← All essays
Xunzi2026-09-02By Anderson#xunzi#hobbes#human-nature#philosophy#political-philosophy#leviathan

Xunzi vs Hobbes: If Human Nature Is Bad, Who Fixes It?

使欲必不窮於物，物必不屈於欲，兩者相持而長。 — Desires must never be exhausted by goods, and goods never depleted by desires; the two support each other and grow together. (《荀子·禮論》, my translation)

Short answer: they agree on the disease and split on whether there's a cure.

Both men open the same way. People want things, the things run short, and fighting follows. Then Hobbes builds a permanent container, because he doesn't think people change. Xunzi builds a training program, because he does. The man who says nature is bad is the one who thinks you can fix it.

Key Takeaways

  • Hobbes says the passions aren't sins at all: "The Desires, and other Passions of man, are in themselves no Sin" (Leviathan, ch. 13).
  • Xunzi says ritual exists 「以養人之欲，給人之求」: to feed desire, not shrink it (《荀子·禮論》).
  • So Hobbes contains a permanent problem. Xunzi treats a curable one.

What's the real difference between Xunzi and Hobbes?

One of them thinks the repair is temporary. The other thinks it's permanent.

Hobbes builds a machine that has to keep running to accomplish anything at all. Take away the enforcement and you are back where you started, which he states plainly enough: "when the Soveraign Power ceaseth, Crime also ceaseth" (ch. 27). Nothing in a person has changed. The pressure changed.

Xunzi builds something you eventually carry yourself. His word for the fix is 禮, usually rendered ritual, and he compares it to a carpenter's marking line and a set of scales. Instruments like those don't restrain you so much as calibrate you, and once the calibration has taken you can't be fooled about what counts as straight.

That's the whole split, and it's why the popular pairing of these two gets awkward fast. Xunzi (荀子, also written Xun Kuang) sounds harsher on the page. His position is the hopeful one.

Do Xunzi and Hobbes actually agree about human nature?

On the mechanism, almost word for word. On the verdict, not at all.

Here's Hobbes on how conflict starts: "if any two men desire the same thing, which neverthelesse they cannot both enjoy, they become enemies" (Leviathan, ch. 13). Scarcity plus appetite equals enmity.

Here's Xunzi opening 《禮論》: 「人生而有欲，欲而不得，則不能無求。求而無度量分界，則不能不爭；爭則亂，亂則窮。」 People are born with desires. Wanting without getting, they can't help seeking. Seeking with no measure or boundary, they can't help fighting. Fighting brings chaos, and chaos brings ruin.

Same chain, same order, roughly nineteen centuries apart, and neither man had any way of reading the other. What makes the overlap worth studying is that it isn't borrowed, so where they diverge you're watching two independent arguments reach a fork rather than one tradition editing another. The fork arrives immediately, and it isn't small.

Does Hobbes say human nature is evil?

No, and he goes out of his way to say so. This is the single most misquoted thing about him. [UNIQUE INSIGHT]

After the famous passage about life being "solitary, poore, nasty, brutish, and short," Hobbes anticipates the reader who thinks he's slandering mankind. He points out that you lock your doors and chests despite the existence of laws and armed officers, then asks whether you aren't accusing your neighbours exactly as he is. Then comes the line nobody quotes: "But neither of us accuse mans nature in it. The Desires, and other Passions of man, are in themselves no Sin."

He goes further still, extending the same verdict to conduct: actions arising from those passions aren't sins either, "till they know a Law that forbids them; which till Lawes be made they cannot know." Before law there is no wrong: "Where there is no common Power, there is no Law: where no Law, no Injustice."

So Hobbes issues no moral verdict on human nature whatsoever. He describes a physics problem. The trouble isn't wickedness, it's arithmetic.

Why does Xunzi call nature bad if he thinks it's fixable?

Because for him the verdict and the prognosis are separate questions, and modern readers keep fusing them.

Xunzi's most famous chapter title is 性惡, nature is bad. That really is a moral verdict, and it's the reason he's cast as the pessimist of his tradition. But look where 《禮論》 lands after pages of argument: 「故學者，固學為聖人也，非特學無方之民也。」 The student's whole purpose is to learn to become a sage, not merely to learn not to be an unmoored person.

Sagehood is the target he sets for an ordinary student, starting from a nature he had just finished calling bad. Hobbes offers nothing remotely like that, and he would think that offering it was a category error rather than a kindness.

Notice what this does to the standard pairing. The man who moralizes human nature turns out to be the optimist. The man who refuses to moralize it turns out to be the one who thinks you're stuck. [UNIQUE INSIGHT]

Xunzi vs Hobbes at a glance

QuestionXunziHobbes
Where does conflict come from?Desire without measure or boundary (《禮論》)Two men wanting one thing (ch. 13)
Is nature morally bad?Yes, 性惡No: passions "are in themselves no Sin" (ch. 13)
What's the remedy?禮, which feeds desire (《禮論》)Covenant, which surrenders claims (ch. 14)
Who enforces it?Eventually you doThe sword, permanently (ch. 17)
Can a person be finished?Yes, 學為聖人 (《禮論》)No such endpoint (ch. 11)
What if enforcement stops?The standard holds"Crime also ceaseth" (ch. 27)

What does Hobbes ask you to give up?

Your claim on everything. That's the price, and he's honest that it is a price.

The second law of nature asks "That a man be willing, when others are so too, as farre-forth, as for Peace, and defence of himselfe he shall think it necessary, to lay down this right to all things; and be contented with so much liberty against other men, as he would allow other men against himselfe" (ch. 14).

Read the verb: lay down, subtract. You had unlimited right, you now hold less than that, and the shortfall is what buys you peace.

What motivates the trade isn't virtue. It's fear plus comfort: "The Passions that encline men to Peace, are Feare of Death; Desire of such things as are necessary to commodious living; and a Hope by their Industry to obtain them" (ch. 13). Appetite doesn't get educated in this story. It gets outbid by a bigger appetite.

What does Xunzi ask you to grow?

Both sides of the equation at once, which is the opposite instruction.

The sentence that decides it comes two lines into 《禮論》: 「先王惡其亂也，故制禮義以分之，以養人之欲，給人之求。」 The former kings hated that chaos, so they instituted ritual and rightness to apportion things, in order to feed people's desires and supply what they seek.

Feed and supply, not curb and not renounce. He then states the target condition in a single sentence: 「使欲必不窮於物，物必不屈於欲，兩者相持而長，是禮之所起也。」 Desires must never be exhausted by goods, and goods never depleted by desires; the two support each other and grow together. That's where ritual comes from.

And he says it flatly a paragraph later: 「故禮者養也。」 Ritual is nourishment. He spends the next passage listing what it nourishes, down to food, scent, sight and sound. Hobbes shrinks the claim to fit the supply. Xunzi grows them into each other. [UNIQUE INSIGHT]

Does Xunzi want you to have fewer desires?

He explicitly says that's the wrong lever, and he says it in a chapter almost nobody reads alongside this one.

In 《正名》 he takes aim at exactly the reading he usually gets: 「凡語治而待去欲者，無以道欲而困於有欲者也。凡語治而待寡欲者，無以節欲而困於多欲者也。」 Anyone who talks about order while waiting on the removal of desires has no way to guide desire and is stumped by having any. Anyone who waits on fewer desires has no way to moderate desire and is stumped by having many.

Then the payoff: 「故治亂在於心之所可，亡於情之所欲。」 Order and disorder lie in what the mind approves, not in what the feelings want.

He makes the consequence explicit both directions. 「心之所可中理，則欲雖多，奚傷於治？」 If what the mind approves hits the mark, what harm do many desires do? And if approval misses, few desires won't stop the mess. Volume was never the variable.

Do they disagree about whether desire has a ceiling?

No. They agree there isn't one, then draw opposite conclusions from it.

Hobbes is emphatic that wanting never resolves: "there is no such Finis Ultimus, (utmost ayme,) nor Summum Bonum, (greatest good,) as is spoken of in the Books of the old Morall Philosophers." Happiness is "a continuall progresse of the desire, from one object to another" (ch. 11).

Xunzi grants the same thing without flinching. 「雖為天子，欲不可盡。」 Even a Son of Heaven can't exhaust his desires. And at the other end: 「故雖為守門，欲不可去，性之具也。」 Even a gatekeeper can't be rid of desire, because it comes with being alive.

Two men reach one observation and take it to two conclusions. Hobbes decides that a thing with no internal stopping point has to be stopped from outside. Xunzi decides that since it can't be stopped at all, you'd better train the faculty that approves. Same premise. Different door.

So who does the fixing?

For Hobbes, someone else, forever. For Xunzi, eventually you.

Hobbes never pretends otherwise: "And Covenants, without the Sword, are but Words, and of no strength to secure a man at all" (ch. 17). Agreement is inert. Force is the active ingredient, and it has to stay switched on.

Xunzi's ritual is external too, at first. But watch the images he picks in 《禮論》: 「故繩墨誠陳矣，則不可欺以曲直；衡誠縣矣，則不可欺以輕重；規矩誠設矣，則不可欺以方圓；君子審於禮，則不可欺以詐偽。」 Once the marking line is truly laid out, you can't be deceived about straight and crooked. Once the scale hangs true, not about heavy and light. Once compass and square are set, not about round and square. And once a gentleman is clear about ritual, he can't be deceived by fraud.

Every one of those is a measuring instrument rather than a weapon, and the difference in how they operate is the whole argument. A sword works on you. A scale works for you, and it goes on working when nobody is watching, including on the occasions when the person doing the deceiving is you. [UNIQUE INSIGHT]

What happens when the enforcement stops?

This is the test that separates them, and both men answer it in their own words.

Hobbes answers in chapter 27: "when the Soveraign Power ceaseth, Crime also ceaseth." He isn't being cynical. His point is that without a power to enforce, there's no law to break, so a person who behaved for twenty years under the sword hasn't accumulated anything. Remove the structure and you get the state of nature back, undiminished.

Xunzi's answer is the reverse, and it follows from the instrument analogy. A person who has genuinely internalized a standard doesn't lose it when the supervisor leaves. He calls ritual 「人道之極」, the furthest reach of the human way, and treats getting there as an achievement that stays got.

If you want a practical version of that difference, we've written one for people who keep breaking promises to themselves, which runs on different chapters and asks the personal version of this question.

Do Xunzi and Hobbes agree about anything surprising?

Yes, and it's the strangest overlap in the pair: both think language is a political problem.

Hobbes is a thoroughgoing conventionalist about words. "For True and False are attributes of Speech, not of things. And where Speech in not, there is neither Truth nor Falshood" (ch. 4). Names are set by agreement, so a disordered vocabulary is a disordered commonwealth.

Xunzi says the same in 《正名》: 「名無固宜，約之以命，約定俗成謂之宜。」 Names have no inherent fitness. Convention assigns them, and what convention settles becomes fitting.

Then both make it the sovereign's business. Xunzi: 「故析辭擅作名，以亂正名，使民疑惑，人多辨訟，則謂之大姦。」 Splitting phrases and coining names to disorder correct naming, so the people grow confused and disputes multiply, is called great criminality. Two conventionalists, two millennia apart, both deciding that whoever controls definitions controls whether the peace holds.

Which argument holds up better?

Each is strongest exactly where the other is weakest, so it depends on what you're trying to do.

Hobbes's advantage is that his system doesn't require anyone to improve. That's a serious virtue in political design. His weakness is that it never finishes, and he never claims it will. You get a durable floor and no ceiling.

Xunzi's advantage is that his fix compounds. His weakness is the one Hobbes would name immediately: the whole thing depends on transformation actually happening, and Xunzi offers no fallback for the person it doesn't take with. He also picked a confusing label. Calling a curable condition 性惡 invited two thousand years of readers to hear a life sentence, and his own student Han Feizi took the diagnosis, dropped the cure, and built a state that assumes nobody ever improves.

That last detail is worth sitting with. The strongest evidence against Xunzi's optimism came from the man he taught.

How is this different from your Mencius and Rousseau comparison?

Different question entirely, and the two pieces are halves of one argument.

Mencius vs Rousseau asks whether the raw material is any good. That's a dispute about the starting point, and those two split on the diagnosis while sounding like allies. Xunzi and Hobbes grant that the starting point is a problem and split on the treatment.

Lay the four out and you get a grid. Rousseau: good material, subtract the corruption. Mencius: good material, add cultivation. Hobbes: bad material, subtract the claims. Xunzi: bad material, add cultivation. The Schwitzgebel study that covers all four reads the pairs the same way.

Also worth separating from our other comparisons: Han Feizi vs Machiavelli asks whether power sits in the person or the post, and Sun Tzu vs Clausewitz asks whether you can know enough before acting. Neither one argues about what people are made of.

How do you use this on a habit you keep failing?

Run the diagnosis before you pick a tool, because the two thinkers fix different failures.

  1. Ask which failure you actually have. If you've never written down what "enough" looks like, you don't have an enforcement problem. You have Xunzi's problem: no 度量分界, no measure or boundary, so there's nothing for approval to hit.
  2. If you genuinely need enforcement, build it outside yourself and expect it to expire. Hobbes's own terms. Deposits, deadlines, other people. Just don't confuse compliance with change, because he tells you not to.
  3. Test which one you've got. Remove the enforcement for a week. If the behavior holds, a standard has been internalized. If it collapses on day two, you were renting compliance, and renting is fine as long as you know that's what the payment is for.

Most people skip step one, install a Hobbesian sword against a Xunzian problem, and conclude they lack willpower. Worth asking Xunzi about the specific case rather than guessing.

Frequently asked questions

Was Xunzi the Chinese Hobbes? Only if you stop at the diagnosis. Both derive conflict from scarce goods and unlimited wanting. But Hobbes says the passions are no sin and offers no path out of the condition, while Xunzi calls nature bad and sends an ordinary student toward sagehood. The labels point in opposite directions from the ones you'd guess.

Did Xunzi think people were evil? He thought unimproved nature was bad and that deliberate human effort makes it good. The chapter 《性惡》 is a claim about starting material, not about what a person has to remain. See the Stanford Encyclopedia entry on Xunzi for how the term gets read.

Did Hobbes and Xunzi know of each other? No. Xunzi died around 238 BCE and Leviathan appeared in 1651 CE, with no line of transmission between them. The similarity is convergent, which is exactly what makes the disagreement informative.

Is ritual just Xunzi's word for laws? Not quite. Laws work by penalty, and Xunzi has those too. Ritual works by shaping what a person approves of, which is why he compares it to scales and marking lines rather than to punishment.

Which one describes modern institutions better? Mostly Hobbes, in the sense that most compliance systems assume nobody improves and keep the enforcement running forever. Whether that's an accurate view of people or a self-fulfilling design is the live question, and it's the same one behind whether incentives really motivate anyone.

Curious how Xunzi would answer a problem you're actually facing? Ask Xunzi directly, or read more about Xunzi's thought.

A note on texts and translation. The Chinese is quoted from 《荀子·禮論篇》 and 《荀子·正名篇》 on Chinese Wikisource, then checked character by character against an independent edition, the 四庫全書 text with 楊倞's commentary (卷十三 and 卷十六). The two agree on wording throughout, differing only in variant graphs and one particle: 窮於物 / 窮乎物, 規矩誠設矣 / 規矩誠施矣, 非特學無方之民也 / 非特學為無方之民也, and 人多辨訟 / 民多辯訟. All English renderings of Xunzi here are my own; for standard scholarly translations see Eric L. Hutton, Xunzi: The Complete Text (Princeton, 2014) and John Knoblock's three-volume Xunzi (Stanford, 1988–1994). Hobbes is quoted verbatim from the public-domain Leviathan text at Project Gutenberg, with the original 1651 spelling preserved.

Sources

Related reading

Xunzi vs Hobbes: If Human Nature Is Bad, Who Fixes It? · Hundred Masters