Why de dicto desires are fetishistic

Internalists and externalists fundamentally disagree about how moral judgments motivate. Internalists think that mor ally good people should have de re desires to do right things, and agents who are motivated by the de dicto desire to do whatever is right are moral fetishists. In contrast, exter nalists accept that morally good people can have both the relevant de re and de dicto desires. My aim is to provide an argument for internalism, which goes beyond the context in which both sides currently debate the relevant de re and de dicto desires. I will first introduce R. Jay Wallace's under standing of a fetish, according to which to have a fetish is to devote interest and attention to objects that do not deserve such responses. Then, I will argue that the relevant de dicto desire in the debate is fetishistic because rightness itself is not a worthwhile goal to have in our moral deliberation and motivation. Moral agents who are motivated by the de dicto desire to do whatever is right, even


|
ZHANG between moral judgments and motivation, which means that moral judgments necessarily motivate. Externalists, in contrast, deny that claim. They argue that there is only a contingent connection between moral judgments and motivation.
In The Moral Problem, Michael Smith (1994) defends his rationalist version of internalism with the famous fetishism argument. Even if this argument was first introduced as an argument for a certain form of internalism, with it, Smith actually has managed to expose a general problem of externalism. According to Smith, internalists can explain the reliable connection between moral judgments and motivation in an intuitively plausible way in which moral agents have de re desires to do what they judge to be the right thing. Yet, as externalists too must explain that same connection in an external way, moral agents within the externalist framework would need to have a de dicto desire to do whatever is right to be motivated. This, according to Smith, means that moral agents are concerned about something that is not of primary moral importance and they thus become so-called moral fetishists. This is why defending the fetishism argument, understood in this way, is not only important because it might provide evidence for a certain specific form of internalism but rather also because it can be used to expose a fundamental problem of externalism.
Of course, the externalists have made many objections to the fetishism argument. Some externalists think that the relevant de dicto desire to do what is right does not amount to a moral fetish (Carbonell, 2013, pp. 463-470;Lillehammer, 1997, p.193), but rather it can even be a part of the motivational structure of the good and strongwilled persons who have both the relevant de dicto and de re desires at the same time (Copp, 1995(Copp, , pp. 212-213, 1997Olson, 2002, p. 91;Svavarsdóttir, 1999, pp. 205-206). Therefore, it will be useful to try to evaluate and respond to some of the externalist objections carefully and convincingly.
The ongoing debate, however, lacks sufficient internalist responses to the externalist claim that moral agents can have both the de re desires to do right things as well as the de dicto desire to do whatever is right. In this paper, I will argue that the relevant de dicto desire is fetishistic itself even in that case. A moral agent would still be a moral fetishist, even if she would only in part be motivated by the de dicto desire to do whatever is right. I will begin from Smith's fetishism argument in Section 2. Then, I will discuss Copp's co-presence objection in Section 3.
As a response to the co-presence objection, I will first introduce R. Jay Wallace's more general understanding of a fetish in Section 4. Then, in Section 5, I will argue that moral agents who are motivated by the relevant de dicto desire still count as moral fetishists.

| Smith's observation
Internalists and externalists have different views about the connection between moral judgments and motivation.
There exists, however, at least one phenomenon about moral motivation, which should be accepted by both sides.
As Michael Smith (1994, p. 71) has put it, we can observe that 'a change in motivation follows reliably in the wake of a change in moral judgment', at least in good and strong-willed persons. 1 This phenomenon is so common that it can be observed in many everyday situations. In The Moral Problem, Smith (1994, p. 71) provides us with an example to illustrate this reliable connection between a change in moral judgment and a change in motivation. 1 Smith (1994, p. 71) originally described that a change in motivation reliably follows a change in moral judgments at least in the good and strong-willed person. The terminology of 'good and strong-willed persons' caused some externalists to misunderstand Smith's view (Copp, 1997, pp. 49-51). Due to this reason, Smith (1996bSmith ( , p. 177, 1997 clarified that the terminology of 'good and strong-willed persons' refers to those who (1) can reliably conform their motivation to their moral judgments and (2) do not suffer from weakness of will and the like that can cause incoherence between their moral judgments and motivation. Because of this, we can accept that the term 'good and strong-willed' means at least roughly the same as 'practically rational' in Smith's works. | 3 ZHANG Let us suppose that you and I are engaged in an argument about which party we should vote for. I have already judged that we should vote for the libertarians and thus I am already motivated to vote for them. But, during an argument, you convince me that voting for the libertarians is wrong and, instead, I should vote for the social democrats. Perhaps you manage to convince me that the social democrats will better promote the values I thought are promoted by the libertarians. You might also be able to convince me that the values I thought would be promoted by the libertarians are themselves essentially misunderstood by them. At this point, if I am a good and strongwilled person, what will happen to my motivation, since I have changed my judgment? According to Smith, it is reasonable to think that I will be motivated to vote for the social democrats following my change in judgment.
The question then is: how can we explain the reliability with which our motivation changes to match our judgments? As internalists and externalists have different views about whether there is an internal, modal connection between moral judgments and motivation, both sides offer different kinds of explanations of the previous reliable connection in the example. In the rest of Section 2, I will compare the internalist explanation and what Smith assumes to be the externalist explanation.

| The de re desire to do the right thing
As internalists believe that there is some kind of an internal connection between moral judgments and motivation, it could be suggested that it is very easy for internalists to explain the previous observation. To see this, let us consider Smith's (1994, p. 61) own formulation of internalism: The Practicality Requirement: [Necessarily], if an agent judges that it is right for her to φ in circumstances C, then either she is motivated to φ in C or she is practically irrational.
Smith's rationalist version of internalism provides an explanation in which the reliable connection between moral judgments and motivation is explained internally. The practicality requirement entails that, when an agent judges that it is right to φ in C, then absent weakness of will and the like she will come to have a direct de re desire to φ in C. This is because rationality disposes one towards coherence between one's judgments and motivation (Smith, 1994(Smith, , pp. 62, 151-161, 1996a. According to the practicality requirement, in the previous case, if I believe that it is right for me to vote for the social democrats, and I am not suffering from weakness of will or the like, I will have a direct de re desire to vote for the social democrats. Smith (1994, p. 75) believes that being motivated in the previous direct way in which moral judgments generate de re desires is how we normally assume virtuous people are motivated. Literally understood, the phrase 'de re' means 'regarding the thing'. In order to gain a better understanding of de re, let us consider this sentence: 'Kalista desires to do what is right' (Dreier, 2000, p. 621). When read de re, we can understand this sentence to mean that Kalista desires to do specific things that are right such as helping the poor and the elderly or taking care of children. Ultimately, Kalista is here directly moved by the right-making features of actions: these actions, for example, make vulnerable individuals better off. Now, according to the internalist account of moral motivation explained above, when Kalista makes judgments about which actions are right for her to do, these judgments produce in Kalista a corresponding, intrinsic desire to perform these actions directly. No further factors or desires are involved in how Kalista comes to acquire her desires to do the right things.

| The de dicto desire to do whatever is right
Although externalism denies that there exists an internal modal connection between moral judgments and motivation, they still believe that moral judgments can contingently motivate (Lillehammer, 1997, p. 1; Svavarsdóttir, 1999, 4 | ZHANG p. 162). Smith (1994Smith ( , p. 73, 1997) then assumes, on behalf of externalists, that there must be something other than the moral judgment itself, for example, a desire to do what an agent judges to be right that can explain the connection between moral judgments and motivation. According to Smith, externalists must think that, if an agent judges that it is right for her to φ in C and she has a desire to do what is right, then the agent will be motivated to φ in C. The agent's desire to φ in C is, on this view, a derivative desire because it derives from a moral agent's judgment and her non-derivative de dicto desire to do what she believes to be right. For instance, if I judge that it is right to vote for the social democrats, then this moral judgment itself and the de dicto desire to do whatever is right will produce a derivative desire to vote for the social democrats in me.
According to Smith, even though the previous externalist account can explain the reliability of the connection, this explanation would still be unacceptable. He has argued that the way in which moral agents would be motivated externally gives us sufficient reason to reject externalism. Let us see why Smith thinks that this is the case.
The previous view suggests that virtuous people are ultimately motivated by their non-derivative de dicto desire to do what is right. Literally, the phrase 'de dicto' means 'about what is said'. To understand why this desire to do what is right is a de dicto desire, let us return to the example we already discussed earlier: 'Kalista desires to do what is right' (Dreier, 2000, p. 621). When we understand this sentence in the de dicto way, we think that Kalista has an abstract desire to do whatever she happens to think is right, under that description as the right thing to do. Because of this desire, Kalista may desire to help the poor and the elderly or to take care of children.
Nevertheless, it is not because these actions are themselves right that Kalista is moved to do these things-it is not the right-making features of these actions that she cares about directly. Rather, the reason why Kalista chooses to do these things is that she has a de dicto desire to do whatever is right and those actions just happen to be the things she believes to be right. The desires to do specific right things such as helping the poor thus here derive from the fundamental desire which is to do whatever is right. Kalista can hold this de dicto desire to do whatever is right even if she has no idea about what the right thing is.
Yet, if we use the de dicto desire to do whatever is right to explain the reliable connection between moral judgments and motivation, we will find it difficult to explain virtuous people's behaviour in a way that would match our intuitions (Smith, 1994(Smith, , p. 75, 1997. It seems that, within this externalist framework, if a moral agent chooses to be honest or to help her friends and family, the desire to be honest and help one's friends and family would derive from the good person's more fundamental desire, which is the de dicto desire to do whatever is right.
We would, of course, hope that good people care non-derivatively about honesty and the well-being of their friends and family. When asking why their moral concerns change because of their moral judgments, we would not expect them to answer: 'although I have no inclination to do these actions in themselves, I really want to do what is right'. We normally think that being motivated merely by the de dicto desire to do whatever is right would be too cold and inhumane-it is not the way a caring moral person would be motivated. It appears that people who are motivated by the relevant de dicto desire would actually care about something that is not primarily important in morality. Thus, it can be argued that, if an agent is motivated by the de dicto desire to do what she believes to be right externally, she has a fetish or moral vice.

| THE CO -PRE S EN CE OBJEC TI ON
In the fetishism argument, Smith claims that externalists can only account for why an agent's motivation changes after a new moral judgment by referring to a de dicto desire to do whatever is right. According to him, being motivated by the de dicto desire to do what is right indicates that an agent does not have direct concerns for the things that really matter. So, the cost of the externalist explanation is that it would turn morally good people into moral fetishists.
The externalist objection to the fetishism argument, which I consider here, is that the relevant de dicto desire is not a problem as Smith argues. Arguably, in addition to the de dicto desire to do whatever is right, we can also find | 5 ZHANG different de re desires from the psychological make-ups of the good and strong-willed people. Hence, it might be a feature of the morally good people that they have a wide range of both the relevant de re desires and the relevant de dicto desire. Let us consider David Copp's (1995, p. 212) objection as an example of this response.
Imagine that Dena is a good person who cares about the well-being, fair treatment, and the mental health of her friends and family. Dena can thus be thought of as having the previous de re desires, which also motivate her to behave morally. Let us then further imagine that, Dena gains an additional de dicto desire to do whatever is right whilst at the same time keeping the previous de re desires. The only difference here is that Dena begins to have an additional desire to do whatever is morally right. Copp thinks that there is no reason for us to regard Dena as someone who has a moral fetish, as Smith seems to suggest. Rather, Copp claims that, as Dena continues to have and be influenced by her de re desires, the additional de dicto desire should make no difference at all to how well we think of her as a moral agent.
In fact, following the previous argument, most externalists would agree with Smith that caring only about doing what is thought to be morally right would not be appropriate for a morally good person. Most externalists would also grant that, in order to count as a morally good person, an agent should have various direct moral concerns. For example, Svavarsdóttir (1999, p. 198) believes that 'a good person is also considerate, compassionate, kind, loyal, and honest' and these direct moral cares and concerns should be assumed to be stable. According to her, usually the moral actions themselves are enough to yield 'comfort, relief, or encouragement', and the result of this is that the moral agent will undertake those actions due to his or her direct concerns (Svavarsdóttir, 1999, p. 199).
What most externalists disagree with Smith about is whether they are committed to explaining the reliable connection between moral judgments and motivation with and only with the relevant de dicto desire. A moral fetishist could be argued to be someone whose only non-derivative desire is to do whatever is morally right. Such a person would not have any direct concerns for things such as honesty, kindness, loyalty, and compassion. In fact, all her desires for performing moral actions would derive from her desire for doing whatever is morally right.
However, according to the externalists, good and strong-willed people are not like this: they also have other direct concerns in addition to the relevant de dicto desire (Carbonell, 2013, p. 469-470;Copp, 1995Copp, , pp. 212-213, 1997Olson, 2002, p. 91;Svavarsdóttir, 1999, pp. 205-206). The relevant de dicto desire could, after all, coexist with many other direct de re desires in the agent's psychological make-up. Many externalists thus claim that explaining the reliable connection between moral judgments and motivation by referring to the relevant de dicto desire 'only commits them to maintaining the desire to be moral is a part of the motivational structure of the good person' (Svavarsdóttir, 1999, p. 199).

| WALL ACE ' S ACCOUNT OF A FE TIS H
According to Smith's understanding of the term 'fetish', a moral fetishist is someone whose only non-derivative desire is the de dicto desire to do whatever is right and, also, whose derivative desires to do the right things are based on that de dicto desire. Yet, as we just noticed, some externalists argue that having the de dicto desire to do whatever is right would not be sufficient to make a moral agent a moral fetishist because morally good people can have both de re desires to do things that are right and the de dicto desire to do whatever is right under that description. Instead of rejecting Smith's definition of 'a fetish', externalists thus deny that moral agents with both kinds of the relevant desires would be moral fetishists even according to Smith's own definition of a moral fetish.
In this section, I will introduce R. Jay Wallace's slightly different, and yet still equally plausible definition of a 'fetish', which allows us to understand what constitutes fetishism more generally, beyond the context in which the relevant de re and de dicto desires are discussed.
Here, I thus want to begin from a definition of a 'fetish', which was first introduced by R. Jay Wallace. According to him, fetishism in this context should be understood as 'the investment of interest and attention in objects that 6 | ZHANG are not intrinsically worthy of such responses' (Wallace, 2006, p. 195). This definition suggests that something is a fetish for an agent if the agent treats the object as more valuable than the object really is.
To illustrate the previous general definition of fetishism, let us consider an example of a shoe fetish. Imagine that Bob is a huge fan of shoes and one of his most favourite things is to collect different kinds of shoes. Many of the shoes in Bob's collection are limited editions that are worth a lot of money and all of them are fashionable, high-quality shoes. Even a person who does not consider shoes a lot would be impressed by Bob's collection. One slightly odd thing about Bob, however, is that, when being asked why he loves collecting different types of shoes so much, he cannot give a plausible reason for his hobby. Bob explains that he treats the shoes in the collection almost as a companion and he even develops an erotic interest in them.
The reason why we call Bob a shoe fetishist then is that he devotes attachment, love, and sexual interest in objects that do not deserve such responses. Usually, we believe that only humans-rather than shoes-deserve such devotion. Nonetheless, Bob chooses to love and to have an intimate relationship with his shoes instead of humans, which makes him a shoe fetishist. As Wallace's definition puts it, he is investing a certain kind of attention to objects that do not deserve it. Yet, notice that, here, we would continue to regard Bob as a shoe fetishist merely on the basis that he is trying to have an intimate relationship with his shoes. Even if at the same time, Bob also cares about other things, this fact would not make Bob any less of a fetishist. 2 It is essential to notice that Wallace's suggestion of what fetishism amounts to makes it irrelevant whether a moral agent has both the de dicto desire to do whatever is right and the relevant de re desires to do the right things at the same time. According to my reconstruction, the crucial point of the fetishism argument is not whether virtuous people act out of their direct concerns (such as their concerns for honesty and the well-being of their family).
Wallace's definition of fetishism enables us to see that the decisive aspect of the fetishism argument is whether an agent cares about the property of rightness, which some actions happen to have. If an agent who has both types of desires counted as a moral fetishist according to Wallace's definition merely in virtue of having the de dicto desire to do whatever is right, she would not be any less fetishistic because of her additional de re desires.
One advantage of this definition is that it is more general than Smith's-the introduced definition applies to both moral and other fetishists. With the help of Wallace's definition, we can not only explain why externalism would make moral agents moral fetishists, but also why individuals such as Bob count as shoe fetishists. As Wallace's definition is more general than Smith's, externalists no longer need to explain the reliable connection between moral judgments and motivation by relying solely on the relevant de dicto desire to do whatever is right.
Because of this, we can grant externalists that good and strong-willed people have both the relevant de re desires and the relevant de dicto desire as externalists argue. Thus, everyone in the debate, including externalists, should be able to accept Wallace's definition of a fetish.
Additionally, the new definition is also more explanatory than Smith's. According to Smith, the reason why moral agents in the externalist framework could be argued to be moral fetishists is that these agents are motivated by the de dicto desire to do whatever is right in a counterintuitive way. This criticism that is based on our moral intuitions is more descriptive than explanatory. It does not say too much about why we should think the externalist account to be counterintuitive. This gives externalists an opportunity to claim that their explanations of the reliable connection between moral judgments and motivation are actually just as plausible as the internalists'. In contrast, our new definition tells us more about the reason why the externalist account of how moral agents are motivated is fetishistic and, thus, Wallace's definition tells us more about why the externalist explanation of moral motivation is counterintuitive. When a moral agent is motivated by the de dicto desire to do whatever is right, he cares too much about the rightness itself. According to our new account of a fetish, when an agent cares too much about the rightness itself, he invests attention in objects that do not deserve it and thus becomes a moral fetishist. | 7

| DE DIC TO DE S IRE S ARE FE TIS HIS TI C
In this section, I will argue that, if an agent has the de dicto desire to do whatever happens to be right (under that description), she cares directly about something that is not intrinsically worth caring about. So, in the light of the new definition of fetishism discussed in the last section, having the de dicto desire to do whatever is right can be argued to be fetishistic. Thus, even if externalists were able to argue that the de re desires to do the right things coexist with the de dicto desire to do whatever is right among rational agents, this argument would not help externalists.
Let us then consider a very fundamental, intuitive view of what happens when we do moral deliberation. When we decide to act, at least at that moment, we believe that our investment and attention is needed for some good reason. We furthermore often assume that the things that motivate us to act are intrinsically worthwhile and because of this they give us practical reasons to act in the considered way. 3 Even externalists cannot deny these simple observations as they too would have to grant that acting for the sake of things that you do not take to be intrinsically worthwhile is unusual and odd.
As externalists always claim that an agent should at least in part be led by the de dicto desire to do whatever is right, an agent who has this desire should be assumed to care about the rightness of actions itself in some abstract sense. The agent should take rightness itself to be reason-providing and thus something that is intrinsically worthwhile to devote time and energy for. Yet, unfortunately, at this point, the externalist view becomes less plausible.
If the rightness of actions itself were reason-providing, then it could be argued that a virtuous person who acts for the right reasons would need to often choose to do certain actions because those actions are right. The problem is that the previous claim just does not seem to be acceptable.
To see why the rightness of actions itself is not reason-providing, we can consider the following basic idea about the reasons behind moral agents' actions. Most philosophers accept the following claim: the fact that certain actions are right is the same as the fact that properties of those actions provide us with good reasons to undertake those exact actions (Dancy, 2000;Stratton-Lake, 2002;Suikkanen, 2005). Thus, 'the reason why a good-willed person does an action, and the reason why the action is right, are the same' (Korsgaard, 1996, p. 60;Stratton-Lake, 2000, p. 16). The previous claim is based on the very natural thought that we are able to find out the reason why a given action is right by finding out why a virtuous person would choose to do that action. The reason that makes an action right should be identical to the reason that counts in favour of an agent performing that action. For example, when a kind and warm virtuous person helps a stranger in need, we will think that the person does a commendable action. If we then know why the virtuous person helps the stranger in need, then we will also know at the same time why the action in question is right.
If the previous compelling idea is true, then what externalists need here-that the rightness of actions itself is reason-providing-leads to an awkward result. Let us begin from the previous simple observation that a virtuous agent usually does a given action because she thinks that there are some good reasons for her to act in that way.
If externalists were right, then one of these reasons that would convince a virtuous person to act in a given case would have to be the fact that the act in question is right. Given the previous plausible thesis that the reasons why actions are right and why virtuous agents do them go hand-in-hand, it would in this situation follow that the fact that a given action is right would be one of the qualities of the action that would make it right. Yet, the fact that some action is right cannot make the action right-the right-making features of an action must be something different, and more basic than the rightness of the action itself. They must be considerations that explain why the action in question is right (for example, because the action saves lives or does not harm anyone), whereas the rightness of the action itself cannot provide such an explanation. 8 | ZHANG I have thus argued, in this section, that we cannot explain why an action is right simply by referring to the fact that the action is right. We then have reason to believe that the rightness of an action itself cannot be an appropriate reason for that right action. Since the rightness itself is unable to give us a proper reason for the action in question, it could be argued that the aim of doing whatever is right itself cannot be an intrinsically worthwhile goal itself. We know that, if an agent has a de dicto desire to do whatever is right, she must care about rightness itself-she must regard rightness itself as a fundamental value. Thus, an agent who is motivated, even in part, by the de dicto desire to do what is right actually cares about something that is not intrinsically worthwhile, given the argument above.
This further entails that, according to Wallace's definition of a 'fetish' which was discussed in Section 4, being motivated by the relevant de dicto desire would be a moral fetish, as an agent who would be motivated by this desire would care about something that is not intrinsically worthwhile. This is the case independently of whether the agent also has other moral de re desires in addition to her de dicto desire to do whatever is right. An agent can still be argued to be a moral fetishist if she is motivated by the de dicto desire to do whatever is right independently of whether she also has many other de re desires at the same time.
Finally, before the conclusion, there is one concern I want to address. In this section, I have shown that, if we accepted the externalist account of moral motivation based on the relevant de dicto desire, then good and strong-willed agents would generally be motivated by a moral fetish (that is understood in the way suggested by Wallace). However, the fetishism argument requires that there is also something bad about being motivated in this way. Furthermore, it cannot simply be assumed that having a fetish is bad or that the badness of fetishes is implicit in Wallace's definition because externalists might reject this assumption by arguing that ordinary fetishes are innocuous in many cases (such as the case of the shoe fetish in Section 4). This is why for the purposes of the argument we also need an explanation of why having a moral fetish would be a bad thing in the case of the externally motivated agents.
I believe that such an explanation can be provided based on Wallace's discussion of the general idea of a fetish. Wallace (2006, p. 195) points out that, if agents were generally motivated by a goal that is not worthwhile, for example, the de dicto desire to do whatever is right, they would be motivated by a goal that is 'arbitrary from the point of view of practical reason'. In this case, if an agent happened to have the de dicto desire to do whatever is right, it would seem mysterious to us why she ought to have the desire and why she ought to act on it. Moreover, the badness of being motivated by a fetishistic desire can be more explicit if we imagine that we were all motivated by the de dicto desire to do whatever is right. In that case, we could be argued to have a collective fetish-as a society, we would be investing a vast amount of emotional energy in a project that just would not merit that response. Yet, the reason why we would devote a huge amount of interest and attention to a project that is not worthwhile remains unjustified to us. This helps us to see the difference between the badness of the moral fetish and the innocuousness of other ordinary fetishes. It is easy to think that morality is a central and core project for us that governs and organises not only our many other projects but also our interactions with other people. In this situation, it would be bad if that project were not a worthwhile one grounded in our practical reason. In contrast, if some of our less central projects are for ends that do not merit such attention, this in itself is less worrying given that these projects are bound to play a smaller role in our lives. This is why I believe that we can explain the badness of having a moral fetish in the way that the fetishism argument requires.

| CON CLUS ION
In this paper, I have argued that de dicto desires are fetishistic. Even if an agent is partly motivated by such a desire, she would still be considered as a moral fetishist. I began by introducing the fetishism argument. Smith originally suggests that to be motivated by the de dicto desire to do whatever is right makes agents moral fetishists. Then, I discussed Copp's objection which claims that morally good people can have both the relevant de re and de dicto desires. As a response, I first introduced Wallace's account of a fetish, which allows us to discuss what constitutes a | 9 ZHANG fetish beyond the context in which the relevant de re and de dicto desires are discussed. Having a fetish now means that agents devote interest and attention to objects that do not intrinsically merit such responses. I then explained that to be motivated by the relevant de dicto desire means that rightness itself would be one right-making feature of an action. This consequence is absurd because the rightness of an action itself cannot explain why the action is right. An agent, who is motivated by the relevant de dicto desire, devotes interest and attention to a goal that is not worthwhile. Thus, the relevant de dicto desire is fetishistic and the moral agent is a moral fetishist.

ACK N OWLED G EM ENTS
Early versions of this article were presented at the University of Birmingham's philosophy postgraduate seminar and an open session of the 92nd Joint Session. Thanks to the audiences for their helpful feedback. I would like to thank in particular Jussi Suikkanen for his detailed comments on earlier drafts of this article. Research for this article was funded by China Scholarship Council, for which I am deeply thankful.