Important Later Mohist Terminology

The terminology invoked in this article relies on translations of technical terms used by the Later Mohists themselves, such as “names and realities/objects/actualities” (ming shi 名實), and dang 當, the notion of what is appropriate, correct, or fitting. The usage of these terms is very specialized in the Later Mohist corpus. The idea behind pairing names (ming 名, linguistic phenomena) with realities (shi 實, conditions of the phenomenal world) in Later Mohist logic is that they each represent components that should be properly joined—like matching tallies—to reveal a standard way of linking the nominal and objective worlds. Ming represent linguistic, ideational formulations created, established, or realized through culture. Shi, on the other hand, are the material, concrete, and allegedly objective manifestations of things and activities in the phenomenal world.Footnote ¹⁶ As a technical, conceptual pair, “names and realities” gained traction in intellectual circles as early as the late fourth century b.c.e., forming the basis of most theories on disputation.Footnote ¹⁷

To link name and object together to express a truth-claim, Later Mohist authors often used the term, dang. Dang tells if a name–object relationship is correct, true, or fitting, or, as Saunders suggests, whether it points to the semantic adequacy between a word and the object it refers to in the world. Fraser goes further, to argue that dang functions like the phrase “… is true” in English, fulfilling “a central function of the truth predicate, namely to express, from the standpoint of the speaker, endorsement of an assertion.”Footnote ¹⁸ Rather than parse the ways in which dang is related to truth, I will use the classical Chinese notion of shi (“reality”) and translate dang as, “fitting the reality.”Footnote ¹⁹ In addition, I suggest that there was often something ethical associated with such a fit in early Chinese circles of debate. This is because the notion of what fits is often tantamount to what is appropriate or not.

A brief examination of the term dang in early Chinese texts reveals this ethical component. In the early third century c.e. dictionary, the Guangya, dang is equated with qi 齊 (“straight” or “even”). Typically, in Confucian ethics, the characteristic of being “straight” is ideal and moral whereas that which is xie 邪 (“slanted”) is perverse and immoral. Dang also means appropriate in the sense of “to match,” or “to suit,” which we see in Xunzi's comments about the gentleman, who excels in making sure “words necessarily fit with patterns and work necessarily fits the task” (言必當理，事必當務).Footnote ²⁰ This notion of matching up theory with practice is reminiscent of the Ruist enterprise of rectifying names, which dictates that human behaviors live up to the normative expectations that are thought to be implied by the names of their roles and positions.Footnote ²¹

The meaning of dang as, “to match,” can be extended to imply equality or equivalence, as in the following statement from the Liji: “The highest minister from a mid-level state has a rank that is equivalent to the middle rank in a large state. A mid-ranking minister has a rank equivalent to the lower rank, and a low-ranking minister has one equivalent to the rank of a high officer (in a large state)” (次國之上卿，位當大國之中，中當其下，下當其上大夫).Footnote ²² As for “suitable” and “appropriate” in an ethical sense, we can look to a quotation from the Mozi to find evidence: “These four kings had been suitably dyed (i.e., influenced), so they ruled as kings under Heaven, were established as Sons of Heaven, and their contributions and good name covered Heaven and Earth” (此四王者所染當，故王天下，立為天子，功名蔽天地).Footnote ²³ These examples show the close proximity between the mechanical act of matching on the one hand, and the ethical imperative of being a suitable or appropriate match on the other. It is precisely this semantic proximity, I think, that allows us to interpret Later Mohist uses of dang in an ethical, and not merely linguistic, sense.

The Later Mohists and Their Writings on Disputation

These chapters, while sometimes fragmented and exceedingly difficult to parse, provide us with a sense of the incredible amount of energy and time invested by the Later Mohists in their study of the meanings and uses of language.Footnote ²⁴ The nature of the writings on disputation are so specialized that one might surmise that they served as textbooks or handbooks for young men training specifically in the field of disputation.

Who were the Later Mohists? Pre-Han sources such as the later chapters of Zhuangzi and the Han Feizi mention a contemporary, derogatory name for these Later Mohists, allegedly used by Mohists of one school against another: “heretical/divergent Mohists” (bie Mo 別墨).Footnote ²⁵ The author in the Zhuangzi passage refers to well-established Mohist groups in his own time that had branched out from different forefathers after Mo Di founded the group.Footnote ²⁶ He also refers to the pressing issue of authenticity within a tradition and the rivalry each Mohist group had with each other. Each school wished to claim authority over the Mohist line of transmission stemming from Mo Di himself, and members of each school did so not by a heated ethical debate, but by debating among groups the most renowned sophistic topics of the day: the “hard and the white,” and “the same and the different.” This suggests that by approximately the late third century b.c.e., the definition of the Mohist school and its identification with the past were shaped by the activity of disputation as a means of claiming the authenticity, truth, and authority of the group.

In the “Xian xue” chapter of Han Feizi, the author informs us that Later Mohist schools were divided into three major branches, but only two of the branches he mentions correspond by name to the ones listed in the Zhuangzi.Footnote ²⁷ As in the Zhuangzi passage that mentions the “heretical Mohists,” one sees here a similar depiction of intra-Mohist disputes. The text's description of each group's self-identification as those who “truly follow Mozi” (zhen Mo 真墨) suggests that each group viewed the other as a rival pretender, and not as the true followers of the Mohist intellectual lineage.Footnote ²⁸ Such a formulation reveals what was at stake for the various Mohist branches: which group held privileged, exclusive access to the “authentic teachings” of Mozi? This of course begs the question: how did the Later Mohists decide who had access to the authentic teachings of Mozi, and on what basis was such a decision made?

The following passage from the Lüshi chunqiu shows the rifts among various Mohist groups and how such rifts were rooted in disagreements about disputation and the use of language:

This passage depicts Mohists in a complicated relationship to the activity of disputation.Footnote ³¹ We see one Mohist, Tang Guguo, making good of a stereotype of untrustworthy disputers in a successful attempt to undermine a fellow Mohist from a different state. The Later Mohist articulation of the importance of making and evaluating proper truth-claims, I think, should be viewed as a reaction to precisely the sort of skepticism towards language that is expressed in this passage.

At the crux of the Later Mohist dispute about authenticity, it seems, was a discussion of methods—in particular, methods of obtaining and proclaiming correct knowledge about the world. Of critical concern to us is the fact that language was at the center of this dispute over access to knowledge, since language was thought to be the tool whereby one might obtain knowledge in the first place.Footnote ³² So although the Later Mohists might have considered themselves very different from each other, it was not difficult for outsiders such as Han Feizi and the author in Zhuangzi to notice their similarities to each other, especially in terms of contentiousness and a fixation on how to use language in disputation.

What was the Later Mohist relationship to the six Mohist chapters on the topic of language and disputation? Of the recent scholars who have speculated on the authorship of these works, Hu Shi makes unequivocal claims that the six chapters were authored by the “heretical Mohists” (he calls them “Neo-Mohists”) mentioned in the “Tian xia” chapter of the Zhuangzi.Footnote ³³ But there are reasons one might question this claim. In the same passage of Zhuangzi mentioned above, the so-called “heretical Mohists” are said to have recited the Mohist “Canons.”Footnote ³⁴ While Hu contends that this reference to “Canons” refers to the writings representing the earlier, ethical doctrines of the Mohists, there is no evidence that this must be so.Footnote ³⁵ If this passage were referring to what is now the “Canons,” however, then the particular “heretical Mohists” mentioned by the Zhuangzi did not compose the “Canons” themselves, as they were treating it already as sacred word.Footnote ³⁶ In such a case, the heretical Mohists in question might have contributed to the creation of the “Explanations,” “Larger Pick” (Daqu) and “Smaller Pick” (Xiaoqu) (Chapters 43–46), but not to the “Canons” themselves. It is also possible that the “Larger Pick” and “Smaller Pick” represent fragments from either one specific group (written at different times) or several rivaling ones. The latter explanation would seem to account for the multiplicity of approaches and topics present in the two chapters (especially the “Larger Pick,” which is famously fragmentary and difficult to understand).Footnote ³⁷

Gongsun Long and the popularity of his debates on names (ming) can be dated approximately to the mid-third century b.c.e. We therefore have good reason to suspect that the Later Mohist preoccupation with similar issues arose and flourished both during and after this time. That some if not most Mohists from the third century b.c.e. onward interested themselves with notions of disputation and the analysis of language does not preclude their engagement in activities that preserved earlier doctrinal underpinnings, nor does it mean that the social, organizational structure of Mohist groups had necessarily changed. So while our late Warring States and early Han literature sources depict the Later Mohist groups as fractious, “heretical Mohists,” it could be the case that the so-called sharp divergences among Later Mohists were more superficial and sprang precisely from their shared use of disputation to make truth-claims about reality. As seen above, even contemporary outsiders were more likely to see their commonalities rather than divergences.

Some scholars assert that Mohist and especially Later Mohist interest in disputation and the powers of persuasive language can be linked to their allegedly low social status and lack of access to traditional forms of power.Footnote ³⁸ Others, like A. C. Graham, suggest that such interest was a product of the times, and that the fields of logic and dialectics were becoming increasingly popular towards the end of the Warring States period.Footnote ³⁹ Certainly, as we have just seen, this latter claim seems to have been true regardless of the class background of Mohist affiliates. Another motivation for the Later Mohist interest in language, I suggest, is their religious imperative to seek and clarify the ethical standards of truth imparted to us by Heaven's Will. Indeed, as I discuss below, their interest in disputation and language seems to have been linked to their earlier focus on ethics and religion, along with the imperative to reveal the correct Will of Heaven on Earth.

The Ethical Goals of Disputation

For the Later Mohists, the act of disputation was considered to be both lofty and fundamental. In general, the Later Mohists placed less explicit emphasis on the art of moral self-cultivation, which was celebrated in many other lineages of thought, replacing it with a concept of cultivating the self through unceasing education and inquisitiveness about the world. In other words, the Later Mohists emphasized the cultivation of an individual's mental—observational and knowledge-based—capacities. Distancing themselves from promoting an individual's active, moral cultivation through virtues and ritual behavior, or even the embodiment of an idealized, natural Dao, Later Mohists supported the obtainment and evaluation of knowledge about the world. They did this, I argue, perhaps not so much because they valued knowledge in and of itself, but because it was considered to be the means whereby humans could uncover and abide in Heaven's Will and Intent: an ethical system expressed through the phenomena and their interactions in the world.Footnote ⁴⁰

For the Early Mohists, ethical standards of behavior could be apprehended by intellectuals through close observations of such natural phenomena as droughts, bad harvests, strange and untimely weather, astral anomalies, etc. In the world of the Later Mohists, however, such knowledge was thought to be accessible through analysis of language usage. A key aspect of Later Mohist preoccupations with language, I contend, is that they thought there existed one “fitting (correct, true) use of language” in the world, one that adequately and truthfully represented an objective, outside reality. Through their proposed definitions (in their “Canons” and “Explanations”) and guidelines for the proper use of language through disputation (in the “Smaller Pick” and “Larger Pick”), the Later Mohists demonstrate how this might be insured in practice. The search for the one, true use of language, I argue, was ultimately an ethical enterprise, one that would establish a proper relationship between humans and the world around them.Footnote ⁴¹ The leap from establishing the proper use of language and taking it to be normative for human behavior is subtle and not often explained in their writings. Nonetheless, there are passages that clearly suggest this, and, furthermore, such an interpretation also helps justify the grounds for the Later Mohists’ avid interest in disputation.

The beginning of the short, Later Mohist treatise, the “Smaller Pick,” outlines various components and objectives of disputation. It idealizes the disputer's role, purpose, and goals, placing him in a lofty role similar to that of the sage, gentleman, or true man, described in other early texts:

Clearly, there is a necessity and nobility associated with disputation. Although the work of the disputer involves tasks such as organizing, arranging, comparing, investigating, clarifying, and marking, the outcomes and goals of such activities are on a much larger scale. The disputer disputes, ultimately, in order to help bring order to the world. If we look to earlier Mohist writings, this grandiose goal was first and foremost associated with the Will of Heaven. It is but a small step to extrapolate from the goal of the disputer, mentioned here, to what the Mohists believed was the fundamental goal of Heaven via humankind.Footnote ⁴³

Unlike many rhetoricians of ancient Greece who focused on the performative and stylistic aspects of oration, and who were often associated with a philosophical relativism,Footnote ⁴⁴ the Mohist disputer (most likely a man in ancient China, so I will use the pronoun “he”) was to occupy himself with the loftiest of goals: the search for and organization of knowledge according to ethical standards. Differentiating between what is correct and incorrect means being able to determine the very ethical standards necessary for evaluating truth-claims and resolving “suspicions and doubts.” In other words, the disputer was to be a philosopher of sorts: he was to discern the very ethical standards that he would express through disputation, and thereby gain the faith of his audience.

Another proposition for disputation or debate, found in the “Upper Canons,” is worth looking at to determine the extent to which it concerned ethical evaluation: “Debate: Vying with another. A winning debate is one that fits [the reality]” (辯，爭彼也。辯勝，當也).Footnote ⁴⁵ Our understanding of this passage hinges on the word, dang, and its connotations. One might argue that dang refers more strictly to the achievement of equality (a good fit or match) between the linguistic, ideational and the objective realm of the world. However, if the meaning is “straight,” as the term is glossed in the Guangya a few centuries later, then it seems that the Later Mohist authors were making a clear argument about the quality of winning debates: that they are winning precisely because they are straight—hinting at moral rectitude and the state of being upright. Even if one takes this dang more narrowly to describe the ideal connection between language and the world, it is difficult to deny the that an “appropriate” fit between the two realms is somehow not good in a weakly aesthetic or ethical sense.

In the Early Mohist core chapters, establishing clear standards for moral behavior is a foundational goal and primary human enterprise. Early Mohists make constant reference to Heaven as the basis for such moral standards, and, it is Heaven's Intent (or Heaven's Will) that serves as a model.Footnote ⁴⁶ They therefore approach the issue of deciphering Heaven's standards with utmost sincerity of purpose. Later Mohists, while in agreement with such goals, are not quite as direct in their approach. They retain their earlier definitions of Heaven's Will and a sense of justice, standards, and norms that derive from it, but they expand on what such standards and norms mean. In so doing, they suggest that we humans might come to understand what is ethically proper by simply understanding, uncovering, and establishing the correct relationship between language and reality, or objects and phenomena in the world. In other words, Heaven's moral standards can be deciphered through the activity of disputation. Understanding disputation as an act of revealing Heaven's moral standards helps us make sense of the claim at the beginning of the “Smaller Pick,” discussed above, that the disputer can ultimately establish good order in the world.

The introductory passage on disputation in the “Smaller Pick,” goes on to further clarify this belief. The author appeals to patterns behind names and realities, along with the claim that disputers organize “what is so of the myriad things,” pairing them up with what is comparable “in clusters of sayings.” But what constitutes a comparable, proper pairing of language and reality? The “Smaller Pick” states: “[The disputer] uses names to refer to objects, phrases to express ideas, explanations to express reasons, and [relationships of] kind to accept and propose [claims]” (夫辯者]以名舉實，以辭抒意，以說出故，以類取，以類予).Footnote ⁴⁷ Of the four devices of disputation mentioned here (names, phrases, explanations, and relationships of kind), at least two can help shed light on what a proper pairing between names and objects is, and how such pairings allow names to fulfill their basic, normative function in human life. In the sections that follow, I focus on two of these key devices: 1) “names and realities,” and 2) relationships of “kind” (lei 纇). A deeper understanding of these phrases will help demonstrate how the Later Mohists thought the correct discernment of language helps to establish an ethical order in the world. Our discussion will ultimately shed light on what is at stake—from a religious and ethical point of view—when the disputer adopts a proto-scientific goal of arranging and organizing “what is so of the myriad things.”Footnote ⁴⁸

Matching Names to Objects (Objective Realities)

As mentioned so far in the “Smaller Pick,” the goal of the disputer is to “observe the patterns behind names and objects” (察名實之理), just as he is to “use names to refer to objects” (以名舉實), during debate.Footnote ⁴⁹ Worth noting here is the fact that these so-called “patterns behind names and objects” are not created but observed in a manner indicating an objective distance. For this reason, I find it appropriate to translate shi not only as “objects” but as “objective realities” as well. Linguistic truth-claims are thus not so much set or constructed by the disputer as they are revealed in their true light. In this context, disputation is a means of using language to ascertain and reveal the intrinsic, Heavenly ordained patterns of objects that transcend the disputer himself. The disputer's job is to clarify what is naturally fitting or not based on what is intrinsic to both language and the world.

But what does it mean for us to observe “the patterns behind names and objective realities”? The Later Mohist “Explanations of the Canons, Upper,” provides a clue, but it also takes us down a rather complicated rabbit hole. In the “Explanations,” the Later Mohists define names and objective realities in the following manner: “That which is used to refer to something is a name. That which is referred to is the objective reality. When names and objective realities are paired, there is a match” (所以謂，名也；所謂，實也。名實耦，合也).Footnote ⁵⁰ This seems simple enough, but in actuality it is complicated and not without problems.

One concern is the fact that language—and, therefore, names—may point to many different objective realities. In the following passage from the “Larger Pick,” we find out what constitutes a correct pairing of names and objective realities, which we might call the correct use of language:

Here, the author points to many objective realities in this passage and only two kinds of descriptive names: “white” and “large.” In one case, the name “white” is a quality that, the Mohists claim, is stable and does not change when used to modify different objects. This type of label is based on a fixed or object-independent similarity. The name, “large,” on the other hand, serves as an example in which there are multiple realities for any given qualifier, “large.” There can be large stones, large houses, large frogs, and large people, but in none of these cases does the name “large” necessarily refer to the same, fixed reality (for it has no fixed size). The reality referred to by such a name is context or object-dependent, and so, there are many different realities for any given name, “large.” This type of label is based on an object-dependent or unfixed similarity.

In order to understand what “names and objective realities do not necessarily always match” means in this passage, we might look more closely at the example of the white, large rock. By “picking” or referring to a white rock vs. a large rock, a disputer must understand that a second or third white rock can be labelled white in the same sense that the current rock is white. But if he takes two rocks and examines them in terms of size instead of color, one may or may not be large in the same way as the other; in fact, one may be small in comparison. Precisely such a situation is an example of names and objective realities not necessarily always matching. A disputer must be carefully trained to clarify that “white” is a stable modifier while “large” is not; the objective reality that “large” describes changes according to the relative context of objects being described.

Why is it important for the disputer to know when a name refers to a fixed or variable reality? Because, if he is to understand “what is so of the myriad things,” he must be clear about the precise reality being referred to in any given utterance or claim. By understanding the nature of how specific names function, whether they might refer to singular/fixed or plural/variable realities, the disputer can uncover the presumed “objective meaning” behind each use of a name. This allows him to match the existence of a particular objective reality with a specific name, thereby securing the appropriate name–object relationship and the proper use of language for any given context.

As seen in the example of the white vs. large rock, names (white, large) point to patterns or similarities among different objective realities. The same names can bring different objects into relationship with each other, not just different rocks. Consider the following example from the “Larger Pick”:

The meaning of “circularity” and “not quite” in these statements is such that a formal congruence is emphasized, rather than the actual size or distance of a modified object. So, for example, whereas for the Later Mohists the name, “white,” points to the fixed qualities of “whiteness,” the qualities outlined by the names “circularity,” and of being “not quite” do not point to fixed, identical realities. Nor do they point to variable realities that are relative to objects evaluated comparatively with one another. Instead, they point to different objective realities that are congruent (tong 同) with each other in their being “circular” or “not quite.” There seem to be two distinct situations here: a new one in which the notion of congruent similarity is introduced (circles of different sizes), and one that reminds us of the modifier “large” in the previous example, where things are similar in their “largeness” but dissimilar in actual size. So, while all things that are circular may be congruent, they are not necessarily identical, just as distances that are “not quite” something point to similarly “not quite” yet not necessarily identical distances. One type of similarity refers to congruent similarity, the other to object-dependent (“unfixed,” or “relative”) similarity.

In these cases, making distinctions concerning names and objective realities exemplifies the disputer's project to “clarify areas of similarity and difference (tong yi 同異) in order to observe the patterns behind names and objective realities.”Footnote ⁵⁵ Understanding congruence and relative similarity is understanding a pattern of similarity. So by fathoming similarity and difference among names (modifiers, qualifiers, or descriptors), the disputer can uncover, identify, and accurately convey the quality of real phenomena in the world. In such a way, he can arrive at the basis for organizing knowledge and obtaining a clearer vision of the myriad things and their relationships in the world. To further explain: by understanding that “large” is completely object-dependent but “white” fixed and “circularity” congruent, the disputer can better evaluate the meaning behind a speaker's descriptions of situations and things. The disputer's intimate familiarity with the scope, limits, and key characteristics of certain types of names—i.e., whether a descriptive label is of fixed (object-independent), congruent, or unfixed (object-dependent) similarity, his understanding of the way names function in relationship to phenomena, and his perceptive observation of similarity and difference in the world allows him to organize reality according to the standards and models provided by language.

What is interesting about this Later Mohist orientation towards language is their non-skeptical attitude towards human perception of reality. For them, while the proper meaning and scope of language is ambiguous, one's perception of reality is not. This is why language, and not our human faculties of perception, is considered to be worthy of suspicion, study, and control. Later Mohist writings impart no sense that our own perceptions might in fact be shaped by our linguistic understanding of the world. Rather, the passages examined so far imply a certain universal determinacy of one's perception of reality while underscoring a tight, normative relationship between names and objective reality. They assume that even though everyone can perceive of reality in similar terms, not everyone uses language in such a way that it highlights and corresponds appropriately with such perception. Worse still, they contend, many people intentionally abuse language, thereby distorting reality to serve their own ends.

Through another passage on names in “Explanations to the Canons,” we might better appreciate these Later Mohist assumptions about one's natural perception of reality:

[By naming something] “thing,” you describe it [literally: arrive at it, or greet it, in a certain way]. There is a reality that necessarily awaits a [more specific] name. By naming it “horse,” you are categorizing it. As long as it corresponds to this object, you must use this name for it. By naming someone “Cang,” you are personalizing it. This name does not go beyond this reality.Footnote ⁵⁶

Here, “names” are explained in terms of their kinds: “categorizing (or generalizing), and personal.”Footnote ⁵⁷ Such names, while they are fundamentally based on what we might call a raw perception of such reality, are picked because they arrive at or greet an object in a certain way. In other words, names label, identify, organize our raw perceptions of reality.

In the example of a horse, the raw perception of a single animal-reality possesses a linguistic counterpart—“horse”—that defines and generalizes such an object beyond the individual instance of perception.Footnote ⁵⁸ This raw perception of an animal-reality necessarily serves as a standard model for the name “horse” in the first place and in all future accounts and comparisons of animal-realities with the name “horse.” Thus, by giving something the name, “horse,” we are categorizing a raw perception in terms of a general case, rather than naming it as an individual. On the other hand, the personal name, Cang, organizes our raw perceptions of reality in a different manner. While it can be assumed that some sort of raw perception precedes the naming of an object, whether we choose to associate this perception with the category of “human” or as Cang, the individual, depends on our choice of (our “picks” in) language. In either case, the perception of Cang or the horse does not appear to be a problem for the Later Mohists, who seem to presume that we are all seeing the same things or raw material. The critical area of discussion is with language and how it can change human understanding via misleading truth-claims, not with our raw perceptions of reality.

The passage from the “Explanations” above is important because it implies the objective existence of realities as thing-like entities and the direct linkage of different kinds of names to such entities. Objective realities await certain names to be picked and affixed to them; and so we may use names, such as “horse,” to test and affirm whether there is a good fit between the category “horse” and the animal-reality or entity that we perceive. In other words, the animal-reality “horse” has an objective reality that exists apart from human overlays of names, categories, descriptions, etc. Humans do not in any way create the reality of a horse by labeling it; they merely ascertain the proper type of linguistic category with which to reference and discuss it.

At some level, then, the Later Mohists believed that humans are able to naturally perceive of such entities as they objectively are. For this assumption to be true, we are required to know a priori the rough boundaries of an object—where it ends and where it begins—as though the object were naturally awaiting human recognition of its thingness. While language might help us decide if we will discuss an object in terms of its particularities and difference or general qualities and similarities (i.e., whether we discuss a “Cang” or a “human” when we see the reality that is Cang-human), the object we encounter is assumed to be present as a bounded entity. We might contrast this with current notions, stemming from semiotics and deconstructionism, that counter this position and assume that humans construct a reality through the very process of naming or constructing a sign for it.Footnote ⁵⁹ According to a typical deconstructionist position, we cannot think of our understandings of “reality” as possessing an objective existence outside of our human linguistic interpretation of it, as it is we who establish the very boundaries of what an object is via our naming of it.

So, how did the Later Mohists understand their concept of “names and objective realities”? We have seen how names are standards that help appropriately describe real objects and phenomena in the world. Of note is that an “appropriate description” points to an accurate picture of reality, one in which the term “accurate” possesses both a moral and physical aspect. To clarify: if a real object or phenomenon concurs with the name applied to it, then the Later Mohists would say that the objective and the linguistic realms overlapped and were dang (fitting or even), or he (合 matching). The physical sense of accuracy is reflected in the sense that something matches or is even with something else (like two parts of a tally), while the moral sense of accuracy is reflected in the sense that something fits appropriately with something else. Perhaps one can go so far as to say that when a name fits appropriately with an object, this constitutes the proper reflection of Heaven's objective truths. Judging from the description of the disputer's lofty goals at the beginning of the “Smaller Pick,” we are justified to argue that the disputer's occupation was nothing short of revealing, through the study and use of language, the divine truths of the world.

Though the Later Mohists have no single term for “truth,” many of their terms point to what we might call truths about the world around us. Consistent with their religious beliefs in the constant presence of Heaven's normative standards in the world, terms such as “Heaven's Intent” (tian yi 天意), “Heaven's Will” (tian zhi 天志), “justice” (yi 義), and “standards/norms” (fa 法), represent ontological and ethical truths rooted in the divine realm.Footnote ⁶⁰ Insofar as the Mohists believe that humans can decipher Heaven's Will by observing its signs throughout history and in everyday phenomena, we might understand these truths to be expressed as “realities” of the world. In other words, it is conceivable that for the Later Mohists, the objective truths of reality (corresponding to shi) were one and the same as the normative truths of Heaven. The disputer needed only find the proper language to match with such realities. Once he did, once he established such a match, he arrived at what we might understand as “truth-claims” (proper pairings of names and their objective realities) about the world.

Relationships of “Kind” or “Category” (lei 纇)

The last of the four methods of debate described prominently by the author of the “Smaller Pick” is the use of “[relationships of] kind to accept and propose [claims]” (以纇取以纇予).Footnote ⁶¹ Surely this is one of the most basic of all methods of Later Mohist disputation, as the authors dedicate the lion's share of discussions in the “Larger Pick” and “Smaller Pick” to the topic of kind.Footnote ⁶² They explicitly assert the importance of kind by saying, “Now, phrases proceed by means of kind. If one establishes phrases without being clear on their kind, then one will surely be in trouble” (夫辭以纇行者也. 立辭而不明於其纇則必困矣).Footnote ⁶³ The very ability of phrases to “proceed,” or, presumably, to attain some measure of persuasive flow and communicative function, is based on one's proper use of kind. But what does it mean to “establish phrases”? And how do Later Mohists relate kind to such phrases?

The notion of kind in Later Mohist writings is intimately linked to similarity and difference (tong yi 同異). In the “Explanations, Upper,” the authors speak of similarity in terms of what one does with it:Footnote ⁶⁴

Lei, when used verbally, can be translated as “to categorize.” By saying that the prerequisite for categorizing things is that they possess something by which they can be likened to each other, Later Mohists posit the existence of an objective state in which similarities might be apprehended. Precisely because there is objectivity in the matter, our perception of similarities is also unproblematic. Indeed, as discussed above, there is no indication in their writings that they consider the heart-mind or any human faculty to be responsible for creating similarities, only for recognizing and perceiving them. Thus, similarities and differences exist naturally in the external world; they are easily recognizable and perceptible in an object's traits, and they can be dealt with in a concrete manner by amassing such traits together and categorizing things accordingly.

The key to understanding the Later Mohist discussion of kind lies in accepting their assumptions that language use, and not reality perception, is the arena in which we must concentrate our attention. One's organization of language, not perception or objective reality, is corruptible and must therefore be studied and fixed in a normative fashion. In order to organize language, one must analyze typologies of kind. As we see in the following passage, this can be done through close observation, analysis, and comparison of the ways in which so-called similar things are the same, along with the ways in which they are ostensibly different:

The logic here may seem opaque to the modern reader, but it is possible to offer an explanation that makes sense. The basis for comparing things, like the attributes of humans, must involve considerations of when to invoke similarity and when to invoke difference. The level of desired comparison must also be clear. Does one compare two things based on whole types (i.e., the example of “human”—whether tall or short, “human” is a kind that shares certain features universally), or does one compare two things based on differences among parts (i.e., the example of “finger” vs. “head”—even though these are both human features, they are different because there are many features in a human body). The text's first example about invoking similarities among features in order to establish the kind or category, “human,” is helpful because it helps us determine whether or not, for example, a little person is a human. The latter example about invoking difference is helpful in keeping one from conflating or holding as equivalent any two parts of the same whole. In other words, even though objects may be parts of the same whole, it does not follow that they are similar in kind.

In seeking criteria for “human” or “not human,” one cannot simply compare the heights of two individuals and on that basis claim that one is not human. Height is not a defining feature in this case. Rather, as Later Mohists show, one must base the name “human” on similarities concerning the possession of certain shared features, and not on the relative differences among one feature. In other words, the fact that humans generally have two arms, a nose, two ears, etc. demonstrates their similarity in participating in the category of “humans,” while similarities or differences among isolated features such as one's height, the size of one's nose, etc., are not relevant to such a categorization. Little people, according to such logic, are humans, just as are really tall people.

So it is that the proper depiction of reality concerns more than one's raw perception of things; it also concerns making a typology based on boundaries of identification—similarity and difference—defined through language, as well as one's ability to organize and make comparisons using perceptual data based on such boundaries. In the passage above, for example, the name “human” is defined in terms of the existence of various features in a single body. Defining “human” as a composite of different features, and not in terms of a fixed form (short or tall), identifies the boundaries of what it means to be human. By scrutinizing phenomena according to correct linguistic categories, units, or levels of comparison, the disputer is able to use language to organize knowledge about reality and further clarify the relationship between both. This results in a refined sense of the precise definitions of names, the extent of their referential scope, and their relationships to other names according to kind.

Another way of clarifying realities through a refined sensibility to kind is by analyzing the form (not just the features) of an object, as defined through names:

“A grabbing sword and drawing sword are different. The sword gets its name by means of form and feature, and since these [two swords] are not of one form, [they are of a] different [kind]’ (將劍與挺劍異。劍，以形貌命者也，其形不一，故異).Footnote ⁶⁷ This passage should be read in light of the previous one on comparing a tall and short human with each other. The two cases differ from each other in that tall and short humans are to be considered the same while grabbing and drawing swords are to be considered different. Mohist logic concerning this determination revolves around the criteria used to define “human” and “sword” respectively. Since the latter definition is based both on “form” and “feature,” rather than just “feature” (as was the case for the name, “human”), a name that modifies “sword” so that it differentiates “sword” according to form serves to differentiate these swords in terms of a difference of kind. Similarly, since the definition of “human” relies merely on a similarity among different features (noses, eyes, arms, etc.) and not on form (giants, hunchbacks and little people are also human despite their different forms), names that modify “human” so as to point to different forms do not serve to differentiate humans in terms of kind. The passage on swords provides us with another example of how names standardize organization based on kind, so that one might rely on names to achieve a better understanding of the patterns behind objective reality.

Other passages also provide examples of how language, through its general and specific referential scope, helps us organize knowledge according to kind: “To conceptualize a pillar is not to conceptualize wood. You are conceptualizing the wood of this pillar. To conceptualize the finger of a person is not to conceptualize the person. [However,] to conceptualize one's prey is to conceptualize a beast” (意楹，非意木也，意是楹之木也。意指之人也，非意人也。意獲也，乃意禽也。志功，不可以相從也).Footnote ⁶⁸ What I am translating here as “to conceptualize” corresponds to the term yi 意, to have a mental picture of. Our yi (conceptualizations, ideas) correspond to certain names and not to others. They have a specific referential scope, one that varies according to the kind of object being specified. In the examples given to demonstrate this phenomenon, we encounter three different types of references to kind. Let us call them “compositional,” “constituent part,” and “overlapping.” These designate the kind of relationship between each of the given pairs of names: pillar/wood, finger/person, and prey/beast, respectively.

In the first example, that of the pillar and wood, when one conceptualizes a pillar (which we must assume was composed of wood in ancient China), its relationship to “wood” is one of composition. Wood is not a feature of pillar; it is what composes it. This is in contrast to the second example, that of the finger and person, in which the relationship described is one of constituent parts. A finger is a feature of a person, not what makes up a person. Therefore, it is not analogous to a person as a pillar is analogous to wood. If we were to create a properly analogous relationship of kind, we might say that wood is to a pillar as cells, water, or flesh are to a person.

In contrast to these two examples, the third example provides us a relationship of kind in which the former (prey) might be identified fully as of the same kind as the latter (beast). A prey is a beast, not something of which it is composed, and not a constituent part of being a beast. Since all prey are by definition also in the category, “beast,” and since the relationship between prey and beast is one of full belonging, it is valid to conceptualize a prey that is at the same time a beast.Footnote ⁶⁹ All of these examples demonstrate correct and incorrect uses of linguistic categories of kind that influence the way we think and make claims about reality.

Not only does kind affect the way we think about our perceived realities; there are clear moral repercussions to understanding reality according to the proper relationship of kind. We see this in the following statement, which describes human relationships normatively according to kind: “Righteousness is when valuable relatives do not act virtuously and you treat them according to their kind” 義厚親，不稱行而顧行.Footnote ⁷⁰ I take this to suggest that there is a continuum of virtuous and not virtuous behaviors, and despite the fact that people may be your valuable relatives, you should still treat them according to the kind of behavior they exhibit as humans. In other words, do not give preferential treatment to misbehaving relatives simply because they are your relatives. “Righteousness” only ensues when one expresses the appropriate kind of response to the behavior of others according to kind, as though there were a one-to-one relationship between type of behavior and proper moral response. One sees here how righteous responses cannot occur without a proper understanding of kind. In order to know how to behave in an appropriate, moral manner, one must be able to correctly apprehend and rank types of behaviors.

The passages above reveal some important Later Mohist presuppositions about language and its relationship to what they see as the objective world. Their statements show how one correctly organizes and parses objective reality by carefully examining, analyzing, and comparing names and objective realities according to similarities and differences in kind, scope, context, etc. For the Later Mohists, there are set rules concerning how things should be named and understood despite the many seeming possibilities.

Since the Later Mohists assume that there is an objective reality behind the names we use, language for them should never be completely arbitrary. By attempting to point to correct levels of comparison of scope and kind, the Later Mohists reveal a belief that language use can and should reveal objective understandings of reality. At the same time, their efforts to set definitions and proper ways of organizing and understanding names show that they were aware of how language might be used to make false, arbitrary, or misleading connections with reality. Later Mohists thus aimed to use language to produce what they thought were accurate reflections of natural entities, units, situations, and phenomena, not arbitrary designations that misleadingly label the manifold appearances in the world. False language, after all, leads to false understandings, claims, and conclusions about the world around us, which in turn lead to decisions about how we act in it.