Human Nature and Moral Education
in Mencius, Xunzi, Hobbes, and Rousseau
Department of Philosophy
951 776 9075
May 22, 2006
Human Nature and Moral Education
in Mencius, Xunzi, Hobbes, and Rousseau
Mencius, Xunzi, Hobbes, and Rousseau were all political philosophers well known for their views on “human nature”. I argue that, to some degree of approximation, their views about human nature can be best understood as views about the proper course of moral education, and that, consequently, a picture of moral development stands near the center of each man’s philosophy. I then suggest that we can explore empirically which philosopher was nearest the truth.
Human Nature and Moral Education
in Mencius, Xunzi, Hobbes, and Rousseau
Mencius, Xunzi, Hobbes, and Rousseau were all political philosophers well known for their views on “human nature”. I will argue in this essay that, to some degree of approximation, their views on human nature can be best understood as views about the proper course of moral education and that, consequently, a picture of moral development stands near the center of each man’s philosophy. I will then suggest that we can explore empirically which philosopher was nearest the truth.
1. The “State of
The dispute between the 17th-century English philosopher Thomas Hobbes and the 18th-century French philosopher Jean-Jacques Rousseau regarding human nature is generally cast – and was indeed by Rousseau himself sometimes cast – as a dispute about what people (or “man”) would be like in the “state of nature”, a state without social structures or government. Hobbes famously writes in the Leviathan that the “naturall condition of mankind” – his condition prior to establishment of the state – is one of misery and “Warre, where every man is Enemy to every man” and life is “solitary, poore, nasty, brutish, and short”. We are propelled into violent competition by the desire for limited goods and for glory, and due to our relative indifference to the suffering of others. When a man in the state of nature sees something he wants – such as the goods or wife of another man – he will try to obtain it, if he can do so consistently with his own safety, regardless of the pain or death it may bring to others. The inevitable result is continual insecurity and strife, and the failure of any stable agriculture or industry, until men are eventually persuaded to submit themselves to a government for their own protection.
Rousseau, equally famously, paints a very different picture of the “state of nature” in his Discourse on the Origin and Foundations of Inequality among Men. Man in the state of nature “breathes only peace and freedom; he wishes only to live and remain idle”. “[H]is heart yearns for nothing; his modest needs are easily within reach”. He is moved only by the urge for self-preservation and by basic biological needs, and by a natural pity for others. Sufficient food is easily enough obtained. Sexual couplings are brief and without social complication. He thinks only of the here and now, not planning for the future, not attempting to elevate himself in the eyes of others, not fearing death, and lacking the bloated desires for prestige and luxury that are nearly universal among civilized men.
It is sometimes suggested that Rousseau is less sanguine about the state of nature in The Social Contract, published seven years later, in 1762. He certainly does, in Chapter 6 of that work, envision that the time may come when the survival of man requires exiting the state of nature and entering some sort of civil society. Perhaps, though, Rousseau is only recognizing here that the state of nature portrayed in his Discourse on Inequality requires a ready abundance of food and may be unsustainable if food becomes scarce. And there may be reasons for thinking that Rousseau felt man’s behavior in conditions of plenty better reflects his “nature” than his behavior in conditions of scarcity – if, for example (to anticipate somewhat the next section of this essay), Rousseau regards the environment of plenty as the “normal” environment of humankind. In any case, in Emile, also published in 1762, Rousseau appears still to hold human nature in high esteem – we will return to Emile later – and in his essay to Christophe de Beaumont, also that same year, he writes that “the fundamental principle of all morality about which I have reasoned in all my Writings ... is that man is a naturally good being, loving justice and order; that there is no original perversity in the human heart, and that the first movements of nature are always right.”
The famous claims about “human nature” in the Leviathan and the Discourse on Inequality appear to pertain, as I have said, to how human beings would behave without government or stable social structures. But it is, in a way, very strange to suppose that our behavior absent social structures is our natural behavior. Biologists do not, for example, separate the ant from the colony or the wolf from the pack to see how they behave “naturally”. The ant and the wolf are naturally social. Their behavior within their social structures is their natural behavior – an isolated ant or wolf is an aberration. Human beings are, of course, the same in this respect.
If this is right it seems also to follow that the “state of nature” thought experiment is, at best, misnamed. If Hobbes’s and Rousseau’s aim in conducting this thought experiment is to begin an inquiry into the value (or disvalue) of society and government, it may have some purpose; but it would be less misleading to call the conditions they imagine “governmentless” or “societyless” than “natural”. With this in mind, it is tempting, perhaps, to dismiss Hobbes’s and Rousseau’s comments on “human nature” (despite their importance as political philosophy) as only marginally related to the biological or psychological or anthropological study of the patterns of thought and behavior truly natural to human beings.
2. A Developmental Approach to the “Natural”.
Let’s think more carefully, then, about what it means to call a human trait “natural”. Consider some intuitive examples: My brown hair is its “natural” color, while my friend Jeanette’s dyed black hair is not. My sexual attraction to women, which began in adolescence (if not earlier), is natural; a crack addict’s indifference is not. My wife’s small feet are natural; a traditional Chinese woman’s feet, if their growth was restricted by tight binding, may be unnaturally small. A trait may be natural to an individual but not, generally speaking, natural for the species: The limited early social attachments of a child with autism may be natural for her but not what we would consider natural for human beings in general. There is some merit in thinking about the “natural” as essentially relative to individuals, as I have in all the examples thus far. However, our four authors aim to make broad generalizations about human nature. Thus (though not only for this reason), it will be useful in this essay to consider what is generally natural to human beings, such as hair color within a certain typical range, adult foot size within a certain typical range, attachments of a certain sort in early childhood, etc.
I propose that we consider a trait 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. A trait is then natural to a species if it is natural to normal members of that species in a broad range of normal environments. A trait needn’t be present at birth to be natural: Adult size, sexual attractions, secondary sexual characteristics, etc., are not present at birth. Nor need a trait be genetically determined to arise in all environments – what phenotypic trait could possibly arise in all environments, anyway? – just in normal, nutritive ones. In abnormal or deficient environments a “natural” trait may generally be absent: For example, normal individuals may have white hair or no hair in environments with enough high-energy background radiation. In environments with severe nutritive deficiencies, people may not grow to their natural heights or develop their natural sexual characteristics. Even in normal, nutritive environments, some aberrant individuals may not acquire a particular “natural” human trait (though the alternative traits they do acquire may be natural for them), as in the case of the autistic child’s social development. A trait may be common among normal individuals yet not natural because, in some sense, it is “externally imposed” – for example, the cropped ears of Dobermans.
This characterization of the natural is both normative and flexible. It assumes that we can distinguish “normal” individuals from abnormal ones; “normal” and “nutritive” environments from abnormal and deficient ones; injury, disease, and malnutrition from healthy processes; and “external imposition” from its absence. A variety of objections may be raised against the unreflective use of such terms. They may often have no determinate applications or their applications may be politically influenced. This, however, is no objection to the proposed definition of “natural”, since the intuitive application of the word “natural” also generates such worries. So, for example, whether we consider homosexuality “natural” depends on whether we think it arises without external imposition in normal members of the population in normal, nutritive environments. Those who deny the naturalness of human homosexuality are apt to assert that it is “imposed” on people, or that homosexual individuals are aberrant, or that homosexuality arises only if the individual’s developmental environment was abnormal or deficient in some way. Or consider the overweight of many Americans. Can a 250 pound man legitimately argue that that is his “natural” weight? The answer depends, in part, on whether the contemporary American nutritional environment, with its superabundant refined sugars and fats, should be considered normal.
I acknowledge the legitimacy of concerns about what might be hidden in decisions about what counts as “normal”, etc. But the proper answer to such concerns, if you find them overriding, is not, I think, to redefine “natural” in some more objective way, but simply to avoid the word. I see no reason to avoid the normative, as long as it’s made explicit; and the present definition does, I think, make the normative dimension of the term nicely explicit. As I see it, problematic uses of the word “natural” derive not so much from the term itself but from the tendency to use it as though what’s “natural” can be determined by objective and apolitical biological measurements alone.
With this understanding of the “natural” in hand, let me reframe the key idea of the previous section. A relatively stable social system is part of a normal, nutritive environment for human beings (and for all social animals). Even in times of political revolution, much of one’s social environment remains the same. The traits that arise among people raised outside such environments are not our natural traits, but often quite the opposite. The most extreme case of this, of course, is the occasional “wild child”, starved of human interaction through much of childhood. Such children may lack language, fear people, and so forth; but that is hardly the natural condition of humankind. The “state of nature” thought experiment thus misses its target completely if its target is the isolation of our natural traits.
Here, then, is how I would prefer to interpret the question about whether human beings are “naturally” violently competitive or placidly compassionate: Look at how those character traits or behavior patterns arise in human development. 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?
These two alternatives are of course too stark, and they are not exhaustive. Among other possibilities, human nature could be mixed, with some elements driving us toward compassion and others driving us toward violence, or different people might have different natures, some naturally inclined toward violence and some toward compassion, so that we can’t say there’s a single “human nature” on this head. Also, by casting the debate here in terms of “violent competition” vs. “placid compassion” (because of the role those ideas play in my simple portrayal of Hobbes and Rousseau), I don’t mean to focus unduly on compassion and violence as the fonts of morality and immorality or assume too simple a relationship between violence, compassion, and morality.
Understood as I recommend – as a claim about development – the question whether “human nature” is good or bad begins to gain some scientifically interesting content.
3. Mencius and Xunzi on Human Nature.
What I have suggested about how to approach questions of human nature – that is, by looking at how development occurs in a normal social environment – certainly seems to be in tension with Hobbes and Rousseau, or at least certain common portrayals of them (though more on this later). It comports more easily with the views of the classical Chinese philosophers Mencius (4th c. BCE) and Xunzi (3rd c. BCE) (and was, in fact, inspired by them). Mencius famously said that human nature is good (xing shan 性善), Xunzi that is bad, ugly, unappealing (xing e 性惡). In this section, I will try to convey the general flavor of their views on human nature. At the end I will suggest what I take to be the unifying thread.
Human nature’s being good is like water’s tending downward. There is no human who does not tend toward goodness. There is no water that does not tend downward. Now, by striking water and making it leap up, you can cause it to go past your forehead. If you guide it by damming it, you can cause it to remain on a mountaintop. But is this the nature of water?! It is that way because of the circumstances. That humans can be caused to not be good is due to their natures also being like this.
The trees of Ox Mountain were once beautiful. But because it bordered on a large state, hatchets and axes besieged it. Could it remain verdant? Due to the rest it got during the day or night, and the moisture of rain and dew, it was not that there were no sprouts or shoots growing there. But oxen and sheep then came and grazed on them. Hence, it was as if it were barren. People, seeing it barren, believed that there had never been any timber there. Could this be the nature of the mountain?! When we consider what is present in people, could they truly lack the hearts of benevolence and righteousness?!
These passages (and others like them, esp. in Book 6A) establish that Mencius regards human nature as good, in some sense. These passages they don’t fully express the content of that view, but one thing is already plain: To say that human nature is good is not to say that all people behave well. Water can be dammed up and kept on a hillside. A mountain that naturally tends to be verdant can be bald. Indeed, Mencius thought the decadent times he lived in were rife with wickedness. Rousseau did also, for that matter. It’s an undergraduate mistake to think that the view that human nature is good is in any straightforward way undermined by the prevalence of evil in the world. The question is not whether evil abounds; it’s whether evil is “natural” or, instead, a perversion.
But what, exactly, is it for us to “tend toward goodness” if many (even most) of us do not achieve it? According to Mencius, just as all (normal) feet are roughly the same and all (normal) palates prefer roughly the same tastes, all normal hearts delight in righteousness (yi 義 – moral rightness). Mencius builds a case for this claim on the basis of what he takes to be normal, spontaneous reactions to circumstances in which what is right or wrong is plain and in one’s face, as it were. The first impulse of the beggar who is given food in an insulting manner is to reject the food, even though doing so may cost him his life. The first reaction of anyone who suddenly sees a child about to fall into a well is a feeling of compassion. The first reaction of people on seeing the dead bodies of their parents eaten by foxes and bugs is to want to bury the bodies. Such universal impulses are the seeds or sprouts (duan 端) of righteousness, benevolence, and propriety. Moral development results from attending to, cultivating, and “extending” these natural moral impulses, noticing and acting upon the heart’s pleasure in right action; evil results from suppressing the heart’s natural desires, subverting them to the desires of lesser parts of oneself such as one’s stomach, eyes, or limbs, or failing to think through the similarities between nearby cases and those farther away.
Xunzi begins his essay “Human Nature is Bad” like this:
People’s nature is bad. Their goodness is a matter of deliberate effort [wei 偽 – deliberate effort, conscious activity, the artificial]. Now people’s nature is such that they are born with a fondness for profit. If they follow along with this, then struggle and contention will arise, and yielding and deference will perish therein. They are born with feelings of hate and dislike. If they follow along with these, then cruelty and villainy will arise, and loyalty and trustworthiness will perish therein. They are born with desires of the eyes and ears, a fondness for beautiful sights and sounds. If they follow along with these, then lasciviousness and chaos will arise, and ritual and the standards of righteousness, proper form and good order, will perish therein. Thus, if people follow along with their inborn nature and dispositions [qing 情 – dispositions, emotions, essence], they are sure to come to struggle and contention, turn to disrupting social divisions and disorder, and end up in violence. (3rd BCE/2001, p. 284, ch. 23).
So for example, Xunzi says that when we are hungry, our natural emotions or dispositions (qing 情) lead us to want to eat without regard for others; it is only by artificial social convention that we come to accept waiting our turn. Similarly, unless you find this essay unusually gripping, your natural emotions or dispositions probably incline you to leave off reading and take a rest. It is only by artificial social means that you are driven to work as hard as you should. Moral rules are an invention of the Sage Kings, a set of artificial constraints imposed on people for the proper functioning of society. With time, one can transform one’s desires to as to align with the proper strictures, but this is a slow, difficult, and unnatural process, one that does not comport with our original impulses.
Despite the starkly different mottoes, it can come to seem unclear where the difference between Mencius and Xunzi lies. They agree on one key point: That people often behave badly when driven by basic bodily impulses, and moral behavior requires the regulation of those impulses by the heart or mind (xin 心). In light of this agreement, it is sometimes suggested that Mencius and Xunzi are much closer in view than it may at first seem: They disagree only (or primarily) in how to define the “natural”. Xunzi holds that anything arising from the “conscious activity” or “deliberate effort” (wei 偽) of the heart is artificial, and thus that morality is artificial. Mencius thinks the best products of the heart are natural and thus that morality is natural. But the difference is merely semantic.
I strongly disagree with this interpretation. But where, then, should we locate their disagreement? The reader will recall my preference for interpreting questions about human nature developmentally. I believe that Mencius and Xunzi implicitly accept this approach (despite Xunzi’s occasionally simplistic remarks about the “natural” being what is present at birth). The core question on which they disagree, I would suggest, is this: Is morality something imposed on people from outside (Xunzi) or something that arises in the normal process of human development if people are encouraged to reflect for themselves (Mencius)? In other words, is moral development a process more of indoctrination or self-discovery?
4. Metaphors for Moral Development.
We can get a better hold of the conflict between Mencius and Xunzi on this point by looking at the different metaphors they use for moral development. Mencius repeatedly compares moral development to the cultivation or growth of a sprout. Xunzi compares moral development to straightening a board or sharpening metal. These metaphors can be made to do a lot of work for both authors.
Both vegetative growth and the straightening of wood are slow processes, suggesting that moral development is also a slow process (unlike, say, some ways of understanding Buddhist enlightenment or Christian conversion and rebirth). Both metaphors imply permanent change and incremental progress, barring toxic or distortive factors in the environment, rather than a pattern of relapse and relearning. So also moral development, in the view of both philosophers.
Environment plays a strikingly different role in the two metaphors, however. A sprout grows into an oak tree (for example) more or less of its own accord, if the environmental conditions are sufficiently nutritive and non-hostile. Crooked, raw timber does not similarly straighten of its own accord: Both the impetus for change and the final shape are imposed from outside. Cultivation and growth work in harmony with the pre-existing inclinations of the sprout, while steaming and pressing work against the hard resistance of the board.
These metaphors thus suggest very different pictures of moral education. The cultivation metaphor suggests what we might call a self-discovery model of education, or an inward-out model of the sort often associated with “liberal” approaches to education in the contemporary West: Learners are encouraged to reflect for themselves, to discover and nurture the values they already have. They needn’t generally be told explicitly what is right and wrong. They are perfectly capable of seeing that for themselves, if they reflect carefully on their pre-existing inclinations and judgments. The inclination toward morality already exists within each of us (as the inclination to grow into an oak tree exists within the sprout), as long as the environment is sufficiently supportive. The environment needn’t be particularly directive: Just as the pattern for the oak tree is in some sense implicit in the sprout, so also a mature sense of right and wrong is in some sense implicit in the small child and will emerge in the normal course of growth, with proper nutrition and protection.
This self-discovery model of education captures, I think, much of the spirit of Mencius’ view, and is seen at work in passages such as 1A7, where Mencius invites the vicious King Xuan to reflect on why he felt an urge to save an ox from slaughter yet allows his innocent subjects to perish. This isn’t to say, of course, that Mencius would be happy with whatever King Xuan decided about such cases. He was a Confucian and adhered to traditional Confucian values. Despite his emphasis on self-discovery, he was not a “liberal” in the contemporary sense (one of the contemporary senses) of tolerating a diverse range of moral perspectives. Mencius takes himself to be encouraging not multiplicity but rather discovery of the one true (Confucian) moral structure already implicit in us and revealed by our impulses – by what pleases and revolts the heart – in obvious and nearby cases.
The straightening and sharpening metaphors of Xunzi suggest, in contrast, a more authoritarian approach to education, more outward-in, more in the style of contemporary Western “conservatives”. Children, and the morally underdeveloped in general, are not to be encouraged to think for themselves. They cannot be expected to know what is right any more than an introductory chemistry student should be expected to know, prior to being taught, the electronegativity of the elements. Free reflection, for the morally immature, is at best a waste of time, and at worst an opportunity for the rationalization of their immoral impulses. While Mencius repeatedly urges us to think (si 思 – think, reflect, ponder, concentrate), Xunzi declares “I once spent the whole day pondering, but it wasn’t as good as a moment’s worth of learning.” This isn’t to say that thinking isn’t of some value in the process of moral education, as thinking is of value in learning the facts of chemistry. Rather, it’s to say that the morally immature cannot discover for themselves right from wrong. Only someone of sagely genius could do that. For the morally immature, reflection can only be effective in the context of outward instruction, in following and understanding rules or a model given by one’s teacher. Morality must be imposed on us from outside – against our original impulses and inclinations and quite possibly contrary to our initial understanding. The process of moral education is not the pleasant matter, as it seems to be for Mencius, of discovering what truly pleases one’s heart. It is instead a matter of being forced against one’s will – and then later forcing oneself, by acts of will – to suppress and redirect one’s natural desires and inclinations.
In Chapters 1 and 2, Xunzi attempts to inspire the reader toward further moral development. He thus seems implicitly to hope that the reader has the desire to improve himself, or can be inspired to that desire; and this may seem to conflict with the picture of moral development just described, on which morality has to be forced from outside. The resolution of this difficulty, I believe, is to read Xunzi here as speaking principally to people who already have come some considerable distance in their development – to the point, perhaps, where their inclinations have some moral merit and they can see the value in further moral development. Adapting the metaphor, one might imagine that Xunzi’s wood, after having been straightened to a considerable extent, can itself contribute to the final part of the straightening process. For the young and the vicious, however, we may still interpret Xunzi as preferring rote conformity (“reciting the classics”) and compulsion; real understanding comes only near the end.
Thus, I read the disagreement between Mencius and Xunzi regarding human nature as principally a disagreement about the proper means of moral education. Moral education is no small thing to them: It was their profession (whether in teaching the young or in attempting to coax virtuous behavior from the vicious rulers of the “Warring States” period they lived in) and their principal concern.
5. Hobbes and Rousseau on Moral Education.
As it happens, Rousseau wrote extensively about moral education in Emile, a story of the idealized education of a boy from birth to adulthood. Hobbes also makes a number of remarks about moral education in the Leviathan. One might thus wonder whether we can recast their claims about “human nature” in developmental terms as I have suggested we do for Mencius and Xunzi. Does Rousseau, who thinks that “human nature is good”, support a vision of moral education as the cultivation of pre-existing, nascent inclinations toward morality? Does Hobbes support a vision of moral education as the external imposition of rules and moral knowledge upon minds without general inclinations in that direction? I’d like to suggest that the answer to both questions is yes.
It is clear in Emile that Rousseau means his claims about “human nature” to pertain not just to the fictional state of nature but also to the developing child. He writes, for example, “a young man raised in happy simplicity is drawn by the first movements of his nature toward the tender and affectionate passions”. Pride and vanity (which Rousseau thinks responsible for a large part of our conflict and unhappiness) does not “have its germ in children’s hearts, cannot be born in them of itself; it is we alone who put it there, and it can never take root except by our fault”. Also:
[T]he first voices of conscience arise out of the first movements of the heart…. [J]ustice and goodness are not merely abstract words – pure moral beings formed by the understanding – but are true affections of the soul enlightened by reason, and hence only an ordered development of our primitive affections.
Rousseau states that his goal in educating Emile is to “cultivate nature” (and not “deprave” it) and “to form the man of nature” but not “a savage [relegated] to the depths of the woods”. In the mouth of the Savoyard Vicar, he puts the view that our soul’s conscience follows the “order of nature” regardless of the laws of man and that in following it, we follow the “impulse of nature”, though it speaks with a quiet voice and “[t]he world and noise scare it”.
That the first impulses of the heart are good, but that they can easily be overridden by louder desires, that they require cultivation – in such matters Rousseau and Mencius agree. These are not remarks simply about how things stand only for the “savage” in the “state of nature” absent society; they concern moral development as it actually proceeds or fails in ordinary acculturated folk. To a considerable extent, the model of moral education offered in Emile resembles the model in Mencius: Specific moral rules are not imposed on Emile. He discovers for himself (in a nurturing, supportive environment – by no means the state of nature) the moral impulses we all share. He wants to act on them, and by acting on them they are nourished, so that a mature moral sense grows from within. Rousseau shares with Mencius, then, what I have called the “self-discovery” model of moral education.
Rousseau and Mencius diverge, however, in important ways. Where Mencius assumes a child always fully embedded in society, Rousseau takes great pains to shield Emile from most of society – so much so that in early adolescence Rousseau can say that
He knows no attachments other than those of habit. He loves his sister as he loves his watch, and his friend as his dog. He does not feel himself to be any sex, of any species. Man and woman are equally alien to him. He does not consider anything they do or say to be related to himself.
Even allowing for some overstatement, this passage is disturbing and seems to me hardly to reflect a process I would call “natural”. Indeed, throughout Emile Rousseau has his tutor take enormous pains to manipulate Emile’s environment. Rousseau appears to think the sprouts of human goodness are so fragile that the slightest chill could cripple them, in contrast to Mencius who sees them as always reasserting themselves (as in the parable of Ox Mountain, quoted above). Rousseau aims, for example, to assure that the infant and the child judge the failures to get what they want to be due only to the resistance of things, never of wills; and it seems to me to require great artifice to ensure this. Without this artifice, Rousseau seems to fear the child will become permanently spoiled. Amour-propre, or the kind of self-love that involves comparing oneself to others, Rousseau calls both “the most natural of all passions” and a “useful but dangerous instrument”. When amour-propre starts its inevitable bloom into vanity, Rousseau’s tutor contrives elaborate humiliations to cut it down. The tutor works to present Emile instead only with situations in which, when he compares himself with others, he finds compassion for others’ suffering and the impulse to improve himself. Similarly, when Emile awakens sexually, the tutor is much exercised, by what seems to me largely artificial imposition, to prevent disasters of vanity, licentiousness, and impulsiveness.
So is Rousseau, after all, not committed to a picture of human nature as good in the sense articulated in this essay? On the one hand, the impulses and forms of morality and found within Emile rather than imposed from without; but on the other, they seem to flourish only in a highly artificial environment. My criteria for the “natural” thus appear to diverge. Or, more properly speaking, although Rousseau clearly avoids the most salient defeater of the view that morality is natural (on my definition of “natural”) – that it is externally imposed – it may seem that the criterion that morality emerge in a broad range of normal environments is not met.
We can avoid this difficulty if we allow Rousseau to suggest that the artifices of his tutor are principally necessary to counteract the highly unnatural toxicity of French civilization, to restore something closer to a normal, as opposed to a distortive and perverted, human environment. This seems a plausible move, given Rousseau’s well-documented disdain for French civilization. Metaphorically speaking, Rousseau’s tutor may be providing a post to a growing vine in an environment stripped of trees. In any case, if we construe the key question as whether moral education should proceed by the cultivation of pre-existing, nascent impulses toward morality, Rousseau clearly thinks it does. If, perhaps, Rousseau’s cultivation is more like the work of a French (or even bonsai) gardener, who constantly prunes and shapes, than like the work of a Chinese rice or barley farmer, there is still much they share in common.
Hobbes, in contrast, seems to envision moral education principally as the imparting of official doctrine:
And (to descend to particulars) the People are to be taught, First, that they ought not to be in love with any forme for Government they see in their neighbor Nations, more than with their own, nor (whatsoever present prosperity they behold in Nations that are otherwise governed than they,) to desire change….
Secondly, they are to be taught that they ought not to be led with admiration of the vertue of any of their fellow Subjects … so as to deferre to them any obedience, or honour, appropriate to the Soveraign onely….
Thirdly, … they ought to be informed, how great a fault it is, to speak evill of the Soveraign Representative … or to argue and dispute his Power….
This education proceeds, not by providing an environment supportive of reflection and self-discovery but rather (for most adults) from the pulpit:
Fourthly, seeing people cannot be taught this, nor when ’tis taught, remember it, nor after one generation past, so much as know in whom the Soveraign Power is placed, without setting a part from their ordinary labour, some certain times, in which they may attend to those appointed to instruct them; It is necessary that some such times be determined, wherein they may assemble together, and (after prayers and praises given to God, the Soveraign of Soveraigns) hear those their Duties told them, and the Positive Lawes, such as generally concern them all, read and expounded…. To this end had the Jewes every seventh day, a Sabbath ….
Indeed, private reflection is condemned:
As for the Means, and Conduits, by which the people may receive this Instruction, wee are to search, by what means so many Opinions, contrary to the peace of Man-kind, upon weak and false Principles, have neverthelesse been so deeply rooted in them. I mean those, which I have in the precedent Chapter specified: as That men shall Judge of what is lawfull and unlawfull, not by the Law it selfe, but by their own Consciences; that is to say, by their own private Judgements….
Hobbes’ metaphor for education in the Leviathan is not, of course, cultivation. He compares education, rather, to writing on paper:
the Common-peoples minds, unlesse they be tainted with dependance on the Potent, or scribbled over with the opinions of their Doctors, are like clean paper, fit to receive whatsoever by Publique Authority shall be imprinted in them.
Paper does not resist writing in the same way a board resists straightening, but neither do the words written on paper gain any particular support from the paper’s antecedent inclinations. Hobbes does perhaps suggest that that which conforms to the principles of Reason is more easily inscribed, but the passage about the pulpit, quoted above – and indeed his authoritarianism generally – suggests that he is not especially sanguine about the common people durably retaining, or even entirely comprehending, what is taught. Perhaps a better metaphor for Hobbes than writing on paper would be writing in sand?
Hobbes’ remarks about the education of children are less extensive, but I interpret them as similar in spirit (especially given Hobbes’ comparison of paternal and maternal dominion to the sovereign’s dominion over the state), emphasizing obedience and the imposition of doctrine – and on the whole being very different from the sort of education envisioned by Rousseau.
Hobbes, like Rousseau, writes repeatedly about the “nature” of humankind outside the context of the “state of nature” thought experiment that has received so much emphasis in interpretations of their work. Hobbes in fact, wrote an entire essay titled Human Nature. This work discusses our “natural faculties”, including both faculties of the body and faculties of the mind. It’s clear that Hobbes takes himself to be treating the normal, mature, adult human being as he (or she) actually develops in a normal societal environment – and not (or not just) as he would develop in a state of anarchy. Of particular interest is Hobbes’ treatment of the passions and what delights the mind in Chapters VII-IX. One sees here more of Xunzi than of Mencius or Rousseau: Hobbes emphasizes self-interest and the sort of desires that would lead to strife without some sort of external or internal suppression or regulation. Where he discusses passions others might see as unselfish, he generally gives them an egoistic interpretation: “Honour” consists not in moral virtue but in signs of power; repentance is not characterized as following from a sense of moral failure but only as “the passion that proceedeth from opinion or knowledge that the action they have done is out of the way to the end they would attain”; charity appears to arise not so much from innate compassion as from the fact that “There can be no greater argument to a man of his own power, than to find himself able, not only to accomplish his own desires, but also to assist some other men in theirs”. Pity appears to involve a genuine desire for (some) others to do well, independently of one’s own condition, but it is counterbalanced by “Laughter” at other men’s “infirmities and absurdity” and is not uncommonly outweighed by the pleasure that arises, in watching others suffer, from the feeling of one’s own security and superior position, as when we watch a battle or a shipwreck. Even Xunzi’s picture of our natural desires might not be quite so relentlessly egoistic.
So when Hobbes and Rousseau speak of human nature as good or as leading to violent strife, I believe we have license to interpret them as speaking not just of how we would behave in the “state of nature”. They are much closer to Mencius and Xunzi than a superficial criticism of a few of their most famous passages suggests. To say that human nature is good, for Mencius or Rousseau, is to say that our first and most basic impulses, if we avoid the corruption of distortive environments, point us in the direction of morality, and that consequently moral education consists in careful attention to and cultivation of those impulses. To take the opposing position, for Xunzi or Hobbes, is to say that the most basic and dominant body of impulses in normal, mature individuals would impel us to conflict and disorder if they were not forcibly restrained, and thus that moral development requires the persistent imposition of rules and doctrines that have little basis in the untutored impulses of ordinary men. That, at least, is my suggestion.
What, then, to
make of Hobbes’ and Rousseau’s thought experiments about the state of
nature? I do think they have an interest
besides that of revealing the consequences of living in a civil society. For Rousseau, the value is also partly in
undermining the idea that contemporary society, especially urban
6. Moral Education and Authoritarian Government.
Given the association between self-discovery models of moral education and “liberalism” in the contemporary West, one might suppose that proponents of the idea that human nature is good (in the sense just explained) would be drawn toward comparatively democratic forms of government in which expressions of popular opinion play a large role; and that, conversely, those who are drawn toward authoritarian models of moral education would inclined more toward authoritarianism in government. Plainly, Rousseau in The Social Contract advocates a less authoritarian style of government than Hobbes in the Leviathan. Likewise, perhaps, the political philosophy in the Mencius has a slightly less authoritarian feel than that in the Xunzi. Both democracy and education by self-discovery turn on the idea that ordinary people are often best encouraged to form their own judgments (perhaps aided and supported in various ways), without the external imposition of doctrines by an authority.
But this is oversimple. Mencius, of course, is no democrat. Like the other major early Confucians, he endorses an enlightened monarchy. Perhaps the better predictor of authoritarian politics is the expense of moral education. If proper moral education requires extensive resources unavailable to the general public, then one might find an aristocratic ruling structure attractive, at least as a utopian possibility. In the early Confucian tradition, moral education is not cheap. Although anyone (or at least any man) can enter it, it requires long devotion to learning ancient ritual and studying classic texts and is incompatible with the life of labor than must be most people’s lot. If the education is successful, those who have run through it will have better moral character and judgment than the masses of people. It is, then, their judgment, and not popular opinion, that should guide the state (though the early Confucians generally thought the masses possessed enough common sense to be attracted to virtuous rulers and to despise the wicked). Plato’s Republic fits a similar model: True knowledge of the good is no common thing (recall the analogy of the cave in Book VII). Those occupied daily with manual labor cannot have the moral education of the philosopher-king, and ideally the guidance of the state should not derive from their judgment.
Rousseau, in contrast, though he imagined the education of Emile to require a private full-time tutor, appeared to believe that, in general, moral development is better attained in a life of rural labor than among the privilege of the elite. Likewise, I suspect, most contemporary dwellers in Western-style democracies see proper moral education as broadly attainable, not requiring the cessation of daily labor – maybe even enhanced by labor. Perhaps also pessimists about moral education, who think a ruling elite will necessarily be corrupt, will be drawn toward more democratic forms of government as best suited to keep our vice in check.
That Hobbes did not go in this last direction, given what some see as a pessimistic strain in his work, has often been held against him. However, he does emphasize that both the gentry and the educators of the masses are to receive their education in proper doctrine from the universities; and perhaps this gives us room to interpret him as thinking that the best moral education requires resources available only to the elite. Maybe he was less than completely pessimistic about the positive moral effect of these more expensive institutions, if properly reformed and subject to the sovereign.
Surely there are counterexamples to this claim about the relationship between authoritarian government and the cost of proper moral education. Supposing the relationship largely holds, however: This is one way in which a view of moral education can drive a political philosophy. Indeed, a take on moral education can motivate a view not only of political and familial authority and subordination, but also of the nature of self-constraint and willpower; of the proper role of ritual, custom, and law; of the origin and character of the emotions; of the ideal structure of society; of the role and value of reason and reflection; of the nature of moral character. Equally, the stances one takes on these issues can motivate thoughts on the proper structure of moral education. Rousseau, more than our other authors, makes explicit the extent to which these issues are entangled: Emile beautifully displays the interrelations, with a program of moral education standing at the center. Indeed, near the end of his life, surveying his work, Rousseau wrote that Emile was the key to understanding all the rest.
7. Who Is Right?
Despite Rousseau’s notorious claim, near the beginning of his Discourse on Inequality to be “setting aside all the facts, for they have no bearing on this question”, claims about human nature are clearly at least partly empirical: They involve assertions about the way human beings are, in fact. This needn’t entail that there is any single, straightforward, empirical test that will definitively reveal whether morality is natural to us or imposed from outside. (Little, indeed, of broad general interest in human psychology can be revealed by a single test.) But the question is empirically explorable. Empirical facts are relevant to it, and various of them may comport more harmoniously with one view or another, fit one picture nicely and require explaining away on another. These empirical facts must be tempered with a normative understanding, not only of what is “good”, but also (as per §2) of what is normal and abnormal, supportive or distortive – but so also must the empirical explorations of all the human and biological sciences.
We can thus attempt some judgment about who is nearest the truth about human nature. I’ve argued that the question is, at core, about the proper course of moral education; so this is an appealing place to begin. If reasonable people can in general more or less agree about standards of moral goodness (as I think they can, pace strong descriptive relativism), we can then ask an empirical question: What sort of moral education best engenders moral maturity – one that imposes morality on children from the outside or one that encourages children to reflect for themselves?
Now in practice, this will be a very difficult assessment, since even without the complications of moral disagreement, the assessment of moral maturity is no easy thing. We can perhaps look at some extreme cases (people convicted of hideous offenses, or moral exemplars of the sort profiled by Colby and Damon), but it seems a mistake to focus only on exceptional people. Long-term, controlled studies are impossible. Short term laboratory tests may be misleading. Nonetheless, there does appear to be a general consensus among the most eminent scholars of moral development that reflection is salutary and its suppression is harmful – that children should be encouraged to think for themselves about right and wrong, in their own terms. If true, this fits nicely with Rousseau and especially Mencius. However, the advocate of a darker view of human nature may legitimately wonder whether structured reflection, with adults nearby who the child knows will approve of one answer and disapprove of another, isn’t really just a form of imposition, more effective for its being subtle and parading as the child’s own independent judgment.
We can look also for other signs of natural goodness or its lack. So, for example, do we see nascent moral impulses – and a comparative lack of nascent immorality – in very young children and in non-human primates, who presumably are less influenced by the imposition of an external code? Do the perpetrators of terrible evil (such as the Holocaust), when they reflect on their deeds, find themselves morally revolted, regardless of their prior doctrines, or are malignant values relatively stable to the reflection of ordinary non-philosophers? When people are encouraged to reflect on their emotional reactions to their own and others’ actions, are they thereafter (at least immediately thereafter) more or less likely to commit misdeeds?
Let me take my stand. I think, overall, the evidence favors a roughly Mencian view, in which ordinary reflection in a supportive but non-directive environment is the best spur to moral development. Besides the researchers on moral development cited above who seem to favor views roughly of this sort, let me mention the work on early childhood sympathy by Zahn-Waxler and others; de Waal’s work on the origins of morality in non-human primates; Arendt’s suggestion that evil tends to flow from a failure to think in her study of Eichmann; and work on juvenile delinquency that suggests reduced recidivism when offenders are encouraged to reflect.
The evidence is by no means unequivocal, and an absolutely pure and uniform goodness in human nature is probably too much to hope for. Young boys seem naturally to delight in the suffering of insects; and a certain kind of pleasure in the small misfortunes of others seems nearly universal, and probably “normal” given inclusive standards of normality. (I think here especially of our reactions to the kinds of mishaps portrayed in shows like America’s Funniest Home Videos.) Ingroups’ aggression against outgroups (from small cliques and sports teams to races and nations) seems too universal and too heartily approved to be anything but natural. (Think of all the “great men” of history, after whom our children are often named, whose principal achievement was in aggressive warfare.) But perhaps counterbalancing this is a natural intolerance of aggression within ingroups. Nazis are often surprisingly unrepentant, even in the face of the vast social disapproval of their actions; but maybe this can be explained by psychological self-defense mechanisms. Ethics professors, despite what seems like ample opportunity for moral reflection, in my experience behave no better than other members of their social class – Rousseau himself famously abandoned his children – but perhaps their reflection is too intellectual, too clever, and too far removed from local particulars to foster their own moral development.
This last concern, actually, is the one that worries me most. Unless one is an absolute pessimist about moral reflection or the value of morality, it seems that one should hope and expect that those who are praised for their talent in thinking through moral issues, who do so in their daily work and who, presumably, extend this at least somewhat into their daily lives, should achieve some moral improvement thereby. Else what is moral reflection for?
 Thomas Hobbes, Leviathan, ed. R. Tuck (Cambridge: Cambridge, 1996 [originally published 1651]), p. 60-62/86-89.
 Jean-Jacques Rousseau, Discourse on the Origin of Inequality, ed. F. Philip, trans. P. Coleman (Oxford: Oxford, 1994 [originally published 1755]), p. 83.
 Ibid., p. 35.
 “Letter to Beaumont”, in Letter to Beaumont, Letters Written from the Mountain, and Related Writings: The Collected Writings of Rousseau, Vol. 9, ed. C. Kelly and E. Grace, trans. C. Kelly and J.R. Bush (Hanover, N.H.: University Press of New England, 2001), p. 935-936/28.
 This includes, of course, 17th- and 18th-century biologists. For example, the eminent naturalist René Antoine Ferchault de Réaumur, in his Natural History of Ants (composed 1742-1743) writes that “all [ants] are born for social life”: The Natural History of Ants, trans. W.M. Wheeler (New York: A.A. Knopf, 1926), p. 47, 135. Nowhere in his extensive discussion of ant behavior does he find it necessary to consider an ant’s life separate from the colony, as indicative of the ant’s true “nature”.
 The present distinction between “natural” and non-natural traits and may seem to conflict with a strong version of Developmental Systems Theory (S. Oyama, The Ontogeny of Information [Cambridge: Cambridge, 1985]) that rejects the “nature” / “nurture” distinction. I largely agree with the standard complaints about that distinction, and I hope that the present understanding of the natural is compatible with at least a moderate version of DST.
 Mengzi 6A2, trans. B.W. Van Norden,
 Mengzi 6A8, ibid., p. 145.
 Mengzi 6A7, ibid., p. 144-145.
 Mengzi 6A10, ibid., p. 146.
 Mengzi 2A6, ibid., p. 125.
 Mengzi 3A5, ibid., p. 130.
 By “righteousness, benevolence, and propriety” here I mean to refer to the canonical Mencian virtues of yi (義), ren (仁), and li (禮), to which 6A10, 2A6, and 3A5 seem respectively to pertain. Scholars of Chinese philosophy will note the conspicuous absence of zhi (智 – wisdom) from this list. I don’t know if there’s a passage in the Mencius that plays the same role vis-à-vis zhi that the cited passages play for yi, ren, and li.
 On “extension” see esp. 1A7, 2A6, 7A15. On causes of moral failure, see esp. 1A7, 6A7, 6A14, 6A15. Van Norden translates all these passages but 6A14 in Readings in Classical Chinese Philosophy. For 6A14 see Mencius, trans. D.C. Lau (London: Penguin, 1970).
 See Hsün Tzu: Basic Writings, trans. B. Watson (New York: Columbia, 1963), p. 159.
 E.g., ibid., p. 159-160.
 Xunzi emphasizes this especially in Ch. 23.
 See especially Ch. 1-2 of the Xunzi.
 Literally the heart, supposed to be the organ of cognition and at least some emotions. See D.B. Wong, “Is There a Distinction Between Reason and Emotion in Mencius”, Philosophy East & West 41 (1991), p. 31-44.
 See D.C. Lau’s introduction to the Mencius, p. 19-22, for a partial endorsement of this view.
 And thus I’m suggesting that we can read Mencius and Xunzi as closer in their sense of what it means to say that a trait is xing (性 – natural) than many commentators now suggest – such as A.C. Graham, Disputers of the Tao (La Salle, IL: Open Court, 1989), P.R. Goldin, Rituals of the Way (Chicago: Open Court, 1999).
 E.g., 2A6, 6A6, 6A8, 6A9, 6A14.
 In emphasizing the importance of
these metaphors, I follow P.J. Ivanhoe, esp. Confucian Moral Self
Cultivation, 2nd ed. (
 Goldin suggests that, contra the present interpretation, a particular passage in Ch. 4 (sec. 12 in Knoblock, trans., p. 87) suggests that we only born ignorant of morality and are naturally attracted to it even on first exposure (“Virtue, Nature, and Moral Agency in the Xunzi”, Journal of the American Academy of Religion 69 , p. 497-498). However, I believe this passage can be accommodated within the present view. What the morally undeveloped might find attractive on first exposure is not morality itself but rather a society that operates according to the proper moral principles (and is thus harmonious). (Thanks to Eric Hutton for discussion of this point.)
 Xunzi also repeatedly stresses, throughout his corpus, the material advantages (such as political power and personal security) that come with the life of virtue. Such remarks may be intended to appeal particularly to the morally undeveloped.
 A number of the essays in T.C. Kline III and P.J. Ivanhoe, eds., Virtue, Nature, and Moral Agency in the Xunzi (Indianapolis: Hackett, 2000) shine excellent light on some of the themes in this section of this paper, esp. Kupperman, Van Norden, Wong, Hutton, and Ivanhoe.
 Emile, trans. A. Bloom (Basic Books: 1979 [originally published 1762]), p. 220.
 Emile, p. 215.
 Emile, p. 235, emphasis added.
 Emile, p. 254-255.
 Emile, p. 267, 286, and 291, respectively.
 Emile, p. 219.
 Emile, p. 66.
 Emile, p. 208 and 244.
 Emile, p. 173-175, 244.
 Leviathan, p. 177-178/233-234.
 Leviathan, p. 178/234-235.
 Leviathan, p. 179/236.
 Leviathan, p. 176/233. See also Human Nature, in The Elements of Natural Law and Politic, ed., J.C.A. Gaskin (Oxford: Oxford, 1994 [originally published 1650]), I.X.8, p. 62-63.
 Leviathan, p. 177/233.
 Leviathan, p. 101-107/138-145 and 178/235. S.A. Lloyd nicely describes Hobbes’ picture of education, and draws a helpful parallel between state-sponsored education and parental instruction, in “Coercion, Ideology, and Education in Hobbes’s Leviathan”, in Reclaiming the History of Ethics, ed. A. Reath, B. Herman, and C.M. Korsgaard (Cambridge: Cambridge, 1997).
 Human Nature, I.VIII.5-6.
 Human Nature, I.IX.7.
 Human Nature, I.IX.17.
 Human Nature, I.IX.10.
 Human Nature, I.IX.13.
 Human Nature, I.IX.19.
 E.g., Knoblock trans.,
 This interpretation conflicts sharply not only with simple “state of nature” interpretations of what Rousseau means by “natural” but also with some of the more textually nuanced or biologically sensitive interpretations in the literature, such as J.C. Hall, Rousseau (London: Macmillan, 1973); J.B. Noone, Jr., Rousseau’s Social Contract (Athens, GA: University of Georgia, 1980); and R. Kukla, “Making and masking human nature: Rousseau’s aesthetics of education”, Journal of the British Society for Phenomenology 29 (1998), p. 228-251. A.M. Melzer’s interpretation, in The Natural Goodness of Man (Chicago: University of Chicago, 1990) nicely incorporates the idea that our first and most basic impulses are good but in my view underplays the developmental aspect.
 It is perhaps worth noting that
Locke, who also famously employs the “state of nature” thought experiment (in
his Second Treatise of Government, in Two Treatises of Government,
ed. I Shapiro [
 See also Aristotle’s Politics, Book III, Ch. 4.
 Leviathan, p. 180/237 and 395/491.
 Hobbes does write that it is easy for common people to learn the basic “Principles of Reason” (Leviathan, p. 176-177/233); but that needn’t conflict with what I say here if Hobbes is willing to draw a distinction between the kind of education necessary to be a good subject in an authoritarian state and that necessary to make governmental decisions.
 Two recent works that argue persuasively for the centrality of views of moral education to one’s overall moral and political outlook are G. Lakoff, Moral Politics, 2nd ed. (Chicago: University of Chicago, 2002) and T. Sowell, A Conflict of Visions (New York: Basic Books, 2002). (Thanks to P.J. Ivanhoe for the Sowell reference.)
 Rousseau Judge of Jean-Jacques, ed. R.D. Masters and C. Kelly, trans. J.R. Bush, C. Kelly, and R.D. Masters (Hanover, NH: University Press of New England, 1990 [originally published 1776]), p. 932-933/211. T. O’Hagan puts Emile near the center of his understanding of Rousseau, and his reading of what Rousseau means by “nature” has an explicitly developmental aspect: e.g., Rousseau (London: Routledge, 1999), p. 63.
 p. 24.
 A. Colby and W. Damon, Some Do Care (New York: Free Press, 1992).
 E.g., J. Piaget, The Moral Judgment of the Child, trans. M. Gabain (New York: Free Press, 1965 [originally published 1932]); L. Kohlberg, Essays on Moral Development, vol. 1 (San Francisco: Harper & Row, 1981); W. Damon, The Moral Child (New York: Free Press, 1988); E. Staub, The Roots of Evil (Cambridge: Cambridge, 1989).
 E.g., C. Zahn-Waxler and M.
Radke-Yarrow, “The origins of empathic concern”, Motivation and Emotion
14 (1990), p. 107-130; C. Zahn-Waxler, M. Radke-Yarrow, E. Wagner, and M.
Chapman, “Development of concern for others”, Child Development 24
(1992), 126-136; M.L. Hoffman, Empathy and Moral Development (
 Esp. F. de Waal, Good Natured (Cambridge, MA: Harvard, 1996); on social animals in general, see also C. Darwin, Descent of Man, ed. J.T. Bonner and R.M. May (Princeton: Princeton, 1981 [originally published 1871]).
 H. Arendt, Eichmann in
 E.g., R. Schwitzgebel (my father!), Streetcorner Research (Cambridge, MA: Harvard, 1964), p. 24-44; S.E. Samenow, Inside the Criminal Mind (New York: Times Books, 1984).
 For helpful conversation and comments, thanks to audiences at U.C. Riverside (in developmental psychology), Cal State San Marcos (in political theory), Cal State Fullerton, Cal Poly Pomona, and the Eastern APA, and to Nick Bunnin, Robert Harada, Eric Hutton, P.J. Ivanhoe, Chris Laursen, Adam Morton, Timothy O’Hagan, Kwong-loi Shun, and Bryan Van Norden. My debt to Kwong-loi Shun’s and especially P.J. Ivanhoe’s view of Mencius is large, as will perhaps be evident to those familiar with their writings.