Names, Actualities, and The Emergence

·期刊原文
Names, Actualities, and The Emergence of
Essentialist Theories of Naming in Classical Chinese Thought
By John Makeham
Philosophy East & West
V. 41 No. 3 (July, 1991)
pp. 341-363
Copyright 1991 by University of Hawaii Press

--------------------------------------------------------------------------------

p. 341

In the classical period of Chinese philosophy -- approximately 500 B.C. to 150 B.C. -- discussion of ming (names) and shi (objects/actualities) and their relationship was common to all the major schools of thought: the Confucians, the Legalists, the School of Names, the Daoists, and the Mohists. In this essay my main thesis is that by the late third century B.C., discussions of the ming-shi relationship evidence a shift from nominalist theories of naming to essentialist theories of naming. According to the former, it is the human being who arbitrarily or conventionally determines which ming should be applied to which shi; there is no proper or correct correspondence between a given ming and a given shi other than what has been artificially determined by man. According to the latter, there is a proper or correct correspondence between a given ming and a given shi, determined, variously, by what is ordained by tian ('Heaven') or by what is 'naturally so'/'so of itself (zi ran). If my interpretation is correct, it would provide the strongest evidence yet that there were thinkers in pre-Buddhist China who entertained realist philosophies.

In Section I, I will describe nominalist views of the ming-shi relationship as described in the Neo-Mohist summa and Xun Qing's (ca. 298-238) "Zheng ming" (On the Correct Use of Names) book (pian), these writings being the most fully developed nominalist theories of naming in classical Chinese thought. In Section II, I will proceed to describe the shift to essentialist theories of naming based on the evidence of late third- and second-century B.C. writings. In particular, although not exclusively, my discussion will focus on the "Xin shu shang" (Method of the Heart 'A'), "Xin shu xia" (Method of the Heart 'B'), and "Bai xin" (Purifying the Heart) books of Guan Zi, and the "Shen cha ming hao" (Make a Thoroughgoing Examination of Names and Appellations) book of Dong Zhongshu's (ca 179-104 B.C.) Chun qiu fan lu.

1. Nominalist Theories of Naming in the Neo-Mohist Summa and Xun Zi
The Neo-Mohists were the first group of thinkers to define ming and shi in the Explanation to Canon A 80:

That by which something is called is its ming; what is so called is a shi. [1]

The word ming encompasses two major semantic components: one, as a name for an entity or state of affairs, and two, as a name in the sense of reputation. Here it means a name in the first sense. The shi graph is composed of a roof with goods below. [2] The primary meaning of this 'full house' image is being 'full of, 'filled with', 'inner substantiality'. [3] This

p. 342

meaning is also implicit in the word fu, 'rich', 'wealth', which Xu Shen uses to gloss shi. He in turn glosses fu as bei, 'to be provided/endowed with'. [4] This meaning is again evident in the case of the word ri, 'sun', which Xu Shen glosses as shi, 'being filled in', as opposed to yue, 'moon', which he glosses as que, lacking', 'diminished'. [5] He also uses shi to gloss the word shi, 'room', conveying the idea that a room is that which is filled. [6] The term ding shi, which means 'the contents of a cauldron', is also a concrete employment of this meaning. [7]

From this primary meaning arose the extended meanings of 'replete', 'complete', 'solidness', 'substantiality', and 'filled out'. These meanings share the common sense of 'substantial manifestation'. Shi, meaning 'fruit', is derived from this sense of substantial manifestation. As an application of this sense of substantial manifestation, throughout the summa shi is used consistently to mean particular objects or entities. This meaning is made most evident in the Explanation to Canon A 78 where the Mohist distinguishes between universal (wu), generic (lei), and private (si) names:

'Thing' is the universal name; any object necessarily requires this name. Naming something 'horse' is an example of a generic name; one necessarily uses this name for those things which are like the object. Naming something 'Cang' is an example of a private name; this name is restricted to this object.

As Graham argues, the Neo-Mohists regard shi as particulars where "a common name is treated as an abbreviation of 'something which is like the object', the object being the particular for which the name is ordained" and "the function of common names is to be explained on purely nominalist grounds." [8]

The Mohist proposes that objects can be known on the basis of observation, explanation, or report (A 80). In other words, it is held that provided certain criteria are adhered to, there can be a correspondence between names and the objects which they 'pick out'. Yet although Canon A 80 lists objects (shi) among the four objects of knowledge -- names, objects, how to relate them, and how to act -- more attention is focused on names than objects. [9]

While initially the naming of an object is implicitly an arbitrary association, the distinctions represented by the man-made names are nevertheless considered to be real (B 68). Once a name has been established, it then serves to refer to that object:

Canon A 32
To speak is to emit references.
Explanation
To inform about this name is to refer to that object. Therefore 'saying' is an emitting of something's characteristics of which
any speaker is capable. To characterize something is like drawing a picture of a tiger, but it is in words. To say a
word, such as 'stone', is to communicate it. [10]

p. 343

Just as a picture may serve as an analogue used to refer to an object -- but is not the object itself -- so, too, names are words used to refer to objects. It is in this sense that names convey meaning. We use words such as 'stone' to communicate the object stone, just as we might communicate an actual tiger by drawing a tiger. [11] "It is by means of names that objects are picked out" (NO 11). [12] On what basis are they picked out? On the basis of similarity:

Canon A 31
To pick out is to present the analogue for the object. [13]

To this end a standard is first established:

Canon A 70
The standard is that in being like which something is so.
Explanation
The image (yi), the compasses, a circle, all three may serve as a standard.
Canon A 71
The criterion is that wherein it is so.
Explanation
Being 'so' is the characteristics being like the standard. [14]

There are three standards which may be appealed to in the construction of a circle: one's image of a circle, a compass, or an actual circle. It is interesting to speculate why the Mohist's first choice as a standard for a circle is an image. Could he be implying that the image of a circle is the most general or universal representation of a circle, even though "all three may serve as standard?" The import of such an interpretation is that here yi signifies a mental image or idea of a circle which has been abstracted from particulars. It would then be tempting to see in this interpretation evidence of universals in the Mohist summa. Somewhat surprisingly, Graham lends indirect support to the plausibility of such a Realist interpretation, [15] by tentatively choosing to follow Liu Chang's [16] proposal that the pu graph in the Explanations to A 58 and A 59 be read pu, "wood (or jade, or earth, or a man) in its crude state," thus yielding the following translation:

Canon A 58
Yuan (circular) is having the same lengths from the centre.
Explanation
The compasses draw it in the rough (?).
Canon A 59
Fang (square) is circuiting in four from a right angle (?).
Explanation
The carpenter's square shows it in the rough (?). [17]

Graham maintains that if Liu's identification is correct, "the meaning is surely that compasses and carpenter's square draw only rough approximations to the true circle and square" (my italics) and even cites the

p. 344

following Sophist's paradox from the Zhuang Zi [18] as further evidence of the plausibility of his interpretation of A 58 and A 59: "The carpenter's square is not square, the compasses cannot make a circle." [19] A tentative textual emendation and one interpretation of a contextless paradox, however, are insufficient to warrant a Realist interpretation of the Mohist's use of yi. Rather, in the summa, yi is best understood, following Hansen, as a term "used of images of memory or imagination." [20]

Returning to the Mohist's claim of being able to identify something as being analogous to an object or "being like an object" (A 78), it will be noted that he has not shown how defining characteristics and mere resemblances are to be differentiated. He does not, however, appear to have regarded this as problematical (assuming that he ever did address the question), presumably being satisfied that his definition of 'standard' -- "that in being like which something is so" -- was sufficient to make the critical distinction between defining characteristics and mere resemblances.

For the Mohist, names, not objects, are of primary importance because they represent distinctions that are then embodied in definitions. Definitions, which function as standards, serve as the final court of appeal for distinguishing between alternative claims as to what name 'picks out' what object. The Mohist's overriding concern is not that a name accurately and faithfully represent some object, but rather that the distinctions invested in that name be maintained.

This concern with maintaining nominal distinctions was shared by Xun Qing, as was his use of shi to mean particular objects. This understanding of shi is made evident in the distinction he draws between 'objects' (shi) and 'things' (wu).

There are things with the same shape but in different locations and things differing in shape but in the same location; these things are to be differentiated. Although those things with the same shape but in different locations can be combined, nevertheless they are said to be two objects. Those things whose shape transforms but still remain the same object are said to be transformed. To be transformed but not to be different is said to be the one object. It is in this way that we examine into objects and determine their number. [21]

An example of 'things with the same shape but in different locations' would be two horses. An example of 'things differing in shape but in the same location' would be a caterpillar which becomes a moth, or a foal which becomes an adult horse. For Xun Qing, wu is used to refer in a general [22] way to the thing which an object is (insect, horse, man, and so forth), while shi is used to refer to particular individual objects. [23]

For Xun Qing, the purpose of naming was to demarcate different objects:

A name has no intrinsic appropriateness; rather, the appropriateness of a particular name is demarcated by being ordained (ming, lit. 'to cause to be

p. 345

brought about by naming'). The demarcation [24] having been fixed and its custom established, then the name is called appropriate. Should a name then differ from custom, it is called inappropriate. A name has no intrinsically corresponding object; rather its corresponding object is demarcated by being ordained. [25] The demarcation having been fixed and its custom established, then it is called the object's name. [26]

This is the core of Xun Qing's nominalist theory of naming. Thus just as the heart is the seat of cognition and overseer of the senses, [27] so, too, the ruler determines what object a name demarcates. By ordaining a particular denomination the ruler establishes boundaries which serve to demarcate one object from another. Only then is a name made a matter of convention. The kingly prerogative to decide how objects should be tailored and the resulting distinctions fixed as names meant that for Xun Qing, consensus regarding a term's usage came after the king had decided what should be named and how.

Among modern linguists, the notion that language is a social convention is commonplace. Hansen argues that similarly, one of the four assumptions that were "implicit in classical [Chinese] thought about language," was "conventionalism," where:

the way of dividing reality into objects to be named (totally apart from what symbol or sounds the community uses) is also a function of common acceptance of a shared and conventional practice of classification or division... Naming is just making the distinctions, and the distinctions themselves are merely conventional -- socially agreed on ways of dividing up the world. [28]

Certainly it is true that Xun Qing recognized the essentially arbitrary and conventional nature of the practice of giving names to objects (for example, what is now called horse could equally have been called an ox). Nevertheless, it is evident that for Xun Qing "the way of dividing reality into objects to be named" was the prerogative of the ruler. "Common acceptance" and "conventional practice" were to be matters subsequent to the ruler's 'ordination' of a name, names being based on his conception, arbitrary or otherwise, of how objects should be differentiated, how the world should be cut up and named.

The following passages record his statement of the purpose of having names:

Thus in order to make distinctions, the wise men (sage kings) instituted names to point out objects. On the one hand noble and base are made clear, and on the other hand the same and the different are distinguished. [29] When noble and base are clearly differentiated and the same and the different distinguished, then there will be no trouble in (the ruler) conveying his intentions nor misfortune arising from being hampered and frustrated in his affairs. This is the purpose in having names. [30]

Names represent standards by which things that are similar may be grouped together and things that are different may be distinguished. The

p. 346

ultimate justification, however, is not philosophical but political:

Hence with regard to names instituted by the [sage] kings, the names having been fixed, then objects were distinguished, the Way was carried out and their will was known everywhere.... Now the sage kings are no more, the upholding of names is remiss, strange words have arisen, names and objects are confused and the shape of what 'is' and what 'is not" is unclear, thus even officials who maintain the laws and Confucians who recite the scriptures [31] are both confused. Should a [sage] king arise, he would certainly continue both to use old names and create new ones. [32]

For Xun Qing it was imperative that the sovereignty to ordain the appropriateness of a name for a given object be invested in the ruler. The names ordained by rulers were normative standards seen as fundamental to the implementation and maintenance of social, political, and ethical objectives and were not conceived as having any inherently appropriate applicability to the objects they named. As with the Mohists, Xun Qing's doctrine of names is based on purely nominalist principles.

II. Essentialist Theories of Naming in the Guan Zi and Chun Qiu Fan Lu
By the late third century B.C. Chinese philosophy had already entered its richest period of cross-fertilization, producing new, hybrid schools of thought. One of the products of this period was the emergence of essentialist theories of naming. The earliest of these theories are evident in late third-early second century syncretic writings, particularly Daoist and Huang-Lao-centered syncretic writings.

One such work is Guan Zi. [33] There are three books in Guan Zi that best evidence essentialist thought: "Xin shu shang," "Xin shu xia," and "Bai xin." Together with "Nei Ye" (Inner Cultivation), these books are now referred to by many Chinese and Japanese scholars homogeneously as the 'Four Books of Guan Zi'. [34] A host of theories has been advanced as to when and by whom these four books were written. The hypothesis advanced by Liu Jie and Guo Moruo that the four books are the writings of two scholars associated with the Academy, Song Xing and Yin Wen (both fl. the latter half of the fourth century B.C.), and their disciples [35] has been influential, although it is difficult to understand why, given the tendentious arguments advanced in its support. [36]

The objections raised by Feng Youlan and Machida Saburo to the spuriousness of the Liu-Guo hypothesis are wholly apposite. [37] Furthermore, Machida's thesis that the four books are not the product of some individual or group of thinkers associated with the Jixia Academy, but rather are the writings of a late Warring States (475-221 B.C.) or early Han (206 B.C.-A.D. 220) school of Daoist thought, is, I believe, the most tenable of the various theories on the dating and authorship of the chapters. [38] A decisive strength of Machida's theory is the plausible account he gives of Confucian and Legalist concepts being hierarchically

p. 347

brought together and subsumed under a Daoist mantle. [39] This type of syncretism is particularly characteristic of late third-century/early second-century philosophical writing. [40]

My exposition of this early essentialism begins with the following passage from "Bai xin":

Trace things back to their origins and determine what their shi (actualities) are; make one's foundation that which gave birth to things. If you want to know something's image, then you search its form; if you follow something's distinguishing marks, then you will come to know its essential qualities; if you search back to its starting point, then you will come to know its name. [41]

In this passage 'form' (xing) is paired with 'image' (xiang); 'distinguishing marks' (li) with 'essential qualities' (qing); and 'starting point' (duan) with name (ming). These pairings provide an important key as to how ming and shi should be understood. Relative to the concepts each is paired with, 'form', 'distinguishing marks', and 'starting point' are all manifest and apparent. For example, distinguishing marks are what lie on the surface of an object while its essential qualities are within and not immediately apparent. Again, the form of an object is its manifest shape while the image we have of it is not. 'Image' here refers to the mental image that one has of an entity. [42] Assumedly a mental picture or model is regarded as being a faithful reflection or representation of the essentials of a given object.

In the same way shi is essential, being the germ or that which is at the origin or core of things. This use of shi is clearly based on its primary meaning such as I have outlined earlier in this article: inner substantiality. To distinguish this use of shi from its conventional sense as 'object/entity', I translate it as 'actuality', understanding it to mean 'that without which an entity would not be what it is' or 'that by virtue of which an entity is what it is'. This meaning is in fact already evident in pre-Han literature, where shi is used synonymously with qing, 'the genuine', 'the essential', when contrasted with xing, 'external shape', or mao, 'visible features', or wei, 'false', 'artificial'. [43] (In this connection it is interesting to note that Graham describes qing as a very close approximation of Aristotelian 'essence' in contexts where it is translatable as 'what X genuinely is'. [44])

In the passage above, the relation between 'name' and 'starting point' follows the same pattern as the other pairs. 'Starting point' means the manifest or discernible beginnings of something. By searching back to a thing's 'starting point', one is able to determine its name. In other words, a thing's name, like its 'essential qualities', is inherent in that thing from its beginning or inception.

This interpretation is supported by the following "Xin shu shang" passage:

p. 348

Things inherently have form and forms inherently have names. If the name corresponds (dang), then such a person is called a sage. [45]

The correspondence referred to here is primarily the correspondence between name and actuality, the basis of the correspondence between an entity and its name. The 'correct' name of an entity, that is, the name that corresponds to the actuality (shi) of that entity, is revealed by the sage by tracing an entity's form back to its inherent actuality.

Thus, far from being understood as being conventionally or arbitrarily determined, 'correct' names are characterized as constituting part of an entity's basic make-up. This same thinking is found in the Mawangdui Cheng text [46]:

'Although the Way has no beginning, it does have Responsiveness'.

While a thing has still not come, do not possess it; after it has come, regard it as it is. When a thing is about to come, its form precedes it. If we establish it in accordance with its form and name it in accordance with its name, then how would we say it? [47]

This final question is rhetorical, as assumedly only a sage could answer it, since only he would be able to discern what name corresponds to a given entity's actuality.

When the appropriate name is used, then order prevails. Thus in "Xin shu xia" we read that:

In all cases, things come bearing names. In accordance with those names the sage tailors [48] things and so the world is well regulated. If actualities are not violated [by employing inappropriate names], there will be no disorder in the world and the world will be well regulated. [49]

Because an entity inherently possesses a name that corresponds to the actuality of that entity, then by correctly discerning what those names are, the sage is able to 'tailor' the world on the basis of 'correct' names.

The following passage from "Bai xin" describes how a sage or sage ruler apprehends what an object's correct name is:

Therefore [the method of] the sage ruler is to make his body tranquil and wait for things. When a thing comes, then he names it. [50] If correct names are employed then things will naturally be in good order; if perverted [51] names are employed then things will naturally be in a state of collapse. [52]

The Huai Nan Zi also reiterates this view:

For this reason, one who has realized the Way is neither sad nor happy, gay nor angry. When he sits he does not think and when he sleeps he does not dream. When things come he names them and when affairs present themselves, he responds. [53]

p. 349

[The sage ruler] does not like or dislike things because they are beautiful or ugly, nor is he pleased or angered by punishments and rewards. He lets each name name itself and each category categorize itself. Affairs proceed from what is so of themselves with no interference from him personally. [54]

By employing the technique of 'passive mindfulness' (yin) [55] the sage ruler's 'clairvoyance and illumination' (shen ming) [56] enable him to apprehend those names that are inherently appropriate to things.

To recapitulate, the conception of names and actualities revealed in the discussion above is premised on a distinction being made between entities and the actualities intrinsic to those entities. Entities are regarded as possessing inherently appropriate names, names which are as intrinsic as an entity's actuality. This understanding of names and actualities can be further elucidated with reference to the concept of "correlative thinking" where the correlation, correspondence, or bond between name and actuality is conceived of as a type of "mysterious resonance." [57] As a corollary to this bond, if the correct name of some entity is not apprehended then the actuality of that entity will not be correctly discerned. And if the correct name for the corresponding actuality is not employed, "then things will naturally be in a state of collapse." [58]

The notion of a correct or true name has a close parallel in classical Greek thought. Leonard Woodbury writes:

It is assumed, in this way of thinking, that there always exists a true name, by which, if we can find it, the truth about things will be revealed.... The true name, which has the power of exposing the revelation of truth, has this power because it was "rightly" given to the object that it signifies. The Greek vocabulary is held to be the product of a name giver and, as in the case of the other arts, the original and authoritative practitioner of the art is held to be divine.... Once the meaning of the true name has been apprehended, all other names, which give or appear to give different information, must be either accommodated to the true name of [sic: 'or'] rejected as false and deceptive. [59]

The Chinese theory of names that I have outlined above shares some obvious parallels with this Greek view. This parallelism is also reflected in popular etymologies where "the meaning of a particular word is disclosed by the perception of its likeness to another word or of its identity with that word, whereby its function as a name is revealed." [60] Dong Zhongshu provides an apt example of this in the following passage from "Shen cha ming hao":

If we make a thoroughgoing examination of the overall meaning of the appellation 'king' (*giwang/wang [61]), we find that it may be divided into five divisions: the august (*g'wang/huang [62]) division, the upright (*piwang/fang [63]) division, the all encompassing (*k'iwang/kuang [64]) division, the central (*g'wang/

p. 350

huang [65]) division and the orientation (*giwang/wang [66]) division. When these five divisions are united and referred to by one word, that word is 'king'. A king is august, upright, all encompassing, central, and the focus of orientation. Therefore if a king's intentions are not universal and [67] august, then his way cannot be correct and upright. If his way cannot be correct and upright, then his potency cannot be universally encompassing. If his potency cannot be universally encompassing then his goodness cannot assume a central position. If his goodness cannot assume a central position, then the four directions will be unable to orient themselves towards him. If the four directions are unable to orient themselves towards him, then as king, he will be incomplete. [68]

In this passage, it is the paronomastic correlation between the key words august, upright, all encompassing, center, and orientation which lends definition to the meaning of wang, 'king'. Using these and other correlations Dong Zhongshu expounds his essentialist theory of naming. This theory of naming, as elucidated principally in "Shen cha ming hao," is centered on three main propositions:

1. That it is Heaven's will (tian yi) which is the ultimate source of all 'correct' names. [69]
2. That names are the criterion for deciding 'what is' (shi) and 'what is not' (fei) the case.
3. That only sages or sage rulers are capable of apprehending which name is intrinsically appropriate to a given actuality.

It is the first proposition that is the most novel and intriguing. Like Confucius and Xun Qing, Dong Zhongshu maintained that names were essential to good sociopolitical order:

The starting point in putting a state in good order lies in putting names in Order. [70]

Where he differs significantly from Confucius and Xun Qing, however, is in proposing that names should be ordered so as to conform with Heaven's will or intentions:

The [standard of] correctness for names and appellations is found in Heaven-and-Earth; Heaven-and-Earth provide the ultimate correctness for names.... [71]

Although names and appellations have different sounds, yet their basis is the same; both are cries and calls which serve to give expression to Heaven's will.... Names are that whereby the sage promulgates Heaven's Will. [72]

According to the standard interpretation, Dong Zhongshu perceived names to be ontologically more fundamental than actualities: actualities rely on names for their identity. Feng Youlan, for example, came to adopt this view in his revised work, Zhongguo zhexueshi xinbian:

p. 351

[Dong Zhongshu] maintained that actualities must conform to names, and that names should be employed to rectify actualities. He attached to names a primary [ontological] status and to actualities, a secondary [ontological] status. [73]

Ron Jiyu similarly comments:

[Dong Zhongshu] turned the relationship between names and actualities on its head, such that names were more fundamental than actualities; names were primary while actualities were secondary. [74]

It is my contention that proponents of the standard interpretation above are mistaken, principally because they misinterpret the following passage from "Shen cha ming hao":

Names are born of 'the genuine' (zhen); if it is not the genuine, then it cannot be used to make a name. Names are that whereby the sage affirms what is the genuine in things. Names are so as to be able to speak of the genuine. [75]

According to the standard interpretation, 'the genuine' does not refer to an entity or its actuality, but rather to 'Heaven's will'. Fang Keli, for example, says:

In fact, what [Dong Zhongshu] calls 'the genuine' does not refer to a thing, but rather refers to Heaven or Heaven's will. "Names are born of the genuine," and "to name things like their 'genuineness'" means that the names of things must felicitously express Heaven's will. In other words, Heaven's will alone is the most real component of a name. [76]

The fallacy of this interpretation derives from the failure to recognize that while ultimately the 'genuine' is indeed a manifestation of Heaven's will, yet more immediately and significantly, the 'genuine' is the embodiment of Heaven's will in a particular entity as that entity's actuality. [77] It is the failure to make this distinction that leads exponents of the standard interpretation to argue that names are seen to be more fundamental than actualities. In the following passage Dong Zhongshu makes it plain that by 'the genuine', he is referring to an entity's actuality:

The genuine is a thing's meaningfulness (yi) [78]; the genuine is a thing's essential qualities. [The sage] thus uses the genuin to name the thing. [79]

In defining 'the genuine' as being an entity's essential qualities and that which is meaning-bestowing, Dong Zhongshu discloses that by 'the genuine' he is in fact talking about an entity's actuality; in doing so he also draws a distinction between an entity and its inherent actuality: entities embody actualities while actualities are those essential qualities which make entities what they are. The distinction is in turn intimately related to his theory of human nature:

p. 352

Some say, "Human nature contains the germ of goodness and the heart-and-mind contains the basic stuff of goodness, so how could it not be good?"

To this I reply, "This is not so. Although the silk cocoon contains silk yet it is not silk, and although an egg contains a chicken, yet it is not a chicken...." [80]

Goodness is like a kernel of rice; human nature is like a stalk of rice. Although a stalk of rice produces kernels of rice, yet it may not be called a kernel of rice. Similarly, although human nature produces goodness, yet it may not be called goodness. Kernels of rice and goodness are the external completion of that which is inherited from Heaven; they do not fall within the realm of that which is done by Heaven. That which is done by Heaven extends to a certain point and then stops. That which lies within this realm is referred to as belonging to Heaven; that which lies outside is referred to as belonging to the 'kingly teachings'. Although the kingly teachings lie outside human nature, yet human nature must advance in that direction. [81]

Human nature is here conceived to be "the natural capacity one is born with," one's "basic stuff" (zhi). [82] It is man's latent capacity to be good that is his nature, and it is this nature that is Heaven-given and so is one aspect of man's 'actuality'.

The relationship between man's actuality and his name is more fully revealed in the following example:

In accordance with that which Heaven has made, [the sage] gives appellations. Thus people are called 'people' (*mien/min [83]), because min serves to verbalize that people are inherently the same as 'being with the eyes closed' (mien/mian [84]). [85]

The actuality or essential quality of 'people' is said to be their state of being asleep. While people have a Heaven-given latent disposition to develop, it is only after they have been educated that they are truly awakened and their potentials are capable of being fully realized. In short, the name 'people' (min) refers to those beings whose actuality is their state of not yet having been awakened (mian). Thus elsewhere Dong Zhongshu writes: "Names are the actualities of natures and actualities are the basic stuff of natures." [86] For the sage, conforming with the will of Heaven involves coining a name in which the meaning 'not yet awakened' is implicit. This name is min.

Accordingly, rather than portraying names as being determinates of actualities or more primary than actualities, Dong Zhongshu presents quite a different picture: names come from actualities.

Investigate into actualities to make names ... [87]

All things come into existence carrying their own name; the sage names a thing in accordance with its image. [88]

p. 353

Each entity, upon its creation, has certain identifiable characteristics which are an expression of that entity's unique actuality. The subsequent image a sage forms of these characteristics provides the basis for naming that entity. Thus the notion that things carry their own name should not be construed too literally. What it means rather is that each entity is sufficiently unique to be able to be given a name that truly represents the qualities of that entity. Ultimately, however, it is Heaven which determines what that name will be:

Each affair accords with its name and each name accords with Heaven. [89]

The manifestation of Heaven's will in a given entity is that which is genuine to an entity. Thus, to accord with Heaven's will requires that the name giver, the sage, apprehend what is the genuine in a given entity. In apprehending what is the genuine, he is in fact apprehending Heaven's will. It is on this basis that the sage proceeds to name.

That at least is the theory; in practice, however, it would have been totally at the sage's discretion to determine how things should be named since only he was capable of 'apprehending' what a particular entity's actuality was. In this respect Dong Zhongshu's doctrine of names is clearly a refinement of Confucian zheng ming political philosophy, where the authority to coin names and prescribe what they should be applied to rests with the sage ruler. Where he differs from earlier Confucians, and indeed also from the Mohists, is that the ontological and epistemological basis of his views on naming is essentialist.

With respect to the name-actuality relationship, Dong Zhongshu's essentialist theory of naming differs from the one adumbrated in Guan Zi. This difference evidences both a continuation and modification of the earlier, more rudimentary theory. In Guan Zi, that which makes a name appropriate to a particular entity is the bond it has with that entity's actuality. There is no notion of names being born of actualities; rather names and actualities are like two sides of the same coin with neither having ontological precedence over the other. As to why certain names correspond with certain actualities, there seems to be no other reason than because this is the way things naturally are (zi ran); names are contrived neither by Heaven nor by man.

For Dong Zhongshu, however, it is Heaven that ultimately determines the appropriateness of names. Initially Heaven's will is made manifest as actualities, and then these actualities produce names. [90] The actual coining of names, however, is left to the sage or sage king. The same principle that applies to kernels of rice and human nature applies analogously to names: "That which is done by Heaven extends to a certain point and then stops. That which lies outside is referred to as kingly teachings." Thus, while in Guan Zi the sage apprehends correct names, in the Chun qiu fan lu the sage also coins names. In both cases, the sage plays the

p. 354

role of 'midwife', but only in Chun qiu fan lu is it evident that he also chooses the particular words to be used in naming.

Nevertheless, both views are patently essentialist. It is a matter of speculation as to why rudimentary essentialist theories of naming began to emerge in late third-century B.C. syncretic writings. Could this emergent realism be a reflection of philosophical influences derived from Chu thinkers? Whatever the reason, the essentialism of these works stands in marked contrast to the nominalist theories of naming expounded by the Neo-Mohists and Xun Qing and eventually became the mainstay of Han philosophy of language, [91] a succinct expression of which is to be found in Liu Xi's (fl. 200) Preface to Shi ming: "In the correspondence between name and actuality, there is in each instance, that which is right and proper." [92] This is certainly a far cry from Xun Qing's doctrine that "a name has no intrinsically corresponding object; rather its corresponding object is demarcated by being ordained."

NOTES
1. In referring to the six dialectical chapters from the Mo Zi (chapters 40-45) I have followed Graham's arrangement and numbering of the Canons and Explanations. Where I have followed a different textual reading, I have indicated so in the appropriate footnote. On the estimated dating of the six dialectical chapters that are identified as the Neo-Mohist summa, A. C. Graham, later Mohist Logic, Ethics and Science (Hong Kong: Chinese University Press, 1978), pp. 23-24; and Watanabe Takashi, "Bokushi shohen no choshaku nendai," Kodai Chugoku no kenkyu (Tokyo: Sobunsha, 1973), pp. 533-534, and 545, date the earliest sections circa late fourth century B.C. and continuing on until the late third century B.C.

2. See Xu Shen's (ca. A.D. 55 - ca. 149) gloss in Shuo wen jie zi (Bering: Zhonghua Shuju, 1979), p. 150, 7B.4b.

3. On this interpretation, see also Kasahara Chuji, "Saden ni arawareta 'jitsu' ji no kenkyu," Ritsumeikan bungaku 157, no. 6 (1958): 8; Chugokujin no shizenkan to bi ishiki (Tokyo: Sobunsha, 1982), p. 106.

4. Shuo wen jie zi, p. 150, 7A.4b.

5. Shuo wen jie zi, p. 137, 7A.1a; p. 141, 8b.

6. Shuo wen jie zi, p. 150, 7B.3a.

7. See Book of Changes, Ding hexagram, Line 2 and Line Text Commentary; Shirakawa Shizuka, Jito (Tokyo: Heibonsha, 1984), p. 390.

p. 355

8. Graham, Mohist Logic, p. 325.

9. Feng Youlan, Zhongguo zhexueshi xinbian (Beijing: Renmin Chubanshe, 1964), Vol. 1, p. 413, has noted the primacy of names as an object of knowledge: 'The arrangement of these four types of knowledge has names first and action last. This indicates that knowledge begins with knowledge of names and is completed with knowledge of action."

10. Graham, Mohist Logic, p. 286, mod.

11. I interpret the 'man' signific that forms part of the word hu to indicate that the drawn tiger is man-made, not a real tiger. Cf. such characters as wei and xiang.

12. Graham, Mohist Logic, p. 482, slightly mod.

13. Ibid., p. 286, slightly mod.

14. Ibid., p. 316, slightly mod.

15. It should be noted that on the issue of the circle, Graham himself did not maintain that the Mohists understood it in a conceptualist sense to name an idea, a charge against which he defends himself in his review article of Hansen's Language and Logic in Ancient China, in Harvard Journal of Asiatic Studies 45, no. 2 (1985): 698-699.

16. Xu Mo Zi jian gu, in Mo Zi jicheng, compiled by Yan Lingfeng, 46 vols. (Taipei: Chengwen Chubanshe, 1975), case 30, 3.6a.

17. Wang Dianji, Mo jing de luoji kexue sixiang fenxi, in Mo Zi jicheng, case 46, p. 307, also interprets this sophism as referring to universals.

18. Zhuang Zi jishi edition (Taipei: Muduo Chubanshe, 1982), bk. 33, p. 1106.

19. Graham, Mohist Logic, p. 309. He even refers the reader to Feng Youlan's Realist interpretation of this paradox.

20. Chad Hansen, Language and Logic in Ancient China (Ann Arbor: The University of Michigan Press, 1983), p. 113. Besides the circle, another example of yi is a pillar (A 39, B 57, and NO 3). Again it is Graham who advances the most philosophically stimulating, if albeit questionable, interpretation, maintaining that the "pillar, the stock example of something we recognize by a mental picture" ... "is known 'a priori' from the definition" (Mohist Logic, pp. 429, 224). This interpretation relies on emending the text at B 57 from wu zhi to xian zhi (on the assumption that wu had originally been written as wu). While I do not contend the plausibility of the amendment, nevertheless, in taking xian zhi to mean 'to know a priori' rather than simply 'to know beforehand', Graham leaves himself open to the objection that one

p. 356

imports more Western philosophical baggage than is warranted on the evidence. Nor has he demonstrated why 'to know beforehand' would be an unsuitable translation. Even if it is granted that the Mohists did use the words xian zhi to mean 'a priori', nevertheless, describing their using it in a "more sophisticated sense" (p. 223) than Wang Chong, who used it in the ordinary sense of 'knowing beforehand', would indeed be a considerable understatement of the philosophical implications of a priorism. The plausibility of this interpretation is further complicated, I believe, by the types of things the Mohist is said to consider to be a priori knowable, such as a pillar.

21. Xun Zi, SBBY, 16.4b.

22. In fact, in "On the Correct Use of Names," 16.4a, wu functions in a special technical sense as an extension of the meaning above: Accordingly, even though the myriad things are numerous, sometimes we wish to refer to them collectively and so we call them 'things'. 'Things' is the most general name.

23. See also Graham, Mohist Logic, pp. 196-197.

24. Following Liu Shipei, "Xun Zi bu shi," in Guo cuixue bao 55 (1909), 4a (in particular); 56 (1909), 1b, 3b, in understanding yue in the sense of 'demarcate/demarcation' or 'define/definition' (to determine or fix the boundary or extent of) rather than the standard rendering of 'convention'.

25. Following Wang Niansun, Du shu za zhi, 3 vols. (Beijing: Beijing Shi Zhongguo Shudian, 1985) (Xun Zi 7), vol. 2, p. 94, in treating shi after ming as superfluous.

26. Xun Zi 16.4b.

27. Xun Zi 11.10a: Ears, eyes, nose, mouth and body* are each connected but cannot perform one another's tasks. They are called the natural senses (lit. 'Heavenly offices'). The heart resides in inner vacuity and governs the five senses. It is called the natural ruler.
(*Following Wang Niansun, Du shu za zhi (Xun Zi 5), vol. 2, p. 61, in emending xing neng to xing tai.)
And if it were not for the heart, then "even if white and black were before one, the eyes would not see it, or a large drum beside one, the ears would not hear it" (Xun Zi 15.1b), because the heart is also the seat of cognition and knowledge:
The heart also recognizes. Because of the faculty of recognition, then by depending on [the sensory impressions of] the ears, sounds are able to be

p. 357

known, and by depending on [the sensory impressions of] the eyes, shapes are able to be known. (Xun Zi 16.3b)

28. Hansen, Language and Logic, pp. 62-63. See also pp. 57-65.

29. Xun Zi 16.2b.

30. Xun Zi 16.3a.

31. Following Yu Yue, Zhu zi ping yi (Taipei: Shijie Shuju, 1973) (Xun Zi 1), p. 132, in his gloss of the sentence gu song shu yi guan zhi from "Quan xue," in understanding shu to mean shuo.

32. Xun Zi 16.2a-2b.

33. In his Guanzi: Political, Economic and Philosophical Essays from Early China, 2 vols. (Princeton: Princeton University Press, 1985), vol. 1, p. 15, W. Allyn Rickett remarks that the "most widely accepted theory concerning the origin of the Guanzi holds that the so-called proto-Guanzi, that is, the core around which much of the present Guanzi finally took shape about 250 B.C., originated with the Jixia Academy." The Jixia Academy began in the reign of King Huan of Qi (374-357) and, despite a couple of periods of decline, was to continue for about one hundred and fifty years until the demise of the Qi state. The Academy was a meeting place for intellectuals and can in some respects be likened to some of the larger government-funded research centers of today; it has also been seen as a precursor of the Han Imperial Academy, albeit without the constraints of orthodoxy that were imposed upon the latter. See Qian Mu, Liang Han jingxue jinguwen pingyi (Taipei: Dongda Tushu Gongsi, 1983), p. 165; Zhou Yutong, "Boshi zhidu he Qin Han zhengzhi," in Zhou Yutong jingxueshi lunzhu xuanji (Shanghai: Shanghai Renmin Chubanshe, 1983), pp. 729-730.
The standard account of the Jixia Academy is that it was established in the reign of King Xuan of Qi (319-301). (Yet this is rather hard to accept, especially as one of the main passages used to support this account (Shi yi (Beijing: Zhonghua Shuju, 1982), 46.1895) explicitly states that after Zou Yan, Chun Yukun, et al. were appointed to the rank of dai fu during King Xuan's reign, then "from this time the schools of Jixia in Qi thrived once again.") Rickett (Guanzi, vol. 1, p. 15) even dates the establishment of the Academy as "about 302 B.C." Nevertheless, Xu Gan (170-217) records that "Duke Huan of Qi established the Jixia Academy and set up the rank of dai fu." See Zhong lun, Longxi jing she edition (Yangzhou: 1907), B.31 b. For a justification of Xu Gan's account, see Sun Yikai, "Jixia xuegong kaoshu," Wen shi 25 (1983): 41-42. For another good overall account of the Academy, see the collection of articles in Qi Lu xuekan, 52

p. 358

(1983): 21-37; 54 (1983): 23-26; and 55 (1983): 21-36. These articles are a significant improvement on Jin Shoushen, Jixia pai zhi yanjiu (Taipei: Shangwu Yinshuguan, 1971; original Preface 1930), which for decades served as the standard study on the Academy.

34. Zhang Nie (1096-1148) is the earliest known scholar to have identified these four books as a homogeneous group. See his "Guan Zi wen ping" in Guan Zi, Guoxue jiben congshu edition (Taipei: Shangwu Yinshuguan, 1967), A.4.

35. See Liu Jie, "Guan Zi zhong suo jian zhi Song Xing yi pai xueshuo" (originally published 1939-1941 in Shuowen yuekan), in Gu shi kao cun (Beijing: Renmin Chubanshe, 1958), pp. 238-258; Guo Moruo, "Song Xing Yin Wen yizhu kao" (dated 1944), in Qingtong shidai, included in Guo Moruo quanji, 8 vols. (Beijing, 1982-1985), vol. 1, pp. 547-572. N.B. Guo also includes "Qu yan" in this group.

36. This same criticism applies to the attempt made by Zhu Bokun to prove that the four books were the work of Shen Dao's later followers. See his article "Guan Zi sipian kao," Zhongguo zhexueshi lunwenji, first series (Jinan: Shandong Renmin Chubanshe, 1979), pp. 107-127. A similar thesis to that of Zhu has also been advanced by Qiu Xigui, "Mawangdui 'Laozi' jia yi ben juan qianhou yi shu yu 'dao fa jia'-jian lun 'Xin shu shang,' 'Bai xin' wei Shen Dao Tian Pian xuepai de zuopin," Zhongguo zhexue 2 (1980): 68-84.

37. See Machida Saburo, "Kan Shi shi hen ni tsuite" (originally published 1961), in his Shin Kan shiso no kenkyu (Tokyo: Sobunsha, 1985), pp. 358-361; Feng Youlan, Zhongguo zhexueshi xinbian, vol. 1, pp. 180-181. More recently, see also Kanaya Osamu, Kan Shi no kenkyu (Tokyo: Iwanami Shoten, 1987), pp. 263-264.

38. See also his article, "Futatabi Kan Shi shi hen ni tsuite," in Tohoku Daigaku kyoyobu kiyo (jimbunka hen) 4 (1966): 170-193. Rickett also gives support for this view by maintaining that the three Guan Zi chapters under discussion were likely to have been among those that:
originated in the old state of Ch'u and entered the Kuan-tzu through a group of scholars centered around the court of Liu An (180-122), the second king of Huai-an, who is best known as the reputed author of the Huai-nan-tzu.
See his article "Kuan-tzu and the Newly Discovered Texts on Bamboo and Silk," in Chinese Ideas About Nature and Society, Studies in Honour of Derk Bodde, ed. Charles Le Blanc and Susan Blader (Hong Kong: Hong Kong University Press, 1987), p. 247.

p. 359

39. "Kan Shi shi hen ni tsuite," pp. 357-384 passim and "Futatabi Kan Shi shi hen ni tsuite," pp. 171-174.

40. It may be objected that the division into Text and Commentary in "Xin shu shang" and the frequent use of the formula "therefore it is said," in quoting passages in "Xin shu xia" and "Bai xin", are indicative of the inclusion of older materials into each of those books. Nevertheless, by showing that portions of the Text itself evidence a syncretism characteristic of late Warring States and early Han Daoism, Machida develops a strong case for his claim that these books were written relatively late. See "Kan Shi shi hen ni tsuite," pp. 362-373.

41. 2:69-3/4.

42. This same meaning is expressed in the following passage from Han Fei Zi, SBBY, 6.9a; Graham, Mohist Logic, p. 213:
Men seldom see a live elephant, but when they find a dead elephant's bones they resort to its picture to imagine it alive. Therefore everything which men use to form an idea or image is called a hsiang [xiang] (elephant/image).

43. For examples and further discussion of this synonymity, see Kasahara, "Saden ni arawareta jitsu ji no kenkyu," pp. 9-10.

44. See his "The Background of the Mencian Theory of Human Nature," Tsing Hua Journal of Chinese Studies 6, nos. 1, 2 (1967): 262. See also Graham's discussion of qing in Mohist Logic, pp. 179-182. In a revised version of his 1967 paper that is included in his Studies in Chinese Philosophy and Philosophical Literature (Singapore: National University of Singapore, 1986), p. 63, Graham writes, "The ch'ing of X is what X cannot lack if it is to be called 'X'; the difference from Aristotelian essence is that it relates to naming, not being." Yet even if qing does relate to naming, this does not mean that is does not relate to being. Just as in early Greek thought so, too, in some early Chinese thought naming and the ability to discern and even determine what an entity genuinely is are intimately related.

45. 2:63-4.

46. It is uncertain when the Mawangdui manuscripts were originally written; I am inclined to accept that the so-called 'Daoist-Legalist' material (of which the Cheng text is an example) is middle-late third century B.C.

47. Mawangdui Han mu boshu (1) (Beijing: Wenwu Chubanshe, 1980), p. 81 (line 143A).

48. Literally, 'cuts'. Following Liu Ji in reading cai, 'wealth', as a loan graph

p. 360

for cai, 'to cut'. See Guan Zi jijiao 6:432. Cf. also Lao Zi 32: "Only when the [Uncarved Block] is cut are there names."

49. 2:66-13/13.

50. Following Tao Hongqing, Du zhu zi zha ji (Taipei: Yiwen Yinshuguan, 1971), p. 177, in emending ming zi zhi zhi to ming zhi.

51. Following Han Fei Zi, "Yang quan," 2.9a in reading yi, not qi.

52. 2:69-8/9. Following Wang Niansun in reading zheng ming zi zhi, qi (yi) ming zi fei. See Guan Zi jijiao, vol. 6, p. 447.

53. Huai Nan Zi, SBBY, 10.1a.

54. 9.1 a.

55. "Xin shu shang" 2:65-8/9 defines 'passive mindfulness' as follows: Passive mindfulness is the way of Non-action. Passive-mindfulness is neither to add to nor detract from anything.

56. Following A. C. Graham's rendering of this pair of elusive terms. See Graham, "A Neglected Pre-Han Philosophical Text: Ho-Kuan-Tzu," Bulletin of the School of Oriental and African Studies 52, no. 3 (1989): 515.

57. Borrowing Joseph Needham's terminology; see his Science and Civilization in China (Taipei: Caves Books, 1985), vol. 2, p. 281.

58. Guan Zi 2:69-8/9. Cf. also 2:66-13/13 . This is quite different from the Greek position as represented by Cratylus. Geoffrey S. Kirk, 'The Problems of Cratylus," American journal of Philology 72 (1951): 242, remarks:
... when Cratylus asked, "How could anyone saying that which he says, not say that which is?", he does not mean that any sound uttered by a man for any object is correct. An object only has one correct name and anyone who tries to call it anything else is not naming it at all but only uttering "a piece of voice." (383A)

59. Leonard Woodbury, "Strepsiade's Understanding: Five Notes on the Clouds," Phoenix 34, no. 1 (1980): 115. See also Kirk, 'The Problems of Cratylus," 239-240.

60. Woodbury, "Strepsiade's Understanding," p. 115. For Greek examples, see Cratylus, 391 D ff. passim.

61. Bernhard Karlgren, Grammata Serica Recensa (Stockholm: Bulletin of the Museum of Far Eastern Antiquities, 1957), No. 739a.

62. Ibid., No. 708a.

63. Ibid., No. 740a.

p. 361

64. Ibid., No. 739m.

65. Ibid., No. 707a.

66. Ibid., No. 739k.

67. Following Yu Yue, Zhu zi ping yi, p. 306, in reading er after da.

68. Su Yu, Chun qiu fan lu yi zheng (Taipei: Heluo Tushu Chubanshe, 1974; photographic reproduction of 1910 edition), 10.4a-4b. In describing the function of language in mythical thinking, Ernst Cassirer, The Philosophy of Symbolic Forms, 3 vols. (New Haven: Yale University Press, 1955), vol. 1, p. 40, writes that "the magical-mythical power of language is truly manifested in articulated sound. The formed word is itself restricted and individual: each word governs a specific realm of being over which it may be said to have unlimited and sovereign power."

69. Ming and hao seem to correspond closely to Xun Qing's san ming, or common names, and gong ming, or general names, respectively. See Chun qiu fan lu yi zheng, 10.3a.

70. Chun qiu fan lu yi zheng, 3.1 b.

71. Following Hsiao Kung-chuan, A History of Chinese Political Thought, vol. 1, trans. F. Mote (Princeton: Princeton University Press, 1979), p. 500, mod.

72. Chun qiu fan lu yi zheng, 10.2a.

73. Feng Youlan, Zhongguo zhexueshi xinbian, vol. 2, p. 86.

74. Ren Jiyu, Zhongguo zhexueshi, 4 vols. (Beijing: Renmin Chubanshe, 1979), vol. 2, p. 75. For similar interpretations, see also Fang Keli, Zhongguo zhexueshi shang de zhixing guan (Beijing: Renmin Chubanshe, 1982), pp. 94-95; Feng Qi, Zhongguo gudai zhexue de luoji fazhan, 3 vols. (Shanghai: Shanghai Renmin Chubanshe, 1984), vol. 2, p. 406; Hou Wailu et al., Zhongguo sixiang tongshi, 6 vols. (Beijing: Renmin Chubanshe, 1980), vol. 2, p. 123; Guan Hanheng, "Guanyu Dong Zhongshu de xiantian gainian shuo," Guangming ribao, 3 May 1964, p. 4; Lin Jie, "Guanyu Dong Zhongshu de renshilun," Guangming ribao, 15 May 1964, p. 4.

75. 10.5b.

76. Zhongguo zhexueshi shang de zhixing guan, p. 94. See also Feng Qi, Zhongguo gudai zhexue de luoji fazhan, vol. 2, p. 406; Ren Jiyu, Zhongguo zhexueshi, vol. 2, p. 75; Hou Wailu et al., Zhongguo sixiang tongshi, vol. 2, p. 123; Guan Hanheng, "Guanyu Dong Zhongshu de xiantian gainian shuo," p. 4; Lin Jie, "Guanyu Dong Zhongshu de renshilun," p. 4. In fact, I am aware of only two commentators who

p. 362

uphold a dissenting view: Su Yu, in his Chunqiu fanlu yi zheng, 10.5b; and Li Min, "Lue tan Dong Zhongshu de renshilun," Shanghai wenhuibao. 2 March 1962, p. 2.

77. For a detailed exposition of the term zhen and its synonymity with shi, 'actuality', see Kasahara Chuji, Chugokujin no shizenkan to bi ishiki, pp. 65-175.

78. Cf. Hall and Ames' characterization of yi as "fundamentally selfassertive and meaning-bestowing," in Thinking Through Confucius (Albany: State University of New York Press, 1987), p. 96.

79. 10.21b.

80. 10.14a.

81. 10.19a-19b.

82. 10.6a, 6b.

83. Bernhard Karlgren, Grammata Serica Recensa, No. 457a.

84. Ibid., No. 457e.

85. 10.10b. This example can also be found in Jia Yi (200-168), Xin shu, SBBY. 9.5b.

86. 10.19a.

87. 10.9b.

88. Chun qiu fan lu yi zheng, 17.11b.

89. 10.4a.

90. A similar, although more detailed, generative process is found in Book 5 of He Guan Zi (Taipei: Shijie Shuju, 1979), A.15a-15b, a syncretic text of the late third century B.C. Following A. C. Graham's rendering ("Ho-Kuan-Tzu," 514), the passage reads:
There being the One there is the ch'i, then the idea, then the picture, then the name, then the shape, then the work, then the covenant. The covenant being decided the time is born, the time being set the thing is born.
Graham rejects the interpretation that the passage presents the rudiments of a Realist theory of naming where "in the cosmological process itself the generation of things [wu] follows ideas, pictures and names emerging from the primal ch'i." The basis of his rejection is that "nothing in the book suggests that Heaven has mental pictures and puts its decrees into words." Yet given that the passage is describing that which issues from "the One," and the One, not Heaven, is portrayed as the ultimate source of things in He Guan Zi (Graham, p. 510), Graham's objection is less than persuasive. As to the interpretation of yi as "mental pictures," I think that here the

p. 363

word is better understood to mean "intention, will" in a similar sense to Dong Zhongshu's usage.

91. See, for example, chapter 2 passim of W. South Coblin's A Handbook of Eastern Han Sound Glosses (Hong Kong: The Chinese University Press, 1983).

92. Wang Xianqian, Shi ming shu zheng bu edition (Shanghai: Shanghai Guji Chubanshe, 1984), Preface, la; slight modification of Roy Andrew Miller's translation. See his "Review of Nicholas Cleveland Bodman, A Linguistic Study of the Shih Ming," T'oung Pao 44 (1956): 281.