Let the ruler be the ruler: aiming at truth in Xunzi’s doctrine of the rectification of names

How should we understand the Confucian doctrine of the rectification of names (zhengming): what does it mean that an object’s name must be in accordance with its reality, and why does it matter? The aim of this paper is to answer this question by advocating a novel interpretation of the later Confucian, Xunzi’s account of the doctrine. Xunzi claims that sage-kings ascribe names and values to objects by convention, and since they are sages, they know the truth. When we misuse names, we are departing from a sagely convention of naming. As sagely convention determines moral truth, departure from the linguistic convention of the sages is a departure from moral truth. On my interpretation of Xunzi, the rectification of names is not a doctrine about what is true, but a doctrine about how we aim at truth. We are aiming at descriptive truth when our language conforms to the correct name of an object according to what I call ‘Confucian conventionalism’. When we correctly aim at descriptive truth we can aim at moral truth. Therefore, I claim that the doctrine of the rectification of names is concerned with discerning the literal accordance of language with an object (what is descriptively, linguistically true), to determine what is normatively, or morally, true. According to Xunzi, moral truth is grounded in linguistic truth.

If the ruler of Wei were to entrust you with governance of his state, what would be your first priority? And Kongzi replies: Most certainly, it would be to rectify names.
This statement, at 13.3 of the Analects, introduces what is commonly known as zhengming (正名), or the doctrine of the 'rectification of names': a principle that states that a thing's (usually an object's) name should be in accordance with its reality.
How should we understand the doctrine of the rectification of names? This question cannot be definitively answered, but we can say via the rectification of names, that the later Confucian Xunzi proposes a procedure for aiming at truth. We are aiming at truth, according to Xunzi, when our language conforms to the correct name of an object or thing, be that in the first instance or by later rectification of a false name. I acknowledge that while I think my interpretation of Xunzi's zhengming is broadly correct, my thesis is not about getting Xunzi's view exactly right, as I am analysing it with the methodology of an analytic Western tradition. My primary objective, therefore, is to develop the best analytic account of what Xunzi's zhengming means. I also happen to think that this is the best account of zhengming simpliciter.
To understand this procedure of aiming at truth, we must be clear about what it means for an object's name to be in accordance with its reality, and why accordance matters. We can delineate at least two different types of function for language: descriptive and normative. The descriptive function concerns how language extensionally engages with non-linguistic objects in the world. A descriptive theory of language should provide answers to questions such as how do names refer? When is a proposition true? The normative function concerns claims about values and morals, for instance, questions regarding how language should guide human action. I leave aside seemingly non-normative, prescriptive language, exemplified in utterances such as 'Walk the dog'. Xunzi's zhengming is a doctrine concerned with accordancing language with reality, and accordance with reality is achieved when our language aims at linguistic truth (the descriptive function), to thereby determine what is morally true (the normative function). This is because according to Xunzi (and Kongzi), moral truth is dependent on linguistic truth. Another way to put this, is to say that moral truth (the normative function) is grounded in linguistic truth (the descriptive function). 2 By 'grounded in', I am employing the notion here as a distinctively metaphysical, non-causal determination relation, where the grounds metaphysically explain the grounded. The grounded (morality) is explicable on the basis of its grounds (language and object). In other words, normative language (where it is expected to conform with descriptive truth) is contingent upon the descriptive use. So according to Xunzi, the doctrine of zhengming is an account of how we aim for normative truth by determining or rectifying its grounds in descriptive truth: aiming at linguistic truth is thereby the means to aim at moral truth.
To justify my interpretation of Xunzi, I will proceed as follows: because the normative account of the zhengming falls out of the descriptive, this entails an investigation of (1) how Xunzi understands the connection between language and objects in the world, before (2) how Xunzi views the connection between language and morality. I therefore divide this paper into three sections. In the first section, I explicate the doctrine of the rectification of names, particularly Xunzi's view. In 'Section 2', I introduce the notion of 'aiming at truth' and employ this to contrast two possible interpretations of the rectification of names. Firstly, that according a name with reality is a linguistic, correspondence picture-theory of language, whereby linguistic truth, and thereby moral truth, is discovered by the ancient sage-kings and conventionally dubbed. I call this view Confucian correspondence-theory. Secondly, that the conventions of these sage-kings preempt and determine what is linguistically true and thereby what is moral. I dub this view Confucian conventionalism. I argue that the second account is the better interpretation. In the third section, I will briefly discuss the mechanics of Xunzi's position on the connection between language, objects, and morality. Less controversially, I argue that the correct reading of the rectification of names is that only the 'legitimate' authority has the right to rectify names, and that this authority is the sage-kings. I conclude that Xunzi's doctrine of the rectification of names is a Confucian conventionalist account of aiming at truth.

On the rectification of names
The significance of zhengming to Confucianism is significant: if Kongzi had an opportunity to enact his philosophy, the first thing he would have done, without hesitation, is to rectify names. 3 13.3 is arguably not the only relevant passage; Kongzi additionally emphasizes the need for people to perform their social roles appropriately at 12.11, and this can be read as complimenting the doctrine. Likewise with passages 3.1 and 3.2, that concern the usurpations of ritual prerogatives and stress the need to have one's conduct mirror social status.
There is substantial translative, historical, and interpretive debate regarding the meaning and significance of 13.3. For instance, there is a vast corpus on how authoritative 13.3 is, what constitutes a Confucian authority, which interpretation is correct, and the role of time and apocrypha on these considerations (Nylan, 2001, 12, 20). Michael Nylan discusses these issues in detail, and she aptly captures the problem when she says that even in Kongzi's lifetime, his disciples fell into 'at least' four distinct groups (Nylan, 2001, 25). To make matters more difficult, according to Carine Defoort, an orthodox understanding of zhengming is taken for granted in 2 Page 4 of 19 contemporary analysis, and so many scholars are blind to the existence of competing interpretations (Defoort, 2021). As both Nylan and Defoort note, the debate regarding the correct interpretation of the rectification of names dates to the time of Kongzi but it was most recently revivified by Hu Shi (胡適, 1891-1962. 4 Carine Defoort holds that it is due to the influence of Hu (2003) that the zhengming is given primacy, and following Hu's lead, passages (such as 3.1, 3.2, 3.14, 4.15, 7.15, 15.3, and 16.4) that do not explicitly discuss names are (perhaps mistakenly) appealed to as auxiliary evidence (Defoort, 2021, 621). Given the number of tomes dedicated to this issue, I cannot do justice to the nuance of the debate regarding zhengming, nor can I decisively settle it. Such historiographical disputation, while of interest, is as prone to engender distraction from, as much as illumination of, the objective of Xunzi's rectification of names. Importantly, my case for the correct interpretation of zhengming is orthogonal to the issue of the preeminence of the doctrine. Additionally, plausible though Defoort's interpretation is, I do think there is a hint in the word 'first' that this doctrine is significant, albeit the degree to which it was significant for Kongzi and Xunzi may be inscrutable. As our interests are primarily concerned with Xunzi, we ought to take as authoritative what he took as authoritative. In this I am in good company, as according to Nylan 'the early followers of Confucius, after all, distinguished themselves from other groups more by the quality of their moral commitments than by their knowledge of old writings' (Nylan, 2001, 351). Thus, to paraphrase Xunzi, what matters most is not Xunzi's knowledge of the old writings nor perhaps my own, but what his doctrine means and can mean. This is not a historical investigation after all: the purpose of this paper is not to somehow grade the doctrine in a hierarchy of importance for Confucianism, but to consider how Xunzi's articulation might be understood as aiming at truth.
What then is the doctrine of the rectification of names? Defoort defines 'rectification' as a liberation from later acquired (aberrant) content (Defoort, 2021, 617). It seems that Kongzi held that to name aright was an epistemic means to grasp reality 4 In addition to highlighting the debate in Kongzi's lifetime about the meaning of the doctrine, Defoort places significant emphasis on the influence of the scholarship of Hu Shi (2003) (Defoort, 2021, 614). Defoort argues that the preeminence of zhengming is a modern preoccupation that owes its origination to Hu Shi. She cites as evidence the uptake in the number of works that treat zhengming in detail post Hu Shi: 'Rectifying names' (Hansen 1992: 65-71), 'Rectification of Names' (Hsiao 1979(Hsiao /1980Feng 1952Feng /1973Nivison 1999: 745-812, esp. 757-758), 'The Ordering of Names' (Hall and Ames 1987: 268-275), 'The Correct Use of Names' (Defoort 1997: 168-174), or 'Correcting Names' (Loy 2014: 146-152).
Defoort's point is well taken. Of course, my argument for the correct interpretation of zhengming has naught to do with the preeminence of the doctrine. The importance of my account of the doctrine to Confucianism is a further issue, one that I leave unexplored. For what it is worth, however, I think Defoort overstates her case for three reasons: i. Hu Shi is not the only scholar who has paid particular attention to zhengming across the millennia, notably this includes Xunzi himself; ii. the proliferation of interest may be nothing more than a correlation of an explosion of research across all academic fields in the modern era with greater preservation of documents and improved transmission of information in the twentieth and twenty-first centuries; and iii. Hu Shi and other scholars who do focus on zhengming are probably justified in doing so, given Kongzi's words about his first priority (or the words attributed to Kongzi by our available sources) at 13.3. aright: we cannot know what is true unless we can describe reality accurately. He explains it thus 5 : If names are not right then speech does not accord with things; if speech is not in accord with things, then affairs cannot be successful; when affairs are not successful, li (禮) [ritual/propriety/righteousness] and music do not flourish; when li and music do not flourish, then sanctions and punishments miss their mark; when sanctions and punishments miss their mark, the people have no place to set their hands and feet.
The thrust of Kongzi's view, as I understand it, is that if names are corrupted and abused, then we cannot correspond our knowledge with reality, and without correct names, social relationships would break down which in turn causes chaos and helplessness. As noted, Kongzi was not the only one to address the rectification of names. Xunzi-a later architect of Confucian philosophy-wrote an entire chapter dedicated to the subject: 'On the Rectification of Names', or 'On Correct Naming'. 6 'On Correct Naming' emphasizes how correct names are grasped by the inseparable relationship between language, understanding, and morality (Wang, 1989, 186-189). Xunzi agrees with Kongzi that if names are corrupted and abused, then we cannot correspond our knowledge and thereby our morals with reality. Like Kongzi, Xunzi is preoccupied with departure from moral truth as explained by departure from linguistic truth. Therefore, when Xunzi discusses regulation of names, he is thinking about how to reconstruct the ideal linguo-normative account of the sage-kings, who were great rulers that possessed the power and authority to name (Bo, 2009, 135). One can appreciate that Xunzi is primarily concerned with emphasizing the normative dimension to language, but this is somehow grounded in a descriptive theory of language that is contingent upon the sages. 7 Before we address the role of the sage-kings and the moral component further in section three, I want us to be clear about what Xunzi's theory of language is. I propose two plausible interpretations: Confucian correspondence-theory: a correspondence picturetheory of language whereby objects, including moral objects, are 'out-in-the-world', to be discovered by the sage-kings; and Confucian conventionalism: the conventions of the sage-kings determine what is linguistically true, and this grounds morality. I will show that Confucian conventionalism is the more plausible explanation.

Aiming at truth
First, let us get clear about my criteria for analysing Confucian correspondence-theory and Confucian conventionalism. I have posited that the rectification of names is not a doctrine about what is true, but a doctrine about how we 'aim at truth'. I adopt this phrase from Bernard Williams (1973). Williams was talking about the function of beliefs, but I think it is applicable to, what is here, the function of a theory of language. What then is the difference between giving an account of truth and aiming at truth? The general idea is this: when aiming at truth one can be agnostic about what it means for it to be the case that something is true. Aiming at truth is concerned with the function of discovering truth but there is no commitment to a particular atheist doctrine. One is talking about how one might come to know what is true, and one can aim at truth even if they are undecided on an exact theory of truth, much like how Columbus could sail for Asia without committing to exact coordinates. On the other hand, an account of truth is concerned about the nature and meaning of 'truth'. The notion is captured by the famous question of Pontius Pilate to Christ: 'What is truth?' (John 18:37-40: 37). An account of truth is committing to the nature, essence, and so on, of what it means for something to be true. A helpful way to think of aiming at truth for our approach is as a 'use based rather than truth based' theoretical approach, per Horwich (2010,. The case of 'belief' provides a helpful example to explain the notion of aiming at truth. According to Tim Crane, beliefs are mental, dispositional attitudes, that aim to provide true or correct representations of the world (Crane, 2016, 3). With notes of Burge (2010), Crane takes belief to be a way of guiding an agent through the world, like a kind of map. We can measure the correctness or accuracy of a belief (at least in principle) and say it aims at truth. Generally, the greater the degree of belief an agent attaches to P, the more likely an agent will act in a manner informed by P (Papineau, 2012, 89-90). Of course, we need a concept of truth as a condition by which to measure correctness and thereby aim at it. After all, if one has the mental state 'I believe that P', and an assertion is the expression of belief, then asserting 'I believe that P' is asserting one's commitment to the truth of P. Yet an agent need not have a settled account of truth from which to derive the concept of truth. Thus, I am treating zhengming as a methodology for aiming at truth in an analogous way to belief. Although discussion of Confucian correspondence-theory and Confucian conventionalism as theories of reference is concerned with the functional question, incidentally this will entail some further discussion of the meaning and content of 'truth'. 8 Let us return to zhengming. The first claim in Xunzi's rectification of names is uncontroversial. He says that historically the ancient sage-kings fixed the reference of names to objects. The sages said something like 'Use N to refer to O'. However, later generations distorted terminology, coined new vocabulary, and thus confused meaning, disabling them from distinguishing right from wrong 9 : Nowadays, the sage kings have passed away, and the preservation of these names has become lax. Strange words have arisen, the names and their corresponding objects are disordered, and the forms of right and wrong are unclear.
This appears to be a straightforward conventionalist account of language, with the requirement that the convention is forged by the authority of sage-kings to reliably guide action and make initial dubbings. For Xunzi, the rectification of names is achieved when the sage-kings choose names that are 'straightforward, simple, and do not conflict, then they are called good names'. 10 Therefore, insofar as the sagely naming reflects reality, then the naming is authoritative. Further evidence that the custom is established by the sage-kings, is that Xunzi calls this 'appropriate'. When a name later diverges from the conventional name, then it must be called 'inappropriate'. 11 Appropriateness here is not a pragmatic, social consideration, but conformity with the linguistic use of the sage-kings. I take this as evidence that for Xunzi, names have no necessary intrinsic object, so they are not bound to reality by any direct reference. They are bound by agreement from the sage-kings to dub an object thus: then the object becomes fixed, the custom is established, and it is called the name of the object (Knoblock, 1994, 130). Therefore, when a word correctly corresponds to the initial baptism of an object by the sage-kings it is a 'good name', and when one uses good names correctly in propositions about the world, they are aiming at truth.
From this information, I argue that the correct reconstruction of Xunzi's account of the rectification of names is as follows: (1) there are real objects (Xunzi, 2014, 373). These are the things sage-kings are naming. (2) There is no inherent, or a priori, connection between names and extensional objects to which they refer (Xunzi, 2014, 239). This is because sage-kings freely chose names. This may entail a posteriori necessary connections between names and extensional objects to which they refer, because we cannot rename objects without creating bad names (Xunzi, 2014, 294). (3) Names are labels attached to objects: they are baptismal, not discovered. Any name could have been attached to any object (but only by the sage-kings) (Xunzi, 2014, 237). (4) Choosing a name is contingent on the custom of the sagekings (Xunzi in Knoblock, 1994, 127). Sage-kings do, however, rely on their senseperception, which is shared by all humans, to choose names. (5) None may change a label (save perhaps a sage-king, but they are no more) because they are lacking in sagely authority (Xunzi, 2014, 293). Therefore, (6) a proposition aims at truth when a word is appropriately connected to a correctly named object-concrete, moral, or otherwise (Kongzi, 2015, 13.3;Xunzi, 2014, 236). Alice is aiming at truth when she says 'It is good to obey one's ruler' where 'ruler' refers to the true king rather than a usurper, for example. 10 Xunzi (trans. Eric L. Hutton) (2014), 239. 11 Xunzi (trans. Eric L. Hutton) (2014), 294: 'Names have no predetermined appropriateness. One forms agreement in order to name things. Once the agreement is set and has become custom, they are called "appropriate," and what differs from the agreed usage is called "inappropriate".'.

Confucian correspondence-theory
I have presented a conventionalist account of zhengming, but if the names the sagekings chose atomically correspond with reality, why not interpret this as a kind of discovery of extensional moral objects and a post ex facto codification of them (like the picture-theory account of language, even if the names have no intrinsic object)? (Wittgenstein, 1922, 6.373-6.522). 12 Alternatively, are the sage-kings dubbing a posteriori, rigid designations? (Kripke, 1981). 13 I think both interpretations can certainly be argued for and lend support to the correspondence view.
One might be sympathetic to the first account of naming, Confucian correspondence-theory, because Xunzi argues that names are social constructs that were created to help society function in an orderly manner. I take this as potential evidence for the correspondence picture-theory because the normative motivation comes before the initial dubbing of appropriate names. Correspondence theories take truth to be a relation between propositions and facts or states-of-affairs (McLeod, 2015, 6). We can interpret Xunzi as saying the sage-kings are searching for extensional moral objects, and their dubbings are at best reflections, or at worst approximations, of real (moral) objects. Arguably, even the sage-kings can misname if they use names that do not conform to the requisite criteria. On Confucian correspondence-theory sagekings may not always discover the correct extensional object of a reference, but this seems to count against conventionalism more than correspondence, as there must be some kind of truth to the object at which they are aiming. 14 Alexus McLeod, for instance, is convinced that the 'truth on offer' in Xunzi is a kind of correspondence, in which ming (fate, name, 命) are made true by their corresponding to shi (actuality, 實) (McLeod, 2015, Chapter 4, 83). On the correspondence-theory picture, the dubbings are only authoritative if they reflect reality: the authority of the sage-kings is downstream from the authority of reality. Certainly, it is difficult to deny that an urge to discover moral truth informs the sage-kings' linguistic enterprise, and this can be read as counting against the moral realism of convention. 12 We might give a definition of a correspondence-theory of truth as the following: x is true if x corresponds to some fact, or state of affairs, that obtains.
x is false if x does not correspond to some fact, or state of affairs, that obtains. Thus, the proposition 'snow is white' is true if there is a fact that, or state of affairs of, snow being white (Armstrong 1997). 13 As noted, there is striking similarity between some of the interpretations of Confucian philosophers and the linguistic theories of the twentieth century analytic tradition, in this case Saul Kripke (1981). 14 There is debate as to whether Xunzi's account is realist or not. See McLeod (2015, chapter 4, 83). On the one hand, Xunzi could be a realist: there are fixed, predetermined, natural patterns to which proper kind distinctions respond, and the sage-kings discern them. Alternatively, Chris Fraser explains how this realism conflicts with Xunzi's depiction of sage-kings 'imposing' orderly patterns on nature with language (Fraser, 2016). Chad Hansen (1992, 308) suggests that there are two Xunzis, on the one hand pragmatic, and on the other hand absolutist. This debate pertains more to the theory of truth than the functional notion of aiming at truth, as discussed earlier in section two. Therefore, I think this debate is too far afield for our concerns regarding whether or not language is a kind of correspondence or convention, as one might be a realist or non-realist on either account. However, the discussion highlights an important parallel debate about how Xunzi understood the linguistic project with which he was engaged. It is also of interest when considering how the sage-kings generate convention. I think the first riposte to the Confucian correspondence-theory interpretation lies in further analysis of Xunzi's account of the sage-kings. Xunzi says that when a name was established, later kings 15 : followed the Shang dynasty in the terminology of criminal law, the Zhou dynasty in the names of titles of rank and dignity, and the Rituals in the names of forms of culture. In applying various names to the myriad things, they followed the established customs and general definitions of the central Xia states.
The later kings did in fact follow a linguistic convention; an inherited system of names, which in turn is their inherited social, political, cultural, and moral system. For the later kings, there is no discussion of intrinsic moral appropriateness, only a binding of morals qua ritual and custom by agreement. They appear less interested in their propositions corresponding with a referent in the world, rather (for the later kings at least) the propositions have no referent until the sage-kings have fixed the referent by an appropriate dubbing. We cannot misname before the sage-kings have dubbed something with a name in the first instance. This is evidence that the convention of the sage-kings precedes any linguistic, and thereby moral, knowledge. For instance, Xunzi appealed to the 'classics' (jing) as the best route to the original teachings of the sage-kings to determine correct names, as transmitted by Confucius (Nylan, 2001, 11). 16 We must be careful not to swing the pendulum too far in the Confucian conventionalist direction, especially if one takes conventionalism to be a kind of relativism about linguistic or moral truth. I do not interpret Xunzi as a subjectivist about morality, nor about the reality of objects, as he seemingly intended his language to be taken literally (McLeod, 2015, 27). I agree with Alexus McLeod that he is at least a semi-realist (McLeod, 2015, Chapter 4, 83). One might even say Xunzi was a confused realist. There is of course a lively debate about whether Xunzi is constructivist, realist, relativist, or pragmatist about linguistic and/or moral truth. 17 Relativist interpretations such as Hagen's (2007) deny that for Xunzi there was a single concept of linguistic or moral truth, but he is also quick to point out that there can still be a true core of moral rites (Hagen, 2007, 113). This is a kind of pluralism about truth, where the predicate < is true > is different in different domains of discourse (McLeod, 2015, 8). Wong (2016) suggests that if Xunzi's theory is relativist, it is a highly constrained relativism: an absolutist theory that accepts variation in language and morality along two or three dimensions. 18 On the other hand, Fraser (2016) proposes that Xunzi's theory is an alternative to either naïve realism or 'unfettered 15 John Knoblock (1994), Xunzi: A Translation and Study of the Complete Works, Stanford University Press, vol. 3, vol. 3, 127. 16 Nylan discusses the nuance of jing in further detail. Jing has a homophonic connection with a word that means a straight path or direct route. So jing is a 'weaving', something that implies definition, order, and utility (Nylan, 2014, 11). 17 For an overview of this debate, see McLeod (2015). 18 Wong (2016) is discussing Xunzi's metaethical theory. However, as I have suggested the moral is grounded in the linguistic for Xunzi, what Wong says about the metaethical theory applies to the linguistic theory, for our purposes. relativism'. He argues that it excludes both a strong realist stance that kind distinctions exist in nature independently of human activity, and a strong relativist stance that kind distinctions are determined by human practices or conventions. Eno (1990, 146) is an advocate of said strong realism, while Bryan Van Norden advocates for a 'weak conventionalism' because it only applies to initial dubbing of the name but does not determine what the objects in the world are (Van Norden, 1993, 376.) Alternatively, Hagen (2007b, 443) claims that there is a compromise position that combines the constructivist and realist interpretations: Xunzi's conception of morality, and therefore language, involves both discovery and invention. Although Eric Hutton (2007, 446) thinks this is mistaken because it concedes too much to the realist interpretation. As we can see, the debate whether Xunzi was a constructivist, realist, relativist, or pragmatist is extensive, and also raises the question of whether he was a subjectivist. Most scholars shy away from declaring him a strong subjectivist. These considerations are important to keep in mind, but they do not bear directly on the debate between Confucian correspondence-theory and Confucian conventionalism, as we are discussing how the rectification of names aims at truth-whether the speaker expresses a belief about the truth of the world when they make assertions-not whether the doctrine should be classified as a kind of moral realism or not, particularly because Xunzi never explicitly discussed his theory of truth in the Xunzi (McLeod,Chapter 4,85). 19 My sympathies to a semi-realist position can be illuminated by reference to Crispin Wright's approach to the separability of truth and assertability conditions (Wright, 1984). Wright explains projective statements by saying that they 'are actually used not to state facts but rather to project various aspects of speakers' attitudes and affective responses' (Wright, 1984, 761). For our purposes, our projective statements are not used to state facts, but project some kind of psychological state that corresponds with the intentions of the sage-kings. 20 I take Xunzi to think that the reality represented by a name used by a sage-king is objective, even if the rectified name is merely conventional. Just as the rituals need to be based on the foundation of Dao (the way 道), the sage-kings' (and later rulers') names, though they can be arbitrary as designations, must correspond to reality. In other words, it is permissible for one to make up the word for 'reality', but they cannot make up reality: 'Names are that by which one defines different real objects.' 21 Indeed, Xunzi makes explicit appeals to sense-data in our approach to objects in the world, so it is unlikely that he would endorse a disjunct between naming and real objects. 22 The sage-kings are therefore aiming truth, even if we cannot know what their theory of truth was, and so I cannot take conventionalism to be a kind of relativism. 19 David Hall and Roger Ames reject the very notion that Chinese thinkers were concerned with truth in anything like the way Western philosophers are (Hall and Ames, 1997, chapter 6). This is further support for my position that we ought to talk about aiming at truth rather than theories of truth. 20 For an account of non-factualism about meaning ascription, see Kripke (1982). 21  A second reason to reject a Confucian correspondence-theory reading in favour of the conventionalist view is that the principle of one name corresponding with one object is untenable on the prior. For example, someone can be both a father and a son, a king and a parent, Hesperus and Phosphorus, Clark Kent and Superman. One person has two roles: there are two names for one object. This violates a condition of atomic one-to-one correspondence. One might try and save the picture-theory view by arguing a Russellian line that Confucian correspondence-theory does not demand one-to-one correspondence; a definite description might instead be a composite of many atomic parts. However, the fact is that there is no evidence of such an account being proposed by Xunzi. It is not far-fetched to think that if Xunzi had correspondence in mind, then he would readily encounter and resolve this distinction between a one-to-one atomic system and multiply realized alternatives. For Xunzi, names are relational, and the sage-kings determined them in accordance with social relationships and roles: they are interdependently defined. On the conventionalist view, a person can possess the li of a father, and the li of a son, and so on (Steinkraus, 1980, 262). The problem of multiple roles does not arise.

Confucian conventionalism
I have presented the case against Confucian correspondence-theory. I acknowledge that Confucian correspondence-theory is a plausible, if flawed, interpretation. As names are defined interdependently through custom and ritual, however, it is less tenable to claim a correspondence theory of language for Xunzi. The rectification of names, therefore, should be considered a doctrine of truth-aiming by linguistic convention. I have already given an account of Confucian conventionalism, but allow me to expand upon it and show why it is preferable. According to the conventionalist interpretation, real objects are dubbed, and this dubbing is fixed, but it is fixed with reference to prior dubbings of the context in which it is adopted. Interestingly, my interpretation of the rectification of names aligns strongly with Michael Devitt's causal-historical account of reference (Devitt, 1981a, b). 23 Devitt classifies a causal theory of designation as follows (Devitt, 1981a, 66) Devitt postulates that semantic (conventional) designation involves shared dispositions to use a name. 25 He says that 'The referential meaning of a description token is its reference-determining relation to the particular object that the speaker has in mind in using the description.' (Devitt, 2004, 282). Whosoever is not a sageking, cannot dub, so the referential meaning of a descriptive token is an application to a particular object in coordination with its dubbing by the sage-kings. What the speaker has in mind is the correct use of the descriptive token by the sage-kings. Of course, the veracity of such a theory would face the same philosophical challenges presented against Devitt's view. 26 The similarities illuminate conventionalism, but it would be a long bow to draw to say this is what Xunzi had in mind. Regardless, it helps us understand that sage-kings designate names that have no inherent connection to the extensional objects to which they refer. Thus, the names are baptisms at the discretion of the sage-kings and customs connected to them (Xunzi, 2014, 237). It follows from this, if it is the correct mechanism for naming, that none may change a label without sagely authority (Xunzi, 2014, 293).
Naturally, one must ask: why are the sage-kings capable of initially dubbing what is true? My answer may disappoint some readers: I don't know. I do not know because Xunzi never explained the reason beyond asserting their legitimacy, just as Kongzi had before him. My inference is that Xunzi and others took the time of the sage-kings to be a kind of religio-historical epoch beyond the empirical judgements or explanations of their contemporary times. 27 This may not satiate some modern, analytic hunger for a definitive answer, but Confucianism straddles the line between religion, philosophy, and politics, after all. This is further evidence for why I take zhengming to be about aiming at truth, not providing an erudite account of what truth is and how the dubbings of the sage-kings are objectively correct.
Finally, one might read Xunzi's theory as a kind of hybrid of causal and descriptivist theories of reference (Genone & Lombrozo, 2012). Indeed, Fraser (2016) raises the problem of determining the grounds for distinguishing similarity and difference among the referents of names for the doctrine of the rectification of names. I take this critique as evidence for a conventionalist interpretation, as a correspondence 25 Drawing upon Lewis (1969). 26 For example, complications arising from 'confused designation, designation change, descriptive names, and, relatedly, the important distinction between speaker, and conventional designation.' (Devitt, 2015, 110). Devitt (2021) outlines and critiques six factors that are commonly proposed to determine reference for these demonstratives: Explicit description, 'Slot' for demonstration, 'Slot' for perceptual cause, 'Slot' for an implicit description, 'Slot' for recognition, 'Slot' for intention. The problem is that any attempt to limit reference of complex demonstrations to a few factors seems to restrict their operation as demonstratives, and thus they are explanatorily insufficient. See also Reimer (1992, 190-191) for an argument that demonstrative utterances can refer to objects that are not the targets of the speaker's directing intentions. 27 Consider the analogous 'Apostolic Age' in the West, wherein miracles were deemed commonplace, but not expected to continue into the present day, by some accounts. Of course, different denominations debate the existence and extent of this alleged age. theory should not fall prey to this problem because the grounds would be determined by reference to the object. Xunzi says the causal basis for distinguishing these differences is the sense organs: shared sense organs are the basis for shared naming conventions, by which people can reach agreement in what they are talking about. This is further evidence of a causal-historical, conventionalist account of aiming at truth. This discussion parallels the debate between the constructivist, realist, and relativist interpretations. And yet, I note again that the objective of this paper is not to determine what 'truth' means in Confucianism, but to elucidate how Xunzi takes the doctrine of the rectification of names to aim at truth.
Per the above discussion, we can conclude that Xunzi's doctrine of the rectification of names is a Confucian conventionalist account. With this in mind, here is my final reconstruction of Xunzi's account of the rectification of names as a function for aiming at descriptive truth: (1) There are real objects (Xunzi, 2014, 373). (2) There is no inherent, or a priori, connection between names and extensional objects to which they refer (Xunzi, 2014, 239). There is a possibility of a posteriori necessary connections between names and extensional objects to which they refer (Xunzi, 2014, 294). (3) Names are labels attached to objects: they are baptismal, not discovered. Any name could have been attached to any object, but only by the sage-kings (Xunzi, 2014, 237). (4) Choosing a name is contingent on the custom of the sage-kings (Xunzi in Knoblock, 1994, 127). (5) None may change a label because they are lacking in sagely authority (Xunzi, 2014, 293). (6) A proposition therefore aims at truth when a word is connected to a correctly named object-concrete, moral, or otherwise (Kongzi, 2015, 13.3;Xunzi, 2014, 236).

The purpose of names
I have explicated Xunzi's doctrine of the rectification of names and proposed that a Confucian conventionalist account of aiming at truth in language is the proper reading. Let us now turn to Xunzi on moral truth. I contend that zhengming is a doctrine that aims at moral truth, which it does by grounding morality in language, per the Confucian conventionalist account of linguistic truth. As discussed, Xunzi seems committed to a kind of conventionalism concerning language: names are not intrinsically appropriate for the objects that they denote, but sagely custom and convention determines their usage. Once a name is dubbed and cemented by convention, we are mistaken to depart from this. And Xunzi seems deeply concerned with any disorder in naming, in truth-telling, in plain speech, in honest speech. Hence why delusions, disorder and departure from Dao are a direct consequence of incorrect naming. 28 For example, when we fail to insist that a father behave as a father should, or a king as a king should, then according to Xunzi not only do we fail to aim at linguistic truth, but we fail to aim at normative truth. Failure to aim at normative truth entails failure to uphold the good. This directly reinforces Kongzi, who said 29 : Let the ruler be ruler, the subject a subject; let the father be father, and the son a son.
According to Kongzi and Xunzi, names are the foundations of civilization, specifically harmonious civilization. According to Hu Shi, this conversation illustrates 'what Confucius considered the inseparable connection between intellectual disorder and moral perversity, between the failure to "rectify names" and the impossibility to establish moral laws and harmony of life' (Hu, 2003, 358). To illustrate the gravity of naming correctly for maintaining civilization, allow me an example. In Hans Christian Andersen's (1837) 'The Emperor's New Clothes', he tells the story of an emperor who is deceptively offered an outfit alleged to be invisible to anyone unfit for their position in society. The tailor gets to work, but in fact creates no such outfit. The emperor, wishing to observe the tailor's progress, but fearing that he will not be able to see the clothes and thereby reveal his own unfitness as emperor, sends his most trusted minister in his stead. The minister cannot see the robes, but fearing for his own station, keeps this fact quiet and reports back rapid sartorial progress. Eventually, the emperor decides to take a look at the outfit himself before it is completed, but of course, he cannot see it either. Fearing he is a fool or unfit to be emperor, he lies and declares the outfit beautiful. Eventually, the emperor parades through the streets adorned in his new clothes. Actually, he is stark naked. Yet none of his subjects admit that they cannot see any clothes, for fear of being accused of being stupid or unworthy of their own respective positions. In the end, a lone child exclaims that the emperor has nothing on! The spell is broken, and the illusion of the emperor's new clothes is dismantled, and all his subjects speak the truth.
If we are to interpret the moral of this tale in accordance with zhengming, we might take two approaches. On the first approach, we appreciate that when a society fails to adhere to the correct ritual of naming, it leads them astray; into chaos and the absurd. The emperor et al. failed to call the outfit by its proper name, and so they were not aiming at linguistic truth. By not aiming at linguistic truth, they could not aim for moral truth, because the moral claim is grounded in a proposition composed of false names that misconstrue reality. If reality is misconstrued, the proposition has no shot at being true, and thus the empire slips into moral absurdity. However, when names are rectified, then we can see that the emperor wears no clothes, and we can guide society back toward the correct Dao. 30 The second approach is that 28 Xunzi (trans. Eric L. Hutton) (2014), 293. For instance, Xunzi says at that time: 'None of their people dared rely on making up strange names so as to disorder the correct names, and so the people were honest and guileless.'. 29 Kongzi (trans. R. Eno) (2015), 12.11. 30 Of course, Xunzi's moral account did not go unchallenged, for he was not alone in considering the moral purpose of language. In chapter 26 of the Zhuangzhi, Zhuangzhi says: 'Words are for getting meaning; after one gets the meaning, one forgets the words. Where can I find people who have forgotten words, and have a word with them?' Zhuangzi (trans. Brook Ziporyn) (2009), Zhuangzhi: The Essential Writings, 112-117. This debate is also paralleled in contemporary analytic discussion of conceptual engi-this tale is a warning against truth as a conventional theory of language, and that it was the innate goodness of the child who correctly identified the truth, and that the others were led astray because of strict, inflexible custom and ritual, that detracted from the antediluvian goodness of humanity. The first approach is Xunzi's, and the second is another Confucian philosopher Mengzi's. For context, many philosophers take Xunzi to be arguing against Mengzi's position that humans are innately good, instead claiming that people require moulding through education and ritual (Fraser, 2018, 229). The followers of Mozi (墨子), the Mohists, thought we needed the correct kind of speech (yan 言) to go along with our Dao, and that how people use language (ming 命) was important, much like for Xunzi. The Mohists held names were insufficient for a correct philosophy without a connection to solid reality (shi 實). For instance, the 'Dialectical' chapters of the Mozi in some ways parallel Xunzi, as they are engaged in trying to fix the referents of terms, to bring order to discourse and argument (Fraser, 2018, 293-296). In at least this regard, Xunzi and the Mohists are similar. However, they were historically, vehemently opposed to Xunzi. This is because of their respective motivations. As we have seen, Xunzi is concerned with the conventionality of language. Xunzi wants to preserve Dao in the face of attacks from the Mohists, thinking their view to be dangerous and damaging to public discourse. It seems on Xunzi's view, preserving the Dao necessitated a defence of his account of the nature of language. One of Xunzi's objections to the Mohists after all, was that they misname. For example, the Mohists state that a robber is not a person, so that killing a robber is not killing a person. Xunzi would take this as a powerful example of how incorrect naming entails incorrect moralizing (Fraser, 2018, 236). This is why we should be interested in the doctrine, because to preserve Dao, we must acknowledge that there is a moral component to language, which in turn compliments Dao. The Mohist rebuttal would be that Xunzi and others may be deferring to rigid (albeit conventional) yet incorrect names. When we assign personhood to a robber, we have confused what personhood is, because of unthinking deference to the dubbing of the sage-kings.
The upshot of this debate between Xunzi and Mengzi, and more so Xunzi and the Mohists, is that it evidences Xunzi's position that the search for moral truth motivates the search for linguistic truth. And to aim at moral truth, one must have a language aligned with reality, which means rectifying names in accordance with the sage-kings. If we accept this interpretation of Xunzi's rectification of names, then one legitimate purpose of language, like that of the state itself, is to serve as the king's tool in propagating moral excellence 31 : When one who is a king determines names, if names are fixed and realities distinguished, if Dao is practiced and his intentions communicated, then he may cautiously lead the people and unify them by this means. 31 Xunzi (trans. Eric L. Hutton) (2014), Xunzi: The Complete Text,Princeton University Press,22.1c,128. neering and social metaphysics, as regards whether normative interests ought to motivate essentialist/ non-essentialist claims, or whether the metaphysics of concepts grounds the normative.
Footnote 30 (continued) The task of determining names belongs to the king (wangzhe 王者) (Nylan, 2001, 156). I interpret 'one who is a king' (wangzhe 王者) as designating not the person, who through the vicissitudes of life, happens to occupy the Dragon Throne, but someone who has cultivated the moral nature necessary for the office, and who thereby rules the world justly and by example. 'Leading and unifying the people' entails morally transforming the world, in line with Confucian values. The rectification of names then, allows the king to promulgate their enterprise, by making their moral objectives understandable to their subjects. This is done by correctly aligning our language with the meanings determined by the sage-kings, which in turn grounds moral truth. When we correctly say that the emperor wears no clothes, then we will not fall into moral error and chaos, but we can only say that emperor wears no clothes because we are interested in whether he ought to be wearing clothes or not in the first place.
We can appreciate the moral dimension to zhengming, and the example of the emperor's new clothes draws out the debate between Mengzi, Mozi, and Xunzi, regarding how we ought to aim at moral truth. There is a final concern I have with Xunzi's account, which is that it does not strictly explain how a new dubbing occurs. While a sage-king may determine the correct name of say a 'father', what about proper names such as 'Kripke', a person born after the epoch of sage-kings, so unnamed by them? One approach is to suggest that Kripke's name is grounded in a system of naming: when his parents named him it was in accordance with the system or mechanism developed by the sage-kings to dub new names. Alternatively, perhaps we must bite the bullet and say that 'Kripke' is not capable of being rectified. Xunzi provides little detail about how sage-kings perform dubbings and what their mechanism is, so this avenue is not promising. We might then concede that Xunzi did not consider this problem, and that the correct reconstruction of Xunzi's system does not provide a definitive answer. This returns to the issue of legitimacy in Xunzi, and his lack of explication about what confers legitimacy on the sage-kings, save they are being sage-kings. Although we cannot say exactly what Xunzi meant by 'legitimacy', nor how he accounts for new dubbings, this does not detract from the thesis that zhengming is a doctrine that aims at moral truth.
We have discussed how the rectification of names is a doctrine that aims at moral truth, and it takes moral truth as grounded in linguistic truth according to an authoritative Confucian conventionalism. Therefore, I interpret the mechanism for applying the rectification of names for Xunzi to be the following: This may strike us as an elitist, authoritarian account, without satisfying answers as to why the sage-kings have a monopoly on legitimacy, and yet this is the best analytic account of what Xunzi's zhengming means.
What is 'truth' in Confucianism according to Xunzi? This question cannot be answered, but we can say that via the rectification of names or zhengming, that Xunzi's Confucianism proposes a procedure for aiming at truth. Xunzi's theory of the rectification of names reveals the normative function of language, emphasizes the relationship between real objects and reference, and commits itself to guiding human behaviour through authoritative ritual and procedure. Descriptively, Xunzi advocates an authoritative conventionalist account of the connection between language and objects in the world. Normatively, morality is known and enforced by the descriptive rectification of names by the legitimate authority, the sage-kings. Thus, to aim at what is true, is to continually use names in accordance with the initial dubbing of the sage-kings, both descriptively and normatively. 32

Funding Open access funding provided by Central European University Private University
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http:// creat iveco mmons. org/ licen ses/ by/4. 0/.