature is good or human nature is bad do not endorse such views, so it’s fitting that a characterization of what it is for human nature to be good or bad does not sort those possibilities in either group.” But it seems to me offhand that this Free Rider possibility might well be acceptable to HN-BAD people (perhaps one of our natural vices is passive mimicry). Actually I think your characterizations agree (earlier I carelessly had a different view). One of my reasons for mentioning the case was to highlight one of the difficulties in the way of empirical exploration: an imposition-based world might allow for lots of free riders. But you’re already alive to such problems, as your paper shows.

    7. Good thinking and good cognitive functioning are often associated with freedom, but can also be characterized as a high degree of efficiency in accepting influences from one’s environment, even becoming a model of one’s environment.

    8. The example of the philosopher wasn’t about me, but thanks! And I hope nobody gets the impression that I don’t really like the paper, because that would be false.

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.