Xúnzǐ on Desire

Next week I’ll be taking part in a roundtable discussion at the Pacific APA with Kurtis Hagen, Eric Hutton, and Aaron Stalnaker, and we’ll be talking about what Xúnzǐ had to say about desire. I thought I’d post here some of the thoughts I’ll be presenting there.

The question I want to raise (and I hope it is a provocative question) is whether Xúnzǐ thought of desire as a kind of motivation.

Here’s the argument that he did not. In “Correct Names” (but not the part of that piān that’s actually about naming) he argues against the view that social order requires eliminating or reducing desires. He claims that how we act depends not on what we desire ( 欲) but on what we approve of ( 可). He writes, “Desire does not wait for what we can get, but seeking (qiú 求) follows what we approve of (欲不待可得,而求 者從所可).” His main piece of evidence for this is that we have a deep desire for life, but are still able to choose death when we do not approve of the course of action that would be necessary to preserve our lives.

Now, when Xúnzǐ talks about desire, he focuses on desires for sensory gratification and for social acclaim. He rarely implies that the attitudes we might call moral desires are actually desires; these attitudes would, according to the passage in “Correct Names,” count as approvals rather than as desires.

Part of what’s going on here is that Xúnzǐ treats desires as constants of human nature: they arise spontaneously from xìng 性 regardless of what we do and regardless of how virtuous we are. Thus, moral excellence requires us to transform not our desires but the way we act on those desires (the behaviour that he calls seeking). His conception of desire, then, is rooted in his conception of nature more generally: just as nature provides a constant context to which human societies must respond, our desires provide a constant context that we must respond to when we act. But it is approval, not desire, that determines how we respond.

How does this constitute an argument that desire is not, for Xúnzǐ, a form of motivation? Two things. First, he seems to be saying that it is approval, not desire, that gives us our ends or goals. For example, in choosing death, one does not choose an end that has any basis in desire. Second, he sees no conflict between desire and approval. It is not as if there is a struggle between them that approval consistently wins—so his contrast of desire and approval is not at all like the traditional western contrast of reason and passion, which constitute two competing sources of motivation. His view is just that approval is the sort of attitude with which we adopt an end, and desire is not. And this to me suggests that he is not thinking of desire as a form of motivation.

This isn’t a knock-down argument; the concept of motivation may well be too vague to support a knock-down argument. Still, it impresses me that there’s a genuine question here.

34 thoughts on “Xúnzǐ on Desire

  1. Dan,

    Very quickly:

    A. f you are right, why not simply conclude instead that yu 欲 does not mean “desire”? That would make things sound less odd, perhaps less provocative however, than that Xunzi does not think of desire as a kind of motivation.

    B. What is a yu 欲 then, on your account, if it does not provide motivational impetus? I know you’ve said what their source is–namely, xing–but what do they *do* then, within Xunzi’s psychological theory?

  2. I’ll start with the second question.

    First, Xunzi certainly thinks that we often do act in such a way that our desires are satisfied. So desire does in a way agree with our actions a lot of the time, even if it’s not providing the motivation.

    (In fact, when he defends ritual in Book 19, he argues that ritual shapes our seeking behaviour in such a way that our desires get satisfied without us getting into conflict. So to some extent desire plays a role in justifying the dao that he thinks we should approve of. But it’s not a direct role. The dao as a whole is justified in large part because it prevents the disorder that would result if people’s desires often went completely unsatisfied. So what justifies my actions is not ultimately the fact that they satisfy my desires, but that they amount to me doing my part in practicing a collective dao that tends to lead to the satisfaction of everybody’s desires. Still, I’m often better satisfied to satisfy my desires than other people are, so often in acting according to that collective dao, I’ll be aiming at the satisfaction of my own desires.)

    Psychologically, I think the key is that the desires that Xunzi concentrates on correspond to what he most likely thought were genuine needs and values: the desire to eat when hungry and so on, and the desire for social acclaim. So a plausible role for the desires is to signal to us when we have a need that requires satisfaction, or when there’s an opportunity to gain social acclaim. You could think of it as a kind of perception, maybe. (And here Xunzi’s comparison of the sense organs—which produce the desires for sensory gratification—with officials may be significant, because one role of officials would be to bring reports to the ruler, i.e. the heart, who would then make the decisions.)

    Another part of the story is Xunzi’s view (at least some of the time) that we naturally tend to try to satisfy our desires in ways that promote disorder; we do this if we have not accumulated artifice through ritual and other education. Together with the desire/approval distinction, this implies that we naturally tend to approve of disorder-producing ways of satisfying our desires. (Caveat: I’m not entirely sure he held these two views at the same time, but that’s an issue for another day.) But whatever we make of it, it means that people without the requisite training typically act as if desires motivated.

    And that’s at least part of my answer to the question of why I think “yu” still means desire for Xunzi. I suppose the rest of my answer is that that’s what it meant in the language Xunzi was speaking, he shows no sign that he’s trying to change the meaning of the term, and indeed that’s what it means when he uses it in less theoretically-stressed contexts.

    There are also cases where I think it’s pretty reasonable to think of desire in about the way I’m suggesting Xunzi did. I’m writing a philosophy paper. I notice a desire for food. I think to myself: that’s distracting, I wish that desire would go away; and, that desire probably indicates that my body requires sustenance, and if I don’t do anything about it, I will lose energy and focus; but, fixing myself something to eat would distract me from my work, so I’d rather put it off as long as possible. (I hope I’m not the only one who reacts to hunger this way when caught up with work.) Here, the desire to eat signals a need, and the recognition of the need can be motivating, but the desire itself is not.

    (Aside: in a case like that, it’s useful to think of a desire as a state that has not satisfaction conditions, but going-away conditions.)

    I hope that’s at least the beginning of an answer to your questions. I confess that right now I’m too hungry to be entirely sure. 😉

  3. Er, I forgot to mention one of the main points in response to your first question: the particular yu he focuses on (such as the yu for food when hungry) are pretty clearly desires.

  4. Dan, this is interesting. I just have some minor comments.

    You say, as part of your argument: Xunzi “seems to be saying that it is approval, not desire, that gives us our ends or goals.” And you mention one’s approval of death, which no one desires.

    But could it not simply be that one desires something else more important than death, and dying is the only or best means of achieving it? That’s what Mencius thinks about rightness, in 6A10: rightness is something that he wants more than life. In that case one could approve death, because it is a means to getting something that one desires. So the example of approving death doesn’t seem decisive to me. It needs to be argued that Xunzi does not share a position like Mencius’s, I think.

    Also, one could alternatively interpret Xunzi as claiming that while it is desires that set our ends, approval determines how to optimally fulfill those ends. Ruling out these alternative interpretations would make your position even more plausible, I’d suggest.

    Enough here… I’d like to spare my time for the other thread (old threads never die, as you say, but they do seem to fade into obscurity!).

  5. To add just something (that may seem ridiculous) I’m wondering about: if what you argue for is indeed Xunzi’s view, how would he think animals are motivated? Do non-human animals also have the capcity for approval, or are humans so different from animals that only humans are not motivated by desires?

  6. Boram, the contrast between what Xunzi says in “Correct Names” and the assumptions about desire in Mencius 6A/10 are certainly interesting. Van Norden discusses this in his paper “Mengzi and Xunzi”; though I think he’s wrong to say that MC 6A/10 presents a theory according to which we always do what we most desire, that view is presupposed there, not asserted, I do agree with Van Norden that Xunzi is disagreeing with it. In any case, I take Xunzi to be saying that in the case of death, we aren’t doing it because of a desire, because life is what we desire most.

    (Van Norden’s paper was reprinted in Kline and Ivanhoe, eds., Virtue, Nature, and Moral Agency in the Xunzi.)

    Your #5 isn’t ridiculous at all, that’s a very interesting question, and I’m not sure how to answer it. Presumably the answer has to go something like this: animals are not capable of accumulating artifice, so they inevitably respond to their desires on instinct. So maybe in the case of animals it makes sense to say that desires motivate? I’m not sure, but I’ll certainly think about it more.

  7. Dan,

    Thanks for the clarification. I think it’s still a bit confusing. It seems to me like if yu are the sorts of things that can be “satisfied” then they are the sorts of things that can motivate. Why wouldn’t you describe what you’re talking about, in more common philosophical parlance, as the difference between *sources* of motivation (desires, or yu) and something like willing (ke)? Isn’t it normal to talk about several conflicting motives, i.e. being moved toward certain ends, but not acting on them all? Take those to be desires; they may not be motivating but one of them could be if you “will” one of the particular ends, but that doesn’t make them an less *sources* of motivation.

    But that’s all anyone wants from the concept of desires: they are potential sources of motivation. They can actually motivate if the end is willed. Or if “will” sounds too spooky–here I’m not talking about for Xunzi, but for contemporary philosophers–then from among the ends “proposed” by desires, the desire for one of them either emerges as strongest (Hume types), or it ends up being the desire that you second-order desire to have fulfilled (Frankfurt types). That’s at least the Humean, Davidsonian, philosophically broad idea of desires, isn’t it? If so, then it seems like the contrast you make–desire as a kind of motivation versus desire as non-motivational–seems based on too strong a reading of “desire as a kind of motivation.”

  8. Just a modification on Manyul’s point about willing and desiring. Dan, you translate the line: “Desire does not wait for what we can get, but seeking (qiú 求) follows what we approve of (欲不待可得,而求 者從所可),”with a shift in meaning between the two ke, as “can” and as “approve.” Couldn’t both mean “can,” so this would be saying we desire something whether or not it can be attained, but we only actually seek something that we think can be attained. So desire would not be necessarily motivating in itself, but desire with a perception/judgment of attainability would. This would still make desire the motivation, and it would avoid taking “ke” as something too close to will. I’m not looking at the context of the passage right now, though, so maybe in context this reading of ke is contradicted.

    If you placed this in a broader context of conflicting desires, we might even say that some things we desire cannot be attained without losing other things we desire, in which case life might still be considered as something we desire but that “cannot be attained 不可得”, and so we could choose death.

    (This is my first post (anywhere) and it feels a bit strange to just cut into the conversation, without saying hello or something. Manyul, we corresponded a bit several years ago — it is nice to cross paths again.)

  9. Hi Dan,

    My concern is similar to Manyul’s; it seems counterintuitive to suggest desires are not a kind or source of motivation for Xunzi. Some Xunzi passages indicate our “approval” can lead us to refrain from pursuing the objects of desire or to do things we don’t desire. But this doesn’t entail that desire alone doesn’t have motivational force or can’t move us to act.

    Consider, for example, the passage toward the end of Book 3 about fixing one’s desires only after weighing different aspects of some potential end (欲惡取舍之權). I think this discussion assumes that people sometimes take action (取) directly on the basis of a desire. A natural way of describing the role of desire there is to say it can motivate action.

    It’s possible, of course, that this passage and the material about “approval” (可) in Book 22 come from different periods in Xunzi’s life and express distinct views. The concept of “approval” may have been a later development.

    Also, I think Xunzi sees some ethically unrefined people as ignoring their own capacity for “approval” and just acting on whatever their nature (xing) prompts them to desire. (Such people would be a Xunzian analogue of Frankfurt’s “wanton,” perhaps.) For such people, it seems appropriate to say desire is a motivation.

  10. Hello Franklin; welcome to the blogosphere. You are now generating “new media.” I hope you’ll keep checking in; the blog certainly gives new meaning to the joy that Confucius refers to in Analects 1.1: 有朋自遠方來,不亦樂乎?

  11. Very interesting. And welcome, Franklin!

    Franklin, the reason the second ke (可) shouldn’t be read as “can,” I think, is because it is preceded by suo (所), which always (usually?) takes a verb. But there might be some way of making yours a viable interpretation.

  12. Hello everybody. I am Siufu from the Chinese University of Hong Kong. Chris referred me to your discussion. Actually I will also present a paper on Xunzi next week at the APA meeting, but at another group program.
    Since I will have a class soon, I will try to be brief.
    I agree that there is some important questions concerning the relation between yu (desire) and ke (approval). However I don’t agree that desire isn’t a form of motivation. I tend to interpet Xunzi as saying that desires, however motivating, do not bring about action without the mediation of approval. I would like to refer to some sentences in the same paragraph of Book 22: “Thus, desires surpass it but the movement falls behind, this is stopped by the heart-mind. …… Desires fall behind but the movement surpasses it, this is prompted by the heart-mind.” (故欲過之而動不及,心止之也。…… 欲不及而動過之,心使之也。) It seems to me that if desire does not have motivating power, there is no need for the heart-mind to stop anything. Also, words like “surpass” (過) and “fall behind” (不及) sound like describing the “force” of desires.

  13. Hi Siufu, welcome to you, too.

    You write: “…if desire does not have motivating power, there is no need for the heart-mind to stop anything.” This seems to me a very strong point.

  14. Manyul, part of the idea is that what’s going on in Xunzi, or at least in the passage from “Correct Names,” is not at all Davidsonian. There’s certainly no suggestion there that what you’re acting on is the desire that turns out to be strongest, and no mention of second-order desires. And here at least, the heart isn’t simply selecting between desires, because Xunzi presents the case of choosing death over life as a case in which you’re doing something you don’t desire at all.

    Again, desires for Xunzi are usually restricted to desires for bodily satisfaction and for social acclaim. These are things that it’s normally reasonable to take as ends, but that doesn’t imply that it’s because we desire them that it’s reasonable to take them as ends, and the contrast between desire and approval implies (this is how I read it anyway) that desire is not the recognition or acceptance of them as ends.

  15. Hi Frank! As Justin points out, these are syntactically distinct uses of “ke 可.” “Ke + [verb]” is “[verb]-able”; “ke” as a transitive verb (which is what it must be after “suo 所”) is “to treat as or consider to be ke,” in which the modality of “ke” could vary from pure possibility through acceptability to correctness. So the way to get your reading is to take it to mean “to consider possible” (which I suspect is what you meant).

    But I don’t think we should read it this way. First, that would have Xunzi saying we do what we consider possible; but there are lots of things that are possible that he thinks we shouldn’t do, and he’s not talking about changing people’s conception of what’s possible. Second, I assume that in the choosing-death example, he really is talking about a choice; in other words, he’s not thinking about a case in which death is unavoidable. (And he also doesn’t say anything about conflicting desires.)

  16. Chris, you’re of course right to raise the possibility that Xunzi changed his mind about relevant issues. But I’m not sure I’d count the passage from Book 3 as evidence that he did.

    In that passage, Xunzi talks about weighing the things that are desirable and hatable (or whatever; 可惡), and weighing the benefits and harms of things. So what you’re weighing here aren’t the desires. Also, I’d take “欲惡取舍” differently from you, I think, as “desiring and hating, selecting and rejecting,” with the second pair parallel to the first (and corresponding to benefit and harm in the rest of the passage). So what you’re weighing doesn’t have to be something revealed to you by desire. (Also, read this way there’s no implication that selecting, or taking action, is on the basis of a desire.)

    The one line in this passage that strikes me as difficult for my interpretation is where he says you “fix desires, hatreds, selections, and rejections.” This suggests to me that the weighing process results in a change to desire, which is something Xunzi normally seems to think isn’t necessary. One possibility: this passage isn’t focusing so narrowly on natural desires as do some of Xunzi’s writings.

    Another passage you could cite in favour of the reading you’re giving this passage is the neat list of action-theoretic glosses near the start of Book 22: “The qing being so and the heart selecting on behalf of them, call it deliberation.” This doesn’t mention desire or approval, but qing and the activity of the heart are close enough that I need to say something about this, I think. And there’s no way out here by saying that he’s not talking about natural desires or emotions, because he explicitly is. Still, this passage is clarifying some terminology, not setting out a theory, so it might leave out certain things. And (as I said in response to Manyul), I agree that in the normal case you are trying to satisfy your desires, and it wouldn’t be too odd if Xunzi concentrates on the normal case here.

    I agree it would make sense for Xunzi to say that bad people ignore their capacity for approval, but I don’t think he leaves himself room to say that. In general, I think Xunzi has trouble explaining exactly what’s wrong in bad people. (If my analysis of Book 23 in Early China 26-27 is correct, this is an issue he changed his mind about more than once.) If it is somehow instinctive for us to do bad things, what does this instinct consist in? Sometimes (e.g., in what I take to be the earliest texts in the Book 23 core) he implies that it just an instinct to obey our desires. But I think his later view is that there’s no such thing as simply obeying our desires (since the desires he talks about can be satisfied in a variety of ways), and that he has to say that our instincts would lead us to try to satisfy our desires in bad ways. If we combine this with the desire/approval distinction, the resulting view is that we instinctively approve of bad courses of action. (And then moral improvement consists in transforming those instincts by forming new habits of evaluation.)

  17. Hi Siufu! I read that line (故欲過之而動不及,心止之也…… 欲不及而動過之,心使之也) as saying that the heart stops and initiates action (dong 動). You take the word “zhi 止 (stop)” to imply that if the heart didn’t act, then one would act differently, presumably following the desire. But I don’t think the rest of the passage leaves this open as a possibility. Or, at least, it doesn’t leave this open as a genuine psychological possibility. Maybe Xunzi is imagining how we would act if (counterfactually) the heart didn’t get involved, and is saying that the heart prevents us from acting that way. Maybe for beings like that desires would motivate, but this does not imply that for beings like us (whose hearts inevitably do get involved) desire motivates. (Taking up Boram’s question in #5, perhaps he’s thinking of animals here.) And in any case, I don’t think we need to place that much weight on the word “zhi 止” here. Xunzi could just be saying that when action starts and stops, it does so because of approval, not desire.

    Notice also that this line implies us that if (counterfactually) we did act from desire, we would not act the way that a good person does act—which is to say that Xunzi is implicitly denying that a good person’s behaviour is motivated by desire. (Of course I think he’s also committed to the same view with respect to bad people. And this comment is directed more towards earlier posts in the thread, not you, Siufu.)

  18. Dan, I don’t disagree with you on this: “the contrast between desire and approval implies (this is how I read it anyway) that desire is not the recognition or acceptance of them as ends.” What I was trying to press is that the scope of the term “desire” is usually taken to include psychological sources of motivation–they “suggest” ends through their foci–not that they are states of acceptance or recognition. That certainly seems to be a job for some other faculty (if we aren’t being Humean, and taking strength of a desire to be all there is), for example ke, or “will,” or some other such thing. The link between desires and motivation in this broad sense doesn’t seem severed by Xunzi just because desires by themselves are not the sort of things that *decide* which ends to pursue. Ke might provide the *actual* ends, but among the *possible* ends are the ones suggested by desires because desires are “end-directed” if not “end-determining.” That’s my point.

    Chris’s point (#9) is similar, though he goes further. Chris is saying that in fact, Xunzi thinks yu are the sorts of things that *could*, though they might not always or even usually, motivate on their own; they are not only potential sources of motivation (end-directed), they sometimes provide motivation (are end-determining) *without* any further approval by ke. I’m not committed to this, and I’ll let Chris defend it.

    I think if your argument about Xunzi were right, then we could say of any philosopher who thinks action requires desire *plus* some other faculty of choice, will, approval, or what have you to determine an end, that they don’t think desire provides sources of motivation. That seems to me to misrepresent the concept of desire.

    It seems to me like you might have an argument to show that Xunzi regards yu as under-determinative of actual motivation; and, Xunzi thinks, as a matter of fact, yu are not the sort of things we could manipulate–they are constant like Heaven, we can only act with knowledge of their predictability. Based on those two premises he regards ke as the only viable focus of moral cultivation. Ke can be changed, yu cannot; it only makes sense to try to change the things you can.

  19. Manyul, it could be that the conclusion that desire isn’t for Xunzi a form of motivation requires too narrow a conception of motivation; that’s why I avoided asserting that conclusion. Still, I’ll try to respond to your objection.

    You say that desires suggest ends, and that this is enough for them to count as motivations (or sources of motivation). However, there are other ways in which ends can be suggested that clearly do not count as motivations or sources of motivation. For example, if someone suggests that you try to achieve some end, the suggestion itself need not be a motivation or source of motivation, but you could still take it up as an end (for example, by approving of it). So the fact that desires suggest ends isn’t enough to qualify them as motivations (or sources of motivation).

    What’s missing, I think, is the idea that desires are a sort of vivid awareness that it would be good (in some sense) to achieve their objects. It’s the vividness of the awareness that I think we’d usually take to supply the motivation, even if we think some selecting faculty also has to get involved. Xunzi’s view (as I’m interpreting it) is different from this; it’s that even when desire suggests an end, it’s the heart that weighs the value of the object of the desire, and that weighing is what supplies the motivation.

  20. Dan,

    A few things and then I’m off to teach:

    “that desire isn’t for Xunzi a form of motivation requires too narrow a conception of motivation; that’s why I avoided asserting that conclusion” — I think you do assert this conclusion (see your post, substitute “form of motivation” for “kind of motivation”).

    I agree with you that a particular desire might not end up being a motivation, but it is always a potential source of motivation. Likewise, I agree that other things could be sources of motivation.

    “Vivid awareness” that something would be good is far too strong. However, desire has as its “constitutive” aim to regard something *as* in some way good. That doesn’t make desire a value judgment per se, it merely inclines one toward something. There’s a good David Velleman paper on this, I think it’s called “The Guise of the Good” or something like that. So, I think at least the contemporary, relatively thin notion of desire is something that (a) Xunzi’s concept of yu seems very much similar to, and (b) admits of a general motivational *role* as potential source of motivation.

    By the way, I don’t think this is just a terminological dispute; it seems to me more like you’re not using the concept of “desire” correctly in analyzing Xunzi’s concept of “yu.”

  21. The post was meant to go like this: here’s an argument that p; but I’m not sure the argument works, because I think a key term (“motivation”) might be too vague to carry the argument. Maybe it wasn’t clear enough.

    I was actually thinking of one of Velleman’s papers when I wrote #20, specifically his argument that desire can be for something conceived of as evil (so “good” wouldn’t be the wrong word). But today is officially devoted to grading and Nagarjuna, so I didn’t spare the time to look for a better formulation.

    Anyway, the key claim in #20 is that desire as motivation has to do more than suggest an end, it has to (as you say) somehow incline you towards that end, and that’s what I take Xunzi to be denying.

    I don’t agree that Xunzi’s conception of yu is similar to a thin (Davidsonian, say) conception of desire. If it were, then ke would trivially count as a kind of yu, which it clearly does not.

    To what extent would it address your worries if I reformulated my claims in terms of yu, allowing that as Xunzi conceives it, yu is not really desire?

  22. Dan,

    “To what extent would it address your worries if I reformulated my claims in terms of yu, allowing that as Xunzi conceives it, yu is not really desire?” — That would make much more sense to me–it would be clearer to me what your thesis is; and it would probably nip similar objections to mine in the bud.

    But I don’t know that Chris, Siufu, or Franklin would agree that your clearer thesis is correct. Anyway, I’ll let them take that up.

  23. Dan,

    It does seem like your account becomes unnecessarily confusing (although more provocative) by using “desire.” Not only does yu seem like it may not be “desire,” while it is clear that ke is not yu, I’m no longer so clear why on your interpretation ke couldn’t be conceived of as “desire.” It may be better to just explain what yu and ke do, perhaps with the basic point being that decision making for Xunzi cannot be conceptualized well in terms of “desire.” I assume you take ke to be a kind of desireless, unattached motivation, which would be interesting to know more about. In the same way, it would be interesting to think more about the difference between knowing you should eat and feeling hungry, a distinction your account seems to make only in terms of vividness.

    I still lean toward taking ke as something guiding desire, something like an ability to see more clearly what will fulfill our desires. I may be too heavily influenced by the passage you refer to in ch. 22: “The qing being so and the heart selecting on behalf of them, call it deliberation.” To come back to your reference to ch 19, I would take it that a person who deliberates well would desire to follow rituals while one who deliberates little or poorly would not desire to follow ritual, even though both are acting on the same natural desires. They would ke differently and act differently, but they would still be motivated by desires. This is the kind of thing I had in mind in my earlier post, taking the second ke (as you note) as something like “consider possible.” The problem (for my reading) is that ke used like that without a verb usually (always?) has a sense of permissibility as well as possibility, and those seem to be blurred in that sentence. It seems odd, though, that the two clauses would have no connection (the first noting only possibility and the second only approval). Anyway, I think ke as the perception of what will in the long run realize our desires makes sense with a lot of the evidence. It may not fit, though, with choosing death or with any example in which doing what we ke clearly contradicts the long term realization of our desires. I’m not sure about that any more.

  24. Dan,

    I do not mean to suggest that “if the heart didn’t act, then one would act differently, presumably following the desire”. What I like to suggest is that desire has a kind of force to prompt the heart to approve in its direction. (I tend to read the sentence of Book 4 “則其心正其口腹也” in this line.) I agree with you that the heart is inevitably involved in actions for beings like us. But I do not take this to imply that the desire then is not a form of motivation.

    I also agree with your reading that Xunzi sees no conflict between desire and approval. For Xunzi desire and approval function at different level in human psychology. I would probably agree with you that “approval is the sort of attitude with which we adopt an end”. Still, desire could be a form of motivation, just that the sort of motivation might be a kind of brute force (here Boram’s suggestion of the case of animals is relevant). If desire is not a form of motivation, what sort of thing it is for Xunzi?

    I like to point out another sentence in Book 22: “Taking what is desired as obtainable and seek for it, that is unavoidable for the feelings; thinking it approvable and giving direction for it, that must come from understanding.以所欲為可得而求之,情之所必不免也;以為可而道之,知所必出也。” I think seeking (qiu求) is different from approval (ke可). Thus I would not translate “而求者從所可” as “seeking (qiú 求) follows what we approve of”. I think “求者” should be “those who seek”. I take it that desire prompts us to seek in a certain way, but whether we will indeed seek in such a way depends on the heart’s approval.

  25. Frank, I would find it it plenty provocative to put it as you suggest, that it’s a mistake to interpret Xunzi’s action theory in terms of desire (not that the aim is to come up with a provocative thesis).

    I take ke to be a normative attitude directed at courses of action, ultimately at the dao. I wouldn’t rule out the possibility that it would or could be vivid—but it would be a vivid awareness of something as right, or somesuch, rather than as desirable (however we spell that out).

    If you’re worried about the different senses of “ke” in that sentence, as I translated it above, you could take “ke-de” as “acceptable to get,” I suppose. But I don’t think that’s necessary; that use of “ke” is as a prefix, and there’s not much reason to presume it’ll mean exactly the same thing as a prefix and as a full verb.

    (I see the case as being quite similar to “able” vs. “-able.” And note that the exact significance of “-able” depends on the verb it is attached to; “gettable” indicates possibility, “desirable” indicates something more than that, often normative correctness; I think there’s probably the same distinction between “ke-de” and “ke-yu” in classical Chinese.)

    Suppose you’re right that ke is a motivating recognition of what will best satisfy my desires over the long term. Then notice: ke is not itself a desire, and the desires it leads me to try to satisfy are for the most part not desires I am now experiencing; so on this picture, too, felt desires are not motivating. (We might say instead that my present beliefs about my future desires motivate me.) So while I don’t think that ke always aims at the satisfaction of desires, this does not entail that felt desires motivate. (And if we are going to say that the future desires motivate, and thus allow that my motivations are not restricted to my current psychological states, then we might just as well say that the objects of those future desires motivate.)

  26. Hi Dan (et al.),

    To me, a natural way to interpret your initial question is: For Xunzi, do people sometimes act directly in response to how they draw the yǜ 欲/wù 惡 (desire/dislike) distinction, without some other psychological state bringing about the action? I think the answer is yes, for Xunzi they sometimes do.

    Saying that the psychological state of yǜ 欲 can motivate action in this way is similar to saying that the state of deeming something shì 是 or shàn 善 can. It doesn’t entail that occurrent yǜ——felt cravings, say——always motivate action, or that in cases when they don’t, it’s because another attitude, such as “approval,” has overpowered them.

    I think the passage I mentioned about 欲惡取舍之權 implies that people will typically act on the outcome of the “weighing” process. The main point of the passage is, as you say, that we should carefully weigh the pros and cons in things before settling on our desires/dislikes, on what we deem beneficial/harmful, and what we select and reject. Selecting/rejecting I think refers to action, since the text talks about avoiding blunders when one “moves” (動) or “acts” (為). A natural way to read the pairing of desire/dislike or benefit/harm with selecting/rejecting is that the latter is typically grounded in either of the former. So I think the passage implies that we sometimes act on yǜ 欲.

    The text does imply that the weighing process might lead us to change what particular things we desire, but I don’t think it suggests that we will change what we consider desirable. The point——which is reiterated, I think, in 22.6b (Knoblock’s numbering)——is that the objects of desire typically include a mixture of desirable and undesirable features, and it is wise to attend to both before determining whether we desire something, all things considered. The result might be that we stop desiring something we originally desired, but our desire for its desirable features could remain unchanged.

    In terms of the general position on Xunzi that I think you want to take, I don’t think you need to argue that yǜ (desires) do not have motivational force—-that is, they never motivate action on their own. I think you need only argue that they do not always motivate us and that we can have them without being motivated by them.

    I agree that your post raises a very interesting, deep question. If yǜ can motivate action, and if, as Xunzi says, becoming good isn’t a matter of eliminating disorderly or bad yǜ, then what is the status of such yǜ in a good person? I myself want to think about my view on this more.

  27. Hi Siufu,

    I’m not sure I understand what you mean by saying that desire and approval function at different levels. Let’s assume that (according to Xunzi) desires do pick out particular ends (or maybe even courses of action) as desired. Then I take Xunzi to be saying that these ends can differ from the ends picked out by approval, but that when they do, this does amount to a conflict—we simply act on approval, not on desire. In that case at least, I don’t know what it would mean to say that desire is a brute force, or, to return to the general issue, what it to mean to say that it is a motivation.

    The line you quote does say that desire inevitably causes the qing to seek, and maybe that implies some kind of push. But it’s a push that’s disconnected from action, or anyway that’s what I take these passages to imply. What actually guides action is, here, understanding or knowledge (zhi 知). This is one of the lines (Chris cited another one above, and I mentioned yet another) in which Xunzi may be implying that the understanding settles on (approves and takes guidance from) some end that has been presented by a desire. But even if this is what’s going on, it’s the understanding that’s weighing the ends presented by desire, so a desire’s degree of push (its strength) doesn’t seem to be relevant to the outcome. And in the passage I quoted in the main post, I take him to be saying that approval can settle on ends that have no support from desire whatsoever.

    A point I haven’t made previously: notice that towards the end of the passage I quoted, Xunzi asks (rhetorically), “If what the heart approves of matches the correct pattern, then even if desires are many, how does that harm order? […] If what the heart approves of misses the correct pattern, then even if desires are few, how does this prevent disorder? (心之所可中理, 則欲雖多,奚傷於治?……心之所可失理,則欲雖寡,奚止於亂?)” Even if we imagine (counterfactually, because Xunzi does not allow this as a real possibility) our desires could change in number, this would have no impact on how we act—not even, apparently, on how we act in the absence of correct education and habituation. I take this to be a pretty strong indication that at least in this passage, Xunzi is saying that the ends selected by approval need not be desired, and that ends selected by desire need not have any pull on approval.

  28. Chris, you didn’t really respond to my suggestion that what’s getting weighed in that passage are not the desires and dislikes, but their objects.

    The question you raise at the end of your comment is a good one. Does Xunzi ultimately want two action theories, one for people without proper education and habituation, and one for people with it? If so, then we could take him to be saying that people without it are motivated by desire, but people with it are not (or at least are not always).

    But I don’t think we should take this line. It would conflict with Xunzi’s view that people do not differ with respect to our natures. And the questions I quoted in #28 seem to me to imply pretty strongly that even in a bad person, who we might expect Xunzi to say is governed by desire, something else is going on.

    What else is going on? I like the suggestion I made in #17, that his view must be that we instinctively approve of bad courses of action, but that approval (unlike desire) is subject to change.

  29. Dan,

    Sorry I didn’t respond to that point directly. In fact, I think it’s clear that what’s being weighed are the desirable and undesirable aspects or features of the object. We agree on that.

    I’d been thinking that a fairly obvious way to draw together Xunzi passages in which “approval” is mentioned and those in which it isn’t is to say that, for Xunzi, people tend to approve their desires by default. Ethically unconscientious or uneducated people, who are “just their mouth and belly,” automatically approve the base desires arising from their nature. (Xunzi doesn’t say this explicitly, though.) So I think we agree about your closing remark in #29 too.

  30. Hi Dan,

    I think underlying our discussion are different assumptions about desire and motivation. There are two interrelated questions: the scope and conception of desire, and the scope and conception of motivation. I have to admit that I don’t have clear answers to these two questions.

    I have reservation over your idea that “ends selected by approval need not be desired, and that ends selected by desire need not have any pull on approval”. This partly depends on the conception of desire we adopt. There is a sense of desire that whenever we want to do something, we have a corresponding desire. Some (such as G. F. Scheuler in Desire: It Role in Practical Reason and the Explanation of Action) call this pro-attitude to distinguish it from desire proper.

    Even if we put aside different conceptions of desire, I don’t think the text warrant the interpretation that desire has no impact on action. I am not sure if this is your position when you suggest that desire is not a form of motivation and that desire, even if has a push, is disconnected from action. As I read the text, Xunzi only suggests that desire is never a sufficient condition of action and that approval is a necessary condition of action. If I understand you correctly, you want to suggest that desire need not be a condition of action, with which I disagree. Xunzi does not state explicitly in the relevant passage what role desire play in action. I like to suggest that he thinks desire is also a necessary condition of action. I might agree with you that the heart stops or initiates action, but I think it significant that in both cases Xunzi mentions desire as well: in one case, desire surpasses; in another case, desire falls behind. Xunzi doesn’t say that when the heart initiates action, desire is completely irrelevant. You are right to point out that Xunzi stresses that the heart’s approval determines order or disorder. But we need not interpret this as saying that desire has no role to play in action.

    I tend to think that approval does not pick out ends by itself, independent of desire (desire in the broad sense, as whatever we want). I rather think that approval is an activity of organising and structuring our desires (again desire in the broad sense). We want different things and we seek in different ways, the heart judges on our wants and our seeking and guide our action through approval. I take his definition of deliberation (lu慮: feelings being so and the heart-mind makes a choice for them) in Book 22 as support for such a view.

    I say desire and approval function at different levels and that there could be no real conflict between them. Yet we do have conflicting desires and also conflicting approvals. We (You, Chris and I) both agree that people sometimes automatically approve base desires. Such approval could conflict with approval after better deliberation. When I say desire does have motivating force, I mean its effect on the heart and the consequent approval. I do not mean that desire could be direct motivation for action. On this I agree with you that action must be mediated by approval. Though I also leave open the possibility that we could move (but not act) under the direct influence of desire. Just that such movement might well be the same as movement in animals’ case and quite rare in human beings.


  31. Siufu, it’s clear isn’t it that Xunzi is not working with the broad conception of desire that you mention? If he were, then approval would be a kind of desire, but it isn’t, and he’d have to allow that desires can change, but he doesn’t.

    But I take your central claim to be that desire (narrowly construed) is necessary for action, in the sense that the ends that we approve of, and thus seek to achieve, must be desired ends. So I’ll talk about that.

    I continue to think that the passage I started out with (especially in the example of choosing death and in the counterfactual questions towards the end, but it’s also my sense of the passage as a whole) strongly suggests that desire isn’t necessary. And I read the passage from Book 3 as telling us that we recognise the ends over which the heart deliberates not only by drawing the yu 欲 / wu 惡 distinction (to adopt Chris’s formulation), but also by drawing the li 利 / hai 害 distinction.

    The bit about deliberation near the beginning of Book 22 does say that the heart chooses on behalf of qing 情; if this is consistent with the passage I’ve been focusing on (as I read it), then this cannot be the whole story. But I don’t think it’s meant to give the whole story (one reason for thinking this is that it doesn’t mention desire at all). There’s also the possibility that the passages really are inconsistent, presumably because Xunzi changed his mind about these issues. But though we should always keep this sort of possibility in mind (and though I’m on the record as arguing that Xunzi did change his mind about related issues), I don’t see that as a way out for me here.

    One important issue is that virtuous people, people who have been correctly educated and habituated and so on, are going to have to end up approving of the dao, and I doubt Xunzi thought that the ends we take on in approving of the dao must be desired ends. He does sometimes say that certain people would get more satisfaction if they adopted the dao, but, first, that’s satisfaction over the long term, not satisfaction of occurrent desires; and, second, it’s natural to take him as saying that they would gain more satisfaction over the long term if they adopted new ends, ends not themselves grounded in desire. (Saying that adopting a new end would lead to greater satisfaction of future desires does not entail that the new end is a desired end, even if we take for granted effective means-ends reasoning.)

    Consider a case of conflict Xunzi mentions in Book 23: a young man is hungry and desires food, but he sees his elder and does not dare eat first. In this passage, desire is entirely on the side of eating right away; Xunzi rules out a desire to yield to elders when he says that if one follows qing 情 and xing 性, one will not yield. Yielding here must result from a commitment to the dao, and in particular to the rituals and duties that require deference.

  32. Hi Dan,

    The case of Book 23 indeed suggests that natural desires are contrary to the requirement of rituals and duties. Yet such desires are desires in their originally natural form. Xunzi does stresses in Book 23 that we need to transform qing and xing so that they would conform to Dao. If desires are irrelevant to action, what need there be for transformation? (As far as I remember, in your paper on Book 23, you suggest that transformation of xing is problematic, because it causes damage to people’s xing. I tend to think that transformation isn’t a kind of damage. Transformation simply changes the form of expression but not the substance of xing: “Where the form is changed but there is no distinction in substance; yet they are deemed different, this is called transformation.”狀變而實無別而為異者,謂之化。)

    I am flying this afternoon. I shall meet you in Pasadena and we could discuss more.


  33. Hi Siu-fu,

    I think transformation of xing is problematic for Xunzi for a few reasons.

    First, he insists (including in one of the passages that mention transforming xing) that sages and wicked people are the same in their xing, so transforming xing can’t result in a new, changed xing.

    Second, when he explains the distinction between xing and artifice (including in a second text that mentions transforming xing), the distinction clearly implies that the means Xunzi advocates for transforming xing cannot result in a new, transformed xing (because the transformation takes effort and study, its results must count as artifice, not xing).

    And third, to the extent that the means he advocate require us to go against xing, his view seems to call for damage to xing (but this is a problem for Xunzi even if he doesn’t think we can literally transform xing; the move to the theory of artifice is, I think, meant to solve this problem).

    And I’ll see you in Pasadena! I should be there Thursday.

Leave a Reply to Dan Robins Cancel 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.